Abstracts 2011 I am agree by gdf57j


									                                                                Meeting of BACR – February 5th, 2011 - Liège

C. Balsat, N. Signolle, F. Goffin, C. Munaut, A. Béliard, S. Blacher, P. Delvenne, K. Delbecque, A. Noël, J-M
Foidart, F. Kridelka
GIGA Research, Laboratory of Tumor and Development Biology, University of Liège, Pathology Tower (B23),
B-4000 Liège, Belgium, c.balsat@ulg.ac.be

Background and objective : Lymphangiogenesis is reported to be triggered when high grade lesions of the
cervix (HGSIL) progress to invasive carcinoma. We aimed at verifying lymphangiogenesis pattern in HGSIL
and microinvasive lesions in comparison to that observed in normal cervical tissue.

Material and Methods : 20 cases of cervical neoplasia (10 CIN3 and 10 FIGO stage 1A1) + 10 cases of normal
tissue were reacted with D2-40, an antibody to M2A, a specific lymphatic endothelial marker. Lymphatic
vessels (LV) were detected using a colorimetric method with further computer- assisted mathematical
morphology post-processing. Whole field of the Transformation Zone (TZ) was assessed from normal
squamous tissue through to normal glandular tissue. Respective LV density, morphology and distance to the
neoplastic and normal epithelium were measured and compared.

Results : In normal cervical tissue, prominent LV density was already detected under the TZ. In HGSIL, the LV
density and caliber were diffusely increased throughout the stroma. In early cervical cancers, LV density and
caliber remained elevated but were closely apposed to the invasive tumor bundles.

Conclusions : For the first time, prominent lymphangiogenesis is described under the TZ of normal cervical
tissue and appears preexisting to the neoplastic event. A further increase in LV density and caliber is observed
through the process of HGSIL development and microinvasion.

I agree.
                                                                Meeting of BACR – February 5th, 2011 - Liège

Bekaert S., Rocks N., Gueders M., Paulissen G., Noël A. and Cataldo D.
Laboratory of Biology of Tumors and Development, GIGA-Cancer, University of Liege and CHU of Liège,

The presence of leucocytes within tumors, observed in the 19th century by Rudolph Virchow, provided the first
indication of a possible link between inflammation and cancer. Yet, it is only during the last decade that clear
evidence has been obtained that inflammation plays a critical role in different stages of tumor development,
including initiation, promotion, malignant conversion, invasion and metastasis. There is increasing evidence
that an inflammatory microenvironment is an essential component of all tumors (Psaila, Lyden 2009).
Cigarette smoke is a complex mixture with over 4000 chemical components, inducing an inflammatory
response in the lower respiratory tract. Cytokines, chemokines and growth factors released by alveolar
macrophages, lymphocytes, neutrophils, endothelials cells and fibroblasts may act to promote epithelial
dysfunction and malignant progression.
In the present study, we assessed in vivo the impact of cigarette smoke (CS) on the tumor cell extravasation in
lungs after tail vein injection of B16F10 melanoma cells. We first characterized airway inflammation obtained
after smoke exposure (reference cigarettes 3R4F) for varied time periods (1, 2, 4, 8 and 12 weeks). Smoke
exposure was performed 5 days a week. Neutrophils, alveolar macrophages, interstitial macrophages, dendritic
cells, T cells and NKT were characterized in lung tissues of mice exposed to CS and AIR using flow cytometry.
In Vitro, the direct effect of cigarette smoke extract (CSE) on proliferation of B16F10 melanoma cells was
determined for 1 to 5 days. In vivo, mice exposed for 2 weeks to cigarette smoke or air were injected with
B16F10 melanoma cells in the tail vein. After 3 weeks, hematoxylin-eosin staining allowed quantification of
lung metastasis (tumor area/ total lung area). An increase of metastasis and implantation site in lungs was
observed in CS exposed group. Conceivably, CS constituents significantly promote extravasation of melanoma
cells in lung tissues. The mechanism or signaling pathway responsible for this dissemination needs to be further

I agree.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Xavier Bisteau, Laurence Bockstaele, Sabine Paternot, and Pierre P. Roger
Institute of Interdisciplinary Research (IRIBHM), Université Libre de Bruxelles, Campus Erasme, B-1070
Brussels, Belgium, xbisteau@ulb.ac.be

The homologous cyclin-dependent kinases (CDK) CDK4 and CDK6 act as master integrators in the G1 phase
coupling with the cell cycle mitogenic and oncogenic signaling cascades. CDK4/6 activity is deregulated
through various mechanisms in many tumors. Their activation requires binding to a D-type cyclin and then T-
loop phosphorylation at T172 and T177 (respectively). At variance with the coexistence of several CDK-
activating kinases in fungi and plants, in animals cells one single CAK constituted of the cyclin H-CDK7-Mat1
complex is considered to be responsible for activating phosphorylation of the various cell cycle CDKs,
including CDK4. At odds with the existing data concerning the constitutive activity of CDK7, we have recently
identified the T172 phosphorylation of cyclin D-bound CDK4 as a crucial cell cycle regulatory target of various
extracellular mitogenic or antimitogenic factors and their signaling cascades, determining CDK4 activity, pRb
phosphorylation and cell cycle progression. Here we show that T172 phosphorylation of CDK4 is conditioned
by its unique proline 173 residue. In contrast to CDK4, CDK6 does not contain such a proline and,
unexpectedly, remained poorly phosphorylated and active in a variety of cells. This difference in the regulation
of the activating phosphorylations of both CDKs was unexpected because of the high degree of homology in
the sequence surrounding their phosphoacceptor sites. Mutation of proline 173 of CDK4 to a serine did not
adversely affect CDK4 activation by CAK (CDK7), in agreement with the conception that recognition of CDKs
by CAK is independent of a consensus sequence around the phosphoacceptor residues. This sharply contrasts
with the complete ablation of CDK4 phosphorylation and activity by P173S mutation in intact cells.
Conversely, substituting a proline for the corresponding residue of CDK6 enforced its complete, apparently
cyclin-independent T177 phosphorylation and dramatically increased its pRb kinase activity. The S178P
mutation thus enforces the activating phosphorylation of CDK6, providing the first example of such a CDK-
activating mutation.
These results and the exquisite regulation of CDK4 phosphorylation lead us to propose that CDK4 might not be
phosphorylated by constitutively active CDK7 in intact cells but is more likely phosphorylated by another,
presumably proline-directed kinase(s). Moreover, they provide a new model of a potentially oncogenic-
activating mutation of a CDK.

Related references:
Paternot S., Bockstaele L, Bisteau X, Kooken H, Coulonval K, Roger PP 2010 Rb inactivation in cell cycle and
cancer: The puzzle of highly regulated activating phosphorylation of CDK4 versus constitutively active CDK-
activating kinase. Cell Cycle 9:689-99
Bockstaele L, Bisteau X, Paternot S, Roger PP 2009 Differential regulation of cyclin-dependent kinase 4
(CDK4) and CDK6, evidence that CDK4 might not be activated by CDK7, and design of a CDK6 activating
Mutation. Mol Cell Biol 29:4188-4200
Paternot S, Roger PP 2009 Combined inhibition of MEK and mammalian target of rapamycin abolishes
phosphorylation of cyclin-dependent kinase 4 in glioblastoma cell lines and prevents their proliferation. Cancer
Res 69:4577-81
Bockstaele L, Coulonval K, Kooken H, Paternot S, Roger PP 2006 Regulation of CDK4. Cell Division 1:25
Bockstaele L, Kooken H, Libert F, Paternot S, Dumont JE, de Launoit Y, Roger PP, Coulonval K 2006
Regulated activating Thr172 phosphorylation of cyclin-dependent kinase 4(CDK4): its relationship with cyclins
and CDK "inhibitors". Mol Cell Biol 26:5070-5085
I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                Meeting of BACR – February 5th, 2011 - Liège

Boeckx C, Wouters A, Deschoolmeester V, Specenier P, Pauwels P, Peeters M, Lardon F, Pauwels B, and
Baay M.
Center for Oncological Research Antwerp, CORE Antwerp.
corresponding author: Carolien Boeckx, University of Antwerp, Laboratory for Cancer Research and Clinical
Oncology, Universiteitsplein 1 (T3.11), B2610 Wilrijk - Belgium, phone +32 3820 2576, e-mail

Background: Formalin-fixed paraffin-embedded (FFPE) tissue is the most common tissue specimen widely
available. In addition, extensive clinical information is accessible. This makes FFPE material a precious source
of material for identifying predictive and/or prognostic biomarkers in cancer research on the basis of gene
expression. However, the main limitation of FFPE tissue is the significant reduction in quality and quantity of
the isolated RNA.
To identify the most optimal RNA isolation procedure for FFPE material, five commercially available RNA
isolation kits were compared and evaluated. We used FFPE sections from cervical cancer patients treated by
Methods: RNA was isolated from three paraffin sections (10 µm) per patient and isolated according to the
manufacturer’s protocol. The following RNA isolation kits were used: Rneasy® FFPE isolation kit (Qiagen),
RecoverAllTM Total Nucleic Acid isolation kit (Ambion), ArrayGradeTM FFPE RNA isolation kit (SA
Biosciences), NucleoSpin® FFPE RNA isolation kit (Macherey-Nagel) and QuickExtractTM FFPE RNA
Extraction Kit (Epicentre® Biotechnologies).
RNA extraction was carried out in RNAse-free environment. The concentration and purity (A260/A280 ratio
and A260/A230 ratio) of extracted RNA were measured using a Nanodrop ND-1000 spectrophotometer. The
integrity of the isolated RNA was assessed by capillary electrophoresis with an Agilent 2100 Bioanalyzer using
Agilent RNA 6000 Series Nano kits, expressed in RNA integrity numbers (RIN).
Results: The mean total RNA eluated by the different kits were as follows: Qiagen 25957ng ± 19417, Ambion
8249ng ± 2898, SA Biosciences 8070ng ± 3700 and Machery-Nagel 622ng ± 394. The purity was measured by
the A260/A280 and A260/A230 ratio. The mean A260/A280 ratios were as follows: Qiagen: 1.81 ± 0.23, SA
Biosciences: 0.66 ± 0.36, Ambion: 1.03 ± 0.37 and Machery-Nagel: 1.04 ± 0.61 and for the mean A260/A230
ratios: 1.88 ± 0.09, SA Biosciences: 1.61 ± 0.32, Ambion: 1.54 ± 0.30 and Machery-Nagel: 1.88 ± 0.61. The
RNA extractions from Epicentre® could not be measured by Nanodrop and, therefore, were excluded from
further analysis.
The mean RIN values were as follows: Qiagen: 2.19 ± 0,37, SA Biosciences: 2.09 ± 0.38, Ambion: 2.44 ± 0.12
and Macherey-Nagel: 2.30 ± 0.44.
Discussion: From the total RNA, the A260/A280 and A260/A230 ratio, it can be concluded that the Rneasy®
FFPE Isolation Kit from Qiagen gave the most consistent results. According to the RIN values, RNA isolated
with Ambion kit gave the highest RIN value.
Taken these results together, the kit from Qiagen appears the most appropriate kit. Therefore, this kit will be
used in our further studies that require RNA isolation from FFPE material.

I agree.
                                                                 Meeting of BACR – February 5th, 2011 - Liège

Brysse A. 1, Polette M. 2, Luczka E. 2, Bonnomet A. 2, Mestdagt M.1, Hunziker W. 3, Noël A. 1, Birembaut P.
2, Foidart J.M. 1and Gilles C1.
1 Laboratory of Developmental and Tumor Biology, GIGA-Cancer, University of Liège, 4000 Liège, Belgium.
2 Unité I.N.S.E.R.M. U514, Laboratoire Pol Bouin, I. F. R. 53, C.H.U. Maison Blanche, 51100 Reims, France.
3 Institute of Molecular and Cell Biology, Epithelial Cell Biology Laboratory, Singapore, Singapore.

Increasing data suggest that the acquisition of migratory and invasive properties by tumor cells is associated
with epithelial-to-mesenchymal transition (EMT) processes. At the molecular level, EMT phenomena involve,
among other mechanisms, a reorganisation of cell-cell adhesion complexes.
ZO-1 is classically known as a submembranous cytoplasmic molecule contributing to the structural
organization of tight junctions (TJs). A concept has recently emerged that, in addition to its structural role in
TJ’s organization, ZO-1 could be involved in signalling pathways favoring tumor progression once delocalized
from TJs.
Here, we first analyzed the subcellular localization of ZO-1 in relationship with the invasive potential of tumor
cells. Comparing different cell lines in vitro by immunostaining, we found ZO-1 mainly localized at the
membrane in non-invasive cell lines (MCF-7), whereas a predominantly diffuse cytoplasmic staining was
observed in invasive tumor cells (BT-549). We also examined ZO-1 distribution in breast tumor biopsies. In in
situ carcinomas, ZO-1 was expressed mostly at the cell membrane whereas a cytoplasmic/nuclear staining was
observed in invasive breast cancers.
These data prompted us to examine the potential pro-metastatic role of ZO-1. We thus employed shRNAs to
downregulate ZO-1 in two different human breast tumor cell lines expressing luciferase as a marker: MDA-
MB-231/LUC and MDA-MB-435/LUC, largely described as highly metastatic cell lines once injected
subcutaneously in mice. Using an in vivo imaging system (IVIS 200, Caliper/Xenogen) allowing the
visualization of luciferase expressing cells in living animals, we obtained preliminary results showing that
transfection of ZO-1 shRNA differentially module the metastatic phenotype in the two cell lines. Indeed, in
MDA-MB-435 shRNA ZO-1 cells, remaining ZO-1 was mainly found in the nucleus in the primary tumors
whereas in the control cells, ZO-1 redistributed at the membrane. In this cell line, ZO-1 shRNA transfection
also increased the metastatic potential. Inversely, in MDA-MB-231 ZO-1 shRNA, the total amount of ZO-1 is
decrease in all cell compartments and no membrane staining was observed either in the control cells or in the
shRNA expressing cells. In MDA-MB-231 cells, shRNA transfection reduced the metastatic potential.
Experiments to understand what triggers the differential ZO-1 localization in the two cell lines in vivo are
ongoing. So far, our data nevertheless suggest a pro-metastatic role of ZO-1 when it is not membrane-
associated. Experiments are also ongoing to determine the potential target genes of ZO-1 signalling in our

I agree.
                                                                      Meeting of BACR – February 5th, 2011 - Liège

Gang Chen*, Peter Kronenberger*§, Erik Teugels* and Jacques De Grève*
*Laboratory of Molecular Oncology and Department of Medical Oncology, Universitair Ziekenhuis Brussel, Vrije
Universiteit Brussel, Brussels, Belgium
§Laboratory for Biotechnology, Department of Gezondheidszorg, Erasmushogeschool Brussel, Brussels, Belgium.
Correspondence to: Prof. Dr. Jacques De Grève.
Address: Laboratory of Molecular Oncology and Department of Medical Oncology, UZ Brussel, Vrije Universiteit
Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium.
Tel: +32 2 4776415 Fax: +32 2 4776210 Email: Jacques.Degreve@uzbrussel.be

Keywords: RNA interference; SiRNA; EGFR; Transfection; Real-time RT-qPCR

Background and objective: The ability of small interfering RNA (siRNA) sequences to modulate gene expression (“RNA
interference”) has provided a powerful tool with which to study gene function, and might in the future be applicable to
clinical treatment of cancer. Double stranded siRNAs can be added to cells grown in vitro, and delivered to the cytosol
using appropriate transfection reagents. The efficiency by which these reagents induce the uptake of siRNA molecules is
cell type specific, varies widely, and is a major parameter affecting the efficiency of gene silencing. To achieve maximum
siRNA delivery and optimal gene knockdown, different siRNA transfection reagents and transfection conditions were
tested in the non-small cell lung cancer cell line H1975, targeting the epidermal growth factor receptor (EGFR).
Materials and methods: The sequence of the siRNA targeting wild type EGFR was designed based on the open reading
frame of the human EGFR reference gene (GenBank sequence NM_005228.3). EGFR siRNA transfection was performed
with different methods to achieve the best transfection efficiency in lung cancer cell line H1975. The following reagents
were tested: Lipofectamine 2000 (Invitrogen), siPORT Amine (Ambion), ICAFectin 442 (Eurogentec), N-TER
nanoparticle siRNA transfection system (Sigma), X-tremeGENE siRNA transfection reagent (Roche),
polyMagnetofection and combiMAGnetofection (Chemicell). All procedures were according to the instructions of the
manufacturer, and assayed in a 24-well plate format. Real time quantitative RT-PCR was performed in the LightCycler®
1.5 instrument with SYBR green at 530 nm to measure EGFR mRNA knockdown, by the delta delta Cq method and
using the GAPDH gene as an internal reference gene. The siRNA transfection reagent with the highest gene silencing
efficiency was further tested by different conditions. siGLO and TOX Transfection controls were obtained from Thermo
Scientific Dharmacon.
Results: The transfection efficiency in the lung cancer cell line H1975 was first compared with different transfection
agents with EGFR siRNA at 48 and 72hrs post transfection. Lipofectamine 2000, siPORT Amine transfection and
combiMAGnetofection all achieved more than 70% EGFR mRNA knockdown at 72hrs with 40nM siRNA, among which
Lipofectamine 2000 gained the highest silencing (77%±3%). To further investigate the optimal concentration of siRNA
and transfection reagent, different ratios of lipofectamine 2000 and siRNA were tested. At 72hrs post-transfection, the
knockdown rate was the most potent with a combination of 1.5 µl lipofectamine 2000 and 200nM EGFR siRNA
(94%±15%). To evaluate effects of transfection efficiency, two approaches were determined: cell fluorescence by siGLO
Transfection Indicators and cell death induced by TOX Transfection Control. In each case, the transfection efficiency was
both higher than 88% at 48hrs and higher than 98% at 72hrs as assessed by the CellTiter-Blue® Cell Viability Assay, or
by fluorescence microscopy. These data suggested that the transfection efficiency and the gene silencing were nearly
optimal with the siRNA tested.

The transfection reagent Lipofectamine 2000 was found to yield the highest percentage of EGFR gene silencing (94%) in
H1975 lung cancer cells (1.5 µl Lipo/ 200nM siRNA / 24 well format). This will allow the study of the biological effects
of EGFR shutdown in this cell line.

I agree.
                                                                        Meeting of BACR – February 5th, 2011 - Liège

Gang Chen*, Ijeoma Adaku Umelo*, Peter Kronenberger*§, Erik Teugels* and Jacques De Grève*
*Laboratory of Molecular Oncology and Department of Medical Oncology, Universitair Ziekenhuis Brussel, Vrije
Universiteit Brussel, Brussels, Belgium
§Laboratory for Biotechnology, Department of Gezondheidszorg, Erasmushogeschool Brussel, Brussels, Belgium.
Correspondence to: Prof. Dr. Jacques De Grève.
Address: Laboratory of Molecular Oncology and Department of Medical Oncology, UZ Brussel, Vrije Universiteit
Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium.
Tel: +32 2 4776415 Fax: +32 2 4776210 Email: Jacques.Degreve@uzbrussel.be

Keywords: BIBW 2992; RNA interference; SiRNA; EGFR; ErbB-2

Background and objective: BIBW 2992 is a representative of a new generation of tyrosine kinase inhibitors (TKIs). This
new compound is a potent and irreversible inhibitor of both the human epidermal growth factor receptor (EGFR; ErbB-1;
HER1) and human epidermal growth factor receptor 2 (ErbB-2; HER2/neu; ERBB2) kinases. Both of these receptors are
involved in cell proliferation, differentiation and apoptosis-inhibition, and play a critical role in tumor growth and spread.
BIBW 2992 irreversibly binds to the ATP binding pocket of the receptors and inhibits the downstream signaling cascade,
which in turn may inhibit cell growth and induce apoptosis in cancer cells. BIBW 2992 is being investigated for various
indications, including non-small cell lung cancer (NSCLC), breast cancer, colorectal cancer and head and neck cancer.
BIBW 2992 is currently in phase IIb/III clinical development in NSCLC. The objective of the present study was to study
the effects of BIBW 2992 combined with small interfering RNA (siRNA) targeting human wild type EGFR on tumor
viability and apoptosis of lung cancer cells.
Materials and methods: Lung cancer cell lines NCI-H292 (wild type for EGFR), NCI-H358 (wild type for EGFR, K-Ras
mutation), NCI-H1650 (E746-A750 deletion in EGFR, PTEN loss), HCC827 (E746-A750 deletion in EGFR, sensitive to
TKI’s ) and NCI-H1975 (L858R mutation , second mutation T790M, in cis, in the kinase domain of EGFR ) were treated
with indicated concentrations of BIBW 2992 for 72hrs. Cell proliferation was measured by CellTiter96 AQueous One
Solution Cell Proliferation Assay (absorbance, Promega) and cell viability was evaluated by CellTiter-Blue Cell viability
Assay (fluorescence, Promega). The induction of apoptosis was evaluated using a fluorescent Caspase-3/7 assay
(Promega), and by Hoechst 33342 and propidium iodide (PI) double fluorescent staining (microscopy). An siRNA
targeting human wild type EGFR was designed and transfected with lipofectamine 2000 into the five different lung cancer
cell lines, with and without BIBW 2992 treatment. The combined effect of proliferation inhibition was detected by
CellTiter96 AQueous One Solution Cell Proliferation Assay and combined effect of apoptosis induction was measured by
Caspase-3/7 assay.
Results: Time and dose-dependent proliferation inhibition and apoptosis induction were produced by BIBW 2992 in all
the five lung cancer cell lines tested. HCC827 cells, (sensitizing E746-A750 deletion), were significantly more responsive
to BIBW 2992 treatment than the other 4 cell lines, i.e. the most potent antiproliferative and apoptotic effect was observed
in this cell line. However, NCI-H1975 cells, which are resistant to reversible TKIs (Gefitinib, Erlotinib), showed more
response in both proliferation inhibition and apoptosis induction than EGFR wild type cell lines NCI-H292, NCI-H358
and NCI-H1650 (E746-A750 deletion, PTEN loss) . Compared to exposure to either siRNA or BIBW 2992 alone, the
combined treatment with EGFR siRNA and BIBW 2992 resulted in better cell growth suppression and greater induction
of apoptosis in all the five NSCLC cell lines, especially in the cell lines that are less sensitive to BIBW 2992.

Conclusions: BIBW 2992, a dual covalent inhibitor of EGFR and HER2 can inhibit proliferation and induce apoptosis in
lung cancer cells that have the T790M resistance mutation. EGFR RNA interference augments the effects of BIBW 2992
leading to inhibition of proliferation and induction of apoptosis, even in lung cancer cell lines that have downstream
resistance mechanisms.

I agree.
                                                                Meeting of BACR – February 5th, 2011 - Liège

Gang Chen*, Peter Kronenberger*§, Erik Teugels* and Jacques De Grève*
*Laboratory of Molecular Oncology and Department of Medical Oncology, Universitair Ziekenhuis Brussel,
Vrije Universiteit Brussel, Brussels, Belgium
§Laboratory for Biotechnology, Department of Gezondheidszorg, Erasmushogeschool Brussel, Brussels,
Correspondence to: Prof. Dr. Jacques De Grève.
Address: Laboratory of Molecular Oncology and Department of Medical Oncology, UZ Brussel, Vrije
Universiteit Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium.
Tel: +32 2 4776415 Fax: +32 2 4776210 Email: Jacques.Degreve@uzbrussel.be

Keywords: Real-time RT-qPCR; primer; RNA interference; siRNA; EGFR

Abstract: RNA interference (RNAi) can mediate a short-term or prolonged silencing of gene expression at the
RNA and protein level. Real-time quantitative RT-PCR (RT-qPCR) is a “gold” standard for measuring steady
state mRNA levels in RNA interference assays. In the present study, we evaluated the RT-qPCR assay used to
measure the siRNA knockdown of EGFR expression. In our experiments, we were interested in determining the
ability of a series of EGFR siRNAs to knock down the levels of endogenous EGFR mRNA expression in a
human lung cancer cell line (H358) expressing wild type EGFR. The knockdown of the epidermal growth
factor receptor (EGFR) gene with eight individual EGFR small interfering RNAs (siRNAs) was estimated by
RT-qPCR using three different RT-qPCR primer sets (q1, q2 and q3). Primer set q2 encompasses the region
targeted by s1247. Our data suggest that the EGFR mRNA fragments amplified by the q1 or q3 primer sets after
transfection of s1247 remain more intact despite RNA cleavage and thus result in an overestimation of the
amount of remaining intact target mRNA. These results thus suggest that s1247 is the most effective of the
siRNA’s tested, which was further corroborated by EGFR protein levels detection by Western blot and
measurement of possible phenotypic consequences of EGFR down-regulation in H358 cells. Cell viability and
caspase-3/7 activity were measured and, in addition, we evaluated the induction of apoptosis. The aggregate
results of these experiments are consistent with the mRNA knock down results obtained in the RT-qPCR
experiments and confirm that of the siRNAs tested, the s1247 is the most powerful siRNA on H358 cells to
down-regulate the EGFR protein level, inhibit the cell viability and induce apoptosis in comparison to other
EGFR siRNAs. To verify whether the effect of primer set choice on RT-qPCR results is also observed with
other messengers, we transfected a positive, validated GADPH siRNA into H358 cells and measured the
GAPDH mRNA knockdown level with 3 different primer sets locating in different positions. The data showed
that with different locations, the primers detected different knockdown efficiencies. The results from GAPDH
RNAi are analogous to the data obtained with the EGFR mRNA suggesting that these findings can be
generalized, i.e., measurements of the knockdown efficiency can be influenced by the primer set used for RT-
qPCR. We thus conclude that there is an unexpected but significant interdependence between the EGFR
targeting siRNA sequence and the RT-qPCR amplification region for assessing the efficacy of the siRNA target
gene knockdown and that this finding can be extended to other mRNAs in lung cancer cells. We therefore
recommend that qPCR primers for siRNA work should span the putative siRNA cleavage site, thus avoiding
the amplification of mRNA molecules that underwent only the initial steps of the siRNA induced degradation
process. It is also recommendable to test more than one primer set at different locations on the mRNA,
particularly for mRNA targets that appear to be refractory to siRNA-mediated cleavage. Ideally, at least one of
the primer sets should amplify a region encompassing the siRNA recognition sequence to ensure optimum
siRNA efficacy readout.
I agree.
                                                                  Meeting of BACR – February 5th, 2011 - Liège

Demoulin Stéphanie*, Herfs Michaël, Joan Somja, Roncarati Patrick, Boniver Jacques, Delvenne Philippe,
Hubert Pascale
Laboratory of Experimental Pathology, GIGA-Cancer, University of Liège, B-4000 Liège, Belgium

The cervical transformation zone is a dynamic area of a few millimeters in which a glandular epithelium has
been replaced by a squamous epithelium through a metaplastic process. Interestingly, a substantial majority
(87%) of cervical (pre)cancerous lesions develops within this peculiar microenvironment. Our previous studies
reported that intrinsic immune features altered in the metaplastic epithelium could contribute to cancer
development by preventing efficient antitumor/antiviral immune response. Plasmacytoid dendritic cells (pDC)
are key effectors in host innate immunity and orchestrate adaptive immune responses. Recently, infiltration by
these subtypes of dendritic cells has been shown in different cancers. However their implication in antitumor
response is largely debated. The present study was performed to determine the implication of pDC in the
cervical “metaplasia-dysplasia-cancer” sequence. We demonstrated that the density of pDC increases in the
epithelium of metaplastic and (pre)cancerous cervical tissues as well as in underlying stroma as compared with
normal exocervical epithelium. This could be partially explained by the increased expression of chemerin, their
chemotactic peptide, observed in those areas.
We developed a method to efficiently generate pDC cells exhibiting morphological and immunohistochemical
features of blood pDC from a limited number of CD34+ cord blood progenitors. Using these in vitro generated
pDC, we demonstrated that medium conditioned by transformed keratinocytes modified the activation status of
pDC, by inducing a decreased expression of costimulatory molecules such as CD86 and HLA-DR. Moreover,
malignant keratinocytes diminished the ability of pDC to produce IFN in response to an oligonucleotide
containing CpG motifs, a defined microbial stimulus for pDC.
These results suggest that pDC could be educated within the metaplastic and/or (pre)cancerous
microenvironment to acquire a tolerogenic phenotype that could promote carcinogenesis. In agreement with
those results, we observed that both metaplastic areas and (pre)cancerous lesions of the cervix are infiltrated by
T regulatory cells.

I agree.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

D. Desmet1 , C. Tellier1, O. Feron2, T. Arnould1, C. Michiels1
(1Laboratory of Biochemistry and Cellular Biology (URBC), University of Namur and 2Unit of Pharmacology
and Therapeutics, University of Louvain)

Cycling hypoxia, that is periods of hypoxia followed by reoxygenation, occurs in solid tumors due to transient
blood flow. These cycles of hypoxia and reoxygenation may influence cancer cell as well as endothelial cell
physiology. However, the effects of cycling hypoxia on endothelial cells are largely unknown. Another important
feature of tumoral microenvironment is inflammation. Among cytokines, the major pro-inflammatory mediator in
tumors is TNFα. Here, we investigated the specific effects of cycling hypoxia (CH), compared to continuous
hypoxia (H) and normoxia (N), combined to a pro-inflammatory environment on tumoral angiogenesis. EAhy926
endothelial cells were incubated under CH (4 cycles of 1 hour hypoxia followed by 30 minutes reoxygenation), or
under H or N for 6 hours. During these incubations, cells were stimulated or not with TNFα at [0.1ng/ml] to
simulate the pro-inflammatory environment.
We showed that cycling hypoxia induced a strong increase in the migratory ability of EAhy926 endothelial cells,
suggesting an effect toward an angiogenic phenotype. The inflammatory environment represented here by TNFα,
the major pro-inflammatory actor in tumors, did not seem to influence this angiogenic phenotype. However, the
tumor micro-environment is complex and we have to test other cytokines, such as IL-1β or IL-6, to confirm these
results. Furthermore, we showed that cycling hypoxia increased the activation of NF-κB initiated by TNFα in
EAhy926 endothelial cells. A pro-inflammatory environment was promoted as the secreted level of IL-8 was
elevated and the protein expression of ICAM-1 was stabilised. Such a pro-inflammatory environment could
favour propensity of endothelial cells to angiogenesis, leading to tumour growth and metastasis.

Presenting author :
Desmet Déborah
Laboratory of Biochemistry and Cellular Biology (URBC)
University of Namur
61, Rue de Bruxelles
5000 Namur
Email : deborah.desmet@fundp.ac.be

Note : I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                  Meeting of BACR – February 5th, 2011 - Liège

Benoît Detry, Silvia Blacher, Charlotte Erpicum, Jenny Paupert, Catherine Maillard, Nor Eddine Sounni and
Agnès Noël.
Laboratory of Tumor and Development Biology, Groupe Interdisciplinaire de Génoprotéomique Appliqué –
Recherche (GIGA-Cancer), University of Liège, Belgium. Benoit.Detry@ulg.ac.be

The formation of new blood vessels (angiogenesis) and lymphatic vessels (lymphangiogenesis) are critical
events in the development of several pathologies including cancer and ocular diseases. In severe corneal
inflammation, blood and lymphatic vessels grow into the avascular cornea reducing visual acuity and increasing
the risk of cornea graft rejection. Here, we studied the development of blood and lymphatic vessels in the
cornea after lesion induction and then evaluated the efficacy of the Sunitinib, a tyrosine kinase receptor
inhibitor, to reduce the neovascular reaction.
Cornea vascularization was induced by thermal cauterization applied in the center of the cornea of C57Bl6
mice. Blood and lymphatic vessels developing from vessels lying in the peripheral zone of the cornea called
limbus were observed after CD31 and Lyve-1 double immunolabelling on cornea whole mounts. Pictures were
analyzed by computer-assisted quantification and relative vascular area, end-point density, node density, length
density and maximal length of the vessels were determined to finely characterize blood and lymphatic vascular
In normal conditions, blood and lymphatic vessels were observed in a peripheral zone called limbus which
represents a junction between the avascular cornea and the broadly vascularised conjunctiva. A parallel growth
of the two vessel types was observed after thermal injury. Nevertheless, some differences were observed.
Indeed, blood vascular network rapidly grew from day 3, up to day 21. Then, a strong regression of blood
vasculature was observed. The lymphangiogenic process was slower and only a slight sprouting on the limbal
vessel was observed during the first days following lesion induction. The growth of lymphatic vessels was
observed between day 7 and 21. A regression of the lymphatic vasculature was also observed after 40 days but
it was less important than the one observed for blood vessels. We then evaluated the capacity of Sunitinib, a
tyrosine kinase receptor inhibitor used in cancer therapy, to inhibit lymphangiogenesis and angiogenesis in this
model. Mice were daily feeded with 40mg/kg Sunitinib or vehicle and sacrificed 6, 11 or 17 days after model
induction. Impact of Sunitinib on angiogenesis was observed after 17 days where relative surface, end-point
density, branching density and length density were 1.8 fold decreased. Maximum length of the blood vessels
was also significantly reduced in the treated group at days 11 and 17. Lymphangiogenesis study revealed a
significant reduction of the different parameters after 6, 11 and 17 days, except for maximum length of the
lymphatic vessels where no difference was observed.
Our results show that angiogenic and lymphangiogenic reactions occur in the cornea following a trauma.
However, blood vessels seem to regress rapidly after invasion of the central cornea. Lymphatic vessels develop
slower but are more stable and are still observed in the centre of the cornea after 40 days. Sunitinib can strongly
affect neovascularisation of the cornea after trauma and could enter in early treatment of such eye lesions to
avoid vision loss and risk of cornea graft rejection.

I agree.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

E. Gengoux, S. Lorquet, E. Hendrick, S. Berndt, A. Noël, J-M. Foidart, C. Munaut, C. Péqueux
University of Liège, GIGA-Cancer, Laboratory of Tumour and Development Biology, Institute of Pathology, CHU-
B23, 4000 Liège, Sart Tilman

Different growth factors and their associated receptors regulate angiogenesis and lymphangiogenesis. These
two processes are associated with normal physiology but also with several pathologies like tumor growth and
metastasis. Among the various growth factors, we are particularly interested in the VEGF family and their
receptors. The two main VEGF receptors implicated in angiogenesis are VEGFR-1 and VEGFR-2. In the other
hand, VEGFR-3 is the major receptor implicated in lymphangiogenesis, but recently, a soluble form of
VEGFR-2 was described as having an antilymphangiogenic role. This soluble form of VEGFR-2, named
sVEGFR-2, is detected in plasma of healthy people, in leukaemia and in systemic lupus erythematosus cases.
However, the mode of generation of this soluble receptor remains controversial and its physiological and
pathological implications are still poorly characterized.
In this context, the aim of this study was to identify the mechanisms leading to sVEGFR-2 release by
endothelial cells (EC) and lymphatic cells (LC).
The ectodomain shedding process was first assessed by evaluating the impact of various protease inhibitors on
sVEGFR-2 release. Among them, BB94 and TIMP-3 induced a partial inhibition of endothelial and lymphatic
sVEGFR-2 release. Additionally, PMA which is described to promote the ectodomain shedding, also promoted
sVEGFR-2 release. Altogether, these observations demonstrated a partial implication of the ectodomain
shedding process in the release of sVEGFR-2 by the two cell types. An RT-PCR strategy was developed to
identify the generation of a splice variant transcript of VEGFR-2. This strategy leads to amplification of a splice
variant from endothelial cells-derived RNA. In the EC, we observed that PMA upregulated the transcription of
the full-length VEGFR-2 but did not significantly modulate the spliced form, on the contrary PMA tend to
downregulate the spliced form production in lymphatic cells. Furthermore, PMA is known to induce a signal
transduction pathway involving the protein kinase C (PKC). The PKC inhibitor, GF109203x, did not affect the
basal release of sVEGFR-2; however, it inhibited in the EC the PMA-related increase of VEGFR-2 expression
and sVEGFR-2 release. In LC, GF109203x has no effects on the VEGFR2 expression.
In conclusion, our results demonstrate that the release of sVEGFR-2 by EC and LC relies at least on two
different cellular mechanisms: ectodomain shedding and alternative splicing. A combination of these two
mechanisms is observed under basal conditions. Altogether, our results suggest that the way by which
sVEGFR-2 is generated in EC and LC could probably closely related to the cell microenvironment context.

I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                Meeting of BACR – February 5th, 2011 - Liège

Ludivine Herman1, Pascale Hubert1, Michaël Herfs1, Gaelle Kustermans1, Yves Henrotin2, Latifa
Bousarghin3, Jacques Boniver1 and Philippe Delvenne1
   Department of Experimental Pathology, University Hospital of Liège, CHU Sart Tilman, Liège, Belgium
   Bone and Cartilage Research Unit, University Hospital of Liège, CHU Sart Tilman, Liège, Belgium
  Department of Microbiology, UPRES-EA 1254, Europeen University of Bretagne, University of Rennes1,
   Rennes, France

Human papillomavirus (HPV) infection, particularly type 16, is causally associated with cancer of the uterine
cervix. The progression of cervical lesions suggests that viral antigens are not adequately presented to the
immune system. The aim of this study was to determine if HPV16 viral particles can influence the trafficking of
human dendritic cells (DC)/ Langerhans cells (LC) either by direct interactions with DC or following
incubation with human normal keratinocytes which are in close contact with LC in the squamous epithelium.
We first demonstrated that HPV16 L1 major capsid protein when self-assembled into virus-like particles (VLP)
is able to induce in DC an over-expression of CXCR4, via the activation of the NF- B signaling pathway and
to enhance DC motility in the presence of CXCL12, suggesting an ability to migrate towards lymph nodes. We
also showed that conditioned media of HPV16 VLP-treated keratinocytes induce a lower LC migration than
those from untreated keratinocytes and that the high levels of prostaglandin E2 (PGE2) found in HPV16 VLP-
treated keratinocyte supernatants may be responsible for the reduced LC recruitment into the squamous
epithelium. Taken together, our data demonstrate that HPV16 VLP may differentially regulate the immune
protective response according to their target cells.

I agree.
                                                                 Meeting of BACR – February 5th, 2011 - Liège

Lorin Host, Alexandra Paye, Catherine Généreux, Nor Eddine Sounni and Agnès Noël.
Laboratoire de Biologie des Tumeurs et du Développement, GIGA-cancer, Université de Liège, tour de
pathologie (B23), +4, 4000 Liège
Tél. : 04366 2559
Fax : 04366 2569
Courriel : L.Host@ulg.ac.be

Tumor growth and metastatic dissemination are associated with an important tissue remodelling that
requires different proteases. The matrix metalloprotease (MMP) are involves in cell migration, cell
proliferation, invasion, apoptosis and also in physiological and pathological remodelling of the
extracellular matrix. In addition to their classical tissue-remodeling functions, MMPs act as processing
enzymes that perform highly limited cleavage of specific substrates including growth factor, cell surface
molecules, cytokines, chemokines and angiogenic factor. Among 24 MMPs known, 6 are anchored to the
plasma membrane (Membrane type-MMP, MT-MMP)
Previous works of our laboratory have shown that MT4-MMP overexpression confers to cancer cell lines
an increase of tumoral growth and metastatic spreading. These works also suggest that MT4-MMP is not
playing a redundant role with the other MMPs and that their molecular mechanisms are different.
Nevertheless, whether MT4-MMP displays pro-tumorigenic and pro-metastatic effects through its
catalytic activity remains to be determined. Furthermore, in vivo substrates of MT4-MMP are not known.
With the aim to investigate the mechanism of action of MT4-MMP (catalytic dependent or not), we have
generated a mutated form of MT4-MMP by site-directed mutagenesis. The mutation (in the highly
conserved domain, HExxHxxGxxH → HAxxHxxGxxH) was performed to inactivate the catalytic
domain of this enzyme and the construction has been transfected into MDA-MB-231. After transfected
cells screening, the in vitro phenotyping of cells did not reveal any differences in the proliferation
between the inactive form of the MT4-MMP (deltaMT4-pop) versus the active enzyme (MT4-pop) and
the control conditions (CTR-pop).
In sharp contrast, after injection of these cells in RAG -/- mice, MT4-MMP overexpression resulted in a
significant increased of tumor growth compared to control condition. Interestingly, the overexpression of
the mutated form showed a significant decrease of the tumoral volume, even when compared to control
We extend actually our research to other kinds of tumors. Different pulmonary epithelial cancer cell lines
(BZR and BZR-T33) who are express endogenous MT4-MMP were transfected with shRNA directed
against transcripts of this protease. These cells will be studied in vitro and in vivo to check if the
inhibition of MT4-MMP expression had an effect on the different process involve in metastatic spreading
(proliferation, migration, apoptosis, invasion and angiogenesis).
These results suggest that the pro-tumoral effect of the MT4-MMP is catalytic-dependent and also
indicate unexpectedly that the inactive form has a “dominant negative” effect.

I agree.
                                                                                       Meeting of BACR – February 5th, 2011 - Liège

Johan Ides1, An Wouters1, Peter Ponsaerts2, Marc Baay1, Greet G.O. Pattyn1, Hilde A.J. Lambrechts1, Marc
Peeters1, Bea Pauwels1, and Filip lardon1
1 Center for Oncological Research (CORE) Antwerp, Laboratory of Cancer Research and Clinical Oncology,
University of Antwerp, Wilrijk, Belgium
2 Laboratory of Experimental Hematology, University of Antwerp, Wilrijk Belgium
Tel: +32 3 265 25 76, Email: Johan.Ides@ua.ac.be

Cellular stress responses are essential for normal cellular growth and development while deregulation of one of
these processes can play a role in tumorigenesis. A lot of stress responses are regulated by mammalian target of
rapamycin (mTOR), which in turn regulates cell growth and cell cycle by inhibiting protein translation. REDD1
is one of the stress response genes that alter the activity status of mTOR. It inhibits mTOR in response to DNA
damage, hypoxia, glucose deprivation, and glucocorticoid treatment. During the last years, hypoxia has become
a big issue in solid tumors due to its adverse effects on tumor behavior. Under anoxic conditions, mTOR will be
rapidly inhibited by activation of REDD1. Inhibiting REDD1 expression could therefore establish a continuous
translation under anoxic conditions. As such, anoxic cancer cells might become more sensitive to certain
cytostatic drugs and/or radiotherapy that target cellular growth and proliferation.
To achieve a continuous silencing of REDD1, human A549 lung carcinoma cells were transfected with short
hairpin RNA targeted against REDD1 (A549shREDD1). After puromycin selection, A549 and A549shREDD1
cells were incubated under anoxic conditions (<0.1% O2) in a Bactron anaerobic IV chamber and harvested at
different time points (24-96h) to determine REDD1 protein expression by western blot analysis. Vindelov
staining was used to analyze cell cycle distribution by flow cytometry. The sulforhodamine B test was used to
determine the IC50 value of gemcitabine under normoxic and anoxic conditions. Cells were treated with
gemcitabine (0 - 100nM) for 24h. 2h prior to gemcitabine treatment, cells were incubated under anoxic
conditions to induce REDD1 upregulation.
REDD1 protein expression was already increased in A549 cells after 2h incubation under anoxia.
Downregulation of REDD1 was confirmed in the A549shREDD1 cells. These A549shREDD1 cells showed an
altered cell cycle distribution after incubation under anoxia, in comparison with the A549 cells. After an anoxic
incubation period, an increase of the percentage G0/1 cells was observed in A549 cells after 72h (A549: 80.68
± 15.06%; A549shREDD1: 65.28 ± 11.23%). Interestingly, the cell fraction in the S phase turned out to be
larger in anoxic A549shREDD1 compared to anoxic A549 cells with active REDD1 function (A549: 14.33 ±
12.96% A549shREDD1: 23.34 ± 6.23%). This suggests a potential role for REDD1 in anoxia-mediated cell
cycle arrest. The difference in REDD1 expression in these cell lines had no influence on the IC50 values of
gemcitabine of the different cell lines (IC50 under normoxia: A549: 17.34 ± 9.21; A549shREDD1: 19.49 ±
2.97; IC50 under anoxia: A549: 17.47 ± 5.79; A549shREDD1: 23.83 ± 1.03). Moreover, no sensitization of
anoxic cells due to the REDD1 inhibition could be observed. Inhibiting REDD1 translation seems to overcome
anoxia-induced G0/1 arrest. This could be an interesting approach to resensitize anoxic cells to conventional
cancer treatments that are targeted against dividing cells. However, although gemcitabine typically interacts
with the cell cycle, no difference in IC50 values could be observed between the wild type A549 and REDD1
silenced A549 cells, neither under normoxia nor under anoxia.

I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                Meeting of BACR – February 5th, 2011 - Liège

Isebaert S1, Van den Bergh L1, Haustermans K1, Lerut E2, De Wever L3, Bogaerts K4, Slagmolen P1, Joniau
S5, Van Poppel H5, Oyen R3
1 Department of Radiation Oncology, University Hospitals Leuven, Leuven, Belgium
2 Department of Histopathology, University Hospitals Leuven, Leuven, Belgium
3 Department of Radiology, University Hospitals Leuven, Leuven, Belgium
4 Leuven Biostatistics and Statistical Bioinformatics Centre, Leuven, Belgium
5 Department of Urology, University Hospitals Leuven, Leuven, Belgium

Purpose: Accurate detection and localization of intraprostatic tumor nodules is a prerequisite for focal therapy
of prostate cancer. The aim of this study was to determine the value of multi-modality imaging for prostate
cancer detection using T2-weighted (T2w) magnetic resonance imaging (MRI), dynamic contrast-enhanced
MRI (DCE-MRI) and diffusion-weighted MRI (DW-MRI), with whole-mount histopathology as a gold
standard reference.
Materials and methods: Forty-nine patients with biopsy-proven prostate cancer were enrolled in a prospective,
single-institution imaging study consisting of T2w-MRI, DCE-MRI and DW-MRI at 1.5T. All patients
underwent radical prostatectomy and extended lymphadenectomy. Histopathological evaluation was performed
by an experienced uropathologist and cancerous regions were delineated on the whole-mount haematoxylin and
eosin (H&E)-stained sections. MR images were evaluated independently by two experienced uroradiologists
who were blinded for the histopathological findings. Tumor regions were outlined on the different MR
modalities in different imaging sessions. Tumor localization was recorded and correlated between the
pathologic and MR imaging findings. Therefore, the prostate was divided into 3 regions (base, mid and apex)
and every region was divided further into octants or quadrants. Moreover, analyses were conducted with a strict
approach and a more clinically relevant approach, in which very small tumor lesions were excluded from the
Results and conclusions: In concordance with the inclusion criteria of this study, all patients presented with
intermediate to high risk prostate cancer (mean age, 64.3 years; median [range] prostate-specific antigen serum
levels, 10.8 [1.5-70.9] ng/mL; median [range] tumor volume, 7.4 [0.9-37] %). Statistical analysis is currently
ongoing to determine the sensitivity, specificity, accuracy, positive predictive value and negative predictive
value of the different imaging modalities for both readers. The inter-observer variability will also be
determined. Next, the histopathological delineated tumor volumes will be correlated with the volumes as
outlined on the different MR imaging modalities. All data will be presented by the time of congress.

*Presenting author:
Sofie Isebaert
Lab of Experimental Radiotherapy
University Hospitals Leuven Campus Gasthuisberg
CDG building, 8th floor, box 815
Herestraat 49
3000 Leuven

I agree.
                                                                Meeting of BACR – February 5th, 2011 - Liège

Heng Jiang, Valeri Verovski, Marieke Vermeersch, Kalun Law, Dirk Van den Berge, Dirk Verellen, Guy Storme and
Mark De Ridder.
UZ-Brussel, Vrije Universiteit Brussel , Department of Radiotherapy, Laarbeeklaan 101, B-1090 Brussels

Purpose : Activation of inducible nitric oxide synthase (iNOS) in hepatocytes is known to inhibit metabolic
oxygen consumption through synthesis of nitric oxide (NO) that interference with the mitochondrial respiratory
activity. This study examined whether NO-producing hepatocytes may radiosensitize neighboring colorectal
cancer cells through oxygen sparing, thus suggesting a novel approach to reverse hypoxia-induced
radioresistance in liver metastases.
Methods and Materials : Mouse hepatocytes and colorectal cancer cells (mouse CT26 and human DLD-1,
HCT116, HT29 and SW480) were exposed to a cytokine mixture (LPS, IFN-γ, TNF-α and IL-1β) and
examined for iNOS expression by RT-PCR and western blotting, and for enzymatic activity by NO/nitrite
production. To model metabolic hypoxia, tumor cells and hepatocytes were co-cultured at a ratio 1:1 in 6-well
plates in a double-layer tissue-mimetic system, which restricted oxygen diffusion through a 300 µm open space.
Oxygen consumption was measured by a fluorescence-based assay, and the activities of aconitase, complex I, II
and IV of the mitochondrial electron transport chain were analyzed by spectrophotometry. Co-cultures were
irradiated at doses of 0, 4, 8 and 12 Gy, and cell survival was measured by an 8-day colony formation assay.
Radiosensitization was determined as a dose enhancement ratio (ER) at the level of 0.1 surviving fraction.
Results : In a tissue-mimetic system, control (untreated) tumor cells alone could not induce metabolic hypoxia
at a cell density up to 0.5x105 cells/cm2. Contrasting, control hepatocytes became hypoxic within 60, 20 and 12
minutes at 0.125, 0.25 and 0.5x105/cm2 respectively. Co-culture of tumor cells and hepatocytes (both at
0.25x105/cm2) revealed similar hypoxia compared with hepatocytes alone, suggesting that hepatocytes were at
least 10-times more active in oxygen consumption. Following exposure to cytokines, hepatocytes showed
significant activation of iNOS (both mRNA and protein) and NO/nitrite production, while all colorectal cancer
cells showed little if any iNOS expression. As the result, the activity of mitochondrial aconitase and complexes
II in NO-producing hepatocytes was inhibited by 81% and 30% respectively, and oxygen consumption was
drastically blocked. The oxygen sparing effect was in line with an increased radiosensitivity of tumor cells,
which displayed ER of 2.3, 2.6, 1.5, 1.6 and 1.8-fold in CT26, DLD-1, HCT116, HT29 and SW480
respectively. The NO donor SNAP could also block oxygen consumption through the inhibition of aconitase
and complex IV (by 85% and 79%), and showed clear radiosensitization of DLD-1 by 1.7-fold. Both oxygen
sparing and radiosensitization could be counteracted by aminoguanidine, a metabolic iNOS inhibitor.
Conclusions : Our study for the fist time provides evidence that NO-producing hepatocytes may radiosensitize
colorectal cancer cells through oxygen sparing and reversal of hypoxia-induced radioresistance.

I agree to have the abstract released on the BACR website.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Mohammad Krayem, Mimoune Berehab, Murielle Wiedig, Renato Morandini, Fabrice Journe, Ghanem
LOCE, Institut Jules Bordet, ULB, Brussels, Belgium

The mutations of BRAF are described in about 50% of melanoma tumors with 90% of these mutations
occurring at a single site, leading to the V600E substitution. This mutation mimics the phosphorylation within
the activation segment of BRAF and results in the constitutive activation of the protein, leading to the
constitutive activation of the MAPK pathway, which is a critical signaling pathway for cancer cell survival.
However, the impact of V600EBRAF mutation on melanoma progression is unclear. In this study, we
compared 3 wild type (WT) melanoma cell lines (HBL, LND1, MM79) and 3 other cell lines with the
V600EBRAF mutation (MM32, MM43, MM74) addressing (i) cell proliferation, (ii) cell differentiation, (iii)
ERK phosphorylation, and (iv) response to MAPK pathway inhibition.
Mutant cells exhibited significantly higher levels of ERK1/2 phosphorylation. But, surprisingly, the
proliferation index (D3/D1 ratio) was significantly lower in the group of mutated cells as compared to the
group of WT cells (1.6 vs 5.0, p=0.01). Moreover, we found that small-molecule inhibitors of MEK (U0126,
PD98) inhibited cell proliferation at lower concentrations in mutated cells than in WT cells, indicating that
mutated cells are more dependent on MAPK signaling for survival. Importantly, the transcription factor MITF
may be activated through a crosstalk with the MAPK pathway, leading to increased transcription of MITF-
dependent genes, such as TYR, TYRP1 and DCT. However, we did not detect TYRP1 (mRNA/protein) in
BRAF mutated cells, indicating a rather limited melanocyte differentiation. To explain these contradictory
observations (high MAPK activation but low proliferation), a series of experiments was conducted. We found
that similar levels of apoptotic cells (annexin-positive cells) were present in both groups, indicating that the
lower growth rate of mutated cells was rather due to a reduced proliferation and not to an increase in cell death.
We also observed that the levels of the dual specificity phosphatase DUSP6, a specific phosphatase of ERK,
were higher in mutated cells, suggesting a blockade of the activated MAPK signaling, and, consequently, an
inhibition of cell proliferation.
In conclusion, in spite of an activating V600EBRAF mutation associated with a hyperphosphorylation of ERK,
mutated melanoma cells exhibit a lower proliferation index along with a decreased differentiation as compared
to WT. DUSP6 overexpression in mutated cells could explain the unexpected consequences of V600EBRAF
mutation on cell proliferation/differentiation. We hypothesize, therefore, that the the regulatory mechanism
involving this particular phosphatase could be critical in the control of melanoma cell proliferation.

I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                               Meeting of BACR – February 5th, 2011 - Liège

Virginie Lamour1, Jérôme Kroonen2, Manuela Dewald1, Aurélie Ory1, Zofia von Marshall3, Larry Fisher4,
Olivier Peulen1, Tieu-Lan Chau4, Alain Chariot4, Bernard Rogister2, Vincent Castronovo1 and Akeila
1Metastasis Research laboratory, GIGA-Cancer, 2Laboratory of Developmental Neurobiology, GIGA-
Neurosciences, University of Liège, Belgium, 3Craniofacial and Skeletal Diseases Branch, NIDCR, NIH,
DHHS, Bethesda, MD, USA, 4Interdisciplinary Cluster for Applied Genoproteomics (GIGA-R), GIGA-Signal
Transduction, Laboratory of Medical Chemistry, University of Liège, Belgium.

Osteopontin (OPN) is one of the rare proteins involved in almost all aspects of tumor progression and
metastasis development. OPN is overexpressed in high grade human glioma tumors. Glioblastoma is the most
common primary brain tumor in adults. A number of reports have demonstrated that these tumors contain a
subpopulation of stem cell-like tumor cells (glioma stem cells, GSCs) implicated in glioma progression,
therapeutic resistance and recurrence. We used a well-characterized human glioblastoma cell line (U87-MG)
comprising a fraction of tumor stem cells that grow as neurospheres in EGF and FGF-enriched culture medium.
The stemness phenotype of U87-MG neurospheres was confirmed by the expression of known stem cell master
genes (Sox2, Nanog and Oct3/4). Neurospheres also expressed OPN and to investigate its role in GSCs, we
performed lentiviral-mediated silencing. Our results demonstrate for the first time that OPN-deficient U87-MG
cells are unable to generate neurospheres in EGF-bFGF medium. Moreover, Sox2, Oct3/4 and Nanog mRNAs
remained at their basal levels in OPN-deficient cells suggesting that OPN is required for the acquisition of a
stem-like phenotype in EGF-FGF medium. Finally, the tumorigenic potential of OPN-deficient GSCs is
completely abolished in an orthotopic mouse model. Together, our data bring the first demonstration of a
greater implication of OPN than expected in glioma pathogenesis. Indeed, our study unveils OPN effects on
cancer stem cells subpopulation let us foresee that anti-OPN therapies may offer much greater benefit for
glioma patients.

I agree.
                                                                Meeting of BACR – February 5th, 2011 - Liège

J. Lecomte1, A. Masset1, S. Blacher1L., M. Jost1, L. Maertens1, A. Gothot², M. Delgaudine², F. Bruyere, M.
Illeman3, Ida K Lund3, Gunilla Høyer-Hansen3, J.-M. Foidart1, Agnès Noel1.
1Laboratory of Tumor and Development Biology, GIGA-Cancer, University of Liège, Tour de Pathologie
(B23), Sart Tilman, 4000 Liège.
²Department of Medicine/Hematology, GIGA-Research, CHU Sart-Tilman, 4000 Liège.
3Finsen Laboratory, Righospitalet, Copenhaguen, Denmark.

Ten years ago, Hanahan and Weinberg described six “hallmarks of cancer” which are essential alterations in
cell physiology that collectively dictate malignant growth [2]. However, this tumor cell-centric view does not
consider the microenvironment where malignant cells evolve. The importance of tumor microenvironment is
now recognized, but the multiple changes in tumor stroma affecting cancer evolution is incompletely
understood. Although roles of inflammatory cells and endothelial cells have been reported to be involved in
tumor immunity and neoangiogenesis, the contribution and the origin of tumor fibroblasts have not been fully
elucidated yet.
Through the years, there has been much debate regarding the extent to which connective tissue is formed from
local fibroblasts, or by mesenchymal cells arising from a distant source. Recently several observations
demonstrated the importance of not only local host cells interacting directly with cancer cells, but also many
other cell types issued from blood, bone marrow or lymphatic system. The cancer disease is now considered as
a systemic disease and not only as a local dysfunction spreading form one tissue.
In the present study, we investigated in vivo the putative contribution of bone marrow-derived cells into
malignant murine keratinocytes (PDVA). Mice were engrafted with bone marrow isolated from transgenic mice
expressing green fluorescent protein (GFP), and cancer cells were subcutaneous injected. There is no doubt that
bone marrow-derived cells (BMDC) are recruited into tumors and participate in cancer progression. The
mechanisms through which they contribute to tumor development are numerous: inflammation, angiogenesis
and also stromal reaction. Interestingly, bone-marrow derived cells were mostly localized in connective tissue
bundles. Some of these cells were fusiforms with a fibroblast-like morphology and were specifically associated
with collagen deposition. GFP+ bone marrow-derived cells express different fibroblastic/mesenchymal
markers, such as α-SMA, Thy1, NG2. Our results suggested that bone marrow-derived cells (GFP+) are
efficiently recruited into tumor, expressed several fibroblastic markers in vivo, and actively take part in the
desmoplastic reaction. We demonstrated that αSMA-expressing myofibroblasts produce MMP13 and
participate in tissue remodelling. This protease plays a key role in the MMP activation cascade by activating
MMP2 and MMP9 which are implicated in angiogenesis. Analyses of the impact of MMP13-producing bone
marrow derived fibroblasts are under process.

I agree.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Shasha Lv, Jan Sadones, Erik Teugels, Jacques De Grève, Bart Neyns
Laboratory of Molecular Oncology and Department of Medical Oncology, Universitair Ziekenhuis Brussel,
Vrije Universiteit Brussel, Brussels, Belgium.

Background: Expression of a constitutively active EGFR mutant, EGFR variant III (EGFRvIII), found in
approximately 20% of glioblastoma is exclusively in cases with EGFR gene amplification and confers a
proliferative and invasive advantage. The aim of this study was to investigate the correlation between EGFRvIII
and IDH1 gene mutation expression status or clinical outcome in of patients with recurrent high-grade glioma
enrolled in a phase II trial with the EGFR blocking monoclonal antibody cetuximab.
Methods: The total RNA was extracted from formalin-fixed and paraffin-embedded tumor tissues of 35
recurrent glioma patients which were treated with cetuximab at the time of recurrence. Rreverse transcription
and nested PCR were was performed to detect the EGFRvIII mutant. IDH1 mutation was detected with DNA
extraction, nested-PCR, DGGE and sequencing. SPSS17.0 software was used to analyze correlation between
EGFRvIII and IDH1 mutation status or available clinical data.
Results: EGFRvIII was found in 11/35(31.4%) patients including 9/26 (34.6%) de novo glioblastoma(dnGB),
1/6 (16.6%) secondary glioblastoma(sGB) and 1/3 (33.3%) grade II and III glioma. We confirmed that
EGFRvIII was exclusively found in patients with EGFR amplification. Moreover, EGFRvIII had a negative
correlation with IDH1 mutation(1/11 with IDH mut. vs 10/11 with IDH wt.). There was no significant
difference between patients with or without EGFRvIII in time to progression(TTP) on cetuximab, or overall
surviva(OS) from initial diagnosis and from the time of study recruitment. Counter intuitively, patients with an
IDH1 mutation had a worse survival from the time of study treatment. as compared to IDH1 wt patients.
Among patients with a IDH1 wt. status, those patient cohort with negative EGFRvIII (n = 19) had a significant
longer survival compared to all the other patients (n=16)in OS from study recruitment(median 169 days vs. 98
days, p=0.049) and TTP(median 76 days vs. 49 days,p =0.032) but not in OS from initial diagnosis(p=0.5).
Conclusions: In this patient population with high-grade glioma, treated with cetuximab at recurrence, EGFRvIII
did not correlate with the survival outcome, while, IDH1 mutation was correlated with a worse surival. Patients
with a wild type IDH1 and negative EGFRvIII had a better survival from the time of cetuximab treatment.

Keywords: EGFRvIII, IDH1, Glioblastoma, Cetuximab

I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Mardaga Julie, Begon Dominique and Delvenne Philippe
Experiemental Pathology Lab (Ulg) Av. de l’hopital n°3 Bat B23, tour de pathologie +4 4000 Liège
Email : Julie.Mardaga@ulg.ac.be

The Yin Yang 1 (YY1) transcription factor modulates the expression of many genes and has fundamental roles
in different biologic processes such as embryogenesis, differentiation, replication and cellular proliferation.
Several studies have shown that YY1 protein could play multiple roles in tumorigenesis. Moreover, some data
suggest that the YY1 gene is overexpressed in colorectal, ovarian and prostatic cancers.
The aim of this study is to characterize the role of YY1 transcription factor in colorectal tumorigenesis.
We first performed immunohistochemistry with an antibody against YY1 on sections from paraffin-embedded
colorectal tumors. We observed a nuclear YY1 expression in epithelial cells. The staining was visible both in
healthy and cancerous tissues. However, in healthy tissues the staining was more intense at the base of the
crypts than at the luminal surface, in contrast to cancerous tissues where the YY1 antibody stained intensely
and more uniformly the neoplastic cells. These results suggest that YY1 is highly expressed in proliferative
cells. To test this hypothesis, we performed immunostaining with antibodies against YY1 and against the
proliferation marker Ki67 on serial sections from the same colorectal tumor. It seems that there is a
colocalisation in the expression of this two stainings.
We also analyzed YY1 functions in HCT-116 colorectal cancer cells by inhibiting its expression using either
transient transfection of siRNAs or stable transfection of lentiviral vector bearing shRNA against YY1. We
confirmed the inhibition of YY1 expression by western blotting.
First, we studied the effect of this inhibition on basal apoptosis and no effect was observed. Secondly, we
analyzed the effect on proliferation. Results of Alamar Blue and WST-1 tests showed that the YY1 inhibition
induces a decrease of proliferation. Then we analyzed the effect of the YY1 inhibition on the ability of HCT-
116 cells to form colonies in soft agar. We observed a decrease in colonies formation in absence of YY1.
Currently, we are testing the effect of the YY1 inhibition in vivo by subcutaneous injection of HCT-116 cells
with or without YY1: settings are made and the first experiment is in progress.
In conclusion, our preliminary results suggest that YY1 expression could have a role in cell proliferation of
colorectal cancer cells and on the ability of the cells to form colonies.
In the future, we will confirm our results in HCT-116 but also in other colorectal cancer cell lines such as HT29
or HTm29 and repeat in vivo experiment.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                    Meeting of BACR – February 5th, 2011 - Liège

Melike Marsan1, Ridha Limame2, Katrien François1, François Bertucci3, Nato Ueno3, Filip Lardon2, Patrick
Pauwels2, Peter Van Dam1, Peter Vermeulen1, Luc Dirix1 and Steven Van Laere1
1 Translational Cancer Research Group Antwerp (Oncology Center, General Hospital Sint-Augustinus,
Wilrijk, Belgium)
2 Laboratory for Cancer Research and Clinical Oncology (University Antwerp, Wilrijk, Belgium)
3 Department of Breast Medical Oncology and Department of Stem Cell Transplantation (The University of
Texas MD Anderson Cancer Center, Houston, Texas, USA)
4 Département d’Oncologie Moléculaire (Centre de Recherche en Cancérologie de Marseille UMR891 Inserm,
Institut Paoli-Calmettes (IPC), Marseille France)

Introduction. We identified a core invasiveness gene (CIG) signature that predicts the invasive properties of
breast cancer cell lines. In the current study, we investigate the translational utility of the current gene signature
in human breast cancer, including inflammatory breast cancer (IBC). The latter is a highly invasive and
metastatic breast cancer subtype.
Materials and methods. Six publicly available gene expression data sets and a data set on 137 IBC and 252
non-IBC samples were analyzed. Each sample was classified according to sets of stromal (N=2), prognostic
(N=2), stem cell (N=3), epithelial-to-mesenchymal transition (EMT) (N=3) and pathway (N=20) gene
signatures. In addition, the cell-of-origin subtype classifier and our own CIG-signature were applied.
Associations between the CIG-signature and clinicopathological or molecular features were searched for using
univariate and multivariate analysis.
Results. We identified associations of the CIG signature (FDR<0.01) with a mammosphere-derived classifier
(median Rs=0.49) and with two EMT-signatures (Core-EMT: median Rs=0.52; EMT/stromal classifier: median
Rs=0.59). In addition EGFR, STAT3 and TGFb were found to be hyperactivated in CIG-positive tumors
(median Rs=0.45; 0.33 and 0.60 respectively). No relations between CIG-expression and stromal gene
expression, molecular subtypes or any of the clinicopathological variables were observed. Compared to non-
IBC, decreased CIG-expression, TGFb- and EGFR-activation were observed in IBC (respectively P<0.0001,
P<0.0001 and P=0.036). As for other EMT-associated gene signatures, both positive (Claudin-low and LYN-
signatures: P=0.004 and P=0.015) and negative (EMT/stromal classifier: P=0.015) associations with IBC have
been observed. When testing the core-EMT signature, no difference was found (P=0.166).
Discussion. We show that the CIG signature in breast cancer is correlated with EMT, stem cell biology and
EGFR-, TGFb- and STAT3-activation. Samples from patients with IBC demonstrate ambiguous EMT-patterns,
suggest that tumor cells from patients with IBC are in a specific state of cell plasticity and EMT as such is not
the main mode of invasion in IBC. On the contrary, lowered TGFb signaling in IBC suggest a form of
collective invasion.

I agree.
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Leroi Natacha1; Boujouf Sarah1; Coucke Philippe2; Noël Agnes1; Martinive Philippe1,2
1.Laboratory of Tumor & Development Biology (LBTD), University of Liège (ULg), 4000 Liège, Belgium
2.Radiotherapy Dept., University Hospital of Liège (CHU), University of Liège ULg, 4000 Liège, Belgium

Purpose: Colorectal cancer is the second leading cancer death and rectal tumors represent 30% of these cancers.
Neoadjuvant Radiotherapy (RT) (i.e. applied before surgery) in Locally Advanced Rectal Cancer (LARC)
decreases drastically local recurrences but has no impact on the overall survival and metastases occurrence. The
best radiotherapy schedule and the right timing of surgery (ST) haven’t been resolved, yet. We purpose to bring
some scientific rationale and molecular basis to solve these questions. Therefore, we hypothesize that
neoadjuvant RT schedules as well as the timing of the ST may influence tumor microenvironment and tumor
Methods: For this study, we developed a unique model of neoadjuvant RT in mice. We injected
subcutaneously, into the flank of the mice, mammary human tumor cells (MDA-MB231); irradiated precisely
the tumor at a total dose of 10Gy based on clinical schedules (2X5Gy and 5X2Gy); removed the tumor at the
4th or the 11th day after the end of RT and then led the mice alive after the surgery for metastases growth. We
quantified lung metastases by immunohistochemistry with human KI-67 labeling. Tumor expression of TIMP-
1, PAI-1, MT1-MMP and HIF-1α were analyzed by RT-PCR.
Results: We observed two different metastatic profiles according to the RT schedule and the time of ST. The
“RT schedule 2X5Gy” drastically decreased lung metastases compared to the unirradiated control mice.
Surgery performed at the 11th day decreased by more than a half the number and the size of metastatic islets
into the lungs compared to the group operated at the 4th day (n=4-5; p<0,05). Surprisingly, for “RT schedule
5X2Gy”, the timing of ST had smaller impacts on lung metastases. We observed a significant reduction of lung
metastases between groups (4th days vs 11th days) only for large metastatic islets (>50 and >100 cells) albeit
we noted the same trend for small metastatic islets (<10 and >10 cells) (n=13-19). Moreover, in 5X2Gy,
surgery performed at the 11th day appeared to favor lung metastases, mirroring the results obtained with RT
2X5Gy. We observed a significant increase of TIMP-1, PAI-1 and MT1-MMP mRNA into the tumor when ST
was performed at the 11th compared to the 4th day in the 2X5Gy but not in the 5X2Gy group. Neither the
tumor size nor the expression of HIF-1α was different between groups.
Conclusions: For the first time, we pointed here up that neoadjuvant RT schedules as well as the timing of the
ST influence tumor dissemination and lung metastases. Underlying mechanisms are still not completely
understood. They are probably dependent on the RT schedule. Further studies are needed for indentifying the
best RT schedule and ST window for improving treatments and the development of new strategies for patients
affected by LARC.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                  Meeting of BACR – February 5th, 2011 - Liège

Masset A.*, Lecomte J.*, Blacher S.,Maertens L., Foidart J-M and Noël A.
Laboratory of Biology Tumor and Developmental, groupe Interdisciplinaire de Génoprotéomique Appliquée
(GIGA-Cancer), Universitysity of Liège, Tour de Pathologie (B23), B-4000 Sart-Tilman, Belgium
* Equally contributed Anne.Masset@student.ulg.ac.be

Mesenchymal stem cells (MSC) are multipotent progenitor cells that contribute to the maintenance and
regeneration of a variety of connective tissues. Several reports proposed that the bone marrow-derived MSC is
a cell type that is recruited in large numbers in the stroma of developing tumors. In addition, bone marrow-
derived MSC have multiple roles in assisting or regulating cancer progression. Recently, Karnoub et al
(Nature,449,2007) showed that weakly metastatic human breast cancer cells acquired an increased ability to
disseminate to the lungs when they are mixed with MSC before being subcutaneously injected. However,
mechanisms by which MSC increase metastasis are not well known. In another hand, matrix metalloproteases
(MMP) are involved in several stages of metastasis process, including the escape of individual tumor cells from
the primary tumor, their intravasation, survival in circulation, and extravasation at the secondary site.
Therefore, the purpose of this study is to determine the implication of proteases in MSC-induced metastatic
dissemination. We established a model to study metastasis in C57BL/6J mice in which Lewis Lung Carcinoma
(LLC-Luc) cells expressing luciferase, were injected subcutaneously into C57BL/6J mice with or without MSC.
Tumors were resected at day 14 after tumor implantation. Lung metastasis development was visualized by
luminescence detection through IVIS® imaging technology. We provide evidence that MSC enhanced lung
metastasis development in mice. Indeed, 35 days after injection, a two-fold increase of mice bearing lung
metastasis was observed when MSC were present. In vitro, MSC increased the proliferation and migration of
LLC. Study of interaction between LLC and MSC in a three dimensional model of invasion, revealed that MSC
increased LLC invasion by the secretion of soluble factors rather than through cell-cell contacts. In addition, an
increase in PAI-1, MMP-13 and TIMP-3 RNA expression were observed in MSC when these cells were
cultured in co-cultures in which LLC and MSC were separated by a permeable membrane.
Altogether, these results suggest that MSC production of proteases could contribute to metastasis development
in mice. Investigations regarding the biological mechanism regulated by MSC in cancer metastasis are currently
under progress.

I agree.
                                                                 Meeting of BACR – February 5th, 2011 - Liège

A. Notte1, L. Flamant1, T. Arnould1, C. Michiels1
1 URBC-NARILIS, University of Namur-FUNDP, 61 rue de Bruxelles, B-5000 Namur, Belgium
E-mail : annick.notte@fundp.ac.be

Hypoxia is a microenvironment that is often associated with tumor resistance and therapy failure. Multiple
mechanisms are responsible for this adaptive response observed in cancer cells when exposed to hypoxia. This
work aims to understand what are the mechanisms activated under hypoxia that induce a resistance to
chemotherapy-induced apoptosis.
In order to investigate the effect of hypoxia on paclitaxel-induced apoptosis, MDA-MB-231 breast cancer cells
were incubated under normoxia or hypoxia with or without paclitaxel at 50 M and caspase 3 activity was
assessed. We showed that paclitaxel (50 M) triggers apoptosis since an increase in caspase 3 activity was
observed. Hypoxia prevents this activation, indicating chemoresistance. We also showed that paclitaxel induces
autophagy since LC3II accumulation and an increased colocalization of the autophagosomes and lysosomes
were observed. In order to investigate if autophagy is modulated by hypoxia and whether it is involved in the
protection observed under hypoxia, Atg5 and Atg7 siRNA were used. After exposure to paclitaxel LC3II
abundance was increased in control cells. However, no variation in LC3 II accumulation via the LC3I/LC3II
ratio was observed in paclitaxel incubated cells, after transfection with Atg5 siRNA suggesting that the
autophagy induced by paclitaxel is activated independently of Atg5 in normoxia. Under hypoxia, the
invalidation of Atg5 induced LC3II accumulation indicating that under hypoxia, Atg5 contributes to autophagy
and that its inhibition blocked the autophagic flow.
Finally, we showed that paclitaxel induced eif2 phosphorylation and CHOP and GRP78 expression probably
via the activation of the unfolded protein response (UPR). This hypothesis needs to be further investigated since
the UPR is known to influence not only adaptation and survival during ER stress but also cell death through
regulation of different effector pathways such as apoptosis or autophagy.

I agree.
                                                                 Meeting of BACR – February 5th, 2011 - Liège

Bea Pauwels1, Filip Lardon1, Jan B. Vermorken1, Johan Ides1, Vanessa Deschoolmeester1, Greet G.O.
Pattyn1, Hilde A.J. Lambrechts1, Paul Meijnders2, Kaye J. Williams3, Marc Peeters1, and An Wouters1
1 Center for Oncological Research (CORE) Antwerp, Laboratory of Cancer Research and Clinical Oncology,
University of Antwerp, Wilrijk, Belgium
2 Dept. of Radiotherapy, University Radiotherapy Antwerp (URA), Antwerpen, Belgium
3 School of Pharmacy and Pharmaceutical Sciences, University of Manchester, United Kingdom
Tel: +32 3 265 25 76, Email: bea.pauwels@ua.ac.be@ua.ac.be

Gemcitabine and difluorodeoxyuridine (dFdU) have apparent radiosensitizing properties under normoxic
conditions. During the past years, it has become evident that solid tumors often contain hypoxic regions and
that hypoxia is one of the causes of resistance or decreased sensitivity to chemotherapy or radiation. Under
hypoxic conditions, the transcription factor, hypoxia-inducible factor-1 (HIF-1), is upregulated. HIF-1 is
responsible for the cellular and adaptive responses of tumors to survive hypoxic conditions. Recently, it has
been demonstrated that gemcitabine retains its radiosensitizing potential under low oxygen conditions. In the
present study, the radiosensitizing potential of dFdU under anoxic conditions is investigated as well as the role
of HIF-1 protein.
In order to explore the role of HIF-1 protein, the human tumor cells included in this study were MDA-MB-231
(breast adenocarcinoma cell line, wt HIF-1), MDA-MB-231 DN-HIF (transfected with dominant-negative HIF-
1α, abrogating HIF-1 function) and MDA-MB-231 EV (empty vector transfected control). Anoxic conditions
(<0.1% O2,) were achieved in a Bactron IV anaerobic chamber. To analyze the radiosensitizing effect of dFdU,
the clonogenic assay was performed. Cells were exposed to normoxic or anoxic conditions and were
simultaneously treated with 0, 2 or 4 µM dFdU for 24h immediately before irradiation (0-8 Gy, room
temperature, linear accelerator). Immediately following radiation, anoxic cells were reoxygenated and cells
were washed with drug-free medium. Using radiosensitizing conditions, cells were collected for cell cycle
A clear concentration-dependent radiosensitizing effect of dFdU was observed under both normoxic and anoxic
conditions (dose enhancement factor (DEF) under normoxia: 1.74-3.38; DEF under anoxia: 2.08-3.81).
Combination index (CI) analysis showed a synergistic to additive interaction between dFdU and radiation under
normoxic conditions (CI 0.651 ± 0.175), and an additive interaction under anoxic conditions (CI 0.784 ±
0.120). Statistical analysis using two-way ANOVA revealed that cell survival was significantly influenced by
the concentration of dFdU, the radiation dose, the oxygen tension and the cell line used. Post hoc analysis
indicated no significant difference between MDA-MB-231 wt, MDA-MB-231 DN-HIF and MDA-MB-231 EV
cells for DEF, ID50, mean inactivation dose and survival fraction at 2 Gy, suggesting that the functionality of
HIF-1 protein has no impact on the radiosensitizing effect of dFdU. Considering the cell cycle distribution after
treatment with dFdU, 24h treatment with dFdU established a significant S-phase block in both normoxic and
anoxic MDA-MB-231 wt and MDA-MB-231 DN-HIF cells.
In conclusion, this study showed that dFdU has a clear concentration dependent radiosensitizing effect in
MDA-MB-231 breast cancer cells using normoxic as well as anoxic conditions. No major role for functional
HIF-1 protein in radiosensitization by dFdU could be demonstrated.
The research described in this abstract was supported by the Vlaamse Liga tegen Kanker and Foundation of
Scientific research (FWO).

I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                                      Meeting of BACR – February 5th, 2011 - Liège

Paye Alexandra, Laurent Host, Nor Eddine Sounni and Agnès Noël
Laboratory of Tumor and Developmental Biology, Groupe Interdisciplinaire de Génoprotéomique Appliquée-
Cancer (GIGA-Cancer), University of Liege, B-4000 Liège, Belgium.

We have previously demonstrated that MT4-MMP (MMP17), a membrane-anchored MMP essentially
expressed by breast tumor cells, promotes primary breast cancer growth and metastases, suggesting that MT4-
MMP plays a key role in metastatic dissemination of breast carcinoma cells to the lung. To determine the
molecular mechanism regulated by MT4-MMP in promoting metastasis of breast carcinoma cells, we assessed
the downstream signaling pathways activated by MT4-MMP in highly metastatic cells in comparison to low
metastatic control cells. By using a global proteomics phosphoantibody array approach (Kinetworks Phospho-
Site Screen 1.3, Kinetworks). We identified increased level of phosphorylated PKCa [S567], PKR1 [T451],
GSK3b [S9], B23 (NPM) [S4], Rb [S809/S811] and decreased level of phosphorylated MKK3/6 [S189/S207],
GSK3a [S21], Adducin g [S693]) in MT4-MMP overexpressing cells. The phosphorylation level of several
targets have been validated by western blot. Among them, the Rb pathway has been particularly investigated.
Indeed, hyperphosphorylation of Rb leads to inhibition of its activity and released of the transcription factor
E2F which transcribes factors essential for cell proliferation. Several targets transcribed by E2F, like cylin E or
cyclin C, were up-regulted in MT4-MMP expressing tumors. Moreover, we have observed an up-regulation of
CDK4 which associates with cyclin D to phosphorylate and activate Rb. Although no increase of proliferation
in MT4-MMP expressing cells compared to control cells has been observed in classical in vitro system, a new
model of cell proliferation on a matrigel layer has been developped and revealed an increased proliferation rate
of MT4-MMP expressing cells. This model will be used to test the implication of the Rb pathway in increased
proliferation, by addition of a specific inhibitor of CDK 4 (PD0332991). In conclusion, our observations point
an implication of the Rb pathway in the proliferating phenotype induced by MT4-MMP in vivo.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                 Meeting of BACR – February 5th, 2011 - Liège

Virginie Renoux1, Inge Langers1, Béatrice Clémenceau2, Marc Thiry3, Bettina Bisig, Estelle Dortu1, Anca
Reschner1, Jacques Boniver1, Philippe Delvenne1 and Nathalie Jacobs1.
1 University of Liège, GIGA-I3 Experimental Pathology, Liège, Belgium
2 INSERM U892, Institut de Recherche thérapeutique de l’Université de Nantes, Nantes, France
3 University of Liège, GIGA-Neurosciences Cellular and Tissular Biology, Liège, Belgium
4Department of Morphology and Pathology, University of Liège, Liège, Belgium,

Persistent infection with oncogenic human papillomavirus (HPV) genotypes is a necessary cause of uterine
cervical cancer. Recombinant HPV L1 protein self-assembles into virus-like particles (VLP) that are
morphologically and immunologically similar to native virions. These HPV-VLP were recently licensed as a
prophylactic vaccine against cervical cancer. Dendritic cell activation by internalisation of HPV-VLP has been
demonstrated, but despite the fact that CD16 have been described as a co-receptor for HPV-VLP in these cells,
their interactions with other CD16+ cells, such as Natural Killer (NK) cells, have not been assessed.
We analysed the presence of NK cells in cervical lesions and observed that NK cells infiltrated low- and high-
grade lesions where viral particles are produced but not squamous cell carcinoma where no virus capsid is
To study HPV-VLP internalisation in NK cells, we used fluorescent HPV-VLP with flow cytometry and confocal
microscopy. A weak entry of HPV-VLP in the NK cell line, NK92 (CD16-) was observed, whereas NK cells
isolated from blood (CD16+) internalised HPV-VLP rapidly. We confirmed these results by electronic
microscopy. Interestingly, CD16 transduction in NK92 cell line restored partially the HPV-VLP internalisation.
We also demonstrated that virus entry into NK cells was mediated by macropinocytosis and was independent of
clathrin and caveolin pathways. To test if NK activity could be influenced by HPV-VLP, we studied the lytic
granule exocytosis by a CD107 assay and observed an increased cell degranulation in the presence of HPV-VLP.
The expression of CD16 seemed necessary to induce the degranulation since only NK92 CD16+, and not NK92
CD16-, showed CD107 positivity. In conclusion, our data demonstrate for the first time that HPV interact with
NK cells and induce lytic granule exocytosis via the CD16.

Mail : v.renoux@ulg.ac.be

I agree.
                                                                  Meeting of BACR – February 5th, 2011 - Liège

Sas L1, Van Laere SJ1, Van der Auwera I1, Trinh XB1, Peeters D1, van Dam P1, Lardon F2, Pauwels P3,
Dierck A4, De Pauw A5, Dirix LY1 and Vermeulen PB1.
1Translational Cancer Research Group Antwerp (TCRG), Laboratory of pathology GZA Hospitals Sint
Augustinus/University of Antwerp, Antwerp, Belgium
2 Department of Medical Oncology, University of Antwerp, Antwerp, Belgium
3 Department of Pathology, University Hospital Antwerp/University of Antwerp, Antwerp, Belgium
4 Laboratory of pathology AZ Klina, Antwerp, Belgium
5 Laboratory of pathology AZ Nikolaas, Antwerp, Belgium

Background: Breast cancer is one of the most important causes of cancer-related mortality in women. Previous
research revealed the predictive role of ER in response to endocrine therapy. This nuclear receptor can be used
to select patients suitable for endocrine treatment. In common daily practice, ER protein expression level is
assessed by immunohistochemistry (IHC) on sections of formalin-fixed and paraffin-embedded tissue (FFPET)
followed by scoring of nuclear immunostaining. Pre-analytic and analytic variability between different
laboratories can influence the interpretations and results, with misclassification as a consequence. The purpose
of this study is to assess quality of the (semi)quantification of the protein expression of the estrogen receptor
alpha by IHC and histomorphometry on breast adenocarcinoma tissue sections (FFPET) in the participating
pathology laboratories.
Material and methods: In this multi-center study two participating pathology labs collected breast carcinoma
samples (N=20) and performed IHC and histomorphometry. The tissue samples and immunostained sections
were sent to the Translational Cancer Research Group (TCRG) for additional analysis. At TCRG the
immunostained sections were re-scored for ER and PR expression using the Allred system. Next, TCRG
performed immunostaining for ER, PR and HER2 of the unstained breast carcinoma tissue sections using the
DAKO pharmDx kit and the DAKO Herceptest, followed by scoring of the slides. HER2 membrane protein
expression was measured following ASCO/CAP guidelines. In a second branch of the study, the RT-PCR
technique was used. RNA was isolated from the tissue samples and used for quantitative RT-PCR for ESR1 and
for 18S and beta-actin as housekeeping genes.
Results: Comparing the immunostaining of the two other laboratories with the results of TCRG revealed similar
results. There was a significant correlation between the scores for ER as well as PR staining. Next, the scoring
procedure of TCRG and the other laboratories were compared. Analysis showed a significant positive
correlation between the sets of scores with one of the laboratories. The scores for ER staining of the other
laboratory showed a borderline correlation with the results of TCRG. Finally, the results obtained by IHC were
compared with those of qRT-PCR. The results of IHC were comparable with those obtained by PCR for both
Conclusions: There are significant correlations between the results obtained by TCRG and the different
laboratories. The pre-analytical and analytical variability in IHC do not seem to affect the results. Validation of
the IHC technique by qRT-PCR also showed a correlation in the expression of ER. The method of IHC for
assessment of ER expression and selection of patients for endocrine therapy is reliable.
Updated results will be presented.

I agree.
                                                                Meeting of BACR – February 5th, 2011 - Liège

1Translational Cancer Research Group Antwerp, St Augustinus Hospital, Antwerp, Belgium
2Department of Gynaecological Oncology, Antwerp University Hospital, Antwerp, Belgium

INTRODUCTION AND OBJECTIVES : Ovarian cancer is the leading cause of mortality among
gynaecological cancers. Identification of relevant pathways may contribute in individualisation of treatment
strategies. The objective of this study was to relate cellular pathway activation to survival outcome. Since
survival is studied, a selection was made of patients with advanced stages of serous papillary carcinoma with
uniform treatment to detect differences without histology-, stage- and treatment- induced biases.
MATERIAL AND METHODS : Two publicly available microarray datasets of advanced stage (III and IV)
ovarian cancer samples (N=165, N=125) were analysed for oncogenic pathways (AKT, BetaCatenin, E2F1,
EGFR, ER, HER2, MYC, INFa, IFNg, p53, p63, PI3K, PR, RAS, SRC, STAT3, TNFa, TGFb, VEGFA) (Gatza
et al. PNAS 2010, Hu et al. BMC Med 2009). These pathway gene signatures were correlated with survival
outcome and 3 prognostic gene signatures (Wound Response Signature WHR, Genomic Grade Index GGI and
the Invasiveness Gene Signature IGS).
RESULTS AND DISCUSSION : Although the WHR, GGI, and IGS have shown prognostic value in breast
cancer and other malignancies, these three prognostic signatures showed opposite survival outcome in this
selected population of advanced serous papillary carcinomas. Patients with a higher genomic grade had a better
survival outcome than lower genomic grade tumours. Since proliferation genes seems to be the driving force
within these prognostic signatures, this observation is concordant with earlier findings that tumours with high
proliferation index are more chemosensitive. We furthermore show that an activated BetaCatenin, RAS and p63
pathway was associated with favourable survival outcome in both datasets (p<0.05) displaying the predictive
value of combination carboplatinum-taxane chemotherapy. Furthermore these oncogenic pathways were
significantly correlated with IGS, WHR and GGI (p<0.001), suggesting that these pathways contribute to
chemosensitivity. These data are in line with findings of recent clinical findings that inhibition of
farnesyltransferase (and downstream RAS pathway) in combination with carbotaxol in first line was not
beneficial or even significantly inferior to carbotaxol alone.
CONCLUSIONS : Microarray analysis of two independent datasets shows that activation of the BetaCatenin,
RAS and p63 oncogenic pathways were consistently of predictive value for carboplatinum-taxane
chemotherapy in advanced stage serous papillary ovarian cancer. Since these pathways may contribute to
proliferation and consequent chemosensitivity, these findings may be of clinical importance for designing
treatment strategies. Validation studies are ongoing.

I agree
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Gert G Van den Eynden1, M Fiorentino2, Steven S Van Laere1, Dieter J Peeters1,3, X Bich Trinh2, Luc Y
Dirix2 and Peter B Vermeulen2
1Translational Cancer Research Group Antwerp, GZA Hospitals Augustinus/University of Antwerp, Antwerp,
2Laboratory of Oncologic and Transplantation Molecular Pathology, "Addarii" Institute of Oncology,
Bologna, Italy
3Dept of Oncology, University Hospital Antwerp, Edegem, Belgium

Introduction : Liver metastases of colorectal cancer (CRC) grow according to different growth patterns (GPs)
with different angiogenic properties. The aim of this study was to characterize these GPs on a molecular level
using genome-wide gene expression analysis.
Materials and Methods: On 18 CRC liver metastases of which genome-wide gene expression data were
available from a previous study (publicly available at GSE 10961), we assessed the GP using a haematoxylin-
eosin and Gordon Sweet’s reticulin stain. 7 metastases had a desmoplastic GP, 6 had a pushing GP and 3 had a
replacement GP. Of 2 metastases the GP could not be assessed due to lack of sufficient material and these were
excluded for further analysis. We used (hierarchical) cluster analysis, principal component analysis,
significance/prediction analysis of microarrays and gene set enrichment analysis to study differences in gene
expression between the different GPs.
Results Principal component analysis and hierarchical cluster analysis showed significant gene expression
differences between metastases with a desmoplastic and pushing GP (p=2.2e-16) and between metastases with
a desmoplastic and replacement GP (p=2.2e-16). Gene set enrichment analysis demonstrated increased
expression of biological processes, molecular functions, cellular components and KEGG pathways related to
immune response, antigen processing, leucocyte activation and cell adhesion molecules in the metastases with a
desmoplastic GP. Similar differences were found when metastases with a pushing GP were compared to
metastases with a replacement GP and when the 3 groups were compared.
Conclusion : These results confirm on a molecular level that liver metastases of patients with CRC have a
heterogeneous biology. Metastases with a desmoplastic GP appear to have upregulation of immunological
processes, compared to metastases with a pushing or replacement GP. Further elucidation of these differences
might lead to better understanding of host-tumour interactions in metastatic biology and new therapeutic
strategies for patients with metastatic CRC.

Presenting author:
Gert G Van den Eynden
TCRG St.-Augustinushospital
Oosterveldlaan 24
B2610 Wilrijk
+32 3 443 48 96

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Steven J Van Laere, Ilse Van der Auwera, Peter van Dam, Luc Dirix, Peter Vermeulen
Translational Cancer Research Group (Oncology Center, General Hospital Sint-Augustinus, Antwerp,

Inflammatory breast cancer (IBC) is one of the most aggressive manifestations of primary epithelial breast
cancer and differs from nIBC both from a clinical and a biological perspective. The increased metastatic
potential of IBC is highlighted by the dismal prognosis. Efforts have been undertaken to unravel the molecular
biology of IBC with high-throughput technology. Our group has generated transcriptome (Affymetrix HGU133
Plus 2.0), miRome (Applied Biosystems Panel A & B low-density PCR-arrays) and methylome
(HumanMethylation27 BeadChip) profiling data, in order to investigate the differential gene expression
between 17 IBC and 32 non-stage matched nIBC samples and to study how gene expression is epigenetically
regulated in this patient series. Each data set was analyzed separately to identify differentially expressed genes,
miRNAs and differentially methylated CpG-islands. In general, about 2379 (15%) genes are differentially
expressed between IBC and a non-stage matched nIBC control group at an FDR level of 0.1, from which
approximately 50% are upregulated. Using a similar stringency, we identified 9 (2%) differentially expressed
miRNAs (miR-29a, miR-29c, miR-30c, miR-30a-3p, miR-30e-3p, miR-24, miR-27b, miR-219-5p and miR-
190b) and 3 (less than 1%) genes with differentially methylated CpG-islands (AGT, CRIM1 and TJP3). All
miRNAs were repressed in IBC and all CpG-islands were more extensively methylated in IBC. When
investigating the mRNA expression of miRNA-processing genes between IBC and nIBC we observed
significant repression for DICER1 (p=0.045) and DROSHA (p=0.018) and significant overexpression of AGO2
(P=0.011) in IBC. As AGO2 is the catalytic subunit of the RISC-complex that is responsible for silencing the
miRNA targets, the latter observation suggests that the regulatory effect of the miRNAs on gene expression is
greater in IBC compared to nIBC. Indeed, we observed a global repression of miRNA-specific target mRNAs
in IBC as compared to nIBC (P<0.05). We also investigated the expression of DNMT1, DNMT3A and
DNMT3B between IBC and nIBC. For DNMT3B, no informative probe set could be indentified. For DNMT3A
a trend towards a significant overexpression in IBC (p=0.054), for DNMT1, no difference was observed.
Interestingly, the average expression of DNMT3A was about 2-fold lower than DNMT1. This suggests that de
novo methylation, which is DNMT3-dependent, is the mechanism repsonsible for the differences in CpG-site
methylation between IBC and nIBC. Finally, we performed an integrated unsupervised cluster analysis using all
three available data structures (transcriptome, mirome and methylome data). This led to 4 tumour subgroups
with different mRNA, miRNA and CpG-island methylation characteristics. In addition, these subgroups
showed a distinct enrichment pattern with respect to IBC and nIBC (P<0.0001).

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Steven Van Laere1,4, Naoto Ueno2,4, Pascal Finetti3,4, Peter Vermeulen1, Anthony Lucci2, Daniel
Brinbaum3, Fredika Robertson2, Takayuki Iwamoto2, Peter Van Dam1, Wendy Woodward2, Patrice Viens3,
Luc Dirix1, James Reuben2, and Francois Bertucci3,4
1Translational Cancer Research Group Antwerp, Onconlogy Center, General Hospital Sint-Augustinus,
Wilrijk, Belgium
2Departments of Breast Medical Oncology and Stem Cell Transplantation, The University of Texas MD
Anderson Cancer Center, Houston, Texas, USA
3Département d'Oncologie Moléculaire, Centre de Recherche en Cancérologie de Marseille, UMR891 Inserm,
Institut Paoli-Calmettes (IPC), Marseille, France
4Authors contributed equally

Introduction : Several studies have applied gene expression profiling to inflammatory breast cancer (IBC).
Most of these studies were underpowered, mainly because IBC is a rare disease and hence, IBC sample sizes
are small. Here, we present an integrated analysis of three distinct gene expression data sets of IBC and non-
IBC samples – thus with enhanced power - to further uncover the IBC-specific molecular biology with
enhanced statistical power.
Materials & Methods : Three Affymetrix gene expression data sets of 137 IBC and 252 non-IBC samples were
integrated. Data were normalized, and genes with high signal-to-noise ratios in at least 1% of the arrays were
filtered in. The samples were classified according to their molecular subtypes. IBC-specific heterogeneity was
investigated using hierarchical clustering, coupled with silhouette score analysis. Supervised analysis,
comparing IBC with non-IBC, was performed in non-stage-matched, stage-matched, and molecular subtype-
matched approaches with global testing. IBC-specific activated pathways, miRNA-families, and transcription
factors were identified with a target gene analysis approach.
Results : Four robust IBC sample clusters were identified, clearly associated with the molecular subtypes, with
a predominance of the combined Basal-like, ErbB2+, and Luminal B subtypes (~70% vs. ~40% in non-IBC;
P<0.0001). When we compared IBC to non-IBC, stage-matched and non-stage-matched differences were
identified (global test, P<0.0001). After comparing IBC with non-IBC samples within each of the molecular
subtypes, differences persisted only within the luminal A and normal-like subtypes (global test, P<0.0001 and
P=0.046, respectively). Target gene analysis identified two molecular pathways (TGFβ and INFα), 16
transcription factors (e.g. NKX2.2, FOXM1), and 10 miRNA families that were differentially activated in IBC
and non-IBC (FDR<0.01) but not in the four identified IBC sample clusters.
Conclusions : This meta-analysis demonstrates that IBC is indeed transcriptionally heterogeneous and that
differences between IBC and non-IBC are predominated by the differences in the distribution of the molecular
subtypes between IBC and non-IBC. Taking this into account, we were able to more accurately spot IBC-
specific changes in pathway, miRNA, and transcription factor activation.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                       Meeting of BACR – February 5th, 2011 - Liège

Eleni Van Schooneveld1, Maartje Wauters, Ilse Van der Auwera, Dieter Peeters, Philippe Huget, Peter Van
Dam, Peter Vermeulen, Steven Van Laere and Luc Dirix
Translational Cancer Research Group (Oncology Center, AZ Sint-Augustinus, Wilrijk, Antwerpen)

Introduction : miRNAs are a group of small non-coding RNAs involved in the regulation of gene expression.
As such, they regulate large number of cellular pathways and dysregulation or altered expression of miRNAs is
associated with tumorigenesis. In the current study we evaluate the feasibility and clinical utility of circulating
miRNAs as biomarkers for the detection and staging of breast cancer.
Materials and methods : miRNAs were extracted from a set of 84 tissue samples from patients with breast
cancer and 8 tissue samples from breast reductive surgery. After reverse transcription and pre-amplification, a
total of 768 miRNAs were profiled using the TaqMan Low-Density arrays. After data normalization, principal
component analysis (PCA) was used to investigate global differences in miRNA expression between cancerous
and normal samples. Using fold-change analysis, the most discriminating miRNAs between both tissue types
were selected to be analysed on serum samples from 20 healthy volunteers and 80 patients with breast cancer.
miRNAs were extracted from 200µL of serum and the selected miRNAs were reverse transcribed and analyzed
in duplo using qRT-PCR.
Results : PCA showed major differences in miRNA expression between tissue samples from patients with
breast cancer and tissue samples from breast reductive surgery (p<0.0001). Generally, miRNA expression in
cancerous samples is repressed when compared to miRNA expression in healthy controls (p=0.0685). The four
most discriminating miRNAs by fold-change (miR-215, miR-299-5p, miR-411, and miR-452) were selected for
further analysis on serum samples. 3/4 miRNAs revealed a differential expression profile between serum
samples from cancer patients and serum samples from healthy controls (miR-411: p=0.004; miR-452: p=0.001
and miR-299-5P: p=0.001). For all these miRNAs, the greatest difference in expression can be observed
between serum samples from healthy volunteers and serum samples from untreated breast cancer patients with
metastatic disease.
Discussion : Our study provides a basis for the establishment of miRNAs as biomarkers for the detection and
eventually staging of breast cancer through bloodborne testing. We identified and tested a set of putative
biomarkers of breast cancer and demonstrated that altered levels of these miRNAs in serum from patients with
breast cancer are particularly associated with the presence of metastatic disease.

    I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                     Meeting of BACR – February 5th, 2011 - Liège

M. Vermeersch, K. Law, A. Hoorens, V. Verovski, A. Sermeus, H. Jiang, G. Storme, D. Verellen and M. De
UZ-Brussel, Vrije Universiteit Brussel, Department of Radiotherapy, Laarbeeklaan 101, B-1090 Brussel.

Purpose : Total mesorectal excision (TME) is the basic treatment for rectal cancer, but results in 5 to 20%
incidence of local recurrence. Additional pre- or postoperative radiotherapy, either alone or in combination with
chemotherapy, has shown to decrease the risk of local relapse. Furthermore, the patient’s immunological status
is complexly related to his prognosis. In this study, the effects of neo-adjuvant pre-operative intensity-
modulated and image-guided radiotherapy in combination with a simultaneous integrated boost on the gross
tumor volume is being investigated in terms of decreased toxicity and oncological safety. In addition, biopsies
from rectal cancer are examined for predictive markers of inflammatory infiltrate in relation to local response.
Methods : Patient biopsies (n=45) were collected during diagnostic colonoscopy and at the moment of surgery
and evaluated on a protein level. IHC staining for proteins was performed on paraffin sections using a labelled
streptavidin-biotin method and quantified by computer-assisted image analysis with Definiens software.
PET/scans were performed before and 5 weeks after neo-adjuvant radiotherapy, and pathological response and
toxicity scores were calculated.
Results : As a first step, computer-assisted analysis of tumor biopsies was optimized and validated. Manual
counts of the number of positive cells per biopsy could be correlated to the % of positive area per biopsy and
the number of positive cells per biopsy. In a second step, all tumor tissue samples were investigated for their
immunological infiltrate by these two different algorithms (% area of positive tumor tissue and number of
positive cells per tumor area). For the given set of samples, no correlation could be found between the number
of CD3, CD4, CD8 and CD68 cells in the pre-operative samples and the local response parameters (∆SUV max,
∆SUV mean, ∆SUV volume, ∆SUV metabolic volume) through IHC study. Likewise, the number of these
subsets of cells was not related to the pre-treatment PET values, an effect that had previously been suggested
for macrophages.
Conclusion : No correlation could be found between the number of immune cells in the pre-treatment biopsies
and the local response after neo-adjuvant radiotherapy. Larger cohorts of samples will be examined to confirm
these findings and post-treatment samples will be included in the study.

* I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                                       Meeting of BACR – February 5th, 2011 - Liège

An Wouters1, Bea Pauwels, Johan Ides1, Marc Baay1, Greet G.O. Pattyn1, Hilde A.J. Lambrechts1, Paul
Meijnders2, Kaye J Williams3, Jan B. Vermorken1, Marc Peeters1, and Filip Lardon1
1 Center for Oncological Research (CORE) Antwerp, Laboratory of Cancer Research and Clinical Oncology,
University of Antwerp, Belgium
2 Dept. of Radiotherapy, University Radiotherapy Antwerp (URA), Belgium
3 School of Pharmacy and Pharmaceutical Sciences, University of Manchester, United Kingdom
Tel: +32 3 265 25 76, Email: An.Wouters@ua.ac.be

Over the last decades, it has been well documented that poor oxygenation is a characteristic pathophysiological
property of the majority of human solid tumors. Hypoxic tumor regions often contain viable cells that are
intrinsically more resistant to anticancer treatment. Recently, it has been demonstrated that the cytotoxic drug
gemcitabine retains its radiosensitizing potential under low oxygen conditions in lung adenocarcinoma cells. As
the transcription factor ‘hypoxia inducible factor 1’ (HIF-1) plays a crucial role in regulating the adaptive
responses of tumor cells to survive under hypoxic conditions, the present study investigated the potential
influence of HIF-1 status on radiosensitization by gemcitabine.
In order to explore the role of the HIF-1 protein, three isogenic human breast adenocarcinoma cell lines with
different HIF-1 status were included in this study: MDA-MB-231 (wt HIF-1), MDA-MB-231 DN-HIF
(transfected with dominant-negative HIF-1α, abrogating HIF-1 function) and MDA-MB-231 EV (empty vector
transfected control, functional HIF-1). Anoxic conditions (<0.1% O2,) were achieved in a Bactron IV anaerobic
chamber. To analyze the radiosensitizing effect of gemcitabine under normoxic versus anoxic conditions, the
clonogenic assay was performed. Cells were exposed to normoxic or anoxic conditions and were
simultaneously treated with 0, 4 or 8 nM gemcitabine for 24h immediately before irradiation (0-8 Gy, room
temperature, linear accelerator). Immediately following radiation, anoxic cells were reoxygenated and all cells
were washed with drug-free medium. Using radiosensitizing concentrations of gemcitabine, cell cycle
distribution was monitored flow cytometrically according to the Vindelov method.
Under both normoxic and anoxic conditions, a clear concentration-dependent radiosensitizing effect of
gemcitabine was observed in all three MDA-MB-231 cell lines (dose enhancement factor (DEF) under
normoxia: 1.02-1.70; DEF under anoxia: 1.11-2.04). Combination index (CI) analysis showed an additive
interaction between gemcitabine and radiation under normal oxygen conditions (CI 0.902-1.148), and a
synergistic interaction under reduced oxygen conditions (CI 0.455-0.889). No significant difference in oxygen
enhancement ratio (OER) was observed between the three isogenic cell lines (OER in MDA-MB-231 wt:
1.47±0.74; OER in MDA-MB-231 EV: 1.34±0.74; OER in MDA-MB-231 DN-HIF: 1.50±0.63). Moreover,
statistical analysis using two-way ANOVA revealed no significant differences in radiobiological parameters
(DEF, ID10, ID50, mean inactivation dose, surviving fraction at 2 Gy) between MDA-MB-231 EV and MDA-
MB-231 DN-HIF cells, suggesting no significant influence of HIF-1 functionality on radiosensitivity.
Considering the cell cycle distribution after treatment with gemcitabine, 24h treatment with 4 or 8 nM
gemcitabine established a significant S-phase block in both normoxic and anoxic MDA-MB-231 wt and DN-
HIF cells.
In conclusion, this study showed that the retained radiosensitizing effect of gemcitabine under anoxic
conditions was not tumor tissue specific and could be observed in MDA-MB-231 breast cancer cells. As HIF-1
proficient and HIF-1 deficient cells were equally radiosensitized, no major role for functional HIF-1 protein in
radiosensitization by gemcitabine could be demonstrated.

The research described in this abstract was supported by the Vlaamse Liga tegen Kanker and the Research
Foundation - Flanders (FWO).
I agree to have the abstract released on the BACR website before the conference in February 2011.
                                                                                      Meeting of BACR – February 5th, 2011 - Liège

Bruno Dumont, Andrei Turtoi, Yannick Greffe, Ying Hong Wang, Akeila Bellahcène and Vincent Castronovo
Metastasis Research Laboratory, University of Liege, Bat. B23, CHU Sart Tilman, B-4000 Liège, Belgium

A typical consequence of the metastasizing breast cancer cells is their colonization of the bone tissue. Up to
date there is no effective specific treatment. Targeted immunotherapy is an option to specifically deliver
cytotoxic agents to the malignant lesions. However, the identification of targetable systemically reachable
cancer antigens precludes the development of novel and effective antibody-based targeted therapies. In this
study, we had the unique opportunity to examine bone metastasis and the corresponding breast cancer primary
lesion obtained simultaneously from a fresh autopsy performed on a patient who died from disseminated breast
cancer. We were interested in identifying accessible protein biomarkers using a procedure which consists of
fresh tissue biotinylation. The biotinylated proteins are captured by streptavidin affinity chromatography and
the peptides derived from tryptic digestion are analyzed using the 2D-HPLC-MS/MS technique. In the present
study 519 proteins were identified in the primary breast cancer and 768 proteins were found in the bone
metastasis lesion. The comparison of the differential expression of the primary breast tumor and the bone
metastasis yielded 234 biomarker proteins of which 78 were found to be located in the extracellular matrix
and/or in the plasma membrane. These proteins are particularly interesting to serve as potential candidates for
antibody targets. 29 of the potential biomarkers were found uniquely expressed in the primary breast cancer
while 27 proteins were detected only in the bone metastasis lesions alone. In particular several proteins
belonging to small leucine rich proteo-glycans, thrombospondin and integrin families were found up-regulated
in the primary breast tumor. Our study has identified potential valuable biomarkers for the selective antibody-
based targeted eradication of bone metastases and the results indicate that the accessible biomarkers for bone
metastases are significantly different than the ones of the corresponding primary breast tumor.
This work is financed by the European Community FP6 project ADAMANT.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                    Meeting of BACR – February 5th, 2011 - Liège

Arnaud Blomme, Bruno Dumont, Andrei Turtoi and Vincent Castronovo
Metastasis Research Laboratory, GIGA Cancer, University of Liege, Bat. B23, CHU Sart Tilman, B-4000
Liège, Belgium

Targeted cancer therapies are nowadays gaining importance in the effort to provide a more specific tool for
patient treatment. Extracellular and membrane cancer protein biomarkers are ideal targets as they bear the
potential to be accessible to systemically administered compounds. However, focusing on one biomarker
assumes its relative homogenous distribution within the lesion. In the frame of the current work we have
explored the heterogeneity of the accessible proteome of liver metastasis from colorectal carcinoma (CRC).
Accordingly, we have ex-vivo biotinylated accessible proteins from several CRC-liver metastases and divided
the specimen in 4 zones: normal, peri-tumoral, tumor-rim and center. The proteins were affinity purified and
analyzed for each zone separately using nano-UPLC-MSe proteomics technique. In total over 1500 unique
proteins were statistically divided into six patterns of expression. Approximately 1/3 was expressed solely in
one of the 4 zones. A further 1/3 was found in all zones. Remaining proteins were present in 2 or 3 regions
studied. Interestingly, significant differences were notable between normal tissue collected far away and the one
sampled in the peri-tumoral zone. Finally, using IHC and more individual samples we have validated several
novel and known proteins for their heterogeneous distribution.
This work is funded through FP7 ADAMANT project granted by European Union.

I agree to have the abstract released on the BACR website before the conference in February 2011
                                                                                     Meeting of BACR – February 5th, 2011 - Liège

Contribution of osteopontin to the acquisition and/or the maintenance of a stem cell phenotype in U87-
MG human glioblastoma cells
Ory Aurélie, Lamour Virginie, Dewald Manuela, Castronovo Vincent and Bellahcène Akeila
Metastasis Research Laboratory, GIGA-Cancer, University of Liège, Belgium

Osteopontin (OPN) is a glycoprotein belonging to the Small Integrin-Binding LIgand N-linked Glycoproteins
(SIBLING) family first described in mineralized tissues. Since several years, the expression of this protein has
been studied in cancer where it correlates with tumor aggresiveness. This correlation has notably been
described in human gliomas which are often incurable tumors. The recent discovery that gliomas contain a sub-
population of cancer stem cells (CSCs) might explain tumor initiation and recurrence.
In this study, we have investigated the potential role of OPN in CSCs which are characterized by multipotency
and self-renewing. For this purpose, we have established, in our laboratory, two models for the study of CSCs.
The first one allows the enrichment in CSCs from the human glioblastoma cell line U87-MG when cells are
cultured in serum free medium comprising two mitogenic factors: the epidermal growth factor (EGF) and the
basic fibroblast growth factor (bFGF). In these conditions, CSCs proliferate and form floating spheres called
neurospheres. At the opposite, the second model consists in the induction of CSCs differentiation by mitogenic
factors suppression and serum addition in the culture medium. Using these two models, we demonstrated that
OPN is highly expressed in CSCs, decreased at day 1 of differentiation and then was re-expressed at day 8 in
differentiated cells.
To further investigate OPN role in CSCs, we used small hairpin RNAs to silence OPN expression in U87-MG
cells. OPN inhibition decreased CSCs growth and neurosphere formation demonstrating the importance of this
protein for the CSC phenotype.
All these results led us to study the transcriptional regulation of OPN gene promoter in U87-MG cells and in
the corresponding CSCs by electrophoretic mobility shift assay (EMSA). We found that some regulator
complexes were differently present in the two cell types suggesting the implication of different regulatory
mecanisms for the expression of OPN in U87-MG cells and in the corresponding CSCs.
Now, we intend to identify the transcriptional factors implicated in these complexes and also to extend our
study to other regions of OPN gene promoter.

Presenting author e-mail: a.ory@student.ulg.ac.be

I agree to have the abstract released on the BACR website before the conference in february 2011

To top