Introduction

Replication-competent, oncolytic adenoviruses represent a new investigational therapy that has been evaluated in human trials both as a single agent and in combination with conventional cancer therapies^{1,2,3,4,5,6,7,8,9,10}. Although the safety of this approach has been demonstrated in multiple settings, it has become clear that oncolytic adenoviruses do not show sufficient activity as single agents to justify their use as a first-line cancer therapy. Nevertheless, anti-tumor activity has been demonstrated and it is possible that further improvements will bring this novel technology to a point at which it will have applicability in the clinic.

One strategy that has been used to increase the anti-tumor activity of replication-competent adenoviruses is to "arm" them with therapeutic genes. The first such armed replication-competent adenovirus, Ad5-CD/TKrep, contained a bacterial cytosine deaminase (bCD)/wild-type herpes simplex virus thymidine kinase (HSV-1 TK) fusion gene in the E1 region in place of the 55-kDa E1B gene¹¹. In preclinical models of human cancer, the anti-tumor activity of Ad5-CD/TKrep could be significantly enhanced by combining it with 5-fluorocytosine (5-FC) and ganciclovir (GCV) prodrugs, which are converted into toxic agents by the CD and HSV-1 TK genes contained in the virus^12,13. Not only do the CD/5-FC and HSV-1 TK/GCV enzyme/prodrug systems provide a chemotherapeutic effect, they also function as potent sensitizers of radiotherapy^{14,15,16,17,18,19,20}. The safety and preliminary efficacy of this combined approach have been evaluated in two Phase I clinical trials of prostate cancer without and with conventional radiotherapy^21,22. In both studies, the combined approach was associated with low toxicity with the vast majority (>90%) of adverse events being mild to moderate and self-limiting. In the absence of radiotherapy²¹, anti-tumor activity was demonstrated by significant (>50%) declines in serum prostate-specific antigen (PSA) and histopathologic evidence of tumor destruction in posttreatment prostate biopsies. When combined with radiotherapy²², better than expected declines in serum PSA were observed and the negative biopsy status at 1 year was better than historical controls (67% vs 45%) with the same patients and radiation doses. Although the results of these small Phase I studies must be interpreted cautiously, they are encouraging and raise the possibility that replication-competent, oncolytic adenoviruses armed with therapeutic genes may demonstrate meaningful clinical activity when applied to the right cancer and in the right setting.

Prior to evaluating the efficacy of this investigational approach in a randomized, prospective, two-arm Phase II trial, we have made two improvements in Ad5-CD/TKrep, generating the second-generation, replication-competent adenovirus Ad5-yCD/mutTK_SR39rep-ADP. First, Ad5-yCD/mutTK_SR39rep-ADP contains an improved yeast cytosine deaminase (yCD)/mutant_SR39 herpes simplex virus thymidine kinase fusion (yCD/mutTK_SR39) gene in place of the bCD/wild-type HSV-1 TK fusion gene. Previous studies have demonstrated that both the yCD and the mutant_SR39 HSV-1 TK enzymes are more catalytically efficient than their bacterial and wild-type counterparts, respectively^23,24,25,26. Second, Ad5-yCD/mutTK_SR39rep-ADP contains the 11.6-kDa adenovirus death protein (ADP) gene that has been shown to enhance the cytolytic activity of replication-competent adenoviruses in vitro^{27,28,29,30,31}. In this report, we compared the toxicity and efficacy of Ad5-yCD/mutTK_SR39rep-ADP versus the parental Ad5-CD/TKrep adenovirus in preclinical models. We demonstrate that Ad5-yCD/mutTK_SR39rep-ADP results in greater tumor cell kill and tumor control without significantly increasing toxicity.

Results

In vitro assays demonstrating therapeutic advantage of yCD/mutTK_SR39 and ADP genes

We made two improvements in the parental Ad5-CD/TKrep adenovirus, generating the second-generation adenovirus, Ad5-yCD/mutTK_SR39rep-ADP. We replaced the prototype bCD/wild-type HSV-1 TK fusion gene located in the E1 region of Ad5-CD/TKrep with a yCD/mutant_SR39 HSV-1 TK (mutTK_SR39) fusion gene. Second, whereas Ad5-CD/TKrep lacks any genes in the E3 region, Ad5-yCD/mutTK_SR39rep-ADP contains the Ad5 ADP gene. Both the yCD/mutTK_SR39 and the ADP genes are under the transcriptional control of the human CMV promoter. Because these identical promoters are separated by approximately 27 kb, no rearrangements in the Ad5-yCD/mutTK_SR39rep-ADP genome have been observed in over a dozen laboratory, and one clinical, preparations.

Western blotting demonstrated that Ad5-yCD/mutTK_SR39rep-ADP expressed the expected 59-kDa yCD/mutTK_SR39 fusion and 11.6-kDa ADP proteins (Figs. 1A and 1B). Ad5-yCD/mutTK_SR39rep-ADP expressed ADP at a level 5- to 10-fold greater than wild-type Ad5. This is likely attributable to the relative strengths of the promoters driving ADP expression (CMV in Ad5-yCD/mutTK_SR39rep-ADP and Ad5 major late promoter in wild-type Ad5).

Figure 1.

Western blotting demonstrating expression of yCD/mutTK_SR39 and ADP proteins. (A) The blot was probed with a mixture of polyclonal antibodies to yCD and bCD. The positions of the 17-kDa yCD, 59-kDa yCD/mutTK_SR39, and 88-kDa bCD/wtTK fusion proteins are indicated. The yCD/mutTK_SR39 fusion protein is susceptible to proteolysis between the yCD and the mutTK_SR39 moieties and the yCD moiety is detected. Ad5-yCD is a replication-competent adenovirus containing only the yCD gene in the E1 region. (B) The blot was probed with an antibody to ADP. The position of the 11.6-kDa ADP protein is indicated.

Full figure and legend (64K)

To demonstrate the therapeutic advantage of the yCD/mutTK_SR39 fusion gene in vitro, we performed 5-FC and GCV prodrug sensitivity assays. To minimize interference due to the ADP cytolytic effect, we infected mouse, rather than human, prostate adenocarcinoma cells with Ad5-CD/TKrep and Ad5-yCD/mutTK_SR39rep-ADP and incubated them with increasing concentrations of 5-FC or GCV. Both Ad5-CD/TKrep and Ad5-yCD/mutTK_SR39rep-ADP replicate inefficiently in mouse TRAMP-C2 cells and the enhanced cytolytic effect of ADP is not observed during the course of the prodrug sensitivity assay. Both the yCD and the mutTK_SR39 genes contained in Ad5-yCD/mutTK_SR39rep-ADP resulted in greater cell kill than the bCD and wt TK genes contained in Ad5-CD/TKrep (Fig. 2). Specifically, at a concentration of 30 g/ml (0.23 mM) 5-FC and 0.03 g/ml (0.1 M) GCV, Ad5-yCD/mutTK_SR39rep-ADP resulted in significantly greater cell kill than Ad5-CD/TKrep. Clonogenic assays demonstrated that the increase in cell kill ranged from two- to fivefold for both the yCD/5-FC and the mutTK_SR39/GCV enzyme/prodrug systems (not shown). These results could not be explained by the presence of ADP in Ad5-yCD/mutTK_SR39rep-ADP because no cell kill was observed in the absence of prodrugs (Fig. 3, 0 viral particles (vp)/cell). Results of the prodrug sensitivity assays were supported by in vitro enzyme assays demonstrating that the yCD and mutTK_SR39 enzymes were five- and twofold better at converting 5-FC and GCV into their corresponding products than their bacterial and wild-type counterparts (not shown).

Figure 2.

Prodrug sensitivity assays. Mouse TRAMP-C2 cells were mock-infected (M) or infected with the Ad5-CD/TKrep or Ad5-yCD/mutTK_SR39rep-ADP adenovirus as indicated. Following infection, cells were incubated in medium containing varying concentrations of 5-FC (1, 3, 10, 30, 100 g/ml) or GCV (0.01, 0.03, 0.1, 0.3, 1.0 g/ml).

Full figure and legend (105K)

Figure 3.

Cytopathic effect assays. LNCaP cells were mock infected or infected with increasing amounts of the Ad5-CD/TKrep or Ad5-yCD/mutTK_SR39rep-ADP adenoviruses. Wells were fixed and stained with crystal violet 3 and 5 days later.

Full figure and legend (87K)

To demonstrate the therapeutic advantage of the ADP gene in vitro, we performed cytopathic effect (CPE) assays (see Materials and Methods for discussion regarding basing adenovirus input on vp versus plaque-forming units (pfu)). We infected human LNCaP prostate adenocarcinoma cells with increasing amounts of Ad5-CD/TKrep and Ad5-yCD/mutTK_SR39rep-ADP and examined cell kill in the absence of prodrugs. At all time points, one-tenth the amount of Ad5-yCD/mutTK_SR39rep-ADP was required to produce the same amount of cell death as Ad5-CD/TKrep (Fig. 3). We obtained similar results with human pancreatic adenocarcinoma cells (PaCa-2 and Panc-1) and human glioma (U251 and U87) cells (not shown).

In vivo assays demonstrating therapeutic advantage of yCD/mutTK_SR39 and ADP genes

To determine whether Ad5-yCD/mutTK_SR39rep-ADP was more efficacious than the parental Ad5-CD/TKrep adenovirus in vivo, we injected human LNCaP tumor xenografts with adenovirus and administered half the animals in each treatment group a 1-week course of 5-FC and GCV prodrug therapy. Ad5-yCD/mutTK_SR39rep-ADP resulted in significantly greater tumor control than Ad5-CD/TKrep both in the absence and in the presence of prodrugs (Fig. 4). In the absence of prodrugs, which reflects the effect of ADP only, Ad5-yCD/mutTK_SR39rep-ADP resulted in a tumor growth delay (relative to PBS-injected controls) of 41 days vs 9 days for Ad5-CD/TKrep (P < 0.05). In the presence of prodrugs, which reflect the combined effects of yCD/mutTK_SR39 and ADP, Ad5-yCD/mutTK_SR39rep-ADP resulted in a tumor growth delay of 85 days vs 18 days for Ad5-CD/TKrep (P < 0.05). Importantly, 5-FC + GCV prodrug therapy significantly enhanced the anti-tumor effects of the Ad5-yCD/mutTK_SR39rep-ADP adenovirus alone (85 days vs 41 days, P < 0.05). Similar results were obtained with the C33a (cervical) tumor model (not shown).

Figure 4.

Comparison of in vivo efficacy of Ad5-CD/TKrep versus Ad5-yCD/mutTK_SR39rep-ADP. LNCaP tumors (150 mm³) were injected with PBS or the Ad5-CD/TKrep or Ad5-yCD/mutTK_SR39rep-ADP adenovirus (10¹⁰ vp per injection) on days 1, 3, and 5 (arrowheads). Half the mice in each treatment group were administered daily injections of 5-FC (500 mg/kg/day) and GCV (30 mg/kg/day) for 1 week (days 5–11, hatched bar). Each data point represents the group mean and the bars the standard deviation. Each group contained 10 animals.

Full figure and legend (90K)

ADP improves local spread of replication-competent adenoviruses in vivo

Previously, we developed a quantitative method to measure the magnitude and volume of transgene expression in vivo with submillimeter resolution^32,33. This method involves quantification of a reporter gene, the human sodium iodide symporter (hNIS), from two-dimensional tissue sections followed by stacking of sections to create a 3D representation of the magnitude and volume of reporter gene expression in vivo. Although it has been demonstrated previously that ADP can enhance adenoviral spread in vitro^{27,28,29,30,31}, to our knowledge, this has never been demonstrated in vivo.

We injected a pair of replication-competent hNIS-containing adenoviruses, one lacking (Ad5-hNISrep) and the other containing (Ad5-hNIS-ADPrep) the Ad5 ADP gene, into mice bearing human tumor xenografts and the naïve canine prostate. Although the latter model lacks malignant tissue, the size and structure of the canine prostate resembles that of humans, making it an excellent preclinical model to examine issues such as adenoviral spread in vivo. In all studies, the dose, concentration, and vp/pfu ratio of adenovirus; the injection volume; the tumor and prostate volumes; and the injection plan were essentially identical so as not to introduce bias into the results. The next day, we administered Na^99mTcO₄ to the animals, removed and sectioned tissue, and quantified the volume of reporter gene expression. Representative autoradiograms of tissue sections depicting the raw data and transgene expression isodose curves depicting computer-generated representation of the raw data are shown in Fig. 5, with the results summarized in Table 1. In the human tumor xenograft model, the presence of ADP increased reporter gene expression volume by an average of 2.5-fold (n = 5). Similarly, in the canine prostate model, the presence of ADP increased transgene expression volume by an average of 3.3-fold (n = 3). These results demonstrate in two preclinical models that ADP increases the local spread of replication-competent adenoviruses in vivo and may provide an explanation for how ADP improves the anti-tumor effects of replication-competent, oncolytic adenoviruses.

Figure 5.

Effect of ADP on reporter gene expression volume in vivo. Replication-competent hNIS-containing adenoviruses, one lacking (Ad5-hNISrep, ADP-) and the other containing (Ad5-hNIS-ADPrep, ADP+) the Ad5 ADP gene, were injected into contralateral sides of the canine prostate. Representative autoradiograms depicting consecutive tissue sections (left) and computer-generated (i.e., digitized) representations of reporter gene expression (right) are shown. A color scale representing different relative levels of reporter gene expression is shown.

Full figure and legend (108K)

Table 1 - Transgene expression volume of ADP-containing versus non-ADP-containing adenoviruses.

Full table

Inclusion of yCD/mutTK_SR39 and ADP genes does not significantly increase toxicity

Two toxicology studies were performed in C57BL/6 male mice to evaluate the toxicity of administering Ad5-yCD/mutTK_SR39rep-ADP intraprostatically (Table 2). The first study compared the toxicity of a single intraprostatic injection of Ad5-yCD/mutTK_SR39rep-ADP versus the parental Ad5-CD/TKrep adenovirus followed by 2 weeks of 5-FC + GCV prodrug therapy and pelvic radiation. For Ad5-yCD/mutTK_SR39rep-ADP treatment groups, we also evaluated every possible mono- and dual-therapy combination for comparison. The second study compared the toxicity of one versus two intraprostatic injections of Ad5-yCD/mutTK_SR39rep-ADP each followed by a 2.6-week cycle of 5-FC + GCV prodrug therapy and pelvic radiation. We recorded general observations daily and measured body weights weekly. We performed necropsies at three time points for histopathological analysis of major abdominal organs, blood cell counts, and blood chemistries.

Table 2 - Design of toxicology studies.

Full table

Overall, the combination of Ad5-yCD/mutTK_SR39rep-ADP with 5-FC + GCV prodrug therapy and concomitant pelvic radiation (trimodal therapy) did not demonstrate significantly increased toxicity relative to the parental Ad5-CD/TKrep adenovirus. There were no deaths attributable to the investigational therapy in either study. In the first study, the most notable histopathological change was the presence of minimal to mild inflammation in all groups that received the Ad5-yCD/mutTK_SR39rep-ADP and Ad5-CD/TKrep adenoviruses (Table 3). The inflammation was often present along the serosal membranes of several abdominal organs (seminal vesicle, urinary bladder, liver) or within the organ parenchyma as was the case with the prostate gland and liver. Subacute inflammation was characterized primarily by the presence of a mixed inflammatory reaction consisting of polymorphonuclear inflammatory cells and mononuclear cells indicative of a more acute and ongoing active response. For instance, when present in the prostate or on serosal surfaces, subacute inflammation was noted only at day 5. By day 17, the inflammatory reaction had become more chronic. Chronic inflammation was characterized by more mononuclear cells, fewer polymorphonuclear cells, and subtle fibrosis. We considered this pattern of serosal inflammation to represent a low-grade, and somewhat regional (seminal vesicles and urinary bladder), form of peritonitis related to leakage of adenovirus into the abdominal cavity following the intraprostatic injection. Subacute inflammation and periportal mononuclear cell infiltrate were two lesions commonly observed within the livers of mice. By day 53, many of the inflammatory reactions noted at days 5 and 17 had resolved.

Table 3 - Results of toxicology studies.

Full table

Prodrug-treated groups (Groups 1, 2, and 7) exhibited leukopenia and lymphopenia on day 17 relative to controls (Table 3). These are expected side effects of the 5-FC + GCV prodrugs and resolved by day 53. Animals injected with Ad5-yCD/mutTK_SR39rep-ADP exhibited statistically higher serum ALT levels on day 17 relative to similarly treated Ad5-CD/TKrep animals and the controls. These elevations were minor (Group 1 vs Group 7—44 8 vs 31 6 IU/L, P < 0.05; Group 4 vs Group 8—34 15 vs 19 5 IU/L, P < 0.05; Group 6—28 8 IU/L); however, because they were observed in the absence of prodrugs and radiation (Groups 4 and 8), they were likely attributable to ADP gene expression. These events resolved by day 53. Animals in both trimodal therapy groups (Groups 1 and 7) developed neutropenia on day 53, which may have been a consequence of the combined myelosuppressive effects of the prodrug and radiation therapies. We noted no other significant differences in the blood chemistries.

The second study demonstrated that two cycles of Ad5-yCD/mutTK_SR39rep-ADP/5-FC + GCV prodrug therapy concomitant with pelvic radiation did not significantly increase toxicity relative to a single treatment cycle (Table 3). As was observed in the first study, the most notable histopathological change was inflammation along the serosal membranes of several abdominal organs or within the organ parenchyma in both treatment groups (Groups 1 and 2). In addition, there was minimal inflammation of the colon and the presence of apoptotic glandular epithelial cells in the small bowel, which are expected side effects of pelvic radiation. As expected, animals in the treatment groups exhibited myelosuppression during the prodrug therapy and radiation course. However, there was no significant difference between Group 1 (one cycle) and Group 2 (two cycles), and all events, except for slight anemia in Group 2, resolved by the end of the study. Interestingly, animals in both treatment groups demonstrated slight elevations in alkaline phosphatase at the end of the study (Group 1—90 15, Group 2—83 9 vs 60 10 IU/L for controls, P < 0.05). Because these events were not accompanied by elevations in hepatic transaminases or bilirubin, indicating damage to the hepatobiliary system, and the animals did not have evidence of bone disease, we believe this event was likely associated with the observed inflammation of the small bowel.

Discussion

The purpose of these studies was to evaluate in preclinical models the toxicity and efficacy of the second-generation Ad5-yCD/mutTK_SR39rep-ADP adenovirus to provide the scientific basis for its evaluation in humans. We demonstrate here that relative to the parental Ad5-CD/TKrep adenovirus, Ad5-yCD/mutTK_SR39rep-ADP exhibited greater tumor cell kill and tumor control without significantly increasing toxicity. Moreover, administering two cycles of the Ad5-yCD/mutTK_SR39rep-ADP adenovirus/5-FC + GCV prodrug therapy combination concomitant with radiotherapy resulted in little added toxicity relative to a single cycle. Together, the results indicate that Ad5-yCD/mutTK_SR39rep-ADP is suitable for evaluation in future clinical trials of human cancer in which the safety and efficacy of combining replication-competent adenovirus-mediated suicide gene therapy with radiotherapy will be compared directly to radiotherapy alone.

The safety of replication-competent adenovirus-mediated suicide gene therapy using the parental Ad5-CD/TKrep adenovirus has been evaluated in two Phase I clinical trials of prostate cancer without and with conventional radiotherapy^21,22. Overall, this investigational therapy has been well tolerated. When combining the results of both studies (31 patients), the most frequent side effects attributable to the investigational therapy were mild (grade 1) to moderate (grade 2) myelosuppression (90% of patients) and mild hepatotoxicity (33% of patients). These events were self-limiting and resolved shortly after completion of the prodrug therapy course. When combined with radiotherapy, there was no significant increase in acute bladder and bowel toxicities relative to conventional radiotherapy alone²². It is comforting that these same toxicities were observed in our preclinical toxicology studies, suggesting that they may accurately forecast what will occur in humans. We are pleased to find that despite having a more catalytically efficient yCD/mutTK_SR39 fusion and the ADP genes, the latter of which is expressed at a level five times that of wild-type Ad5, Ad5-yCD/mutTK_SR39rep-ADP resulted in little added hepatotoxicity relative to Ad5-CD/TKrep, which has a proven safety profile in humans^21,22. Indeed, Ad5-yCD/mutTK_SR39rep-ADP has already been administered intraprostatically to humans (10¹¹, 10¹² vp) followed by 2.6 weeks of 5-FC + GCV prodrug therapy and 76-Gy intensity-modified radiotherapy, and there have been only minor elevations in liver transaminase levels. These results are consistent with our previous studies in the dog, which demonstrated that following direct intraprostatic injection, little of the injected adenovirus disseminates beyond the prostate gland³². Nevertheless, because hepatotoxicity can be life-threatening, it will be monitored closely in all future trials, particularly those in which the injected tumor is in close proximity to the liver.

A major limitation of human cancer gene therapy, however, is not excessive toxicity but rather a lack of significant efficacy. The limited efficacy observed in human studies to date is likely attributable to poor in vivo delivery of the therapeutic agent. Both replication-competent, oncolytic adenoviruses and the CD/HSV-1 TK suicide gene systems are very efficient at killing tumor cells in vitro and it is possible that if a high fraction of tumor cells could be transduced in vivo, significant outcomes would be observed in humans. In an attempt to improve the local spread of these therapeutic agents in vivo, we generated an armed replication-competent adenovirus containing the ADP gene. As demonstrated here for the first time, inclusion of the ADP gene increased reporter gene expression volume approximately threefold in vivo. Our results confirm and extend those of Wold and colleagues, who demonstrated that ADP increases adenovirus spread in vitro^{27,28,29,30,31}. In our studies, an ADP-dependent increase in adenoviral spread was detectable at 24 h based upon evidence obtained in two animal models. It is possible that the observed increase in adenoviral spread at this early time point might not be explained solely on the basis of ADP's known cytolytic effect. Oncolytic adenoviruses expressing ADP might elicit a more robust immune response, which, in turn, might facilitate adenoviral spread through faster cell death. We have observed robust immune responses to replication-competent adenoviruses in the dog prostate as early as 24–48 h postinfection³². The immune response is characterized by a massive infiltration of neutrophils and macrophages, which are also present in athymic mice. Thus, the observed increase in adenoviral spread in vivo may reflect ADP's cytolytic effect, which can be observed in vitro, and the possible effect of ADP on the magnitude of the immune response. If these effects of ADP can be reproduced in humans, it should result in greater tumor destruction via the cytolytic effects of the Ad5-yCD/mutTK_SR39rep-ADP adenovirus, the chemoradiosensitization effects of the CD/HSV-1 TK suicide gene systems, and the resulting immune response.

Prior to these studies, it could have been argued that a potential drawback of including ADP in an armed replication-competent adenovirus is that it may shorten the duration of therapeutic gene expression. Replication-competent adenoviruses expressing ADP spread much faster, resulting in greater cell death when observed at a fixed time point. Although this effect of ADP clearly enhances the anti-tumor activity of the adenovirus itself (^29,30,31, this paper), it may weaken the chemoradiosensitization effects of the CD/HSV-1 TK suicide gene systems. Our preclinical studies indicate that ADP does not circumvent the chemoradiosensitization effects of the CD/HSV-1 TK suicide gene systems. As demonstrated here, the addition of 5-FC + GCV prodrug therapy significantly improved tumor control over that of the Ad5-yCD/mutTK_SR39rep-ADP adenovirus alone. Similar results have been obtained with the C33a (cervical) tumor model (unpublished). Although not examined here in the prostate model, we have observed marked synergy when combining Ad5-yCD/mutTK_SR39rep-ADP + 5-FC + GCV with radiation in a preclinical model of human pancreatic cancer. Moreover, as demonstrated here, significant tumor regression, rather just a slowing of tumor growth, was observed in the Ad5-yCD/mutTK_SR39rep-ADP-treated groups only when the 5-FC + GCV prodrugs were administered. Such effects were not observed with Ad5-yCD/mutTK_SR39rep-ADP itself, and after completion of the prodrug therapy course, tumor growth resumed. Taken together, these observations demonstrate that ADP does not circumvent the chemotherapeutic or radiosensitization effects of the CD/HSV-1 TK suicide gene systems and it can provide a therapeutic benefit even in the context of an armed replication-competent, oncolytic adenovirus. Based on these results, we believe that Ad5-yCD/mutTK_SR39rep-ADP will demonstrate greater efficacy than Ad5-CD/TKrep in human cancer trials and therefore is a better agent to test the hypothesis that replication-competent adenovirus-mediated suicide gene therapy can improve the outcome of conformal radiotherapy.

Materials and methods

Cell lines and adenoviruses

LNCaP cells were obtained from Dr. Leland Chung (Emory University, Atlanta, GA, USA). TRAMP C2 cells were obtained from the American Type Culture Collection (Manassas, VA, USA). All cell lines were grown in Dulbecco's modified essential medium with 10% fetal bovine serum.

The replication-competent Ad5-CD/TKrep adenovirus has been described previously^11,12,13. Details regarding the construction of Ad5-yCD/mutTK_SR39rep-ADP will be provided upon request. The entire DNA sequence (35,180 bases) of Ad5-yCD/mutTK_SR39rep-ADP has been determined and is in 99.92% agreement with the predicted sequence. There were no alterations in the yCD/mutTK_SR39 fusion or ADP genes.

Comparing the in vitro and in vivo properties of the Ad5-CD/TKrep and Ad5-yCD/mutTK_SR39rep-ADP adenoviruses, all studies were performed with an equal input of adenovirus based on viral particles. As previously described by Wold and colleagues^27,28, adenoviruses containing ADP develop plaques at a much faster rate than do adenoviruses lacking ADP. Therefore, the determination of plaque-forming units is very time dependent and can be highly variable depending on the condition of the infected cells. When plaque formation is scored 5 days postinfection, Ad5-yCD/mutTK_SR39rep-ADP typically has a vp/pfu 10 times that of Ad5-CD/TKrep. However, because Ad5-CD/TKrep (which lacks ADP) plaques develop much more slowly, the final vp/pfu of the two adenoviruses is essentially identical (10–20) when observed over a protracted period of time (i.e., 3 weeks). Thus, an equal input of vp is equivalent to an equal input of final pfu (21 days).

Western blotting

Western blotting was performed as previously described¹¹. Rabbit polyclonal antibodies to bCD, yCD, and ADP were obtained from C. Richards (Glaxo-Wellcome, Inc., Research Triangle Park, NC, USA), A. Rehemtulla (University of Michigan, Ann Arbor, MI, USA), and W. Wold (St. Louis University, St. Louis, MO, USA), respectively. ECL sheep anti-rabbit IgG horseradish peroxidase was used as secondary antibody (Amersham Biosciences, Piscataway, NJ, USA).

In vitro prodrug sensitivity, cytopathic effect, and enzyme assays

Prodrug sensitivity and CPE assays were performed as described previously¹¹. Briefly, LNCaP cells were mock infected or infected with adenovirus. One hour later, the medium was changed and cells were maintained thereafter in growth medium. For prodrug sensitivity assays, 5-FC and GCV prodrugs were maintained throughout. Wells were fixed and stained with 0.4% crystal violet at specific time points thereafter.

CD and HSV-1 TK assays were performed as described previously¹¹. Briefly, LNCaP cells were mock infected or infected with adenovirus at a vp/cell of 1000. Forty-eight hours later, cells were harvested and a cell extract was prepared by three cycles of freeze–thawing. Clarified cell extracts containing an equal amount of protein were incubated with [¹⁴C]cytosine and 5-[¹⁴C]fluorocytosine for CD assays and [³H]ganciclovir for HSV-1 TK assays at 37°C for 1 h. Substrates and products were resolved by thin-layer chromatography, the chromatogram was exposed to X-ray film, and the amount of radioactivity in each reaction species was quantified using a Bio-Rad FX molecular phosphoimager.

In vivo efficacy and toxicology studies

Animal studies were approved by the Institutional Animal Care and Use Committee of the Henry Ford Health System and Good Laboratory Practices were used throughout. LNCaP tumors were established by inoculating 2 10⁶ cells prepared in 0.9% NaCl and 50% Matrigel in the right gastrocnemius muscle (im) of male SCID mice (20–22 g). Upon reaching 150 mm³, tumors were injected with PBS or adenovirus (10¹⁰ vp, 50 l) on days 1, 3, and 5. Half the mice in each treatment group were administered daily intraperitoneal (ip) injections of 5-FC (500 mg/kg/day) and GCV (30 mg/kg/day) for 1 week (days 5–11). Tumor dimensions were measured every 2–3 days.

Toxicology studies were designed with input from the FDA. In the first study, 102 male C57BL/6 mice (20–22 g) were injected intraprostatically with 10¹⁰ vp (10 l) Ad5-yCD/mutTK_SR39rep-ADP (Groups 1–4), 10¹⁰ vp (10 l) Ad5-CD/TKrep (Groups 7 and 8), or PBS (Groups 5 and 6) on day 1. Animals that were scheduled to receive prodrugs (Groups 1, 2, and 7) were given 14 daily ip injections of 5-FC (500 mg/kg/day) and GCV (30 mg/kg/day) (days 3–16). Other groups received saline in place of prodrugs. Animals that were scheduled to receive radiation (Groups 1, 3, 5, and 7) received 4 Gy of -irradiation to their left (2 Gy) and right (2 Gy) pelvic region twice per week for 7 weeks for a total dose of 56 Gy using a ¹³⁷Cs irradiator.

In the second study, 60 male C57BL/6 mice (20–22 g) were injected intraprostatically with 10¹⁰ vp (10 l) Ad5-yCD/mutTK_SR39rep-ADP on day 1 (Groups 1 and 2) and day 22 (Group 2) or injected with PBS (Group 3). Animals in treatment groups received 13 ip injections (weekdays only) of 5-FC (500 mg/kg/day) and GCV (30 mg/kg/day) in one (Group 1) or two (Group 2) 2.6-week cycles following each adenovirus injection. Prodrugs were administered on weekdays only to mimic future clinical trials. Animals in treatment groups (Groups 1 and 2) received 4 Gy of -irradiation to their left (2 Gy) and right (2 Gy) pelvic region twice per week for 7 weeks for a total dose of 56 Gy using a small-animal linear accelerator.

In both studies, animals were examined for a number of toxicological parameters: general observations were noted daily; body weights were taken weekly and at each necropsy; gross observations were noted at time of necropsy; and at necropsy either a partial or a full set of major abdominal tissues was taken for histopathological analyses. A blood sample was taken at every necropsy for complete blood cell counts (CBC) and blood chemistries. CBC included absolute counts for white blood cells (WBC), red blood cells (RBC), lymphocytes, neutrophils, eosinophils, basophils, and platelets. Blood chemistries included determination of hemoglobin (HGB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), bilirubin, alkaline phosphatase, and albumin. CBC was performed by Laboratory Corporation of America (Burlington, NC, USA). Histopathology was performed by Experimental Pathology Laboratories, Inc. (Herndon, VA, USA). Statistical analyses were conducted by the Department of Biostatistics and Epidemiology at the Henry Ford Health System.

Quantification of reporter gene expression volume in vivo

In the mouse studies, tumor cells were implanted into both hind legs of CD-1 (nu/nu) athymic mice. When tumors reached 150 mm³, 1 10¹⁰ vp (50 l) of replication-competent hNIS-containing adenoviruses, one lacking (Ad5-hNISrep) and the other containing (Ad5-hNIS-ADPrep) the Ad5 ADP gene, was injected into contralateral tumors of each mouse. One day later, animals were administered 0.15 mCi Na^99mTcO₄ ip. One hour later, tumors were excised, frozen, and completely sectioned in 100-m-thick sections, and sections were exposed to X-ray film for 60 h. In the dog studies, 1 10¹¹ vp (100 l) of the replication-competent hNIS-containing adenoviruses (one containing and one lacking the ADP gene) was injected into contralateral sides of the naïve canine prostate. The next day, animals were administered 20 mCi Na^99mTcO₄ ip. One hour later, the prostate was excised, frozen, and completely sectioned in 100-m-thick sections, and sections were exposed to X-ray film for 60 h. In both mouse and dog studies, autoradiograms were developed and digitized and the volume of reporter gene expression was determined using methods previously described^32,33.

References

1. Ganly, I., et al. (2000). A phase I study of ONYX-015, an E1B attenuated adenovirus, administered intratumorally to subjects with recurrent head and neck cancer. Clin. Cancer Res. 6: 798–806. | PubMed | ISI | ChemPort |
2. Nemunaitis, J., et al. (2001). Phase II trial of intratumoral administration of ONYX-015, a replication-selective adenovirus, in subjects with refractory head and neck cancer. J. Clin. Oncol. 19: 289–298. | PubMed | ISI | ChemPort |
3. Nemunaitis, J., et al. (2000). Selective replication and oncolysis in p53 mutant tumors with ONYX-015, an E1B-55kD gene-deleted adenovirus, in subjects with advanced head and neck cancer: a phase II trial. Cancer Res. 60: 6359–6366. | PubMed | ISI | ChemPort |
4. Khuri, F., et al. (2000). A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in subjects with recurrent head and neck cancer. Nat. Med. 6: 879–885. | Article | PubMed | ISI | ChemPort |
5. Hecht, J., et al. (2003). A phase I/II trial of intratumoral endoscopic ultrasound injection of ONYX-015 with intravenous gemcitabine in unresectable pancreatic carcinoma. Clin. Cancer Res. 9: 555–561. | PubMed | ISI | ChemPort |
6. Mulvihill, S., et al. (2001). Safety and feasibility of injection with an E1B-55 kDa gene-deleted, replication-selective adenovirus (ONYX-015) into primary carcinomas of the pancreas: a phase I trial. Gene Ther. 8: 308–315. | Article | PubMed | ChemPort |
7. Reid, T., et al. (2001). Intra-arterial administration of a replication-selective adenovirus (dl1520) in subjects with colorectal carcinoma metastatic to the liver: a phase I trial. Gene Ther. 8: 1618–1626. | Article | PubMed | ISI | ChemPort |
8. Vasey, P., et al. (2002). Phase I trial of intraperitoneal injection of the E1B-55-kd-gene-deleted adenovirus ONYX-015 (dl1520) given on days 1 through 5 every 3 weeks in subjects with recurrent/refractory epithelial ovarian cancer. J. Clin. Oncol. 20: 1562–1569. | Article | PubMed | ISI | ChemPort |
9. Chiocca, E. A., et al. (2004). A phase I open-label, dose-escalation, multi-institutional trial of injection with an E1B-attenuated adenovirus, ONYX-015, into the peritumoral region of recurrent malignant gliomas, in the adjuvant setting. Mol. Ther. 10: 958–966. | Article | PubMed | ISI | ChemPort |
10. DeWeese, T., et al. (2001). A phase I trial of CV706, a replication-competent, PSA selective oncolytic adenovirus, for the treatment of locally recurrent prostate cancer following radiation therapy. Cancer Res. 61: 7464–7472. | PubMed | ISI | ChemPort |
11. Freytag, S. O., Rogulski, K. R., Paielli, D. L., Gilbert, J. D. and Kim, J. H. (1998). A novel three-pronged approach to selectively kill cancer cells: concomitant viral, double suicide gene, and radiotherapy. Hum. Gene Ther. 9: 1323–1333. | PubMed | ISI | ChemPort |
12. Rogulski, K. R., Wing, M., Paielli, D. L., Gilbert, J. D., Kim, J. H. and Freytag, S. O. (2000). Double suicide gene therapy augments the therapeutic efficacy of a replication-competent lytic adenovirus through enhanced cytotoxicity and radiosensitization. Hum. Gene Ther. 11: 67–76. | Article | PubMed | ISI | ChemPort |
13. Freytag, S. O., et al. (2002). Efficacy and toxicity of replication-competent adenovirus-mediated double suicide gene therapy in combination with radiation therapy in an orthotopic mouse prostate cancer model. Int. J. Radiat. Oncol. Biol. Phys. 54: 873–886. | Article | PubMed | ChemPort |
14. Kim, J. H., Kim, S. H., Brown, S. L. and Freytag, S. O. (1994). Selective enhancement by an antiviral agent of the radiation-induced cell killing of human glioma cells transduced with HSV-tk gene. Cancer Res. 54: 6053–6056. | PubMed | ISI | ChemPort |
15. Kim, J. H., Kim, S. H., Kolozsvary, A., Brown, S. L., Kim, O. B. and Freytag, S. O. (1995). Selective enhancement of radiation response of herpes simplex virus thymidine kinase transduced 9L gliosarcoma cells in vitro and in vivo by antiviral agents. Int. J. Radiat. Oncol. Biol. Phys. 33: 861–868. | Article | PubMed | ISI | ChemPort |
16. Khil, M., Kim, J. H., Mullen, C. A., Kim, S. H. and Freytag, S. O. (1996). Radiosensitization by 5-fluorocytosine of human colorectal carcinoma cells in culture transduced with cytosine deaminase gene. Clin. Cancer Res. 2: 53–57. | PubMed | ChemPort |
17. Rogulski, K. R., Kim, J. H., Kim, S. H. and Freytag, S. O. (1997). Glioma cells transduced with an E. coli CD/HSV-1 TK fusion gene exhibit enhanced metabolic suicide and radiosensitivity. Hum. Gene Ther. 8: 73–85. | PubMed | ISI | ChemPort |
18. Rogulski, K. R., Zhang, K., Kolozsvary, A., Kim, J. H. and Freytag, S. O. (1997). Pronounced anti-tumor effects and tumor radiosensitization of double suicide gene therapy. Clin. Cancer Res. 3: 2081–2088. | PubMed | ChemPort |
19. Kim, J. H., Kolozsvary, A., Rogulski, K. R., Khil, M. and Freytag, S. O. (1998). Double suicide fusion gene: a selective radiosensitization of 9L glioma in rat brain. Can. J. Sci. Am. 4: 364–369. | ChemPort |
20. Xie, Y., Gilbert, J. D., Kim, J. H. and Freytag, S. O. (1999). Efficacy of adenovirus-mediated CD/5-FC and HSV-1 TK/GCV suicide gene therapies concomitant with p53 gene therapy. Clin. Cancer Res. 5: 4224–4232. | PubMed | ChemPort |
21. Freytag, S. O., et al. (2002). Phase I study of replication-competent adenovirus-mediated double suicide gene therapy for the treatment of locally recurrent prostate cancer. Cancer Res. 62: 4968–4976. | PubMed | ISI | ChemPort |
22. Freytag, S. O., et al. (2003). Phase I study of replication-competent adenovirus-mediated double suicide gene therapy in combination with conventional-dose three-dimensional conformal radiation therapy for the treatment of newly diagnosed, intermediate- to high-risk prostate cancer. Cancer Res. 63: 7497–7506. | PubMed | ISI | ChemPort |
23. Kievit, E., et al. (1999). Superiority of yeast over bacterial cytosine deaminase for enzyme/prodrug gene therapy in colon cancer xenografts. Cancer Res. 59: 1417–1421. | PubMed | ISI | ChemPort |
24. Hamstra, D., Rice, D., Fahmy, S., Ross, B. and Rehemtulla, A. (1999). Enzyme/prodrug therapy for head and neck cancer using a catalytically superior cytosine deaminase. Hum. Gene Ther. 10: 1993–2003. | Article | PubMed | ISI | ChemPort |
25. Black, M., Newcomb, T., Wilson, H. and Loeb, L. (1996). Creation of drug-specific herpes simplex virus type 1 thymidine kinase mutants for gene therapy. Proc. Natl. Acad. Sci. USA. 93: 3525–3529. | Article | PubMed | ChemPort |
26. Black, M., Kokoris, M. and Sabo, P. (2001). Herpes simplex virus-1 thymidine kinase mutants created by semi-random sequence mutagenesis improve prodrug-mediated tumor cell killing. Cancer Res. 61: 3022–3026. | PubMed | ISI | ChemPort |
27. Tollefson, A., Scaria, A., Hermiston, T., Ryerse, J., Wold, L. and Wold, W. S. (1996). The adenovirus death protein (E3-11.6 K) is required at very late stages of infection for efficient cell lysis and release of adenovirus from infected cells. J. Virol. 70: 2296–2306. | PubMed | ISI | ChemPort |
28. Tollefson, A., Ryerse, J., Scaria, A., Hermiston, and Wold, W. S. (1996). The E3-11.6 kDa adenovirus death protein (ADP) is required for efficient cell death: characterization of cells infected with adp mutants. Virology. 220: 152–162. | Article | PubMed | ISI | ChemPort |
29. Doronin, K., Toth, K., Kuppuswamy, M., Ward, P., Tollefson, A. and Wold, W. S. (2000). Tumor-specific, replication-competent adenovirus vectors overexpressing the adenovirus death protein. J. Virol. 74: 6147–6155. | Article | PubMed | ISI | ChemPort |
30. Doronin, K., et al. (2001). Tissue-specific, tumor-selective, replication-competent adenovirus vector for cancer gene therapy. J. Virol. 75: 3314–3324. | Article | PubMed | ISI | ChemPort |
31. Toth, K., et al. (2003). Radiation increases the activity of oncolytic adenovirus cancer gene therapy vectors that overexpress the ADP (E3-11.6K) protein. Cancer Gene Ther. 10: 193–200. | Article | PubMed | ChemPort |
32. Barton, K., et al. (2003). GENIS: gene expression of sodium iodide symporter for non-invasive imaging of gene therapy vectors and quantification of gene expression in vivo. Mol. Ther. 8: 508–518. | Article | PubMed | ChemPort |
33. Barton, K., et al. (2004). A quantitative method for measuring gene expression magnitude and volume delivered by gene therapy vectors. Mol. Ther. 9: 625–631. | Article | PubMed | ChemPort |