Background and aims:Emerging evidence suggests that highly treatment-resistant tumour-initiating cells (TICs) play a central role in the pathogenesis of pancreatic cancer. Tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) is considered to be a novel anticancer agent; however, recent studies have shown that many pancreatic cancer cells are resistant to apoptosis induction by TRAIL due to TRAIL-activated nuclear factor-κB (NF-κB) signalling. Several chemopreventive agents are able to inhibit NF-κB, and favourable results have been obtained—for example, for the broccoli compound sulforaphane—in preventing metastasis in clinical studies. The aim of the study was to identify TICs in pancreatic carcinoma for analysis of resistance mechanisms and for definition of sensitising agents.Methods:TICs were defined by expression patterns of a CD44+/CD24−, CD44+/CD24+ or CD44+/CD133+ phenotype and correlation to growth in immunodeficient mice, differentiation grade, clonogenic growth, sphere formation, aldehyde dehydrogenase (ALDH) activity and therapy resistance.Results:Mechanistically, specific binding of transcriptionally active cRel-containing NF-κB complexes in TICs was observed. Sulforaphane prevented NF-κB binding, downregulated apoptosis inhibitors and induced apoptosis, together with prevention of clonogenicity. Gemcitabine, the chemopreventive agents resveratrol and wogonin, and the death ligand TRAIL were less effective. In a xenograft model, sulforaphane strongly blocked tumour growth and angiogenesis, while combination with TRAIL had an additive effect without obvious cytotoxicity in normal cells. Freshly isolated patient tumour cells expressing markers for TICs could be sensitised by sulforaphane for TRAIL-induced cytotoxity.Conclusion:The data provide new insights into resistance mechanisms of TICs and suggest the combination of sulforaphane with TRAIL as a promising strategy for targeting of pancreatic TICs.

Little is known about the factors that enable the mobilisation of human mesenchymal stem cells (MSC) from the bone marrow into the blood stream and their recruitment to and retention in the tumour. We found specific migration of MSC towards growth factors present in pancreatic tumours, such as PDGF, EGF, VEGF and specific inhibitors Glivec, Erbitux and Avastin interfered with migration. Within a few hours, MSC migrated into spheroids consisting of pancreatic cancer cells, fibroblasts and endothelial cells as measured by time-lapse microscopy. Supernatant from subconfluent MSC increased sprouting of HUVEC due to VEGF production by MSC itself as demonstrated by RT-PCR and ELISA. Only few MSCs were differentiated into endothelial cells in vitro , whereas in vivo differentiation was not observed. Lentiviral GFP-marked MSCs, injected in nude mice xenografted with orthotopic pancreatic tumours, preferentially migrated into the tumours as observed by FACS analysis of green fluorescent cells. By immunofluorescence and intravital microscopic studies, we found the interaction of MSC with the endothelium of blood vessels. Mesenchymal stem cells supported tumour angiogenesis in vivo , that is CD31 + vessel density was increased after the transfer of MSC compared with siVEGF-MSC. Our data demonstrate the migration of MSC toward tumour vessels and suggest a supportive role in angiogenesis.

Genetic modification of human bone marrow mesenchymal stem cells (MSC) is highly valuable for their exploitation in basic science and therapeutic applications, for example in cancer. We present here a new, fast and easy-to-use method to enrich a functional population of lentiviral (LV)-transduced MSC expressing enhanced green fluorescent protein (eGFP). We replaced the eGFP gene by a fusion gene of puromycin acetyltransferase and eGFP. Upon LV gene transfer and puromycin selection, we quickly obtained a pure transduced MSC population, in which growth, differentiation capacity and migration preferences were not compromised. Furthermore, we are the first to report the migration velocity of MSC among which 30% were moving and velocity of about 15 μm h −1 was not altered by LV transduction. Manipulated MSC underwent senescence one passage earlier than non-transduced cells, suggesting the use for therapeutic intervention in early passage numbers. Upon tail vein application in nude mice, the majority of LV-transduced MSC could be detected in human orthotopic pancreatic tumor xenografts and to a minor extent in mouse liver, kidney and lung. Together, LV transduction of genes to MSC followed by puromycin selection is a powerful tool for basic research and improves the therapeutic prospects of MSC as vehicles in gene therapy.