Mesenchymal stem cells expressing TRAIL lead to tumour growth inhibition in an experimental lung cancer model

Abstract

Lung cancer is a major public health problem in the western world, and gene therapy strategies to tackle this disease systemically are often impaired by inefficient delivery of the vector to the tumour tissue. Some of the main factors inhibiting systemic delivery are found in the blood stream in the form of red and white blood cells (WBCs) and serum components. Mesenchymal stem cells (MSCs) have been shown to home to tumour sites and could potentially act as a shield and vehicle for a tumouricidal gene therapy vector. Here, we describe the ability of an adenoviral vector expressing TRAIL (Ad.TR) to transduce MSCs and show the apoptosis-inducing activity of these TRAIL-carrying MSCs on A549 lung carcinoma cells. Intriguingly, using MSCs transduced with Ad.enhanced-green-fluorescent-protein (EGFP) we could show transfer of viral DNA to cocultured A549 cells resulting in transgenic protein production in these cells, which was not inhibited by exposure of MSCs to human serum containing high levels of adenovirus neutralizing antibodies. Furthermore, Ad.TR-transduced MSCs were shown not to induce T-cell proliferation, which may have resulted in cytotoxic T-cell-mediated apoptosis induction in the Ad.TR-transduced MSCs. Apoptosis was also induced in A549 cells by Ad.TR-transduced MSCs in the presence of physiological concentrations of WBC, erythrocytes and sera from human donors that inhibit or neutralize adenovirus alone. Moreover, we could show tumour growth reduction with TRAIL-loaded MSCs in an A549 xenograft mouse model. This is the first study that demonstrates the potential therapeutic utility of Ad.TR-transduced MSCs in cancer cells and the stability of this vector in the context of the blood environment.

Fig. 1

Adenovirally expressed TRAIL induces higher levels of apoptosis than recombinant TRAIL in A549 cells. (A) Replication-defective adenoviral vectors with deletions of the E1 and E3 regions (Ad. ΔE1 and Ad. ΔE3) were used in our study. They expressed TRAIL (Ad.TR), β-galactosidase (Ad.BGal) or EGFP (Ad.EGFP) from the cytomegalovirus (CMV) promoter (P-CMV). Other elements are the 5′ and 3′ inverted-terminal-repeats (ITR) and the thymidine kinase polyA-addition signal (TKpA). (B) A549 cells were seeded at 10⁵ cells per well overnight. Adenovirus (expressing TRAIL or β-galactosidase, Ad.TR and Ad.BGal, respectively) was added at 1, 10 or 100 pfu/cell or recombinant TRAIL (rTR) was added at 1, 5 or 10 ng/ml. As positive control for the effects of rTR human colorectal cancer HCT116 cells were treated with 10 ng/ml rTR. Cells were assayed for apoptosis after 48 hrs by Nicoletti hypodiploidity assay. Samples were analysed by flow cytometry. Numbers represent mean values of five samples ± standard deviation. **P < 0.001.

Fig. 2

Transgene expression after adenovirus transduction of MSCs. (A) MSCs were transduced with Ad.EGFP and Ad.TR at 100 pfu and were analysed by fluorescent and phase contrast microscopy 48 hrs later. The images show that green fluorescent expression was established in Ad.EGFP-transduced cells. Transduction with Ad.EGFP and Ad.TR did not alter the morphology of MSCs as compared to untransduced cells. (B) EGFP transgene expression was also quantified by FACS analysis. MSCs were transduced with Ad.EGFP at 100 pfu/cell. Almost all cells were transduced after 48 hrs. (C) Cell lysates from MSCs transduced with Ad.BGal or Ad.TR were probed for TRAIL expression by Western blotting and compared to signals from rTR that was loaded at 5, 20 and 30 ng on the same gel. The Ad.TR-encoded full-length membrane-bound form runs at 30 kD, whereas the soluble rTR, representing only the ectodomain of TRAIL, runs at 20 kD. A CuZnSOD Western blot serves as loading control for the protein lysates from Ad.BGal and Ad.TR-transduced cells (D) Neither treatment with rTR nor transduction with Ad.TRAIL causes apoptosis in MSCs. MSCs were either treated with rTR at concentration of 10 ng or transduced with Ad.TRAIL (100 pfu/cells) or Ad.BGal (100 pfu/cells). Specific apoptosis was measured 48 hrs later by Nicoletti hypodiploidity assay. HCT116 cells (white bar) treated with rTR serve as positive control.

Fig. 3

MSCs transduced with Ad.TR can induce apoptosis in A549 cells. (A) MSCs were transduced with Ad.BGal (100 pfu/cell) or Ad.TR (100 pfu/cell). After 48 hrs MSCs were trypsinized and 10⁴ of these cells were added to wells that had been seeded the day before with A549 cells at 10 ⁵ cells per well. After 48 hrs, wells were trypsinized and the complete mixed cell population was measured by Nicoletti apoptosis assay. A549 cells and MSCs cultured alone as well as A549 cells mixed with MSCs transduced with Ad.BGal showed only background apoptosis of below 5%. In contrast, in A549 mixed with MSCs transduced with Ad.TR apoptosis rates of almost 30% could be measured. This apoptosis could be inhibited by TRAIL neutralizing antibodies (α-TR-ab) and the pan-caspase inhibitor zVAD. Numbers represent mean values of three samples ± standard deviation. **P < 0.001. (B) MSCs cocultured with A549 cells represent approximately 1% of the final cell population. 10⁴ untransduced or Ad.BGal transduced (100 pfu/cell) MSC cells were added to wells that were unseeded or had been seeded the previous day with 10⁵ A549 cells. After 48 hrs, cells were detached from wells using trypsin free buffer (Specialty Media, NJ, USA). Cells were stained with anti-CD105 and analysed by flow cytometry. The isotype control for a 1:10 MSC:A549 mix (MSC+A549 ISO) was, as expected, negative for CD105, whereas 98% of all cells in a pure MSC culture (MSC) were CD105 positive. A549 cells only (A549) showed a weak CD105 signal in 0.5% of all cells. In 1:10 mixes with MSCs (MSC+A549) or MSCs transduced with Ad.BGal (A549+MSCβGal) the CD105 signal increased to about 1.5% 48 hrs after mixing and coculturing. Please note that graph is depicted in log-scale. (C) The experiment was performed as described above (Fig. 3A), however, untransduced and transduced (Ad.BGal and Ad.TR) MSC cells were added at 10⁴ (1:10) (black bars) and 3.3 × 10³ (1:30) (white bars) cells per well. After 48 hrs, cells were analysed by Nicoletti apoptosis assay proving that even at a lower ratio against the target cancer cells, MSCs expressing TRAIL can trigger significant levels of apoptosis. Numbers represent mean values of 12 samples ± standard deviation. **P < 0.001. (D) MSCs were transduced with Ad.EGFP (100 pfu/cell) (black bars) or Ad.TR (100 pfu/cell) (white bars). After 48 hrs, MSCs were harvested and 10⁴ of these cells were mixed with 10⁵ pre-seeded primary human fibroblasts (Fib). After 48 hrs, apoptosis was determined by Nicoletti apoptosis assay. HCT116 cells (HCT) served as positive controls in this experiment.

Fig. 4

Ad.EGFP-transduced MSCs transfer adenoviral-encoded transgene expression onto A549 cells. (A) MSC cells were transduced with Ad.EGFP at 10, 100 or 1000 pfu/cell. After 48 hrs MSCs were trypsinized and prepared as previously described. Before addition to A549 seeded wells the Ad.EGFP-transduced MSC cell suspension was centrifuged. The supernatant was removed from the cell pellet and also added to A549 cells (white bars, which are hardly visible except for 1000 pfu). This resulted in almost no transduction, indicating that no virus was left in the supernatant medium or produced in the MSCs. The MSCs themselves were re-suspended in a volume of medium equivalent to that of the aspirated supernatant. 10⁴ Ad.EGFP-transduced MSCs (black bars) were added to wells seeded with A549 cells. Samples were assayed for EGFP expression by flow cytometry after a further 48 hrs. (B) Adenovirus-transduced MSC cocultured with 293 cells results in novel virus genome production. MSCs were transduced with Ad.BGal (100 pfu/cell) or left untreated. After 48 hrs, untransduced (MSC+293) and Ad-transduced MSCs (MSC Ad.+293) were cocultured with 293 cells that had been seeded the previous day at 10⁵ cells/well or Ad.BGal-transduced MSCs were replated into wells containing no 293 cells (MSC Ad.). After 24 hrs, cells were harvested and DNA was extracted. Cells were analysed for viral DNA content by quantitative PCR using primers specific for the knob turn region of the virus and the SYBR Green quantitative PCR system. Samples were run on an ABI Prism 7000 sequence detector. Please note that the graph is depicted in log-scale. (C) In order to rule out that our MSCs expressed adenoviral E1A and could facilitate virus replication, we performed a Western blot on protein lysates from MSCs, MSCs transduced with Ad.EGFP (0, 48, 72 and 96 hrs after transduction) and 293 cells as positive control. Only 293 cells showed a signal for E1A. A CuZnSOD Western blot was carried out as loading control.

Fig. 5

Transduction of MSCs with adenovirus does not induce T-cell proliferation. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll gradient from blood samples taken from a healthy donor. PBMCs were then stained with carboxyfluorescein diacetate succin-imidyl ester (CFSE) dye and left overnight in PBMC culture medium. MSCs and fibroblasts were untransduced or transduced with Ad.BGal (100 pfu/cell) or Ad.TR (100 pfu/cell) (MSC, MSCβGal, MSCTR and Fib, FibβGal, FibTR) and harvested after 48 hrs. 10⁴ MSCs were cocultured with 10⁵ PBMCs for 5 days. As a positive control, PBMCs were also stimulated by the addition of 2.5 μl CD3/CD28 beads (Dynal, Invitrogen). Cells were then collected and analysed by flow cytometry for cell proliferation, gating on the lymphocyte population. Please, note that graph is depicted in log-scale.

Fig. 6

Ad.TR-transduced MSCs induce apoptosis in A549 cells in the presence of washed blood (erythrocytes) and white blood cells at physiological concentrations. MSCs were untransduced or transduced with Ad.BGal (100 pfu/cell) or Ad.TR (100 pfu/cell) (MSC, MSCβGal, MSCTR) and harvested after 48 hrs. A549 cells were stained with cell tracker green and seeded into wells at 10⁵ cells per well. Once these cells had adhered to the tissue culture wells (A) white blood cells were added at a concentration of 3 × 10⁶ cells/ml or (B) washed blood cells were added at approximately 5 × 10⁹ cells/ml. 10⁴ untransduced or Ad-transduced MSC cells were added to each well. After 48 hrs, washed blood cells (A) or white blood cells (B) and medium alone containing wells were harvested and prepared for Nicoletti assay. The wells containing washed blood cells were harvested and the red blood cells were lysed by exposure to a hypotonic solution before these samples were prepared for Nicoletti assay. All samples were assayed by flow cytometry gating on the cell tracker green positive population alone. Overall, the level of apoptosis in white blood cell and washed blood exposed wells was increased, however, samples exposed to Ad.TR-transduced MSCs were consistently higher than background or controls. Numbers represent mean values of three experiments ± standard deviation. *P < 0.05; **P < 0.001.

Fig. 7

Comparison of neutralization properties of different human sera. (A) 10⁷ Ad.EGFP pfu were incubated in complete RPMI medium (white bar) or human serum (black bars) diluted 1/10 or 1/100 in complete RPMI medium for 30 min. at 37°C. Data shown are samples with high (Serum A) and low (Serum B) adenovirus-neutralizing activities chosen from a panel of 18 donors. After pre-incubation of Ad.EGFP with medium and the two sera, the samples were added to wells containing A549 cells seeded the previous day at 10⁵ cells per well. After 48 hrs, the cells were harvested and analysed for EGFP expression by flow cytometry. While A549 cells are readily transduced with Ad.EGFP pre-incubated in RPMI medium (white bar), pre-incubation with Serum A at both concentrations completely neutralizes the transduction activity of the adenoviral vector. Serum B, on the other side allows transduction of A549 cells after pre-incubation Numbers represent mean values of duplicate experiments. (B) Exposing MSCs transduced with Ad.EGFP to neat human serum with high levels of adenovirus neutralizing antibodies (Serum A) does not inhibit the ability of these cells to ‘post-transduce’ A549 cells. MSC cells were transduced with Ad.EGFP (100 pfu/cell). After 48 hrs, the cells were harvested and incubated in neat serum or serum diluted 1:10 or 1:100 in complete DMEM medium for 30 min. at 37°C. Samples were prepared using serum from a high neutralizing donor (Serum A; white bars) and a low-neutralizing donor (Serum B; black bars). 10⁴ MSCs including the serum were then added to wells containing A549 cells seeded the previous day. After 48 hrs, samples were prepared for analysis by flow cytometry. This analysis shows that A549 cells were transduced by the Ad.EGFP-loaded MSCs regardless of the concentration of adenovirus-neutralizing antibodies. Numbers represent mean values of three samples ± standard deviation. (C) Exposing MSCs transduced with Ad.TR to neat human serum with high levels of adenovirus neutralizing antibodies does not inhibit the ability of these cells to induce apoptosis in A549 cells. MSC cells were untransduced (MSC) or transduced with Ad.BGal (MSCβGal) or Ad.TR (MSCTR) with 100 pfu/cell. A549 cells that were treated with serum only are shown on the left (serum). After 24 hrs, the cells were harvested and incubated in neat serum for 30 min. at 37 C. The cell suspension (10⁴ MSC cells/well) and serum were then added to wells containing A549 cells seeded the previous day. After 48 hrs, samples were prepared for Nicoletti assay and analysed by flow cytometry. Numbers represent mean values of duplicate samples ± standard deviation. *P < 0.05.

Fig. 8

MSCs transduced with Ad.TR give rise to tumour growth reduction in vivo. (A) Immune-deficient mice were injected with 5 × 10⁶ A549 cells into the right flank. After 1 week when tumours were approximately 100 mm³ in size, 1 × 10⁶ MSCs either transduced with Ad.TR or Ad.EGFP were directly injected into the tumour. The tumour growth was followed over 3 weeks. This is depicted in the graph. Numbers represent mean values of five animals in each group ± standard deviation. **P < 0.001 (B) Representative tumours of two animals are shown after dissection clearly showing the smaller size in the animal from the group treated with Ad.TR-transduced MSCs.