Coadministration of Epithelial Junction Opener JO-1 Improves the Efficacy and Safety of Chemotherapeutic Drugs

One of the key features of epithelial tumors is the presence of intercellular junctions, which link cells to one another, and act as barriers to the penetration of molecules with a molecular weight of 500 Da (1–3). Given that many chemotherapy drugs are larger than 500 Da, intercellular junctions represent a barrier to the penetration of these therapeutic agents into tumor. Several studies have shown that upregulation of epithelial junction proteins correlated with increased resistance to therapy, including therapy with monoclonal antibodies and chemotherapeutics (4, 5). One of these junction proteins is desmoglein 2 (DSG2). DSG2 is upregulated in malignant cells (6, 7). We found higher DSG2 immunoreactivity in breast cancer cells than in the surrounding normal epithelial tissue or tumor stroma cells (Supplementary Fig. S1).

Recently, we developed a recombinant protein (JO-1) that transiently triggers the opening of intercellular junctions in epithelial tumors. This work is based on our finding that DSG2 is a high-affinity receptor for a number of human adenoviruses (Ad), including Ad serotype 3 (8, 9). JO-1 is a self-dimerizing recombinant protein derived from the Ad3 fiber (10). JO-1 has a molecular weight of approximately 60 kDa and binds with picomolar avidity to DSG2. It can be readily produced in Escherichia coli and purified by affinity chromatography.

In mouse xenograft tumor models, we have shown that intravenous administration of JO-1 mediated cleavage of DSG2 dimers (between epithelial tumor cells) and activated intracellular signaling pathways, which reduced the expression of epithelial junction proteins in tumors (11). The morphologic changes triggered by JO-1 occurred within 1 hour after intravenous JO-1 injection and allowed for increased intratumoral penetration of the anti-Her2/neu monoclonal antibody trastuzumab (Herceptin) as well as for improved access to its target receptor, which is partly trapped in epithelial junctions (11). The effects of JO-1 on epithelial junctions translated into increased therapeutic efficacy of monoclonal antibodies (e.g., trastuzumab, cetuximab/Erbitux) against several xenograft tumor models, including breast, colon, ovarian, gastric, and lung carcinoma models (11).

Ad3 and its derivative JO-1 do not bind to mouse cells, implying that mouse DSG2 is not recognized (12). For safety studies with JO-1, we therefore used human DSG2 (hDSG2) transgenic mice that we recently generated. These mice express hDSG2 in a pattern and at a level similar to humans (12). For JO-1 efficacy studies, we also created a mouse epithelial breast cancer line that expressed hDSG2 and formed tumors in hDSG2-transgenic mice.

Using human xenograft and mouse tumor models, we showed that JO-1 increases the efficacy of a number of chemotherapy drugs that are widely used in the treatment of cancer patients.

The production of JO-1 in E. coli and its purification have been described previously (10).

Breast cancer BT474-M1 cells and MDA-MB-231 [American Type Culture Collection (ATCC), HTB-26] were cultured in Dulbecco's Modified Eagle's Medium/F12 with 10% FBS, 1% penicillin/streptomycin and 2 mmol/L l-glutamine. Breast cancer HCC1954 (ATTC; CRL-2338), lung cancer A549 (ATCC; CCL-185), prostate cancer 22Rv1 (ATCC; CRL-2505), and mouse mammary carcinoma (MMC) and MMC-hDSG2 cells (12) were cultured in RPMI with 10% FBS and 1% penicillin/streptomycin. To achieve cell polarization, 1.4 × 10⁵ T84 cells (ATCC; CCL-248) were cultured in collagen-coated 6.5 mm Transwell inserts (0.4 μm pore size; Costar Transwell Clears) for a period of 14 to 20 days until transepithelial resistance was stable (8).

The preparation of optimum cutting temperature and paraffin sections has been described previously (12). Bones were decalcified before embedding in paraffin. A list of antibodies used for immunohistochemistry and flow cytometry is available in the Supplementary Information. Images were taken with a Leica DMLB microscope (Wetzlar), using Leica DFC300 FX Digital camera and Leica Application Suite Version 2.4.1 R1 software.

All experiments involving animals were conducted in accordance with the institutional guidelines set forth by the University of Washington. Breast cancer xenografts were established by injecting tumor cells into the mammary fat pad of CB17 SCID-beige mice. Prostate and lung cancer models were created by subcutaneous injection. hDSG2-transgenic mice were generated by knock-in of approximately 90 kb of human DNA containing the DSG2 gene together with its regulatory regions. The mice contain 2 copies of the hDSG2 gene and express hDSG2 in a pattern similar to humans (12). MMC-hDSG2 tumors were generated by subcutaneous injection of MMC-hDSG2 cells mixed with Matrigel. JO-1 was intravenously injected 1 hour before the application of chemotherapy drugs. Tumor volumes were measured 3 times a week. Each treatment group consisted of a minimum of 5 mice. Animals were sacrificed, and the experiment terminated when tumors in one of the groups reached a volume of 800 mm³ or tumors displayed ulceration. To produce anti–JO-1 antibodies, hDSG2-transgenic mice received 3 subcutaneous injections of 5 μg JO-1 at days 0, 3, and 14. Serum was collected 4 weeks later and analyzed for JO-1–specific antibodies by Western blot.

To measure liposomal doxorubicin/Doxil concentrations, mice were sacrificed and blood was flushed from the circulation with 10 mL PBS. Tissues were homogenized in PBS/0.1% Tween 20/protease inhibitors. Anti-PEG (PEG, polyethylene glycol) antibody AGP4 (13) was used as a capture antibody. Binding was detected with anti-PEG antibody 3.3.-biotin (13) followed by a streptavidin–HRP (HRP, horseradish peroxidase) conjugate. To measure JO-1 in tissue lysates, an anti-Ad3 fiber knob antibody (8) was used as capture antibody. JO-1 binding was detected using a mouse monoclonal antibody against the 6xHis tag of JO-1, followed by anti-mouse IgG-HRP (IgG, immunoglobulin G).

All results are expressed as mean ± SD. Two-way ANOVA for multiple testing was applied. Animal numbers and P values are indicated in the figure legends.

The goals of this study were to show that: (i) JO-1 cotherapy specifically increases the antitumor efficacy of chemotherapeutic drugs in epithelial tumors and (ii) JO-1 cotherapy allows dose sparing of chemotherapeutics, which in turn reduces adverse side effects associated with chemotherapy. Our study included the following chemotherapy drugs: (i) paclitaxel/Taxol (molecular weight 856.9 Da). Paclitaxel is used to treat patients with lung, ovarian, breast, head, and neck cancer and advanced forms of Kaposi sarcoma. (ii) Nanoparticle albumin–bound paclitaxel nab-paclitaxel/Abraxane (effective size of the particle: ∼130 nm). nab-Paclitaxel has recently been approved for the treatment of recurrent breast cancer. (iii) Irinotecan/Camptosar (molecular weight 586.7 Da). Irinotecan is used to treat colon and lung cancer. The most significant adverse effects of irinotecan are severe diarrhea and immunosuppression. (iv) Liposomal doxorubicin/Doxil (effective size of the particle: ∼90 nm) is a PEG coated, liposome-encapsulated form of doxorubicin. Liposomal doxorubicin is used in the treatment of several epithelial tumors, including ovarian and breast cancers. To test the effect of JO-1 on chemotherapy, we studied 5 xenograft models using immunodeficient CB17-SCID-beige mice carrying tumors derived from human tumor cell lines. The epithelial phenotype and existence of tight junctions in the xenograft tumors was confirmed by immunohistochemical detection of the junction proteins DSG2, E-cadherin, ZO-1, and claudin 7 (8, 14; data not shown). Furthermore, we used an hDSG2-transgenic mouse model with orthotopic breast tumors expressing hDSG2. The doses of chemotherapy drugs used in our studies were based on information reported in previous publications (15). These doses reflected doses used in patients converted through allometric scaling to the weight of mice. Mice with preestablished tumors were intravenously injected with either JO-1 (2 mg/kg) or PBS 1 hour before intravenous/intraperitoneal administration of the chemotherapeutic agents. This regime had been found to be optimal in the combination therapy of JO-1 with monoclonal antibodies (14).

Cotherapy of JO-1 and paclitaxel was first tested in the BT474-M1 breast cancer model (Fig. 1A). These tumors are resistant to paclitaxel. JO-1 cotherapy was able to halt tumor growth and thus overcame resistance to paclitaxel treatment. JO-1/paclitaxel cotherapy was then evaluated in the MDA-MB-231 breast cancer model (Supplementary Fig. S2). JO-1 injection alone had no significant therapeutic effect. Antitumor effect of paclitaxel alone was found, which was significantly enhanced by preinjection of JO-1. The MDA-MB-231 tumor model was also used to study the effect of JO-1 on nab-paclitaxel therapy (Fig. 1B). Nab-paclitaxel was not significantly more effective than paclitaxel in decreasing tumor growth in this tumor model. Similar to paclitaxel, combining JO-1 with nab-paclitaxel therapy increased its therapeutic effects, whereby the enhancing effect was significantly more pronounced for nab-paclitaxel. This suggests that the enhancement in therapeutic efficacy through JO-1 is stronger for larger, particle-based drugs. Similarly, JO-1 was shown to enhance the antitumor effects of liposomal doxorubicin in the prostate cancer model (Fig. 1C) and the model with subcutaneous tumors derived from primary ovarian cancer cells (ref. 16; Supplementary Fig. S3A). The residual tumor mass seen in JO-1 plus chemotherapy treated mice in Fig. 1A and C only contained tumor stroma and infiltrating leukocytes upon histologic inspection. Furthermore, in these studies, no tumor growth relapse was observed when mice were followed for a total of 60 days after the start of the experiment.

Paclitaxel, nab-paclitaxel, and liposomal doxorubicin all have molecular weights of more than 500 Da; therefore, intercellular junctions could represent physical barriers to the penetration of these drugs into tumors. If our hypothesis is correct, JO-1–mediated junction opening should have less effects on drugs smaller than 500 Da. To test this hypothesis, we evaluated the effects of JO-1 cotherapy on cisplatin, which has a molecular weight of 300 Da in models of breast (Fig. 1D), lung (Fig. 1E), and ovarian (Supplementary Fig. S3B). Notably, it has been shown before that the cell lines used in this study were sensitive to cisplatin in culture (16, 17). Further support for the hypothesis that larger drugs benefit more from JO-1 cotherapy comes from a study with cyclophosphamide (molecular weight 260), in which JO-1 also did not significantly enhance the therapeutic effect (Fig. 1F).

Our second hypothesis is that JO-1 cotherapy would allow decreasing the doses of chemotherapeutics required to achieve therapeutic efficacy. To test this, we injected chemotherapeutic agents at different dose levels into tumor-bearing mice and monitored tumor volumes as well as toxicity parameters, including blood cell counts, blood chemistry, and tissue histology. First, we tested JO-1 in combination with the topoisomerase inhibitor irinotecan in the A549 lung cancer model described above (Fig. 2A–D). As observed for the other drugs, JO-1 significantly increased the therapeutic efficacy of irinotecan. A combination of a low dose of irinotecan (37.5 mg/kg) and JO-1 was significantly more effective than “standard” dose (75 mg/kg) irinotecan alone, indicating that JO-1 allows for lowering the effective drug dose (Fig. 2A). Irinotecan treatment caused thrombocytopenia and leukopenia as a result of myelotoxicity, although there was no significant difference in the severity of leukopenia or thrombocytopenia observed in animals treated at the 2 different doses (Fig. 2B). Coadministration of JO-1 had a myeloprotective effect. It prevented the decrease in platelet and white blood cell counts, which was especially noteworthy at the higher drug dose. Compared with control animals, spleens were significantly smaller in irinotecan-treated animals but not in animals that received irinotecan in combination with JO-1. Splenic atrophy in irinotecan-treated animals is also visible in microscopic analysis of spleen sections (Fig. 2C). JO-1 also reduced irinotecan-mediated damage to the intestinal epithelium (Fig. 2D).

Next, we evaluated nab-paclitaxel in the MDA-MB-231 breast cancer model. nab-Paclitaxel was administered at 2 different dose levels (5 and 10 mg/kg) alone or in combination with JO-1 (Fig. 2E–F). JO-1 cotherapy significantly increased the antitumor effects of nab-paclitaxel (Fig. 2E). The combination of JO-1 plus nab-paclitaxel (5 mg/kg) was significantly more effective than nab-paclitaxel alone given at a dose of 10 mg/kg. JO-1 plus nab-paclitaxel (10 mg/kg) halted tumor growth. Compared with paciltaxel, nab-paclitaxel showed a better safety profile based on blood cell analysis (data not shown). However, the highest dose of nab-paclitaxel (10 mg/kg) still decreased the white blood cell count, which was prevented by JO-1 pretreatment (Fig. 2F). The myeloprotective effect was also shown in sternum sections. In contrast to the depletion of bone marrow cells seen in mice treated with ABX alone, animals that received JO-1/ABX cotherapy displayed a normal histology (Fig. 2G). Notably, no toxic side effects of nab-paclitaxal were found in the other tissues analyzed, that is, intestine, kidney, and spleen.

The third model involved liposomal doxorubicin in the A549 lung cancer model (Fig. 2H–K). Similar to the results above, JO-1 pretreatment significantly improved liposomal doxorubicin therapy at both dose levels of the chemotherapy drug (1 and 3 mg/kg; Fig. 2H). JO-1 mitigated adverse side effects induced by liposomal doxorubicin. For example, markers for liver damage, including aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) were significantly decreased in animals treated with JO-1 and liposomal doxorubicin than in mice treated with liposomal doxorubicin alone (Fig. 2I). Severe tissue damage, caused by liposomal doxorubicin, found in the bone marrow and intestinal epithelium was greatly reduced in mice that received JO-1 injections (Fig. 2J and K).

Importantly, in all 3 studies shown in Fig. 2, we found that JO-1 cotherapy greatly reduced the toxic side effects associated with chemotherapy. We speculate that the ability of JO-1 to open up intercellular junctions in tumors and increase the uptake of chemotherapeutics reduces the drug exposure to normal tissue thus providing a larger therapeutic window. In support of this speculation, using an ELISA to measure PEGylated compounds in tissues (13), we found significantly more liposomal doxorubicin in tumors and less in other tissues in mice that received JO-1 prior to intravenous liposomal doxorubicin injection (Fig. 3A). Immunofluorescence analysis of tissue sections revealed more Doxil signals in tumors of JO-1 + Doxil-treated mice (Fig. 3B). In these animals, Doxil was found more dispersed over a greater distance from blood vessels, suggesting better intratumoral penetration and absorption by tumor tissue (Fig. 3B).

A more adequate model for safety and efficacy studies with JO-1 are hDSG2-transgenic mice with MMC-hDSG2 tumors. JO-1 binds to MMC-DSG2 cells and triggers reorganization of junctions in a similar way as seen in human epithelial cells, indicating that JO-1 signaling through hDSG2 can override junction regulation in mouse cells (12). Transplantation of MMC-hDSG2 cells into the mammary fat pad of hDSG2-transgenic mice resulted in tumors that resembled key features of breast cancer in humans, for example, nests of hDSG2^bright epithelial cells that were surrounded by tumor stroma (Fig. 4A). Intravenous injection of JO-1 resulted in efficient accumulation of JO-1 in MMC-hDSG2 tumor nests as seen on tumor sections analyzed 6 hours after injection (Fig. 4A). Immunoreactive JO-1 was also found in epithelial cells of the intestine in hDSG2-transgenic mice (Fig. 4B), which is consistent with our recent biodistribution study with Ad3-GFP virus (12). We measured JO-1 concentrations in tissue and tumor lysates by ELISA at 6 hours after injection in the following 4 groups of animals: (i) hDSG2-transgenic mice with MMC-hDSG2 tumors (Fig. 4C; “hDSG2-pos”), (ii) hDSG2-transgenic mice without tumors (Fig. 4D; “hDSG2-pos”), (iii) nontransgenic littermates with MMC tumors (Fig. 4C; “hDSG2-neg”), and (iv) nontransgenic littermates without tumors (Fig. 4D; “hDSG2-neg”). We found highly selective, hDSG2-dependent accumulation of JO-1 in MMC-hDSG2 tumors established in hDSG2-transgenic mice. About 10-fold less JO-1 was found in the intestine and colon of hDSG2-transgenic mice. JO-1 was also detected in the liver and spleen of both hDSG2-positive and hDSG2-negative mice, suggesting an hDSG2-independent uptake. Immunohistochemistry analyses showed that JO-1 in the liver is mostly present in sinusoidal spaces and F4/80-positive Kupffer cells (Supplementary Fig. S4) and not taken up by parenchymal liver cells. Similarly, in the spleen, JO-1 signals appear in the peripheral zone of germinal centers (Supplementary Fig. S4) and are mostly likely the result of uptake by splenic macrophages. The amount of JO-1 in the intestine and colon was more than 2-fold higher in hDSG2-transgenic mice without MMC-hDSG2 tumors, supporting our hypothesis that the tumor acts as a “sink” for JO-1. From the JO-1 biodistribution studies in MMC-hDSG2 tumor-bearing hDSG2 mice, the question arose why JO-1 predominantly accumulates in the tumor. Immunofluorescence studies on tumor and intestine sections of hDSG2-transgenic mice, showed that epithelial tumor cells lack strict polarization, in contrast to normal epithelial cells (Fig. 4E and F). As a consequence, in tumors, hDSG2 appears to be readily accessible, while in intestinal epithelial cells most of the hDSG2 is trapped in junctions, reflected by overlapping signals for hDSG2 and the junction marker claudin 7 (Fig. 4F).

We also measured clearance of JO-1 from blood after intravenous injection (Supplementary Fig. S5). These studies show that the half-life of JO-1 in hDSG2-transgenic mice is approximately 6 hours. JO-1 clearance from blood is slower in hDSG2-transgenic mice than in nontransgenic littermates, which could be due to the fact that hDSG2 is expressed on platelets and subfractions of leukocytes (12).

As JO-1 is a viral protein, adaptive immune responses might develop in humans, particularly after repeated injection. Despite the fact that approximately one third of humans have neutralizing antibodies against Ad3 (8), our studies showed that these antibodies did not interact with JO-1 (Fig. 5A). Furthermore, polyclonal anti–JO-1 antibodies generated by vaccination of mice did not affect the enhancing effect of JO-1 transepithelial transport of PEG 4000 in polarized colon cancer cultures (Fig. 5B). Our hDSG2-transgenic mouse/MMC-hDSG2 model allowed us to test the effect of anti–JO-1 antibodies on JO-1 activity in vivo (Fig. 5C). When JO-1 antibodies were detectable in serum of JO-1–vaccinated mice, MMC-hDSG2 tumor cells were injected into the mammary fat pad. As seen in the xenograft tumor models, JO-1 increased the efficacy of Doxil therapy in vaccinated, hDSG2-transgenic mice with preestablished MMC-hDSG2 tumors (Fig. 5C). Interestingly, JO-1 injection alone had a significant therapeutic effect in this immunocompetent tumor model. Because this effect was not seen in the immunodeficient tumor models (Fig. 1), we speculated that JO-1 facilitates antitumor immune responses. In support of this speculation, we found significantly more tumor-infiltrating leukocytes in MMC-hDSG2 tumor nests in JO-1–injected mice (Fig. 5E and F). Clearly, detailed T-cell studies are required to support this speculation. [Notably, MMC-hDSG2 cells express rat Neu (18), which could serve as an antigen.]

As seen before in hDSG2-transgenic mice that received intravenous Ad3-GFP injection, JO-1 injection also caused a mild diarrhea that started within 3 hours after injection and subsided by day 2. There were no significant changes in blood parameters at 24 hours (Supplementary Fig. 6A and B) and 3 days (not shown) after injection of JO-1 at a dose of 2 mg/kg, that is, a dose that conferred an enhancing effect on Doxil therapy. After injection of JO-1 at a dose of 10 mg/kg, blood analysis detected leucopenia, decrease in serum glucose levels, and increase in serum GPT levels. Liver damage is also reflected by the presence of steatosis in the liver (Supplementary Fig. S6C). Notably, the potential danger of provoking a serious immune response as a consequence of the administration of multiple doses of JO-1 to patients has to be taken into consideration.

Overall, our studies in hDSG2-transgenic mice show that JO-1 (2 mg/kg) cotherapy is safe and effective in the presence of anti–JO-1 antibodies.

We showed that JO-1 cotherapy allows doses of chemotherapeutics to be decreased without compromising antitumor effects and that it provides protective effects to normal tissues resulting in an improved safety profile for chemotherapy. We also showed that JO-1 can overcome resistance of xenograft tumors to a chemotherapy drug. Our data indicate that JO-1 predominantly accumulates in tumors. A number of factors could account for this, including (i) overexpression of DSG2 on tumor cells, (ii) better accessibility of DSG2 on tumor cells because of a lack of strict cell polarization, and (iii) a high degree of vascularization and vascular permeability in tumors. Because of its preferential action on epithelial junctions of tumors, JO-1 appears to create as “sink” for chemotherapy drugs, thereby reducing the exposure of normal tissue to these drugs.

Before JO-1 cotherapy is used in cancer treatment, certain issues will need to be addressed.

This might not be a critical issue if JO-1 is used in combination with chemotherapy, which suppresses immune responses to foreign proteins (even at doses that are markedly lower than drug doses used for cancer treatment). This expectation is supported by studies with oncolytic adenovirus vectors in which immunosuppression allowed for repeated vector application (19). Furthermore, we have showed that JO-1 remains active in vitro and in vivo even in the presence of anti–JO-1 antibodies generated by vaccination of mice. This might be due to the fact that the JO-1 interaction with DSG2 is of very high avidity and cannot be disrupted by polyclonal anti–JO-1 antibodies. [Notably JO-1 is a dimer of a trimeric fiber knob (10)]. It can however not be excluded that repeated JO-1 injection into immunocompetent patients results in the development of competing, high-affinity antibodies. This problem can potentially be addressed by generating next generation JO-1 proteins that have immunodominant epitopes removed or higher affinity to DSG2. Furthermore, we have recently identified the DSG2-interacting residues within JO-1 and based on this knowledge, it might be possible to transplant the DSG2-interacting Ad3 fiber loops into less immunogenic scaffolds.

In agreement with other studies (6, 7), we have found a higher DSG2 expression in malignant tissues than in the surrounding normal epithelial tissue (Supplementary Fig. S1). Furthermore, a recent study on squamous cell cancer showed that, in contrast to other desmosomal proteins, DSG2 was upregulated in invasive cancer (20). An National Center for Biotechnology Information website reports DSG2 expression in esophageal, colorectal, head and neck, kidney, uterine, bladder, prostate, gastrointestinal tract, breast, cervical, lung, ovarian, pancreatic, and skin tumors, whereas DSG2 was absent in adrenal tumors, sarcoma, glioma, leukemia, lymphoma, retinoblastoma, and soft tissue/muscle tissue tumors. (http://www.ncbi.nlm.nih.gov/UniGene/ESTProfileViewer.cgi?uglist = Hs.412597). There are, however, also studies reporting a reduction in the amounts of DSG2 in invasive pancreatic or gastric cancer (21, 22), and it has been argued that epithelial features, including tight junctions, are lost in metastatic cancer. These arguments do not consider that in advanced carcinomas mesenchymal cells can regain characteristics of epithelial cells via mesenchymal-to-epithelial transition (23). The epithelial phenotype of cancer cells and their ability to form physical barriers represent a protective mechanism for cancer cells. We have analyzed more than 60 primary and metastatic breast cancer samples and found overexpression of DSG2 in all of these samples. A fraction of these biopsies are shown in Supplementary Fig. S1.

JO-1 binding to DSG2 on tumor cells triggers pathways involved in epithelial-to-mesenchymal transition (EMT), a process which has been associated with tumor metastasis. However, to date none of our in vivo studies has shown any evidence of increased tumor growth or metastasis after treatment with JO-1. Furthermore, at day 3 after JO-1 injection into mice-bearing Her2/neu-positive HCC1954 tumors, there was no significant increase in the percentage of circulating Her2/neu-positive cells in the blood (Supplementary Fig. S7). Tumor metastasis requires more than transient activation of EMT pathways. Detachment from epithelial cancers and migration of tumor cells is only possible after long-term cross-talk between malignant cells and the tumor microenvironment, resulting in changes in the tumor stroma and phenotypic reprogramming of epithelial cells into mesenchymal cells (24).

Except for a mild, transient diarrhea, intravenous injection of JO-1 (at a dose of 2 mg/kg) had no critical side effects on other tissues or hematologic parameters in hDSG2-transgenic mice. We speculate that DSG2 in tissues other than the tumor and certain epithelial cells in the gastrointestinal tract is not accessible to intravenously injected JO-1. Clearly, the demonstration of safety of JO-1 in combination with chemotherapy drugs in non-human primates is required before this approach can be considered in humans. In this context, we have recently reported the DSG2 in macaques is expressed in the same patterns as in humans and that it is recognized by JO-1 (12).

Overall, we showed that JO-1 improves the efficacy and safety of a number of cancer chemotherapy drugs, implying that JO-1 cotherapy has the potential to improve the life quality of patients that receive cancer treatment. Moreover, it may allow patients to continue chemotherapy, which has been halted or delayed due to toxicity. Finally, cancer treatment with newer drugs, such as nab-paclitaxel and liposomal doxorubicin is expensive and presents a substantial burden to both patients and the costs of healthcare in general. JO-1 has potential to decrease costs of cancer therapy by reducing the amount of chemotherapy required as well as costs associated with treatment-related toxicity. These savings are likely to exceed the additional cost of JO-1 treatment.

No potential conflicts of interest were disclosed.

The authors thank Melanie Wurm for helpful advice.

The study was supported by NIH grant R01 CA080192 (A. Lieber), R01 HLA078836 (A. Lieber), the Pacific Ovarian Cancer Research Consortium/Specialized Program of Research Excellence in Ovarian Cancer grant P50 CA83636, by a grant from the Fred Hutchinson Cancer Research Center, and a grant from the Marsha Rivkin Center for Ovarian Cancer Research. I. Beyer is a recipient of a postdoctoral fellowship award from “Deutsche Krebshilfe” (108988).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.