Gene-directed enzyme prodrug therapy for localized chemotherapeutics in allograft and xenograft tumor models

Abstract

Most chemotherapy regimens rely on systemic administration of drugs leading to a wide array of toxicities. Using viral-vector-mediated gene modification of muscle tissues, we have developed a method for gene-directed enzyme prodrug therapy that allows for localized drug administration. An inactive prodrug of geldanamycin was activated locally for inhibition of tumor growth without systemic toxicities. A recombinant adeno-associated virus (rAAV) was used to deliver β-galactosidase (LacZ) to the treatment group and green fluorescent protein to the control group. After 1 week, both groups received adenocarcinoma cells in the same location as the previous rAAV injection. The geldanamycin prodrug was administered 1 h later via intraperitoneal injection. Tumor growth was significantly suppressed in animals whose muscles were gene modified to express β-galactosidase compared with the control. Serum assay to access hepatotoxicity resulted in no significant differences between the animals treated with the inactive or activated form of geldanamycin, indicating minimal damage to non-target organs. Using gene-directed enzyme prodrug therapy, in combination with novel recombinant AAV vectors, we have developed a method for localized activation of chemotherapeutic agents that limits the toxicities seen with traditional systemic administration of these potent drugs.

Figure 1

Molecular structure of the geldanamycin prodrug, Compound 25: C₃₆H₅₃N₃O₁₄ (17-demethoxy-17[(2-β-galactopyranosylethyl) amino]-geldanamycin).

Figure 2

Recombinant adeno-associated viruses (rAAVs) were used to gene modify the quadriceps femoris muscles of mice by direct injection. In the treatment group, gene modification was accomplished with rAAV-LacZ to express β-galactosidase. In the control group, muscles were injected with rAAV-GFP to express GFP protein. After 1 week, colon cancer cells were implanted into the muscle. After 1 h, animals were treated with intraperitoneal injection of the geldanamycin prodrug. Mice were then retreated with the prodrug on days 3 and 7 following tumor injection. GFP, green fluorescent protein.

Figure 3

At 2 weeks after administration of prodrug, all mice were examined and tumor volume was measured using topical calipers. Measurements were recorded and repeated every 3 days until the mice were killed at day 26. In both the (a) allograft and (b) xenograft models, tumor mass began to increase more quickly for the control group (rAAV-GFP and prodrug) than the treated group (rAAV-LacZ and prodrug) and a significant difference was achieved by 26 days. *P ≤ 0.05 demonstrated a significant result. Data are mean ± s.d.; n = 3 for each group; Student's t-test: P_allograft = 0.007, P_xenograft = 0.004. GFP, green fluorescent protein; rAAV, recombinant adeno-associated virus.

Figure 4

Mice were killed 26 days after administration of prodrug and tumors were surgically exposed. Mice treated with a combination of (a) green fluorescent protein (GFP) and prodrug (control) demonstrated a tumor that was visibly larger than the mice treated with a combination of (b) LacZ viral vector and prodrug (gene-directed enzyme prodrug therapy (GDEPT) treated). Images shown are from the allograft tumor model.

Figure 5

At 26 days postimplantation, the average tumor mass in the control group (rAAV-GFP and prodrug) was significantly greater than the average tumor mass for the treated group (rAAV-LacZ and prodrug) for both the allograft and xenograft tumor models. *P ≤ 0.05 demonstrated a significant result. Data are mean ± s.d.; n = 3 for each group. Student's t-test: P_allograft = 0.003, P_xenograft = 0.002. GFP, green fluorescent protein; rAAV, recombinant adeno-associated virus.

Figure 6

In vitro cellular proliferation assay (MTS) was performed on both the MC38 and SW620 cell lines to quantify the rate of cellular doubling in the control (plain media and prodrug) and treated (β-galactosidase and prodrug) groups. The control cells proliferated at a significantly greater rate than the cells that were treated with the activated prodrug. *P ≤ 0.05 demonstrated a significant result. Data are mean ± s.d.; n = 2 for each group. Student's t-test: P_allograft = 0.004, P_xenograft = 0.003.

Figure 7

Tumors were excised, cleaned of excess uninvolved tissue and weighed using a calibrated gram scale. All tumor weights were measured and recorded to the nearest 0.1 g. Averages and standard deviations were calculated for (a) the control group (rAAV-GFP and prodrug) and (b) the treated group (rAAV-LacZ and prodrug). Images shown are from the xenograft tumor model. Note: Images do not represent the actual size of the specimens, only the relative difference in tumor mass between the two populations. GFP, green fluorescent protein; rAAV, recombinant adeno-associated virus.