Oncolytic Activity of a Recombinant Measles Virus, Blind to Signaling Lymphocyte Activation Molecule, Against Colorectal Cancer Cells

Oncolytic virotherapy is a distinctive antitumor therapy based on the cancer-cell-specific infectivity and killing activity of viruses, which exert a considerable antitumor effect with only a few treatments. Because colorectal cancer cells often acquire resistance to the molecular-targeted therapies and alternative treatments are called for, in this study, we evaluated the oncolytic activity against colorectal cancer cells of a recombinant measles virus (rMV-SLAMblind), which is blind to signaling lymphocytic activation molecule (SLAM) and infects target cells via nectin-4/poliovirus receptor-related 4 protein. We examined 10 cell lines including 8 cell lines that were resistant to epidermal-growth-factor-receptor (EGFR) targeted therapy. rMV-SLAMblind infected and lysed the nectin-4-positive cell lines dependently on nectin-4 expression, in spite of mutation in EGFR cascade. Tumour progression in xenograft models was also abrogated by the virus, and the infection of cancer cells in vivo by the virus was demonstrated with both flow cytometry and a histological analysis. Therefore, rMV-SLAMblind is considered a novel therapeutic agent for colorectal cancers, including those resistant to molecular-targeted therapies.

Scientific RepoRts | 6:24572 | DOI: 10.1038/srep24572 the pathogenesis of wild-type MV is mediated by the infection of immune cells via SLAM. Nectin-4 expression in the normal human body is observed in the placenta and is slightly detected in the epithelial cells of the trachea, where it forms adherens junctions together with E-cadherin [17][18][19] . rMV-SLAMblind caused no pathogenicity in rhesus or cynomolgus monkeys 13 . Recently, Noyce et al. 16 reported that some colorectal adenocarcinoma cell lines express nectin-4. In this study, we examined the antitumor effects of rMV-SLAMblind on colorectal cancer cells to investigate whether rMV-SLAMblind is an effective agent for treatment of colorectal cancer, especially with resistance to molecular targeted therapies.

Results
Nectin-4 expression in colorectal cancer cell lines. A flow-cytometric analysis was conducted to examine the expression of nectin-4 in colorectal cancer cell lines. Among the 10 cell lines examined (CaCo-2, DLD1, HT29, LS174T, SW48, SW948, HCT116, LoVo, RKO, and SW480), the CaCo-2, DLD1, HT29, LS174T, SW48, and SW948 cell lines expressed nectin-4, whereas the others did not (Fig. 1a, Table 1). Among these, nectin-4 expression in SW48 cells was heterogeneous, of which approximately 15% cells only express nectin-4 on the cell surface (Fig. 1a). The wild-type MV strains, including the HL strain, use SLAM as their receptor, whereas MV vaccine strains use CD46 14,20 , which is a recognition molecule expressed ubiquitously in human nucleated cells. We also analysed the expression of these receptors and observed that all the cell lines tested were negative for SLAM and positive for CD46 (Fig. 1a). To ascertain the expression of nectin-4 at the messenger RNA (mRNA) level, reverse transcription and polymerase chain reaction (RT-PCR) were performed. Higher expression of nectin-4 mRNA was observed in the cells that were positive for nectin-4 in the flow-cytometric analysis than in those that were nectin-4-negative on flow cytometry (Fig. 1a,b). Regarding SW48 cells, nectin-4 mRNA expression was as high as other nectin-4-positive cells in spite of their heterogeneous nectin-4 expression (Fig. 1b).

Infectivity and cytotoxicity of rMV-SLAMblind in colorectal cancer cell lines.
To investigate the susceptibility of the colorectal cancer cells to rMV-SLAMblind, each cell line was inoculated with the virus at a multiplicity of infection (MOI) of 2 and examined with fluorescence microscopy at 3 days post-infection (dpi). To visualize viral infection, enhanced green fluorescent protein (EGFP)-expressing rMV-SLAMblind (rMV-EGFP-SLAMblind) was used, based on the previous observations that the insertion of EGFP does not affect the growth kinetics of rMVs 21,22 . As shown in Fig. 2a, the replication of rMV-EGFP-SLAMblind was only observed in the nectin-4-positive cells. A water-soluble tetrazolium salt (WST) assay was performed to determine the killing activity of rMV-EGFP-SLAMblind in nectin-4-positive colorectal cancer cell lines. The inoculation of nectin-4-positive cells with rMV-EGFP-SLAMblind caused a time-dependent reduction in cell viability compared with that of the control (Fig. 2b). In contrast, the viabilities of nectin-4-negative cells were not altered after their inoculation with rMV-EGFP-SLAMblind (Fig. 2c).

Antitumor effects of rMV-SLAMblind in vivo. The antitumor effects of rMV-EGFP-SLAMblind in vivo
were examined using xenograft models. DLD1 and HT29 cells were transplanted into C.B-17/Icr-scid/scidJcl (SCID) mice. When the tumours reached 200 mm 3 , they were inoculated with rMV-EGFP-SLAMblind three times at weekly intervals. As shown in Fig. 3a, the administration of rMV-EGFP-SLAMblind exerted potent antitumor effects on the DLD1 cells, resulting in a reduction in the tumour volume of approximately 55% compared with the control. Similar results were obtained with the HT29 cell transplantation model, resulting in a reduction in the tumour volume of approximately 60% compared with the control (Fig. 3b). Twenty days after the first inoculation, each mouse was euthanized and their tumours weighed. The mean tumour weight in the virus-treated group was significantly lower than that in the control group for each type of tumour (Fig. 3c,d). We also performed a flow-cytometric analysis to investigate whether rMV-EGFP-SLAMblind remained within the tumour a week after the last administration of the virus. The live cell population was selected based on the incorporation of 7-amino-actinomycin D (7-AAD), detected with forward/side scatter (FSC/SSC; Fig. 3e). The mouse-derived H2K d -positive cells were gated out to focus on the live tumour cells (Fig. 3e). The proportion of EGFP-positive cells within the gate, which represented the rMV-EGFP-SLAMblind-infected live tumour cells, was 1-6% (2.9% on average) in the DLD1 cells and 0.2-2% (1.0% on average) in the HT29 cells (Fig. 3e,f). In contrast, the tumours in the control group were negative for EGFP (Fig. 3f). To analyse the distribution of the virus-infected cells, a histopathological analysis was performed on tumour tissues treated with rMV-EGFP-SLAMblind. Large necrotic regions were observed histopathologically in the tumour masses from both xenograft models, established with DLD1 and HT29 cells (Fig. 4a, H-E). Positive EGFP fluorescence, indicating the presence of the virus in growing cells, was observed in the area adjacent to the necrotic region (Fig. 4b). MV-N protein was immunostained in an analysis of serial sections, and the MV-N-positive area corresponded to the necrotic region and the EGFP-positive area in both xenograft models (DLD1 and HT29 cells) (Fig. 4a,b).
Unique pattern of nectin-4 expression in SW48 cells. The results described above clearly demonstrate the nectin-4-dependent infectivity and killing activity of rMV-EGFP-SLAMblind against colorectal cancer cells. Interestingly, the virus showed strong killing activity against SW48 cells (Fig. 2b), even though only approximately 15% of the population expressed nectin-4 in SW48 cells (Fig. 1a). To confirm lower MOI efficiently kills such cells with low expression of nectin-4, we inoculated the cells with three MOIs (0.1, 0.5, and 2) and found that, even at a lowest MOI, SW48 was killed efficiently (Fig. 5a). The heterogeneity of nectin-4 expression on SW48 cells was also confirmed with an immunofluorescence assay (Fig. 5b,c). To analyse whether rMV-EGFP-SLAMblind infects only the nectin-4-positive sub-population of SW48 cells, the nectin-4-positive and nectin-4-negative populations were sorted, and a WST assay was performed. The purity of either nectin-4-positive or -negative cells was > 99% respectively. Surprisingly, not only the sorted nectin-4-positive single cells but also the nectin-4-negative cell population were infected with the virus, and were killed by it to the same extent (Fig. 5d,e,f). Based on the  nectin-4-dependent infectivity of rMV-EGFP-SLAMblind in other cell lines, we hypothesized that SW48 cells constitutively express intracellular nectin-4 and intermittently express it on the cell surface, although most of the protein localizes in the cytoplasm. To examine this possibility, the total nectin-4 expression (both intracellular and cell-surface expression) and its cell-surface expression were determined. Live cells were stained with a mouse anti-nectin-4 monoclonal antibody (mAb) and an anti-mouse secondary Ab, fixed, permeabilized, and then stained with goat anti-nectin-4 polyclonal Ab (pAb), followed by an anti-goat secondary Ab. As expected, the total expression of nectin-4 was detected, regardless of the surface expression of nectin-4 (Fig. 5g). To directly determine whether nectin-4-negative cells become nectin-4-positive cells, the nectin-4-negative fraction of SW48 cells was sorted and cultured in foetal bovine serum (FBS)-containing medium for 2 h, and the extracellular nectin-4 expression was reanalysed. As shown in Fig. 5h, the population of cells expressing surface nectin-4 increased after culture. To examine whether this unique nectin-4 expression pattern in SW48 cells is attributable to the amino acid sequence of nectin-4, a sequence analysis was performed with mRNA obtained from SW48 cells. However, the sequence of the nectin-4-coding region, including its signal peptide, was identical to that in the GenBank database (accession number NM_030916.2) and no cell-line-specific mutation was observed (data not shown).

Discussion
In this study, we have demonstrated the antitumor effects of rMV-SLAMblind on colorectal cancer cells both in vitro and in vivo. Nectin-4 expression was observed in six of the 10 cancer cell lines tested. Targeted cancer therapy is one of the major treatments currently used for cancer. Resistance to EGFR inhibitors, which are among the most commonly used targeted therapies for colorectal cancer, occurs with mutations in EGFR-related molecules. Eight of the 10 tested cell lines had various mutations in the KRAS, BRAF, and/or PI3KCA oncogenes (Table 1), which confer resistance to anti-EGFR therapies [5][6][7][8][9] . Half of these cell lines (four of eight) expressed nectin-4 and were infected and killed by rMV-SLAMblind in this study. In addition, rMV-SLAMblind showed antitumor effect in xenograft models, even of HT29 to which cytotoxicity of rMV-SLAMblind in vitro was not high. These results suggest that rMV-SLAMblind is a novel therapeutic tool for the treatment of nectin-4-positive colorectal cancers, including those that are refractory to molecular-targeted therapies. Interestingly, serum nectin-4 levels have been used as prognostic markers in breast, ovary, and lung cancer 19,[23][24][25] . Moreover, because nectin-4 boosts the anchorage-independent growth of epithelial cells 26 , treatment with rMV-SLAMblind may reduce the number of cancer cells with a malignant phenotype, which drive cancer invasion and metastasis.
One of the therapeutic strengths of oncolytic viruses is their capacity to replicate within tumour cells. Theoretically, the number of oncolytic viral particles can increase beyond the number initially injected, exerting a relatively long-lasting antitumor effect after one treatment. The data from both our flow-cytometric and histological analyses exemplify this, demonstrating the presence of rMV-SLAMblind within the tumour cells in vivo at least 7 days in this study. MV is also known to induce strong cell-mediated immune responses 27 , which target MV-infected cells. Therefore, rMV-SLAMblind-infected cells should be the targets of the cell-mediated immunity of the host and will also be killed by it. Adequate in vivo models using immunocompetent animals must be established to predict more accurately the clinical efficacy of rMV-SLAMblind in cancer patients, and to understand the involvement of the host immune responses in virotherapies.
Interestingly, the nectin-4 expression pattern was distinctive in SW48 cells. Our data suggest that nectin-4 is constitutively expressed intracellularly and its surface expression is rapidly turned over in SW48 cells. Because no SW48-cell-line-specific mutation in the nectin-4 mRNA was detected, including in the region coding the signal peptide, a further investigation is necessary to clarify the mechanism of this nectin-4 turnover. In addition, SW48 cells were efficiently killed by rMV-SLAMblind. This raises the possibility that rMV-SLAMblind is even effective in some tumours that express nectin-4 intermittently. On the other hand, it was reported that there is a cell line that was partially nectin-4-positive but not susceptible to MV infection (SCaBER, urinary bladder squamous cell line) 15 , and that nectin-4 is necessarily complexed with afadin to work as entry receptor for MV 28 . Thus, effects of expression level of nectin-4 associated proteins on susceptibility of cancer cells to rMV-SLAMblind should be investigated in future.
In conclusion, we have demonstrated the antitumor effects of rMV-SLAMblind against nectin-4-positive colorectal cancer cells, including cells resistant to EGFR inhibitors, indicating that it is a potential novel therapy for colorectal cancers.

Materials and Methods
Cell culture. All colorectal cancer cell lines, which expressed various kinds of KRAS, BRAF and PIK3CA status 29,30 (Table 1), were obtained from the American Type Culture Collection (Manassas, VA) 31 . MCF7 human breast cancer cells were obtained from the RIKEN Cell Bank (Tsukuba, Japan). DLD1, HT29, and SW48 cells were cultured in RPMI1640 medium (Life Technologies, Gaithersburg, MD) supplemented with 10% FBS (AFC Biosciences, Lenexa, KS) and antibiotics. All other cells, were maintained in Dulbecco's modified Eagle's medium (Life Technologies) supplemented with 10% FBS and antibiotics.
To determine the whole coding region sequence of nectin-4, PCR was conducted with LA Taq DNA Polymerase (Takara) and the following primer set: forward primer 5′-GGTCAGTTCCTTATTCAAGTCTGC-3′ and reverse primer 5′-GCTAAAATCTCCCATGTCAACAG-3′. The PCR products were cloned into a TA cloning vector (pGEM-T; Promega, Madison, WI), and then sequenced on an ABI 3130 Genetic Analyzer (Life Technologies). The sequence was compared with the GenBank reference (accession number NM_030916.2). Virus. rMV-EGFP-SLAMblind were grown in MCF-7 as described previously 13 , and the virus stocks were kept at − 70 °C. The titres of rMV-EGFP-SLAMblind were determined as 50% tissue culture infectious doses (TCID 50 ) in MCF7 cells based on the Reed-Muench method 32 . Briefly, MCF-7 cells in 96-well plates were inoculated with virus suspensions, which were serially 10-fold diluted. Then plates were incubated for 7 days and viral titres were  determined. All MOIs used in this study were determined based on this value. The titres calculated using MCF7 cells were almost identical to the one using nectin-4-expressing Vero cells (data not shown).

Viral infection of each colorectal cancer cell line with rMV-EGFP-SLAMblind. Monolayers of cells
in 96-well plates were infected with rMV-EGFP-SLAMblind at MOI 2 and the EGFP expression in the cells was detected at 3 dpi with an FV1000 microscope.
WST assay. Cell viability was determined with the WST-1 Cell Proliferation Kit (Takara), according to the manufacturer's instructions. Briefly, 0.5-2 × 10 4 cells in 96-well plates were infected with rMV-EGFP-SLAMblind and cultured for the indicated days in FBS-containing medium. For the evaluation of cell viability, 10 μl WST-1 solution was added in each well, incubated for 2-4 h, and then the absorbance at 450 nm was measured at the indicated dpi using a Model 450 Microplate Reader (Bio-Rad, Hercules, CA). The viability of the cells was determined as described previously 13 .
Antitumor effects of rMV-SLAMblind in vivo. All experiments with animals both complied with the standards specified in the guidelines of the Experimental Animal Committee of The University of Tokyo, and were reviewed and approved by the institutional committee. A total number of 5 × 10 6 cells were suspended in a 50% concentration of growth factor-reduced Matrigel (BD Biosciences) and injected subcutaneously into the right flanks of 6-week-old female C.B-17/Icr-scid/scidJcl mice (Clea Japan, Tokyo, Japan). The mice were carefully monitored for the development of palpable or visible tumours at the sites of injection. The tumours were measured with a calliper every 2 or 3 days. The tumour volume (V) was calculated with the formula V = ab 2 /2, where a and b are the length and width of the tumour mass (in mm), respectively. The mice were administered 10 6 TCID 50 of rMV-EGFP-SLAMblind or plain RPMI 1640 medium intratumorally three times at weekly intervals. Twenty days after the first inoculation, the mice were killed and the tumour tissues analysed immunohistochemically.
Flow-cytometric analysis of in vivo tumours. Tumour tissues were digested with HBSS containing 5 mM HEPES, 2% FBS, 1 mg/mL collagenase (Wako Pure Chemical Industries, Osaka, Japan) and 0.1% DNase I (Life Technologies). The cells were stained with 7-AAD and phycoerythrin-Cy7-conjugated anti-mouse H2K d Ab (clone SF 1-1.1; BD Biosciences), and then fixed with 4% PFA. The cells were analysed with a BD FACSVerse flow cytometer, and the data were processed using the FlowJo FACS analysis software ver. 9.5.3.
Immunohistochemistry and fluorescence microscopy. Tumour tissues were fixed in 4% PFA overnight at 4 °C, embedded in Tissue-Tek OCT compound (Sakura Finetechnical Co., Ltd, Tokyo, Japan), and frozen in liquid nitrogen. The tissues were cut into 6 μm thick sections, fixed in acetone, washed in phosphate-buffered saline, and blocked with hydrogen peroxide. The sections were then incubated with anti-MV-nucleocapsid (N) protein pAb, which was produced in our laboratory 33 , and secondary antibody (Envision HRP kit; Dako). Positive reactions were identified by incubating them with DAB reaction solution. As a negative control, species-matched and filtered sera were used instead of primary antibodies. Images were captured with a Nikon microscope (Nikon, Melville, NY). The sections were fixed for fluorescence microscopy in acetone and stained with Hoechst 33342 (Cambrex Bio Science Walkersville Inc., Walkersville, MD) diluted 1:10,000, and were mounted in Dako Glycerol Mounting Medium. Images were taken on an FV1000 microscope.
Statistical analysis. Two-tailed Welch's t-test were used for the statistical analysis of the data, and p values of < 0.05 were considered statistically significant.