Abstracts 2004.doc

M. Baay1
,*, V. Verhoeven2, J. Weyler3, F. Lardon1, D. Avonts2, P. Van Royen2 and
J.B. Vermorken1.
1 Dept. Medical Oncology, UA, 2 Centre for General Practice, UA, 3 Dept.
Epidemiology and Community Medicine, UA, Antwerp, Belgium.
* Tel. 03/8202576, fax 03/8202248, e-mail marc.baay@ua.ac.be
Introduction: The strong association between human papillomavirus (HPV) and
cervical cancer has led to the suggestion that the weaker association between
Chlamydia trachomatis (Ct) and cervical cancer is merely a reflection of sexual
behaviour. The aim of this study was to investigate whether the association between
Chlamydia and HPV is independent of sexual behaviour within the frame of a case-
control study.
Methods: Twenty-two Ct-positive cases from a previous study on the prevalence of Ct
in the general practice were included. Two controls were matched to each case for
age, sexual behaviour (coitarche, number of partners in the past year and condom
use). HPV detection was performed on DNA isolated from residual cells of vaginal
swabs by overnight Proteinase K digestion. DNA integrity was checked by beta-
globin PCR and HPV DNA was detected by the GP5+/6+ general primer PCR,
followed by non-radioactive enzyme immunoassay detection. Probes for all high-risk
or probably high-risk HPV types were available, as well as probes for 8 low-risk types
(Munoz, 2003).
Results: Eight of 22 cases (36.4%) had a high- risk HPV infection compared to 7 of 44
controls (15.9%). Furthermore, 5 cases had a low -risk HPV infection (22.7%), versus
4 cases (9.1%). This resulted in an odds ratio (Mantel- Haenszel estimate) for HPV
positivity of 8.5 (95% confidence interval, 2.0 - 35.3).
Discussion: We have shown in this study that the increased prevalence of HPV
infections amongst Ct-positive women is independent of sexual behaviour. It has
recently been shown that Ct infection is an independent risk factor for cervical cancer
development (Wallin, 2002). Our results are suggestive of a facilitating effect of the
Ct infection on HPV infection. Since it is well-known that only persistent HPV
infections potentially lead to cervical cancer development, we are currently
performing follow -up of the women in this study, to investigate whether prior or
current Ct infection has an effect on the outcome (persistence or clearance) of the
HPV infection.
Literature: Munoz, N., et al. Epidemiologic classification of human papillomavirus
types associated with cervical cancer. N Engl J Med., 348: 518- 527, 2003.
Wallin, K.L., et al. A population-based prospective study of Chlamydia trachomatis
infection and cervical carcinoma. Int J Cancer., 101: 371-374, 2002.
Zakia Belaid1, Pierre mineur2, Frédérique Hubin1, Chantal Humblet1, Myriam
Verlaet3, Charles M Lapière2, Jacques Boniver1, Betty V Nusgens2 and Marie - Paule
1- Dpts of Cytology & Histology and Pathological Anatomy, 2-Laboratory of Connective
Tissues Biology, 3- Research Center in Cellular and Molecular Neurobiology, * Grant FNRS -
Télévie, CRCE,Ulg.

Bone marrow is a tissue consisting of heterogeneous populations of cells including
hematopoietic stem cells, endothelial cells and other stromal cells ( adipocytes,
fibroblasts, osteoclasts etc….)
The bone marrow microenvironment plays a critical role in regulating the growth and
differentiation of hematopoietic cells by secretion of growth factors and cytokines as
well as by direct cell-to-cell contacts. Recent evidence suggests that vascular growth
factor (VEGF) is involved in these interactions.We have thus compared human iliac
crest bone marrow ( hematopoietic bone marrow ) and femoral bone marrow (fatty
bone marrow) for the expression of VEGF , its receptors (VEGFR-1, VEGFR-
2/KDR) and neuropilin-1 (NRP- 1), a VEGF165 specific co-receptor and have shown
that adipocytes are responsible for the high expression of NRP-1 in fatty bone
marrow. NRP-1 has previously been reported to be implicated in angiogenesis ; our
observation suggest that it could also be implicated in the regulation of hematopoiesis.
Its high expression in non hematopoietic bone marrow seems to be inconsistent with
the hypothesis that NRP-1 exerts a postive effect on hematopoiesis (R.Torjman and
al ;Blood 1999, 94 :2301-2309). To identify the role of NRP -1 and the role of
adipocytes in the regulation of hematopoiesis, adipocytes isolated from femoral bone
marrow have been studied in vitro. In culture, adipocytes ( unilocular fat cells) adopt
fibroblastic morphology ( fibroblast-like fat cells), they still express NRP-1 , have lost
the expression of the adipocytes marker aP2 ( adipocyte lipid binding protein), they
express PPAR gamma 2 (peroxisome proliferator activated receptor) and LPL
(lipoprotein lipase) at low levels , and display a mitotic activity after stimulation by
VEGF165. The use of adipogenic medium allows the fibrolast-like fat cells to
differentiate into unilocular fat cells which reexpress adipocytes marker aP2, PPAR
gamma 2 and LPL at high levels but have down regulated the expression of NRP- 1.
This down regulation seems to be induced by steroids in the adipogenic medium
and seems to be cell type specific as it is not observed in endothelial cells. Since the
fibroblast-like fat cells still express NRP- 1, their influence on hematopoietic
progenitors (CD34+) in vitro is under investigation ; studies are also in progress to
analyse their properties in an in vivo hematopoiesis model developped in our
laboratory (F.Hubin).
Agnès Brouet
, Philippe Martinive, Pierre Sonveaux and Olivier Feron
UCL Medical School, Unit of Pharmacology and Therapeutics (FATH 5349), 53 Ave
E. Mounier, B-1200 Brussels, Belgium (agnes.brouet@mint.ucl.ac.be)
Caveolin- 1, the structural protein of the plasmalemmal invaginated microdomains
named caveolae, is particularly abundant in endothelial cells (EC) where it regulates
various functions including transcytosis and permeability. The well- established
inhibitory interaction between caveolin and the endothelial NO synthase (eNOS)
recently led us to examine whether the modulation of the stoichiometry of this
complex could impact on angiogenesis. In agreement with the pro-angiogenic
properties of NO, we documented that synthetic peptides derived from the caveolin
scaffolding domain could block angiogenesis. In tumor development, however,
caveolin is thought to constitute a key switch through its function as a tumor
suppressor and as a promoter of metastasis and chemoresistance. We therefore
reasoned that to exploit the anti-tumor effects of caveolin without increasing tumor
cell dissemination, the increase in caveolin abundance should mainly target EC lining
tumor blood vessels. For this purpose, we used cationic liposomes which are delivery
systems effective at targeting tumor versus normal vascular networks.
We first validated that in vitro EC transfection with caveolin DNA-cationic
lipocomplex led to anti-angiogenic effects. Accordingly, both VEGF-induced EC
migration determined in a scratch assay and tube formation on Matrigel were
dramatically reduced following caveolin lipofection (-73% and -61% at 48h,
respectively, versus empty plasmid transfection; P<0.05, n=4). The DNA-
lipocomplex was then administered through the tail vein of tumor-bearing mice. A
dramatic tumor growth delay was observed in mice treated with caveolin DNA
(doubling time = 10±1.5 vs 5±1 days with empty vector; P<0.01, n=8). That a direct
interaction between recombinant caveolin and native eNOS occured was validated by
immunoprecipitating eNOS from tumor extracts. The much larger proportion of
caveolin-1 in the immunoprecipitates from tumors exposed to caveolin DNA
confirmed that the transgene was efficiently delivered in the tumors, and since eNOS
is exclusively expressed in EC, that the caveolin expression occurred in tumor
vascular structures. Similar co-immunoprecipitation experiments from other organs
including lung and liver did not reveal any change in the caveolin-eNOS interaction.
Finally, we used Laser Doppler imaging and microprobes, and documented in the
early time after lipofection (e.g. when macroscopic effects on tumor growth are not
detectable) that caveolin overexpression led to a significantly reduced tumor blood
flow likely to reflect tumor vasculature collapse.
In conclusion, our study reveals that by exploiting the exquisite regulatory interaction
between eNOS and caveolin and the propensity of cationic lipids to target EC lining
tumor blood vessels, caveolin plasmid delivery appears to be a safe and efficient way
to block angiogenesis in solid tumors.
*I agree to have the abstract released on the BACR web site before the conference in Jan2004
Erik Bruyneel
, Filip Ryniers1, Carol Box2, Quang-Dé Nguyen3, Fons Verdonck4,
Bart Van Haesebroeck2, Christian Gespach2, Jean Willems4 and Marc Mareel1.
1Laboratory of Experimental Cancerology, Ghent University, B-9000 Ghent,
Belgium; 2Cell Signalling Group, The Ludwig Institute for Cancer Research, London,
UK; 3INSERM U482, Signal Transduction and Cellular Functions in Diabetes and
Digestive Cancers, Hôpital Saint-Antoine, F-75571 Paris Cedex 12, France;
4Interdisciplinary Research Center, KULAK, B-8500 Kortrijk, Belgium.
Cancer is a disease of growth, differentiation, invasion and ectopic survival. The cell-
cell adhesion and signal-transducing molecule E-cadherin is essential for maintaining
correctly epithelial intercellular adhesion (1). Opistoporin (OC44) and parabutoporin
(PS45) are a- helical pore-forming peptides from scorpion venom.
The E- cadherin function of colonic HCT -8/E11and kidney MDCKts.srcCl2 (40°)
cells is inhibited in the Collagen I (CI) and the Fast Aggregation Assay.
Administration of the pharmacological inhibitor of c-src, M475, and E-cadherin
antibody MB2 restores the E-cadherin function. 3-dimensional Collagen invasion of
CT5.3 myofibroblasts is stimulated by both scorpion peptides and prevented by
inactivation of c-src.
PS45 and OC44 stimulate also Collagen I invasion of kidney MDCK-T23, enteric
Caco-2, thyroid KAT -10, colonic PC and not of HCT-8/E41 cells. Addition of an
antibody against PS45 prevents stimulation of invasion by PS45 but not by OC44.
Both scorpion peptides signal via Gao/i, RhoA/ROCK, TXA-R, PKA, PKC and not
via PI3K.
HCT-8/E11and kidney MDCKts.srcCl2 invade via activation of c-src, Rac1 and
formation of lamellipodia.
The micro-ecosystem of cancer cells, polymorphonuclear cells and myofibroblasts is
stimulated by both scorpion peptides. Chemotaxis and activation of Rac1 in
polymorphonuclear cells is stimulated by PS45 and OC44 and dependent on c-src.
Reference :
1) Mareel and Bracke , Encyclopedia of Cancer, Sec. Edition, Academic Press,
S. Califice
, V. Castronovo and F.A. van den Brûle
Metastasis Research Laboratory, Experimental Cancer Research Center (C.R.C.E.),
Pathology B23/ -1, B-4000 Sart Tilman, University of Liège, Belgium. E-mail:
Introduction. Our research project aims to determine the functions played by
galectin -3 (Gal-3), a β-galactose-binding lectin, in the nucleus of prostate cancer cells.
Gal-3, a 30 kDa protein, is involved in several physiological and pathological events.
Extracellular Gal-3 mediates cell-cell and cell- matrix interactions, and cytosolic Gal-3
mediates apoptosis control. Several studies , including ours, have demonstrated
decreased Gal-3 expression in cancer cells compared with normal cells. Cytoplasmic
expression and nuclear exclusion are characteristics of invasive cancer cells from
endometrium, colon and prostate carcinomas. Our project aims to better understand
the biological significance of nuclear Gal-3 in cancer cells.
Methodology. We have cloned the Gal-3 ORF cDNA upstream of 3 nuclear
localization sequences (NLS) in an expression vector, that was transfected in LNCaP
prostate carcinoma cells that do not express endogenous Gal-3, in order to verify the
hypothesis according to which the expression of Gal-3 in the nucleus could decrease
or prevent the expression of cancer phenotype. Resistant clones were screened using
Western blotting and immunohistochemistry. Apoptosis was induced by actinomycin
D treatment and measured by TUNEL and annexin-V and propidium iodide staining
analyses. Anchorage-independent growth experiments were performed by counting
colonies formed in soft agar after one month. 106 cells were injected in nude mice and
tumors were measured twice a week.
Results. Matrigel invasion assays revealed the nuclear Gal- 3-expressing (nGal-3)
clones to be less invasive than the cytoplasmic Gal-3-expressing (cGal-3) clones and
the parental cell line. Less apoptosis was observed in the cGal-3 clones than in the
parental cell line, consistent with the reported antiapoptotic role of cytoplasmic Gal- 3.
Conversely, the nGal- 3 clones displayed increased sensitivity to apoptosis compared
to control vector-transfected clones and the parental cell line, suggesting a specific
role for nuclear Gal-3. Increased cas pase-8 expression and PARP cleavage levels
were observed for the nGal-3 clones. cGal-3 clones were characterised by increased,
whereas nGal-3 clones presented with decreased anchorage-independent growth,
when compared to a control clone. In vivo xenograft experiments showed larger
tumors for the cGal-3 clones and control clones than for the nGal-3 clones. Moreover,
tumors developed from cGal-3 clones demonstrated higher vascular densities than
those developed from control and nGal-3 clones, as detected by Factor VIII
immunostaining. The TUNEL method allowed the detection of apoptotic cells only at
the border of tumors developed from nGal- 3 clones.
Conclusions. These data demonstrate that nGal- 3, contrary to cGal-3, could decrease
the invasive phenotype in prostate cancer. Further experiments will better characterize
the functions of nGal-3 and cGal-3 in prostate cancer cells.
I agree to have the abstract released on the BACR web site before the conference in January 2004.
Thibaut Dassesse1,
Xavier de Leval*, Pascale Jackers1, Jean-Michel Dogné* and
Vincent Castronovo1
1Metastasis Research Laboratory, CRCE - Experimental Cancer Research Center,
University of Liège.
*Natural and Synthetic Drugs Research Center, University of Liège.
Cyclooxygenase (COX) inhibitors, especially COX- 2 selective inhibitors, exhibit anti-
angiogenic activity in both in vitro and in vivo models. This activity is, at least in part,
associated to their ability to inhibit prostanoid biosynthesis in endothelial cells (EC).
Among these prostanoids, thromboxane A2 (TXA2) has been demonstrated to be
involved in angiogenesis. Furthermore, drugs able to inhibit the thromboxane pathway
such as COX inhibitors, thromboxane synthase inhibitors or thromboxane A2 receptor
antagonists exert anti- angiogenic activities, leading to the suggestion that modulation
of TXA2 production by EC should be an interesting anti-angiogenic approach. The
aim of this work was to evaluate original thromboxane modulators displaying both
thromboxane receptor antagonist and thromboxane synthase inhibitory properties in
the rat aortic ring assay for angiogenesis. Evaluation of the microvascular outgrowth
was performed by computer assisted images analysis using the NIH Image 1,62
software. Results obtained demonstrated that TXA2 agonist acts as a pro-angiogenic
agent and that several of the investigated original TXA2 dual inhibitors displayed a
significant dose-dependent inhibition of both spontaneous and VEGF-induced
In this study, we showed the involvement of TXA2 in ex-vivo angiogenesis.
Therefore, further ongoing investigations will investigate the role of TXA2 in in vitro
human endothelial cells proliferation, adhesion and migration assays.

I agree - don't agree" to have the abstract released on the BARC web site
before the conference in January 2004

Bram De Craene
(1), Erik Bruyneel (3), Frans Van Roy (2) and Geert Berx (1)
1: Unit for Molecular and Cellular Oncology and 2: Molecular Cell Biology Unit,
Department for Molecular Biomedical Research, VIB-Ghent University, B-9052 Zwijnaarde, Belgium 3: Department of Experimental Cancerology, University Hospital Ghent, B-9000 Malignancy of carcinoma cells is characterized by loss of both cell-cell adhesion and cellular differentiation. The epithelial cell-cell adhesion protein E-cadherin is a genuine tumor suppressor as well as an established invasion suppressor. Transcriptional downregulation of E-cadherin in various epithelial tumors appears to be a main event during tumorigenesis. A variety of transcriptional regulators important in development and tumor progression have been described like Snail, Slug, SIP1, dEF-1, E12/E47. Members of the Snail family seem to be involved in the transcriptional regulation of epithelial mesenchymal transitions (EMT) in both vertebrates and invertebrates. Snail proteins are expressed during embryonic development in mesoderm-derived tissues and their expression correlates with loss of adhesion and induction of cell migration. Snail is able to bind to E-boxes in the E-cadherin promotor; mutation of these E-boxes impairs transcriptional repression by Snail (Batlle et al., 2000). Binding to the E-box sequences is mediated through an evolutionary conserved cluster of zinc fingers at the carboxyterminus of the protein. Upon Snail induction epithelial cells get converted to a fibroblastic phenotype, concomitantly with loss of epithelial markers and gain of mesenchymal markers (Cano et al., 2000). Slug is a close relative of Snail, and expression of this protein is also correlated with the loss of E-cadherin transcription. Slug expression seems to show a much stronger correlation with loss of E-cadherin in breast cancer cell lines than Snail (Hajra et al., 2002). Furthermore, some Slug family members play an important anti-apoptotic role. We generated human colon cancer cell lines with conditional expression of Snail and Slug. In these cell lines induction of the different transcription factors resulted in an EMT-like process, with a dramatic change in morphology. Induced cells showed loss of intercellular adhesion coinciding with induction of invasiveness. Comparative differential gene expression analysis using cDNA microarrays was performed for the different model cell systems. Gene expression profiles indicated a group of commonly regulated genes, but also distinct groups of genes specifically modulated by only one of the Snail family members. Identification of possible target genes in addition to E-cadherin may explain in more detail the role of the different transcriptional repressors in EMT and invasion by cancer cell. References Batlle, E., Sancho, E., Franci, C., Dominguez, D., Monfar, M., Baulida, J., and de Herreros, A.G. 2000. The transcription factor Snail is a repressor of E-cadherin gene expression in epithelial tumour cells. Nat. Cell Biol. 2:84-89. Cano, A., Perez-Moreno, M.A., Rodrigo, I., Locascio, A., Blanco, M.J., del Barrio, M.G., Portillo, F., and Nieto, M.A. 2000. The transcription factor Snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat. Cell Biol. 2:76-83. Hajra KM, Chen DY, and Fearon ER. 2002. The SLUG zinc-finger protein represses E-cadherin in breast cancer. Cancer Res. 2002 Mar 15;62(6):1613-16188. presenting author: Bram De Craene address : DMBR - VIB-Ghent University - FVMS building - Technologiepark 927 - 9052 email address: Bram.DeCraene@dmbr.ugent.be

foot note: I agree to have the abstract released on the BACR web site before the conference in January
M. De Ridder
, V. Verovski, D. Van den Berge, A. Sermeus, C. Monsaert, N.
Wauters, I. Beirnaert, M. Darville, D. Eizirik & G. Storme
Cancer Research Unit, Vrije Universiteit Brussel

Background: The developme nt and progression of solid tumors is dependent on a complex
network of different cell types including tumor cells, endothelial cells and macrophages.
Tumor-associated macrophages (TAM) are recruited from peripheral blood monocytes by
chemokines and respond to the tumor microenvironment by secreting growth factors,
cytokines and NO, which are pro -angiogenic and promote tumor growth (1). However,
activated TAM may produce a high level of TNF- α and NOthereby causing tumor
regression. Till now, no attempts have been undertaken to exploit TAM as a source of
radiomodulatory cytokines, although TNF-α and other cytokines are known to radiosensitize
tumor cells through the apoptotic pathway (2) and by activating NF-κB signaling towards
inducible nitric oxide synthase (iNOS; 3, 4). In this study, we examined the possibility to
activate macrophages by lipid A, a non-toxic derivate of LPS, in order to radiosensitize tumor
Methods: To explore this hypothesis, we exposed RAW 264.7 and J774A.1 macrophage
cultures to lipid A for 24h. Afterwards, the conditioned medium (CM), which contains the
cytokines secreted by the macrophages, was collected and analyzed for TNF-α, IL-6, IL-1β
and interferon-? ?by ELISA.
Next, EMT-6 mammary carcinoma cells were exposed to CM and examined for the
expression of iNOS by Western blotting, for the production of NO by the Griess assay and for
their hypoxic cell radioresponse by clonogenic assay. To clarify the mechanism of iNOS
induction, we finally used a series of iNOS gene vectors with mutations of the proximal and
distal NF-κB binding sites in transient transfection experiments.
Results: RAW 264.7 and J774A.1 macrophages were activated with a plasma relevant
concentration of lipid A (3 µg/ml) and analyzed for the secretion of cytokines. Both
macrophage cell lines secreted TNF-α and IL-6, but not IL-1β or interferon-γ.
EMT-6 tumor cells displayed a low constitutive expression and activity of iNOS, which
drastically increased after treatment with CM. CM caused a 6 and a 3 times increase in
luciferase activity using NF-κB tandem and wild type iNOS promoter containing plasmids
respectively. Mutation of the proximal NF-κB binding site significantly decreased the
luciferase activity, whereas mutation of the distal NF-κB binding site did not influence it.
Finally, CM caused a 2,4 times enhancement in tumor cell radiosensitivity, which could be
largely reconstituted by a mixture of TNF-α and IL-6.
Conclusion: This study for the first time demonstrates that activated macrophages may
drastically increase the radiosensitivity of tumor cells through the release of pro -inflammatory
cytokines, which trigger NF-αB signaling towards the iNOS gene.
(1) Bingle et al, J Pathol. 196(3):254-65, 2002
(2) Kimura et al, Cancer Res. 59(7):1606-14, 1999
(3) De Ridder et al, Br. J. Cancer, 88(1):120-4, 2003
(4) De Ridder et al, Int J Radiat Oncol Biol Phys, 57(3):779-86, 2003
Corresponding author: Mark.De.Ridder@vub.ac.be
I agree to have the abstract released on the BACR website before the conference in January
Deroanne C, Hamelrijckx D., Lambert Ch.A., Lapière Ch.M. and Nusgens B. V.
Laboratory of Connective Tissues Biology, GIGA/CRCE, University of Liège, Tour de
Pathologie, B23/3, 4000 Liège, Sart Tilman, Belgium.

In fibroblasts, matrix metalloproteinase (MMP)-1 expression is modulated by
integrin - mediated changes in cell shape and by disruption of the cytoskeleton. The
small GTPases of the Rho family are key intermediates in cellular signaling triggered
by clustered cell adhesion receptors. The implication of these GTPases in the control
of MMP-1 expression has been tested by means of constitutively active-, or dominant
inactive forms of these molecular switches. However, these tools are not adequate to
use with primary cells and they are probably not absolutely specific for one protein.
The interfering RNA (siRNA) technology is best suitable for investigating the
involvement of RhoA, Rac1 and Cdc42 in the control of MMP-1 expression in
primary human skin fibroblasts (HSF). The target sequences of Rac1 and Cdc42 were
chosen in the same region as for RhoA that was successfully ablated in primary cells
(Deroanne et al., 2003). Transient transfection of HSF by the calcium phosphate
procedure with the respective specific siRNA induced a 90% reduction in the RhoA
and Cdc42 protein level and an 80 % reduction in the Rac1 protein. The repression of
these proteins was maintained for up to 7 days post-transfection and confirmed with
pull-down assays. Ablation of RhoA did not induce any significant morphological
alterations while ablation of Rac1 decreased lamellipodia formation and ablation of
Cdc42 induced a "dendritic" morphology. The reduction of Cdc42 protein level
induced a 5 fold overexpression of MMP-1 which was detected for up to 7 days post-
transfection. In parallel, a slight increase of interleukin 1beta was also observed but it
does not seem to be related to MMP-1 overexpression. Finally, simultaneous
repression of Rac1 and Cdc42 or RhoA and Cdc42 did not reverse the MMP-1
overexpression suggesting that it is independent of the Rac1 and RhoA pathways.
The siRNA technology revealed an unexpected role of Cdc42 in the control of MMP-
1 expression.
Deroanne C., Vouret- Craviari V., Wang B. and Pouysségur J. EphrinA1 inactivates
integrin - mediated vascular smooth muscle cell spreading via the Rac/PAK pathway.
J. Cell Sci. 116(2003), 1367-1376.

Presenting author : Christophe Deroanne, PhD
Laboratory of Connective Tissues Biology, GIGA/CRCE, University of Liège, Tour de Pathologie, B23/3, B-4000 Sart-Tilman, Liège 1, Belgique I agree to have the abstract released on the BACR web site before the conference in January 2004 ABSTRACT n° 10
Cédric Detry
, Michael Chaplet, Viviane Bougnet, Vincent Castronovo, and Akeila
Metastasis Research Laboratory, University of Liège, BELGIUM

Bone sialoprotein (BSP) is a secreted glycoprotein primarily found in the mineral
compartment of developing bones. BSP is detected in a variety of human cancers and
particularly those that metastasize to the skeleton. High expression of BSP in breast
and prostate primary carcinomas is associated with progression and bone metastases
development. This discovery led us to study the regulation of BSP transcription in
human breast cancer cells compared with cells from the osteoblastic lineage. We
focused our study on MCF7 breast cancer cells and Saos- 2 human osteosarcoma cells
which constitutively express BSP. As a negative control we used CaOV-3 ovary
cancer cells because they have a low metastatic impact on bones and do not express
BSP. We first constructed a series of promoter deletions cloned upstream of a
luciferase reporter gene that were transiently transfected in both cell lines.
Interestingly, we found that a -84 bp human BSP promoter- luciferase construct is
sufficient for maximal reporter gene expression in the MCF- 7 breast cancer cell line.
This promoter region analysis allowed us to identify a CRE element (TGACATCA).
Using mutated promoter constructs, we evaluated the impact of this element on the
activity of the human BSP promoter. Mobility shift assays using probes containing the
CRE element revealed a major DNA-protein complex attributed to the CRE element.
Our study demonstrated that three transcription factors, CREB-1, Jun D and Fra-2
were responsible for BSP gene transactivation in MCF-7 cells as well as in Saos-2
osteoblastic -like cell line. We also showed by Western blot that MCF- 7 and Saos-2
cells express a high level of Fra-2 and Jun D while CaOV-3 cells express only low
amounts of these transcriptional factors. As such, BSP transcriptional regulation
through the CRE element represents a new demonstration of a common regulatory
mechanism utilized by both osteoblasts and breast cancer cells and supports the
hypothesis of osteomimetism. Indeed, our results suggest that the expression of
critical transcription factors, such as Fra-2 and Jun D in breast cancer cells, confers to
these cells a phenotype that mimics that of osteoblasts.
I agree to have the abstract r eleased on the BACR web site before the conference in January 2004
Devoogdt Nick
, Hassanzadeh Gh. Gholamreza, Revets Hilde, and De Baetselier
Department of Molecular and Cellular Interactions, Vlaams interuniversitair Instituut
voor Biotechnologie, Vrije Universiteit Brussel, Belgium
Secretory Leukocyte Protease Inhibitor (SLPI), an epithelial-specific serine protease
inhibitor, exerts broad tissue-protec tive functions: on one hand it counteracts
inflammatory responses and invading pathogens, and on the other repairs the damaged
tissue. Here we present data investigating the role of SLPI in cancer.
We report that subcutaneous growth of low-malignant 3LL-S cells coincides with
infiltration of inflammatory cells and an increase in the malignancy of these cancer
cells. Using TNF-alpha deficient mice we demonstrated that the increased malignancy
upon subcutaneous growth of 3LL-S cells is partially dependent on the presence of
this cytokine. Applying the SSH method to this cancer-model, we identified SLPI as
one of the genes whose expression correlates with malignancy of the cancer cells.
Using transfection experiments, we demonstrated that SLPI plays a causative role in
the malignant behavior of these cells, and that this is dependent on its capacity to
inhibit serine proteases (1). We also show that both in vivo and in vitro SLPI
expression in 3LL-S cells is induced by TNF-alpha, suggesting that SLPI expression,
and eventually malignant behavior, is, at least in part, regulated by this inflammatory
cytokine. In conclusion, this study establishes a new link between TNF- alpha and
cancer cells, whereby macrophage-derived TNF-alpha contributes to tumor
progression and metastasis via induction of SLPI expression in cancer cells.
Furthermore, using cDNA dot blot hybridizations we show that in human
gynecological cancer tissue, but not in other types tested, SLPI expression is
frequently increased as compared to normal tissue.
1. Secretory Leukocyte Protease Inhibitor promotes the tumorigenic and metastatic potential of cancer cells; Proc Natl Acad Sci USA; 100:5778-5882,2003; Devoogdt N., Hassanzadeh Ghassabeh G., Zhang J., Brijs L., De Baetselier P., Revets H. I agree to have the abstract released on the BACR website before the conference in January 2004 ABSTRACT n° 12
Dierickx K.,
Morandini R., Ghanem G.
L.O.C.E, J. Bordet-institute, Free university of Brussels, Belgium
Karen Dierickx
L.O.C.E - Institut J. Bordet
Rue Héger-Bordet 1
1000 Bruxelles
PSF is an alkylating oligopeptid e containing m-sarcolysin (m-SL), an isomer of the
anticancer drug melphalan. It showed promising antitumor activity both in vivo and
in vitro. However its mechanism of action needs to be uncovered. To this aim, we
conducted a study consisting in successive phases; drug Formulation, in the presence
of blood components and HBL melanoma cells (MC) over time. By HPLC analysis,
we found that PSF disappears very quickly in the blood along with the generation of
m-SL. The latter had time kinetics compatible with enzymatic catalysis, presumably
by some RBC proteases associated to the RBC membranes. In order to check this
hypothesis, the reaction was examined under conditions inhibiting the enzymatic
activities. We obtained a significant inhibition of the main product generation at zero
degrees as well as with metal chelators and various other enzyme inhibitors. In order
to establish the saturation conditions for the RBCs we detected, in the supernatant of
different amounts of RBC incubated with the same concentration of PSF: (1) PSF and
its metabolites (HPLC analysis), (2) the alkylating activity (Epstein reagent), (3) the
cytotoxic effect (MTT-essay).
Melanoma cells showed a significantly higher capacity than RBCs to bind and
degrade PSF. In addition, conditioned medium from melanoma cell culture was able
to efficiently degrade PSF. PSF displayed the same toxicity pattern in both short and
long term cell exposures illustrating a rapid penetration into the cells. A dose finding
study in nude mice allowed us to administer up to 20 mg/kg IP single dose that was
able to induce, unlike an equivalent dose of melphalan or m-SL, not only inhibition of
growth but also tumor regression without any significant weight loss. Dose-response
curves where achieved and fractionated doses had significant better effects on tumor
regression and regrowth than a single dose. These results strongly suggest that
melanoma tumor cells and tumour sites, may be a target for the toxic activity of PSF
owing to the much higher load in proteolytic enzymes and at the same time, it is less
toxic to normal tissues displaying poor or no proteolytic activity thus limiting side-
M. Fournier
, M. Lopez y Cadenas, V. Bours and C. Herens.
Dept. Human Genetics, University of Liege, Belgium
BACKGROUND : Chronic myeloid leukaemia is a clonal hematologic malignancy
characterized by the Philadelphia chromosome, which results from the translocation
t(9;22)(q34;q11). At the molecular level, the translocation juxtaposes 3'- DNA
sequences of the ABL oncogene with 5'-DNA sequences of the BCR gene, giving rise
to two fusion genes: BCR- ABL on the derivative 22 and ABL- BCR on the derivative
9. Inclusion of fluorescence in situ hybridisation in the diagnosis of the pathology has
led to the observation of a deletion of the 5'-ABL region in about 10% of the patients.
Losses of the 3'- BCR region are simultaneously observed in more than 90% of the
OBJECTIVES AND METHODS : As the deletions in 5'- ABL and 3'- BCR span the
breakpoints on chromosome 9, we studied the expression of the reciprocal ABL- BCR
gene. We adapted the RT- PCR assay previously described by Uphoff et al (1999) to
detect the ABL- BCR 1b transcript.
RESULTS : Our results show that 36% of the tested patients do not express the ABL-
BCR transcript. However, the distribution of non- expressing ABL- BCR patients is
different according to the deletion status. 100% of patients with FISH deletions and
26,1% of non-deleted patients do not express this gene.
CONCLUSION : our results demonstrate that all the cases with deletions lacked
expression of ABL- BCR but also that 6/14 patients lacking ABL- BCR expression do
not have a deletion detectable by FISH. This observation suggests the existence of
other mechanisms, which can hinder the transcription. First, microdeletions of the 5'-
ABL region may be under the level of detection of FISH. Secondly, about 10% of
CML- patients have a breakpoin t upstream of ABL exon 1b, removing the site at
which ABL transcription is initiated. Thirdly, complex translocations have 5'- ABL
and 3'-BCR sequences on separate chromosomes preventing the formation of the
reciprocal gene fusion.
, HOSTENS Jeroen, VAN DEN HEMEL Maureen and
Molecular Pathophysiology and Experimental Therapy Unit, Department of
Molecular Biomedical Research, VIB, Ghent University, Belgium.

TNF is a pleiotropic cytokine, mainly produced by macrophages. Beside the ability to
destroy solid tumours, both in patients and animals, TNF has also shock-inducing
properties, which limits the use of TNF as an antitumour drug to locoregional
treatment through isolated limb perfusion. It is necessary to elucidate the downstream
pathways to expand the use of this cancer treatment by a selective mimicking of the
antitumour activities.
In order to determine whether these effects are due to cytotoxicity directed onto the
tumour cells or are host-mediated, we overexpressed a dominant negative TNF-RI in
B16BL6 cells. The expression and dominant negative activity of the mutant TNF-RI
were analysed with different assays: immunofluorescence confirmed expression, a
cytotoxicity assay, EMSA, and FACS analysis for MHCI upregulation and confirmed
the dominant negative character. These cells showed no difference in WT and TNF-
RII-/- mice in tumour regression when treated with TNF (+IFN-γ), which indicated
that only host-mediated effects occur during tumour destruction by TNF.
By using C57Bl6 (control), B6TNF-RI-/- and B6TNF-RII-/- m ice we tried to determine
which TNF-receptor is involved. Tumour-bearing control and B6TNF- RII-/- showed
complete regression of the tumour after 1 week treatment with TNF (+IFN-γ), in
contrast to B6TNF-RI-/-, where no effect of mTNF (+IFN-γ) on the tumour growth
was seen. These results confirmed our hypothesis that TNF mediates its antitumour
effect through a host -mediated mechanism, instead of acting directly onto the tumour
cells. Also toxic actions of TNF were only in WT and B6TNF-RII-/- m ice observed,
which indicated that not only the antitumour effect, but also the toxicity are due to
TNF-RI triggering. These data were confirmed with other tumours, namely PG19,
LLC and EL4.
To analyse the role of TNF-RI stimulation in immune cells in the tumour destructing
capacity of TNF, we performed a bone marrow transplantation in C57Bl6 and
B6TNF- RI-/- mice with bone marrow derived from B6TNF-RI-/- mice or C57Bl6 mice.
B16BL/6 tumours were as effectively destroyed in mice containing B6TNF-RI-/- bone
marrow cells, as mice with WT bone marrow cells. This indicated that although
certain immune cells are necessary in the antitumour activity of TNF, as has been
described several time, TNF- RI stimulation on these cells is dispensable.
Tumour-experiments in endothelial specific TNF-RI reactivation knock-out mice
showed that endothelial cells are most probably the target cells of TNF.
An Goethals
Department of Molecular Biomedical Research
Technologiepark 927
9052 Ghent (Zwijnaarde), Belgium
Tel: 09/3313714
Fax: 09/3313609
e-mail: An.Goethals@dmbr.UGent.be
Christian Herens
1, Frédéric Baron2, Mauricette Jamar1, Anne-Cécile Hellin1,
Mickaël Fournier1 and Vincent Bours1.
(1) Department of Human Genetic, University of Liège, CHU, Liège, Belgium.
(2) Department of Medicine, division Haematology, University of Liège, CHU,
The Philadelphia (Ph) chromosome is the hallmark of chronic myeloid leukemia (CML) and results from the reciprocal translocation t(9;22)(q34;q11). At the molecular level, the translocation juxtaposes 3'DNA sequence of the ABL oncogene with 5'DNA sequences of the BCR gene, resulting in the generation of a chaemeric BCR- ABL fusion gene. The BCR- ABL tyrosine kinase protein has enhanced activity compared to the wild type ABL tyrosine kinase and is responsible for the pathogenesis of the disease. Conventional therapies include a- interferon, hydroxyurea and progenitor stem cell transplantation. Recently, a specific inhibitor of the tyrosine kinase domain of the BCR- ABL protein, imatinib mesylate (trade name: gleevec, glivec; experimental name: STI 571), has been shown to induce growth arrest or apoptosis in BCR-ABL expressing hematopoietic cells. Imatinib entered clinical trials in 1998 and since, have been shown to induce dramatic hematologic and cytogenetic responses. While imatinib has remarkable clinical efficacy in CML, emergence of clonal chromosome aberrations in Ph- negative cells during the treatment has been recently described. We report here two additional cases of CML patients in chronic phase who developed new clonal cytogenetic anomalies in Ph- negative cells during imatinib mesylate therapy. Case 1. A 64-year-old man with CML in chronic phase underwent a non myeloablative stem cell transplantation from his sister but relapsed (accelerated phase). Imatinib treatment (600mg/day) was initiated. The daily dose was reduced to 400mg/day due to moderate anemia (Hb: 12,2g/dL; hematocrit : 38,7%) and thrombopenia (155 109/ L). Complete cytogenetic response (CCR) was observed 6 months after the beginning of the treatment and persisted. However, a new abnormal clone consisting of 46,XY,inv(1)(q12q32), del(7)(q22) emerged with 16% (9 months), 33% (12 months) and 44% (15 months) abnormal mitoses. Case 2. A 59-year-old woman with CML in chronic phase received Imatinib mesylate after inefficient autograft marrow transplantation and interferon therapy. CCR was observed 6 months after the treatment. The first cytogenetic evaluation (3 months) already evidenced a new abnormal clone in Ph-negative cells : 46, XX, dup(1)(q21q42). This clone progressively developed with 13% (6 months), 33% (9 months), 44% (12 months) and 50% (15 months) abnormal metaphases. We will discuss the pathological significance of these findings and review the literature data. Address and E-mail address of presenting author : Christian Herens Service de Génétique, Cytogénétique, CHU, Tour de Pathologie, B23, Sart Tilman, 4000 LIEGE, E-mail : christian.herens@chu.ulg.ac.be ABSTRACT n° 16
Leander Huyghe
, An Goethals, Jeroen Hostens, Maureen Van den Hemel and Peter
Tumor necrosis factor (TNF) is a pleiotropic cytokine with a very potent antitumor
activity. The severe shock-inducing effects res ulting from the systemic administration
of the high doses needed to obtain tumor destruction, however, limit its clinical
application to locoregional treatment using isolated limb perfusion. An expansion of
the use of TNF in cancer treatment could be achieved by lowering the dose of TNF
required for tumor destruction. Previous studies have demonstrated that the antitumor
effect of TNF can be explained by the interaction of this cytokine with the TNF-RI on
the endothelium of the tumor neovasculature. We reasoned that we could sensitize the
tumor endothelium to this cytokine by inhibiting the survival signals for these cells.
Since the PI3-kinase plays an important role in cell survival and this cytokine is
activated by many angiogenic growth factors (VEGF, FG F, EGF, PDGF), we decided
to use the selective PI3-kinase inhibitor wortmannin for this study.
Tumor bearing C57BL/6 mice were injected paralesionally with this inhibitor (0.25
mg/kg) daily for 10 days. One hour later these mice were injected with TNF. Control
mice were injected with PBS, wortmannin or TNF only. In a first experiment human
TNF (1.5 mg/kg) was used because this has a tumoristatic effect in our model. In
combination with wortmannin however a complete tumor regression was obtained. In
B6TNF- RI-/- mice the treatment with wortmannin and human TNF had no antitumor
effect, while in wild -type mice bearing a B16Bl6 tumor overexpressing a dominant-
negative TNF-RI this treatment induced a complete tumor destruction. This shows
that the synergistic antitumor activity of wortmannin and human TNF is not due to a
direct sensitization of the tumor cells to the TNF-cytotoxicity, but is a host-mediated
effect. Next we performed a dose-respons experiment with mouse TNF to determine
if the dose required for a complete tumor regression could be lowered. In B16Bl6
melanoma bearing C57BL/6 mice a complete tumor regression is usually obtained
with daily administration of a dose of 0.35 mg/kg mouse TNF. In combination with
the inhibitor wortmannin a complete tumor regression could be obtained with a dose
of mouse TNF as low as 0.04375 mg/kg.
These results show that the use of PI3-kinase inhibitors can potentiate the antitumor
activity of TNF in vivo, and that interfering with survival signals may be a
worthwhile approach to sensitize the tumor neovasculature.
Leander Huyghe, PhD Student
Dept. for Molecular Biomedical Research
Flanders Interuniversity Institute for Biotechnology / Ghent University
Fiers-Schell- Van Montagu Building, Technologiepark 927, B-9052 Ghent, Belgium
Tel: +32-9-33 13 715
Email: leander.huyghe@dmbr.ugent.be
Comijn Joke1, Cindy Vandewalle1, Frans Van Roy2 and Geert Berx1
1 Unit of Molecular and Cellular Oncology,Department of Molecular Biology
2 Molecular Cell Biology Unit, Department of Molecular Biology
SIP1 was isolated as a Smad Interacting Protein and acts as a transcriptional repressor.
It is a multi zinc finger protein that is known to bind CACCT sequences (E2-boxes) in
the Xenopus Xbra2 promoter, the human α4- integrin promoter and the E-cadherin
promoter. The cell adhesion protein E-cadherin is a direct target of SIP1. SIP1
expression result s in the transcriptional downregulation of E-cadherin and this
suppression of E- cadherin functionality was accompanied by loss of aggregation and
acquisition of invasive properties. During embryonic development, SIP1 expression is
essential for the migratory behaviour of cranial neural crest cells, indicating an active
role for SIP1 in processes triggering epithelial- mesenchymal transition. In human
epithelial cells (DLD1Tr21/WTSIP1), SIP1 expression results in a clear
morphological change from an epithelia l phenotype to a more fibroblast-like
Using this conditional SIP1 expressing DLD1Tr21/WTSIP1 cell line, we examined
the mechanism involved in the SIP1-induced epithelial dedifferentiation by cDNA
microarray- based differential gene expression analysis. We found that SIP1-induced
morphological changes are accompanied with downregulation of a set of genes,
known to be functional important for different cell-cell junction complexes in
epithelial cells. Furthermore, we identified E2-boxes in the promoters of these genes
and transfection experiments revealed that SIP1 expression results in the
downregulation of their transcriptional activity. Here, we demonstrate that during the loss of epithelial differentiation upon SIP1 expression besides E-cadherin also other genes coding for crucial proteins of the adherens junctions, as well as tight junction, desmosomes and gap junctions are simultaneously transcriptionally repressed. ABSTRACT n° 18
Jost M
.1*, Lambert V. 1,2*, Maillard C.1*, Van Overstraeten-Schlögel N.3, Gothot A. 3,
Motte P.4, Humblet Ch.5*, Defresne M.P.5*, Rakic J.M. 2, Foidart J.M.1* and Noël A. 1*
1 Laboratory of Tumor and Development Biology, Ulg, Belgium
2 Department of Ophthalmology, Ulg, Belgium 3 Laboratory of Hematology, Ulg, Belgium 4 Unité de Biologie Cellulaire Végétale, Ulg, Belgium 5 Laboratory of Cytology and Histology, Ulg, Belgium * Center of Experimental Cancer Research (CRCE), Ulg, Belgium M.Jost@student.ulg.ac.be Angiogenesis, the formation of new blood vessels by sprouting from pre-existing vessels, is an important feature of various diseases such as cancer and ocular pathologies (age-related macular degeneration). In addition, some angioblasts (precursor endothelial cells) from bone marrow (BM) can be activated and then can migrate to the angiogenic lesion. Angiogenesis involves different proteases including serine proteases - urokinase-type (uPA), tissue-type (tPA) plasminogen activators - and their specific inhibitors (plasminogen activator type 1 and 2 : PAI-1 and PAI-2). To study pathological vascularization, we have developed an in vivo model : the choroidal neovascularization induced by impact laser. The application of this model to PAI-1 deficient mice have demonstrated the essential role of this inhibitor. Indeed, PAI -1 deficiency in host mice prevented choroidal pathological vascularization (Lambert et al., 2001 FASEB J. 15:1021- 1027). Several cell types are producers of PAI -1, such as endothelial cells, inflammatory cells, mast cells, but the essential cellular type is not yet established. The aim of this work was to determine the putative role of bone marrow stem cells (hematopoietic stem cells and/or angioblasts) in the pathological neovascularization in PAI-1 KO mice. Therefore, the model was applicated in PAI-1 deficient mice sublethally irradiated and engrafted with bone marrow (BM) from Wild -Type (WT) mice. A complete restoration of the angiogenic phenotype by WT BM transplantation was observed after laser-induced choroidal neovascularization. This result suggests that vasculogenesis is associated with angiogenesis (from pre-existing vessels) in this model of pathological vascularization. To observe the localization of precursor cells in choroidal lesion, we have engrafted BM from GFP mice to WT mice, and then observe the cells GFP+ in vascularized lesion. GFP expression was often but not always co-located with blood vessels. This suggests the presence of different types of prec ursor cells from BM into the lesion : inflammatory cells, pericytes, endothelial cells. I agree to have the abstract released on the BACR web site before the conference in January 2004 ABSTRACT n° 19
Frédéric KESTELOOT, Charles LAPIERE, Betty NUSGENS and Alain COLIGE
Laboratory of Connective Tissues Biology - GIGA/CRCE - Tour de Pathologie
B23/3, University of Liège, B-4000 Sart Tilman, Belgium.
Angiogenesis is required for development, growth, tissue remodeling, and wound
healing. Pathologies such as diabetic retinopathy, rheumatoid arthritis and cancer
would benefit from therapies controlling and reducing angiogenesis. Enzymes of the
ADAMTS family are closely related to MMPs and ADAMs. They contain however
some specific features, such as a variable number of domains known as
"ThromboSpondin type I repeat" (TSPI). ADAMTS- 1 and -8 are 20-fold more anti-
angiogenic than angiostatin and endostatin, two potent inhibitors of angiogenesis.
The primary function of ADAMTS-2 is the maturation of collagen type I and II
molecules by excising the amino-propeptide. In addition, ADAMTS-2 could also
modulate angiogenesis, as it contains the same sequences than those responsible for
the anti-angiogenic activity of ADAMTS-1 and -8. This hypothesis was investigated
in vitro in different experimental models such as cell proliferation and formation of
capillary structures by human endothelial cells. An ex vivo angiogenesis model was
also used. It consists in mice or rat aorta pieces embedded in a collagen gel in order
to allow the growth of capillaries from the vascular endothelium. As compared to
control mice (TS2+/+), angiogenesis was slightly increased, in absence of ADAMTS-
2, from aortas of ADAMTS-2 KO mice (TS2-/-). Using rat aortas, addition of
recombinant ADAMTS- 2 reduced the formation of capillary structure, also
confirming the anti-angiogenic activity of ADAMTS- 2. Finally, an in vivo model of
angiogenesis was also used. Biocompatible sponges (Ivalon) were implanted under
the skin of TS2+/+ or TS2-/- mice in order to evaluate the formation of capillaries in the
granulation tissue invading the sponge. In absenc e of ADAMTS-2, angiogenesis and
granulation tissue formation were both reduced. Additional investigations are being
performed in order to identify the underlying mechanism(s) inducing these
Presenting author:
Email: Frederic.Kesteloot@ulg.ac.be.
I agree to have the abstract released on the BACR web site before the conference in January 2004.
Vincent Lambert 1,
Ben Wielockx2, Carine Munaut1, Maud Jost1, Patrick Motte8,
Takeshi Itoh3, Zena Werb4, Andrew Baker5, Claude Libert2, Hans-Willi Krell6, Jean-
Michel Foidart 1, Agnès Noël1 and Jean-Marie Rakic 7.
1Laboratory of Tumor and Development Biology, University of Liège, Pathology Tower (B23), Sart -Tilman, B-4000 Liège, Belgium; 2Department of Molecular Biology, Flanders Interuniversity Institute for Biotechnology and University of Ghent, Ghent, Belgium; 3Shionogi Institute for Medical Science, Shionogi & Co., Osaka, Japan; 4Department of Anatomy, University of California at San Francisco, San Francisco, California 94143, USA; 5Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, Marlborough Road, Bristol BS2 8HW, United Kingdom; 6Roche-Diagnostics GmbH, Pharma- Research Penzberg, 82372 Penzberg, Germany,; 7Department of Ophthalmology, University Hospital, Sart -Tilman, B-4000 Liège, Belgium; 8 Unité de Biologie Cellulaire végétale, Dept of Life Sciences, University of Liège. Vincent.lambert@ulg.ac.be Matrix metalloproteinase (MMP) 2 and MMP-9 are increased in human choroidal neovascularization (CNV) occurring during the exudative most aggressive form of age-related macular degeneration (AMD), but their precise role and potential interactions remain unclear. To address the question of MMP-2 and MMP-9 functions, mice deficient in the expression of MMP-2 (MMP-2-KO), MMP-9 (MMP-9-KO), and both MMP-2 and MMP-9 (MMP- 2,9-KO) with their corresponding wild-type mice (WT) underwent CNV induction by laser-induced rupture of the Bruch's membrane. Both the incidence and the severity of CNV were strongly attenuated in double deficient compared to single gene deficient mice or corresponding WT controls. The reduced neovascularization was accompanied by fibrinogen/fibrin accumulation. Furthermore, overexpression of the endogenous MMP inhibitors TIMP-1 or TIMP-2 (delivered by adenoviral vectors) in WT mice or daily injection of a synthetic and gelatinase selective MMP inhibitor (Ro 26-2853) significantly decreased the pathological reaction. These findings suggest that MMP-2 and MMP-9 may cooperate in the development of AMD and that their selective inhibition represents an alternative strategy for the treatment of choroidal neovascularization. ABSTRACT n° 21
Philippe Martinive
, Pierre Sonveaux, Chantal Dessy, Vincent Grégoire and Olivier
UCL Medical School, Unit of Pharmacology and Therapeutics (FATH 5349), 53 Ave
E. Mounier, B-1200 Brussels, Belgium (philippe.martinive@rbnt.ucl.ac.be)
Although derived from the host tissue, the tumor vasculature is under the influenc e of the tumor microenvironment and needs to adapt to the resistance to blood flow inherent to the dynamics of tumor growth. Such vascular remodeling can offer selective targets to pharmacologically modulate tumor perfusion and thereby improve the efficacy of conventional anti-cancer treatments. Radiotherapy and chemotherapy can, indeed, take advantage of a better tumor oxygenation and drug delivery, respectively, both partly dependent on the tumor blood supply. Here, we showed that isolated tumor arterioles mounted in a pressure myograph have the ability, contrary to size-matched healthy arterioles, to contract in response to a transluminal pressure increase. This myogenic tone was exquisitely dependent on the endothelin-1 pathway since it was completely abolished by the selective ET A antagonist BQ123. This selectivity was further supported by the large increase in endothelin-1 abundance in tumors (5 to 15-fold according to tumor models) and the higher density of the ET A receptors in tumor vessels. We also documented by using laser doppler microprobes and imaging that administration of the ET A antagonist led to a significant increase in tumor blood flow whereas the perfusion in control healthy tissue was not altered. Finally, we provided evidence that acute administration of the BQ123 could significantly stimulate tumor oxygenation, as determined by EPR oximetry, and increase the efficacy of low-dose, clinically-relevant fractionated radiotherapy. The dose-dependency of the ETA antagonist effects and the consistency of these findings in various mouse tumor models further emphasized the relevance of our data. Thus, blocking the tumor-selective increase in the vascular endothelin -1/ET A pathway unravels an important reserve of vasorelaxation which can be exploited to selectively increase tumor response to radiotherapy. *I agree to have the abstract released on the BACR web site before the conference in Jan2004 ABSTRACT n° 22
NDLOVU 'Matladi1
1Lab. Of Eukaryotic Gene Expression and Signal Transduction, Department of Molecular Biology, University of Gent, K.L. Ledeganckstraat 35, Gent 9000, Belgium. Email: Matladi.Ndlovu@.ugent.be 2Laboratory of Molecular Virology, Institute of Biology and Molecular Medicine, University of Brussels, Charleroi, Gosselies. Heterochromatin and euchromatin typically have a different chromatin structure and accessibility, especially in the promoter regions. Generally, domains of chromatin that are associated with transcriptionally active genes (euchromatin) are less compact relative to silenced surrounding regions (heterochromatin). Furthermore, cis-acting elements that are functional in a cell-dependent manner often form nuclease-hypersensitive sites (HS) in chromatin (GROSS AND GARRARD, 1988). These sites can either be constitutively formed before the activation of a gene, or else be created upon inducible binding of transcription factors. Deoxyribonuclease-I (DNase-I), micrococcal nuclease and restriction enzyme accessibility assays have been used extensively to detect these hypersensitive sites and, consequently, remodelled chromatin ( VAN LINT et al., 1996; VERDIN et al., 1993). We have studied the chromatin organization of the Interleukin -6 (IL-6) gene promoter in two breast cancer cell lines that differentially express this cytokine, namely MDA-MB-231 cells, which produce high levels of IL- 6 in response to TNF-α, and MCF-7 cells, which produce very low levels of IL- 6 in response to the same stimulus. The promoter region of the IL-6 in the two cell lines was analysed for the presence of hypersensitive sites with DNase-I and for nucleosome positions with micrococcal nuclease (MNase). Upon comparison of the MDA-MB-231 and MCF-7 breast cancer cell lines, reduced IL-6 expression levels did correlate with reduced number of hypersensitive sites. These results suggest that the nucleosomal structure of the IL-6 promoter is a major determining factor in establishing a transcription-competent enhanceosome for optimal gene expression (SMALE AND FISHER, 2002; ARMENANTE et al., 1999). In this respect, we are currently investigating the role of the NF-κB and H3 kinase MSK1, in selective nucleosome remodelling at the IL- 6 promoter in breast cancer cells (Vermeulen et al. 2003). REFERENCES 1. ARMENANTE, F., MEROLA, M., FURIA, A., TOVEY, M. AND PALMIERI, M. (1999) Nucleic Acids Res. 27: 4483-4490.
2. GROSS, D.S AND GARRARD W.T. (1988) Annu. Rev. Biochem. 57: 157-197.
3. SMALE, S.T. AND FISHER, A.G. (2002) Annu. Rev. Immunol. 20: 427-462.
4. VAN LINT, C., EMILIANI, S., OTT, M. AND VERDIN, E. (1996) EMBO J. 15: 1112-
5. VERDIN, E., PARAS, P. JR AND VAN LINT, C. (1993) EMBO J. 12: 3249-3259.
HAEGEMAN G. (2003) EMBO J. 22: 1313-1324. ABSTRACT n° 23

Nguyen, N- Q-N
, Rentier-Delrue, F., Martial, J.A. and Struman, I.
Laboratory of Biology Molecular and Genetic Engineering, B6, University of Liège,
Sart-Tilman 4000 Liège, Belgium.
We previously demonstrated that the 16 kDa N-terminal fragment of the human
prolactin (16K hPRL) is a potent antiangiogenic factor that inhibits the endothelial
cell function in vitro (1) and the neovascularization in vivo (2). Moreover, 16K hPRL
was shown to inhibit tumor growth by decreasing angiogenesis in a xenograph animal
model (3).
In this project, we search for minimal peptides that retain the antiangiogenic and
antitumoral properties of the 16K hPRL. Based on the 3D model of the full-length
hPRL (4) we designed three peptides (called 16K peptides). We first tried to produce
these peptides by transfecting 293 eukaryotic cells with the recombinant vectors
encoding the 16K peptides and showed, by RT-PCR, that mRNA was synthesized.
However, no peptide production was detected in the conditioned medium of the
transfected cells.
We therefore decided to turn to a gene therapy strategy using adenoviruses. We first
constructed four Ad5 adenovirus vectors expressing the wild-type 16K hPRL and the
16K peptides. We then engineered recombinant adenoviruses and produced high
titered and CsCl purified preparations of these viruses. We are currently testing the
ability of the peptides to inhibit endothelial cell proliferation. The more efficient
peptide will then be assessed for its ability to inhibit growth of tumor implanted in
immunodeficient mice.

1. Clapp C, Martial JA, Guzman RC, Rentier-Delrue F, and Weiner R 1993 The 16-

kilodalton N-terminal fragment of human prolactin is a potent inhibitor of angiogenesis. Endocrinology 133:1292- 9. 2. Struman, I., Bentzien, F., Lee, H., Mainfroid, V., D'Angelo, G., Goffin, V., Weiner, R. I. and Martial, J. A. 1999. Opposing actions of intact and N- terminal fragments of the human prolactin/growth hormone family members on angiogenesis: an efficient mechanism for the regulation of angiogenesis.Proc Natl Acad Sci U S A 96(4): 1246-51. 3. Bentzien F, Struman I, Martini J-F, Martial JA, and Weiner R 2001 Expression of the antiangiogenic factor 16K hPRL in human HCT116 colon cancer cells inhibits tumor growth in Rag1-/- mice. Cancer Research 61:1-7. 4. Goffin V, Martial JA, Summers NL 1995 Use of a model to understand prolactin and growth hormone specifities. Protein Eng 8(12):1215-31. I agree to have the abstract released on the BACR web site before the conference in January 2004 ABSTRACT n° 24
Marc Pauly
, Manon Bosseler, Brigitte Schroell, Valérie Palissot, Carlo Faber,
Jacques Kayser, Mario Dicato, and Guy Berchem.
Blood and Cancer Foundation, Centre Hospitalier de Luxembourg, and Clinique Ste
Thérèse, Luxembourg, Luxembourg.
Colorectal cancer is a major cancer death cause in many European countries. A
crucial step towards the advanced stages of this disease is the occurrence of a point
mutation mostly in codon 12 and 13 of the K-ras 1proto-oncogene which confers a
selective growth advantage to the target cell. The detection of these mutations is
therefore of great importance forthe early diagnosis, prognosis and planning of a
therapy. In our study, weperformed the rapid cycle real-time fluorescent polymerase
chain reactionmethod based on the principle of increased resonance energy transfer
byfluorogenic probe hybrization while screening for such mutations in genomic DNA
extracted from minimal surgical resected colorectal tumour samples of 30
luxembourgish patients with advanced colorectal cancer. The use of a peptid -nucleic
acid wild type probe greatly helped to increase the sensitivity of this method by
efficiently eliminating the wild type genomic DNA contamination back- ground in the
samples. We showed that this method represents a useful, sensitive and rapid tool for
the genetic analysis of clinical tumour samples.
Dr. Sc. Marc Pauly (Ph. D.)
Laboratoire de Recher che
sur le Cancer et les Maladies du Sang (RCMS)
(Laboratory of Research on Cancer and Blood Diseases)
Bâtiment des Sciences (Science Building)
Centre Universitaire de Luxembourg
162-A Avenue de la Faïencerie
L-1511 Luxembourg
Grand-Duché de Luxembourg
Tél: (00 352) 466644 432
Bea Pauwels
, Annelies E.C. Korst, Veronique Andriessen, Marc F.D. Baay, Hilde
A.J. Lambrechts, Greet G.O. Pattyn, Christel M.J. De Pooter, Filip Lardon & Jan B.
Gemcitabine (dFdC) is an active antitumour agent with radiosensitising properties.
The mechanism of this enhancement effect is not fully elucidated yet. Both radiation
and gemcitabine interfere with the cell cycle checkpoint mechanism. The aim of this
study is to examine the role of p53 protein in the radisensitising effect of gemcitabine.
The cell lines used in this study were A549 (wt p53), a human lung adenocarcinoma
cell line; A549 E6 and A549 LXSN: A549 cells transduced with the retrovirus vector
LXSN containing the neo resistance gene with or without HPV type 16 E6 gene.
A549 E6 cells have a functionally inactive p53 gene product. A549 LXSN cells
maintain their p53-dependent responses. These cells were created in the laboratory of
Denise Galloway (Fred Hutchinson Cancer Research Centre, Seatle). The cells were
treated with moderate cytotoxic concentrations gemcitabine (4- 15 nM) during 24 h,
prior to radiation (RT) (γ- Co60, 0-8 Gy, room temperature). Cell survival was
determined 7 days after RT by the sulforhodamine B test. In addition, cell cycle
perturbation was determined by flow cytometry and p53 expression was determined
by Western blot analysis. Experiments were performed at least three times. Radiation
parameters were calculated from the survival curves, fitted according to the linear-
quadratic model: survival=exp( -αD-βD2). The radiosensitising effect was represented
by the dose enhancement factor (DEF): ID50control/ID50pretreated, with the ID50 the dose
radiation resulting in 50% growth inhibition. The concentration gemcitabine causing a
specific percentage of cell kill is presented by the IC value. A two-sample t-test and
one-way ANOVA were used to determine significance.
A549 LXSN and A549 E6 were less sensitive for the cytotoxic effect of dFdC than
the parent A549 cells: IC50 values were 9.01 ± 0.89, 15.64 ± 1.39 and 17.86 ± 3.17
nM for A549, A549 LXSN and A549 E6, respectively. However, dFdC (at
concentrations around IC20-IC45) caused a radiosensitising effect in all cell lines:
DEFs ranged from 1.90-2.58, 1.34-2.06 and 1.53- 2.27 for A549, A549 LXSN and
A549 E6 cells and were not significantly different at equitoxic concentrations. Cell
cycle analysis after treatment during 24 h with different concentrations dFdC (2-20
nM) showed no difference between the three cell lines. P53 expression was dose- and
time-dependent after dFdC or RT. The expression of p53 increased after the combined
treatment with dFdC and RT. No p53 expression was observed in A549 E6 cells.
In conclusion, the A549 sublines were less sensitive to gemcitabine, probably due to
transfection. However, since both the radiosensitising effect at equitoxic
concentrations and the cell cycle effect of dFdC were independent of p53 expression,
it seems that p53 protein does not play a crucial role in the radiosensitising
mechanism of gemcitabine.
R. Salgado
1,2, I. Benoy1,2, G. Van Den Eynden1,2 , I. Van der Auwera1,2 , S. Van Laere1,2, P. Van Dam1,
C. Colpaert 1, P. Vermeulen1,2 , L. Dirix 2, E. Van Marck1.
Translational Cancer Research Group Antwerp (1 Lab Pathology University of Antwerp/ University
Hospital Antwerp, Edegem; 2Oncology Centre, General Hospital Sint-Augustinus, Wilrijk) - e- mail:
Approximately 50% of breast cancer patients have no metastatic involvement of axillary lymph nodes
at diagnosis, and may be considered cured after primary locoregional treatment. Yet, approximately
20% will experience distant relapse within 5 years. This suggests outgrowth of disseminated tumour
cells present at diagnosis and undetectable with current routine techniques.
Currently, approximately 2500 patients with breast cancer have been analysed in 5 prospectively
designed clinical trials in order to verify associations between the presence of immuno-stained tumor
cells in bone marrow and prognosis. In a large study performed by Braun and colleagues, 552 patients
with stage I, II and III breast cancer were included in a prospective trial to verify whether
immunohistochemically detected tumour cells in bone marrow predicted prognosis. Tumour cells
stained with the antibody A45- B/B3 independently predicted distant metastasis but not locoregional
recurrence in their patient series.
The quantitative importance of angiogenesis is, considering the methodological, study design and-
population heterogeneities, still subject to debate as to the strength of it as a predictive and/or
prognostic factor in human cancer. A less observer-dependent way of quantifying angiogenesis is a
Chalkley point overlap morphometry estimation of relative microvessel area. Significant associations
between Chalkley counts and axillary lymph node metastasis, large tumour size, high histological
malignancy grade and histological type have been reported in breast cancer patients. Moreover, an
independent prognostic value has been demonstrated in patients with primary, unilateral invasive breast
Considering the independent prognostic value of Chalkley estimation of relative vascular area in
patients with primary unilatera l invasive breast cancer, present study aimed at assessing the correlation
between Chalkley estimation of surface area and the presence of micrometastases in a series of 68
patients with localised untreated breast cancer. When the maximum Chalkley count (Chalkley max)
and CK-19 RGE-values (RGE: Relative Gene Expression) were categorised according to their
respective cut -off values 7 (median) and 0.77 (95% percentile of control patients), those patients who
had a CK-19 RGE >0.77 had a mean Chalkley count of 7.5 ± 1.7 (median: 7; S.E.: 0.30) compared to
6.5 ± 1.7 in patients with Chalkley counts <7 (median: 6; S.E.: 0.3) (Mann -Whitney U-test; p= 0.04).
Chalkley max, considered as a continuous variable, has an independent predictive value for CK-19
RGE when dichotomised according to the cut -off value of 0.77 (p= 0.04; odds ratio: 1.38; 95% C.I.:
Our data are in support of a quantitative relation between vessel area and bone marrow
micrometastases in breast cancer patients with operable disease and provide further evidence that both
high relative microvessel area and tumour cell lodging in bone marrow are rate-limiting steps in
determining prognosis in patients with primary breast cancer.
I agree.
Cindy Simoens1
, Annelies E.C. Korst1, Christel M.J. De Pooter2, Greet G.O. Pattyn1,
Hilde A.J. Lambrechts1, Fabienne Breillout3, Filip Lardon1 & Jan B. Vermorken1
1Dept. Of Medical Oncology, University of Antwerp, Wilrijk, Belgium
2St. Augustinus hospital, Antwerp, Belgium
3Institut de Recherche Pierre Fabre, Boulogne, France Tel. :+32 3 820 25 76, Fax. : +32 3 820 22 48, E-mail : Cindy.Simoens@ua.ac.be Vinflunine is a novel Vinca alkaloid, synthesized from vinorelbine using superacidic chemistry. It has shown markedly superior antitumour activity to vinorelbine in various experimental tumour models. Since vinflunine treatment results in an accumulation of cells in G2/M phase, the most radiosensitive phase of the cell cycle, this new chemotherapeutic agent might have radiosensitising properties. Therefore, the aim of the present study was to investigate the interaction between vinflunine and radiation, in vitro. Four human tumour cell lines were used: ECV304, a bladder cancer cell line; CAL-27, a head and neck cancer cell line; MCF-7, a breast cancer cell line and H292, a lung cancer cell line. To investigate the interaction between vinflunine and radiation, cells were incubated with three different concentrations of vinflunine during 24h immediately before radiation (Co -60γ rays, 0-8 Gy, room temperature). Cell survival was determined by the sulforhodamine B assay 7 or 8 days after radiation. The dose-survival curves were fitted according to the linear quadratic model to calculate the ID50 (radiation dose causing 50% growth inhibition). A possible radiosensitising effect was represented by the Dose Enhancement Factor (DEF = ID50 control / ID50 vinflunine treated). The combination index (CI) by the Chou and Talalay equation was calculated to define synergism (synergism : CI<0.7, moderate synergism : 0.7<CI<0.9, nearly additive effect : 0.9<CI<1.1). The concentration vinflunine causing a specific percentage of cell kill is represented by the IC-value. All experiments were performed at least three times. In ECV304 cells, treatment with vinflunine resulted in a clear increase in radiosensitivity. The DEF-values were 1.57, 1.53 and 1.93 for 30nM (=IC10), 40nM (=IC25) and 50nM vinflunine (=IC30), respectively. Moderate synergism was observed for all tested concentrations, with mean CI-values of 0.84, 0.81 and 0.71, respectively. In CAL-27 cells, DEF-values of 1.41, 1.43 and 2.29 were observed for 25nM (=IC40), 27.5nM (=IC45) and 30nM vinflunine (=IC70), respectively. For 30nM vinflunine, a clear synergistic effect was observed (CI-value = 0.57), whereas the CI-values of 25nM and 27.5nM mainly showed additivity. In MCF-7 cells, a significant increase in radiosensitivity was observed after treatment with 35nM (=IC50) and 40nM vinflu nine (=IC60), DEF-values were 1.57 and 2.24, respectively. Mean CI -values were 0.69 and 0.54, showing a synergistic effect. Although 30nM vinflunine (=IC40) did not result in a significant decrease in ID50 (DEF = 1.42), the combination with radiation was moderately synergistic according to the CI analysis (CI -value = 0.77). In H292 cells, concentrations of 30nM (=IC10) and 35nM vinflunine (=IC25) resulted in a DEF of 1.29 and 1.22, respectively; CI analysis showed only an additive effect. Only with a concentration of 40nM vinflunine (=IC40, DEF = 1.53), a CI-value of 0.77 was observed, a moderate synergistic effect. Vinflunine caused radiosensitisation in all four cell lines tested, most pronounced in ECV304 cells, with radiosensitising effects with non-toxic concentrations of vinflunine (around IC10). For MCF-7 and H292, concentrations around IC40 were needed to cause radiosensitisation and in CAL-27 cells, rather toxic concentrations, around IC70, were needed to result in a radiosensitising effect. These positive results warrant further study. ABSTRACT n° 28
Pierre Sonveaux
, Agnès Brouet, Philippe Martinive and Olivier Feron
UCL Medical School, Unit of Pharmacology and Therapeutics (FATH 5349), 53 Ave
E. Mounier, B-1200 Brussels, Belgium (pierre.sonveaux@mint.ucl.ac.be)
The vascular network is a highly accessible target for tumor therapy. However, as for
any cancer treatment regimen, the primary goal is to deliver sufficient amounts of
drug to the targeted tissue while minimizing damage to healthy organs. Cationic
liposomes have been identified as delivery systems significantly more effective at
targeting tumor versus normal vascular networks. Though, the liposome uptake by the
liver restricts their potential as shuttles to selectively target the tumor endothelium.
Here, we reasoned that additional selectivity could be found in the nature of the
delivered gene and by combining another anti-tumor therapy. Accordingly, the pro-
survival PI3K /Akt pathway is thought to be activated by ionizing radiations and a
dominant negative Akt would therefore mostly target tumors (versus non-irradiated
organs). Furthermore, we have recently documented that irradiation led to NO-
mediated tumor vessel dilation which could thereby enhance the access of liposomes
to the tumor. In this study, we have therefore examined whether radiotherapy and the
use of dominant-negative Akt delivered as a plasmid by cationic liposomes could
mutually improve their efficacies.
We first used cultured endothelial cells and isolated tumor microvessels to
demonstrate that low dose irradiation led to the stimulation of both Ser473 Akt and
Ser1177 eNOS phosphorylations (e.g., activation); the use of a PI3K inhibitor further
indicated that the former largely accounted for the increased NO/cGMP production.
We then examined whether these effects of irradiation could impact in vivo on the
gene delivery into the tumor. Using a reporter-encoding plasmid, we found that
irradiation dramatically enhanced the expression of the tagged protein in the tumor
endothelium. Also using L-NAME and eNOS deficient mice, we documented the key
role of NO in mediating the adjuvant effects of X-Ray on plasmid delivery, likely
through an increase in tumor blood flow. We then combined local irradiation to the
administration of the liposome-dominant negative Akt DNA complex. In two
different protocols associating gene therapy with either a single 6 Gy dose or a 5 x 2
Gy fractionated scheme, we consistently observed synergistic effects of the
combinatory treatments. In fact, the dominant-negative Akt transgene when
administered alone, did not reveal any tumor response and the tumor growth delay
after irradiation represented less than 50% of the gain obtained when combining both
approaches (n=8; P<0.01); these findings were obtained in two different mouse tumor
In conclusion, the combination of low dose radiotherapy and liposome-cargoed
dominant-negative Akt gene therapy appears particularly well suited to selectively
target tumor vasculature. Besides the intrinsic tumor specificity of local X- Ray
administration and the propensity of cationic liposomes to bind tumor vessels, we
have further identified a "treatment-driven" selectivity, e.g. the ability of radiotherapy
to induce Akt activation in tumor vasculature and to increase the liposome access to
the tumor.
*I agree to have the abstract released on the BACR web site before the conference in Jan2004
Laboratory of Tumor and Development Biology, C.R.C.E., University of Liège, Sart
Tilman, B-4000 Liège, Belgium
* Departement of Oncology, Cambridge Institute for
medical Reasearch, Univerity of Cambridge, United Kingdom
MT1-MMP and VEGF are two key molecules involved in pericellular proteolysis and
cell proliferation, two processes involved during tumor growth and angiogenesis. Our
previous data showed that MT1-MMP overexpression in human breast carcinoma
MCF7 cells induces an up-regulation of VEGF expression. This effect is associated in
with accelerated tumor growth and angiogenesis (Sounni et al. FASEB J. 16,
555-564 - 2002). We provide now evidence that MT1- MMP overexpression
specifically affects VEGF- A production and failed to influence t hat of other members
of the VEGF family (VEGF, B, C or PLGF). In this study, we also investigated the
molecular mechanism by which MT1-MMP regulates VEGF expression and
production. The transcriptional regulation of VEGF by MT1-MMP was completely
blocked by a synthetic MMP inhibitor. Deletion of MT1- MMP at catalytic or
cytoplasmic domains was associated with a decreased VEGF production. We also
investigated the signaling pathway that mediates the effect of MT1-MMP on VEGF
expression. The VEGF mRNA up-regulation is significantly suppressed by genistein,
herbimycin A and 4- amino-5-(-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d] pyrimidine
(PP2), but not (MAPK) kinase (MEK)1/2 inhibitor (PD98059), phosphatidylinositol
3-kinase inhibitor (Wortmannin) or p38 MAPK inhibitor (SB203580). In conclusion
our results suggest that the full enzymatic activity of MT1-MMP is required for up-
regulation of VEGF-A through an activation of Src-type tyrosin kinase pathways,
leading to angiogenic response in breast cancer cells.
Sébastien P. Tabruyn
, Catherine Sorlet, Angélique Georges, Françoise Rentier-
Delrue, Joseph A. Martial and Ingrid Struman.
Laboratoire de Biologie Moléculaire et de Génie Génétique, Université de Liège, B6,
Sart Tilman, B-4000 Liège, Belgium.

We previously demonstrated that the 16 kDa N-terminal fragment of the human
prolactin (16K hPRL) has antiangiogenic properties 1,2 and could prevent tumor
growth in xenograph mouse model 3,4. To better understand the mechanisms by which
16K hPRL inhibits angiogenesis, we first focus on the pathway leading to the
endothelial cell death. We have demonstrated that 16K hPRL requires the activation
of the transcription factor NF-κB to induce apoptosis 5. Indeed, this induction leads,
by the activation of two upstream caspases, caspase-8 and caspase-9, to the cell
apoptosis via the activation of caspase-3. We have also showed that 16K hPRL
inhibits basic fibroblast growth factor (bFGF)- and vascular endothelial growth factor
(VEGF)-stimulated proliferation of human and bovine endothelial cells 6. Based on
this initial observation, we have first confirmed that 16K hPRL induces cell cycle
arrest in adult bovine aortic endothelial (ABAE) cells by 3H-thymidine or BrdUrd
incorporation analysis. We have showed that 16K hPRL is able to inhibit both basal
and bFGF-induced proliferation. Using labeling technique with propidium iodide, we
have demonstrated, by flow cytometry analysis, that 16K hPRL induces a cell cycle
arrest at the G1/S transition in a dose dependant manner. To better understand the
mechanisms by which 16K hPRL triggers this process, we then investigated the
molecular events implicated in the cell cycle arrest. By transfecting the ABAE cells
with a reporter plasmid that allows luciferase expression under the control of E2F, we
have shown that 16K hPRL causes inactivation of the transcription factor E2F, a key
factor essential for the progression in S phase. All these results reveal novel important
mechanisms involved in the antiangiogenic action of 16K hPRL, clearly further
studies remain necessary to better understand the mechanism of action of this
promising anticancer drug.

Clapp C, Martial JA, Guzman RC, Rentier-Delrue F, Weiner RI. The 16-kilodalton N-terminal fragment of human prolactin is a potent inhibitor of angiogenesis. Endocrinology. 1993;133:1292-9. Struman I, Bentzien F, Lee H, Mainfroid V, D'Angelo G, Goffin V, Weiner RI, Martial JA. Opposing actions of intact and N-terminal fragments of the human prolactin/growth hormone family members on angiogenesis: an efficient mechanism for the regulation of angiogenesis. Proc Natl Acad Sci U S A. 1999;96:1246-51. Bentzien F, Struman I, Martini JF, Martial J, Weiner R. Expression of the antiangiogenic factor 16K hPRL in human HCT116 colon cancer cells inhibits tumor growth in Rag1( -/-) mice. Cancer Res. 2001;61:7356-62. Kim J, Luo W, Chen DT, Earley K, Tunstead J, Yu -Lee LY, Lin SH. Antitumor activity of the 16-kDa prolactin fragment in prostate cancer. Cancer Res . 2003;63:386-93. Tabruyn SP, Sorlet CM, Rentier- Delrue F, Bours V, Weiner RI, Martial JA, Struman I. The antiangiogenic factor 16K human prolactin induces caspase-dependent apoptosis by a mechanism that requires activation of nuclear factor- kappaB. Mol Endocrinol. 2003;17:1815-23. Ferrara N, Clapp C, Weiner R. The 16K fragment of prolactin specifically inhibits basal or fibroblast growth factor stimulated growth of capillary endothelial cells. Endocrinology. 1991;129:896-900. ABSTRACT n° 31
Zhen Zhang2, Xiao-Ying Meng1, Pete Tornatore1, Robert Bast3, Yinhua Yu3, Daniel
Chan2, Scot Weinberger1, Eric T. Fung1
1Ciphergen Biosystems, Fremont, CA; 2Johns Hopkins Medical Institutions,
Baltimore, MD; 3MD Anderson Cancer Center, Houston, TX
Presenting scientist : Davy T'Jampens.
Ovarian cancer is diagnosed in 23000 women each year in the United States. 14000
women die from this disease each year, and the high mortality associated with ovarian
cancer can be attributed to the late stage at which most cases are detected. In contrast,
survival for early stage ovarian cancer is greater than 90%. Consequently, biomarkers
that aid in the detection of early stage ovarian cancer are needed. Protein expression
profiling offers the opportunity to identify novel biomarkers and to develop
combinations of biomarkers that can be used for the diagnosis of early stage ovarian
cancer. Increased sensitivity and specificity can be obtained using these combinations
of biomarkers, and their identification can reveal pathways and mechanisms of
disease that may be exploited for additional diagnostic, therapeutic, and prognostic
Methods : Protein expression profiling was performed using SELDI protein biochips
and a Ciphergen PBSII, laser desorption/ionization, time- of-flight mass spectrometer.
Serum from 502 samples (ovarian cancer, healthy controls, and benign disease)
obtained from four institutions were fractionated on a filter plate containing Q Hyper
DF anion exchange resin. Individual fractions were profiled on a variety of SELDI
arrays. A support vector machine algorithm (UMSA, 3Z Informatics) was performed
to determine optimal biomarker candidates among peaks with signal/noise ratios>5.
Data were analyzed independently for samples obtained from two institutions and the
candidate biomarkers were validated on samples obtained from the other two
institutions. A multi-marker panel comprising these three biomarkers with CA125
was created based on a subset of samples and independently validated on the
remaining samples. Biomarkers were purified using spin column micro-
chromatography and monitored on SELDI biochip arrays. Biomarkers were identified
by tryptic peptide ms analysis usin g a PBSII reader and ms/ms analysis using an
AB/Sciex QStar.reg. equipped with a Ciphergen ProteinChip.reg. Interface.
Preliminary Data : Three biomarkers were found to be significant individually as well
as synergistically when applied in a multi- marker panel. Two of these markers could
be seen on the ProteinChip Biomarker System only, while the other could be seen on
both the ProteinChip Biomarker System as well as the ProteinChip Interface QStar-
MS/MS. Indvidually, each biomarker had p<.00001 when comparing healthy controls
against all ovarian cancer, while two of the biomarkers each had p<.00001 when
comparing healthy controls against early stage ovarian cancer. These scores were
comparable to that of CA125. A multi-marker panel comprising these three
biomarkers combined with CA125 was created based on its performance on a training
set. When tested on an indpendent validation set, this model obtained a specificity of
93% as compared to 52% for CA125 alone for early stage ovarian cancer, at a fixed
sensitivity of 83%. The three biomarkers have been purified and identified. Two are
cleavage products of mature proteins. Immunoassays were developed so that these
analytes could be measured simultaneously using the SELDI biochip approach.
Ilse Van der Auwera
, Steven Van Laere, Gert Van den Eynden, Peter Van Dam, Eric
Van Marck, Cecile Colpaert, Peter Vermeulen, Luc Dirix
Translational Cancer Research Group Antwerp (Lab Pathology University of
Antwerp/ University Hospital Antwerp, Edegem; Oncology Centre, General Hospital
Sint- Augustinus, Wilrijk) - e-mail: ilse-vda@pandora.be
Inflammatory breast cancer (IBC) is a distinct and aggressive form of locally
advanced breast cancer with unique clinical and pathological features. Prognosis of
IBC patients remains poor due to the rapid onset of disease and its high metastatic
potential. Recently, Colpaert et al. (2003) found histological evidence of intense
angiogenesis in IBC specimens.
The aim of this study was to confirm the angiogenic phenotype of IBC and to
investigate its potential to induce lymphangiogenesis. Real-time RT -PCR was used to
measure mRNA levels of a group of tumour angiogenesis and lymphangiogenesis-
related factors (VEGF-A, VEGF- C, VEGF- D, Flt - 1, KDR, Flt-4, Ang- 1, Ang-2, Tie-
1, Tie-2, COX-2, bFGF, Egr-1, Prox-1 and LYVE-1) in specimens of 17 IBC and 20
non-IBC patients.
Statistical analysis showed that IBC tumour specimens had higher expression levels
of the following mRNAs than non-IBC specimens: Ang-1 (P=0.0004), Tie-1
(P=0.001), Tie -2 (P=0.003), bFGF (P=0.007), VEGF-C (P=0.003), VEGF- D
(P=0.035), Flt -4 (P=0.001), Prox-1 (P=0.004) and LYVE-1 (P=0.012). VEGF mRNA
expression was positively correlated with the fractions of tumour cells present in the
IBC and non-IBC samples used for analysis (rs =0.379, P=0.025). In contrast, bFGF
mRNA (rs = - 0.483, P = 0.003), Ang-1 mRNA (rs = -0.527, P = 0.001), Tie-2 mRNA
(rs = -0.394, P = 0.019), VEGF-D mRNA (rs = -0.480, P = 0.003), Flt -4 mRNA (rs = -
0.450, P = 0.007) and LYVE-1 mRNA (rs = -0.405, P = 0.016) expression was
negatively correlated with the fractions of tumour cells present in the samples. After
correction for the fraction of tumour-associated stroma present in the IBC and non-
IBC samples used for analysis, expression levels of LYVE- 1 mRNA (P=0.035) and
VEGF- D mRNA (P=0.008) remained significantly increased in IBC specimens, while
Ang- 1 mRNA (P=0.074), Tie-1 mRNA (P=0.01) and Tie-2 mRNA (P=0.066) were
almost significantly increased in IBC specimens.
Using real-time RT -PCR we confirmed the increased angiogenic activity in IBC. We
also demonstrated the presence of active lymfangiogenesis in IBC. This explains the
high metastatic potential of IBC by lymphatic and hematogenous route. Both
pathways are potential targets for the treatment of IBC.

SJ Van Laere
, I Van der Auwera, GG Van den Eynden, HJ Elst, I Benoy, SB Fox,
AL Harris, CG Colpaert, PB Vermeulen, P Van Dam, EA Van Marck, LY Dirix
Translational Cancer Research Group (GH Sint-Augustinus, Wilrijk; University
Hospital Antwerp, Edegem, Belgium), Weatherall Institute of Molecular Medicine
and Nuffield Dept. of Clinical Laboratory Sciences, John Radcliffe Hospital,
Headington Oxford, UK.
e-mail: Steven.Van.Laere@GVAGroup.be
Inflammatory Breast Cancer ( IBC ) is a specific type of breast cancer with unique
clinical and pathological features. Approximatly 6% of all breast cancer patients
suffer from IBC. IBC is very aggressive and highly metastatic. Despite a high
lethality and the fact that IBC is clinically well characterised, little is known about the
parameters that determine the rapid progression of the disease. Several studies
indicate an increase in angiogenic activity when comparing IBC with non-IBC. The
aim of this study was to investigate whether a difference between IBC and non- IBC
was present based on gene expression profiling.
RNA extracted from 17 IBC and 22 non-IBC patient was used. The RNA
concentration and the purity (A260/A280) was measured spectrophotometrically. The
quality (28S/18S) was assessed using the Agilent Bioanalyzer 2100. RNA with high
concentration (>100ug/mL) , high purity (A260/A280 > 1.8) and high quality
(28S/18S > 2.0) was used for microarray experiments. These experiments were
conducted at the Nuffield Department of Clinical Laboratory Sciences, John Radcliffe
Hospital, Headington Oxford. cDNA slides were obtained from the Sanger Institute
(Hver1.2.1). This slide contains 10752 cDNA fragments derived from various
collections. Before hybridisation, RNA was amplified and labelled using the Ambion
Amino Allyl MessageAmp aRNA Amplification Kit. RNA from each patient was
hybridised individually against a reference (Universal Human Reference RNA,
Stratagene). Patient RNA was labelled with Cy5, reference RNA was labelled with
Cy3. Hybridisation was executed following the Sanger Institute protocol. The slides
were scanned and analysed using three different software programs : ScanArray,
Quantarray and GeneSpring. An unsupervised hierarchical clustering was executed
using 10 non-IBC patients and 8 IBC patients and resulted in two groups.
Unsupervised hierarchical clustering (class discovery) was able to exactly classify all
IBC patients. This clearly indicates that both types of cancer are different when
compared by their gene expression profiles. The IBC cluster can be further divided in
two subclusters. A recent study by Lerebours et al. has shown that two prognostic
groups of inflammatory breast cancer have distinct genotypes. Supervised class
comparison will be done and might reveal some parameters that are important for the
highly metastatic behaviour of IBC.

Source: http://www.bacr.be/Abstracts2004.pdf

Microsoft word - abstract toc-titles authors only to ccfa 11-4-08

Table of Contents O-0001. SONIC: A randomized, double-blind, controlled trial comparing Infliximab and Infliximab plus Azathioprine to Azathioprine in patients with Crohn’s Disease naive to immunomodulators and biologic therapy Sandborn W 1, Rutgeerts P2, Reinisch W3, Mantzaris G4, Kornbluth A5, Rachmilewitz D6, Lichtiger S5, D'Haens G7, van der Woude C8, Diamond R9, Broussard D9, Heged

Metabolic and cardiovascular effects of very-low-calorie diet therapy in obese patients with type 2 diabetes in secondary failure: outcomes after 1year

VLCD therapy in obese patients with diabetes in secondary failure P. Dhindsa et al. Metabolic and cardiovascular effects of very-low-calorie diet therapy in obese patients with Type 2 diabetes in secondary failure: outcomes after 1 year Abstract School of Medical & Surgical Sciences, University of To evaluate the short-term and 1-year outcomes of an intensive very-low-Nottingham, an

Copyright © 2018 Predicting Disease Pdf