Cloning of carrier cells infected with oncolytic adenovirus driven by midkine promoter and biosafety studies

Abstract Background A549 carrier cells infected with oncolytic adenovirus can induce complete tumor reduction of subcutaneous ovarian tumors but not intraperitoneal disseminated ovarian tumors. This appears to be a result of the insufficient antitumor effect of A549 carrier cells. Therefore, in the present study, we cloned a novel carrier cell with the aim of improving the antitumor effects. Methods Carrier cells infected with oncolytic adenovirus AdE3‐midkine with a midkine promoter were cloned by limiting dilution. We examined the antitumor effects of these cells on subcutaneous and intraperitoneal OVHM ovarian tumors in a syngeneic mouse model. Biosafety tests were conducted in beagle dogs and rabbits. Results We cloned EHMK‐51‐35 carrier cells with 10‐fold higher antitumor effects compared to A549 carrier cells in vitro. EHMK‐51‐35 carrier cells co‐infected with AdE3‐midkine and Ad‐mGM‐CSF induced a 100% complete tumor reduction in subcutaneous tumors and a 60% reduction of intraperitoneal disseminated tumors. Single‐dose acute toxicity test on beagle dogs with EHMK‐51‐35 carrier cells co‐infected with AdE3‐midkine and Ad‐cGM‐CSF showed no serious side effects. Biologically active adenoviruses were not detected in the blood, saliva, feces, urine or whole organs. In a chronic toxicity test, VX2 tumors in rabbits were injected five times with EHMK‐51‐35 carrier cells infected with AdE3‐midkine and these rabbits showed no serious side effects. Conclusions Significant antitumor effects and safety of cloned EHMK‐51‐35 carrier cells were confirmed in intraperitoneal ovarian tumors and toxicity tests, respectively. These findings will be extended to preclinical efficacy studies using dogs and cats, with the aim of conducting human clinical trials on refractory solid tumors.


| INTRODUCTION
More than 1000 clinical trials of cancer gene therapies have been conducted to date, although encouraging clinical results have yet to be obtained. 1 Replication-competent viral vectors have been developed to improve antitumor activity. Talimogene laherparepvec is a genetically modified herpes simplex virus type 1 that expresses granulocyte-macrophage colony-stimulating factor (GM-CSF) and was shown to improve the durable response rate in a phase III single-agent registration study. 2 This viral therapy is the first gene therapy drug approved by the US Food and Drug Administration.
An oncolytic, nonpathogenic ECHO-7 virus without genetic modification (Rigvir) was shown to significantly prolong survival in patients with early-stage melanoma without any side effects. 3 However, these viral vectors do not show a significant clinical effect with respect to improving the survival rate of patients with advanced cancers because they result in high titers of neutralizing antibodies that inhibit repetitive viral infections. 4 Adenovirus infection is highly likely to be inhibited even at initial administration because adenovirus type 5 seroprevalence is 80%-90% in humans. 5 To evade such antiadenovirus immunity, adenovirus can be coated with liposomes or polymers. However, the infectivity of these modified adenoviruses is effective only in low antibody titers, and these conditions have little antitumor effect in vivo and fail to induce complete tumor reduction. 6,7 Furthermore, because the adenovirus may induce fatal side effects as a result of a cytokine surge, 8 it cannot be administered intravenously. However, carrier cells infected with oncolytic adenovirus can be safely administered intravenously with significant antitumor effects. 9 Many studies of replication-competent virus-infected carrier cells have been described, including PA-1 ovarian cancer cells infected with oncolytic HSV-1, 10 mesenchymal stem cells infected with oncolytic adenovirus, 11 myeloma cells infected with oncolytic measles and vaccinia viruses 12 and autologous CD8 + lymphocytes infected with oncolytic vesicular stomatitis virus. 13 However, the anti-tumor potency of these carrier cells has been insufficient because they cannot produce sufficiently high virus titers and are vulnerable to damage even before targeting cancer cells.
Human non-small cell lung cancer A549 cells have been conventionally used to produce various viruses including adenovirus because of their high virus production capacity. A previous study showed that A549 carrier cells infected with oncolytic adenovirus exhibited a significant antitumor effect in immunocompromised mice. 14 Adenoviral particle-containing cell fragments derived from these A549 carrier cells were shown to be engulfed by target cancer cells. 14 This novel non-receptor-mediated adenoviral infection system circumvents neutralization by anti-adenovirus antibodies and enhances antitumor activity by inducing anti-adenoviral cytotoxic T lymphocyte (CTL) responses after pre-immunization with adenovirus in immunocompetent mice, thus inducing an anti-tumoral immune response. However, although A549 carrier cells infected with oncolytic adenovirus could completely reduce subcutaneous ovarian tumors, they were unable to reduce intraperitoneally disseminated ovarian tumors.
Biosafety tests for ovarian cancer-specific IAI.3B promoter-driven oncolytic adenovirus-infected A549 carrier cells for human clinical trial of recurrent solid tumors were reported in mice and rabbits. 15 However, biosafety tests for carrier cells co-infected with oncolytic adenovirus and adenovirus-GM-CSF have yet to be reported.
Midkine is overexpressed in the malignant solid tumors of humans, dogs and cats. More than one hundred million dogs and cats are bred in developed countries such as Japan, the USA and Europe, and half of animal deaths are the result of cancers. 16 Because treating cancers in companion animals by surgery, radiation and chemotherapy is impractical and uneconomical, more convenient and less invasive treatment methods should be developed. Complete treatment of tumors in companion animals by injection of carrier cells might be a potential strategy to circumvent these problems.
In the present study, to induce complete tumor reduction of intraperitoneally disseminated ovarian tumors using carrier cells infected with oncolytic adenovirus, we cloned a new carrier cell from cells that were established in our laboratory and characterized the antitumor activity and biosafety of these carrier cells. We injected the newly developed cloned carrier cells infected with midkine promoter-driven oncolytic adenovirus into mice, beagle dogs and rabbits aiming to examine antitumor efficacy and biosafety. These efficacy and biosafety tests could comprise a preliminary study for a clinical efficacy trial regarding recurrent canine and feline solid tumors and potentially provide proof-of-concept for their use as a pre-clinical efficacy trial for testing in humans.

| Cell lines and adenoviruses
Human ovarian cancer HEY and non-small cell lung cancer A549 cells were cultured in RPMI, and murine ovarian carcinoma OVHM cells were cultured in Dulbecco's modified Eagle's medium with high glucose. All cells were cultured with 10% heat-inactivated fetal calf serum (FCS), 5% antimycotics and antibiotics in 5% CO 2 at 37°C.
The construction and the purification of adenoviruses were performed as described previously. 14,15,17,18

| Cloning of EHMK cell lines
The EHMK cell line was established from human lung adenocarcinoma in Ehime University (Ehime, Japan) and cultured in RPMI with 10% heat-inactivated FCS, 5% antimycotics and antibiotics in 5% CO 2 at 37°C. EHMK cells were cloned by limiting dilution. EHMK cells were seeded at 0.25 cells/well in 96-well plates and incubated in 100 μl/well of RPMI medium, with another 100 μl/well of RPMI medium added into each well every 1 or 2 weeks. We selected cells that had spread from an initial single cell/well and showed stable growth.
These cells were cultured and analyzed as potential carrier cells.

| Cytotoxic assay
To evaluate the cytotoxic effects of AdE3-midkine-infected cell clones, EHMK cells grown in 96-well plates were replated in 5 × 10 6 cells/well in six-well plates, infected with AdE3-midkine at 200 multiplicity of infection (MOI) for 3 h in RPMI medium without FCS and cultured overnight in RMPI medium with 10% FCS. The infected clones were trypsinized and added to HEY cells at 2000 cell/well in 96-well plates with or without high titer (600×) of anti-adenovirus antibodies (Takeda Pharmaceutical, Osaka, Japan) for 5 days. Cells were fixed and stained with 0.5% crystal violet in 20% ethanol and absorbance was measured by a spectrophotometer at 550 nm (n = 5). The cytotoxic effects of AdE3-midkine-infected clones against HEY cells were calculated from the 50% inhibition level of cell growth (IC 50 ).

| Establishment of subcutaneous and intraperitoneal ovarian tumor model in syngeneic mice and treatments
Murine OVHM cells (10 6 ) were subcutaneously and intraperitoneally injected into female (C57BL/6 × C3H/He) F1 mice (CLEA Japan Inc., Tokyo, Japan) (n = 10 for each injected substance). In total, 140 mice  and used for the analyses. The extraction method for each sample is shown in Table 1. qPCR was performed using the LightCycler® Nano  Table 2.   carrier cells were infected with AdE3-midkine and Ad-cGM-CSF at 20 MOI for 33 h each and then 10 4 , 10 5 , 10 6 or 10 7 cells were injected subcutaneously into the thigh of four beagle dogs (n = 1 for each concentration) ( Table 3). EHMK-51-35 carrier cells were irradiated at 200 Gy, cryopreserved in liquid nitrogen and thawed before injection. Body weight, clinical symptoms and body temperature were observed every day, and saliva, feces and urine were collected every day. Blood was collected before and 1, 3, 5, 7 and 14 days after injection ( Figure 4A). Dogs were euthanized 14 days after injection and the organs of each dog were harvested (Table 4).

| Acute toxicity test of AdE3-midkine in beagle dogs
The experimental protocol of animal studies was approved by the Institutional Animal Care and Use Committee at Ehime University and Tokyo University of Agriculture and Technology. AdE3-midkine at 10 10 PFU was injected subcutaneously into three beagle dogs (Table 5). Body weight, clinical symptoms, body temperature, and saliva, urine and blood of the first beagle dog (No. 5) were observed and collected before and at 0, 3, 6, 12 and 24 h after injection ( Figure 5A). Feces were collected before and 24 h after injection.
The dog was euthanized 24 h after injection and the organs were harvested ( Table 6). Body weight, clinical symptoms and body temperature of the second dog (No. 6) were observed every day, and saliva, feces, urine and blood were collected before and every day after injection ( Figure 5A). The dog was euthanized 4 days after injection and the organs were harvested (Table 6). Body weight, clinical symptoms and body temperature of the third dog (No. 7) were observed every day; saliva, feces, urine and blood were collected before and 0, 1, 2, 3, 4, 5, 7, 9 and 11 days after injection ( Figure 5A). The dog was euthanized 11 days after injection and the organs were harvested

| Statistical analysis
Data are expressed as the mean ± SD and were analyzed using an unpaired t-test, the Welch tests, one-way analysis of variance and   Ad-cGM-CSF showed no significant antitumor effect compared to controls ( Figure 3A). A549 carrier cells infected with AdE3-midkine or co-infected with AdE3-midkine and Ad-mGM-CSF induced 30% or 60% of complete tumor reduction, respectively (p < 0.05). Notably, EHMK-51-35 carrier cells infected with AdE3-midkine or co-infected with AdE3-midkine and Ad-mGM-CSF induced 80% or 100% of complete tumor reduction, respectively, which increased the survival rates significantly compared to A549 carrier cells (p < 0.05).

| Antitumor effect of EHMK-51-35 carrier cells on intraperitoneal tumor-bearing mice
We performed similar analyses on the antitumor effect of EHMK-51- mice. Three intraperitoneal injections with adenovirus alone of AdE3-midkine or AdE3-midkine with Ad-cGM-CSF showed no significant antitumor effect compared to controls ( Figure 3B). In addition,

A549 carrier cells infected with AdE3-midkine or co-infected with
AdE3-midkine and Ad-mGM-CSF also showed no significant antitumor effect compared to controls. However, EHMK-51-35 carrier cells infected with AdE3-midkine or co-infected with AdE3-midkine and Ad-mGM-CSF induced 20% or 60% of complete tumor reduction, respectively, which increased the survival rates significantly compared to controls and A549 carrier cells (p < 0.05).

| Acute toxicity test of carrier cells infected with AdE3-midkine-and ad-cGM-CSF in beagle dogs
To evaluate the acute safety of AdE3-midkine-and Ad-cGM-CSF-

| qPCR and plaque assay
We constructed primer sets for each adenovirus (see Supporting information, Figure S1A) and performed qPCR to determine specificity. AdE3-midkine was amplified by qPCR in 10 to 10 8 PFU in a dose-dependent manner with a specific melting curve (see Supporting information, Figure S1B). qPCR of 10 8 PFU of AdE3, AdE3-midkine, Ad-cGM-CSF and Ad-fGM-CSF, using specific primers for AdE3-midkine, amplified only AdE3-midkine with a specific melting curve for AdE3-midkine (see Supporting information, Figure S1C). Similarly, qPCR with specific primers successfully amplified AdE3, Ad-cGM-CSF and Ad-fGM-CSF with specific melting curves for each adenovirus (see Supporting information, Figure   S1D-F).
We AdE3-midkine DNA and biologically active adenovirus (plaque assay) were not detected in blood, feces, urine, saliva or removed organ tissues in the three dogs injected with AdE3-midkine in the acute toxicity test.

| Acute toxicity test of AdE3-midkine in beagle dogs
To evaluate acute safety of AdE3-midkine, 10 10 PFU of AdE3-midkine was injected subcutaneously into the thighs of three beagle dogs.
Dogs were sacrificed at 24 h, 4 days or 11 days later; organs were  AdE3-midkine (Table 6). Although no change was observed in white blood cell (WBC) levels, α1AG increased over days 2-4 after injections ( Figure 5B). CRP started to increase 3 h after injection of AdE3-midkine, peaked on day 2 and returned to normal on day 9 ( Figure 5B). Other blood tests did not change after injections (see Supporting information, Figure S2). No abnormal clinical symptoms were detected in any of the three beagle dogs. A summary of abnormal clinical laboratory findings of the acute toxicity test of AdE3-midkine in beagle dogs is shown in Table 8.  AdE3-midkine were injected five times into VX2 tumors of two rabbits.
RBC, Ht and HGB decreased, and WBC, platelets, triglyceride and activated partial thromboplastin time (APTT) increased in the injected rabbits compared to controls (Figure 6), whereas other blood tests showed no changes (see Supporting information, Figure S3). Both  Table 9.  Because CRP started to rise more than 3 h after injection of AdE3-midkine, whereas α1AG and WBC did not rise, we consider CRP to be a more sensitive inflammatory response marker than Because disseminated intravascular coagulation (DIC) was mild, platelets might not be consumed and increased reactively; this is consistent with our previous results obtained in a rabbit safety test in which platelets increased after low-dose injections of A549 carrier cells and decreased after high-dose injections. 15 Increased P-LCR, MPV and PDW were considered to reflect an increased platelet production response. We found that RBC, Ht and HGB did not decrease after injections of AdE3-midkine alone but decreased in a dose-dependent manner after carrier cell injections into beagle dogs and decreased in rabbits whose tumors were injected five times. APTT was extended in rabbits treated with carrier cells, probably because DIC was caused by cytolysis of the injected carrier cells. These results are also consistent with our previous safety testing results in rabbits treated with A549 carrier cells. 15 Almost no blood chemistry abnormality was detected after injection of AdE3-midkine alone. No electrolyte abnormalities were detected after carrier cell injections and no abnormal values were detected in liver function or pancreatic function of beagle dogs that received high-dose carrier cells. These results are consistent with our previous safety testing results in rabbits treated with A549 carrier cells. 15 We observed decreased daily activity in beagle dogs treated with the highest dose and in tumor-bearing rabbits injected five times. Because injected tumor-bearing rabbits had relatively minor changes in clinical symptoms and laboratory findings compared to non-tumor bearing dogs that received the highest-dose of carrier cells, the changes of clinical symptoms and laboratory findings might be even smaller upon administration of carrier cells to dogs or cats with cancer. These animals might therefore be able to accept the highest dose of carrier cells safely. Dose limiting toxicity in carrier cell therapy was a decrease in daily activity and anemia as a result of RBC, HT and HGB reduction. These side effects are considered to be a result of DIC by cell lysis of carrier cells; it appears that the nuclear membrane was already broken at the time of freezethaw and the carrier cells were not alive before administration. 15 By reducing the infectious dose and infection time of AdE3-midkine and Ad-GM-CSF or by using fresh carrier cells without freezing, it may be possible to administer living carrier cells, thereby reducing side effects and increasing the dose.
With respect to future studies, after obtaining approval from the Ministry of Agriculture, Forestry and Fisheries of Japan, we plan to perform carrier cell therapy for cancers of dogs and cats and examine its safety and effectiveness. These results will inform the potential start of human clinical trials for refractory solid cancers. Because anti-PD-1 antibody [19][20][21][22] and anti-CTLA4 antibody 23 that suppress suppressor T cells can be formulated and effectively used in human cancers and might markedly activate carrier cell-induced CTLs, we expect that the antitumor effect of carrier cell therapy could be further increased by combination with these preparations.