We have demonstrated previously the oncolytic effects of a systemically delivered, replicating vaccinia virus. To enhance the tumor specificity of this vector, we have developed a combined thymidine kinase-deleted (TK−) and vaccinia growth factor-deleted (VGF−) vaccinia virus and investigated its properties in vitro and in vivo. The gene for enhanced green fluorescent protein (EGFP) was inserted into the TK locus of a VGF− vaccinia virus by homologous recombination creating a double-deleted mutant vaccinia virus (vvDD-GFP). Infection of resting and dividing NIH3T3 cells with vvDD-GFP yielded reduced viral recovery compared with wild-type (WT), TK−, or VGF− viruses from resting cultures but equivalent virus recovery from dividing cultures. Eight days after nude mice were injected i.p. with 107 plaque-forming units (pfu) of WT, TK−, VGF−, or vvDD-GFP vaccinia virus, tissues and tumor were harvested for viral titer determination. No virus was recovered from the brains of mice injected with vvDD-GFP compared with the other viruses, which ranged from 130 to 28,000 pfu/mg protein; however, equivalent amounts were recovered from tumor. There was no toxicity from vvDD-GFP because nude mice receiving 108 pfu of IP vvDD-GFP lived >100 days, whereas mice receiving WT, VGF−, or TK− virus had median survivals of only 6, 17, and 29 days, respectively. Similar results were seen when 109 pfu of vvDD-GFP were given. Nude mice bearing s.c. murine colon adenocarcinoma (MC38) had significant tumor regression after treatment with 109 pfu of systemic (i.p.) vvDD-GFP compared with control (mean tumor size, 180.71 ± 35.26 mm3versus 2796.79 ± 573.20 mm3 12 days after injection of virus). Our data demonstrate that a TK− and VGF− mutant vaccinia virus is significantly attenuated in resting cells in vitro and demonstrates tumor-specific replication in vivo. It is a promising vector for use in tumor-directed gene therapy, given its enhanced safety profile, tumor selectivity, and the oncolytic effects after systemic delivery.

Successful tumor-directed gene therapy is dependent upon a high percentage of tumor cells expressing a large quantity of gene product after systemic injection of a vector. Transduction efficiencies and levels of gene expression of currently available vectors remain low and limit their potential therapeutic effects, despite modification of these vectors to allow tumor targeting and tumor-specific gene expression (1, 2, 3, 4).

Recently, replicating viruses have been explored and offer several advantages. Levels of gene expression are higher, transduction efficiency is improved through replication and infection of surrounding cells, and antitumor effects are seen attributable to virus-mediated cell death (5, 6, 7). Currently, viruses such as adenovirus, herpes virus, Newcastle disease virus, and vaccinia virus are being used as replicating vectors (6, 8, 9, 10, 11, 12). Vector-associated toxicity is a concern, and various modifications have been explored in an effort to improve their tumor specificity and safety. Previously, herpes virus vectors have been modified for gene therapy in a similar fashion using mutant vectors with inactivation of the thymidine kinase or ribonucleotide reductase genes (10, 13, 14, 15). This is the first report of a mutant vaccinia virus with multiple selective mutations to enhance tumor specificity.

Vaccinia virus has been used as a live vaccine in the smallpox eradication program and recently as a vaccine against cancer (16). It has not been widely accepted as a potential tumor-directed gene therapy vector because of concerns regarding the safety of a systemically administered replicating virus. Vaccinia has been modified previously to carry various antigens, cytokines, and immunostimulatory molecules including carcinoembryonic antigen, gp100, MART-1, granulocyte/macrophage-colony stimulating factor, B7-1, IL3-1β, IL-2, and IL-12 (17, 18, 19, 20, 21, 22, 23, 24, 25, 26). These studies have provided insights into the potential toxicities of systemically delivered replicating vaccinia virus. Although it is generally a safe vector, case reports of generalized vaccinia and vaccinia-associated encephalitis have been described, usually in the immunosuppressed population (27, 28, 29, 30, 31). Strategies toward improving the safety of this vector have been described (16, 24, 25, 32, 33, 34).

Previously, deletion of either the TK gene or VGF genes was shown to significantly decrease pathogenicity compared with WT virus (32, 33). A TK− virus requires TTP for DNA synthesis from the nucleotide pool present in dividing cells. This leads to preferential viral replication in dividing cells and is the presumed explanation for the observed tumor specificity. We have shown previously that a systemically delivered TK− vaccinia virus expressing the firefly luciferase gene resulted in up to 3 logs higher gene expression in murine tumors compared with normal tissues (7, 35, 36). Further improvement in both tumor specificity and safety of vaccinia is required before its use as a systemic gene therapy vector in humans.

VGF is a secreted protein produced early in viral infection and acts as a mitogen to prime surrounding cells for vaccinia infection (37). Deletion of this growth factor causes decreased viral replication in resting cells and a 1000-fold increase in the LD50 of intracranial vaccinia (33). The combined effect of TK and VGF deletions on tumor specificity should be synergistic. In the absence of TK, viral replication will require TTP from dividing cells. The normal stimulation of surrounding cells to divide will not occur in the absence of VGF; hence, replication will occur only in actively dividing cells. As well as decreasing pathogenicity, this is expected to maintain or enhance the tumor selectivity reported previously (7, 35, 36). Here we describe a unique tumor selective vaccinia gene therapy vector with deletions of both the TK gene and VGF genes. We examine in vitro replication, in vivo biodistribution, viral pathogenicity, and antitumor effects of this new vector in a mouse model.

Cell Lines.

Human cervical adenocarcinoma (HeLa S3), monkey kidney fibroblasts (CV1), and murine NIH3T3 cells were obtained from the American Type Culture Collection (Manassas, VA). HUTK-143B cells were originally obtained from Dr. K. Huebner (Kimmel Cancer Center, Philadelphia, PA). MC38, a nonmetastatic colon adenocarcinoma cell line from C57BL/6 mice, was originally induced with oral dimethylhydrazine. It has been passaged over many years and is used extensively in our branch. All cell lines were grown in DMEM or MEM (HUTK-143B) supplemented with 10% heat-inactivated FCS, 2 mm glutamine, 1% penicillin/streptomycin (10,000 units/ml) and 0.2% Fungizone (250 μg/ml; all Biofluids, Rockville, MD). Cell lines were maintained in an incubator at 37°C with 5% CO2 and serially passaged every 3–4 days.

Vaccinia Viruses.

The recombinant (WR strain) vaccinia viruses F13L+ (WT virus with lacZ gene insertion; Ref. 38), VSC20 (VGF− virus with lacZ gene insertion; Ref. 33), and VJS6 (TK− virus with lacZ gene insertion; Ref. 35) were used in these experiments.

Creation of a TK Gene and Vaccinia Growth Factor Gene Deleted Virus.

VSC20, a vaccinia virus that has the lacZ gene inserted into its VGF sites (33), was used as the background virus. A vaccinia shuttle plasmid was created that allowed for homologous recombination of EGFP into the TK locus of VSC20, creating vvDD-GFP. Briefly, pEGFP (Clontech, Palo Alto, CA) was digested with SalI and SpeI. The segment containing EGFP was ligated into the SalI and SpeI sites of our shuttle plasmid pCB023-II (12), creating pSEL-EGFP (Fig. 1 A). This placed the EGFP gene under the control of the vaccinia synthetic early/late promoter (39). It is flanked by portions of the vaccinia TK gene, which allows for homologous recombination into this locus. Confluent wells of CV1 cells were infected for 2 h at 37°C with 1.4 × 105 pfu of VSC20 in 1.0 ml of MEM-2.5% FCS. Supernatants were removed, and a liposomal transfection (Superfect; Qiagen, Santa Clarita, CA) of pSEL-EGFP was performed using 0.7 ml of DMEM-10% FCS containing 2 μg of plasmid DNA and 10 μl of liposomes/well at 37°C for 4 h. The transfection medium was removed and replaced with 3.0 ml of DMEM-10% FCS. After 3 days of incubation, cells were collected and sonicated in MEM-2.5% FCS. Serial dilutions (1.0 ml; 10−2 to 10−4 in MEM-2.5% FCS) were used to infect HUTK-143B cells at 37°C. After 2 h, 2 ml of MEM-10% FCS containing bromodeoxyuridine (final concentration, 25 μg/ml; Sigma Chemical Co. Chemical Co, St. Louis MO) were added for selection. After 24–48 h, cells were observed for green fluorescence and viral plaque formation. Six positive plaques were isolated, resuspended in MEM-2.5% FCS, and used to reinfect further HUTK-143B cells. After three to four cycles of selection, all plaques were positive for GFP.

DNA Extraction.

Confluent CV1 cells were infected with unpurified recombinant vaccinia virus. After 2–3 days when complete viral cytopathic effect was seen, supernatant was removed, and cells were washed and harvested in 750 μl of PCR buffer [50 mm KCl, 10 mm Tris-Cl, 2.5 mm MgCl2, 0.1 mg/ml gelatin, and 0.45% NP40 (Sigma Chemical Co.), 0.45% Tween 20 (Bio-Rad, Richmond CA)]. Proteinase K (Life Technologies, Inc., Rockville MD; 4.5 μl of 10 mg/ml) was added, and DNA was prepared by incubation for 1 h at 55°C. Proteinase K was inactivated (10 min at 95°C) prior to PCR, as described previously (40).

PCR.

A standard PCR was performed using primers external to the site of recombination (P1, sense 5′-ATCGCATTTTCTAACGTGATGGAT-3′; P2, antisense 5′-TATCTAACGACACAACATCCATT-3′), within the newly recombined section (P3, EGFP sense 5′-ATGGTGAGCAAGGGCGAGGAGC-3′; P4, GPT sense 5′-ATACATCGTCACCTGGGACATG-3′) and within the VGF gene (sense 5′-TGACGGGGTCACCCACACTGTGCCCATCTA-3′ and antisense 5′-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3′). In brief, 25 μl of the digested vaccinia DNA, 1 μl of each primer, 1 μl deoxynucleotide triphosphates (10 mm; Life Technologies, Inc.), 2.5 units Taq polymerase (Promega Corp., Madison WI), and PCR buffer to 50 μl final volume were added. PCR parameters consisted of 15 s denaturing (94°C), 30 s annealing (55°C), and 2 min extension (72°C) for 35 cycles (GeneAmp PCR System 9700; Perkin-Elmer, Norwalk CT).

In Vitro Virus Yield.

Dividing NIH3T3 cells were infected with 200 pfu (multiplicity of infection, ∼0.0005) of either F13L+, VJS6, VSC20, or vvDD-GFP in 1 ml of MEM-2.5% FCS for 2 h at 37°C. DMEM-10% FCS was added, and cells were incubated until harvesting at 24, 48, and 72 h after infection. To establish resting cultures, confluent NIH3T3 cells were washed with PBS and incubated for 5 days in DMEM with 5% FCS (33). These resting cultures were infected as above and harvested at the same time points after infection. After a single freeze-thaw cycle, virus was quantified by plaque titering on CV1 cells as described previously (41).

Mice.

Female athymic (C57BL/6) and C57BL/6 immunocompetent mice, 6 weeks of age, were obtained from the NIH small animal facility (Frederick, MD). They were housed in standard conditions and given food and water ad libitum. All animal studies were approved by the Animal Care and Use Subcommittee of the Animal Sciences Branch, National Cancer Institute.

In Vivo Viral Pathogenicity.

Viral pathogenicity was assessed with tissue histology, in vivo viral replication, and mouse survival. Seven days after i.p. injection of 107 pfu of F13L+, VJS6, VSC20, or vvDD-GFP, whole sections of brain, liver, spleen, testes, bone marrow, ovary, and tumor were homogenized in HBSS (Biofluids) and kept at −70°C until use. Five hundred μl of the homogenate were incubated on CV1 cells at 37°C in 5% CO2, and titers were determined as described previously. Viral titers were standardized to total protein.

Survival studies were performed on 6-week-old nude or immunocompetent mice. Non-tumor-bearing mice were injected i.p. with 108 to 109 pfu of F13L+, VJS6, VSC20, or vvDD-GFP in 2 ml of HBSS and followed for survival.

Immunohistochemistry.

Five days after i.p. injection of 108 pfu of F13L+, VJS6, VSC20, vvDD-GFP, and HBSS control into nude mice, whole sections of brain, liver, spleen, ovary, and tumor were harvested, fixed in 10% formalin (Fisher Scientific, Pittsburgh, PA), embedded in paraffin, and stained with H&E stain. Selected tissues were used for vaccinia virus antigen immunohistochemistry (Fig. 3). A rabbit polyclonal anti-vaccinia antibody was used at a dilution of 1:8,000–16,000 with the Vectastain Rabbit Elite kit (Vector Laboratories, Inc., Burlingame, CA). Diaminobenzidine was the chromogen. All results were interpreted by a pathologist (J. M. W.) blind to the virus treatment of the sections.

Antitumor Effect.

MC38 tumor cells (105) in 100 μl of DMEM were injected s.c. into the right flanks of 6-week-old female athymic (C57BL/6) mice and allowed to grow for 7–10 days. When the tumors reached 75–125 mm3 in volume, 109 pfu of vvDD-GFP or HBSS control were injected i.p. (in 2 ml of HBSS/0.1% BSA; Calbiochem, La Jolla, CA). Tumors were measured by an investigator blind to the treatment of the animals. Tumor volume was calculated as [(width)2 × length]0.52 (42).

Statistics.

Statistical analysis was performed using the Mann-Whitney test for nonparametric data when appropriate. Tumor volumes between groups were assessed using the ANOVA for repeated measures. Survival analysis was performed using the method of Kaplan-Meier (43), and differences between curves were assessed using the log-rank test (44). All statistics were generated using StatView Software (Abacus Concepts, Inc., Berkeley, CA), and Ps <0.05 were considered significant.

Creation of a TK and Vaccinia Growth Factor Deleted Vaccinia Virus.

A shuttle plasmid containing the gene for EGFP (Clontech, Palo Alto, CA) was created (Fig. 1,A). The parental VGF-deleted virus (VSC20) was created previously by the insertion of the lacZ gene into the VGF sites (33). The shuttle plasmid was used to insert EGFP into the TK locus of the VGF-deleted vaccinia virus by homologous recombination as described above, creating the double-deleted virus, vvDD-GFP. PCR using primers designed to amplify the TK gene, VGF gene, and spanning the sites of recombination confirmed the deletion of these genes (Fig. 1 B). Staining of vvDD-GFP infected cells with 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (ICN Biomedicals, Inc., Aurora, OH) demonstrated the presence of the lacZ gene and fluorescent microscopy confirmed the presence of the EGFP gene (data not shown).

In Vitro Replication of vvDD-GFP.

Synchronous infection of dividing (nonconfluent) NIH3T3 cells with WT, VGF−, TK−, or vvDD-GFP vaccinia yielded similar amounts of virus at various time points after infection (Fig. 2,A). However, when resting (confluent) NIH3T3 cells were used, the yield of vvDD-GFP was significantly less than the other viruses at all time points (Fig. 2 B). This supports the hypothesis that a vaccinia virus deleted of both TK and VGF would be attenuated in resting cells but show comparable growth to WT virus in dividing cells that have adequate sources of nucleotides for DNA synthesis.

In Vivo Replication of vvDD-GFP.

Tumor-bearing (s.c. MC38) and non-tumor-bearing nude mice were injected with 107 pfu i.p. of WT, VGF−, TK−, or vvDD-GFP vaccinia. Eight days after injection of virus, samples of brain, liver, spleen, testes, bone marrow, ovary, and tumor were harvested, titered on CV1 cells, and viral yield was calculated per milligram protein (Table 1). vvDD-GFP was recovered at significantly lower titers in brain compared with WT virus and markedly reduced titers in spleen and bone marrow. It was also recovered at lower titers compared with the TK-deleted or VGF-deleted viruses, particularly in the brain. Notably, all four viruses were equally infective in tumor and ovary. Previously, we have shown that although virus was recovered from the bone marrow, no myelosuppression was apparent (7). Patterns of infectivity were the same for the non-tumor-bearing group (data not shown). When this experiment was repeated in immunocompetent mice, similar patterns were seen; however, the peak infectivity occurred earlier, and most organs were negative for all four viruses by day 8 (data not shown).

Histology of these organs, 5 days after injection with virus, supported the viral recovery data (Table 2). Although all sections of brain were found to have a mild vasculitis on H&E staining, this did not necessarily correspond to the degree of viral staining on immunohistochemistry. Immunohistochemistry demonstrated more dramatic differences with less viral antigen in the brain and ovary after vvDD-GFP infection compared with the other viruses, whereas tumor staining was remarkably similar for all four viruses (Fig. 3). In the ovary, viral antigens were found in follicles and in adjacent ovarian stromal tissues. Oviducts were often immunoreactive. Primary tumors had strong immunoreactivity within viable tumor cells. Necrotic areas were less immunoreactive or nonimmunoreactive.

In Vivo Pathogenicity of vvDD-GFP.

Nude mice were injected with systemic (108 pfu i.p.) WT, TK−, VGF−, or vvDD-GFP vaccinia virus and followed for survival (Fig. 4,A). As expected, WT virus was extremely virulent, with nude mice surviving a median of 5 days. Both the VGF− and TK− virus were attenuated compared with the WT virus (median survival in nude mice, 17 and 29, days respectively; P < 0.0001 compared with WT) as expected from previous reports (32, 33). vvDD-GFP was highly attenuated in nude mice, and all lived >100 days (P < 0.0001). When this experiment was repeated using a 10-fold higher dose (109 pfu) of vvDD-GFP, similar results were seen with an insignificant left shift of the curve (Fig. 4,B). 109 pfu were still significantly less toxic than 108 of the TK− virus (P < 0.0001). Interestingly, when this study was repeated in C57BL/6 immunocompetent mice, less toxicity was seen in all groups except the group receiving WT virus, which remained very toxic (Fig. 4 C). In experiments using nude mice, the animals receiving 109 pfu of vvDD-GFP began to die from viral pathogenicity before 100 days, as shown. This was not seen in the immunocompetent mice. Likely the virus is never eliminated by the immune system in athymic mice.

Antitumor Effect.

As demonstrated above, vvDD-GFP is able to selectively replicate in tumor tissues. Nude mice bearing s.c. MC38 tumors were injected systemically (i.p.) with 109 pfu of vvDD-GFP or HBSS control. A significant antitumor effect was seen in the mice treated with vvDD-GFP (P < 0.001), including one complete response (Fig. 5). Remarkably, this antitumor effect is attributable to the replication of virus alone because no therapeutic genes have been included.

The creation of a TK- and VGF-deleted vaccinia virus offers several advantages for use as a tumor-directed vector for cancer gene therapy. We have demonstrated decreased replication of this virus both in vitro and in vivo in nondividing cells. Decreased viral pathogenicity was demonstrated, including minimal recovery of vvDD-GFP from the brain tissue of nude mice. As well, a significant antitumor effect was seen after systemic injection because of selective replication of this virus in tumor cells.

The inefficiency of currently available vectors underscores the need for a highly efficient and safe vector such as the one described. Vaccinia virus has several advantages for tumor-directed gene therapy (34, 45). It is a large, cytoplasmic virus, and up to 25 kb of foreign DNA can be inserted without the need for viral deletions (46). A major advantage of vaccinia virus is that it uses viral factors for DNA and RNA synthesis in the cytoplasm and is therefore less dependent on the cellular factors than other viruses. Remarkably, the native and synthetic vaccinia promoters in use are very strong and contribute to its efficiency (39, 47, 48).

The most important feature of this virus, however, is the ability to selectively replicate and express genes in tumor tissues compared with normal tissues. We have demonstrated previously a 3-log higher expression of luciferase in tumors compared with normal tissue in a s.c. tumor model after systemic administration of the virus (7). The complete explanation for this tumor selectivity has yet to be elucidated; however contributing factors include the enhanced ability of macromolecules to extravasate through permeable tumor vasculature (49) and the replication selectivity shown by vaccinia within the preferred metabolically active environment of tumor cells. Evidence for the contribution of permeable vasculature comes from smallpox literature. Smallpox was known to replicate preferentially in areas of increased vascular permeability secondary to injury and histamine release (50). The biodistribution of vaccinia (7, 35, 36), predominantly to tumor and ovarian follicles (both sites of increased vascular permeability; Refs. 49, 51), is also suggestive. Recently, we have shown that hyperthermia, which increases vascular permeability, leads to increased uptake of vaccinia virus.4 Of interest, ovaries have been shown to have high levels of VEGF (which also increases vascular permeability) and may explain the propensity for the virus to localize there as well as in tumor (52). Tumor-selective replication has been demonstrated previously (7, 53) and is thought to be largely attributable to the rapid division of tumor cells that provide nucleotides, specifically TTP, to complement a TK-deleted vector. The hypothesis for the high level of tumor selectivity shown by the current vector is a combination of the two. vvDD-GFP travels intravascularly and escapes at sites of increased vascular permeability, such as the tumor and ovary. Because of its ultimate reliance on dividing cells for replication, it is only able to replicate efficiently within tumor cells or ovarian follicles. Other sites of cellular replication such as bone marrow and gastrointestinal mucosa do not demonstrate the same levels of vaccinia infection, possibly because of a lack of leaky vasculature.

In addition to the high levels of gene expression resulting from the tumor-specific replication, we have shown that vaccinia alone, in the absence of a therapeutic gene, is capable of causing an antitumor effect from viral replication and subsequent cell death. This was dramatic in the rapidly growing MC38 cell line described above and will likely be similar in other cancer cell lines. The fact that a single systemic injection resulted in a complete antitumor response is an illustration of the efficient replication of this vector in tumor tissue. Despite a small percentage of cells initially infected, within 12 days the virus spreads from cell to cell until a complete response is achieved. The effect in slow-growing human tumors and in immunocompetent models is currently under investigation.

The potential for serious infection leading to disease or death is a major concern when using any replicating virus. Extensive human trials with vaccinia virus as a smallpox vaccine showed mainly local side effects in normal subjects. However, encephalitis in infants and spreading infection in immunodeficient individuals, specifically those with deficits in cellular immunity were recognized (28, 30, 54, 55). The latter corresponds with current thinking regarding the importance of cell-mediated immunity in recovery from poxvirus infections (54). Currently, no human trials exist delivering replicating vaccinia systemically, and the development of a safe, tumor-selective vaccinia virus for this purpose is timely. Vaccinia has been used successfully in humans as an intratumoral vector with no reported toxicity (53, 56, 57, 58).

The major advantage of the currently described vector is the marked decrease in in vivo toxicity demonstrated. The use of a nude mouse model stringently tested the pathogenicity of this vector. Of interest, viral replication in tumor was preserved, and infection of the brain with vaccinia was significantly abrogated. This was demonstrated by immunohistochemistry for vaccinia in the organs of interest as well as viral recovery from these organs. The decrease in vaccinia recovered from the brain may explain the decreased pathogenicity because postvaccinal encephalitis has been a complication after smallpox vaccination (31). Decreased pathogenicity from all but the WT vaccinia was seen in immunocompetent mice as expected because there is an excellent immune response to vaccinia virus in mice. This is unlike the situation seen with adenoviral vectors, where the immune response seems to enhance toxicity.

Disadvantages to vaccinia and other viral vectors for cancer gene therapy still remain. The rapid immune response and clearance of viruses limit their utility in immunocompetent hosts, although the immune response to a tumor-specific virus may be an advantage if it leads to antitumor immunity as well. The initial injection results in the rapid formation of circulating, neutralizing antibodies, which limits the use of repeated injections; however, strategies for masking viruses from these neutralizing antibodies, such as coating them in liposomes, polyethylene glycol, or other biological agents, are under investigation (59, 60). As well, the previous immunization program for smallpox created a population that is generally immune to vaccinia. However, smallpox immunization was discontinued in 1978; therefore, future cancer patients will not have preformed immunoreactivity against this vector. Finally, the propensity of vaccinia to infect ovaries may lead to sterility if this treatment were to be considered for female cancer patients; however, this risk is currently unknown, and the virus may target only the corpus luteum.

We have created a tumor-selective vaccinia virus by deletion of both its TK gene and VGF genes. It is capable of selective tumor replication and is significantly less pathogenic than other forms of the virus. The potential utility of this vector is broad in that it may be used as an antitumor agent on its own or by expressing suicide or cytokine genes. It may be useful in immunotherapy trials expressing tumor-associated antigens and costimulatory molecules. Perhaps most importantly, given its large capacity for foreign gene insertion, it can be used for all of the above. Its efficiency and selectivity compares favorably with other replicating viral vectors currently in clinical trials (61). This is a new vector for the tumor-directed gene therapy of cancer and worthy of consideration as a systemic vector in human cancer trials.

Fig. 1.

A, linear diagram of the pSEL-EGFP shuttle plasmid and parental vaccinia VSC20 used to create vvDD-GFP. Sites of left (vTK-L) and right (vTK-R) segments of vaccinia TK gene, synthetic early/late promoter (Pse/I), 7.5 promoter (p7.5), ampicillin resistance gene, and xanthine-guanine phosphoribosyl transferase (gpt) gene (a gene for viral selection, not used in this study) are indicated. The VGF sites within the inverted terminal repeats of VSC20 are interrupted by the insertion of the lacZ gene. - - - -, sites of predicted homologous recombination. P1–P4, approximate location of primers used in B. B, PCR analysis of viral DNA from vvDD-GFP confirming homologous recombination. Lane L, 100-bp ladder. Lane 1 (vvDD-GFP), P1/P2 primers that amplify a 680-bp fragment across the region of recombination, which demonstrate absence of TK. The TK positive (Lane 5, vSC20) and negative (Lane 7, H2O control) controls are shown. Lane 2 (vvDD-GFP), P1/P3 amplify EGFP across the 5′ site of recombination, giving a 1220-bp band. Negative control is shown (Lane 8, H2O control). Lane 3, P2/P4 amplify gpt across the 3′ site of recombination, giving an 826-bp band. Negative control is shown (Lane 9, H2O control). Lane 4 (vvDD-GFP), VGF primers, which amplify a 382-bp fragment, demonstrate absence of VGF in vvDD-GFP but a positive band for VGF in the control virus (Lane 6, F13L+). Negative control is shown in Lane 10 (H2O control).

Fig. 1.

A, linear diagram of the pSEL-EGFP shuttle plasmid and parental vaccinia VSC20 used to create vvDD-GFP. Sites of left (vTK-L) and right (vTK-R) segments of vaccinia TK gene, synthetic early/late promoter (Pse/I), 7.5 promoter (p7.5), ampicillin resistance gene, and xanthine-guanine phosphoribosyl transferase (gpt) gene (a gene for viral selection, not used in this study) are indicated. The VGF sites within the inverted terminal repeats of VSC20 are interrupted by the insertion of the lacZ gene. - - - -, sites of predicted homologous recombination. P1–P4, approximate location of primers used in B. B, PCR analysis of viral DNA from vvDD-GFP confirming homologous recombination. Lane L, 100-bp ladder. Lane 1 (vvDD-GFP), P1/P2 primers that amplify a 680-bp fragment across the region of recombination, which demonstrate absence of TK. The TK positive (Lane 5, vSC20) and negative (Lane 7, H2O control) controls are shown. Lane 2 (vvDD-GFP), P1/P3 amplify EGFP across the 5′ site of recombination, giving a 1220-bp band. Negative control is shown (Lane 8, H2O control). Lane 3, P2/P4 amplify gpt across the 3′ site of recombination, giving an 826-bp band. Negative control is shown (Lane 9, H2O control). Lane 4 (vvDD-GFP), VGF primers, which amplify a 382-bp fragment, demonstrate absence of VGF in vvDD-GFP but a positive band for VGF in the control virus (Lane 6, F13L+). Negative control is shown in Lane 10 (H2O control).

Close modal
Fig. 2.

Viral recovery 24, 48, and 72 h after in vitro infection of NIH3T3 cells with WT (F13L+; • ), TK− (VJS6; ▴), VGF− (VSC20; ○ ), and vvDD-GFP (▪) vaccinia viruses. Equivalent recovery of vvDD-GFP was seen compared with the other viruses when dividing cells were used (A); however, decreased virus was recovered after infection of resting (confluent) cells (B).

Fig. 2.

Viral recovery 24, 48, and 72 h after in vitro infection of NIH3T3 cells with WT (F13L+; • ), TK− (VJS6; ▴), VGF− (VSC20; ○ ), and vvDD-GFP (▪) vaccinia viruses. Equivalent recovery of vvDD-GFP was seen compared with the other viruses when dividing cells were used (A); however, decreased virus was recovered after infection of resting (confluent) cells (B).

Close modal
Fig. 3.

Immunohistochemistry with a polyclonal vaccinia antibody of nude mouse brain (left column), ovary (middle column), and tumor (right column), 5 days after i.p. injection of HBSS control (a–c), or 108 pfu of WT (F13L+, d–f), TK− (VJS6, g–i), VGF− (VSC20, j–l), or vvDD-GFP (m–o) vaccinia viruses. Note focal meningeal immunoreactivity after WT (d) or VGF− (j) infection (left column), limited immunoreactivity of ovarian follicles and adjacent tissues after vvDD-GFP infection (n), with more diffuse immunoreactivity after infection with the other viruses (middle column), and diffuse tumor reactivity after infection with each of the viruses (right column). Centers and edges of tumors are necrotic (composed of dead tumor cells) and less immunoreactive. Brain, ×75; ovary, ×30; tumor, ×15.

Fig. 3.

Immunohistochemistry with a polyclonal vaccinia antibody of nude mouse brain (left column), ovary (middle column), and tumor (right column), 5 days after i.p. injection of HBSS control (a–c), or 108 pfu of WT (F13L+, d–f), TK− (VJS6, g–i), VGF− (VSC20, j–l), or vvDD-GFP (m–o) vaccinia viruses. Note focal meningeal immunoreactivity after WT (d) or VGF− (j) infection (left column), limited immunoreactivity of ovarian follicles and adjacent tissues after vvDD-GFP infection (n), with more diffuse immunoreactivity after infection with the other viruses (middle column), and diffuse tumor reactivity after infection with each of the viruses (right column). Centers and edges of tumors are necrotic (composed of dead tumor cells) and less immunoreactive. Brain, ×75; ovary, ×30; tumor, ×15.

Close modal
Fig. 4.

A, survival of nude mice after treatment with 108 pfu of replicating WT (F13L+, solid line), TK− (VJS6, dotted line), VGF− (VSC20, medium dash), or vvDD-GFP (long dash) by i.p. injection (n = 10). Both the TK− and VGF− infected mice had significantly prolonged survival (P < 0.0001) compared with WT. The double-deleted vaccinia virus (vvDD-GFP) had no evidence of toxicity by day 100 (P < 0.0001 compared with WT). B, survival of nude mice after treatment with 108 pfu (solid line) or 109 pfu (dashed line) of vvDD-GFP compared with 108 pfu of the TK-(VJS6, dotted line). Both groups of mice infected with vvDD-GFP had significantly prolonged survival (P < 0.001) compared with the TK− infected mice. C, survival of C57BL/6 (immunocompetent) mice after treatment with 109 pfu of replicating WT (F13L+, solid line), TK− (VJS6, dotted line), VGF− (VSC20, medium dash), or vvDD-GFP (long dash) by i.p. injection (n = 10). All mice died after treatment with WT vaccinia; however, less toxicity was seen in the groups receiving the TK− or VGF− virus compared with nude mice. No toxicity was seen in the mice receiving vvDD-GFP.

Fig. 4.

A, survival of nude mice after treatment with 108 pfu of replicating WT (F13L+, solid line), TK− (VJS6, dotted line), VGF− (VSC20, medium dash), or vvDD-GFP (long dash) by i.p. injection (n = 10). Both the TK− and VGF− infected mice had significantly prolonged survival (P < 0.0001) compared with WT. The double-deleted vaccinia virus (vvDD-GFP) had no evidence of toxicity by day 100 (P < 0.0001 compared with WT). B, survival of nude mice after treatment with 108 pfu (solid line) or 109 pfu (dashed line) of vvDD-GFP compared with 108 pfu of the TK-(VJS6, dotted line). Both groups of mice infected with vvDD-GFP had significantly prolonged survival (P < 0.001) compared with the TK− infected mice. C, survival of C57BL/6 (immunocompetent) mice after treatment with 109 pfu of replicating WT (F13L+, solid line), TK− (VJS6, dotted line), VGF− (VSC20, medium dash), or vvDD-GFP (long dash) by i.p. injection (n = 10). All mice died after treatment with WT vaccinia; however, less toxicity was seen in the groups receiving the TK− or VGF− virus compared with nude mice. No toxicity was seen in the mice receiving vvDD-GFP.

Close modal
Fig. 5.

Mean tumor volume after i.p. treatment of s.c. MC38 in nude mice; bars, SE. On day 0, 8 days after inoculation with tumor, mice were injected with 108 pfu of replicating vvDD-GFP (•, n = 10) or HBSS control (▪, n = 10). Control HBSS had no effect on tumors.

Fig. 5.

Mean tumor volume after i.p. treatment of s.c. MC38 in nude mice; bars, SE. On day 0, 8 days after inoculation with tumor, mice were injected with 108 pfu of replicating vvDD-GFP (•, n = 10) or HBSS control (▪, n = 10). Control HBSS had no effect on tumors.

Close modal

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

3

The abbreviations used are: IL, interleukin; TK, thymidine kinase; VGF, vaccinia growth factor; WT, wild type; TTP, thymidine triphosphate; EGFP, enhanced green fluorescent protein; pfu, plaque forming unit.

4

E. Chang, J. Friedl, H. Xu, G. Phan, S. Mocellin, F. Marincola, R. Alexander, and D. Bartlett. Targeting vaccinia to solid tumors with local hyperthermia, submitted for publication.

Table 1

Median viral recovery from nude mouse tissues

Median (range) viral pfu/mg total protein of tissues 8 days after infection with WT (F13L+), TK− (VJS6), VGF− (VSC20), or vvDD-GFP. Nude mice (n = 3–5) were infected with 107 pfu of virus. On day 8 after infection, tissues were harvested and homogenized, and a standard plaque assay was performed.
 WT TK− VGF− vvDD-GFP 
Brain 2.8 (2.4–4.9) × 104a 1.3 (.21–20) × 102 1.5 (.76–4.3) × 102 0 (0–8)a 
Liver 3 (.8–11) 7 (.6–13) 1 (.24–1.1) 0.1 (0–.2) 
Spleen 5.1 (.59–21) × 102 12 (6–16) 23 (16–308) 8 (0–16) 
Testesb 54 (0.4–2800) 12 (0.13–24) × 102 0.6 (0.4–0.8) 6.8 (0.7–28) 
Bone marrow 1.0 (.08–10) × 104 3.0 (.075–7.6) × 103 1.1 (.41–2100) × 103 5.0 (0–12) × 102 
Ovary 7.1 (2.6–9.7) × 106 9.3 (2.3–15) × 106 2.1 (.41–3.9) × 107 8.6 (.6–172) × 106 
Tumor 17.0 (1.2–14) × 106 4.6 (.3–6.6) × 106 2.3 (.05–2.6) × 107 6.5 (.4–6.5) × 106 
Median (range) viral pfu/mg total protein of tissues 8 days after infection with WT (F13L+), TK− (VJS6), VGF− (VSC20), or vvDD-GFP. Nude mice (n = 3–5) were infected with 107 pfu of virus. On day 8 after infection, tissues were harvested and homogenized, and a standard plaque assay was performed.
 WT TK− VGF− vvDD-GFP 
Brain 2.8 (2.4–4.9) × 104a 1.3 (.21–20) × 102 1.5 (.76–4.3) × 102 0 (0–8)a 
Liver 3 (.8–11) 7 (.6–13) 1 (.24–1.1) 0.1 (0–.2) 
Spleen 5.1 (.59–21) × 102 12 (6–16) 23 (16–308) 8 (0–16) 
Testesb 54 (0.4–2800) 12 (0.13–24) × 102 0.6 (0.4–0.8) 6.8 (0.7–28) 
Bone marrow 1.0 (.08–10) × 104 3.0 (.075–7.6) × 103 1.1 (.41–2100) × 103 5.0 (0–12) × 102 
Ovary 7.1 (2.6–9.7) × 106 9.3 (2.3–15) × 106 2.1 (.41–3.9) × 107 8.6 (.6–172) × 106 
Tumor 17.0 (1.2–14) × 106 4.6 (.3–6.6) × 106 2.3 (.05–2.6) × 107 6.5 (.4–6.5) × 106 
a

P = 0.011.

b

Testes samples obtained in a separate experiment.

Table 2

Nude mouse pathology after viral infection

Nude mice (n = 2) received injections of HBSS or 108 pfu of WT (F13L+), TK− (VJS6), VGF− (VSC20), or vvDD-GFP. Five days after injection, organs were harvested, placed in 10% formalin, and embedded in paraffin. Histological exam was performed in a blinded fashion after staining with H&E and graded according to the above key. Ovaries showed minimal to severe necrosis and/or fibrinous inflammation and hemorrhage. Brain lesions represent focal meningeal vasculitis. Liver necrosis was focal.
Saline/Virus Tumor necrosisa Ovarian necrosisa Brain lesionsa Liver necrosisa 
HBSS − − − − 
vvDD-GFP +++ − − 
VGF− +++ +/− − 
TK− +++ ++ +/− 
WT +++ +++ 
Nude mice (n = 2) received injections of HBSS or 108 pfu of WT (F13L+), TK− (VJS6), VGF− (VSC20), or vvDD-GFP. Five days after injection, organs were harvested, placed in 10% formalin, and embedded in paraffin. Histological exam was performed in a blinded fashion after staining with H&E and graded according to the above key. Ovaries showed minimal to severe necrosis and/or fibrinous inflammation and hemorrhage. Brain lesions represent focal meningeal vasculitis. Liver necrosis was focal.
Saline/Virus Tumor necrosisa Ovarian necrosisa Brain lesionsa Liver necrosisa 
HBSS − − − − 
vvDD-GFP +++ − − 
VGF− +++ +/− − 
TK− +++ ++ +/− 
WT +++ +++ 
a

Degree of the lesion: −, not present; +/−, minimal; +, mild; ++, moderate; +++, severe.

We thank S. A. Rosenberg and C. L. Nutt for critical reading of the manuscript and the NIH Medical Arts and Photography Branch for help with the photography.

1
Siders W. M., Halloran P. J., Fenton R. G. Transcriptional targeting of recombinant adenoviruses to human and murine melanoma cells.
Cancer Res.
,
56
:
5638
-5646,  
1996
.
2
Parr M. J., Manome Y., Tanaka T., Wen P., Kufe D. W., Kaelin W. G., Jr., Fine H. A. Tumor-selective transgene expression in vivo mediated by an E2F-responsive adenoviral vector.
Nat. Med.
,
3
:
1145
-1149,  
1997
.
3
Dmitriev I., Krasnykh V., Miller C. R., Wang M., Kashentseva E., Mikheeva G., Belousova N., Curiel D. T. An adenovirus vector with genetically modified fibers demonstrates expanded tropism via utilization of a coxsackievirus and adenovirus receptor-independent cell entry mechanism.
J. Virol.
,
72
:
9706
-9713,  
1998
.
4
Park B. J., Brown C. K., Hu Y., Alexander H. R., Horti J., Raje S., Figg W. D., Bartlett D. L. Augmentation of melanoma-specific gene expression using a tandem melanocyte-specific enhancer results in increased cytotoxicity of the purine nucleoside phosphorylase gene in melanoma.
Hum. Gene Ther.
,
10
:
889
-898,  
1999
.
5
Advani S. J., Chung S. M., Yan S. Y., Gillespie G. Y., Markert J. M., Whitley R. J., Roizman B., Weichselbaum R. R. Replication-competent, nonneuroinvasive genetically engineered herpes virus is highly effective in the treatment of therapy-resistant experimental human tumors.
Cancer Res.
,
59
:
2055
-2058,  
1999
.
6
Kirn D. H., McCormick F. Replicating viruses as selective cancer therapeutics.
Mol. Med. Today
,
2
:
519
-527,  
1996
.
7
McCart J. A., Puhlmann M., Lee J., Hu Y., Libutti S. K., Alexander H. R., Bartlett D. L. Complex interactions between the replicating oncolytic effect and the enzyme/prodrug effect of vaccinia-mediated tumor regression.
Gene Ther.
,
7
:
1217
-1223,  
2000
.
8
Wildner O., Blaese R. M., Morris J. C. Therapy of colon cancer with oncolytic adenovirus is enhanced by the addition of herpes simplex virus-thymidine kinase.
Cancer Res.
,
59
:
410
-413,  
1999
.
9
Chase M., Chung R. Y., Chiocca E. A. An oncolytic viral mutant that delivers the CYP2B1 transgene and augments cyclophosphamide chemotherapy.
Nat. Biotech.
,
16
:
444
-448,  
1998
.
10
Walker J. R., McGeach K. G., Sundaresan P., Jorgensen T. J., Rabkin S. D., Martuza R. L. Local and systemic therapy of human prostate adenocarcinoma with the conditionally replicating herpes simplex virus vector G207.
Hum. Gene Ther.
,
10
:
2237
-2243,  
1999
.
11
Gnant M. F. X., Puhlmann M., Alexander H. R., Jr., Bartlett D. L. Systemic administration of a recombinant vaccinia virus expressing the cytosine deaminase gene and subsequent treatment with 5-fluorocytosine leads to tumor specific gene expression and prolongation of survival in mice.
Cancer Res.
,
59
:
3396
-3404,  
1999
.
12
Puhlmann M., Gnant M., Brown C. K., Alexander H. R., Bartlett D. L. Thymidine kinase deleted vaccinia virus expressing purine nucleoside phosphorylase as a vector for tumor directed gene therapy.
Hum. Gene Ther.
,
10
:
649
-657,  
1999
.
13
Heise C., Sampson-Johannes A., Williams A., McCormick F., Von Hoff D. D., Kirn D. H. ONYX-015, an E1B gene-attenuated adenovirus, causes tumor-specific cytolysis and antitumoral efficacy that can be augmented by standard chemotherapeutic agents.
Nat. Med.
,
3
:
639
-645,  
1997
.
14
Martuza R. L., Malick A., Markert J. M., Ruffner K. L., Coen D. M. Experimental therapy of human glioma by means of a genetically engineered virus mutant.
Science (Wash. DC)
,
252
:
854
-856,  
1991
.
15
Galmiche M. C., Rindisbacher L., Wels W., Wittek R., Buchegger F. Expression of a functional single chain antibody on the surface of extracellular enveloped vaccinia virus as a step towards selective tumour cell targeting.
J. Gen. Virol.
,
78
:
3019
-3027,  
1997
.
16
Moss B. Genetically engineered poxviruses for recombinant gene expression, vaccination, and safety.
Proc. Natl. Acad. Sci. USA
,
93
:
11341
-11348,  
1996
.
17
McAneny D., Ryan C. A., Beazley R. M., Kaufman H. L. Results of a Phase I trial of a recombinant vaccinia virus that expresses carcinoembryonic antigen in patients with advanced colorectal cancer.
Ann. Surg. Oncol.
,
3
:
495
-500,  
1996
.
18
Kantor J., Irvine K., Abrams S., Snoy P., Olsen R., Greiner J., Kaufman H., Eggensperger D., Schlom J. Immunogenicity and safety of a recombinant vaccinia virus vaccine expressing the carcinoembryonic antigen gene in a nonhuman primate.
Cancer Res.
,
52
:
6917
-6925,  
1992
.
19
Overwijk W., Lee D. S., Surman D. R., Irvine K. R., Touloukian C. E., Chan C-C., Carroll M. W., Moss B., Rosenberg S. A., Restifo N. P. Vaccination with a recombinant vaccinia virus encoding a “self” antigen induces autoimmune vitiligo and tumor cell destruction in mice: requirement for CD4+ T lymphocytes.
Proc. Natl. Acad. Sci. USA
,
96
:
2982
-2987,  
1999
.
20
Qin H., Chatterjee S. K. Cancer gene therapy using tumor cells infected with recombinant vaccinia virus expressing GM-CSF.
Hum. Gene Ther.
,
7
:
1853
-1860,  
1996
.
21
Hodge J. W., Abrams S., Schlom K. J., Kantor J. A. Induction of antitumor immunity by recombinant vaccinia viruses expressing B7-1 or B7-2 costimulatory molecules.
Cancer Res.
,
54
:
5552
-5555,  
1994
.
22
Peplinski G. R., Tsung K., Meko J. B., Norton J. A. In vivo gene therapy of a murine pancreas tumor with recombinant vaccinia virus encoding human interleukin-1β.
Surgery
,
118
:
185
-191,  
1995
.
23
Meko J. B., Yim J. H., Tsung K., Norton J. A. High cytokine production and effective antitumor activity of a recombinant vaccinia virus encoding murine interleukin 12.
Cancer Res.
,
55
:
4765
-4770,  
1995
.
24
Flexner C., Moss B., London W. T., Murphy B. R. Attenuation and immunogenicity in primates of vaccinia virus recombinants expressing human interleukin-2.
Vaccine
,
8
:
17
-22,  
1990
.
25
Ramshaw A., Andrew M. E., Phillips S. M., Boyle D. B., Coupar B. E. H. Recovery of immunodeficient mice from a vaccinia virus/IL-2 recombinant infection.
Nature (Lond.)
,
329
:
545
-546,  
1987
.
26
Kim C. J., Cormier J., Roden M., Gritz L., Mazzara G. P., Fetsch P., Hijazi Y., Lee K. H., Rosenberg S. A., Marincola F. M. Use of recombinant poxviruses to stimulate anti-melanoma T cell reactivity.
Ann. Surg. Oncol.
,
5
:
64
-76,  
1998
.
27
Lane J. M., Millar J. D. Risks of smallpox vaccination complications in the United States.
Am. J. Epidemiol.
,
93
:
238
-240,  
1971
.
28
Robinson M. J., Murugasu R., Thong Y. H., Chen S. T. A fatal case of progressive vaccinia—clinical and pathological studies.
Aust. Paed. J.
,
13
:
125
-130,  
1977
.
29
Turkel S. B., Overturf G. D. Vaccinia necrosum complicating immunoblastic sarcoma.
Cancer (Phila.)
,
40
:
226
-233,  
1977
.
30
Keane J. T., James K., Blankenship M. L., Pearson R. W. Progressive vaccinia associated with combined variable immunodeficiency.
Arch. Dermatol.
,
119
:
404
-408,  
1983
.
31
Gurvich E. B., Vilesova I. S. Vaccinia virus in postvaccinal encephalitis.
Acta Virol.
,
27
:
154
-159,  
1983
.
32
Buller R. M., Smith G. L., Cremer K., Notkins A. L., Moss B. Decreased virulence of recombinant vaccinia virus expression vectors is associated with a thymidine kinase-negative phenotype.
Nature (Lond.)
,
317
:
813
-815,  
1985
.
33
Buller R. M. L., Chakrabarti S., Cooper J. A., Twardzik D. R., Moss B. Deletion of the vaccinia virus growth factor gene reduces virus virulence.
J. Virol.
,
62
:
866
-874,  
1988
.
34
Peplinski G. R., Tsung K., Norton J. A. Vaccinia virus for human gene therapy.
Surg. Oncol. Clin. N. Am.
,
7
:
575
-588,  
1998
.
35
Puhlmann M., Brown C. K., Gnant M., Huang J., Libutti S. K., Alexander H. R., Bartlett D. L. Vaccinia as a vector for tumor directed gene therapy: biodistribution of a thymidine kinase deleted mutant.
Cancer Gene Ther.
,
7
:
66
-73,  
2000
.
36
Gnant M., Puhlmann M., Bartlett D. L., Alexander H. R. regional verses systemic delivery of recombinant vaccinia virus as suicide gene therapy for murine liver metastases.
Ann. Surg.
,
230
:
352
-361,  
1999
.
37
Buller R. M. L., Chakrabarti S., Moss B., Frederickson T. Cell proliferative response to vaccinia virus is mediated by VGF.
Virology
,
164
:
182
-192,  
1988
.
38
Roper R. L., Moss B. Envelope formation is blocked by mutation of a sequence related to the HKD phospholipid metabolism motif in the vaccinia virus F13L protein.
J. Virol.
,
73
:
1108
-1117,  
1999
.
39
Chakrabarti S., Sisler J. R., Moss B. Compact, synthetic, vaccinia virus early/late promoter for protein expression.
BioTechniques
,
23
:
1094
-1097,  
1997
.
40
Higuchi R. simple and rapid preparation of samples for PCR Erlich H. A. eds. .
PCR Technology: Principals and Applications for DNA Amplifications
,
:
31
-38, W. H. Freeman and Company New York  
1992
.
41
Earl P. L., Moss B. expression of proteins in mammalian cells using vaccinia viral vectors Ausubel F. M. Kinston R. Kingston R. E. Moore D. D. Seidman J. G. Smith J. A. Struhl K. eds. .
Current Protocols in Molecular Biology
,
:
16.15.1
-16.18.11, Greene/Wiley Interscience New York  
1998
.
42
O’Reilly M. S., Boehm T., Shing Y., Fukai N., Vasios G., Lane W. S., Flynn E., Birkhead J. R., Olsen B. R., Folkman J. Endostatin. An endogenous inhibitor of angiogenesis and tumor growth.
Cell
,
88
:
277
-285,  
1997
.
43
Kaplan E. L., Meier P. Non-parametric estimation from incomplete observation.
J. Am. Stat. Assoc.
,
53
:
457
-481,  
1958
.
44
Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration.
Cancer Chem. Rep.
,
50
:
163
-170,  
1966
.
45
Carroll M. W., Moss B. Poxviruses as expression vectors.
Curr. Opin. Biotechnol.
,
8
:
573
-577,  
1997
.
46
Smith G. L., Moss B. Infectious poxvirus vectors have capacity for at least 25,000 base pairs of foreign DNA.
Gene (Amst.)
,
25
:
21
-28,  
1983
.
47
Davison A. J., Moss B. Structure of vaccinia virus early promoters.
J. Mol. Biol.
,
210
:
749
-769,  
1989
.
48
Davison A. J., Moss B. Structure of vaccinia virus late promoters.
J. Mol. Biol.
,
210
:
771
-784,  
1989
.
49
Kohn S., Nagy J. A., Dvorak H. F., Dvorak A. M. Pathways of macromolecular tracer transport across venules and small veins. Structural basis for the hyperpermeability of tumor blood vessels.
Lab. Investig.
,
67
:
596
-607,  
1992
.
50
Ricketts T. F. .
Diagnosis of Smallpox
,
6
-13, Cassell and Company, Ltd. London  
1966
.
51
Goede V., Schmidt T., Kimmina S., Kozian D., Augustin H. G. Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis.
Lab. Investig.
,
78
:
1385
-1394,  
1998
.
52
Reynolds L. P., Grazul-Bilska A. T., Redmer D. A. Angiogenesis in the corpus luteum.
Endocrine
,
12
:
1
-9,  
2000
.
53
Palumbo G. J., Higginbotham J. N., Toland B., Ramsey J., Blaese R. M. A replication competent recombinant vaccinia vector expressing HSV-TK for the treatment of tumors in vivo.
Proc. Am. Soc. Gene Ther.
,
1
:
169a
1998
.
54
Fenner F., Wittek R., Dumbell K. R. The pathogenesis, pathology, and immunology of orthopoxvirus infections.
The Orthopoxviruses
,
85
-141, Academic Press, Inc. New York  
1989
.
55
Karupiah G., Coupar B., Ramshaw I., Boyle D., Blanden R., Andrew M. Vaccinia virus-mediated damage of murine ovaries and protection by virus-expressed interleukin-2.
Immunol. Cell Biol.
,
68
:
325
-333,  
1990
.
56
Mastrangelo M. J., Maguire H. C., Jr., Eisenlohr L. C., Laughlin C. E., Monken C. E., McCue P. A., Kovatich A. J., Lattime E. C. Intratumoral recombinant GM-CSF-encoding virus as gene therapy in patients with cutaneous melanoma.
Cancer Gene Ther.
,
6
:
409
-422,  
1998
.
57
Mukherjee S., Haenel T., Himbeck R., Scott B., Ramshaw I., Lake R. A., Harnett G., Phillips P., Morey S., Smith D., Davidson J. A., Musk A. W., Robinson B. Replication-restricted vaccinia as a cytokine gene therapy vector in cancer: persistent transgene expression despite antibody generation.
Cancer Gene Ther.
,
7
:
663
-670,  
2000
.
58
Mastrangelo M. J., Maguire H. C., Jr., Lattime E. C. Intralesional vaccinia/GM-CSF recombinant virus in the treatment of metastatic melanoma.
Adv. Exp. Med. Biol.
,
465
:
391
-400,  
2000
.
59
Beer S. J., Matthews C. B., Stein C. S., Ross B. D., Hilfinger J. M., Davidson B. L. Poly (lactic-glycolic) acid co-polymer encapsulation of recombinant adenovirus reduces immunogenicity in vivo.
Gene Ther.
,
5
:
740
-746,  
1998
.
60
Chillon M., Lee J. H., Fasbender A., Welsh M. J. Adenovirus complexed with polyethylene glycol and cationic lipid is shielded from neutralizing antibodies in vitro.
Gene Ther.
,
5
:
995
-1002,  
1998
.
61
Rosenberg S. A., Blaese R. M., Brenner M. K., Deisseroth A. B., Ledley F. D., Lotze M. T., Wilson J. M., Nabel G. J., Cornetta K., Economou J. S., Freeman S. M., Riddell S. R., Oldfield E., Gansbacher B., Dunbar C., Walker R. E., Schuening F. G., Roth J. A., Crystal R. G., Welsh M. J., Culver K., Heslop H. E., Simons J., Wilmott R. W., Habib N. A. Human gene marker/therapy clinical protocols.
Hum. Gene Ther.
,
10
:
3067
-3123,  
1999
.