Characterization and Cytotoxic Effects on Pancreatic Cancer Cells

ORIGINAL ARTICLE TRAIL-engineered pancreas-derived mesenchymal stem cells: characterization and cytotoxic effects on pancreatic cancer cells

MR Moniri1,4, X-Y Sun2,4, J Rayat1, D Dai3,ZAo1,ZHe1, CB Verchere3, L-J Dai1 and GL Warnock1

Mesenchymal stem cells (MSCs) have attracted great interest in cancer therapy owing to their tumor-oriented homing capacity and the feasibility of autologous transplantation. Currently, pancreatic cancer patients face a very poor prognosis, primarily due to the lack of therapeutic strategies with an effective degree of speciﬁcity. Anticancer gene-engineered MSCs speciﬁcally target tumor sites and can produce anticancer agents locally and constantly. This study was performed to characterize pancreas- derived MSCs and investigate the effects of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-engineered MSCs on pancreatic cancer cells under different culture conditions. Pancreas-derived MSCs exhibited positive expression on CD44, CD73, CD95, CD105, negative on CD34 and differentiated into adipogenic and osteogenic cells. TRAIL expression was assessed by both enzyme-linked immunosorbent assay and western blot analysis. Different patterns of TRAIL receptor expression were observed on the pancreatic cancer cell lines, including PANC1, HP62, ASPC1, TRM6 and BXPC3. Cell viability was assessed using a real-time monitoring system. Pancreatic cancer cell death was proportionally related to conditioned media from MSCnsTRAIL and MSCstTRAIL. The results suggest that MSCs exhibit intrinsic inhibition of pancreatic cancer cells and that this effect can be potentiated by TRAIL-transfection on death receptor-bearing cell types.

Cancer Gene Therapy (2012) 19, 652--658; doi:10.1038/cgt.2012.46; published online 6 July 2012 Keywords: cytotherapy; gene therapy; mesenchymal stem cells; pancreatic cancer

INTRODUCTION comes from the tumor necrosis factor (TNF)-related apoptosis- According to 2012 cancer statistics, pancreatic cancer is ranked inducing ligand (TRAIL), a member of the TNF super-family. TRAIL 10th in the United States in terms of new cancer cases and 4th in was originally identified and cloned on the basis of its sequence 19,20 number of cancer deaths.1 Its 5-year survival rate is 4% after homology to the extracellular domain of CD95 ligand and TNF. diagnosis.2,3 At the time of diagnosis, B40% of patients with TRAIL is one of few anticancer proteins that selectively causes pancreatic cancer have metastatic disease, 40--50% have local apoptosis of transformed or tumor cells through the activation of 21 advanced-stage disease that is not amenable to surgery, and only death receptors (DR), with no effects on healthy cells. Five TRAIL B10% of patients can undergo a potentially curative resection. receptors are known: TRAIL receptor 1 (DR4), 2 (DR5), 3 (decoy However, a curative outcome from surgery for pancreatic cancer is receptor 1, DcR1), 4 (DcR2) and a soluble receptor, osteoproteger- 22 rare, with an 80% recurrence rate, and the 5-year survival rate for in. As a component of developing a dual-targeted therapeutic cases involving total resection is only 10--24%.4--6 These statistics strategy, which includes both homing to the tumor site and show that pancreatic cancer is highly malignant and intractable. A delivery of anticancer agents, this study assessed the cytotoxic major obstacle limiting the effectiveness of conventional therapies effects of TRAIL-engineered pancreas-derived MSCs on human for pancreatic cancer, as well as other types of cancer, is the lack of pancreatic cancer cells. tumor specificity. Thus, it is critical to explore therapeutic strategies specifically targeting tumor tissue. Since the discovery of the tumor-oriented homing capacity of MATERIALS AND METHODS mesenchymal stem cells (MSCs), the application of specific Cells and culture conditions anticancer gene-engineered MSCs has held great potential for Human pancreas-derived MSCs (hpMSCs) were isolated from human cancer therapies.7 Tumor-directed migration and incorporation of pancreas ductal tissue and expanded ex vivo as previously described.23,24 MSCs have been demonstrated in a number of preclinical studies Human pancreases were obtained (with consent) from adult heart-beating in vitro using transwell migration assays, and in vivo using animal cadaver organ donors through the organ procurement program at the tumor models. The homing capacity of MSCs has been demon- British Columbia Transplant Society (BCTS, Vancouver, Canada). Pancreatic strated with almost all tested human cancer cell lines, including ductal tissue taken from the Ricordi chamber during islet isolation was lung cancer,8 malignant gliomas,9--11 Kaposi’s sarcomas,12 breast utilized as the starting material. Primary culture was initiated by seeding cancer,13,14 colon carcinoma,15 pancreatic cancer,16,17 melanoma18 chopped ductal tissue onto 100-mm culture dishes (CellBind, Corning, and ovarian cancer.13 Engineered with tumor-specific anticancer Acton, MA, USA). The hpMSCs were cultured in minimum essential medium genes, MSCs are capable of producing anticancer agents locally with 10% fetal bovine serum, 2 mML-glutamine and 1% penicillin-- and constantly at the tumor site. One such anticancer agent streptomycin solution (all from Invitrogen, Carlsbad, CA, USA) and

1Department of Surgery, University of British Columbia, Vancouver, British Columbia, Canada; 2Institute of Transplant Medicine, 303 Hospital of PLA, Nanning, PRC and 3Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, British Columbia, Canada. Correspondence: Dr L-J Dai, 400-828 West 10th Avenue, Department of Surgery, University of British Columbia, Vancouver, BC V5Z 1L8, Canada. E-mail: [email protected] 4These authors contributed equally to this work. Received 22 February 2012; revised 1 June 2012; accepted 6 June 2012; published online 6 July 2012 MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 653 incubated at 37 1C in a humidified, 5% CO2 atmosphere. The cells used in events were recorded and the mean fluorescence intensity of each this study were limited to passages 5--7. Subculture was performed once receptor was assessed on the live-cell population. The reference gating newly grown cells reached subconfluence. On the basis of the minimal location was determined by both isotype IgG and secondary antibody criteria for defining human MSCs, which was established by the controls with the aid of Flowjo software purchased from Tree Star, International Society of Cellular Therapy (ISCT),25 these hpMSCs were (Ashland, OR, USA). verified by both membrane biomarker determination and functional differentiation. They fulfilled the characteristics of human MSCs as descried TRAIL-bearing vector transfection of MSCs in sections ‘Differentiation culture conditions, ORO and AR staining and TRAIL-bearing expression plasmids were used for the transfection, thereby Flow cytometry (FACS) analysis’. avoiding viral vectors in view of potential clinical studies. Two types of The human pancreatic cancer cell lines (BXPC3, ASPC1 and PANC1) were plasmids were used: non-secreting TRAIL (nsTRAIL) and secreting TRAIL purchased from American Type Culture Collection (ATCC, Manassas, VA, (stTRAIL) as previously shown.24 The transfections were performed with USA), and the transformed human islet cell lines, HP62 and TRM6 were TransIT-2020 (Mirus, Madison, WI, USA) as suggested by the manufacturer. gifts from Ou et al. (University of British Columbia, Vancouver, Canada). The The transfected MSCs were labelled as MSCnsTRAIL and MSCstTRAIL, respectively. cell lines were maintained as suggested by ATCC and culture conditions were kept consistent with the hpMSCs. Real-time assessment of cell viability and migration using the xCelligence real-time cell analyzer (RTCA) Differentiation culture conditions 5 The impedance-based RTCA system (Roche) was used for label-free and MSCs were seeded at a density of 1 Â 10 cells per well in 6-well culture real-time monitoring of cell adherence properties, viability and migration. plates that were precoated with poly-L-lysine (ScienCell Research The RTCA system uses specially designed microtiter plates containing À2 Laboratories, Carlsbad, CA, USA) at a concentration of 2 mgcm , for 1 interdigitated gold microelectrodes to noninvasively monitor the viability 1 day before the seeding. The cells were then seeded and incubated (37 C of cultured cells using electrical impedance as the readout.26--28 The and 5% CO2) until they reached 95% confluency with regular media interaction of cells with electrodes impedes the current and generates the changes. The media was then changed to either MSC adipogenic impedance signal, which is displayed as an arbitrary unit referred to as cell differentiation medium (ScienCell Research Laboratories) or the MSC index (CI). CI value is defined as relative change in measured impedance to osteogenic differentiation medium (ScienCell Research Laboratories) with background impedance and represents cell status, and is directly the addition of supplements accordingly. The media was changed every 3 proportional to the quantity, size and attachment forces of the cells. days for the period of 18 days and on day 21, specific staining was performed using Oil Red O (ORO) and Alizarian Red (AR) solutions as Titration of cancer cells. An initial titration of different cell densities described below. (20 000, 10 000, 5000, 2500 cells per well) was performed for each cell line in order to find the ‘optimal’ cell density in this experiment. Cells were ORO and AR staining trypsinized and counted using the trypan blue exclusion method and To detect adipogenesis, 0.5% stock solution of ORO (Polysciences, hemacytometer and then were resuspended in culture medium. Back- Warrington, PA, USA) was combined with a solution of isopropanol ground measurements were taken from the wells by adding 100 ml of the (99.94%). We then mixed three parts of the stock solution with two parts same medium to the E-Plate 16. A volume of 100 ml of cell suspension was phosphate-buffered saline (PBS) to produce the working solution, and then then added to the wells to make a final volume of 200 ml. All cells were let stand for 30 min at room temperature (RT). The working solution was allowed to settle at the bottom of the wells at RT for 15 min, and then were filtered using a 0.45 mm filter and applied to cells using a 0.2 mm filter. The incubated at 37 1C and 5% CO2. The impedance signals were recorded cells were washed twice with PBS. To detect osteogenesis, we used 1% AR every 5 min for the first 25 sweeps (2 h) and every 20 min until the end of

Solution (Polysciences) in double-distilled water (ddH2O), which was pH the experiment (up to 70 h for cell titrations). adjusted to 4.21 with 0.1% ammonium hydroxide. The solution was then filtered into an amber bottle (light protected) with a Whatman #1 filter Cytotoxicity assessment. Following 20 h as described above, different (VWR, Alberta, Canada), and finally it was injected onto the cells using a volumes (50, 100 or 190 ml of 25%, 50% or 100% conditioned medium 0.2 mm filter. Cells were washed before and after staining with ddH2O. Both (CM), respectively) were removed from each well and replaced with the staining phases were followed by a 20 min incubation period at RT and appropriate volume of CM, and the experiment was run for 70 h. The were followed by three washes in PBS (ORO) or ddH2O (AR) accordingly. impedance signals were recorded using the same time intervals for up to The cells were covered with PBS or ddH2O for analysis using a light 20 h. The culture medium was then replaced with CM and a recording microscope. performed every 5 min until the end of the experiment (up to 70 h).

Flow cytometry (FACS) analysis Migration capacity of MSCs. The RTCA system was also applied to monitor cell migration by using cell invasion and migration (CIM) Plates, The cluster of differentiation (CD) marker expression of Phycoerythrin (PE- which contain fenestrated membranes mounted on the upper chamber CD44, CD95, CD105 and CD34)- and Allophycocyanin (APC-CD73)- with interdigitated gold microelectrodes. Cancer cells were prepared as conjugated antibodies and their corresponding isotypes were purchased speciﬁed above but in serum-free media in a total volume of 160 ml. Cells from eBiosciences (San Diego, CA, USA) and used according to the were then seeded in the lower chamber of plates and left to settle for manufacturer’s protocol. TRAIL receptor expression on pancreatic cancer 30 min at RT in sterile conditions. The upper chamber was then mounted cells was detected using FACS. Subconﬂuent cells were detached with and 25 ml of corresponding media was added to each well and left to 0.25% trypsin-EDTA and washed with PBS. A total of 1--5 Â 105 cells were equilibrate the RTCA system for 1 h at 37 1C and 5% CO . After this resuspended in 200 ml PBS for each reaction, and then 10 ml of primary 2 incubation, a background reading was taken for each well. As per the antibody solution (25 mg/ml) (antibodies to human DR4, DR5, DcR1, DcR2 manufacturer’s protocol, 40 000 MSCs were prepared in serum-free media or isotype IgG, R&D Systems, Minneapolis, MN, USA) was added to each and plated into the upper chamber of the CIM plates to a total volume of eppendorf tube and incubated at 4 1C for 30 min. The cells were washed 160 ml. Readings were recorded initially at 25 sweeps every 5 min and then three times with PBS and resuspended in 200 ml PBS, and the followed every 10 min until the end of the experiment (up to 24 h). phycoerythrin-conjugated secondary antibody (R&D Systems) was added to each reaction. The cells were incubated for 30 min at 4 1C under darkened conditions and again washed three times with PBS. The cells Statistical analysis were resuspended with 500 ml PBS for FACS analysis (FACSCalibur, Becton Cell index calculations for real-time dynamic cytotoxicity assessments Dickinson, Heidelberg, Germany). For all of the FACS experiments, 15 000 (n ¼ 3) and slope calculations for the migration assessments as well as

& 2012 Nature America, Inc. Cancer Gene Therapy (2012), 652 -- 658 MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 654 normalizations, were performed automatically by the RTCA software capacity of MSCs to adipogenic and osteogenic cells. Lastly package 1.2 for the xCelligence system, the area under curve was Figure 1c shows the biomarker characterization of MSCs using calculated using the trapezoid rule. Data are expressed as mean±s.e. or FACS, exhibiting positive expression of CD44 þ , CD73 þ , CD95 þ , mean±s.d.. Statistical differences between means for the different groups CD105 þ and negative of CD34À. These MSCs fulfilled the criteria were evaluated with Prism 5.0 (GraphPad software, La Jolla, CA, USA) using for defining human MSCs established by ISCT.25 either Student’s t-test or one-way ANOVA with Bonferroni’s correction. Level of significance was set at Po0.05. TRAIL receptor expression in pancreatic cancer cells Figure 2a displays the expression of four TRAIL receptors on all tested pancreatic cancer cells. The percentage of positive cell RESULTS population for each individual receptor is summarized in Pancreas-derived MSCs’ isolation and characterization Figure 2b. These results enable us to examine the type of Figure 1a displays the isolation and propagation of our pancreas- responses observed from each cell type after CM treatments in derived MSCs. Figure 1b depicts the positive differentiation response to their DR expression.

Figure 1. Pancreatic ductal-derived mesenchymal stem cell’s (MSC’s) isolation and characterization. (a) Cell propagation and isolation. (1) Purified ductal tissue is shown. (2) MSCs grew out of the fragmented ductal tissue on day 3 and became subconfluence on day 7. (3) MSCs reached confluence within 3 days after P0. (b) Differentiation capacity of MSCs. (1) Control. (2) Oil Red O (ORO) staining for the adipogenic cells. (3) Alizarian Red (AR) staining for the osteogenic cells. (c) Biomarker characterization using flow cytometry (FACS). Black lines, cell population; blue lines, isotype IgG; red plottings, anti-human CD IgG. From left to right: top row CD105, 73, and 90 bottom row CD34, and 44.

Cancer Gene Therapy (2012), 652 -- 658 & 2012 Nature America, Inc. MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 655

Figure 2. Expression of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) receptors in pancreatic cancer cells. (a) Flow cyto- metric analysis of cell surface expression of DR4, DR5, DcR1 and DcR2 on BXPC3, TRM6, ASPC1, HP62 and Panc1 cells. Red lines represent the reference gates that were set according to isotype IgG and secondary antibody controls. The graph represents one of the three parallel assessments. (b) Summary of TRAIL receptor expression in pancreatic cancer cells.

Dynamic observation of MSC migration toward pancreatic cancer cells The direct co-culture studies showed that most dead cancer cells were observed around MSCs (data not shown). This phenomenon is presumably attributed to MSC’s tumor-tropic capacities. This view- point could be, at least partially if not totally, confirmed by the use of RTCA technology. Figure 4a displays the dynamic migration of MSCs toward each tested pancreatic cancer cell line. MSCs showed remarkable chemotaxis migration to pancreatic cancer cells compared with normal human cells. The most active cell migration was observed from the conditioned media of cancer cells during the first 3 h of the experiment. The chemotactic effects of CM from each cell type are shown in Figure 4b, which is presented as slope (dc/dt) value of migration; the bigger the slope of the real-time curve, the higher the rate of migration. Conditioned media from all tested pancreatic cancer cells induced significantly higher rates of MSC migration compared with regular medium control.

Real-time detection of CM-induced cytotoxicity Figure 3. Cell viability profiles of pancreatic cancer cells. Cell index For these experiments, it was vital to ensure the appropriate cell was automatically recorded with real-time cell analyzer (RTCA) every density and time point of treatment to avoid false-positive or 20 min until the end of the experiment (70 h). Each tracing was false-negative results. In the titration experiments, as introduced collected from the titration test for each individual cell types re- in the section ‘Migration capacity of MSCs’, four densities of each presenting an average of three parallel assessments. cell type were determined. On the basis of the results of titration test, the applied cell density was set at 20 000 cells per well, and 20 h was determined as the time point for the treatment, which is Cell viability profiles of pancreatic cancer cells in the middle of the exponentially increasing portion of the CI Figure 3 demonstrates the growth profile of the cancer cells at (Figure 3). Figure 5a presents the real-time dynamic changes of CI 20 000 cells per well. Different densities (2500, 5000, 10 000, 20 000 with the application of various CM concentrations for all studied cells per well) were assessed to determine the ‘optimal’ testing cell lines. The corresponding calculation of area under curve for conditions for each cell type, only the last density (20 000 cells per these types of cells was summarized in Figure 5b where it is well) is shown in Figure 3. It is worth noting that this approach not shown that ASPC1, TRM6 and HP62 cells displayed significant only explains the binding behavior and growth patterns of each reduction in cell viability of CM-treated groups compared with cell type having HP62 being the most aggressive and ASPC1 the controls. There was no significant inhibition of BXPC3 cell viability least, but also enables us to verify a time point for cytotoxicity during the observation period. However, these cells responded at assessment where no false-positive can be observed from over- a later time beginning at approximately 50 h after application of populated wells. CM (data not shown).

& 2012 Nature America, Inc. Cancer Gene Therapy (2012), 652 -- 658 MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 656 Recent studies have demonstrated TRAIL-induced apoptosis in pancreatic cancer cells.32,33 Increased number of MSC-mediated anticancer studies have shown MSCs to be an effective therapeutic tool and gene carrier for the treatment of pancreatic cancer.34 To use MSCs as effective anticancer agent vehicles, we need a viable and practical source of MSCs and a method to successfully transfect MSCs with anticancer genes. To this end, we have successfully isolated and characterized MSCs from the ductal tissue of human donor pancreases and as well as, we have successfully transfected and expressed TRAIL in MSCs. Further- more, the presence of DRs in cancer cells is a prerequisite determinant for TRAIL to induce apoptosis in targeted cancer cells.35 However, the presence of DRs does not necessarily indicate the sensitivity of tumor cells to TRAIL and functional assessments are needed toward developing MSC-mediated targeted therapeutic strategies. It is essential to understand that many possible factors could affect the interactions between MSCs and tumor cells. These factors include the forms of transfected anticancer genes, required MSC numbers (ratio), direct and/or indirect effects, as well as the effects of native MSCs. As previously shown,24 native MSC-induced moderate cell death was observed in the low MSC/HepG2 ratio co-cultures, whereas marked cell death was exhibited in the high MSC/HepG2 ratio co- cultures. These results indicate that MSCs alone could have an important role in HepG2 cell death. In the present direct co-culture studies, the majority of dead cells were observed around MSC occupied areas, consistent with our previous findings with liver cancer cells.24 The innate antitumor effects of MSCs have been previously recognized in experimental Kaposi’s sarcoma,12 hepa- tocellular carcinoma,36 and colorectal carcinoma37 studies. The direct effects of MSCs on tumor cell viability are mainly attributed to their intrinsic antitumor properties.7 Our data suggest that cancer cell-induced chemotactic effects on MSCs have an Figure 4. Migration kinetics of mesenchymal stem cells (MSCs). (a) important role in the cell-to-cell interactions. Related chemotactic Cell-directed migration MSCs (40 000 per well) were seeded into factors are to be identified in forthcoming studies. insert wells on CIM-Plate 16 and pancreatic cancer cells (20 000 per MSCs transfected with the secreting form of TRAIL may be more well) were plated at the bottom of the transwell. The cell index was automatically recorded with real-time cell analyzer (RTCA) every efficient than the MSCs transfected with the non-secreting form of 20 min for 30 h. Results shown are mean±s.d. (n ¼ 3). Human TRAIL in terms of inhibiting tumor cell growth. Our indirect co- fibroblast cells (A1F8) were assessed as a reference of normal cells. culture studies support this point, as CM-induced inhibition of (b) The chemotactic effects of cancer cells conditioned medium ASPC1 cell viability was observed at two different concentrations (CM) on MSC migration. MSCs (40 000 per well) were seeded into of 50 and 100% CM, and TRM6 and HP62 cells in only 100% CM insert wells on CIM-Plate 16 and regular medium (RM) or CM from during the period of observation. Compared with CM from various pancreatic cancer cells were plated at the bottom of the MSCnsTRAIL, CM from MSCstTRAIL showed more potent inhibition at transwell. The MSC migration rate (Slope, dc/dt) was calculated from high concentrations on these cell types. Viability of BXPC3 cells the first 3 h segment of the migration tracings. Results shown are was not affected during the same time course, but viability did mean±s.d. (n ¼ 3). *P 0.05 vs RM. o decrease at a later time point (data not shown). Both cell titration and cytotoxicity data for the BXPC3 cells indicate that they would need a higher density of cells due to their slow pattern of growth, DISCUSSION with a doubling time of almost 1 week. There was no difference TRAIL and its receptors are important components of the extrinsic between the two concentrations of CM on cytotoxicity on PANC1 pathway of apoptosis. A death domain in the intracellular region cells, whereas these cells showed more sensitivity to native MSC- of DR4 and DR5 can recruit a death-inducing signaling complex derived medium compared with other cell types. Moreover, upon TRAIL stimulation, activating a downstream caspase cascade growth spurts were observed in the lower CM conditions for both leading to cell death by apoptosis. No intact death domain is BXPC3 and HP62 cells. Further in vivo studies are in process to present in the intracellular region of DcR1, DcR2 and osteopro- investigate the therapeutic potential of engineered MSCs in a tegerin, so they are unable to induce apoptosis, even though they physiologically relevant microenvironment for pancreatic cancer. could compete with DR4 or DR5 for binding with TRAIL,29 and TRAIL-based therapy is highly dependent on DR expression of overexpression of DcR1 and/or DcR2 blocks TRAIL-mediated tumor cells. TRAIL receptor expression profiles in pancreatic apoptosis in some cell types.30 The DRs (DR4 and DR5) are cancer patients may predict the feasibility of using TRAIL gene frequently expressed in cancer and transformed cells.21,31 In the therapy as a treatment option. However, considerable numbers of present study, the highly expressed DR5 was identified in human cancer cells, especially some highly malignant tumors, are pancreatic adenocarcinoma (BXPC3 and ASPC1), insulinoma resistant to apoptosis induction by TRAIL, and some cancer cells (HP62) and SV40 viral DNA-transformed human islet cell line that were originally sensitive to TRAIL-induced apoptosis can (TRM6). However, human pancreatic epithelioid carcinoma cell line become resistant after repeated exposure.38--40 Mueller et al.41 (PANC1) showed remarkably low expression of DRs, which is in recently demonstrated that TRAIL-transduced MSCs overcome contrast to a recent report.31 These cell lines with different TRAIL resistance in selected colorectal carcinoma cell lines. patterns of TRAIL receptor expression can serve as constructive Therefore, the anticancer properties of TRAIL-engineered MSCs tools for other TRAIL-related studies. potentially involves three aspects: MSC’s intrinsic anticancer effect,

Cancer Gene Therapy (2012), 652 -- 658 & 2012 Nature America, Inc. MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 657

Figure 5. Conditioned medium (CM)-induced cytotoxicity on pancreatic cancer cells. (a) Real-time monitoring of conditioned media-induced cytotoxicity in pancreatic cancer cells. Cells (20 000 per well) were seeded in E-Plate 16 with normal minimum essential medium (MEM) medium. Cell toxicity was monitored using the real-time cell analyzer (RTCA) instrument. After B20 h, culture medium was replaced with CM from MSCstTRAIL or CM from MSCnsTRAIL. The tracings represent the real-time dynamic changes of cell index with the application of various CM concentrations for BXPC3, ASPC1, TRM6, HP62 and PANC1 cells. Arrows indicate the time points of CM replacement and the density of the colors represents the different CM concentrations, i.e. 100, 50 and 25%. (b) Area under curve (AUC) analysis. The AUC was calculated using Trapezoid Rule. The controls show the values obtained from the treatment of CM from native mesenchymal stem cells (MSCs) and normalized with the values from regular normal medium. Black lines represents the comparisons of regular MSC CM with MSCstTRAIL or MSCnsTRAIL CM. Red lines represent the comparison of the same concentration of CM between the MSCstTRAIL or MSCnsTRAIL. Results shown are mean±s.d. (n ¼ 3). *Po0.05, **Po0.01, ***Po0.001.

TRAIL-induced speciﬁc tumor cell death and induction of tumor treatment. hpMSCs could serve as an allogeneic cell source when cell’s TRAIL sensitivity. Combined with the use of real-time autologous MSCs are not available for MSC-mediated cancer detection of cytoxicity, the most effective anticancer genes can therapy especially for the treatment of pancreatic cancer. be identiﬁed from different anticancer gene-engineered MSCs In conclusion, the intrinsic properties of MSCs have an within a short period of time. As outlined in our recent review,42 important role in the induction of pancreatic cancer cell death the present study provides a useful platform toward the under co-culture conditions. MSC-mediated cell death is poten- development of practically personalized therapy for cancer tiated by TRAIL gene transfection on DR-expressing pancreatic

& 2012 Nature America, Inc. Cancer Gene Therapy (2012), 652 -- 658 MSC-mediated pancreatic cancer cytotherapy MR Moniri et al 658 cancer cells. The impedance-based RTCA system is a very reliable 18 Studeny M, Marini FC, Champlin RE, Zompetta C, Fidler IJ, Andreeff M. Bone technology in continuously evaluating cell viability and cell marrow-derived mesenchymal stem cells as vehicles for interferon-beta delivery migration in real time. This study provides a practical platform into tumors. Cancer Res 2002; 62: 3603--3608. for the development of MSC-based personalized treatments for 19 Pitti RM, Marsters SA, Ruppert S, Donahue CJ, Moore A, Ashkenazi A. Induction of pancreatic cancer. apoptosis by Apo-2 ligand, a new member of the tumor necrosis factor cytokine family. J Biol Chem 1996; 271: 12687--12690. 20 Wiley SR, Schooley K, Smolak PJ, Din WS, Huang CP, Nicholl JK et al. Identification CONFLICT OF INTEREST and characterization of a new member of the TNF family that induces apoptosis. The authors declare no conflict of interest. Immunity 1995; 3: 673--682. 21 Wu GS. TRAIL as a target in anti-cancer therapy. Cancer Lett 2009; 285:1--5. 22 Zhu DM, Shi J, Liu S, Liu Y, Zheng D. HIV infection enhances TRAIL-induced cell ACKNOWLEDGEMENTS death in macrophage by down-regulating decoy receptor expression and generation of reactive oxygen species. PloS One 2011; 6: e18291. This work was supported by the JDRF, the VGH and UBC Hospital Foundation and the 23 Zhang ZX, Guan LX, Zhang K, Zhang Q, Dai LJ. A combined procedure to deliver Guangxi Ministry of Science and Technology. We also would like to extend our autologous mesenchymal stromal cells to patients with traumatic brain injury. gratitude to Dr E LeBlanc, Ms K Frewin, Dr Melissa Brierley and Jeffrey Helm, who Cytotherapy 2008; 10: 134--139. have been tremendous resources throughout our work. We are grateful for Mrs 24 Sun XY, Nong J, Qin K, Lu H, Moniri MR, Dai LJ et al. MSCTRAIL-mediated HepG2 Crystal Robertson’s assistance for manuscript editing. cell death in direct and indirect co-cultures. Anticancer Res 2011; 31: 3705--3712. 25 Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini F, Krause D et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The REFERENCES International Society for Cellular Therapy position statement. Cytotherapy 2006; 1 Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the impact of 8: 315--317. eliminating socioeconomic and racial disparities on premature cancer deaths. CA 26 Diemert S, Dolga AM, Tobaben S, Grohm J, Pfeifer S, Oexler E et al. Impedance Cancer J Clin 2011; 61: 212--236. measurement for real time detection of neuronal cell death. J Neurosci Methods 2 Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA: A Cancer J Clin 2012; 2012; 203: 69--77. 62: 10--29. 27 Ke N, Wang X, Xu X, Abassi YA. The xCELLigence system for real-time and label- 3 Jemal A, Tiwari RC, Murray T, Ghafoor A, Samuels A, Ward E et al. Cancer statistics, free monitoring of cell viability. Methods Mol Biol 2011; 740: 33--43. 2004. CA Cancer J Clin 2004; 54: 8--29. 28 Stander XX, Stander BA, Joubert AM. In vitro effects of an in silico-modelled 4 Warshaw AL, Gu ZY, Wittenberg J, Waltman AC. Preoperative staging and 17beta-estradiol derivative in combination with dichloroacetic acid on MCF-7 and assessment of resectability of pancreatic cancer. Arch Surg 1990; 125: 230--233. MCF-12A cells. Cell Prolif 2011; 44: 567--581. 5 Conlon KC, Klimstra DS, Brennan MF. Long-term survival after curative resection 29 Pan G, O0Rourke K, Chinnaiyan AM, Gentz R, Ebner R, Ni J et al. The receptor for for pancreatic ductal adenocarcinoma. Clinicopathologic analysis of 5-year the cytotoxic ligand TRAIL. Science 1997; 276: 111--113. survivors. Ann Surg 1996; 223: 273--279. 30 Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, Baldwin D et al. Control 6 Yeo CJ, Cameron JL, Lillemoe KD, Sohn TA, Campbell KA, Sauter PK et al. of TRAIL-induced apoptosis by a family of signaling and decoy receptors. Science Pancreaticoduodenectomy with or without distal gastrectomy and extended 1997; 277: 818--821. retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: 31 Stadel D, Mohr A, Ref C, MacFarlane M, Zhou S, Humphreys R et al. TRAIL-induced randomized controlled trial evaluating survival, morbidity, and mortality. Ann Surg apoptosis is preferentially mediated via TRAIL receptor 1 in pancreatic carcinoma 2002; 236: 355--366 discussion 366--368. cells and profoundly enhanced by XIAP inhibitors. Clin Cancer Res 2010; 16: 7 Dai LJ, Moniri MR, Zeng ZR, Zhou JX, Rayat J, Warnock GL. Potential implications 5734--5749. of mesenchymal stem cells in cancer therapy. Cancer Lett 2011; 305:8--20. 32 Jacob D, Davis JJ, Zhang L, Zhu H, Teraishi F, Fang B. Suppression of pancreatic 8 Loebinger MR, Kyrtatos PG, Turmaine M, Price AN, Pankhurst Q, Lythgoe MF et al. tumor growth in the liver by systemic administration of the TRAIL gene driven by Magnetic resonance imaging of mesenchymal stem cells homing to pulmonary the hTERT promoter. Cancer Gene Ther 2005; 12: 109--115. metastases using biocompatible magnetic nanoparticles. Cancer Res 2009; 69: 33 Lemke J, Noack A, Adam D, Tchikov V, Bertsch U, Roder C et al. TRAIL signaling is 8862--8867. mediated by DR4 in pancreatic tumor cells despite the expression of functional 9 Sasportas LS, Kasmieh R, Wakimoto H, Hingtgen S, van de Water JA, Mohapatra G DR5. J Mol Med 2010; 88: 729--740. et al. Assessment of therapeutic efficacy and fate of engineered human 34 Mohr A, Albarenque SM, Deedigan L, Yu R, Reidy M, Fulda S et al. Targeting of mesenchymal stem cells for cancer therapy. Proc Natl Acad Sci USA 2009; 106: XIAP combined with systemic mesenchymal stem cell-mediated delivery of 4822--4827. sTRAIL ligand inhibits metastatic growth of pancreatic carcinoma cells. Stem Cells 10 Sonabend AM, Ulasov IV, Tyler MA, Rivera AA, Mathis JM, Lesniak MS. 2010; 28: 2109--2120. Mesenchymal stem cells effectively deliver an oncolytic adenovirus to intracranial 35 Johnstone RW, Frew AJ, Smyth MJ. The TRAIL apoptotic pathway in cancer onset, glioma. Stem Cells 2008; 26: 831--841. progression and therapy. Nat Rev Cancer 2008; 8: 782--798. 11 Yang B, Wu X, Mao Y, Bao W, Gao L, Zhou P et al. Dual-targeted antitumor effects 36 Qiao L, Xu Z, Zhao T, Zhao Z, Shi M, Zhao RC et al. Suppression of tumorigenesis by against brainstem glioma by intravenous delivery of tumor necrosis factor- human mesenchymal stem cells in a hepatoma model. Cell Res 2008; 18: 500--507. related, apoptosis-inducing, ligand-engineered human mesenchymal stem cells. 37 Luetzkendorf J, Mueller LP, Mueller T, Caysa H, Nerger K, Schmoll HJ. Growth Neurosurgery 2009; 65: 610--624 discussion 624. inhibition of colorectal carcinoma by lentiviral TRAIL-transgenic human mesench- 12 Khakoo AY, Pati S, Anderson SA, Reid W, Elshal MF, Rovira II et al. Human ymal stem cells requires their substantial intratumoral presence. J Cell Mol Med mesenchymal stem cells exert potent antitumorigenic effects in a model of 2010; 14: 2292--2304. Kaposi0s sarcoma. J Exp Med 2006; 203: 1235--1247. 38 Zhang L, Fang B. Mechanisms of resistance to TRAIL-induced apoptosis in cancer. 13 Kidd S, Spaeth E, Dembinski JL, Dietrich M, Watson K, Klopp A et al. Direct evidence Cancer Gene Ther 2005; 12: 228--237. of mesenchymal stem cell tropism for tumor and wounding microenvironments 39 Ibrahim SM, Ringel J, Schmidt C, Ringel B, Muller P, Koczan D et al. Pancreatic using in vivo bioluminescent imaging. Stem Cells 2009; 27: 2614--2623. adenocarcinoma cell lines show variable susceptibility to TRAIL-mediated cell 14 Patel SA, Meyer JR, Greco SJ, Corcoran KE, Bryan M, Rameshwar P. Mesenchymal death. Pancreas 2001; 23: 72--79. stem cells protect breast cancer cells through regulatory T cells: role of 40 Sanlioglu AD, Dirice E, Elpek O, Korcum AF, Ozdogan M, Suleymanlar I et al. High mesenchymal stem cell-derived TGF-beta. J Immunol 2010; 184: 5885--5894. TRAIL death receptor 4 and decoy receptor 2 expression correlates with 15 Menon LG, Picinich S, Koneru R, Gao H, Lin SY, Koneru M et al. Differential gene significant cell death in pancreatic ductal adenocarcinoma patients. Pancreas expression associated with migration of mesenchymal stem cells to conditioned 2009; 38: 154--160. medium from tumor cells or bone marrow cells. Stem Cells 2007; 25: 520--528. 41 Mueller LP, Luetzkendorf J, Widder M, Nerger K, Caysa H, Mueller T. TRAIL- 16 Kidd S, Caldwell L, Dietrich M, Samudio I, Spaeth EL, Watson K et al. Mesenchymal transduced multipotent mesenchymal stromal cells (TRAIL-MSC) overcome TRAIL stromal cells alone or expressing interferon-beta suppress pancreatic tumors in vivo, resistance in selected CRC cell lines in vitro and in vivo. Cancer Gene Ther 2011; 18: an effect countered by anti-inflammatory treatment. Cytotherapy 2010; 12: 615--625. 229--239. 17 Zischek C, Niess H, Ischenko I, Conrad C, Huss R, Jauch KW et al. Targeting tumor 42 Sun XY, Nong J, Qin K, Warnock GL, Dai LJ. Mesenchymal stem cell-mediated stroma using engineered mesenchymal stem cells reduces the growth of cancer therapy: a dual-targeted strategy of personalized medicine. World J Stem pancreatic carcinoma. Ann Surg 2009; 250: 747--753. Cells 2011; 3: 96--103.

Cancer Gene Therapy (2012), 652 -- 658 & 2012 Nature America, Inc.