ã
Oncogene (1999) 18, 4554 ± 4563
1999 Stockton Press All rights reserved 0950 ± 9232/99 $15.00
http://www.stockton-press.co.uk/onc
Overexpression of urokinase-type plasminogen activator in pancreatic
adenocarcinoma is regulated by constitutively activated RelA
Weixin Wang1, James L Abbruzzese2, Douglas B Evans1 and Paul J Chiao*,1,3
1
Department of Surgical Oncology, The University of Texas MD Andersen Cancer Center, Houston, Texas, TX 77030, USA;
Department of Gastrointestinal Medical Oncology and Digestive Diseases, The University of Texas MD Andersen Cancer Center,
Houston, Texas, TX 77030, USA; 3Department of Tumor Biology, The University of Texas MD Anderson Cancer Center,
Houston, Texas TX 77030, USA
2
The Rel/NF-kB transcription factors regulate the
expression of many genes. The activity of RelA, a
member of the Rel/NF-kB transcription factor family, is
constitutively activated in the majority of pancreatic
adenocarcinomas and cell lines. We report that the
urokinase-type plasminogen activator (uPA), one of the
critical proteases involved in tumor invasion and
metastasis, is overexpressed in pancreatic tumor cells
and its overexpression is induced by constitutive RelA
activity. The uPA promoter contains an NF-kB binding
site that directly mediates the induction of uPA
expression by RelA. Expression of a dominant-negative
IkBa mutant inhibits kB site-dependent transcriptional
activation of a uPA promoter-CAT reporter gene.
Treating the pancreatic tumor cell lines with the known
NF-kB inhibitors, dexamethasone and n-tosylphenyalanine chloromethyl ketone (TPCK), abolishes constitutive
RelA activity and uPA overexpression. These results
show that uPA is one of the downstream target genes
induced by constitutively activated RelA in human
pancreatic tumor cells, and suggests that constitutive
RelA activity may play a critical role in tumor invasion
and metastasis. Inhibition of constitutive RelA in
pancreatic tumor cells may reduce their invasive and
metastatic potential.
Keywords: urokinase-type plasminogen
NF-kB; pancreatic adenocarcinoma
activator;
Introduction
Rel/NF-kB is a family of pleiotropic transcription
factors that control the expression of numerous genes
involved in the immune response, embryonic development, lymphoid di€erentiation, oncogenesis and
apoptosis (Verma et al., 1995; Baldwin, 1996). The
Rel/NF-kB family consists of relA (p65), c-rel, v-rel,
relB, nfkb1 and nfkb2, which can form heterodimers
and homodimers among themselves (Verma et al.,
1995; Baldwin, 1996). In most cell types, Rel/NF-kB
proteins are sequestered in the cytoplasm in an inactive
form through their noncovalent association with the
inhibitor IkB (Baeuerle and Baltimore, 1988; Grilli et
al., 1993). This association masks the nuclear
*Correspondence: PJ Chiao, Department of Surgical Oncology and
Tumor Biology, Box 107, The University of Texas MD Anderson
Cancer Center, 1515 Holcombe Blvd., Houston, Texas 77030, USA
Received 6 November 1998; revised 11 March 1999; accepted 11
March 1999
localization signal of Rel/NF-kB, thereby preventing
Rel/NF-kB nuclear translocation (Baeuerle and Baltimore, 1988; Grilli et al., 1993). Activation of Rel/NFkB is controlled by IkBa and does not require protein
synthesis, thereby allowing rapid and ecient gene
regulation (Baeuerle and Baltimore, 1988; Grilli et al.,
1993; Verma et al., 1995; Baldwin, 1996). Stimulation
of cells by various inducers leads to phosphorylation of
IkBa at serine residues 32 and 36, which triggers the
rapid degradation of the inhibitor. Consequently, Rel/
NF-kB proteins are released and translocated into the
nucleus, where they activate the expression of target
genes (Brown et al., 1995; Chen et al., 1995; DiDonato
et al., 1995).
Several reports suggest that members of the Rel/NFkB and IkB families are involved in tumorigenesis and
metastasis (Gilmore et al., 1997). Carried by a highly
oncogenic retrovirus, v-rel causes aggressive tumors in
young birds and is able to transform avian lymphoid
cells and ®broblasts (Moore and Bose, 1988a,b). The
mutated c-rel oncogene transforms cells (Sylla and
Temin, 1986). The Tax protein from the human T-cell
leukemia virus (HTLV-1) is a potent activator of Rel/
NF-kB, and the growth of Tax-induced tumors in mice
is inhibited by antisense RelA constructs (Kitajima et
al., 1992). Furthermore, the expression of IkBa
antisense results in constitutive activation of RelA
and oncogenic transformation of NIH3T3 cells
(Beauparlant et al., 1994), suggesting that RelA is a
proto-oncogene. The genes encoding c-Rel, Bcl-3, p105
(p50) and p100 (p52) have been shown to be located at
sites of recurrent genomic rearrangements in lymphoid
cancers (Ohno et al., 1990; Lu et al., 1991; Neri et al.,
1991; Liptay et al., 1992; Gilmore et al., 1997). It has
been shown that NF-kB can regulate the expression of
several genes important to cellular adhesion (i.e. genes
encoding ICAM-I, ELAM-I, VCAM-I and type IV
collagenase) (Voraberger et al., 1991; Whelan et al.,
1991; Ghersa et al., 1992; Sato and Seiki, 1993;
Iademarco et al., 1992). These results suggest that a
dysregulation of NF-kB activity in tumor cells could
lead to tumorigenesis and promote the development of
metastasis.
The expression of urokinase-type plasminogen
activator (uPA) can be regulated by ras (Testa et al.,
1989; Stacey et al., 1991), by the tyrosine kinaseencoding oncogenes v-src and v-yes (Bell et al., 1990,
1993), by the serine-threonine kinase-encoding v-mos
oncogene (Lengyel et al., 1995b) and by polyomavirus
middle-T antigen (Besser et al., 1995b). uPA is a serine
protease that cleaves extracellular matrix/basement
membrane (ECM/BM) proteins (e.g. ®bronectin,
laminin and collagen) directly; it also can degrade
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
ECM/BM proteins by converting the ubiquitous
extracellular zymogen plasminogen into plasmin,
another serine protease with broader proteolytic
activity. Subsequently, plasmin can convert procollagenases into collagenases necessary for the degradation
of basement membrane collagen and can activate
metalloproteinases, another class of proteases with
more potent ECM/BM-degrading capability (Dano et
al., 1985; Pollanen et al., 1991; Birkedal-Hansen et al.,
1993; Stetler-Stevenson et al., 1993). Thus, overexpression of uPA by neoplastic cells at the invading front of
cancerous tissue leads to a cascade of proteolytic
activities critical to local tumor growth and metastasis
(Ginestra et al., 1997).
Little information is available concerning the
expression and regulation of the uPA gene and its
possible role in mediating the aggressive biology of
pancreatic cancer. This disease is the ®fth leading cause
of cancer death with an overall 5-year survival rate of
less than 5% (DiGiuseppe et al., 1996; Rosewicz and
Wiedenmann, 1997), and has the worst prognosis of all
gastrointestinal malignancies. Most patients develop
hepatic metastasis and die secondary to the deleterious
physiologic e€ects of progressive metastatic disease.
Current therapy has little e€ect on the development or
growth of metastases. Therefore, it is important to
understand the mechanism of pancreatic cancer
metastasis and develop strategies to improve therapeutic response.
We reported that RelA was constitutively activated
in pancreatic adenocarcinoma and that the constitutive
activity of RelA was inhibited by dominant-negative
mutants of IkBa, ras, and MEKK1 (Wang et al., 1999).
These results show that MAP kinase signaling cascades
are key components regulating RelA in pancreatic
adenocarcinoma cells. However, downstream target
genes of RelA that are relevant to tumorigenesis and
tumor metastasis remain unclear. Interestingly, Hansen
et al. (1992) had reported that phorbol ester-inducible
uPA expression was mediated by two NF-kB heterodimeric binding complexes. One consisted of Rel/p65
and the other p65/p50. We therefore undertook a study
to de®ne the expression of uPA in pancreatic
adenocarcinoma cells and to determine whether uPA
expression is induced by constitutive RelA activity in
this cancer. As shown in this report, we found that: (1)
uPA was overexpressed in pancreatic adenocarcinoma
tissue and cell lines; (2) uPA overexpression was
induced by constitutive RelA activity in pancreatic
adenocarcinoma cells; and (3) inhibition of constitutive
RelA activity suppressed uPA expression in pancreatic
adenocarcinoma cell lines.
detected in 83% of (25 of 30) pancreatic adenocarcinomas but no staining was detected in normal human
pancreatic ductal epithelial cells (Figure 1). No
nonspeci®c binding or cross reactivity of this antibody
to other proteins were detected in subsequent Western
blot analysis (data not shown). These results suggested
that overexpression of uPA may contribute to the
metastatic potential of human adenocarcinoma.
To further examine the overexpression of uPA in
human pancreatic adenocarcinoma, levels of uPA
mRNA and protein in pancreatic adenocarcinoma cell
lines were studied. Northern blot analysis was carried
out using RNA extracted from 11 pancreatic cell lines.
Figure 2 shows that MiaPaCa-2, Hs-766T, CFPAC-1,
CAPAN-2, MDAPanc48, MDAPanc28, MDAPanc3
and Panc-1 exhibited much higher levels of uPA
mRNA, compared to BxPC-3, AsPC-1 and CAPAN1 cells that displayed low levels of uPA mRNA. The
percentage of pancreatic adenocarcinoma cell lines with
high level of uPA mRNA (73%, eight of 11) is
consistent with the results obtained from the immunohistochemical analysis of pancreatic tumor tissues. The
high- and low-uPA expressing pancreatic tumor cell
lines provided useful tools for studying the regulation
of uPA expression.
Results
uPA is overexpressed in pancreatic adenocarcinoma and
pancreatic tumor cell lines, and its expression is related
to tumor metastatic potential
To determine the level of uPA protein in pancreatic
adenocarcinoma, paran sections of 30 human
pancreatic adenocarcinoma samples and ®ve normal
pancreatic tissue samples were subjected to immunohistochemical analysis with anity-puri®ed anti-human
uPA monoclonal antibody. Speci®c uPA staining was
Figure 1 Immunohistochemical detection of uPA in normal and
tumor pancreatic tissue. Normal and tumor pancreatic tissue
samples were subjected to immunohistochemical analysis using
anti-uPA monoclonal antibody as described in Materials and
methods. (a) normal pancreas tissues and (b) pancreatic
adenocarcinoma
4555
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
4556
To further analyse the level of uPA protein and
enzymatic activity, four human pancreatic tumor cell
lines (CFPAC-1, BxPC-3, AsPC-1 and CAPAN-1)
representing di€erent levels of uPA expression were
chosen for subsequent analyses. Cytoplasmic extracts
from those four cell lines were prepared for Western
blot analysis. As shown in Figure 3a, the cytoplasmic
uPA protein level in CFPAC-1 cells was much higher
relative to BxPC-3, AsPC-1 and CAPAN-1 cells.
Using these four cell lines, Western blot analysis was
also carried out to determine the level of secreted uPA
in conditioned media collected from cultures of equal
numbers of cells. As shown in Figure 3b, conditioned
medium from CFPAC-1 cells had the highest uPA
level at 24 and 48 h; an intermediate level of uPA
activity was observed in conditioned medium from
BxPC-3 cells, and the lowest levels of uPA were
detected in conditioned media from AsPC-1 and
CAPAN-1 cells.
To determine the uPA enzymatic activity secreted
by these four pancreatic cell lines, zymography, with
casein and plasminogen as the in-gel substrate, was
carried out using the conditioned medium collected
from cultures of equal numbers of cells. uPA activity
was demonstrated as an unstained band in the
Coomassie blue-stained gel. As shown in Figure 3c,
CFPAC-1 cells had the highest level of uPA activity,
BxPC-3 cells had an intermediate level and AsPC-1
and CAPAN-1 cells had a very low level of
proteolytic activity. Taken together, these results
indicate that uPA was overexpressed with very potent
enzymatic activity in CFPAC-1 cells, that the uPA
protein level and activity in BxPC-3 cells was
intermediate, and that in AsPC-1 and CAPAN-1
cells both the uPA protein level and uPA activity
were almost undetectable. The expression patterns of
uPA mRNA, cytoplasmic and secreted uPA protein
and uPA enzyme activity were consistent in these
analyses.
To determine whether high expression and activity
of uPA contributed to the invasive and metastatic
potential of pancreatic tumor cells, we performed
adhesion assays using the high- and low-uPA-expressing pancreatic tumor cell lines CFPAC-1 and
CAPAN-1, respectively. As shown in Figure 3d, the
numbers of CAPAN-1 cells that adhered to laminin-
Figure 2 uPA mRNA level in pancreatic adenocarcinoma cell
lines. Total RNA (10 mg) extracted from di€erent cell lines was
subjected to Northern analysis. The blots were hybridized with
radioisotope 32P-dCTP-labeled human the 1.2-kb PstI uPA
cDNA probe, exposed, stripped, and rehybridized with the
cDNA probe for glyceraldehyde-3-phosphate dehydrogenase
(GAPDH) for loading control
coated plates were four times the numbers of CFPAC-1
cells that adhered. Similar adhesion pro®les were
observed on ®bronectin- and collagen-coated plates:
Figure 3 uPA protein level and enzyme activity and in vitro
adhesion assays in pancreatic adenocarcinoma cell lines. uPA
protein level in cytoplasm (a), and in conditioned medium (b) as
shown by Western blotting. Recombinant human uPA (rhuPA)
was loaded as a positive control. (c) uPA enzyme activity in
conditioned medium at 24 and 48 h as shown by zymography. (d)
In vitro adhesion assay showing that di€erent adhesion
capabilities of CFPAC-1 and CAPAN-1 cells on plates coated
with laminin, ®bronectin or collagen. Data are expressed as the
mean+s.d. of four identical experiments
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
CAPAN-1 cells showed much more adhesiveness to
®bronectin and collagen than did CFPAC-1 cells
(Figure 3d). Taken together, these data show that
overexpression of the uPA gene decreased the
capabilities of pancreatic cancer cells to adhere to
extracellular matrix and basement membrane, suggesting that of uPA overexpression plays a critical role in
the invasive and metastatic potential of pancreatic
adenocarcinoma.
Overexpression of uPA is induced transcriptionally
To determine whether the overexpression of uPA is
caused by uPA gene ampli®cation or by transcriptional up-regulation, we carried out Southern blot and
nuclear run-on analyses. The Southern blot analysis
using DNA from human normal pancreas as a control
showed that the overexpression of uPA in these cell
lines did not result from uPA gene ampli®cation
suggesting that overexpression of uPA may be
regulated at the transcriptional level (Figure 4a).
Nuclei from CFPAC-1 and CAPAN-1 cells were
isolated, and a nuclear run-on assay was performed
to compare the RNA transcription rates of these two
cell lines. The results of nuclear run-on assay clearly
indicated that overexpression of uPA in CFPAC-1
cells resulted from transcriptional up-regulation
(Figure 4b).
a
4557
Transcription factor RelA plays an important role in
regulating uPA expression in pancreatic adenocarcinoma
cells
We previously showed that the transcription factor
RelA was constitutively activated in pancreatic
adenocarcinoma tissue samples and cell lines (Wang
et al., 1999). It has been reported that the kB enhancer
element between 71592 bp and 71570 bp of the uPA
promoter mediated phorbol ester-inducible urokinase
gene expression and the kB binding complexes consist
of Rel/p65 and p65/p50. (Hansen et al., 1992). We
undertook the following studies to speci®cally determine the possible role of constitutive RelA activity in
inducing uPA gene overexpression in pancreatic cancer
cells.
In Figure 5 (lane 1), electromobility shift assay
using a probe containing the uPA kB site con®rmed
the constitutive RelA activity in CFPAC-1 cells. The
speci®city of RelA DNA binding activity was
determined by competition with double-stranded
oligonucleotides containing mutant (mu) or wildtype (wt) uPA kB site and HIV wild-type kB site
(Figure 5, lanes 2 ± 6). To con®rm that the uPA kB
binding complex consists of RelA/p50, supershift
experiments were performed with anti-RelA (p65),
anti-p50, and anti c-Rel antibodies. The results
showed that RelA and p50 were present in the
uPA kB binding complex, but c-Rel protein could
not be detected in the uPA kB binding complex
(Figure 5, lanes 7 ± 9). These data are consistent with
our previous report that constitutive kB binding
activity in pancreatic adenocarcinoma tissue samples
and cell lines contained RelA(p65)/p50 proteins
(Wang et al., 1999). To determine whether the
RelA protein can activate the kB binding element
in the uPA promoter, RelA (p65) expression vectors,
HIV kB-CAT reporter plasmids containing wild-type
or mutant kB sites (Chiao et al., 1994), a uPAwtkB-
1
Figure 4 Southern blot and nuclear run-on analyses of the uPA
gene in pancreatic cancer cell lines. In (A) southern blot of the
uPA gene in pancreatic cell lines: (a) ethidium bromide-staining of
agarose gel before blotting and (b) hybridization pattern of
EcoRI-digested genomic DNA with 32P-labeled human uPA
cDNA fragment (1.2-kb PstI). In (b), nuclear run-on analysis of
uPA gene transcription rate from CFPAC-1 and CAPAN-1 cells.
Nylon-immobilized cDNAs corresponding to uPA, GAPDH and
plasmid vector DNA were probed with 32P-labeled total RNA
2
3
4
5
6
7
8
9
Figure 5 Constitutive NF-kB DNA binding activity in CFPAC-1
cells. EMSA showing that NF-kB DNA binding activity in
CFPAC-1 cells (Control) and the speci®city of RelA NF-kB DNA
binding activity was determined by competition with doublestranded oligonucleotides containing mutant (mu) or wild-type
(wt) uPA kB site and HIV wild-type kB site. The supershifts were
performed with anti-p65 (RelA) (ap65), anti-p50 (ap50) and anti
c-Rel (ac-rel) antibodies The supershifted bands are indicated by
arrows. The nonspeci®c DNA bindings were indicated as asterisks
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
4558
CAT reporter plasmid (Hansen et al., 1992; Lengyel
et al., 1995c) and a uPAmukB-CAT reporter plasmid
were cotransfected into AsPC-1 cells that express
much lower level of uPA in comparison with that in
CFPAC-1 cells (Figure 2) and have constitutive
RelA activity (Wang et al., 1999). CAT assay
results showed that constitutive RelA activity in
AsPC-1 cells activated the HIVwtkB-CAT reporter
D
Figure 6 Activation of uPA promoter activity by NF-kB and
inhibition of its activity by IkBa.M. (a) Wild-type (HIV[wt]CAT)
or mutant (HIV[mu]CAT) kB reporter gene construct was
cotransfected with p65(RelA) (pCMX-p65) expression plasmids
or pBSCMV plasmid into AsPC-1 cells as controls. (b) uPA
promoter containing wild-type kB binding motif (uPA[wt]kBCAT) or mutated one (uPA[mu]kB-CAT) was cotransfected with
p65(RelA) expression plasmid or pBSCMV plasmid as indicated.
(c) uPA[wt]kB-CAT or uPA[mu]kB-CAT reporter gene was
cotransfected with IkBaM expression vector (CMV-IkBaM) or
pBSCMV plasmid into CFPAC-1 cells. (d) b-actin-Fire¯y
luciferase and TK-Rellina luciferase reporter genes were
cotransfected with IkBaM expression vector (CMV-IkBaM) or
pBSCMV plasmid into CFPAC-1 cells as indicated. Dual
luciferase assay was carried out to determine the ratio of the
two luciferases
gene (Figure 6a, lane 3) and overexpression of RelA
by transient transfection did not signi®cantly increase
the HIVwtkB-CAT reporter gene activity (Figure 6a,
lane 4). Little reporter gene activity was detected
when the HIVmukB-CAT was used (Figure 6a, lanes
1 and 2). In Figure 6b, a 2-kb uPA promoter
containing a wild-type kB binding motif (uPA[wt]kBCAT) or a mutated kB motif (uPA[mu]kB-CAT)
was cotransfected into AsPC-1 cells with the
p65(RelA) expression plasmid or the pBSCMV
plasmid as indicated. Overexpression of RelA by
transient transfection induced the uPA[wt]kB-CAT
reporter gene activity (Figure 6b lanes 1 and 2) by
250% over the control and was consistently obtained
in three independent experiments. However, the
similar basal levels of CAT activities between the
2-kb uPA(wt)kB promoter- and 2-kb uPA(mu) kB
promoter-CAT reporter gene constructs that were
consistently obtained in three independent experiments may be due to other endogenous transcription
factors such as c-Jun or Ets-1 and Ets-2 (Janulis et
al., 1999; Watabe et al., 1998). In contrast,
overexpression of RelA increased uPA[mu]kB-CAT
reporter gene activity to a much lesser extent (70%,
Figure 6b lanes 3 and 4). The small increase seen
with the uPA[mu]kB-CAT reporter gene may be due
to the persistent weak interaction of overexpressed
RelA with seven out of 10 remaining wild-type
nucleotides in the mutated kB site in the uPA
promoter. Nevertheless, these results demonstrate
that the kB site in uPA promoter is the major
determinant for the enhanced expression seen in the
presence or overexpressed RelA in the transient
transfection experiments. To con®rm that the overexpression of the uPA gene in pancreatic tumor cells
was speci®cally induced by constitutive RelA, the
plasmids expressing dominant-negative IkBa(IkBaM)
(Van Antwerp et al., 1996) were included in
cotransfection experiments using CFPAC-1 cells that
express high level of uPA (Figures 2 and 3) and
RelA activity (Wang et al., 1999). The phosphorylation mutant of IkBa(IkBaM) was generated by
converting the inducible phosphorylation sites of
serine 32 and 36 (which are responsible for NF-kB
activation) to alanines. The carboxyl-terminal phosphorylation sites in the PEST sequence implicated in
constitutive turnover of IkBa, were also mutated
(Van Antwerp et al., 1996). It has been shown that
ectopically expressed IkBaM completely inhibits kBspeci®c DNA binding activity in extracts from cell
lines stimu- lated with TNFa (Van Antwerp et al.,
1996). The results show that dominant-negative IkBa
eciently blocked uPA promoter activation in
CFPAC-1 cells (Figure 6c).
To rule out that overexpression of IkBaM did not
inhibit gene expression nonspeci®cally, we repeated the
transfection experiments using a TK-Renilla luciferase
reporter gene for normalization of transfection
eciency, a b-actin-promoter ®re¯y luciferase reporter
gene for determining the speci®city of IkBaM protein,
and uPA-CAT reporter genes with wild-type and
mutated kB sites for determining the inhibition of
RelA activity by the IkBaM protein. The results
presented in Figure 6d showed that b-actin-promoter
was not a€ected by overexpression of IkBaM protein.
We therefore conclude that the constitutive RelA
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
activity plays a critical role in overexpression of uPA in
pancreatic adenocarcinoma cells.
RelA activity suppresses uPA expression in pancreatic
adenocarcinoma cell lines.
uPA overexpression can be repressed by NF-kB
inhibitors
Discussion
TPCK (N-tosyl-L-phenylalanine chloromethyl ketone)
and dexamethasone have been shown to inhibit NF-kB
activation (Chiao et al., 1994; Auphan et al., 1995;
Scheinman et al., 1995; Wu et al., 1996). To further test
whether RelA regulates uPA gene expression in
pancreatic cells and whether uPA overexpression can
be inhibited by NF-kB inhibitors, CFPAC-1 cells,
which overexpress uPA and possess constitutive RelA
activity, were treated with TPCK and dexamethasone.
Nuclear extracts were prepared from CFPAC-1 cells
with or without treatment with these NF-kB inhibitors
to determine whether RelA DNA binding activity was
inhibited. Simultaneously, media were collected from
the CFPAC-1 cell cultures after TPCK and dexamethasone treatments to determine the levels of
secreted uPA protein. As shown in Figure 7a, after
treatment with TPCK and dexamethasone, RelA/NFkB DNA binding activity was clearly inhibited in the
CFPAC-1 cells. The speci®c RelA/NF-kB DNA
binding activity was con®rmed in the nuclear extracts
of control CFPAC-1 cells by competition studies using
oligonucleotides containing the wild-type (wt) or
mutant (mu) NF-kB binding site and by supershifting
assays using anti-RelA and anti-p50 antibodies (Figure
5). Consistent with the inhibition of RelA NF-kB
DNA binding activity in the nuclei of CFPAC-1 cells,
levels of secreted uPA protein were reduced in
conditioned media from cultures of CFPAC-1 cells
treated with TPCK and dexamethasone (Figure 7b).
Taken together, our data indicate that uPA is
frequently overexpressed in pancreatic adenocarcinoma tissues and cell lines, uPA overexpression is induced
by the constitutive RelA activity in pancreatic
adenocarcinoma cells and inhibition of constitutive
Figure 7 Inhibition of constitutive NF-kB activity in CFPAC-1
cells. (a) EMSA showing that NF-kB DNA binding activity in
CFPAC-1 cells, untreated or treated with DMSO as controls, or
treated with TPCK (100 mM for 4 h) or dexamethasone (DEX)
(1 mM for 8 h). (b) Western blot results showing that the level of
uPA protein secreted into conditioned medium from control
(control and DMSO) and from TPCK or dexamethasone (DEX)
treated CFPAC-1 cells. The percentages of the secreted uPA levels
of that in control CFPAC-1 cells were indicated
In this report, we have shown the key role played by
RelA in inducing uPA gene overexpression. In turn,
uPA is capable of increasing the invasion and
metastatic potential of pancreatic ductal adenocarcinoma cells. Rel/NF-kB belongs to a family of pleiotropic
transcription factors that control the expression of
numerous genes (Lenardo and Baltimore, 1989, Grilli
et al., 1993). The downstream target genes of Rel/NFkB include the p53 tumor suppressor gene, several
proteins important to cellular adhesion (ICAM-I,
ELAM-I, VCAM-I) and extracellular matrix protease
(type IV collagenase) (Wu and Lozano, 1994; Ghersa
et al., 1992; Iademarco et al., 1992; Voraberger et al.,
1991; Whelan et al., 1991; Sato and Seiki, 1993). A
recent report showed that the promoter of the major
histocompatibility complex (MHC) Class I-encoding
gene contains an NF-kB site and that NF-kB p50
homodimers repress H-2Kb expression in metastatic
tumor cells (Plaksin et al., 1993). It was previously
shown that downregulation of the MHC Class Iencoding gene's expression in tumors is associated with
reduced immunogenicity and with the development of
metastatic phenotype (Cordon-Cardo et al., 1994).
Taken together, these results suggest that dysregulation of NF-kB activity in tumor cells may play a
critical role in the development of cancer metastasis.
uPA plays a central role in catalyzing extracellular
and basement membrane degradation. Several investigators have studied the regulation of uPA expression
and they have found that induction of uPA expression
is detected during G0/G1 transition and correlates with
the proliferative state of normal mouse cells (Grimaldi
et al., 1993). uPA expression is required in in vitro
assays of basement membrane invasion and in
metastasis of model tumors and there is a strong
association between uPA expression and the invasivemetastatic phenotype. High expression of uPA is a
signi®cant prognostic marker for various human
tumors (Blasi, 1990). For example. uPA is a
prognostic marker in breast cancers (Du€y et al.,
1990). Axelrod et al. (1989) reported H-ras-transformed NIH3T3 cells stably expressing a transfected
uPA gene invaded more rapidly than parent cells
through a basement membrane and caused a high
incidence of pulmonary metastatic lesions after
intravenous injection into nude mice. Conversely,
inhibition of uPA can reduce the invasion and
metastasis in H-ras-transformed NIH3T3 cells (Axelrod et al., 1989; Rao et al., 1993). It has also been
reported that inhibition of uPA can reduce the size of
prostate cancer tumors or even cause complete
remission of cancers in SCID mice (Harvey et al.,
1988; Jankun et al., 1997). These ®ndings provide
direct evidence for a causal role for uPA in local tumor
invasion and metastasis. However, the role of uPA and
its regulation in pancreatic cancer metastasis has not
been previously reported. In this report, we have
provided evidence that uPA is overexpressed in most
pancreatic adenocarcinomas and is related to
the metastatic potential of pancreatic tumors
4559
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
4560
(Figures 1 ± 3). The overexpression of uPA is induced
transcriptionally by constitutively activated RelA
activity (Figures 4 ± 6). Inhibition of constitutive RelA
activity by cotransfection with dominant-negative
mutant of IkBa and by treatment with the known
NF-kB inhibitors TPCK and dexamethasone represses
the overexpression of uPA in pancreatic tumor cell
lines (Figures 6 and 7).
Several inducers have been reported to induce uPA
promoter activity by acting at its upstream PEA3/AP1
sites and via the Ras/Raf-1/MEK/ERK pathway.
These inducers include polyomavirus middle-T antigen
(Besser et al., 1995a), macrophage colony-stimulating
factor (Stacey et al., 1995), ®broblast growth factor 2
(Besser et al., 1995b), LFB3 (Marksitzer et al., 1995),
hepatocyte growth factor (Je€ers et al., 1996; Mars et
al., 1996), tumor necrosis factor-a (Lengyel et al.,
1995a), c-Ha-ras (Lengyel et al., 1995b), cyclic
adenosine monophosphate (Grimaldi et al., 1993) and
tissue plasminogen activator (Nerlov et al., 1992).
Recently, it was reported that MEKK1 kinase can
phosphorylate the IkBa kinase complex and that
CHUK is one of the IkBa kinases (Lee et al., 1997;
Regnier et al., 1997; DiDonato et al., 1997). It will be
important to determine whether dominant negative
mutants of CHUK, MEKK1 and other components of
the MAP kinase pathway can inhibit uPA overexpression in human pancreatic tumor cells. Interestlingly, our data show that additional transcription
factors may involved in overexpression of uPA in the
pancreatic cancer cell lines. Previously, we have shown
that RelA is constitutively activated in pancreatic
adenocarcinoma cell lines CAPAN-1, ASPC-1 and
CFPAC-1 (Wang et al., 1999). Figures 2 and 3 show
that the level of uPA expression in both CAPAN-1 and
ASPC-1 is much lower than those in CFPAC-1 and
other cell lines, suggesting that constitutive RelA
activity is necessary but may not be sucient to
induce uPA overexpression. It is possible that other
transcription factors may cooperate with RelA protein
to induce uPA overexpression in CFPAC-1 cells,
whereas the activities of such transcription factors
that cooperate with RelA protein to induce uPA
overexpression may not present in CAPAN-1 and
ASPC-1. Given the role of uPA in tumor invasion and
metastasis, it will be important to identify the
transcription factors that cooperate with RelA protein
to induce uPA overexpression.
In this report, we also showed that uPA expression
in pancreatic cell lines can be inhibited by using a
dominant-negative mutant of IkBa to inhibit inducible
phosphorylation in NF-kB activation or by using
known NF-kB inhibitors such as TPCK and
dexamethasone. Although both TPCK and dexamethasone reduced NF-kB activity to a similar level,
there is a di€erence in reduction of the levels of uPA
between TPCK and dexamethasone treated cells. This
di€erence may re¯ect a nonspeci®c e€ect or the high
toxicity of TPCK. These known uPA inhibitors are
unlikely to be used in anticancer therapy because of
their weak inhibitory activity or high toxicity. Thus,
these ®ndings may help guide studies in which
di€erent NF-kB inhibitors are analysed for their
ability to prevent liver metastasis of pancreatic
cancer in animal models by inhibiting overexpression
of uPA.
Materials and methods
Cell lines and cell culture
The human pancreatic adenocarcinoma cell lines AsPC-1,
BxPC-3, CAPAN-1, CAPAN-2, Hs-766T, CFPAC-1, Panc-1
and MiaPaCa-2 were obtained from the American Type Cell
Culture (Rockville, MD, USA). MDAPanc-3, MDAPanc-28
and MDAPanc-48 cells were obtained from Dr Marsha L
Frazier (MD Anderson Cancer Center). All cells were grown
in 10% fetal bovine serum in Dulbecco's modi®ed Eagle's
medium in a 378C incubator containing a 5% CO2
atmosphere.
Immunohistochemistry
Formalin-®xed, paran-embedded adenocarcinoma tissues
were obtained at the time of pancreatectomy from patients
at our institution. The monoclonal anti-human uPA antibody
was purchased from American Diagnostica (Greenwich, CT,
USA). Immunohistochemical staining was performed as
described previously (Sinicrope et al., 1996).
Western blotting and zymography
Conditioned media were collected from cultures containing
equal numbers of cells. Cytoplasmic extracts were prepared
as described by Andrews and Faller (1991). The amount of
urokinase in cytoplasm and conditioned medium was
determined by Western blotting. Samples were denatured in
the absence of reducing agent and subjected to 10% sodium
dodecyl sulfate-polyacylamide gel electrophosesis (SDS ±
PAGE). The resolved proteins were transferred to an
Immobilon-P membrane (Millipore, Bedford, MA, USA).
The membrane was blocked with 5% nonfat milk and
incubated with polyclonal antibody to urokinase (American
Diagnostica). Proteins were detected using a chemoilluminescence kit (Amersham, Arlington Heights, IL, USA).
Zymography was performed as described (Gogly et al.,
1998) to determine the uPA enzyme activity. Conditioned
medium collected from cultures containing equal numbers of
cells was denatured in the absence of reducing agent and
subjected to 7.5% SDS ± PAGE on a gel containing 0.2%
casein and 6 mg/ml plasminogen. The gels were incubated at
378C for 2 h in the presence of 2.5% Triton X-100 and then
for 1 h in a bu€er containing 10 mM CaCl2, 0.15 M NaCl and
50 mM Tris (pH 7.5). The gel was then stained for protein
with 0.25% Coomassie blue and photographed on a light
box. Plasminogen-dependent proteolysis was detected as a
white zone in a dark ®eld.
Adhesion assay
In vitro adhesion assays were performed as described by
Weinel et al. (1995). Cells were grown in complete medium
until reaching subcon¯uence and then were detached by
trypsin-EDTA (Gibco ± BRL, Gaithersburg, MD, USA;
0.05% trypsin). Cells were washed once with phosphatebu€ered saline and suspended in F-10 Nutrient Mixture
(Ham) (Gibco ± BRL) containing 0.5% bovine serum
albumin. Cell viability was determined by trypan blue
exclusion to be 495%. Flat-bottom 96-well enzyme-linked
immunosorbant assay (ELISA) plates (Corning, Corning,
NY, USA) were coated with 10 mg/ml ®bronectin (Sigma, St.
Louis, MO, USA), collagen type IV (Sigma) or laminin
(Sigma), or 10 mg/ml BSA or 100% serum overnight at 48C.
The wells were washed twice and blocked with 1 ± 5% BSA in
PBS for 2 h at 378C. After removal of BSA, 16105 of
CFPAC-1 and CAPAN-1 cells were seeded into each well of
the 96-well plate. Attachment was allowed for 1 h at 378C.
Nonadherent cells were carefully removed by washing three
times with PBS. Adherent cells were ®xed and stained with
0.1% crystal violet for 5 min. Cell density was determined by
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
measuring the absorbance at 630 nm on an ELISA reader
(Scimetrics Inc., USA). Actual cell numbers were determined
by normalizing dye uptake values with a calibration curve.
Northern and Southern blotting
Total RNA was isolated from the pancreatic cancer cell lines
by the acid guanidium thiocyanate phenol chloroform
(AGPC) extraction (Chomczynski and Sacchi, 1987). Ten
micrograms of total RNA was subjected to electrophoresis in
1% denaturing agarose gels, transferred to nylon membranes
(Stratagene, La Jolla, CA, USA) and ultraviolet radiationcrosslinked and baked. The membranes were hybridized with
32
P-dCTP-labeled 1.2-kb PstI uPA cDNA fragment and
glyceraldehyde-3-phosphate
dehydrogenase
(GAPDH)
cDNA probes.
Genomic DNA was prepared from pancreatic cell lines
according to the method described by Sambrook et al. (1989),
digested to completion with an excess amount of EcoRI
enzymes and then run in a 0.8% agarose gel. Southern
blotting was then performed as described (Sambrook et al.,
1989). The 1.2-kb PstI uPA cDNA fragment was used as
probe in Southern blot analyses.
Nuclear run-on assay
Nuclear run-on assay was performed as described (McKnight
and Palmiter, 1979; Groudine et al., 1981; Murdoch et al.,
1982). Cells (16108) were harvested and washed with PBS
and resuspended in a hypotonic bu€er containing 10 mM
HEPES, pH 7.9, 0.75 mM spermidine, 0.15 mM spermine,
0.1 mM EDTA, 0.1 mM ethylene glycol-bis (b-aminoethyl
ether) N,N,N'-tetraacetic acid (EGTA), 1 mM DL-dithiothreitol (DTT) and 0.5% Nonidet P-40. The cells were lysed in a
Dounce homogenizer. After centrifugation, the nuclear pellets
were resuspended in transcription bu€er (150 mM KCl, 5 mM
MgCl2, 1 mM MnCl2, 20 mM HEPES, pH 7.9, 10% glycerol
and 5 mM DTT). Nuclei were incubated at 208C for 30 min
in the presence of 1 mM each adenosine triphosphate (ATP),
guanidine triphosphate (GTP) and cytidine triphosphate
(CTP) and 200 mCi [32P]deoxyuridine triphosphate. Nuclei
were then collected and treated twice with deoxyribonuclease
(DNase) I and proteinase K. RNA was extracted with
phenol/chloroform and precipitated. Radioactive RNA was
treated again with DNase I and proteinase K followed by
phenol/chloroform extraction. Newly transcribed RNA
(66107 c.p.m./reaction) was incubated at 428C for 48 h
with nylon-immobilized cDNAs (3 mg/mm2) corresponding to
uPA, GAPDH and linearized plasmid vector. After
hybridization, the ®lter was washed at 658C with 0.16SSC
in the presence of 1% SDS and exposed on X-ray ®lm.
Nuclear extracts and electrophoretic mobility shift assay
(EMSA)
Nuclear extracts were prepared by the method of Andrews
and Faller (1991). After preparations 5 mg of nuclear extract
was incubated with 1 mg of poly(dI-dC) (Pharmacia) in
10 ml of binding bu€er (75 mM NaCl, 15 mM Tris-HCl,
pH 7.5, 1.5 mM EDTA, 1.5 mM DTT, 25% glycerol and
20 mg/ml BSA) for 30 min at 48C. 32P-labeled doublestranded oligonucleotides containing the kB site (under-
lined) found in the Igk gene (5'-CTCAACAGAGGGGACTTTCCGAGAGGCCAT) were incubated as probes.
Unlabeled double-stranded oligonucleotides (50-fold molar
excess) containing the Igk wild-type kB site and mutant kB
site (5' - CTCAACAGAGTTGACTTTCGAGAGGCCAT 3') (Chiao et al., 1994) were used for competition studies.
For antibody supershifts, 2 ml of the polyclonal antibodies
against p65 and p50 (Santa Cruz Biotechnology, Santa Cruz,
CA, USA) were preincubated for 30 min at room
temperature before probe was added. The probe was
allowed to bind for 20 min at room temperature. Reaction
mixtures were analysed on 4% polyacrylamide gels containing 0.256TBE (22.5 mM Tris, 22.5 mM borate, 0.5 mM
EDTA, pH 8.0) bu€er.
Site-directed mutagenesis
By using PCR-mediated site-directed mutagenesis, the wildtype kB element (5'-GGGAAAGTC-3') in uPA-CAT reporter
gene were substituted with a mutant kB site (5'CGGAAAATT-3' where underlined bases are the mutated
sequences). PCR-mediated site-directed mutagenesis were
performed using double-stranded, site-directed mutagenesis
kit from Stratagene (La Jolla, CA, USA).
Transient transfections, CAT assays and dual luciferase assay
Transient transfection was performed as described previously
(Chiao et al., 1994). Brie¯y, 1 ± 2 mg of the wild-type
(HIV[wt]CAT), or mutant (HIV[mu]CAT) kB plasmid that
contains two kB sites derived from the HIV long terminal
repeat region (30 bp) and the TATA box region of the TK
gene in the promoter the CAT reporter construct, or CAT
reporter plasmid driven by 2345 bp of uPA promoter
(uPAwtkBcat) (Hansen et al., 1992; Lengyel et al., 1995a)
or (uPAmukBcat) and 5 or 10 mg of NF-kB expression
plasmids (CMV-p65) (Chiao et al., 1994) or expression
plasmid of dominant-negative IkBa (CMV-IkBaM) (Van
Antwerp et al., 1996) were cotransfected into AsPC-1 or
CFPAC-1 cells with 1 ± 5 mg of b-actin LacZ, or b-actin
Fire¯y-luciferase and TK-Rellina luciferase reporter genes.
Forty-eight hours after transfection, cells were harvested and
the CAT reaction was performed as previously described
(Chiao et al., 1994). Brie¯y, the relative transfection
eciencies were determined by using the cotransfected LacZ
expression plasmid (1 mg, b-actin-LacZ), and subsequent bgalactosidase activities in the cell extracts were used to
normalize the transfection eciencies. Luciferase assay was
performed using Dual-LuciferaseTM Reporter Assay System
(Promega) and according to the instruction from the
manufacturer.
Acknowledgements
This work was supported in part by a MD Anderson
Cancer Center PRS grant and NIH grant (CA73675-01A1)
to PJ Chiao, and by NIH grant (CA71836-01) to DB
Evans. We thank Dr Marsha L Frazier for providing us
with pancreatic cancer cell lines, Dr Dan Van Antwerp and
Dr Inder M Verma for providing pIkBaM plasmid, Dr
Douglas Boyd for providing uPA promoter, Di Shen and
Lillie Larry for technical assistance, and Leslie Wildrick
and Pat Thomas for editorial assistance.
References
Andrews NC and Faller DV. (1991). Nucl. Acids Res., 19,
2499.
Auphan N, DiDonato JA, Rosette C, Helmberg A and Karin
M. (1995). Science, 270, 286 ± 290.
Axelrod JH, Reich R and Miskin R. (1989). Mol. Cell. Biol.,
9, 2133 ± 2141.
Baeuerle PA and Baltimore D. (1988). Cell, 53, 211 ± 217.
Baldwin Jr AS. (1996). Ann. Rev. Immunol., 14, 649 ± 681.
4561
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
4562
Beauparlant P, Kwan I, Bitaar R, Chou P, Koromilas AE,
Sonenberg N and Hiscott J. (1994). Oncogene, 9, 3189 ±
3197.
Bell S, Brackenbury RW, Leslie ND and Degen JL. (1990). J.
Biol. Chem., 265, 1333 ± 1338.
Bell SM, Connoly DC, Maihle NJ and Degen JL. (1993).
Mol. Cell. Biol., 13, 5888 ± 5897.
Besser D, Presta M and Nagamine Y. (1995a). Cell Growth
Di€., 6, 1009 ± 1017.
Besser D, Urich M, Sakaue M, Messerschmitt A, BallmerHofer K and Nagamine Y. (1995b). Oncogene, 11, 2383 ±
2391.
Birkedal-Hansen H, Moore WGI, Bodden MK, Windsow
LJ, Birkedal-Hansen B, DeCarlo A and Engle JA. (1993).
Crit. Rev. Oral Biol. Med., 42, 197 ± 250.
Blasi F and Verde P. (1990). Semin Cancer Biol., 2, 117 ± 126.
Brown K, Gerstberger S, Carlson L, Franzoso G and
Siebenlist U. (1995). Science, 267, 1485 ± 1488.
Chen ZJ, Hagler J, Palombella VJ, Melandri F, Scherer D,
Ballard D and Maniatis T. (1995). Genes Dev., 9, 1586 ±
1597.
Chiao P, Miyamoto S and Verma IM. (1994). Proc. Natl.
Acad. Sci. USA, 91, 28 ± 32.
Chomczynski P and Sacchi N. (1987). Anal. Biochem., 162,
156 ± 159.
Cordon-Cardo C, Fuks Z, Drobnjak M, Moreno C,
Eisenbach L and Feldman M. (1994). Cancer Res., 51,
6372 ± 6376.
Dano K, Andreasen PA, Grondahl-Hansen J, Kristensen P,
Nielsen LS and Skriver L. (1985). Adv. Cancer Res., 44,
139 ± 266.
DiDonato JA, Mercurio F and Karin M. (1995). Mol. Cell
Biol., 15, 1302 ± 1311.
DiDonato JA, Hayakawa M, Rothwarf DM, Zandi E and
Karin M. (1997). Nature, 388, 548 ± 554.
DiGiuseppe JA, Yeo CJ and Hruban RH. (1996). Adv. Anat.
Pathol., 3, 139 ± 155.
Du€y MJ, Reilly D, O'Sullivan C, O'Higgins N, Fenelly JJ
and Andreasen P. (1990). Cancer Res., 50, 6827 ± 6829.
Ghersa P, Hooft van Hiujsduinen R, Whelan J and
DeLamarter JF. (1992). J. Biol. Chem., 267, 19226 ±
19232.
Gilmore TD, Koedood M, Pi€at KA and White D. (1997).
Oncogene, 14, 1367 ± 1378.
Ginestra A, Monea S, Seghezzi G, Dolo V, Nagase H,
Mignatti P and Vittorelli ML. (1997). J. Biol. Chem., 272,
17216 ± 17222.
Gogly B, Groult N, Hornesbeck W, Godeau G, Pellat B.
(1998). Anal. Biochem., 15, 211 ± 216.
Grilli M, Chiu JJ-S and Lenardo MJ. (1993). Intl. Rev.
Cytol., 143, 1 ± 62.
Grimaldi P, Piscitelli D, Albanesi C, Blasi F, Geremia R and
Rossi P. (1993). Mol. Endo., 7, 1217 ± 1225.
Groudine M, Peretz M and Weintraub H. (1981). Mol. Cell.
Biol., 1, 281 ± 288.
Hansen SK, Nerlov C, Zabel U, Verde P, Hohnsen M,
Baeuerle PA and Blasi F. (1992). EMBO J., 11, 205 ± 213.
Harvey SR, Lawrence DD, Madeja JM and Abbey SJ.
(1988). Clin. Exp. Metastasis, 6, 431 ± 450.
Iademarco MF, McQuillan JJ, Rosen GD and Dean DC.
(1992). J. Biol. Chem., 267, 16323 ± 16329.
Jankun J, Keck RW, Skrzypczak-Jankun E and Swiercz R.
(1997). Cancer Res., 57, 559 ± 563.
Janulis M, Silberman S, Ambegaokar S, Gutkind JS and
Schultz RM. (1999). J. Biol. Chem., 274, 801 ± 813.
Je€ers M, Rong S and Vande Woude GF. (1996). Mol. Cell.
Biol., 16, 1115 ± 1125.
Kitajima I, Shinohara T, Bilakovics JDAB, Xu X and
Nerenberg M. (1992). Science, 258, 1792 ± 1795.
Lee FS, Hagler J, Chen ZJ and Maniatis T. (1997). Cell, 88,
213 ± 222.
Lenardo MJ and Baltimore D. (1989). Cell, 58, 227 ± 229.
Lengyel E, Stepp E, Gum R and Boyd D. (1995a). J. Biol.
Chem., 270, 23007 ± 23012.
Lengyel E, Singh B, Gum R, Nerlov C, Sabichi A, Birrer M
and Boyd D. (1995b). Oncogene, 11, 2639 ± 2648.
Lengyel E, Klostergaard J and Boyd D. (1995c). Biochim.
Biophys. Acta, 1268, 65 ± 72.
Liptay S, Schmid RM, Perkins ND, Meltzer P, Altherr MR,
McPherson JD, Wasmuth JJ and Nabel GJ. (1992).
Genomics, 13, 287 ± 292.
Lu D, Thompson JD, Gorski GK, Rice NR, Mayer MG and
Yunis JJ. (1991). Oncogene, 6, 1235 ± 1241.
Marksitzer R, Stief A, Menoud P-A and Nagamine Y.
(1995). J. Biol. Chem., 270, 21833 ± 21838.
Mars WM, Kim T-H, Stolz DB, Liu M-L and Michalopoulos
GK. (1996). Cancer Res., 56, 2837 ± 2843.
McKnight GS and Palmiter R. (1979). J. Biol. Chem., 254,
9050 ± 9058.
Minden A, Lin A, McMahon M, Lange-Carter C, Derijard
B, Davis RJ, Johnson GL and Karin M. (1994). Science,
266, 1719 ± 1723.
Moore BE and Bose HRJ. (1988a). Mutat Res., 195, 79 ± 80.
Moore BE and Bose HRJ. (1988b). Virology, 162, 377 ± 387.
Murdoch GH, Potter E, Nicolaisen AK, Evans RM and
Rosenfeld MG. (1982). Nature, 300, 192 ± 194.
Neri A, Chang CC, Lombardi L, Salina M, Corradini P,
Maiolo AT, Chaganti RSK and Dalla-Favera R. (1991).
Cell, 67, 1075 ± 1087.
Nerlov C, De Cesare D, Pergola F, Caracciolo A, Blasi F,
Johnsen M and Verde P. (1992). EMBO J., 11, 4573 ±
4582.
Ohno H, Takimoto G and McKeithan TW. (1990). Cell, 60,
991 ± 997.
Plaksin D, Baeuerle PA and Eisenbach L. (1993). J. Exp.
Med., 177, 1651 ± 1662.
Pollanen J, Stephens RW and Vaheri A. (1991). Adv. Cancer
Res., 57, 273 ± 328.
Rao JS, Steck PA, To®lon P, Boyd D, Ali-Osman F, StetlerStevenson WG, Liotta LA and Sawaya R. (1993). J.
Neurooncol., 18, 129 ± 138.
Regnier CH, Song HY, Gao X, Goeddel DV, Cao Z and
Rothe M. (1997). Cell, 90, 373 ± 383.
Rosewicz S and Wiedenmann B. (1997). Lancet, 349, 485 ±
489.
Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular
Cloning: A Laboratory Manual, 2nd edn. Cold Spring
Harbor Laboratory Press: Cold Spring Harbor, NewYork.
Sato H and Seiki M. (1993). Oncogene, 8, 395 ± 405.
Scheinman RI, Cogswell PC, Lofquist AK and Baldwin Jr
AS. (1995). Science, 270, 283 ± 286.
Sinicrope FA, Evans DB, Leach SD, Cleary KR, Fenoglio
CJ, Lee JJ and Abbruzzese JL. (1996). Clin. Cancer Res., 2,
2015 ± 2022.
Stacey DW, Roudebush M, Day R, Mosser SD, Gibbs JB
and Feig LA. (1991). Oncogene, 6, 2297 ± 2304.
Stacey KJ, Fowles LF, Colman MS, Ostrowski MC and
Hume DA. (1995). Mol. Cell. Biol., 15, 3430 ± 3441.
Stetler-Stevenson WG, Aznavoorian S and Liotta LA.
(1993). Ann. Rev. Cell Biol., 9, 541 ± 573.
Sylla BS and Temin HM. (1986). Mol. Cell Biol., 6, 4709 ±
4716.
Testa JE, Medcalf RL, Cajot JF, Chleuning WD and Sorbat
B. (1989). Int. J. Cancer, 43, 816 ± 822.
Van Antwerp DJ, Martin SJ, Kafri T, Green DR and Verma
IM. (1996). Science, 274, 787 ± 789.
Verma IM, Stevenson JK, Schwarz EM, Van Antwerp D and
Miyamoto S. (1995). Genes Dev., 9, 2723 ± 2735.
Voraberger G, Schafer R and Stratowa C. (1991). J.
Immunol., 1147, 2777 ± 2786.
Wang W, Abbruzzese JL, Evans DB, Larry L, Cleary K and
Chiao PJ. (1999). Clin. Cancer Res., 5, 119 ± 127.
uPA gene regulation by NF-kB in pancreatic cancer
W Wang et al
Watabe T, Yoshida K, Shindoh M, Kaya M, Fujikawa K,
Sato H, Seiki M, Ishii S and Fujinaga K. (1998). Int. J
Cancer, 77, 128 ± 137.
Weinel RJ, Rosendahl A, Pinschmidt E, Kisker O, Simon B
and Santoso S. (1995). Gastroenterol., 108, 523 ± 532.
Whelan J, Ghersa P, Hooft van Hiujsduinen R, Gray J,
Chandra G, Talabot F and DeLamarter JF. (1991). Nucl.
Acids Res., 19, 2645 ± 2653.
Wu H and Lozano G. (1994). J. Biol. Chem., 269, 20067 ±
20074.
Wu M, Lee H, Bellas RE, Schauer SL, Arsura M, Katz D,
FitzGerald MJ, Rothstein TL, Sherr DH and Sonenshein
GE. (1996). EMBO J., 15, 4682 ± 4690.
4563