Adenovirus vectors have a number of advantages for gene therapy. However, because of their lack of tumor tropism and their preference for liver infection following systemic administration, they cannot be used for systemic attack on metastatic disease. Many epithelial tumors (e.g., colon, lung, and breast) express carcinoembryonic antigen (CEA). To block the natural hepatic tropism of adenovirus and to “retarget” the virus to CEA-expressing tumors, we used a bispecific adapter protein (sCAR-MFE), which fuses the ectodomain of the coxsackie/adenovirus receptor (sCAR) with a single-chain anti-CEA antibody (MFE-23). sCAR-MFE untargets adenovirus-directed luciferase transgene expression in the liver by >90% following systemic vector administration. Moreover, sCAR-MFE can “retarget” adenovirus to CEA-positive epithelial tumor cells in cell culture, in s.c. tumor grafts, and in hepatic tumor grafts. The sCAR-MFE bispecific adapter should, therefore, be a powerful agent to retarget adenovirus vectors to epithelial tumor metastases. [Cancer Res 2007;67(11):5354–61]

Gene therapy presents a novel therapeutic approach for the treatment of cancer, in which corrective or toxin genes are delivered to neoplastic cells. Many gene therapy strategies, however, are limited due to the lack of “effective gene delivery” (efficient and specific delivery of genes to the target cell). Both non-viral and viral vectors for gene therapy have been proposed. Several viral species have been adapted for gene delivery purposes (e.g., retrovirus, herpes simplex virus, and adenovirus). Adenovirus vectors have several characteristics that make them attractive gene delivery vehicles: high levels of transgene expression, large DNA insertion capacity, high stability in vivo, and low pathogenicity in humans. However, therapeutic application of adenovirus vectors is currently limited to i.t. delivery for local/regional neoplastic disease; current generation adenovirus gene delivery vectors cannot achieve selective transgene delivery to tumor cells in vivo.

Adenovirus binding to target cell surfaces is mediated by interaction between virus fiber protein “knob” domains and cellular coxsackie/adenovirus receptors (CAR; refs. 13). Hepatocytes express high CAR levels; consequently, i.v. adenovirus vector administration results in extensive liver infection, which can lead to liver toxicity (46). In contrast, most human tumor cells produce relatively low CAR levels. Moreover, tumor cells with low CAR levels are often more invasive and have greater propensity to metastasize (7, 8). The differential expression of CAR between cancer cells and normal cells makes it difficult to target tumors with adenovirus vectors. Thus, both hepatic CAR-dependent binding resulting in liver infection and reduced tumor cell infection by adenovirus are serious limitations to systemic use of adenovirus vectors for therapy of metastatic disease. Application of cancer gene therapy approaches for disseminated disease will require adenovirus modifications to allow selective transgene delivery, based upon tumor targeting, in combination with virus untargeting for liver.

One method to selectively target adenovirus vector transgene expression to tumors is “transductional untargeting and retargeting,” in which the virus particle is physically prevented from binding to CAR on normal cells and simultaneously redirected to receptors expressed preferentially or exclusively on tumor cells. To achieve transductional targeting, bispecific adapter molecules have been developed. Bispecific adapters bind with one domain to the virus, usually the “knob” of the viral fiber protein, thereby blocking native tropism, and bind with their other domain to target receptors, thus bridging the adenovirus vector with the target cell. Bispecific adapters also mediate infectivity via CAR-independent cellular pathways. Both sCAR, the soluble CAR ectodomain, and antibodies to the adenovirus fiber knob have been used as “untargeting” components of bispecific adapters (913). Ligands for cell surface receptors [e.g., fibroblast growth factor-2, epidermal growth factor (EGF), folate] and antibodies to cell surface antigens are used as “retargeting” components of bispecific adapters (911, 14). Transductional targeting of adenovirus vectors to cell receptors/antigens with bispecific adapters facilitates increased adenovirus infectivity, both in vitro (9, 11, 1417) and, most importantly, in vivo (15, 18, 19).

Carcinoembryonic antigen (CEA), a cell surface antigen, is expressed in nearly all colorectal cancer tumors, in 70% of non–small-cell lung cancers, and in 50% of breast cancers. In contrast, CEA expression is substantially restricted in normal adult tissues; there is little or no detectable CEA expression in the liver. Bispecific adapters with a CEA-binding domain should untarget adenovirus vectors from normal tissues and retarget them to CEA-positive tumors. MFE-23, a single-chain antibody with high affinity for CEA, is currently in phase I studies both as an imaging agent for radioimmunoguided surgery and as a tumor-targeting agent for antibody-directed enzyme prodrug therapy (2022). To retarget CEA-positive cells, MFE-23 has been fused to sCAR to form sCAR-MFE, a bispecific transductional reagent. sCAR-MFE successfully targets adenovirus to cells that express CEA (19). sCAR-MFE should therefore be useful in retargeting adenovirus vectors to CEA-positive hepatic metastases. In this report, we show that sCAR-MFE can dramatically reduce hepatic infection following systemic administration and can redirect adenoviral gene therapy vectors to CEA-positive epithelial tumor cells in cell culture, in s.c. tumor grafts, and in hepatic tumor grafts.

Cell culture. AML12 cells were from the American Type Culture Collection. MC38 and MC38-cea-2 cells (23) were provided by Dr. Jeffrey Schlom (NIH). A427 and H2122 cells were provided by Dr. Steven Dubinett (University of California at Los Angeles). LS174T cells were provided by Dr. Anna Wu (University of California at Los Angeles). AML12, LS174T, MC38, and MC38-cea-2 were grown in DMEM (Life Technologies) containing 10% fetal bovine serum (FBS; Cellegro) with penicillin-streptomycin (Life Technologies), 0.1 mmol/L nonessential amino acids (Life Technologies), and 1.0 mmol/L sodium pyruvate (Life Technologies). A427 and H2122 were grown in RPMI 1640 (Life Technologies) containing 10% FBS with penicillin-streptomycin.

Virus construction and production. The adenovirus vector encoding firefly luciferase (fLuc) under transcriptional control of the constitutively active cytomegalovirus (CMV) promoter Ad.CMVfLuc was constructed as described by Liang et al. (24). Ad.CMVgfp, an adenovirus in which the green fluorescent protein (GFP) is expressed from the CMV promoter, was a gift from Dr. Sanjiv Gambhir (Stanford University). Virus was prepared in 293 cells by double cesium chloride (CsCl) gradient centrifugation. Cells were infected in medium containing 2% FBS. After overnight incubation, cells were shifted to medium containing 10% serum and incubated until a total cytopathic effect was observed. Cells were harvested, frozen, and thawed thrice, and virus was purified using standard CsCl purification methods. Viral particle number was determined by measuring absorbance at 260 nm, using a conversion factor of 1.1 × 1012 viral particles (vp) per absorbance unit. Viral titers were determined with the Adeno-X Rapid Titer kit (BD Clontech). “vp” is used to indicate virus particles when considering virus mass interactions with bispecific adapter; “pfu” (for plaque forming units) is used when viral infectivity is described.

Stable transfection. To establish stable cell lines that over express Renilla luciferase (rLuc) or red fluorescence protein (RFP), LS174T cells were transfected with pcDNA3 expression vectors encoding rLuc or RFP and a neomycin selectable marker, using LipofectAMINE 2000 (Invitrogen). Neomycin-resistant cells were selected using medium containing G418 (1 mg/mL). G418-resistant cells were screened for rLuc expression with the Luciferase assay system (Promega) or for RFP by fluorescence microscopy.

Immunoblot analysis. Cells were washed twice with PBS and incubated on ice for 10 min in lysis buffer containing 50 mmol/L Tris-HCl (pH 7.4), 1% NP40, 1% Triton X-100, 1% sodium deoxycholate, 150 mmol/L NaCl, 1 mmol/L EDTA, and protease inhibitors (Complete Tablet; Roche). After 10,000 × g centrifugation for 10 min, supernatant protein concentrations were measured with the Bio-Rad Protein Assay (Bio-Rad). To examine CEA protein expression, protein extracts were denatured in loading buffer. To examine the expression of CAR protein, the protein extracts were heated at 37°C for 10 min in non-reducing conditions before loading. Equal amounts of protein (30 μg) were loaded on SDS-polyacrylamide gels (8% for CEA and 12% for CAR) for electrophoresis. Proteins were subsequently transferred to nitrocellulose membranes. The membranes were probed with anti-CEA antibody, cT84.66 (1:10,000 dilution; ref. 25), anti-CAR antibody (RmcB; 1:2,000 dilution; Upstate Biologicals), or anti-14-3-3𝛉 antibody (1:3,000 dilution; Santa Cruz Biotechnology) overnight at 4°C followed by incubation at room temperature for 60 min with horseradish peroxidase–conjugated secondary antibody. Immunoreactivity was determined by enhanced chemiluminescence (Amersham).

Immunofluorescence staining. Cells were seeded in four-well chamber slides (1 × 105 per chamber) and cultured overnight. Anti-CEA primary antibody cT84.66 (1:1,000) was added for 60 min at room temperature. Antibody was aspirated, and cells were fixed (37% formaldehyde diluted 1:10 with PBS) for 20 min. After aspirating, cells were washed thrice with washing buffer [1 mmol/L CaCl2, 140 mmol/L NaCl, 3 mmol/L KCl, and 25 mmol/L Tris-Cl (pH 7.4)] for 10 min. Cells were incubated for 20 min with blocking buffer [3% dry milk, 0.1% Triton X-100, 1 mmol/L CaCl2, and 50 mmol/L Tris-Cl (pH 7.5)]. After aspirating, cells were incubated with FITC-conjugated goat anti-human IgG secondary antibody (1:500 dilution in blocking buffer; Jackson ImmunoResearch) for 30 min in the dark. After incubation, cells were washed thrice for 10 min with washing buffer. Fluorescence was observed with a Zeiss AIOSKOP fluorescence microscope.

Adenovirus untargeting and retargeting on cultured cell lines. sCAR-MFE was constructed and produced as described by Everts et al. (19). Ad.CMVfLuc or Ad.CMVgfp (3 × 108 vp) were mixed with sCAR-MFE (1.3 mg/mL) in concentrations ranging from 0 to 500 ng in 0.5 μL and incubated for 60 min at room temperature. The sCAR-MFE:Ad complexes were diluted to 150 μL with medium containing 2% FBS and then added to cell monolayers in 24-well plates (2 × 105 per well). Virus-treated cells were incubated at 37°C for 90 min. Medium was aspirated, and cells were washed with PBS. After a 40-h incubation at 37°C in medium containing 10% FBS, cells were lysed, and luciferase activity was measured. Ad.CMVgfp-infected cells were observed using fluorescence microscopy. For competition experiments, recombinant CEA (rCEA) protein (Protein Sciences Corp.) and albumin (Sigma) were added to the cultures immediately before addition of [Ad.CMVfLuc][sCAR-MFE].

Adenovirus untargeting and retargeting on mixed cell cultures. Ad.CMVfLuc (1.2 × 109 vp) was mixed with sCAR-MFE (500 ng) or PBS (control) and incubated for 60 min at room temperature. To mimic hepatic tumor metastasis, plates containing LS174T(RFP) cell colonies were seeded with mouse hepatocyte–derived AML12 cells in six-well plates. Mixtures containing Ad.CMVfLuc and [Ad.CMVfLuc][sCAR-MFE] were diluted to 600 μL with medium supplemented with 2% FBS and added to the mixed cell cultures. After incubation at 37°C for 90 min, virus-containing media were aspirated, and the cells were washed with PBS. After a 40-h incubation in medium, fLuc-mediated bioluminescence and RFP fluorescence were observed with an IVIS Optical Imaging System (Xenogen). Before imaging, culture medium was replaced with PBS. Cells were first scanned for 10 s to obtain an RFP fluorescence image. d-Luciferin (150 μg/mL; Xenogen) was then added to the wells, and repeated 1-min bioluminescence scans were acquired until maximum photon accumulation in a 1-min period was obtained.

Systemic adenovirus administration to measure hepatic untargeting. For in vivo transductional hepatic untargeting studies, 5 × 108 pfu per mouse Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE] were administered by tail vein injection. Before the injection, viruses were incubated for 1 h either with sCAR-MFE or with PBS. Injection volumes were 100 μL in all cases. On the 3rd day after virus administration, the mice were injected i.p. with d-luciferin (250 μL; ∼125 mg/kg body weight) and scanned to image adenovirus-directed fLuc activity. Immediately after imaging, mice were sacrificed, and the livers were removed and imaged for fLuc (adenovirus dependent) bioluminescence. After optical imaging, liver extracts were prepared and assayed for fLuc activity.

Adenovirus untargeting and retargeting following i.t. injection in s.c. tumor grafts. MC38 and MC38-cea-2 s.c. tumors were prepared in nude mice (nu/nu; Charles River Laboratories) by mixing 2 × 106 cells in 50 μL of PBS with Matrigel (50 μL; BD Biosciences) and injecting the mixtures s.c. into the flanks. Each mouse received MC38 cells on one flank and MC38cea-2 cells on the opposite flank. When tumors reached ∼0.5 cm in diameter (in 11–14 days), Ad.CMVfLuc (1 × 108 pfu; about 5 × 109 vp) or [Ad.CMVfluc][sCAR-MFE] (3 μg) in 20 μL was injected into the centers of tumors. Three days after i.t. virus injection, the mice were anesthetized with ketamine/xylazine (80/4 mg/kg; Phoenix Pharmaceutical), injected i.p. with d-luciferin (250 μL; ∼125 mg/kg body weight), and imaged with the Xenogen Optical Imaging System. Sequential 1-min scans were collected until maximum photon accumulation in a 1-min period was obtained. After optical imaging, mice were sacrificed. Luciferase activity in tumor extracts was measured, and protein concentrations were determined with the Bio-Rad Protein Assay (Bio-Rad).

Adenovirus untargeting and retargeting following i.v. injection of mice carrying hepatic tumor grafts. Eight-week-old (nu/nu) nude mice were anesthetized with ketamine/xylazine (100/10 mg/kg). A transverse incision was made across the xyphoid process and extended ∼2 cm. MC38 and MC38-cea-2 cells (1 × 106 per mouse) in 15 μL were injected into the front of the left upper liver lobe, using a 27-gauge needle. The lobe was returned to the abdomen, and the incision was closed with sutures and wound clips. Buprenorphine was administered every 12 h for 48 h. Wound clips were removed after 5 days. Experiments were done 5 days after surgery. For transductional untargeting and retargeting, 5 × 108 pfu per mouse of AdCMVfLuc or [Ad.CMVfLuc][sCAR-MFE] in 100 μL were administered by tail vein injection. Before injection, viruses were incubated for 1 h either with 10 μg sCAR-MFE per mouse or with PBS. Mice were sacrificed on the 4th day after virus injection. Tumor and liver extracts were prepared and assayed for luciferase activity.

Bioluminescence quantitation. Bioluminescence images were analyzed with Living Image software version 2.20 (Xenogen). Regions of interest (ROI) were drawn over the tumor or liver area, and total photons of the ROI were calculated. ROI in all images of an experiment were kept with a consistent area (24).

Statistical analysis. All experiments were done at least in triplicate. Data are presented as means ± SE and compared by Student's t test.

sCAR-MFE increases adenovirus infection of CEA-positive cells. To evaluate the retargeting efficacy of sCAR-MFE to CEA-positive, CAR-expressing tumor cells, we used cell lines with differing levels of CEA and CAR expression. Levels of CEA and CAR protein in various cell lines were determined by immunoblotting (Fig. 1A). CEA protein is not detectable in MC38, AML12, and A427 cells. LS174T and H2122 express a 180-kDa full-length CEA protein; MC38-cea-2 expresses a 70-kDa truncated CEA isoform (23). CAR is detectable by Western blotting in A427 and LS174T cells but not in H2122 cells. Thus, human A427 cells have relatively high hCAR levels but low CEA protein; LS174T and H2122 cells have intermediate and low CAR expression and express CEA protein.

Figure 1.

sCAR-MFE increases adenovirus infection of CEA-positive cells. A, CEA and CAR protein levels expressed in cell lines, analyzed by Western blotting. 14-3-3 was used as a loading control. Arrow, full-length CEA; arrowhead, truncated CEA. B, cell surface CEA expression. Unfixed cells were incubated with anti-CEA antibody to detect CEA on the cell surface. C, adenovirus targeting with sCAR-MFE in CEA-positive and CEA-negative cell lines. The cell lines were infected with Ad.CMVfLuc (3 × 108 vp per well) or with [Ad.CMVfLuc][sCAR-MFE] (50 ng) as described in Materials and Methods. Cell extracts were assayed for fLuc activity and protein concentration. Luciferase activities (a); fold increases in luciferase activity by sCAR-MFE targeting (b). Columns, averages for four cultures; bars, SE. *, P < 0.05; **, P < 0.01, for cells infected with [Ad.CMVfLuc][sCAR-MFE] versus cells infected with Ad.CMVfLuc. D, titration of sCAR-MFE retargeting activity on cell lines expressing varying CEA levels. Ad.CMVfLuc (3 × 108 vp per well) was pre-incubated with increasing amounts of sCAR-MFE (0–500 ng). Cells were then infected with the viral preparations. Cell extracts were assayed for luciferase activity and protein concentration. Points, averages of four cultures; bars, SE. Luciferase activity (a); fold increase in luciferase activity by sCAR-MFE retargeting (b).

Figure 1.

sCAR-MFE increases adenovirus infection of CEA-positive cells. A, CEA and CAR protein levels expressed in cell lines, analyzed by Western blotting. 14-3-3 was used as a loading control. Arrow, full-length CEA; arrowhead, truncated CEA. B, cell surface CEA expression. Unfixed cells were incubated with anti-CEA antibody to detect CEA on the cell surface. C, adenovirus targeting with sCAR-MFE in CEA-positive and CEA-negative cell lines. The cell lines were infected with Ad.CMVfLuc (3 × 108 vp per well) or with [Ad.CMVfLuc][sCAR-MFE] (50 ng) as described in Materials and Methods. Cell extracts were assayed for fLuc activity and protein concentration. Luciferase activities (a); fold increases in luciferase activity by sCAR-MFE targeting (b). Columns, averages for four cultures; bars, SE. *, P < 0.05; **, P < 0.01, for cells infected with [Ad.CMVfLuc][sCAR-MFE] versus cells infected with Ad.CMVfLuc. D, titration of sCAR-MFE retargeting activity on cell lines expressing varying CEA levels. Ad.CMVfLuc (3 × 108 vp per well) was pre-incubated with increasing amounts of sCAR-MFE (0–500 ng). Cells were then infected with the viral preparations. Cell extracts were assayed for luciferase activity and protein concentration. Points, averages of four cultures; bars, SE. Luciferase activity (a); fold increase in luciferase activity by sCAR-MFE retargeting (b).

Close modal

To retarget cells with recombinant bispecific adapters, the retargeting protein must be expressed on the cell surface. The cell surface location of CEA was confirmed by immunofluorescence on unfixed cells. Cell surface staining with anti-CEA antibody is observed for MC38-cea-2, LS174T, and H2122 cells; no CEA is detectable on MC38, AML12, or A427 cells (Fig. 1B).

To initially evaluate sCAR-MFE targeting efficacy, cells were infected with Ad.CMVfLuc, an adenovirus carrying a fLuc reporter gene driven by the CMV promoter, and pre-incubated either with incubation buffer or with a single concentration of sCAR-MFE. fLuc activity of the cell extracts reflects the degree of virus infection. Prior incubation with sCAR-MFE increased Ad.CMVfLuc infectivity between 5- and 55-fold for CEA-positive MC38-cea-2 (P < 0.01), H2122 (P < 0.01), and LS174T (P < 0.05) cells but did not increase infection of CEA-negative MC38 and A427 cells (Fig. 1C).

To determine an optimal sCAR-MFE:Ad ratio for retargeting efficacy, AdCMVfLuc virus was incubated with various amounts of sCAR-MFE before addition to cultured cells. sCAR-MFE:Ad retargeting is optimal at ∼100 ng sCAR-MFE per 3 × 108 vp. At this ratio, sCAR-MFE can increase the Ad.CMVfLuc infectivity ∼50-fold on MC38-cea-2 and H2122 cells and 10-fold on LS174T cells (Fig. 1D).

To show the specificity of sCAR-MFE retargeting to CEA-positive cells, the ability of rCEA protein to block retargeting was examined. Increasing concentrations of rCEA protein, but not albumin, can block sCAR-MFE–mediated retargeting of Ad.CMVfLuc infection to MC38-cea-2 (CEA positive) and LS174T (CEA positive) cells (Fig. 2). In contrast, rCEA protein has no effect on Ad.CMVfLuc infection of MC38 (CEA negative) cells.

Figure 2.

CEA blocks sCAR-MFE–directed retargeting of adenovirus infection to CEA-positive cells. LS174T (CEA+), MC38-cea-2 (CEA+), and MC38 (CEA) cells were infected with AdCMVfLuc (3 × 108 vp per well) or with [Ad.CMVfLuc][sCAR-MFE] (100 ng) in the presence of increasing amounts of recombinant CEA or bovine serum albumin (BSA). Cell extracts were assayed for fLuc activity and protein concentration. Fold retargeting is the ratio of luciferase activity in the presence of sCAR-MFE relative to activity in cells infected with Ad.CMVfLuc. Columns, averages (n = 4); bars, SE. **, P < 0.01, comparing the fold retargeting in the cells infected by [Ad.CMVfLuc][sCAR-MFE] in the absence versus the presence of rCEA.

Figure 2.

CEA blocks sCAR-MFE–directed retargeting of adenovirus infection to CEA-positive cells. LS174T (CEA+), MC38-cea-2 (CEA+), and MC38 (CEA) cells were infected with AdCMVfLuc (3 × 108 vp per well) or with [Ad.CMVfLuc][sCAR-MFE] (100 ng) in the presence of increasing amounts of recombinant CEA or bovine serum albumin (BSA). Cell extracts were assayed for fLuc activity and protein concentration. Fold retargeting is the ratio of luciferase activity in the presence of sCAR-MFE relative to activity in cells infected with Ad.CMVfLuc. Columns, averages (n = 4); bars, SE. **, P < 0.01, comparing the fold retargeting in the cells infected by [Ad.CMVfLuc][sCAR-MFE] in the absence versus the presence of rCEA.

Close modal

sCAR-MFE untargets adenovirus from CAR-positive, CEA-negative cells and retargets adenovirus to CEA-positive cells. To visualize sCAR-MFE untargeting and retargeting in cultured cells, we used Ad.CMVgfp, an adenovirus in which the CMV promoter drives GFP. A427, LS174T, and H2122 cells were infected with Ad.CMVgfp or Ad.CMVgfp pre-incubated with sCAR-MFE. For A427 cells, which express high hCAR levels and low CEA levels (Fig. 1A), sCAR-MFE reduces Ad.CMVgfp infection (Fig. 3A,, a and b). In contrast, for LS174T and H2122 cells, which express CEA, sCAR-MFE pre-incubation substantially increases infection by Ad.CMVgfp (Fig. 3A , c–f). Thus, sCAR-MFE can untarget adenovirus from CAR-mediated infection and can retarget adenovirus to CEA-positive cells.

Figure 3.

sCAR-MFE untargets adenovirus from CAR-positive cells and retargets adenovirus to CEA-positive cells. A, sCAR-MFE untargets CAR-dependent adenovirus infection and retargets infection to CEA-positive cells. Tumor cells were infected with Ad.CMVgfp or with [Ad.CMVgfp][sCAR-MFE]. GFP fluorescence reflects viral infection. B, adenovirus is untargeted by sCAR-MFE from CAR-positive, CEA-negative cells and retargeted to CEA-positive cells in cell mixtures. LS174T(RFP) CEA+ cell and A427 CEA cell mixtures were infected with Ad.CMVgfp or [Ad.CMVgfp][sCAR-MFE]. Red fluorescence reflects RFP expression in the LS174T(RFP) cells; green fluorescence reflects Ad.CMVgfp infection. Yellow signal, colocalization of GFP and RFP fluorescence (c and f). C, adenovirus is targeted to CEA-positive colonies by sCAR-MFE. Colonies of LS174T(RFP) CEA+ cells (arrowheads) in a monolayer of AML12 CEA cells (arrows) were infected with Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE]. RFP fluorescence and luciferin-dependent bioluminescence were monitored with a CCD camera. RFP fluorescence shows LS174T(RFP) colonies; fLuc (FL) bioluminescence reflects Ad.CMVfLuc infection. D, Quantitation of RFP fluorescence and luciferase activity for colonies (C). ROIs for each colony were kept constant for fluorescent and bioluminescent analyses. Luciferase activity, measured by bioluminescence, was normalized by the RFP fluorescent ROI data for each colony. Columns, averages; bars, SE. **, P < 0.01.

Figure 3.

sCAR-MFE untargets adenovirus from CAR-positive cells and retargets adenovirus to CEA-positive cells. A, sCAR-MFE untargets CAR-dependent adenovirus infection and retargets infection to CEA-positive cells. Tumor cells were infected with Ad.CMVgfp or with [Ad.CMVgfp][sCAR-MFE]. GFP fluorescence reflects viral infection. B, adenovirus is untargeted by sCAR-MFE from CAR-positive, CEA-negative cells and retargeted to CEA-positive cells in cell mixtures. LS174T(RFP) CEA+ cell and A427 CEA cell mixtures were infected with Ad.CMVgfp or [Ad.CMVgfp][sCAR-MFE]. Red fluorescence reflects RFP expression in the LS174T(RFP) cells; green fluorescence reflects Ad.CMVgfp infection. Yellow signal, colocalization of GFP and RFP fluorescence (c and f). C, adenovirus is targeted to CEA-positive colonies by sCAR-MFE. Colonies of LS174T(RFP) CEA+ cells (arrowheads) in a monolayer of AML12 CEA cells (arrows) were infected with Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE]. RFP fluorescence and luciferin-dependent bioluminescence were monitored with a CCD camera. RFP fluorescence shows LS174T(RFP) colonies; fLuc (FL) bioluminescence reflects Ad.CMVfLuc infection. D, Quantitation of RFP fluorescence and luciferase activity for colonies (C). ROIs for each colony were kept constant for fluorescent and bioluminescent analyses. Luciferase activity, measured by bioluminescence, was normalized by the RFP fluorescent ROI data for each colony. Columns, averages; bars, SE. **, P < 0.01.

Close modal

To show the ability of sCAR-MFE to specifically target CEA-positive cells in the presence of CAR-positive cells, Ad.CMVgfp retargeting was done with a mixture of CEA-positive and CEA-negative cells. Mixtures of CAR-positive, CEA-negative A427 cells and CEA-positive LS174T cells expressing RFP, LS174T(RFP), were infected with Ad.CMVgfp (Fig. 3, a–c) or [Ad.CMVgfp][sCAR-MFE] (Fig. 3, d–f). Red fluorescence identifies LS174T(RFP) cells in the cell mixture (Fig. 3B, a and d); green fluorescence identifies Ad.CMVgfp-infected cells (Fig. 3B, b and e). The RFP and GFP images are merged in Fig. 3,c and f. Ad.CMVgfp preferentially infects A427 cells, which express a high level of CAR and a low level of CEA protein, and does not infect LS174T(RFP) cells, which have low CAR levels and high CEA levels, very well (Fig. 3B, a–c). In contrast, [Ad.CMVgfp][sCAR-MFE] preferentially infects LS174T(RFP) cells in the A427/LS174T(RFP) mixtures (Fig. 3B, d–f). Thus, in cell mixtures, sCAR-MFE can both untarget adenovirus from CAR-positive cells and retarget the virus to CEA-positive cells.

To more closely model hepatic metastases in cell culture, we created colonies of LS174T(RFP) cells in a lawn of transformed mouse hepatocyte AML12 cells. LS174T(RFP) colonies were grown on culture dishes. The remaining space was filled with CEA-negative AML12 cells to mimic surrounding liver tissue (Fig. 3C, a and d). The cell cultures were infected either with Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE]. Location of LS174T(RFP) colonies was observed by RFP optical imaging (Fig. 3C, b and e). Following luciferin addition to the cultures, bioluminescence indicates the degree of Ad.CMVfLuc infection (Fig. 3C, c and f). Comparing ROI bioluminescence and fluorescence measurements of the LS174T(RFP) colonies infected with Ad.CMVfLuc and [Ad.CMVfluc][sCAR-MFE] shows that sCAR-MFE transductional retargeting increases Ad.CMVfLuc infectivity 6-fold for CEA-positive LS174T(RFP) colonies cultured in the presence of AML12 cells (Fig. 3D).

sCAR-MFE retargets adenovirus to CEA-positive s.c. tumor grafts. CEA-negative MC38 cells were injected s.c. in the left flanks of nude mice, and CEA-positive MC38-cea-2 cells were injected in the right flanks. Because adenovirus injected systemically does not reach s.c. tumors effectively (data not shown), Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE] (20 μL) were injected i.t. when the tumors reached ∼0.5 cm in diameter. Three days later, the mice were anesthetized, injected with d-luciferin, and imaged to monitor bioluminescence.

CEA-negative MC38 tumors and CEA-positive MC38-cea-2 tumors injected with Ad.CMVfLuc show essentially equivalent virus infection (Fig. 4A,, top). CEA-negative MC38 tumors injected with [Ad.CMVfLuc][sCAR-MFE] show substantially reduced (untargeted) bioluminescence when compared with MC38 tumors injected with Ad.CMVfLuc (compare tumors on the left flank in Fig. 4A,, bottom with tumors on the left flank in Fig. 4A,, top). In contrast, CEA-positive MC38-cea-2 tumors expressed substantially more bioluminescence following [Ad.CMVfluc][SCAR-MFE] injection than following Ad.CMVfLuc injection (compare tumors on the right flank in Fig. 4A,, bottom with tumors on the right flank in Fig. 4A , top).

Figure 4.

Adenovirus infection is decreased in CEA-negative MC38 s.c. tumors and increased in CEA-positive MC38-cea-2 s.c. tumors by sCAR-MFE, following i.t. injection. A, MC38 (left flank) and MC38-cea-2 (right flank) s.c. tumors were injected with Ad.CMVfLuc (1 × 108 pfu per tumor; top) or [Ad.CMVfLuc][sCAR-MFE] (3 μg per tumor; bottom). Luciferin-dependent bioluminescence was imaged 3 days after virus administration. B, quantitation of luciferase activity from the tumors of (A). Luciferase activity was quantitated from ROIs of the tumors (a and b). After bioluminescent imaging, the mice were sacrificed; tumor extracts were prepared; and luciferase activity was assayed by conventional assays (c and d). Columns, averages (n = 3); bars, SE. *, P < 0.05; **, P < 0.01, comparing tumors injected with Ad.CMFfLuc versus [Ad.CMVfLuc][sCAR-MFE].

Figure 4.

Adenovirus infection is decreased in CEA-negative MC38 s.c. tumors and increased in CEA-positive MC38-cea-2 s.c. tumors by sCAR-MFE, following i.t. injection. A, MC38 (left flank) and MC38-cea-2 (right flank) s.c. tumors were injected with Ad.CMVfLuc (1 × 108 pfu per tumor; top) or [Ad.CMVfLuc][sCAR-MFE] (3 μg per tumor; bottom). Luciferin-dependent bioluminescence was imaged 3 days after virus administration. B, quantitation of luciferase activity from the tumors of (A). Luciferase activity was quantitated from ROIs of the tumors (a and b). After bioluminescent imaging, the mice were sacrificed; tumor extracts were prepared; and luciferase activity was assayed by conventional assays (c and d). Columns, averages (n = 3); bars, SE. *, P < 0.05; **, P < 0.01, comparing tumors injected with Ad.CMFfLuc versus [Ad.CMVfLuc][sCAR-MFE].

Close modal

The tumor untargeting and retargeting results from quantitative ROI analyses of bioluminescence data obtained from the living mice are displayed in Fig. 4B (top). Following optical imaging, the mice were sacrificed. Tumors were excised, and extracts were assayed for fLuc activity, using a conventional in vitro assay. The “untargeting” effect of sCAR-MFE on CEA-negative MC38 cells and the “retargeting” effect on CEA-positive MC38-cea-2 cells, using data from the conventional luciferase assay, are shown in Fig. 4B (c and d). sCAR-MFE significantly reduces Ad.CMVfLuc infection of CEA-negative MC38 tumors. In contrast, SCAR-MFE increases Ad.CMVfLuc infection of CEA-positive MC38-cea-2 tumors ∼12-fold. sCAR-MFE thus untargets adenovirus infection of CEA-negative s.c. tumors and retargets adenovirus infection of CEA-positive tumors following i.t. injection.

sCAR-MFE untargets adenovirus from liver. To examine the efficacy of sCAR-MFE hepatic untargeting, Ad.CMVfLuc and [Ad.CMVfLuc][sCAR-MFE] were injected systemically via the tail vein into groups of three mice. Three days after viral injection, the mice were anesthetized, injected i.p. with d-luciferin, and imaged. Immediately following imaging, mice were sacrificed, and the livers were removed and imaged (Fig. 5A). The livers were then homogenized, and extracts were assayed by conventional luciferase assays. Adenovirus-mediated, luciferin-dependent bioluminescence was quantitated from the hepatic CCD images (Fig. 5B,, a) and from the conventional luciferase assays (Fig. 5B , b). Hepatic Ad.CMVfLuc infection following i.v. administration is decreased ∼80% to 90% by 5 μg per mouse of sCAR-MFE and ∼95% to 97% by 10 μg per mouse of sCAR-MFE.

Figure 5.

sCAR-MFE hepatic untargeting of adenovirus injected i.v. A, Ad.CMVfLuc (5 × 108 pfu per mouse; a–c), [Ad.CMVfLuc][sCAR-MFE] (5 μg per mouse; d–f), and [Ad.CMVfLuc][sCAR-MFE] (10 μg per mouse; g–i) were injected by tail vein into groups of three mice. The living mice and their isolated livers were imaged by bioluminescence 3 d after virus injection. B, quantitation of luciferase activity from the livers of (A). After imaging, liver extracts were prepared. Luciferase activity was quantitated by ROI analysis of luciferin-dependent bioluminescent activity in (a) and by conventional luciferase assays of liver extracts (b). Columns, averages (n = 3); bars, SE. *, P < 0.05, comparing mice injected with Ad.CMFfLuc versus [Ad.CMVfLuc][sCAR-MFE].

Figure 5.

sCAR-MFE hepatic untargeting of adenovirus injected i.v. A, Ad.CMVfLuc (5 × 108 pfu per mouse; a–c), [Ad.CMVfLuc][sCAR-MFE] (5 μg per mouse; d–f), and [Ad.CMVfLuc][sCAR-MFE] (10 μg per mouse; g–i) were injected by tail vein into groups of three mice. The living mice and their isolated livers were imaged by bioluminescence 3 d after virus injection. B, quantitation of luciferase activity from the livers of (A). After imaging, liver extracts were prepared. Luciferase activity was quantitated by ROI analysis of luciferin-dependent bioluminescent activity in (a) and by conventional luciferase assays of liver extracts (b). Columns, averages (n = 3); bars, SE. *, P < 0.05, comparing mice injected with Ad.CMFfLuc versus [Ad.CMVfLuc][sCAR-MFE].

Close modal

sCAR-MFE retargets systemically administered adenovirus to CEA-positive hepatic tumor grafts. To show effective Ad.CMVfLuc retargeting to CEA-positive tumors following systemic virus administration, we used mice bearing hepatic grafts of MC38 and MC38-cea-2 tumors. Five days after tumor cell injection into the liver, mice were injected i.v. with Ad.CMVfLuc or [Ad.CMVfLuc][sCAR-MFE] (10 μg). Four days later, mice were sacrificed, and tumor and liver extracts were assayed for luciferase activity. Liver infectivity is reduced ∼80% by sCAR-MFE untargeting for mice carrying both MC38-cea-2 and MC38 tumors (Fig. 6, a and b). A nearly 2-fold increase in Ad.CMVfLuc infection of MC38-cea-2 tumors occurs if sCAR-MFE is used for virus retargeting (Fig. 6, a). In contrast, sCAR-MFE reduces Ad.CMVfLuc infection of hepatic MC38 hepatic tumors (Fig. 6, b), presumably as a result of sCAR-MFE untargeting of CAR-dependent infection of the tumor cells.

Figure 6.

Adenovirus infection is untargeted in liver and MC38 hepatic tumors but increased in MC38-cea-2 hepatic tumors, following i.v. injection with sCAR-MFE. Mice bearing MC38-cea-2 (A) and MC38 (B) hepatic tumors were injected i.v. with Ad.CMVfLuc (5 × 108 pfu per mouse) or [Ad.CMVfLuc][sCAR-MFE] (10 μg per mouse). Four days later, mice were sacrificed, and luciferase activity and protein concentration were determined in tumor and liver extracts. C and D, tumor/liver luciferase ratios following injection with Ad.CMVfLuc (open columns) and [Ad.CMVfLuc][sCAR-MFE] (closed columns). Columns, averages (n = 4); bars, SE. *, P < 0.05, comparing tumor/liver luciferase ratios for infection by [Ad.CMVfLuc][sCAR-MFE] versus infection by AdCMVfLuc; **, P < 0.01, for luciferase activity in MC38-cea-2 tumors versus liver following systemic infection with [Ad.CMVfLuc][sCAR-MFE].

Figure 6.

Adenovirus infection is untargeted in liver and MC38 hepatic tumors but increased in MC38-cea-2 hepatic tumors, following i.v. injection with sCAR-MFE. Mice bearing MC38-cea-2 (A) and MC38 (B) hepatic tumors were injected i.v. with Ad.CMVfLuc (5 × 108 pfu per mouse) or [Ad.CMVfLuc][sCAR-MFE] (10 μg per mouse). Four days later, mice were sacrificed, and luciferase activity and protein concentration were determined in tumor and liver extracts. C and D, tumor/liver luciferase ratios following injection with Ad.CMVfLuc (open columns) and [Ad.CMVfLuc][sCAR-MFE] (closed columns). Columns, averages (n = 4); bars, SE. *, P < 0.05, comparing tumor/liver luciferase ratios for infection by [Ad.CMVfLuc][sCAR-MFE] versus infection by AdCMVfLuc; **, P < 0.01, for luciferase activity in MC38-cea-2 tumors versus liver following systemic infection with [Ad.CMVfLuc][sCAR-MFE].

Close modal

The combined effect of CAR-dependent hepatic untargeting and CEA-dependent MC38-cea-2 tumor retargeting by sCAR-MFE, as determined by the tumor/liver luciferase activity ratio, is ∼10-fold (Fig. 6, c). sCAR-MFE–mediated untargeting of both liver and MC38 tumor is essentially equivalent; consequently, the tumor/liver infectivity ratio is not altered by sCAR-MFE (Fig. 6, d).

In this report, we show effective tumor cell gene delivery using an adapter retargeted adenovirus, both in cell culture and in tumor graft models. Specifically, we have employed an sCAR-based fusion adapter system to achieve vector retargeting. In addition to tumor cell–selective enhanced targeting, this recombinant adapter also accomplished liver untargeting of the adenovirus gene delivery vector. Given the overwhelming natural hepatotropism of adenovirus, liver untargeting is particularly noteworthy. Indeed, liver untargeting provides the major practical basis for improved “target cell specificity” in vivo for adenoviral vectors, ablating the major site of vector sequestration. Consequently, tumor retargeting with sCAR-based fusion adapters should be simultaneously coupled with reduction of potentially limiting liver-associated vector toxicities. The combination of tumor retargeting and liver untargeting allows a dramatic enhancement of the targeting index achievable in vivo via adenovirus vectors.

The key accomplishment of our study is the achievement of tumor-selective targeting in combination with liver untargeting. In the first instance, the sCAR adapter molecule provides a functional cross-link between the adenovirus particle and a target “receptor” on the tumor cell. In particular, the sCAR-MFE adapter places the anti-CEA scFv tumor-targeting moiety at the natural locus of cellular recognition within the adenovirus capsid. The adapter thus allowed viral-target cell interaction to occur primarily on the basis of the association of MFE with CEA. We have previously shown successful targeting of adenovirus vectors to CEA artificially expressed in murine pulmonary vasculature, using sCAR-MFE as an adapter molecule (19). The data presented herein show successful targeting to CEA expressing tumors, both in vitro and in vivo. Our previous studies with sCAR-EGF retargeting of adenovirus to EGF-expressing tumor cells, both in culture and in tumor xenografts (15), suggest that sCAR-ligand fusions are likely to be a generalizable class of reagents for properly presenting adenoviral vectors to tumor cell “receptors” in vivo.

Liver is the most common site of metastasis for a number of epithelial cancers. Dramatic reduction in adenovirus liver infection, >90%, is observed for sCAR-MFE adapter–treated virus following systemic i.v. infection. An 80% to 90% reduction in adenovirus liver infection, following i.v. injection of virus pre-incubated with an sCAR-EGF adapter (15), suggests that sCAR-ligand fusions are also likely to be a generalizable class of reagents for achieving substantial hepatic untargeting of systemically administered therapeutic adenovirus vectors. These results provide a strong rational for further study of sCAR adapter–based targeting methods for adenovirus vectors, particularly in the context of hepatic metastatic disease, using systemic injection as the route of administration.

The precise basis of adenovirus liver untargeting observed with the use of the adapter systems remains obscure. In this regard, it has been shown in recent years that liver uptake of adenovirus vectors is not, as first postulated, solely mediated by fiber interaction with CAR or penton interaction with integrins on hepatocytes. This was illustrated by the observation that genetic mutations that abolish both CAR and penton interactions do not eliminate liver transduction (26, 27). These results exemplify that, in vivo, adenovirus hepatic transduction is a complex process. For example, a major effect on adenovirus liver tropism was achieved via ablation of binding to heparan sulfate proteoglycans (27, 28). Moreover, recent observations show that systemically administered adenovirus particles can interact with serum factors that foster their uptake and sequestration by the liver. This process may be based upon the presence of serum factor binding motifs in the fiber knob (29, 30). We can speculate that the sCAR adapter molecules mask the fiber domains that otherwise interact with either heparan sulfate proteoglycans or serum factors involved in hepatic sequestration. Irrespective of the mechanism, our data with both sCAR-MFE and sCAR-EGF (15) clearly show that the adapter-based approach can effectively untarget the liver from adenovirus transduction while simultaneously promoting tumor targeting.

The demonstrations of tumor antigen selective targeting (shown here) and tumor cell receptor overexpression targeting (15) provide a clear basis to advance adenovirus-based gene therapy/virotherapy approaches for disseminated neoplastic disease. In this regard, a synergy of effect has been observed with the combination of tumor transductional targeting and tumor-enhanced transcriptional targeting. The latter modality is based upon tumor-selective promoter control methods in which gene expression is enhanced in tumors and restricted in normal tissues. Indeed, in our groups, we have improved the targeting index several orders of magnitude with this combination, in the context of pulmonary vascular endothelial–selective gene delivery (31) and ovarian cancer therapy (32). Clearly, it would thus be logical to apply next such a double targeting approach in metastatic cancer models to further build upon the gains described herein. Additionally, cancer-selective gene expression that exploits cancer-specific patterns of translation (33) has recently been reported. In this schema, the employment of selective mRNA 5′ untranslated regions can provide expression differentials between tumor and non-tumor targets. These alternate targeting technologies provide a practical basis for combinational approaches to tumor targeting of adenoviruses, to build upon the targeting gains we have shown in the current report.

Grant support: National Cancer Institute grants R01 CA84572 and P50 CA86306 (H.R. Herschman) and NIH and Department of Defense grants R01CA111569 and W81HWH-05-0035 (D.T. Curiel).

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.

We thank the members of the Herschman and Curiel laboratories for helpful discussions and David Stout and the members of the University of California at Los Angeles Small Animal Imaging Shared Resource for technical advice and experimental assistance.

1
Louis N, Fender P, Barge A, Kitts P, Chroboczek J. Cell-binding domain of adenovirus serotype 2 fiber.
J Virol
1994
;
68
:
4104
–6.
2
Bergelson JM, Cunningham JA, Droguett G, et al. Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5.
Science
1997
;
275
:
1320
–3.
3
Bergelson JM, Krithivas A, Celi L, et al. The murine CAR homolog is a receptor for coxsackie B viruses and adenoviruses.
J Virol
1998
;
72
:
415
–9.
4
Kass-Eisler A, Falck-Pedersen E, Elfenbein DH, Alvira M, Buttrick PM, Leinwand LA. The impact of developmental stage, route of administration and the immune system on adenovirus-mediated gene transfer.
Gene Ther
1994
;
1
:
395
–402.
5
Yee D, McGuire SE, Brunner N, et al. Adenovirus-mediated gene transfer of herpes simplex virus thymidine kinase in an ascites model of human breast cancer.
Hum Gene Ther
1996
;
7
:
1251
–7.
6
Terazaki Y, Yano S, Yuge K, et al. An optimal therapeutic expression level is crucial for suicide gene therapy for hepatic metastatic cancer in mice.
Hepatology
2003
;
37
:
155
–63.
7
Okegawa T, Li Y, Pong RC, Bergelson JM, Zhou J, Hsieh JT. The dual impact of coxsackie and adenovirus receptor expression on human prostate cancer gene therapy.
Cancer Res
2000
;
60
:
5031
–6.
8
Okegawa T, Pong RC, Li Y, Bergelson JM, Sagalowsky AI, Hsieh JT. The mechanism of the growth-inhibitory effect of coxsackie and adenovirus receptor (CAR) on human bladder cancer: a functional analysis of car protein structure.
Cancer Res
2001
;
61
:
6592
–600.
9
Dmitriev I, Kashentseva E, Rogers BE, Krasnykh V, Curiel DT. Ectodomain of coxsackievirus and adenovirus receptor genetically fused to epidermal growth factor mediates adenovirus targeting to epidermal growth factor receptor-positive cells.
J Virol
2000
;
74
:
6875
–84.
10
Pereboev AV, Asiedu CK, Kawakami Y, et al. Coxsackievirus-adenovirus receptor genetically fused to anti-human CD40 scFv enhances adenoviral transduction of dendritic cells.
Gene Ther
2002
;
9
:
1189
–93.
11
Douglas JT, Rogers BE, Rosenfeld ME, Michael SI, Feng M, Curiel DT. Targeted gene delivery by tropism-modified adenoviral vectors.
Nat Biotechnol
1996
;
14
:
1574
–8.
12
Goldman CK, Rogers BE, Douglas JT, et al. Targeted gene delivery to Kaposi's sarcoma cells via the fibroblast growth factor receptor.
Cancer Res
1997
;
57
:
1447
–51.
13
Gu DL, Gonzalez AM, Printz MA, et al. Fibroblast growth factor 2 retargeted adenovirus has redirected cellular tropism: evidence for reduced toxicity and enhanced antitumor activity in mice.
Cancer Res
1999
;
59
:
2608
–14.
14
Kashentseva EA, Seki T, Curiel DT, Dmitriev IP. Adenovirus targeting to c-erbB-2 oncoprotein by single-chain antibody fused to trimeric form of adenovirus receptor ectodomain.
Cancer Res
2002
;
62
:
609
–16.
15
Liang Q, Dmitriev I, Kashentseva E, Curiel DT, Herschman HR. Noninvasive of adenovirus tumor retargeting in living subjects by a soluble adenovirus receptor-epidermal growth factor (sCAR-EGF) fusion protein.
Mol Imaging Biol
2004
;
6
:
385
–94.
16
Haisma HJ, Pinedo HM, Rijswijk A, et al. Tumor-specific gene transfer via an adenoviral vector targeted to the pan-carcinoma antigen EpCAM.
Gene Ther
1999
;
6
:
1469
–74.
17
Kelly FJ, Miller CR, Buchsbaum DJ, et al. Selectivity of TAG-72-targeted adenovirus gene transfer to primary ovarian carcinoma cells versus autologous mesothelial cells in vitro.
Clin Cancer Res
2000
;
6
:
4323
–33.
18
Reynolds PN, Zinn KR, Gavrilyuk VD, et al. A targetable, injectable adenoviral vector for selective gene delivery to pulmonary endothelium in vivo.
Mol Ther
2000
;
2
:
562
–78.
19
Everts M, Kim-Park SA, Preuss MA, et al. Selective induction of tumor-associated antigens in murine pulmonary vasculature using double-targeted adenoviral vectors.
Gene Ther
2005
;
12
:
1042
–8.
20
Mayer A, Tsiompanou E, O'Malley D, et al. Radioimmunoguided surgery in colorectal cancer using a genetically engineered anti-CEA single-chain Fv antibody.
Clin Cancer Res
2000
;
6
:
1711
–9.
21
Chester KA, Mayer A, Bhatia J, et al. Recombinant anti-carcinoembryonic antigen antibodies for targeting cancer.
Cancer Chemother Pharmacol
2000
;
46
Suppl:
S8
–12.
22
Chester K, Pedley B, Tolner B, et al. Engineering antibodies for clinical applications in cancer.
Tumour Biol
2004
;
25
:
91
–8.
23
Robbins PF, Kantor JA, Salgaller M, Hand PH, Fernsten PD, Schlom J. Transduction and expression of the human carcinoembryonic antigen gene in a murine colon carcinoma cell line.
Cancer Res
1991
;
51
:
3657
–62.
24
Liang Q, Yamamoto M, Curiel DT, Herschman HR. Noninvasive imaging of transcriptionally restricted transgene expression following intratumoral injection of an adenovirus in which the COX-2 promoter drives a reporter gene.
Mol Imaging Biol
2004
;
6
:
395
–404.
25
Neumaier M, Shively L, Chen FS, et al. Cloning of the genes for T84.66, an antibody that has a high specificity and affinity for carcinoembryonic antigen, and expression of chimeric human/mouse T84.66 genes in myeloma and Chinese hamster ovary cells.
Cancer Res
1990
;
50
:
2128
–34.
26
Martin F, Chowdhury S, Neil SJ, Chester KA, Cosset FL, Collins MK. Targeted retroviral infection of tumor cells by receptor cooperation.
J Virol
2003
;
77
:
2753
–6.
27
Smith TA, Idamakanti N, Rollence ML, et al. Adenovirus serotype 5 fiber shaft influences in vivo gene transfer in mice.
Hum Gene Ther
2003
;
14
:
777
–87.
28
Nicol CG, Graham D, Miller WH, et al. Effect of adenovirus serotype 5 fiber and penton modifications on in vivo tropism in rats.
Mol Ther
2004
;
10
:
344
–54.
29
Shayakhmetov DM. Binding of Adenovirus Fiber Knob to Blood Coaglation Factors Mediates CAR-independent Liver Tropism.
Mol Ther
2003
;
7
:
S164
–1.
30
Shayakhmetov DM, Gaggar A, Ni S, Li ZY, Lieber A. Adenovirus binding to blood factors results in liver cell infection and hepatotoxicity.
J Virol
2005
;
79
:
7478
–91.
31
Reynolds PN, Nicklin SA, Kaliberova L, et al. Combined transductional and transcriptional targeting improves the specificity of transgene expression in vivo.
Nat Biotechnol
2001
;
19
:
838
–42.
32
Barker SD, Dmitriev IP, Nettelbeck DM, et al. Combined transcriptional and transductional targeting improves the specificity and efficacy of adenoviral gene delivery to ovarian carcinoma.
Gene Ther
2003
;
10
:
1198
–204.
33
Mathis JM, Williams BJ, Sibley DA, et al. Cancer-specific targeting of an adenovirus-delivered herpes simplex virus thymidine kinase suicide gene using translational control.
J Gene Med
2006
;
8
:
1105
–20.