Systemic delivery of fusogenic membrane glycoprotein-expressing neural stem cells to selectively kill tumor cells

Abstract

Intravenously injected neural stem cells (NSCs) can infiltrate both primary and metastatic tumor sites; thus, they are attractive tumor-targeting vehicles for delivering anticancer agents. However, because the systemic distribution of the injected NSCs involves normal organs and might induce off-target actions leading to unintended side effects, clinical applications of this approach is impeded. Given that the vesicular stomatitis virus glycoprotein (VSV-G) can promote the formation of multinucleated syncytia to kill cells in a pH-dependent manner, we engineered a pH sensor of VSV-G and generated a novel VSV-G mutant that efficiently promotes syncytium formation at the tumor extracellular pH (pHe) but not at pH 7.4. Using transduced NSCs derived from induced pluripotent stem cells (iPSCs), the VSV-G mutant was delivered into mice with metastatic breast cancers in the lung through tail vein injection. Compared with the conventional stem cell-based gene therapy that uses the herpes simplex virus thymidine kinase (HSVtk) suicide gene, this treatment did not display toxicity to normal non-targeted organs while retaining therapeutic effects in tumor-bearing organs. Our findings demonstrate the effectiveness of a new approach for achieving tumor-selective killing effects following systemic stem cell administration. Its potential in stem cell-based gene therapy for metastatic cancer is worthy of further exploration.

Figure 1

Vesicular stomatitis virus glycoprotein (VSV-G) mutant expression promotes cell fusion between neural stem cells (NSCs) and tumor cells. (a) Selection of a VSV-G mutant that can be expressed on NSCs. Immunofluorescence staining demonstrating the expression of wild-type (WT) and mutated VSV-G in living NSCs. NSCs were transfected with plasmids encoding VSV-G WT or mutants and collected after 24 hours for staining with mouse anti-VSV–G monoclonal antibody (clone P5D4) to detect surface VSV-G expression. (b) Syncytium formation between NSCs expressing VSV-G WT or H162R and 4T1 breast cancer cells. NSCs were mixed with 4T1 cells, and the co-cultures were treated with a pH 6.8 or 7.4 fusion buffer for 1 minute. Hoechst 33342 staining of the co-cultured cells 1 hour later was used to show syncytia (circled by dash line). (c) Dual color syncytium formation assay using NSCs and 4T1 cells. NSCs expressing the VSV-G H162R mutant were stained with calcein red-orange AM, whereas 4T1 cells were stained with calcein AM before syncytium formation assays. (d) Dual color syncytium formation assays using NSCs expressing H162R and U87 glioma cells (left) or Hepa 1-6 hepatocellular carcinoma cells (right). Bars in b–d: 100 μm.

Figure 2

Vesicular stomatitis virus glycoprotein (VSV-G) mutant expression promotes cell death. (a,b) Effects of transfection on cell viability. Neural stem cells (NSCs) were transfected with the indicated plasmids. Twenty-four hours post-transfection, the NSCs were mixed with 4T1 breast cancer cells (in a), U87 glioma cells (in b) or Hepa 1-6 hepatocellular carcinoma cells (in b). The co-cultures were treated with a fusion buffer at pH 6.8 or 7.4 for 1 minute. In the NSC-thymidine kinase (TK) group, ganciclovir (GCV) was added. After 48 hours, cell viability was measured by the MTS assay in a and b and caspase 3/7 activity was measured using a luminogenic caspase-3/7 substrate in a. (c) VSV-G H162R expression mediated by baculoviral transduction. NSCs were transduced with a baculoviral vector encoding VSV-G H162R at a multiplicity of infection of 100 plaque-forming unit per cell. H162R expression was analyzed with reverse transcriptase-PCR (top) and western blotting (bottom). (d) Effects of baculoviral transduction on cell viability. Twenty-four hours post-transduction, NSCs were mixed with 4T1 cells, and the mixtures were treated as described above. Cell viability was measured after 48 hours. NSC-eGFP: NSCs transfected with an eGFP control plasmid or transduced with a baculoviral vector encoding eGFP; NSC-VSVG, NSCs transfected with the VSV-G H162R plasmid or transduced with a baculoviral vector encoding VSV-G H162R; NSC-TK, NSCs transfected with the HSVtk plasmid. The results are given as the mean ± SD of three independent experiments, and each was performed in triplicate. Error bars represent SD. **P < 0.01 and ***P < 0.001 versus the NSC-eGFP control by analysis of variance. eGFP, enhanced green fluorescent protein; HSVtk, herpes simplex virus thymidine kinase; MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium.

Figure 3

Tumor tropism of neural stem cells (NSCs). (a) Effects of H162R expression on NSC migration. Boyden chamber migration assays were performed 24 hours after baculovirus transduction in NSCs. 4T1 cells were seeded in the lower chamber, and blank medium in the lower chamber was used for the negative control group. NSCs were stained with calcein AM and seeded in the upper chamber. After 24 hours, the percentage of cell migration was evaluated (n = 6). Bottom: Fluorescence images show the migration of NSCs toward the lower chambers. (b) Whole animal imaging with the IVIS imaging system. Immunodeficient Balb/c nude mice were inoculated with 4T1-luc breast cancer cells (Luc-4T1) by tail vein injection to establish a disseminated breast cancer model. Three days after tumor inoculation, DiR-labeled, human iPSC-derived NSCs were injected through the tail vein into the tumor-bearing mice. Imaging was performed on the indicated days. Imaging at day 0 was performed 5 minutes after NSC tail vein injection. Three mice without tumors (Balb/c) and three tumor-bearing mice (4T1 Balb/c) were imaged ventrally. Dual color imaging was performed for the tumor-bearing mice: DiR near-infrared fluorescence imaging was used to show the distribution of systemically injected NSCs, whereas bioluminescence imaging was used to document tumor growth in the same set of the mice. (c) Ex vivo organ imaging to demonstrate the tumor tropism of iPSC-NSCs. Organs were harvested from the animals 7 days after NSC injection. DiR fluorescence images show the organ distribution of systemically injected NSCs in normal and tumor-bearing mice. Bioluminescence images show Luc-4T1 metastases in tumor-bearing mice. The organs shown in each panel from left to right: lung, liver, spleen, kidney, heart, brain, stomach, spinal cord, and femur. (d) Quantitative evaluation of changes in organ distribution of iPSC-NSCs in tumor-bearing mice. DiR fluorescence signal values from c were used for analysis. Fluorescence intensity is shown (mean ± SD, n = 3 mice per group). Error bars represent SD. *P < 0.05 and **P < 0.01 versus normal Balb/c by analysis of variance. AU, arbitrary unit; iPSC, induced pluripotent stem cell; VSV-G, vesicular stomatitis virus glycoprotein.

Figure 4

In vivo gene therapy for 4T1 metastatic breast cancer using NSC-VSVG and NSC-TK/GCV. (a) The protocol used for the in vivo experiment. 4T1-luc cells were inoculated into mice by tail vein injection. Three days after 4T1 inoculation, PBS, NSC, NSC-eGFP, NSC-TK, and NSC-VSVG were inoculated by tail vein injection (n = 10 mice per group). GCV (50 mg/kg body weight) was intraperitoneally administered daily to the NSC-TK group, and PBS was given to other groups. (b) Bioluminescence images of tumor growth in representative animals from each group. The heat map represents the tumor area, and color represents the intensity. D: dorsal; V: ventral. (c) Quantitative analysis of bioluminescence signals. Error bars represent SD. *P < 0.05; **P < 0.01; and ***P < 0.001 versus the NSC-eGFP control group by analysis of variance. (d) Survival curves. Mice were observed until day 29 after 4T1 inoculation. eGFP, enhanced green fluorescent protein; GCV, ganciclovir; NSC, neural stem cell; PBS, phosphate-buffered saline; TK, thymidine kinase; VSV-G, vesicular stomatitis virus glycoprotein.

Figure 5

Hepato- and nephrotoxicities: NSC-VSVG versus NSC-TK/GCV treatment. (a) Alanine aminotransferase (ALT), (b) aspartate aminotransferase (AST), (c) blood urea nitrogen (BUN), and (d) creatinine assays demonstrating liver and kidney toxicity in each treatment group (n = 3 mice per group). Blood samples were collected from each mouse using orbital sinus blood sampling 1 week after NSC injection and used in the above assays. Error bars represent SD. *P < 0.05 and **P < 0.01 versus the PBS group by analysis of variance. eGFP, enhanced green fluorescent protein; GCV, ganciclovir; NSC, neural stem cell; PBS, phosphate-buffered saline; ppm, parts per million; TK, thymidine kinase; VSV-G, vesicular stomatitis virus glycoprotein.