Incorporation of a mutated E1A and a chimeric 5/3 fiber improved the therapeutic efficacy of a novel CRAd (AdF512) on disseminated ovarian cancer

VERONICA M LOPEZ; ANGEL A RIVERA; LORENA BENEDETTI; DIEGO L VIALE; KRISTOPHER J KIMBALL; MINGHUI WANG; ALICIA I BRAVO; CHRISTINE DORVAULT; JOANNE DOUGLAS; ZENG B ZHU; RONALD D ALVAREZ; DAVID T CURIEL; OSVALDO L PODHAJCER

Congreso; 13th Annual Meeting of the American Society of Gene & Cell Therapy; 2010

We have previously developed a new CRAd where E1A activity is driven by a 0.5 kb fragment of the SPARC gene promoter (SPPr). This novel CRAd (termed AdF512) was designed to target both the malignant and the tumor-associated stromal compartment of the tumor mass (Lopez et al, PlosOne 2009). In order to further improve specificity and retargeting, we prepared different versions of the CRAd where SPPr is driven mutated E1A variants. Moreover, the CRAd was pseudotyped with a chimeric Ad5/3(shaft/knob) fiber. Ovarian cancer is one of the leading gynecologic malignancies with no effective treatment at advanced stages if conventional chemotherapy fails. In the present study we assessed the different CRAd variants in ovarian cancer models. Initial studies were designed to establish the infective capacity of the new vectors. We observed in 3 out of 4 ovarian cancer cell lines a 2-50 fold increase when Ad-SV40-luc 5/3 was compared to Ad-SV40-luc 5/5. Moreover, Ad-F512-luc 5/3 was as active as Ad-SV40-luc 5/3. Next, we constructed four CRAds where SPPr drove the activity of wild type E1A, E1A deleted in the Rb binding motif, in the p300 motif or both and measured the efficacy in the 4 ovarian cancer cell lines. By using the MTS assay we observed that the better lytic capacity was obtained with the CRAd containing the Rb deletion (Ad-F512-E1Rb 5/3).  In order to establish the potential clinical utility of Ad-F512-E1Rb 5/3 we also assayed its replication capacity in a most rigorous preclinical system, which is slices obtained from fresh tissue explants of human ovarian carcinomas compared to normal ovary. Ad-F512-E1Rb 5/3 replicated in three out of four human ovarian carcinomas but showed no replication in 3 normal ovary samples. It is important to mention that Ad-wt 5/3 could replicate in both malignant and normal samples. Histochemical analysis for SPARC expression levels, presence of stroma and virus replication was also performed on the fresh explants. Finally, we examined the in vivo lytic capacity of Ad-F512-E1Rb 5/3 in a xenograft model of disseminated intraperitoneal ovarian cancer by using a bioluminescent imaging follow up. We found that four i.p. injections of 1010 v.p inhibited almost 50% of tumor growth, (assessed using caliper, by signal intensity and by weight) in animals treated with Ad-F512-E1ARb 5/3 compared to the controls.  In conclusion, Ad-F512-E1ARb 5/3 demonstrated a strong killing effect on ovarian cancer cells in vitro, and in vivo on disseminated tumors as well as a high replication capacity in fresh human ovary cancer explants. Ad-F512-E1ARb 5/3 appears safe since no replication was observed in normal human ovary raising its potential use in the clinics.