Determination of molecules regulating gene delivery using adenoviral vectors in ovarian carcinomas

Understanding and addressing barriers to successful adenovirus-based virotherapy for ovarian cancer

Ovarian cancer is the leading cause of death among women with gynecological cancer, with an overall 5-year survival rate below 50% due to a lack of specific symptoms, late stage at time of diagnosis and a high rate of recurrence after standard therapy. A better understanding of heterogeneity, genetic mutations, biological behavior and immunosuppression in the tumor microenvironment have allowed the development of more effective therapies based on anti-angiogenic treatments, PARP and immune checkpoint inhibitors, adoptive cell therapies and oncolytic vectors. Oncolytic adenoviruses are commonly used platforms in cancer gene therapy that selectively replicate in tumor cells and at the same time are able to stimulate the immune system. In addition, they can be genetically modified to enhance their potency and overcome physical and immunological barriers. In this review we highlight the challenges of adenovirus-based oncolytic therapies targeting ovarian cancer and outline recent advances to improve their potential in combination with immunotherapies.

Paclitaxel resistance increases oncolytic adenovirus efficacy via upregulated CAR expression and dysfunctional cell cycle control

Resistance to paclitaxel chemotherapy frequently develops in ovarian cancer. Oncolytic adenoviruses are a novel therapy for human malignancies that are being evaluated in early phase trials. However, there are no reliable predictive biomarkers for oncolytic adenovirus activity in ovarian cancer. We investigated the link between paclitaxel resistance and oncolytic adenovirus activity using established ovarian cancer cell line models, xenografts with de novo paclitaxel resistance and tumour samples from two separate trials. The activity of multiple Ad5 vectors, including dl922-947 (E1A CR2-deleted), dl1520 (E1B-55K deleted) and Ad5 WT, was significantly increased in paclitaxel resistant ovarian cancer in vitro and in vivo. This was associated with greater infectivity resulting from increased expression of the primary receptor for Ad5, CAR (coxsackie adenovirus receptor). This, in turn, resulted from increased CAR transcription secondary to histone modification in resistant cells. There was increased CAR expression in intraperitoneal tumours with de novo paclitaxel resistance and in tumours from patients with clinical resistance to paclitaxel. Increased CAR expression did not cause paclitaxel resistance, but did increase inflammatory cytokine expression. Finally, we identified dysregulated cell cycle control as a second mechanism of increased adenovirus efficacy in paclitaxel-resistant ovarian cancer. Ad11 and Ad35, both group B adenoviruses that utilise non-CAR receptors to infect cells, are also significantly more effective in paclitaxel-resistant ovarian cell models. Inhibition of CDK4/6 using PD-0332991 was able both to reverse paclitaxel resistance and reduce adenovirus efficacy. Thus, paclitaxel resistance increases oncolytic adenovirus efficacy via at least two separate mechanisms - if validated further, this information could have future clinical utility to aid patient selection for clinical trials.

Species D human adenovirus type 9 exhibits better virus-spread ability for antitumor efficacy among alternative serotypes

Species C human adenovirus serotype 5 (HAdV-C5) is widely used as a vector for cancer gene therapy, because it efficiently transduces target cells. A variety of HAdV-C5 vectors have been developed and tested in vitro and in vivo for cancer gene therapy. While clinical trials with HAdV-C5 vectors resulted in effective responses in many cancer patients, administration of HAdV-C5 vectors to solid tumors showed responses in a limited area. A biological barrier in tumor mass is considered to hinder viral spread of HAdV-C5 vectors from infected cells. Therefore, efficient virus-spread from an infected tumor cell to surrounding tumor cells is required for successful cancer gene therapy. In this study, we compared HAdV-C5 to sixteen other HAdV serotypes selected from species A to G for virus-spread ability in vitro. HAdV-D9 showed better virus-spread ability than other serotypes, and its viral progeny were efficiently released from infected cells during viral replication. Although the HAdV-D9 fiber protein contains a binding site for coxsackie B virus and adenovirus receptor (CAR), HAdV-D9 showed expanded tropism for infection due to human CAR (hCAR)-independent attachment to target cells. HAdV-D9 infection effectively killed hCAR-negative cancer cells as well as hCAR-positive cancer cells. These results suggest that HADV-D9, with its better virus-spread ability, could have improved therapeutic efficacy in solid tumors compared to HAdV-C5.

Adenovirus serotype 11 causes less long-term intraperitoneal inflammation than serotype 5: implications for ovarian cancer therapy

In a phase II/III clinical trial intraperitoneal (i.p.) administration of a group C adenovirus vector (Ad5) caused bowel adhesion formation, perforation and obstruction. However, we had found that i.p. group B, in contrast to group C adenoviruses, did not cause adhesions in nude BALB/c ovarian cancer models, prompting further investigation. Ex vivo, group B Ad11 caused lower inflammatory responses than Ad5 on BALB/c peritoneal macrophages. In vivo, i.p. Ad11 triggered short-term cytokine and cellular responses equal to Ad5 in both human CD46-positive and -negative mice. In contrast, in a long-term study of repeated i.p. administration, Ad11 caused no/mild, whereas Ad5 induced moderate/severe adhesions and substantial liver toxicity accompanied by elevated levels of IFNγ and VEGF and loss of i.p. macrophages, regardless of CD46 expression. It appears that, although i.p. Ad11 evokes immediate inflammation similar to Ad5, repeated administration of Ad11 is better tolerated and long-term fibrotic tissue remodelling is reduced.

A simple detection system for adenovirus receptor expression using a telomerase-specific replication-competent adenovirus

Adenovirus serotype 5 (Ad5) is frequently used as an effective vector for induction of therapeutic transgenes in cancer gene therapy or of tumor cell lysis in oncolytic virotherapy. Ad5 can infect target cells through binding with the coxsackie and adenovirus receptor (CAR). Thus, the infectious ability of Ad5-based vectors depends on the CAR expression level in target cells. There are conventional methods to evaluate the CAR expression level in human target cells, including flow cytometry, western blotting and immunohistochemistry. Here, we show a simple system for detection and assessment of functional CAR expression in human tumor cells, using the green fluorescent protein (GFP)-expressing telomerase-specific replication-competent adenovirus OBP-401. OBP-401 infection induced detectable GFP expression in CAR-expressing tumor cells, but not in CAR-negative tumor cells, nor in CAR-positive normal fibroblasts, 24 h after infection. OBP-401-mediated GFP expression was significantly associated with CAR expression in tumor cells. OBP-401 infection detected tumor cells with low CAR expression more efficiently than conventional methods. OBP-401 also distinguished CAR-positive tumor tissues from CAR-negative tumor and normal tissues in biopsy samples. These results suggest that GFP-expressing telomerase-specific replication-competent adenovirus is a very potent diagnostic tool for assessment of functional CAR expression in tumor cells for Ad5-based antitumor therapy.

Combination of a MDR1-targeted replicative adenovirus and chemotherapy for the therapy of pretreated ovarian cancer

PURPOSE

Targeted oncolytic adenoviruses capable of replication selectively in cancer cells are an appealing approach for the treatment of various cancer types refractory to conventional therapies. The aim of this study was to evaluate the effect of Ad5/3MDR1E1, a multidrug resistance gene 1 (MDR1)-targeted fiber-modified replication-competent adenovirus for the therapy of platinum-pretreated ovarian cancer in combination with cytostatic agents.

METHODS

MDR1-specific tumor cell killing of Ad5/3MDR1E1 was systematically evaluated in chemotherapy naïve and pretreated ovarian cancer cells in vitro. Combinations of Ad5/3MDR1E1 and cytostatic agents were studied in vivo and in vitro. An in vivo hepatotoxicity model was used to evaluate liver toxicity.

RESULTS

We demonstrate efficient oncolysis of Ad5/3MDR1E1 in chemotherapy-resistant ovarian cancer cells as well as therapeutic efficacy in an orthotopic mouse model. Further, combining Ad5/3MDR1E1 with paclitaxel resulted in greater therapeutic benefit than either agent alone.

CONCLUSION

These preclinical data suggest that a fiber-modified adenovirus vector under the control of the MDR1 promoter represents a promising treatment strategy for platinum-pretreated ovarian cancer as a single agent or in combination with conventional anticancer drugs.

Treatment of chemotherapy resistant ovarian cancer with a MDR1 targeted oncolytic adenovirus

OBJECTIVE

Multidrug resistance gene 1 (MDR1) mediated resistance to chemotherapeutic agents is a major obstacle for the therapy of various cancer types. The use of conditionally replicating adenoviruses (CRAds) is dependent on molecular differences between tumor cells and non tumor cells. Transcriptional targeting of CRAd replication is an effective way to control replication regulation. The aim of this study was to evaluate the effect of a MDR1 targeted fiber-modified CRAd against chemotherapy resistant ovarian cancer.

METHODS

MDR1 expression was evaluated in chemotherapy naïve and pretreated ovarian cancer cells and various control cells. We constructed 2 variants of a fiber-modified CRAd, Ad5/3MDR1E1 and Ad5/3MDR1E1∆24 containing the MDR1 promoter to control viral replication via the E1A gene. The MDR promoter activity and cell killing efficacy were evaluated in vitro. Orthotopic murine models of peritoneally disseminated ovarian cancer were utilized to evaluate the preclinical efficacy of MDR targeted CRAds in vivo. To evaluate the liver toxicity of MDR1 targeted CRAds, we compared Ad5/3MDR1E1 with Ad5/3∆24, a CRAd that replicates in cancer cells inactive in the Rb/p16 pathway by use of an in vivo hepatotoxicity model.

RESULTS

We demonstrate efficient oncolysis of Ad5/3MDR1E1 in both chemotherapy resistant ovarian cancer cell lines and in primary tumor cells from pretreated patients as well as therapeutic efficacy in an orthotopic mouse model. Ad5/3MDR1E1 demonstrated significantly decreased liver toxicity compared to other 5/3-fiber modified control vectors examined.

CONCLUSIONS

In summary, Ad5/3MDR1E1 is an efficient and safe gene therapy approach for specific targeting of chemotherapy resistant cancer cells.

An adenoviral vector expressing human adenovirus 5 and 3 fiber proteins for targeting heterogeneous cell populations

Human adenovirus serotype 5 (HAdV-5) attaches to its primary receptor, the coxsackie and adenovirus receptor (CAR) as the first step of infection. However, CAR expression decreases as tumors progress, thereby diminishing the utility of HAdV-5-based vectors for cancer therapy. In contrast, many aggressive tumor cells highly express CD46, a cellular receptor for HAdV-3. We hypothesized that a mosaic HAdV vector, containing two kinds of fiber proteins, would provide extensive transduction in a heterogeneous population of tumor cells with varying expression levels of HAdV receptors. We therefore generated a fiber-mosaic HAdV vector displaying both a chimeric HAdV-3 fiber and the HAdV-5 fiber protein. We verified the structural integrity of purified viral particles and confirmed that the fiber-mosaic HAdV vector has expanded tropism. We conclude that the use of fiber-mosaic HAdV vectors is a promising approach for transducing a heterogeneous cell population with different expression levels of adenovirus receptors.

p21 Promotes oncolytic adenoviral activity in ovarian cancer and is a potential biomarker

The oncolytic adenovirus dl922-947 replicates selectively within and lyses cells with a dysregulated Rb pathway, a finding seen in > 90% human cancers. dl922-947 is more potent than wild type adenovirus and the E1B-deletion mutant dl1520 (Onyx-015). We wished to determine which host cell factors influence cytotoxicity. SV40 large T-transformed MRC5-VA cells are 3-logs more sensitive to dl922-947 than isogenic parental MRC5 cells, confirming that an abnormal G1/S checkpoint increases viral efficacy. The sensitivity of ovarian cancer cells to dl922-947 varied widely: IC₅₀ values ranged from 51 (SKOV3ip1) to 0.03 pfu/cell (TOV21G). Cells sensitive to dl922-947 had higher S phase populations and supported earlier E1A expression. Cytotoxicity correlated poorly with both infectivity and replication, but well with expression of p21 by microarray and western blot analyses. Matched p21+/+ and -/- Hct116 cells confirmed that p21 influences dl922-947 activity in vitro and in vivo. siRNA-mediated p21 knockdown in sensitive TOV21G cells decreases E1A expression and viral cytotoxicity, whilst expression of p21 in resistant A2780CP cells increases virus activity in vitro and in intraperitoneal xenografts. These results highlight that host cell factors beyond simple infectivity can influence the efficacy of oncolytic adenoviruses. p21 expression may be an important biomarker of response in clinical trials.

Role of cell surface molecules and autologous ascitic fluid in determining efficiency of adenoviral transduction of ovarian cancer cells

Adenovirus is the most frequently used virus in gene therapy clinical trials. There have been conflicting reports on the ability of adenovirus to transduce primary ovarian cancer samples and the expression of relevant cell surface molecules. These factors were examined using primary ovarian cancer cells cultured from ascites and solid tumor to gain insights into the clinical use of adenovirus in ovarian cancer. The level of transduction of primary cultures was much higher than uncultured cells and established cell lines, and correlated with higher levels of coxsackie-adenovirus receptor (CAR) and integrin expression. Growth of primary cultures in autologous ascitic fluid prevented an increase in CAR expression and inhibited transduction compared with cells treated in supplemented RPMI. Cells at the periphery of solid tumor samples were transduced using a replication-incompetent virus and correlated with CAR expression. However, transduction was abolished by autologous ascitic fluid, despite the expression of CAR. We conclude that the use of adenoviruses for ovarian cancer gene therapy will require testing in the presence of inhibitory factors in ascitic fluid. The clinical use of adenoviral vectors may require circumvention of such inhibitory factors and the use of replication competent adenovirus to enable efficient viral penetration of the cancer.

Dilated cardiomyopathy alters the expression patterns of CAR and other adenoviral receptors in human heart

Gene therapy trials for heart failure have demonstrated the key role of efficient gene transfer in achieving therapeutic efficacy. An attractive approach to improve adenoviral gene transfer is to use alternative virus serotypes with modified tropism. We performed a detailed analysis of cardiac expression of receptors for several adenovirus serotypes with a focus on differential expression of CAR and CD46, as adenoviruses targeting these receptors have been used in various applications. Explanted hearts from patients with DCM and healthy donors were analyzed using Q-RT-PCR, western blot and immunohistochemistry. Q-RT-PCR and Western analyses revealed robust expression of all receptors except CD80 in normal hearts with lower expression levels in DCM. Immunohistochemical analyses demonstrated that CD46 expression was somewhat higher than CAR both in normal and DCM hearts with highest levels of expression in intramyocardial coronary vessels. Total CAR expression was upregulated in DCM. Triple staining on these vessels demonstrated that both CAR and CD46 were confined to the subendothelial layer in normal hearts. The situation was clearly different in DCM, where both CAR and CD46 were expressed by endothelial cells. The induction of expression of CAR and CD46 by endothelial cells in DCM suggests that viruses targeting these receptors could more easily gain entry to heart cells after intravascular administration. This finding thus has potential implications for the development of targeted gene therapy for heart failure.

Cetuximab retargeting of adenovirus via the epidermal growth factor receptor for treatment of intraperitoneal ovarian cancer

Gene and virotherapy of ovarian cancer, using type 5 adenovirus (Ad5), has demonstrated good activity in preclinical animal studies, particularly after intraperitoneal administration of virus; however, success in clinical trials has been limited by poor infectivity of ovarian cancer cells and inflammatory responses to Ad5. We previously demonstrated that covalent modification of Ad5 with reactive copolymers on the basis of poly(hydroxypropylmethacrylamide) can shield the virus, offering protection from neutralizing antibodies and enabling retargeting to cancer-upregulated receptors with peptide ligands (basic fibroblast growth factor [bFGF] and murine epidermal growth factor [EGF]). These ligands may be less than ideal for clinical use, however, because they are potential mitogens. Accordingly, in this study we investigated the use of an anti-EGF receptor (EGFR) antibody, cetuximab, to retarget adenoviral transduction of EGFR-positives in vitro and in vivo. Cetuximab retargeting altered the physicochemical characteristics of Ad5, although it did not cause particle aggregation. Although cetuximab stimulated internalization of EGFR, similarly to EGF, it inhibited EGFR phosphorylation. Adenoviral transduction was inhibited after polymer coating, but was rescued in EGFR-positive cells (and not in EGFR-negative cells) by cetuximab retargeting. Cetuximab retargeting of wild-type adenovirus serotype 5 (Ad5WT) prolonged survival in an animal model of human ovarian cancer, similar to unmodified Ad5WT, but polymer coating ameliorated stimulation of adhesion formation. We conclude that polymer coating and covalent attachment of cetuximab successfully retargeted adenovirus to EGFR-positive cells, retained in vivo efficacy of an oncolytic adenovirus, and ameliorated side effects caused by unmodified adenovirus.

Oncolytic adenoviral gene therapy in ovarian cancer: why we are not wasting our time

Preclinical gene-therapy studies in the past 15 years have repeatedly raised hopes that we were about to enter a brave new era. However, many clinical trials have disappointed. For tumor types with poor response rates to first-line conventional cytotoxic chemotherapy and/or high rates of chemorefractory disease, there remain very few treatment options. In this article we review gene therapy within the context of ovarian cancer. We examine why clinical data have been discouraging and discuss how the lessons learned from earlier trials are being applied to current research.

Directed evolution generates a novel oncolytic virus for the treatment of colon cancer

Background

Viral-mediated oncolysis is a novel cancer therapeutic approach with the potential to be more effective and less toxic than current therapies due to the agents selective growth and amplification in tumor cells. To date, these agents have been highly safe in patients but have generally fallen short of their expected therapeutic value as monotherapies. Consequently, new approaches to generating highly potent oncolytic viruses are needed. To address this need, we developed a new method that we term “Directed Evolution” for creating highly potent oncolytic viruses.

Methodology/Principal Findings

Taking the “Directed Evolution” approach, viral diversity was increased by pooling an array of serotypes, then passaging the pools under conditions that invite recombination between serotypes. These highly diverse viral pools were then placed under stringent directed selection to generate and identify highly potent agents. ColoAd1, a complex Ad3/Ad11p chimeric virus, was the initial oncolytic virus derived by this novel methodology. ColoAd1, the first described non-Ad5-based oncolytic Ad, is 2–3 logs more potent and selective than the parent serotypes or the most clinically advanced oncolytic Ad, ONYX-015, in vitro. ColoAd1's efficacy was further tested in vivo in a colon cancer liver metastasis xenograft model following intravenous injection and its ex vivo selectivity was demonstrated on surgically-derived human colorectal tumor tissues. Lastly, we demonstrated the ability to arm ColoAd1 with an exogenous gene establishing the potential to impact the treatment of cancer on multiple levels from a single agent.

Conclusions/Significance

Using the “Directed Evolution” methodology, we have generated ColoAd1, a novel chimeric oncolytic virus. In vitro, this virus demonstrated a >2 log increase in both potency and selectivity when compared to ONYX-015 on colon cancer cells. These results were further supported by in vivo and ex vivo studies. Furthermore, these results have validated this methodology as a new general approach for deriving clinically-relevant, highly potent anti-cancer virotherapies.

Optimization of conditionally replicative adenovirus for pancreatic cancer and its evaluation in an orthotopic murine xenograft model

BACKGROUND

The full realization of the therapeutic potential of conditionally replicative adenoviruses (CRAds) in the field of pancreatic cancer has been hindered by limited tumor transduction and suboptimal replication control.

METHODS

We optimized infectivity enhancements and tumor-specific promoters (tsps) for pancreatic cancer. Infectivity was enhanced both by incorporating an RGD motif and by substituting the knob region with Ad serotype 3 knob (Ad5/Ad3). An optimized CRAd was tested in an orthotopic pancreatic cancer model by systemic administration.

RESULTS

Among a panel of 8 tsps, the 1.5-kb cyclooxygenase-2 (Cox-2L) promoter profile was most advantageous in the pancreatic cancer cell lines, whereas 4 more promoters were also promising. An infectivity-enhanced Ad5/Ad3 CRAd controlled with Cox-2L promoter was found to safely exhibit replication within a tumor in this model and was found to suppress tumor growth after systemic delivery.

CONCLUSIONS

The infectivity-enhanced, promoter-controlled CRAd promises useful clinical applications for pancreatic cancer gene therapy.

Virotherapy of ovarian cancer with polymer-cloaked adenovirus retargeted to the epidermal growth factor receptor

Adenovirus gene therapy for intraperitoneal (IP) cancer is limited in clinical trials by inefficient tumor cell transduction and development of peritoneal adhesions. We have shown previously that normal virus tropism can be ablated by physically shielding the virus surface with reactive hydrophilic polymers and that linkage of novel ligands enables virus "retargeting" through chosen receptors. To achieve tumor-selective infection, polymer-coated virus was retargeted using murine epidermal growth factor (mEGF). The resulting mEGF-polymer coated adenovirus lost its normal broad tropism and transduced cells selectively via the EGF receptor (EGFR). We assessed whether this approach could be used to target lytic "virotherapy" using wild-type adenovirus (Ad5WT) in a peritoneal xenograft model of human ovarian cancer. Oncolytic activity of Ad5WT was retained following polymer coating and mEGF-retargeting. Importantly, adhesion formation was markedly decreased compared with the unmodified virus, and no dose-limiting toxicities were observed following treatment with mEGF-retargeted polymer-coated virus. Restricting virus tropism by physical coating, coupled with tumor-selective retargeting promises to combine good anticancer efficacy with acceptable toxicity, enabling application of elevated virus doses leading to an improved therapeutic outcome.

Soluble isoforms but not the transmembrane form of coxsackie-adenovirus receptor are of clinical relevance in epithelial ovarian cancer

The coxsackie-adenovirus receptor (hCAR) has been extensively studied in context of adenoviral-based gene therapy for cancer. However, there is strong evidence that besides its decisive role in coxsackie and adenovirus cell-entry, hCAR is a component of epithelial tight junctions and involved in cell-cell adhesions in normal and cancer cells. Furthermore, this adhesion molecule behaves like a cell surface receptor endowed with tumor suppressive properties via signal transduction. Moreover, 3 truncated soluble isoforms of hCAR were recently identified. We investigated the quantitative expression of all known CAR isoforms in a training set of 140 ovarian cancer samples and 21 controls by RT-PCR. The expression levels of the various isoforms were compared with clinicopathologic parameters and their prognostic significance was assessed. Expression levels of all CAR isoforms were elevated in ovarian carcinomas as compared with those of non-malignant controls. mRNA-expression correlated with protein levels. Moreover, expression of the soluble isoforms CAR 3/7 and CAR 4/7 but not that of hCAR was significantly increased in advanced ovarian cancer as revealed by a highly significant correlation with FIGO stage and residual disease > 2 cm in diameter after debulking surgery. High expression of CAR 3/7 and 4/7 was shown to be of independent prognostic relevance for progression-free (CAR 4/7) and overall survival (CAR 3/7 and CAR 4/7). In conclusion, soluble CAR isoforms 3/7 and 4/7 may play a pivotal role in ovarian cancer biology, possibly by counteracting migration- and growth-inhibitory properties of the membranous hCAR and thus favoring cancer cell dissemination throughout the peritoneal cavity.

A conditionally replicating adenovirus with strict selectivity in killing cells expressing epidermal growth factor receptor

Virotherapy of cancer using oncolytic adenoviruses has shown promise in both preclinical and clinical settings. One important challenge to reach the full therapeutic potential of oncolytic adenoviruses is accomplishing efficient infection of cancer cells and avoiding uptake by normal tissue through tropism modification. Towards this goal, we constructed and characterized an oncolytic adenovirus, carrying mutated capsid proteins to abolish the promiscuous adenovirus native tropism and encoding a bispecific adapter molecule to target the virus to the epidermal growth factor receptor (EGFR). The new virus displayed a highly selective targeting profile, with reduced infection of EGFR-negative cells and efficient killing of EGFR-positive cancer cells including primary EGFR-positive osteosarcoma cells that are refractory to infection by conventional adenoviruses. Our method to modify adenovirus tropism might thus be useful to design new oncolytic adenoviruses for more effective treatment of cancer.

Head and neck cancer cells are efficiently infected by Ad5/35 hybrid virus

BACKGROUND

Clinical gene therapy trials using standard Ad5-based vectors have thus far demonstrated limited efficacy, most likely due to low expression levels of adenoviral receptors on tumor cells. We sought to analyze adenoviral receptor expression levels on primary head and neck squamous cell carcinoma (HNSCC) cells and to determine whether adenoviral retargeting to the CD46 receptor via the Ad5/35 system would increase therapeutic potential for HNSCC.

METHODS

We used flow cytometric analyses to determine adenoviral receptor expression levels on nine primary HNSCC cells collected from cancer patients. Adenoviruses Ad5.LacZ and Ad5/35.LacZ were used to analyze the differences in viral transduction both in vitro and in a HNSCC tumor mouse model.

RESULTS

Flow cytometric analyses demonstrated uniformly high CD46 expression in all cells studied (85-99%). In contrast, coxsackievirus and adenovirus receptor (CAR) expression was substantially lower and highly variable (1.6-62%). Alpha(v) integrin expression was between 39-98%. In situ stainings for beta-galactosidase gene expression demonstrated that Ad5/35.LacZ was clearly more effective than Ad5.LacZ in transducing primary HNSCC cells. Quantification of beta-galactosidase expression revealed up to 65 times higher transgene expression from Ad5/35.LacZ than Ad5.LacZ. In vivo, beta-galactosidase expression was detected in a substantial area after a single intratumoral injection of Ad5/35.LacZ, whereas injection with Ad5.LacZ resulted in gene expression only in a few cells.

CONCLUSIONS

Our results demonstrate that the low and variable CAR expression levels limit the therapeutic efficacy of Ad5-based strategies for HNSCC. In contrast, the effective in vivo transduction capacity of Ad5/35 warrants further development of this vector for the treatment of head and neck cancer.

Adenovirus type 5 E1A gene therapy for ovarian clear cell carcinoma: a potential treatment strategy

Resistance of ovarian clear cell carcinoma (CCC) to platinum-based chemotherapy is associated with poor prognosis, and an effective treatment for advanced disease is urgently needed. HER2/neu is up-regulated more often in CCC than in other histologic types of epithelial ovarian cancer. The purpose of this study was to assess possible treatment for ovarian CCC with the anti-HER2 antibody trastuzumab or human adenovirus type 5 E1A. We treated 10 CCC cell lines with trastuzumab or E1A and assessed cell viability, proliferation, and colony formation and the expression of HER2 and wild-type p53 proteins and molecules downstream of those signaling pathways. HER2 protein was detected at various levels in all 10 cell lines by Western blotting and in 5 CCC cell lines by immunohistochemical staining; HER2 gene amplification was detected (by fluorescence in situ hybridization) in only one cell line (RMG-I). Trastuzumab did not inhibit proliferation in any of the four CCC cell lines tested (RMG-I, SKOV-2, OVTOKO, and OVSAYO). However, transfection with E1A (as compared with control vectors) reduced colony formation in all 10 CCC cell lines regardless of HER2 expression level. Infection of RMG-I and SMOV-2 cells with an adenoviral vector encoding E1A led to significant (P < 0.05) suppression of proliferation and enhancement of cell death; this effect required stabilization of p53 (but not p73) protein and was associated with the up-regulation of Bax and the cleavage of caspase-9. Other mechanisms, such as p53-independent apoptosis, may also be involved in E1A-mediated cell death in CCC. Finally, treatment with E1A prolonged survival in a CCC xenograft model (P < 0.001). E1A gene therapy, because of its ability to stabilize wild-type p53, is worth exploring as a treatment modality for women with ovarian CCC, which typically expresses wild-type p53.

Deletion of the intracellular domain of coxsackie and adenovirus receptor (CAR) enhances the expression of itself and boosts the efficiency of current adenovirus-mediated gene therapy in ovarian cancer cell lines in vitro

The failure of adenovirus-mediated gene therapy often derives from the absence of coxsackie and adenovirus receptor (CAR) expression in target cells. We hypothesize that the slight up-regulation of CAR expression might boost the effect of adenovirus-mediated gene therapy in ovarian cancer. To test this hypothesis, we transfected full-length and intracellular-domain-deleted (tailless) CAR plasmids into CAR-deficient ovarian cancer cell line SKOV3. We observed significant elevations of the in vitro killing effect of Adv-TK and oncolytic adenovirus-mediated cytopathic effect (CPE) in transfected sub-clones, and tailless-transfected SKOV3 showed higher CAR expressions than full-length CAR-transfected cells. We conclude that the extracellular domain of CAR is essential for adenovirus-based gene therapy and, furthermore, that its intracellular domain might play an important role in the regulation of its own expression.

p53 as a target for anti-cancer drug development

Loss of p53 function compromises genetic homeostasis in cells exhibiting deregulated DNA replication and/or DNA damage, and prevents normal cytotoxic responses to cancer therapies. Genetic and pharmacological approaches are being developed with the ultimate goal of restoring or controlling p53 functions in cancer patients. Progress has recently been made in the clinical use of replication-deficient virus carrying wt-TP53 (Ad5CMV-p53) and/or cancer-selective oncolytic adenoviruses (ONYX-015). These strategies demonstrated clinical activity as monotherapy and were synergistic with traditional chemotherapy agents in the treatment of some types of cancer. In addition, pharmacological methods are under development to either stimulate wild-type p53 protein function, or induce p53 mutant proteins to resume wild-type functions. These methods are based on small chemicals (CP-31388, PRIMA-1), peptides (CDB3) or single-chain Fv antibody fragments corresponding to defined p53 domains. Here, we discuss the mechanisms underlying these approaches and their perspectives for cancer therapy.

Sodium iodide symporter-mediated radioiodide imaging and therapy of ovarian tumor xenografts in mice

Ovarian cancer represents the fifth leading cause of cancer death among women in the United States, with >16 000 deaths expected this year. This study was carried out to investigate the potential of sodium iodide symporter (NIS)-mediated radioiodide therapy as a novel approach for ovarian cancer treatment. Radioiodide is routinely and effectively used for the treatment of benign and malignant thyroid disease as a result of native thyroidal expression of NIS, which mediates iodide uptake. In vitro gene transfer studies in ovarian cancer cells revealed a 12- and five-fold increase in iodide uptake when transduced with Ad/CMV/NIS or Ad/MUC1/NIS, respectively. Western blot/immunohistochemistry confirmed NIS protein expression. In vivo ovarian tumor xenografts were infected with the adenoviral constructs. (123)I imaging revealed a clear image of the CMV/NIS-transduced tumor, with a less intense image apparent following infection with MUC1/NIS. Therapeutic doses of (131)I following CMV/NIS infection caused a mean 53% reduction in tumor volume (P<0.0001). MUC1/NIS-transduced tumors did not regress, although at 8 weeks following therapy, tumor volume was significantly less that of control animals (166 versus 332%, respectively, P<0.05). This study represents a promising first step investigating the potential for NIS-mediated radioiodide imaging and therapy of ovarian tumors.

The effect of hypoxia on the uptake, replication and lytic potential of group B adenovirus type 3 (Ad3) and type 11p (Ad11p)

Replicating, tumor selective viruses are being tested as potential treatments for human cancers. Hypoxia is a pathophysiological cancer condition that alters the lytic potential of the replication-competent adenovirus serotype 5 (Ad5) virus by a mechanism independent of receptor levels or internalization rates. We extend these initial studies to examine the potential effects of hypoxia on the group B adenoviruses (Ads), adenovirus type 3 (Ad3) (group B1) and adenovirus type 11p (Ad11p) (group B2). Receptor expression (CD46) is not altered by hypoxia. However, the lytic potential is compromised in a cell-dependent fashion. Consequently, our study suggests that group B replicating Ad-based treatments, like the group C Ad-5-based viruses, will need to be modified in order to effectively treat hypoxic components of human tumors.

Variable expression of coxsackie-adenovirus receptor in thyroid tumors: implications for adenoviral gene therapy

Adenoviral gene therapy represents a novel approach for the treatment of aggressive thyroid carcinomas. Both coxsackie-adenovirus receptor (CAR) and integrins have been shown to be the major determinants for adenoviral infectivity in many types of cancer cells, yet conflicting results have been reported. In this report we examine these factors mediating adenoviral infection in thyroid cells and to evaluate CAR expression in various types of thyroid cancer. We found that neither expression levels of CAR nor integrins are solely predictive of adenoviral infectivity in thyroid cells. However, the absence of CAR was associated with poor adenoviral infectivity in immortalized rat FRTL-5 cells. Moreover, preincubation with alpha-CAR antibody decreased infectivity in FTC 238 cells, a human thyroid tumor line. These results indicate that CAR does play a role in adenoviral infection of thyroid cells. Immunohistochemical analysis revealed that CAR is expressed at the cell surface in the majority of malignant thyroid tumors. We further show that adenoviral infectivity in some thyroid cancer cells can be improved by poly-L-lysine. Our study warrants a functional method to evaluate adenoviral infectivity should be developed and instituted prior to clinical trials of adenoviral gene therapy in patients with advanced thyroid cancer.

The tumor suppressor gene TP53: implications for cancer management and therapy

The p53 protein is an inducible transcription factor with multiple anti-proliferative roles in response to genotoxic damage; unprogrammed proliferative stimuli; and deprivation of oxygen, nutrients, or ribonucleotides. Inactivation of the TP53 gene by mutation or deletion is the most common event in human cancer. Loss of p53 function compromises genetic homeostasis in cells exposed to mutagens and prevents normal cytotoxic responses to cancer therapies. Genetic and pharmacological approaches are being developed with the ultimate goal of restoring or controlling p53 functions in cancer patients. Genetic interventions aiming at expressing wild-type TP53 in cancer cells, either by retroviral or adenoviral transfer, have met limited clinical success. However, recently, the use of a defective adenovirus (ONYX-015) that selectively kills p53-incompetent cells has shown promising effects in pre-clinical and clinical studies. Pharmacological methods are under development to either stimulate wild-type p53 protein function or induce p53 mutant proteins to resume wild-type functions. These methods are based on small chemicals (CP-31388, PRIMA-1), peptides (CDB3), or single-chain Fv antibody fragments corresponding to defined p53 domains. In addition, detection of mutant TP53 may also serve as a marker for early cancer detection, prediction, and prognosis. In this review, we discuss the mechanisms underlying these approaches and their perspectives for cancer therapy.

Histone deacetylase inhibitors upregulate expression of the coxsackie adenovirus receptor (CAR) preferentially in bladder cancer cells

Studies on bladder cancer cell lines have shown that low adenoviral (Ad) infectivity is associated with low-level coxsackie adenovirus receptor (CAR) expression. Recently, we and others demonstrated a tumor stage- and grade-dependent downregulation of CAR expression in a large series of clinical bladder cancer specimens. Here, we demonstrate adenoviral gene transfer can be markedly enhanced in bladder cancer cells by upregulation of CAR through the use of certain differentiating agents, including the histone deacetylase inhibitors (HDACI) trichostatin A and sodium phenylbutyrate. CAR upregulation to supraphysiologic levels was demonstrated by quantitative rt-PCR, Western blotting, flow cytometry and adenoviral gene transfer. Normal urothelial cells and CAR-positive papilloma cells (RT4) failed to demonstrate upregulation under the same conditions. Upregulation was cell cycle dependent, associated with increased adenoviral gene transfer and persisted for at least 7 days after a single treatment. Such upregulation, however, appears to be tumor cell specific, as other CAR-negative cell lines failed to demonstrate enhanced adenoviral gene transfer with the same treatments. These results provide a rational basis for combining HDACI therapy with gene therapy as a method of augmenting activity in bladder cancer, but this strategy may not be universally applicable to other cell types.

A Fiber-Modified, Secretory Leukoprotease Inhibitor Promoter-Based Conditionally Replicating Adenovirus for Treatment of Ovarian Cancer

Purpose: The use of conditionally replicating adenoviruses (CRAD) is dependent on molecular differences between tumor cells and nontumor cells. Transcriptional targeting of CRAD replication via tumor-specific promoters is an effective way to control replication regulation. Genetic fiber pseudotyping is an approach for circumventing low expression of the primary adenovirus serotype 5 (Ad5) receptor by using the distinct adenovirus serotype 3 (Ad3) receptor for entry into and subsequent killing of ovarian cancer cells.

Experimental Design: In this study, we constructed a fiber-modified CRAD containing the secretory leukoprotease inhibitor (SLPI) promoter to control viral replication via the E1A gene (Ad5/3SLPI). To evaluate the liver toxicity of chimeric 5/3 fiber-modified CRADs, we compared Ad5/3SLPI with Ad5/3Cox-2L, a CRAD with E1A under control of the Cox-2 promoter, and Ad5/3Δ24, a CRAD that replicates in cancer cells inactive in the retinoblastoma/p16 pathway by use of an in vivo hepatotoxicity model and by a model system that uses slices of human liver.

Results: We show efficient viral replication and oncolysis of Ad5/3SLPI in both multiple ovarian cancer cell lines and primary tumor cell spheroids as well as therapeutic efficacy in an orthotopic mouse model of peritoneal carcinomatosis. Ad5/3SLPI showed significantly decreased liver toxicity compared with other 5/3 fiber-modified control vectors examined.

Conclusions: In summary, Ad5/3SLPI is a promising vector candidate for treating metastatic ovarian cancer and showed robust virus replication, oncolysis, and in vivo therapeutic efficacy. Ad5/3SLPI showed comparatively low liver toxicity and therefore holds potential for patient use in the clinic.

Novel agents in epithelial ovarian cancer

The gold standard chemotherapy for previously untreated patients with ovarian cancer is currently a combination of taxane and platinum. However, most patients still suffer relapse, and less than 20% of the patients with stage III or IV disease survive long term. With more advanced technology, newer cytotoxic agents have been identified and are currently being tested in patients with ovarian cancer. Recent advances in the understanding of ovarian cancer biology have also led to the identification of multiple molecular targets that may soon change the standard treatment of ovarian cancer. Several of these targeted agents have entered clinical trials. Small molecular-weight inhibitors, monoclonal antibodies, antisense therapy, and gene therapy are all being evaluated alone and in combination with cytotoxic chemotherapy. Several of these cytotoxic and targeted therapies are reviewed here. Ultimately, the success of ovarian cancer therapy lies not just in the availability of new agents but in the ability to identify patients with biomarkers that may predict their response to these agents.

A tumor-targeted and conditionally replicating oncolytic adenovirus vector expressing TRAIL for treatment of liver metastases

We have constructed a new capsid-modified adenovirus (Ad) vector that specifically replicates in tumor cells and expresses TNF-related apoptosis-inducing ligand (TRAIL). The Ad capsid contains short-shafted fibers derived from Ad serotype 35, which allow for efficient infection of malignant tumor cells, and largely avoids innate toxicity after intravenous application. Replication-dependent homologous recombination in Ad genomes was used to achieve tumor-specific expression of Ad E1a (to mediate viral replication) and TRAIL (to mediate apoptosis and enhance release of progeny virus from infected cells). We demonstrated that our oncolytic vector (Ad5/35.IR-E1A/TRAIL) induced apoptosis in human tumor cell lines derived from colorectal, lung, prostate, and liver cancer. Both in vitro and in vivo tumor models showed efficient intratumoral spread of this vector. In a model for metastatic colon cancer, tail vein infusion of Ad5/35.IR-E1A/TRAIL resulted in elimination of preestablished liver metastases. Intravenous injection of this vector caused a transient elevation of serum glutamic pyruvic transaminase in tumor-bearing mice, which we attributed to factors released from apoptotic tumor cells. Liver histology analyzed at day 14 after virus injection did not show signs of hepatocellular damage. This new oncolytic vector represents a potentially efficient means for gene therapy of metastatic cancer.

Genetic replacement of the adenovirus shaft fiber reduces liver tropism in ovarian cancer gene therapy

Approaches to alter the native tropism of adenoviruses (Ads) are beneficial to increase their efficacy and safety profile. Liver tropism is important with regard to potential clinical toxicity in humans. Ad5/3 chimeras in which the Ad5 knob is substituted by the Ad3 knob, such as Ad5/3luc1, have been recently shown to increase infectivity of ovarian cancer cell lines and primary tumor cells, which express low levels of the coxsackie-adenovirus receptor (CAR), without increasing infectivity of liver cells. A novel strategy to address the problem of liver uptake and improve the tumor/liver ratio is genetic replacement of the Ad fiber shaft. Ad5.Ad3.SH.luc1 is an Ad5-based vector that contains the fiber shaft from Ad serotype 3 but the fiber knob from Ad serotype 5. To compare tumor/liver of Ad5.Ad3.SH.luc1 and Ad5/3luc1 in vivo, we created three different tumor and treatment models of ovarian cancer in mice, simulating intraperitoneal and intravenous administration of tumors. Ad5.Ad3.SH.luc1 displayed the lowest liver tropism of all viruses in all models tested. Intravenous administration of all viruses resulted in higher tumor transduction rates compared to intraperitoneal administration. Genetic shortening of the Ad5 fiber shaft significantly increases relative tumor/liver gene transfer. This could improve the effective tumor dose and reduce side effects, thereby increasing the bioavailability of therapeutic agents.

Infectivity enhanced adenoviral-mediated mda-7/IL-24 gene therapy for ovarian carcinoma

OBJECTIVE

Melanoma differentiation associated gene-7 [mda-7/Interleukin (IL)-24] has been identified as a novel anti-cancer agent, which specifically induces apoptosis in cancer cells but not in normal epithelial, endothelial and fibroblast cells. The objective of this study was to evaluate the anti-tumor effect of adenovirus-mediated mda-7/IL-24 (Ad.mda-7) gene therapy in ovarian carcinoma and further improve anti-tumor effect by enhancing infectivity of Ad.mda-7.

METHODS

A panel of human ovarian carcinoma cells, OV-4, HEY, SKOV3, SKOV3.ip1 and control normal human mesothelial cells, were infected by a replication deficient recombinant adenovirus encoding mda-7/IL-24 and control virus Ad.CMV.Luc. After 72 h, apoptosis was evaluated by TUNEL and Hoechst staining and further quantified by fluorescent activated cell sorter (FACS) analysis. Infectivity of Ad.mda-7 was enhanced by retargeting it to CD40 or EGF receptors overexpressed on ovarian cancer cells. Subsequently, enhancement in apoptosis of CD40- or epidermal growth factor receptor (EGFR)-retargeted Ad.mda-7 was evaluated.

RESULTS

Adenoviral-mediated delivery of mda-7 induces apoptosis ranging from 10-23% in human ovarian cancer cells tested with the highest percentage of apoptosis noted in SKOV3 cells. Minimal apoptosis was noted in normal mesothelial cells. CD40- or EGFR-retargeted Ad.mda-7 increased apoptosis by 10-32% when compared to that achieved with untargeted Ad.mda-7.

CONCLUSION

Ad.mda-7 exhibits ovarian cancer-specific apoptosis, but does not affect normal human mesothelial cells. Infectivity enhanced CD40- and EGFR-retargeted Ad.mda-7 augments apoptosis induction, thus increasing the therapeutic index and translational potential of Ad.mda-7 gene therapy.

Tissue-specific promoters for cancer gene therapy

Adenoviral cancer gene therapy approaches have resulted in promising recent results. Following only a decade of intense development, some of the crucial obstacles are now being overcome. Insufficient transduction has been the main limitation of earlier approaches. A new approach for increasing transduction of tumour cells is utilisation of replication-competent oncolytic agents, such as conditionally replicating adenoviruses (CRADs). The anti-tumour effect is caused by replication of the virus per se and, thus, replication must be restricted to tumour cells to protect normal tissues from damage. Tissue-specific promoters (TSPs) represent a powerful tool for decreasing the toxicity of cancer gene therapy to normal tissues and have previously been utilised for specific mutation compensation or delivery of prodrug-converting enzymes. However, TSPs can also be used for controlling crucial viral replication regulators and consequent restriction of replication to tumour cells. Initial clinical trials have demonstrated the safety and suggested efficacy for TSP-controlled CRADs as a novel approach for cancer gene therapy.

Preclinical evaluation of a class of infectivity-enhanced adenoviral vectors in ovarian cancer gene therapy

Ovarian carcinoma cells are often infected inefficiently by adenoviruses (Ad) due to low expression of coxsackie-adenovirus receptors (CAR), hindering the application of adenovirus-mediated gene therapy in ovarian cancer. In this study, we explored a class of infectivity-enhanced Ad vectors, which contain CAR-independent targeting motifs RGD (Ad5.RGD), polylysine (Ad5.pK7), or both (Ad5.RGD.pK7), for their utility in ovarian cancer gene therapy using in vitro and in vivo model systems. We found that these vectors infected established ovarian carcinoma cell lines and primary ovarian cancer cells with significantly enhanced infectivity. Among them, Ad5.RGD.pK7 appeared to be most efficient. Further, we evaluated their gene delivery efficiency using two different ovarian cancer mouse models--subcutaneous and intraperitoneal human ovarian cancer xenografts. All of the modified vectors appeared to be more efficient than the unmodified Ad5 vector in both models, although some of the differences are not statistically significant. Of these, Ad5.RGD.pK7 exhibited the highest efficacy in the subcutaneous tumor model, while Ad5.pK7 worked most efficiently in the intraperitoneal tumor model. These preclinical results suggest that Ad5.RGD.pK7 and Ad5.pK7 may be very useful in ovarian cancer gene therapy.

Ablating CAR and Integrin Binding in Adenovirus Vectors Reduces Nontarget Organ Transduction and Permits Sustained Bloodstream Persistence Following Intraperitoneal Administration

To create tumor-targeted Ad vectors, ablation of native CAR and integrin receptor binding is crucial to enhance the specificity of tumor transduction. Toward this aim, we have previously created base vectors in which binding to CAR (single-ablated) or to both CAR and integrins (double-ablated) has been ablated. In this study, the biodistribution of the conventional (CAR and integrin binding intact), single-ablated, and double-ablated vectors was evaluated following intraperitoneal administration. The mesothelial lining of the peritoneal organs was the principle site of CAR-dependent gene transfer by the conventional vector. Surprisingly, the single-ablated vector strongly transduced the liver parenchyma rather than the mesothelium, while the double-ablated vector did not significantly transduce the parenchyma or mesothelium. The high level of parenchymal transduction by the single-ablated vector suggested that it efficiently entered the bloodstream from the peritoneal cavity. Consistent with this hypothesis, a large proportion of active particles distributed and persisted in the bloodstream following intraperitoneal administration of either the single- or the double-ablated vector. The above results suggest that the double-ablated vector backbone may not only significantly improve targeting to cancers located in the peritoneal cavity, but may also significantly improve targeting to metastatic tumors located throughout the body by virtue of its enhanced bloodstream persistence.

Conditionally replicative adenovirus expressing a targeting adapter molecule exhibits enhanced oncolytic potency on CAR-deficient tumors

Conditionally replicative adenoviruses (CRAds) are potentially useful agents for anticancer virotherapy approaches. However, lack of coxsackievirus and adenovirus receptor (CAR) expression on many primary tumor cells limits the oncolytic potency of CRAds. This makes the concept of targeting, that is, redirecting infection via CAR-independent entry pathways, relevant for CRAd development. Bispecific adapter molecules constitute highly versatile means for adenovirus targeting. Here, we constructed a CRAd with the Delta24 E1A mutation that produces a bispecific single-chain antibody directed towards the adenovirus fiber knob and the epidermal growth factor receptor (EGFR). This EGFR-targeted CRAd exhibited increased infection efficiency and oncolytic replication on CAR-deficient cancer cells and augmented lateral spread in CAR-deficient 3-D tumor spheroids in vitro. When compared to its parent control with native tropism, the new CRAd exhibited similar cytotoxicity on CAR-positive cancer cells, but up to 1000-fold enhanced oncolytic potency on CAR-deficient, EGFR-positive cancer cells. In addition, EGFR-targeted CRAd killed primary human CAR-deficient brain tumor specimens that were refractory to the parent control virus. We conclude, therefore, that CRAds expressing bispecific targeting adapter molecules are promising agents for cancer treatment. Their use is likely to result in enhanced oncolytic replication in cancerous tissues and thus in more effective tumor regression.

Targeted therapy for epithelial ovarian cancer: Current status and future prospects

Despite advances in surgery and chemotherapy, less than 20% of patients with stage III or IV ovarian cancer survive long-term. In the past, cytotoxic regimens have been developed empirically, combining active agents at maximally tolerated doses, often without a clear rationale for their interaction. Advances in understanding the biology of ovarian cancer have identified multiple molecular targets that differ in normal and malignant cells. Targets include cell cycle regulators, growth factor receptors, signal transduction pathways, molecules that confer drug resistance, and angiogenic mechanisms. A number of targeted agents have entered clinical trials. Small molecular weight inhibitors, monoclonal antibodies, and antisense and gene therapy are all being evaluated alone and in combination with cytotoxic drugs. In contrast to earlier studies, the impact of each agent on the designated target can be assessed and agents can be matched to the genotype and phenotype of malignant and normal cells. In the long run, this should facilitate individualization of more effective, less toxic therapy for women with ovarian cancer.

Gene therapy for ovarian cancer: progress and potential

Gene therapy remains a promising therapeutic modality for ovarian cancer. Yet much work remains to be done to see gene therapy realize its full potential in elucidating the complex genetic interactions of delivered genes within target cancer cells and in the development of improved vector systems. Because most neoplasms involve multiple mutations, the targeting of a single mutation is unlikely to achieve total tumor control: gene therapy strategies that target multiple cellular processes or invoke various antitumor approaches need to be investigated. Additionally, current vector systems do not transduce ovarian cancer cells efficiently and are hampered by immune responses that further limit their efficacy. Additionally, limitations in vector specificity lead to transduction of normal cells and subsequent toxicity. Investigators are developing refinements to current gene therapy approaches that would address these limitations and that are soon to be incorporated into clinical trials. It is hoped that these advances will lead to improvements in the therapeutic index for ovarian cancer gene therapy and provide another effective therapeutic tool for this deadly disease.

Why did p53 gene therapy fail in ovarian cancer?

Promising preclinical and clinical data led to the initiation of an international randomised phase II/III trial of p53 gene-therapy trial for first-line treatment of patients with ovarian cancer. In that trial, replication-deficient adenoviral vectors carrying wild-type p53 were given intraperitoneally in combination with standard chemotherapy to patients with ovarian cancers harbouring p53 mutations. The study was closed after the first interim analysis because an adequate therapeutic benefit was not shown. In this review, we discuss the possible reasons for failure of p53 gene therapy, which include the multiple genetic changes in cancer and epigenetic dysregulations leading to aberrant silencing of genes. These complex interactions lead us to conclude that repair of single genes might not be a suitable strategy for the treatment of cancer. Moreover, dominant negative cross talk between ectopic wild-type p53 and recently identified dominant p53 mutants and splice variants of p63 and p73--which are frequently overexpressed in ovarian cancers--could seriously compromise the effectiveness of p53 gene therapy. Other substantial problems in targeting tumour cells with adenoviral vectors are the heterogeneity or lack of expression of coxsackie-adenovirus receptors and integrin co-receptors in ovarian tumours and the presence of adenovirus-neutralising antibodies in ovarian cancer-related ascites.

An adenoviral vector containing an arg-gly-asp (RGD) motif in the fiber knob enhances protein product levels from transgenes refractory to expression

Genetic manipulation of the adenovirus type 5 represents one strategy to modify viral transduction properties in vitro and in vivo. In the majority of studies to date, reporter gene activity has been monitored to assess transduction efficiency. BRCA1 is a gene whose protein product is clinically important, biologically toxic, difficult to overexpress, and difficult to detect as an untagged protein species. Thus, it represents an attractive candidate from which to evaluate the efficacy of a gene delivery system. In the present study, transgene expression was assessed employing otherwise isogenic viruses, which differed only in the presence or absence of an RGD integrin-binding motif in the HI loop of the Ad fiber knob. We utilized a combination of BRCA1 expression level comparisons among several human BRCA1/mutant BRCA1/murine Brca1 constructs and reporter gene activity following transduction of a panel of human breast and ovarian tumor cell lines representative of both sporadic and hereditary cases. A general overall concordance in efficiency was observed, whether the biological readout measured was reporter gene activity or steady-state level of ectopic BRCA1 protein produced. Importantly, the expression of full-length wild-type BRCA1 protein, clinically relevant mutant BRCA1 proteins or murine Brca1 was superior when the gene was delivered via the RGD-modified Ad. The ectopic BRCA1 stabilized endogenous BARD1 and this functional effect was evident at lower input viral doses when BRCA1 was delivered via the RGD-modified Ad. Quantitative, noninvasive, real-time image analysis of reporter gene function in nude mice harboring human ovarian tumor xenographs demonstrated a similar enhancement of expression in vivo by the RGD fiber modification, with low levels of transduction of normal mouse mesothelium. These results provide additional evidence supporting the concept that rational modification of viral vectors can result in the delivery of functionally active therapeutic proteins such as BRCA1 that present with technical difficulties with regard to their expression.