Materials and Methods

Cell lines and primary cultures—The cell lines, CMT-1 and CMT-2, were used in the study. These are established primary tumor cell lines that originated from resected specimens of CMTs obtained from the Veterinary Teaching Hospital at the National Taiwan University. The tumors were identified as carcinomas via histopathologic classification. Primary cultures were prepared and purified according to the protocol described previously.²²

The nonneoplastic canine mammary gland epithelial cells for primary cultures were isolated from mammary glands that were surgically removed from 1 Beagle. Surgery was performed routinely. Briefly, preoperative medication included cefazolin^a (40 mg/ kg, IV) and meloxicam^b (0.2 mg/kg, SC), followed by sedation with acepromazine^c (0.02 mg/kg, IV). Anesthesia was induced by use of propofol^d (3 mg/kg, IV) and maintained with isoflurane in oxygen. Esophageal temperature, ECG data, pulse rate, indirect arterial pressure, oxygen saturation, and end-tidal CO₂ concentration were monitored during surgery. Two (unilateral fourth and fifth glands) of 10 mammary glands were removed from the dog. After the mammary tissues were removed, the subcutaneous abdominal layers were apposed by use of 4-0 absorbable polyglactin 910 suture^e and the skin at incision sites was apposed by use of 4-0 nonabsorbable nylon suture.^f Subsequent to completion of surgery, the dog received carprofen^g (2 mg/kg, PO, q 12 h) and cephalexin^a (20 mg/kg, PO, q 12 h) for 10 days. The mammary tissue specimens were sent for histologic examination to confirm that the tissue was apparently normal (without evidence of neoplasia) prior to use in molecular studies.

A 1-cm³ sample of the nonneoplastic mammary tissue was minced by use of surgical scissors and filtered through a stainless-steel, 100-μm cell strainer.^h The tissue pieces were then dissociated by gentle agitation at 37°C for 2 hours in RPMI mediumⁱ containing collagenase^j (type IV; 1,000 U/mL), hyaluronidase (300 U/mL), DNaseI (100 μg/mL), and 2% fetal bovine serum.^k After filtration through a stainless-steel, 100-μm cell strainer to remove undissociated tissue and debris, the cells were collected via centrifugation at 1,000 X g for 5 minutes and washed twice with PBS solution. The isolated cells were then cultured in RPMI medium containing 10% fetal bovine serum, insulin (10 μg/mL), 1% penicillin-streptomycin,^l and 1% glutamine in an incubator at 37°C.

Vector construction—The VP3 gene was amplified from a Taiwan-Ilan strain of CAV. The sequence data of the nucleotides and deduced amino acids of this Taiwan-Ilan strain of CAV are deposited in the GenBank database (accession No. AF242190).²³ The plasmids pEGFP-C1,^m pCA4,ⁿ and pSIF-H1^o were used to construct VP3 protein expression vectors. Plasmid pEGFPC1 contains an EGFP gene that allows the inserted gene to be expressed as a fusion protein at the C-terminus of GFP. Plasmid pCA4 supports the expression of native VP3 protein. Plasmid pSIF-H1 was used to generate FIV-based lentivirus.²⁴ All pEGFP-C1, pCA4, and pSIF-H1 vectors are driven by a cytomegalovirus promoter. To introduce a XhoI restriction enzyme site for subcloning VP3 cDNA into the expression vectors, VP3 sequence specific PCR primers were designed and designated as LC1 as follows: 5′ LC1: 5′–CCG CTC GAG CTA TGA ACG CTC TCC AAG–3′ (forward) and 3′ LC1: 5′–CGG GAT CCT TAC AGT CTT ATA CAC CT–3′ (reverse). To introduce SacII and BamHI restriction enzyme sites for subcloning full-length and mutant cDNA of VP3 mRNA into the pEGFP-C1 expression vector, VP3 PCR primers for full-length VP3 cDNA were designed and designated as FL as follows: 5″FL: 5′–ATT AAC CGC GGG AAT GAA CGC TCT CCA AGA ′–3′; 3″FL: 5′–GGA TCC TTA CAG TCT TAT ACG CCT–3′; 5′ N-terminal (1 to 80 amino acids): 5′–ATT AAC CGC GGG AAT GAA CGC TCT CCA AGA A–3′; and 3″N–terminal (1 to 80 amino acids): 5′-ATA TAG GAT CCT TAT TGA TCG GTC CTC AAG TC–3′; 5″ C-terminal (66 to 121 amino acids): 5′–ATT AAC CGC GGG AGA AAG CAC TGG TTT CAA G–3′; and 3″ C-terminal (66 to 121 amino acids): 5′–GGA TCC TTA CAG TCT TAT ACG CCT–3′.

The full-length and mutant VP3 genes were amplified by use of Pfx DNA polymerase^p from a previously constructed plasmid (pIRES-EGFP-VP3²³) in a DNA thermal cycler.^q The PCR products were analyzed via electrophoresis and purified from agarose gel by use of a DNA extraction kit.^r To construct the pEGFP-VP3, XhoI-digested purified PCR product of the VP3 gene was inserted into the EGFP gene sequence downstream. For pCA4-VP3, the PCR product was inserted into the XhoI site of the vector. To analyze the function of deletion mutants of VP3 gene (N-terminal [EGFP-VP3N] and C-terminal [EGFP-VP3C] mutants) through construction of pEGFP-VP3 with mutant VP3, SacII- and BamHI-digested purified PCR products were inserted into 3′ multiple cloning sites of the EGFP gene to generate inframe fusion. For pSIF-EGFP-VP3, the EGFP-VP3 expression cassette was subcloned into pSIF-H1 vector by a 5″ NheI site and 3″ SalI site of the vector. After transformation, individual colonies were collected and the plasmid DNA extracted. The restriction enzyme XhoI was used to confirm the insertion of the VP3 DNA fragment for pEGFP and pCA4 vectors and the insertion of the EGFP-VP3 DNA fragment for the pSIF-H1 vector. The restriction enzymes SacII and BamHI were used to confirm the insertion of the full-length and mutant VP3 DNA fragments for the pEGFP vector. Clones containing DNA of the expected size insert were confirmed as wild type after DNA sequence analysis. Clones in the correct orientation and reading frame (for the pEGFP vectors) were chosen and designated as pEGFP-VP3, pEGFP-VP3, pEGFP-VP3N, pEGFP-VP3C, pCA4-VP3, and pSIF-EGFP-VP3 for purposes of the study.

VP3 gene transfection and lentiviral infection— Purification of plasmid DNA was performed by use of plasmid kits^s according to the manufacturer's protocols. For transfection, mammary gland cells were plated onto circular cover glasses^t (15 mm in diameter) coated with poly-lysine^u (50 mg/mL). The cells were trypsinized 1 day before transfection, counted, and plated at 1 X 10⁵ cells/well, so that 30% to 40% confluence could be achieved on the day of transfection. Liposomes were used in transfection (lipofection) of VP3 DNA into canine mammary gland cells. Liposomes are vesicles that can easily merge with the cell membrane because they are composed of a phospholipid bilayer. For each well to be transfected, 1 μg of plasmid DNA was diluted into 25 μL of buffer (HEPES, 20mM; NaCl, 100mM [pH, 7.4]) and 0.7 μL of liposome preparation was added into 25 μL of the same buffer. Once the second liposome (1,2-dioleoyl-3-trimethylammonium propane [DOTAP], 10 nmole/mL) was diluted in 50 μL of the semicomplex, it was combined with the plasmid DNA and incubated for 15 minutes at room temperature (approx 25°C) to form the DNA complexes. Cells were washed once with a serum-reducing medium.^v The DNA complexes (100 μL) and 2 volumes (200 μL) of the serum-reducing medium were added to each well. After incubation at 37°C for 1 hour, the transfection mixture was removed and new medium containing 10% fetal bovine serum was added and incubated at 37°C in a CO₂ incubator for different time courses until transgene expression assay.

Mammary gland cells without transfection were termed mock specimens.

To improve the ectopic VP3 gene expression efficiency in CMT cells, lentiviruses were generated by use of an FIV-based system^w according to the manufacturer's instructions. Briefly, human kidney 293T cells were plated to approximately 70% confluence in a 10-cm² culture dish and then cotransfected with lentiviral shuttle vector (10 μg) and packaging vectors pFIV-34N (10 μg) and pVSVG (vesicular stomatitis virus, protein G; 10 μg) by use of a calcium phosphate reagent. Sixteen hours after transfection, 5 mL of medium was removed and replaced with fresh medium. Supernatants were collected 48 hours after transfection, filtered through a 0.4-μm membrane, and then used directly to infect cells. Viral transduction was performed on 24-well plates. Cells transduced with pSIF-EGFP were used as transduction control specimens, and cells without transduction were used as mock specimens. After 24 hours of incubation, medium was removed from the plates and lentiviruses were added onto the plates in appropriate medium containing 10% fetal bovine serum and polybrene^x (8 μg/mL). All transductions were carried out with a multiplicity of infection of 4. Cells were incubated with viruses for 18 hours, followed by addition of 1 mL of medium to each well. Cells were collected for determination of apoptosis via TUNEL analysis 3 days after viral transductions.

Fluorescence microscopy analysis—At 24, 48, and 72 hours after transfection, the transfected and untransfected (mock) cells were collected and fixed with 4% paraformaldehyde in PBS solution (pH, 7.4) for 15 minutes at room temperature, washed twice with PBS solution, and stored at 4°C until analysis. Fixed cells were rinsed with PBS solution followed by treatment with 100% methanol for 15 minutes at −20°C. To stain intracellular DNA, cells were counterstained with a double-stranded DNA-binding fluorescent dye^y (50 mg/ mL) for 15 minutes in the dark at room temperature. The green fluorescent cells and cells stained with the dye were detected by use of fluorescence microscopy at 450 to 490 nm and 330 to 380 nm, respectively.

Western blot analysis—Western blot analysis was used to confirm the expression of VP3 and GFPVP3 proteins and the apoptosis-related enzymes of caspases-8, -9, and -3 and PARP.^z Every 12 hours until 72 hours after transfection, cells transfected with the pEGFP-C1 (control vector), pEGFP-VP3, pCA4 (control), or pCA4-VP3 plasmid were pelleted and lysed with protein lysis buffer (ethylene glycol tetraacetic acid, 0.19 g; NaCl, 2.62 g; NaF, 0.013 g; octylphenol-poly[ethyleneglycolether], 3 mL; and 1M Tris-Cl [pH, 7.4], 25 mL in 500 mL of distilled deionized water). After addition of 100mM phenylmethylsulphonyl fluoride (0.52 g of phenylmethylsulphonyl fluoride in 30 mL of isopropanol), the cells were kept on ice and shaken for 1 hour. Canine mammary tumor cells exposed to UV light (as much as 120 J) were used as positive control specimens for cellular apoptosis evaluations; cells without transfection were considered mock specimens. Acrylamide gel^aa was prepared according to the manufacturer's instructions. The separated protein was transferred to nitrocellulose membrane^s by use of a semidry system.^bb The membrane was immersed in 5% milk–PBST solution to block nonspecific background staining and then washed 3 times in PBST solution, followed by the addition of anti-GFP polyclonal antibody^cc (1:1,000), rabbit anti-VP3 polyclonal antibody (1:2,500), caspase-8 (1:1,000), cleaved caspase-9 (1:1,000), caspase-3 (1:1,000), PARP (1:1,000), or β-actin^dd (1:10,000). The horseradish peroxidase–labeled secondary antibody^ee (1:5,000) for generating chemiluminescence was added after washing the membrane with PBST solution. A substrate reagent was added to the membrane and incubated for 1 minute. The membrane was placed on a radiographic film^ff in a dark room for optimal exposure (approx 3 to 5 minutes) before final development.

Indirect immunofluorescence analysis—Cells were seeded in 12-well plates to a final concentration of 2 X 10⁴/well, cultured overnight, and then transfected with pCA4 (control) and pCA4-VP3 DNA. Conditions for transfection were similar to those described previously. Cells were incubated in blocking solution (PBS solution containing 5% fetal bovine serum) for 15 minutes at room temperature. After removal of the blocking solution, cells were incubated in rabbit anti-VP3 antibody (1:2,500) at room temperature for 1 hour. Following several rinses with PBS solution, cells were incubated with fluorescein isothiocyanate–labeled anti-rabbit IgG^gg (1:200) at room temperature for 1 hour; cells were counterstained with the double-stranded DNA-binding fluorescent dye as described. Photographic images of western blots were analyzed by use of a computer program^hh to obtain quantitative data regarding protein expression. Protein expression blots were measured and calculated in relation to intrinsic control expression of β-actin from each sample. The results were analyzed and represented graphically by use of a computer program.ⁱⁱ

Detection of apoptosis—Apoptotic cells were quantified by use of a TUNEL assay involving the TUNEL reaction mixture^jj according to the manufacturer's instructions. Cells were counterstained with blue fluorescent nuclear stain^kk (1 μg/mL) and examined by use of a fluorescence microscope. At least 500 cells in different fields/sample were evaluated. Apoptosis was assessed as the fraction of cells with nuclear condensation and fragmentation that were positive for TUNEL staining.

Tumor cells were also harvested after 48 hours of transfection and stained with propidium iodide to determine the DNA indexes after flow cytometry analysis to evaluate the changes in cell cycle after VP3 transfection or pretreatment for 3 hours with 100μMzVAD.fmk^ll (a pan-caspase inhibitor) in VP3 gene or vector control-transfected cells. Cells were collected via trypsinization and fixed in ice-cold methanol–PBS solution (2:1 [vol/vol]). Fluorescence emitted from the propidium iodide–DNA complexes was quantitated after laser excitation of the fluorescent dye by a fluorescence-activated cell sorter with specialized computer software.^mm The fluorescence-activated cell sorter uses laser light to excite fluorescent-labeled molecules (such as antibodies or proteins) attached to or located inside cells and subsequently detects the fluorescent light emitted from cells that are passing a sensor in single file in a fluid stream. The ratio of cells in the sub-G1 phase was determined by calculating the percentage of cells with less-than-expected diploid (2N) DNA content by the cell population at all phases (G1, S, G2+M, and Sub-G1).

Statistical analysis—All of the experiments were performed in triplicate. Results were expressed as mean ± SD. Statistical comparisons in percentage of apoptotic cells between experimental and control groups were made by use of a Student t test. A value of P < 0.05 was considered significant.

Results

Expression patterns of VP3 fusion protein in CMT cells—The extent of apoptosis induced by the VP3 protein in CMTs was determined by use of the CMT cell lines. The VP3 gene was fused with EGFP for real-time detection of VP3 transgene expression. The expressions of both pEGFP-VP3 and control pEGFP vectors were first confirmed via western blot analysis; by use of polyclonal antibody against EGFP, the EGFP and EGFP-VP3 fusion proteins were detected at 30 and 40 kd, respectively, 48 hours after transfection (Figure 1). The expression patterns for both EGFP and EGFP-VP3 were further examined via fluorescence microscopy. At 48 hours after transfection, EGFP protein was distributed evenly in the cell cytoplasm, whereas the EGFP-VP3 fusion protein was localized to the nuclei with a granular distribution.

Results of western blot analysis and fluorescence microscopy performed on lysates of CMT cells and whole cells, respectively, 48 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid. A—Protein expression patterns were determined via western blot analysis. The EGFP protein (30 kd) and EGFP-VP3 fusion protein (40 kd) are detectable in the cell lysates by use of polyclonal antibody against EGFP. Protein markers ranging from 17 to 62 kd are shown on the right of the panel. B—Photomicrograph of CMT cells stained for detection of EGFP protein via fluorescence microscopy. At 48 hours after transfection, EGFP protein is distributed evenly in the cell cytoplasm. C—Photomicrograph of CMT cells stained for detection of EGFP-VP3 fusion protein via fluorescence microscopy. At 48 hours after transfection, the distribution of the fusion protein is localized to the cell nuclei. Bar (applies to panels B and C) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of western blot analysis and fluorescence microscopy performed on lysates of CMT cells and whole cells, respectively, 48 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid. A—Protein expression patterns were determined via western blot analysis. The EGFP protein (30 kd) and EGFP-VP3 fusion protein (40 kd) are detectable in the cell lysates by use of polyclonal antibody against EGFP. Protein markers ranging from 17 to 62 kd are shown on the right of the panel. B—Photomicrograph of CMT cells stained for detection of EGFP protein via fluorescence microscopy. At 48 hours after transfection, EGFP protein is distributed evenly in the cell cytoplasm. C—Photomicrograph of CMT cells stained for detection of EGFP-VP3 fusion protein via fluorescence microscopy. At 48 hours after transfection, the distribution of the fusion protein is localized to the cell nuclei. Bar (applies to panels B and C) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of western blot analysis and fluorescence microscopy performed on lysates of CMT cells and whole cells, respectively, 48 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid. A—Protein expression patterns were determined via western blot analysis. The EGFP protein (30 kd) and EGFP-VP3 fusion protein (40 kd) are detectable in the cell lysates by use of polyclonal antibody against EGFP. Protein markers ranging from 17 to 62 kd are shown on the right of the panel. B—Photomicrograph of CMT cells stained for detection of EGFP protein via fluorescence microscopy. At 48 hours after transfection, EGFP protein is distributed evenly in the cell cytoplasm. C—Photomicrograph of CMT cells stained for detection of EGFP-VP3 fusion protein via fluorescence microscopy. At 48 hours after transfection, the distribution of the fusion protein is localized to the cell nuclei. Bar (applies to panels B and C) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Changes in EGFP-VP3 fusion protein-expressing cells with time—As the duration of expression increased, microscopic examination revealed that cells expressing EGFP-VP3 fusion protein became rounded and eventually detached from the culture surface (Figure 2). At 72 hours after transfection, untransfected cells were predominant in the culture plates and the fluorescent cells were barely evident. In comparison to the fusion protein–expressing cells, which decreased in size and became rounded, cells in the control (EGFP-C1 transfected) group retained their normal parental shape. Western blot analysis was performed to evaluate the quantity of EGFP and EGFP-VP3 in cells expressing those proteins. In comparison with the control group, in which substantial and relatively consistent EGFP expression was detected throughout the 96-hour posttransfection period, detectable quantities of the EGFP-VP3 fusion protein decreased significantly during the same time period; the VP3 fusion protein was barely detectable at 72 hours after transfection and almost undetectable at 96 hours (Figure 3). Quantitative analysis revealed relatively stable EGFP protein expression in the control group, whereas there was marked decrease in protein expression in the EGFP-VP3–expressing cells as a result of extensive cell death after VP3 expression.

Photomicrographs of CMT cells at intervals after transfection with the pEGFP-VP3 plasmid (A–F) and CMT cells at 96 hours after transfection with the pEGFP-C1 plasmid (control group; G and H) obtained via fluorescence microscopy (A, C, E, and G) and direct light microscopy (B, D, F, and H). The images in each column represent the same cells. The cells expressing the fusion protein become rounded with time (A–F), whereas the control cells expressing the EGFP protein retain an apparently normal shape (G and H). At 96 hours after transfection, cells expressing the EGFP-VP3 protein are barely detectable via fluorescence microscopy (E), despite the abundance of cells in the culture plates (F). Bar (applies to all panels) = 20 mm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photomicrographs of CMT cells at intervals after transfection with the pEGFP-VP3 plasmid (A–F) and CMT cells at 96 hours after transfection with the pEGFP-C1 plasmid (control group; G and H) obtained via fluorescence microscopy (A, C, E, and G) and direct light microscopy (B, D, F, and H). The images in each column represent the same cells. The cells expressing the fusion protein become rounded with time (A–F), whereas the control cells expressing the EGFP protein retain an apparently normal shape (G and H). At 96 hours after transfection, cells expressing the EGFP-VP3 protein are barely detectable via fluorescence microscopy (E), despite the abundance of cells in the culture plates (F). Bar (applies to all panels) = 20 mm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Photomicrographs of CMT cells at intervals after transfection with the pEGFP-VP3 plasmid (A–F) and CMT cells at 96 hours after transfection with the pEGFP-C1 plasmid (control group; G and H) obtained via fluorescence microscopy (A, C, E, and G) and direct light microscopy (B, D, F, and H). The images in each column represent the same cells. The cells expressing the fusion protein become rounded with time (A–F), whereas the control cells expressing the EGFP protein retain an apparently normal shape (G and H). At 96 hours after transfection, cells expressing the EGFP-VP3 protein are barely detectable via fluorescence microscopy (E), despite the abundance of cells in the culture plates (F). Bar (applies to all panels) = 20 mm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of western blot analysis to determine expression of EGFP (A) and EGFP-VP3 (B) proteins in CMT cells at 48, 72, and 96 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid, respectively, and quantitative comparison of those protein expression levels (C). Relatively substantial EGFP protein expression is evident in cells throughout the 96-hour posttransfection period (A), whereas expression of the EGFP-VP3 fusion protein progressively decreases and is barely detectable and undetectable at 72 and 96 hours after transfection, respectively (B). B-Actin was used as a control sample for each blot analysis. Quantitative comparison of expression of the fusion protein (black bars) and EGFP protein (white bars) performed via computer analysis of photographic images of western blot images identified a marked decrease in fusion protein expression level with time (C).

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of western blot analysis to determine expression of EGFP (A) and EGFP-VP3 (B) proteins in CMT cells at 48, 72, and 96 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid, respectively, and quantitative comparison of those protein expression levels (C). Relatively substantial EGFP protein expression is evident in cells throughout the 96-hour posttransfection period (A), whereas expression of the EGFP-VP3 fusion protein progressively decreases and is barely detectable and undetectable at 72 and 96 hours after transfection, respectively (B). B-Actin was used as a control sample for each blot analysis. Quantitative comparison of expression of the fusion protein (black bars) and EGFP protein (white bars) performed via computer analysis of photographic images of western blot images identified a marked decrease in fusion protein expression level with time (C).

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of western blot analysis to determine expression of EGFP (A) and EGFP-VP3 (B) proteins in CMT cells at 48, 72, and 96 hours after transfection with the pEGFP-C1 or pEGFP-VP3 plasmid, respectively, and quantitative comparison of those protein expression levels (C). Relatively substantial EGFP protein expression is evident in cells throughout the 96-hour posttransfection period (A), whereas expression of the EGFP-VP3 fusion protein progressively decreases and is barely detectable and undetectable at 72 and 96 hours after transfection, respectively (B). B-Actin was used as a control sample for each blot analysis. Quantitative comparison of expression of the fusion protein (black bars) and EGFP protein (white bars) performed via computer analysis of photographic images of western blot images identified a marked decrease in fusion protein expression level with time (C).

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Expression of native VP3 gene in CMT cells—To rule out the possible influence of EGFP protein on induction of cell death, the VP3 gene was reconstructed into the eukaryotic expression vector, pCA4. This vector supports expression of native VP3 gene by use of the same cytomegalovirus promoter required by pEGFP-C1. The expression of VP3 protein was first confirmed via western blot analysis with anti-VP3 protein antibody. Cell lysates of CMT cells that were transfected with pCA4, pCA4-VP3, and the 2 constructs used in previous experiments (pEGFP-C1 and pEGFP-VP3) were also analyzed. The results indicate that only cells with plasmids containing the VP3 gene (eg, pEGFP-VP3 and pCA4-VP3) were detected. The size of the expressed native VP3 protein was 17 kd and that of the EGFP-VP3 fusion protein was 40 kd, whereas no protein band was detected from vectors without VP3 gene insertion (Figure 4).

Results of western blot analysis (A) and immunofluorescence staining (B–G) of CMT-1 cells transfected with the VP3 gene to assess the influence of EGFP protein on induction of cell death and characterize cell death caused by the VP3 protein. The VP3 gene was reconstructed into the eukaryotic expression vector, pCA4; CMT-1 cells were transfected with the pCA4, pCA4-VP3, pEGFP-C1, or pEGFP-VP3 plasmid. Via western blot analysis (A), the VP3 protein is detectable in pEGFP-VP3– and pCA4-VP3–transfected cells (bands at 17 kd [native VP3 protein] and 40 kd [EGFP-VP3 fusion protein]); no VP3 protein band can be detected in cells transfected with pCA4 or pEGFP-C1. Protein markers ranging from 3 to 43 kd are shown to the right of the panel. By use of immunofluorescence staining, the nuclear morphology of cells expressing VP3 protein was examined at intervals after transfection with pEGFP-VP3. At 24 and 48 hours after transfection, DNA staining in the nuclei of the control cells (transfected with pEGFP-C1) is even and regular (data not shown). The staining of the nuclei (arrows) of cells expressing VP3 protein is patchy and uneven (B and D); the nuclei (arrowheads) become irregularly shaped and weakly stained (C and E). At 72 hours after transfection, nuclei of cells expressing the VP3 protein are fragmented and condensed (F and G). Bar (applies to panels B through G) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of western blot analysis (A) and immunofluorescence staining (B–G) of CMT-1 cells transfected with the VP3 gene to assess the influence of EGFP protein on induction of cell death and characterize cell death caused by the VP3 protein. The VP3 gene was reconstructed into the eukaryotic expression vector, pCA4; CMT-1 cells were transfected with the pCA4, pCA4-VP3, pEGFP-C1, or pEGFP-VP3 plasmid. Via western blot analysis (A), the VP3 protein is detectable in pEGFP-VP3– and pCA4-VP3–transfected cells (bands at 17 kd [native VP3 protein] and 40 kd [EGFP-VP3 fusion protein]); no VP3 protein band can be detected in cells transfected with pCA4 or pEGFP-C1. Protein markers ranging from 3 to 43 kd are shown to the right of the panel. By use of immunofluorescence staining, the nuclear morphology of cells expressing VP3 protein was examined at intervals after transfection with pEGFP-VP3. At 24 and 48 hours after transfection, DNA staining in the nuclei of the control cells (transfected with pEGFP-C1) is even and regular (data not shown). The staining of the nuclei (arrows) of cells expressing VP3 protein is patchy and uneven (B and D); the nuclei (arrowheads) become irregularly shaped and weakly stained (C and E). At 72 hours after transfection, nuclei of cells expressing the VP3 protein are fragmented and condensed (F and G). Bar (applies to panels B through G) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of western blot analysis (A) and immunofluorescence staining (B–G) of CMT-1 cells transfected with the VP3 gene to assess the influence of EGFP protein on induction of cell death and characterize cell death caused by the VP3 protein. The VP3 gene was reconstructed into the eukaryotic expression vector, pCA4; CMT-1 cells were transfected with the pCA4, pCA4-VP3, pEGFP-C1, or pEGFP-VP3 plasmid. Via western blot analysis (A), the VP3 protein is detectable in pEGFP-VP3– and pCA4-VP3–transfected cells (bands at 17 kd [native VP3 protein] and 40 kd [EGFP-VP3 fusion protein]); no VP3 protein band can be detected in cells transfected with pCA4 or pEGFP-C1. Protein markers ranging from 3 to 43 kd are shown to the right of the panel. By use of immunofluorescence staining, the nuclear morphology of cells expressing VP3 protein was examined at intervals after transfection with pEGFP-VP3. At 24 and 48 hours after transfection, DNA staining in the nuclei of the control cells (transfected with pEGFP-C1) is even and regular (data not shown). The staining of the nuclei (arrows) of cells expressing VP3 protein is patchy and uneven (B and D); the nuclei (arrowheads) become irregularly shaped and weakly stained (C and E). At 72 hours after transfection, nuclei of cells expressing the VP3 protein are fragmented and condensed (F and G). Bar (applies to panels B through G) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Assessment of VP3 protein-associated apoptosis among CMT cells—After expression of VP3 protein had been confirmed in CMT cells, functional consequences of that protein expression were further analyzed by direct comparison of the transfected cells. At 24, 48, and 72 hours after transfection, cells were removed from the incubator and directly observed by use of an inverted microscope. At 24 hours after transfection, limited cell rounding and floating were evident in both the pCA4-VP3 and pCA4 plasmid-transfected cells, which may have resulted from the toxic effect of the liposome used to transport plasmid DNA into the cells. With time, the proportion of dead cells in the VP3 protein–expressing groups became more prominent, compared with the control group (data not shown). To further characterize the mechanism underlying the cell death caused by VP3 protein expression, the nuclear morphology of cells that expressed VP3 protein after transfection was examined at various stages of the cell cycle stages by use of DNA-specific dyes. The percentage of cells in different cell cycle stages between experimental groups and control group was also compared. Although most of the control group cell nuclei were stained evenly and regularly at 24 and 48 hours after transfection, staining of the VP3 protein–expressing cell nuclei was patchy and uneven (Figure 5). With time, the nuclei of cells expressing VP3 protein became irregularly shaped with weak DNA staining. At 72 hours after transfection, fragmentation and condensation of the nuclei (characteristics of cellular apoptosis) were prominent in cells that expressed VP3 protein. Results of TUNEL staining confirmed that cells were undergoing apoptosis after VP3 gene transfection; VP3-transfected CMT cells with nuclear chromatin condensation and fragmentation yielded positive results via TUNEL analysis. The percentage of cells undergoing apoptosis was approximately 31 ± 4%. Among vector-transfected cells, apoptosis was only rarely detected (4 ± 1% of cells).

Results of immunofluorescence staining with a double-stranded DNA-binding fluorescent dye in CMT-1 cells at 48 hours after transfection with pCA4-VP3. Although the cells transfected with control vector (pCA4) did not show changes of DNA staining via fluorescence microscopy, the nuclei of cells expressing VP3 protein became irregularly shaped with weak DNA staining (A). At 72 hours after transfection, fragmentation and condensation of the nuclei were prominent in the cells that expressed VP3 protein (B). Mock = Untransfected cells. Bar (applies to both images) = 50 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of immunofluorescence staining with a double-stranded DNA-binding fluorescent dye in CMT-1 cells at 48 hours after transfection with pCA4-VP3. Although the cells transfected with control vector (pCA4) did not show changes of DNA staining via fluorescence microscopy, the nuclei of cells expressing VP3 protein became irregularly shaped with weak DNA staining (A). At 72 hours after transfection, fragmentation and condensation of the nuclei were prominent in the cells that expressed VP3 protein (B). Mock = Untransfected cells. Bar (applies to both images) = 50 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of immunofluorescence staining with a double-stranded DNA-binding fluorescent dye in CMT-1 cells at 48 hours after transfection with pCA4-VP3. Although the cells transfected with control vector (pCA4) did not show changes of DNA staining via fluorescence microscopy, the nuclei of cells expressing VP3 protein became irregularly shaped with weak DNA staining (A). At 72 hours after transfection, fragmentation and condensation of the nuclei were prominent in the cells that expressed VP3 protein (B). Mock = Untransfected cells. Bar (applies to both images) = 50 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Assessment of apoptosis pathway associated with VP3 protein expression—The biochemical features that resulted in apoptosis induced by overexpression of VP3 protein in CMT cells were investigated. Western blot analysis was performed to examine the activation of caspases after VP3 protein overexpression in CMT cells that were transfected with pCA4-VP3. Samples collected at 12, 24, 48, and 72 hours after transfection were blotted with monoclonal antibodies that recognize proand cleaved forms of caspases-3, -8, and -9 and PARP. Cleaved caspase-9 was detected at 12 hours; peak amounts of this enzyme were detected at 24 hours after VP3 transfection (Figure 6). Cleaved caspase-3 and PARP were also detected at 12 hours after transfection. Caspase-8 activation was not involved in VP3 protein–induced apoptosis because the amounts of this enzyme remained unmodified, compared with amounts of caspases-3 and -9. Furthermore, pretreatment of cells with zVAD. fmk, a pan-caspase inhibitor, completely abrogated the induction of apoptosis by VP3 protein. The percentage of sub-G1 cells in the VP3 transfected group was 33 ± 5%, whereas the percentage was markedly reduced to 5 ± 1% in the zVAD.fmk-pretreated VP3 transfected group. The difference in percentage of sub-G1 cells between these 2 groups was significant (P < 0.05); thus, it appears that VP3 protein-induced apoptosis among CMT cells requires the activation of the caspase-9 mediated apoptosis pathway.

Results of western blot analysis of CMT cells at 12, 24, 48, and 72 hours after transfection with the pCA4 (vector control; left panels) or pCA4-VP3 (right panels) plasmid to assess the activation of caspases after VP3 protein overexpression. Gels were blotted with monoclonal antibodies that recognize pro- and cleaved forms of caspases-3, -8, and -9 and PARP. In pCA4-VP3–transfected cells, cleaved caspase-9 is detectable after 12 hours and peak amount is present at 24 hours. Cleaved caspase-3 and PARP are also detectable after 12 hours. There is no evidence of cleaved caspase-8 in pCA4-VP3–transfected cells. Data obtained from CMT-1 cells exposed to UV light are illustrated, representing the mitochondrial-mediated apoptosis pathway that results in activation of caspase-9 and -3 and PARP (far right lane). B-Actin expression was used as quantitative control of protein expression. Values to the left are molecular weights in kilodaltons. Casp = Caspase.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Results of western blot analysis of CMT cells at 12, 24, 48, and 72 hours after transfection with the pCA4 (vector control; left panels) or pCA4-VP3 (right panels) plasmid to assess the activation of caspases after VP3 protein overexpression. Gels were blotted with monoclonal antibodies that recognize pro- and cleaved forms of caspases-3, -8, and -9 and PARP. In pCA4-VP3–transfected cells, cleaved caspase-9 is detectable after 12 hours and peak amount is present at 24 hours. Cleaved caspase-3 and PARP are also detectable after 12 hours. There is no evidence of cleaved caspase-8 in pCA4-VP3–transfected cells. Data obtained from CMT-1 cells exposed to UV light are illustrated, representing the mitochondrial-mediated apoptosis pathway that results in activation of caspase-9 and -3 and PARP (far right lane). B-Actin expression was used as quantitative control of protein expression. Values to the left are molecular weights in kilodaltons. Casp = Caspase.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Results of western blot analysis of CMT cells at 12, 24, 48, and 72 hours after transfection with the pCA4 (vector control; left panels) or pCA4-VP3 (right panels) plasmid to assess the activation of caspases after VP3 protein overexpression. Gels were blotted with monoclonal antibodies that recognize pro- and cleaved forms of caspases-3, -8, and -9 and PARP. In pCA4-VP3–transfected cells, cleaved caspase-9 is detectable after 12 hours and peak amount is present at 24 hours. Cleaved caspase-3 and PARP are also detectable after 12 hours. There is no evidence of cleaved caspase-8 in pCA4-VP3–transfected cells. Data obtained from CMT-1 cells exposed to UV light are illustrated, representing the mitochondrial-mediated apoptosis pathway that results in activation of caspase-9 and -3 and PARP (far right lane). B-Actin expression was used as quantitative control of protein expression. Values to the left are molecular weights in kilodaltons. Casp = Caspase.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Apoptosis induction potentials of VP3 protein in CMT cell lines and non-neoplastic mammary gland epithelial cells—To analyze the extent of apoptosis induced by VP3 protein in CMT-1 and CMT-2 cell lines and non-neoplastic mammary gland epithelial cells, the EGFP only and EGFP-VP3 fusion lentiviruses were generated. The CMT cells and nonneoplastic mammary gland epithelial cells were transduced with pSIF-EGFP and pSIF-EGFP-VP3 and then harvested at 72 hours after transduction. The results indicate that > 90% of the CMT cells and approximately 85% of the nonneoplastic mammary gland epithelial cells were infected with pSIF-EGFP (multiplicity of infection, 4; Figure 7). Infection efficiency associated with pSIF-EGFP-VP3 was similar. The pSIF-EGFP-VP3 was associated with nuclear expression and caused considerable rounding up and detachment of CMT cells, whereas nonneoplastic mammary gland epithelial cells expressed EGFP-VP3 only in the cytoplasm and remained attached to the culture surface. The percentage of apoptotic cells in the 2 CMT cell lines and nonneoplastic mammary gland epithelial cells after either EGFP or EGFP-VP3 lentivirus infection was assessed via TUNEL analysis. The results indicate that > 70% of the tumor cells underwent apoptosis (78 ± 10% in the CMT-1 cell line and 73 ± 8% in the CMT-2 cell line) after infection with lentivirus encoding EGFP-VP3, which was significantly (P < 0.05) greater than findings in EGFP control groups (12 ± 5% and 8 ± 3% in the CMT-1 and CMT-2 cell line, respectively). In nonneoplastic mammary gland epithelial cells, transduction of EGFP-VP3 lentivirus resulted in minimal induction of apoptosis (13 ± 5%), which was not significantly different from findings in the EGFP control transduced group (10 ± 3%).

Photomicrographs of CMT cells and nonneoplastic canine mammary gland epithelial cells lines obtained via direct light microscopy (A, C, E, G, I, and K) and fluorescence microscopy (to determine expression of VP3 protein; B, D, F, H, J, and L) at 72 hours in cells without transduction (A–D) or with transduction by lentiviral-vector infection with pSIF-EGFP (E–H) or pSIF-EGFP-VP3 (I–L). Transduction efficiency of pSIF-EGFP in CMT cells is 90% and no apoptosis is observed. The VP3 protein is expressed in the nucleus of the pSIF-EGFP-VP3–transduced CMT cells and causes marked apoptosis (J); the protein is expressed in the cytoplasm of pSIF-EGFP-VP3–transduced nonneoplastic mammary gland cells and causes only minimal apoptosis (L) with no notable change, compared with the pSIF-EGFP–transduced control cells (H). Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photomicrographs of CMT cells and nonneoplastic canine mammary gland epithelial cells lines obtained via direct light microscopy (A, C, E, G, I, and K) and fluorescence microscopy (to determine expression of VP3 protein; B, D, F, H, J, and L) at 72 hours in cells without transduction (A–D) or with transduction by lentiviral-vector infection with pSIF-EGFP (E–H) or pSIF-EGFP-VP3 (I–L). Transduction efficiency of pSIF-EGFP in CMT cells is 90% and no apoptosis is observed. The VP3 protein is expressed in the nucleus of the pSIF-EGFP-VP3–transduced CMT cells and causes marked apoptosis (J); the protein is expressed in the cytoplasm of pSIF-EGFP-VP3–transduced nonneoplastic mammary gland cells and causes only minimal apoptosis (L) with no notable change, compared with the pSIF-EGFP–transduced control cells (H). Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Photomicrographs of CMT cells and nonneoplastic canine mammary gland epithelial cells lines obtained via direct light microscopy (A, C, E, G, I, and K) and fluorescence microscopy (to determine expression of VP3 protein; B, D, F, H, J, and L) at 72 hours in cells without transduction (A–D) or with transduction by lentiviral-vector infection with pSIF-EGFP (E–H) or pSIF-EGFP-VP3 (I–L). Transduction efficiency of pSIF-EGFP in CMT cells is 90% and no apoptosis is observed. The VP3 protein is expressed in the nucleus of the pSIF-EGFP-VP3–transduced CMT cells and causes marked apoptosis (J); the protein is expressed in the cytoplasm of pSIF-EGFP-VP3–transduced nonneoplastic mammary gland cells and causes only minimal apoptosis (L) with no notable change, compared with the pSIF-EGFP–transduced control cells (H). Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Analysis of the proapoptosis activity caused by deletion mutants of VP3 protein—The extent of apoptosis induced by full-length and deletion mutants of VP3 protein was assessed. The cDNA of the full-length VP3 protein and 2 deletion mutants (N-terminal [EGFP-VP3N] and C-terminal [EGFP-VP3C] mutants) was each fused with EGFP for real-time detection of transgene expression. Full-length EGFP-VP3 protein was detected exclusively in the nucleus of CMT-1 cells (Figure 8). The EGFP-VP3N protein was detected in the cytoplasm, whereas EGFP-VP3C protein was localized predominantly in the nucleus of CMT-1 cells. The percentages of apoptotic cells induced by the full-length protein and deletion mutants of the VP3 protein were evaluated. The percentages of cells undergoing apoptosis induced by full-length EGFP-VP3 and the C-terminal mutant of EGFP-VP3 were significantly (P < 0.05) higher (35 ± 8% and 23 ± 6%, respectively) than findings with the EGFP vector and N-terminal mutant of EGFP-VP3 (5 ± 2% and 7 ± 4%, respectively).

Photomicrographs of CMT cells obtained via direct light microscopy (A, C, E, G, and I) and fluorescence microscopy (B, D, F, H, and J) 48 hours after transfection with full-length (FL) VP3 cDNA and cDNA of VP3 deletion mutants (N-terminal and C-terminal mutants; EGFP-VP3N and EGFP-VP3C) fused with pEGFP for real-time detection of transgene expression. The EGFP-VP3FL protein is localized exclusively in the nucleus of CMT cells (F). The EGFP-VP3N protein is restricted to the cell cytoplasm (H), whereas EGFP-VP3C protein is localized predominantly in the nucleus (J) of the CMT cells. Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photomicrographs of CMT cells obtained via direct light microscopy (A, C, E, G, and I) and fluorescence microscopy (B, D, F, H, and J) 48 hours after transfection with full-length (FL) VP3 cDNA and cDNA of VP3 deletion mutants (N-terminal and C-terminal mutants; EGFP-VP3N and EGFP-VP3C) fused with pEGFP for real-time detection of transgene expression. The EGFP-VP3FL protein is localized exclusively in the nucleus of CMT cells (F). The EGFP-VP3N protein is restricted to the cell cytoplasm (H), whereas EGFP-VP3C protein is localized predominantly in the nucleus (J) of the CMT cells. Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Photomicrographs of CMT cells obtained via direct light microscopy (A, C, E, G, and I) and fluorescence microscopy (B, D, F, H, and J) 48 hours after transfection with full-length (FL) VP3 cDNA and cDNA of VP3 deletion mutants (N-terminal and C-terminal mutants; EGFP-VP3N and EGFP-VP3C) fused with pEGFP for real-time detection of transgene expression. The EGFP-VP3FL protein is localized exclusively in the nucleus of CMT cells (F). The EGFP-VP3N protein is restricted to the cell cytoplasm (H), whereas EGFP-VP3C protein is localized predominantly in the nucleus (J) of the CMT cells. Bar (applies to all panels) = 20 μm.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.411

• Download Figure
• Download figure as PowerPoint slide

Discussion

The VP3 protein of CAV is known to have an ability to induce apoptosis, specifically among tumor cells of human origin.²⁵ Results of the study reported here indicate that the VP3 protein obtained from a Taiwanese isolate (Taiwan-Ilan) of CAV can induce death among tumor cells of canine origin in vitro, which to our knowledge, has not been explored in the field of veterinary oncology.

The VP3 gene used in our study was cloned from a CAV isolate of the Taiwan-Ilan strain. The apoptosis-inducing ability of this virus has been confirmed previously; the DNA isolated from virus-infected cells has an apoptosis-specific laddering pattern and yields positive results via TUNEL analysis.¹⁶ The alignment of the VP3 gene from the Taiwan-Ilan isolate has an overall homology of 98.1% to 100% with other known CAV strains, and it is identical to the American CIA-1 strain.²⁶ However, compared with the sequence of the German Cuxhaven-1 strain, a difference of 2 bases at the 3′ end (C-terminus) has been identified.²³ These base differences result in 2 amino acid changes at residue 116 (from arginine to lysine) and residue 118 (from cysteine to arginine).²³ The first substitution retains the basic character of the VP3 protein in both the German Cuxhaven-1 strain and the Taiwan-Ilan strain. However, in the second substitution, a sulfur side-chain amino acid (cysteine in the Taiwan-Ilan strain VP3 protein) replaces a basic side-chain amino acid (arginine in the German Cuxhaven-1 strain VP3 protein). As the VP3 protein C-terminus is reported²¹ to be critical to apoptotic activity, the consequence of the substitution of amino acids in the Taiwan-Ilan strain remains to be determined.

In our study, most of the VP3 protein in CMT cells was expressed only in the nucleus, where it aggregated to form a fine granular pattern, a finding that is consistent with results of other research.^18,21 This nuclear localization of VP3 protein strongly suggests an association between VP3 protein and cell-killing activity in tumor cells.^27,28 The tumor cell–specific activity of the CAV-associated VP3 protein is believed to be dependent on its ability to localize in the nuclei of transformed cells, but not in nuclei of primary or nontransformed cells. In recent studies,^29–32 the signaling responsible for the nucleocytoplasmic translocation of VP3 protein has been characterized.

The nuclear targeting abilities of VP3 protein appear to be associated with a tumor cell–specific nuclear targeting signal in the C-terminus; the targeting signal is determined by amino acid residues 74 to 121, which contain the basic NLS sequences (NLS1 and NLS2).³¹ In the VP3 protein of Taiwan-Ilan CAV, 2 stretches of basic amino acids that resemble the NLS sequences have been identified; NLS1 is located at residues 82 to 88, and NLS2 is located at residues 111 to 121.¹⁶ There are putative sites in the VP3 gene of the Taiwan-Ilan strain of CAV despite gene sequence differences from other clones. Thus, its nuclear localization ability would appear to remain intact and unaffected. The different subcellular localizations of the full-length VP3 protein and the mutant fusion protein in CMT and nonneoplastic mammary gland cells are thought to be mediated by the NES and NLS of VP3. In addition, similar results regarding subcellular localization of VP3 protein have been found in the MB157 human breast cancer cell line (data not shown). In the present study, the use of deletion mutants of VP3 protein revealed that the C-terminus of VP3 protein is the major tumor-killing domain for CMT cells.

In addition to the tumor cell–specific nuclear targeting signal, 2 stretches of basic amino acids in the VP3 protein C-terminus have been identified as an NES, which is highly efficient in tumor cells, compared with nonneoplastic cells. The NES (amino acids 97 to 105) operates uniquely in nonneoplastic cells and not in tumor cells through the action of the threonine 108 phosphorylation site adjacent to the NES, which inhibits its action.³¹ The threonine 108 is phosphorylated specifically in tumor but not in nonneoplastic cells.²⁹ The VP3 protein is a nucleocytoplasmic shuttling protein whose localization is mediated by an NES and NLS.³² Both signals are required for cell type–specific localization because VP3 protein fragments containing either the NES or NLS fail to differentially localize in transformed and primary cells.³² In addition, a leucine-rich sequence (amino acids 33 to 46) assists nuclear accumulation of VP3 protein by functioning as a nuclear retention sequence. This combination of signals contributes to the tumor cell–specific nuclear targeting abilities of VP3 protein.^29–32

A CMT was chosen to determine the potential of VP3 protein for tumor gene therapy in veterinary medicine because this type of tumor is the most common form of malignancy in female dogs; sexually intact females have a reported lifetime risk of 2% to 20%.^1,33 Like most malignant tumors, treatment modalities for CMTs are typically ineffective once local recurrence and metastasis occur.³⁴ The study of novel treatments to successfully manage these devastating diseases has been a major issue in the field of veterinary oncology. Gene therapy has been explored in an attempt to prevent development of metastatic lesions, which often causes failure in patient management. The CMT cell lines used in the present study originated from surgically resected mammary gland cancer specimens that were identified as carcinomas. Among all types of CMT in dogs, carcinoma is most frequently identified, with an incidence of 15.4%.³⁵ Forty-one percent to 53% of mammary gland tumors are malignant,³⁶ and more metastasis or death is associated with this type of tumor.³⁷ Positive results of immunohistochemical staining of the intermediate cytokeratin filaments in CMT cells have proven their epithelial origin.³⁸ The CMT cell lines retain the important characteristic of highly abundant expression of secreted frizzled-related protein 2, an antiapoptotic protein. This characteristic, which is present in most primary CMT tissues but not in nonneoplastic mammary gland tissues of dogs, clearly indicates the association of apoptosis dysregulation with mammary gland tumorigenesis.^7,22,39 In addition, secreted frizzled-related protein 2 modulates apoptotic function by both autocrine and paracrine mechanisms within the extracellular matrix of malignant CMT cells and human breast cancer cell lines.²²

To analyze the possibility of improvement in tumorkilling efficiency of the VP3 protein, we constructed a lentiviral vector carrying the VP3 gene to infect CMT cells and nonneoplastic mammary gland epithelial cells. The percentage of transduced CMT cells (in both CMT-1 and CMT-2 cell lines) undergoing apoptosis was increased to > 70% by lentivirus with the EGFP-VP3 fusion gene. The tumor-killing efficiency of the VP3 protein after viral vector infection of cells was significantly higher than that achieved by lipofection (only 30% of transfected CMT cells were apoptotic). The induction of apoptosis was also low to minimal in viral vector-infected, nonneoplastic mammary gland epithelial cells; however, the percentage of cells undergoing apoptosis in the EGFP- and EGFP-VP3–infected nonneoplastic mammary gland cells was not significantly different. The apoptosis detected in nonneoplastic mammary gland epithelial cells may be attributable to high lenti-virus loads (multiplicity of infection of 4).

Incorporation of the VP3 gene into various viral vectors has been explored to evaluate the viability of this protein as a therapeutic agent for use in gene therapies directed against tumors in humans. A recombinant adenovirus carrying the VP3 gene has been used to infect a hepatoma cell line and has resulted in rapid induction of programmed cell death. Furthermore, it has been determined that intratumoral transfer of VP3 recombinant virus results in regression of xenografted human hepatomas⁴⁰ as well as effectively killing human biliary tract cancer cells.²⁵ Results of the present study of VP3 gene expression in CMT cell lines have provided strong evidence that the VP3 protein is also capable of inducing apoptosis in CMT cells, but not in nonneoplastic canine mammary gland cells. Furthermore, the subsequent activation of caspase-9 but not caspase-8 has confirmed that VP3 protein-induced apoptosis occurs primarily through the intrinsic mitochondrial death pathway.²⁰ The finding that antiapoptotic members of the Bcl-2 protein family appear to counteract VP3 protein–induced cell death has further strengthened the proposed relationship between VP3 protein and the mitochondrial apoptosis pathway.²⁰ In the mechanism suggested by Teodoro et al,⁴¹ VP3 protein interferes with the function of the anaphase-promoting complex, which greatly disrupts the process of mitosis and activates the mitochondrial death pathway.⁴¹ Differences in distribution of VP3 protein between neoplastic and nonneoplastic mammary gland cells are also indicative of nuclear trafficking and tumor-specific phosphorylation of VP3 protein at threonine 108.⁴² Nuclear receptor 77 (Nur77) facilitates the translocation of the VP3 protein from nucleus to mitochondria and is required for the protein's toxic actions.²⁰ On the basis of evidence to date, the primary routes of VP3 protein–induced apoptosis appear to mainly involve the intrinsic mitochondrial apoptosis pathway. Given this finding, it seems reasonable to propose that the VP3 gene or VP3 protein may be a potential therapeutic agent for future use in veterinary cancer gene therapies. Furthermore, in the field of veterinary oncology, the use of interleukin-2 as an antitumor treatment against canine tumors (albeit not mammary gland tumors) has been recently reported.^43–45 In addition, for a variety of human tumors, there are synergistic antitumor and enhanced therapeutic effects associated with local interleukin-2 injection in combination with other treatments, such as chemotherapy, radiotherapy, endocrinotherapy, and gene therapy.^46–53 To achieve optimal therapeutic efficacy of VP3 protein in the treatment of CMTs, assessment of the synergistic antitumor effects achieved by VP3 protein with interleukin-2 administration or in combination with other treatment modalities is warranted. On the basis of data obtained in the present study, which have indicated that the VP3 gene of CAV induces apoptosis in malignant CMT cells but not in nonneoplastic mammary gland epithelial cells, it is suggested that the VP3 protein has great potential in the development of antitumor treatments in veterinary medicine because of its selective toxic effects against transformed or cancer cells.