Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

As a well-established cancer chemotherapeutic drug, bleomycin is favored for its effectiveness in treating squamous cell carcinomas, lymphomas, and testicular cancers (1). Although bone marrow suppression is uncommon for bleomycin, pulmonary injury develops in ~ 10% of the patients, with a mortality of 1% (1). Clearly, pulmonary toxicity is the major adverse reaction limiting the widespread use of this drug.

The primary cellular targets of bleomycin-induced pulmonary injury are the cellular components of the alveolar walls, including capillary endothelial cells and type I and type II alveolar epithelial cells (2). Alveolar vascular damage, type I alveolar epithelial cell necrosis, and type II cell hyperplasia and/or metaplasia occur following bleomycin treatment (3, 4). Injury to capillary endothelial cells by bleomycin is often quickly repaired (3, 5). In contrast, severe injury of alveolar epithelial cells typically results in inadequate repair; this failure appears to be a requirement for the development of pulmonary fibrosis (3).

Bleomycin directly attacks DNA of target cells via bleomycin-iron complex in the presence of molecular oxygen (1, 2, 6, 7). It specifically oxidizes the deoxyribose moiety of DNA. Subsequently, the attack either releases the base attached to the sugar, leaving an apurinic/apyrimidinic (AP) site, or excises the base together with part of the sugar ring, resulting in a strand break with a 3'-phosphoglycolate termini, which blocks DNA polymerase to initiate repair (7).

AP endonucleases are DNA repair enzymes that participate in the repair of AP sites and 3' blocking termini (8). First, AP endonucleases are able to cleave at the 5' side of AP sites, leaving a 3'-hydroxyl group and a 5'-deoxyribose phosphate (dRP) terminus at each side of the cut. DNA polymerase  (-pol) will subsequently remove the 5'-dRP and insert an appropriate nucleotide. DNA ligase will eventually seal the gap (9). Furthermore, the AP endonucleases are capable of hydrolyzing 3'-blocking fragments, as occurs with ionizing radiation and bleomycin, and creating 3'-hydroxyl termini, which permit -pol to initiate repair (8, 9, 11). Thus, the two DNA lesions associated with bleomycin toxicity, AP sites and strand breaks with 3'-blocking fragments, are the primary targets for the repair functions of the hydrolytic AP endonucleases. In fact, some previous studies have shown AP endonucleases facilitate repair of DNA damage caused by bleomycin (12). Therefore, we hypothesize that overexpression of AP endonucleases in lung cells will significantly reduce the toxicity associated with bleomycin-induced DNA damage.

The major AP endonuclease of Saccharomyces cerevisiae, APN1, plays an important role in maintaining genetic stability (8). APN1-deficient strains not only have much higher sensitivities to alkylating and oxidating agents, but also demonstrate a dramatic increase in the spontaneous mutation rate (15, 16). With a proven ability to remove 3'-blocking fragments in bleomycin-treated bacterial DNA (9), without any mammalian counterpart found so far (17), APN1 is an excellent AP endonuclease to test for its ability to protect lung cells from bleomycin toxicity. Thus, we introduced yeast APN1 into A549 cells, a cell line derived from human alveolar epithelium. The expression of APN1 was determined at mRNA, protein, and enzyme activity levels. The protective effect of APN1 against bleomycin was demonstrated in APN1 expressing A549 population and clones.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture

The human alveolar epithelial cell line A549 (ATCC #CCL-185; American Type Culture Collection, Manassas, VA) was maintained in Dulbecco's modified Eagle's medium (DMEM; BioWhittaker, Walkersville, MA) with 10% fetal bovine serum (Hyclone, Logan, UT), containing 1,000 U/ml penicillin and 100 U/ml streptomycin (Biosource International, Rockville, MD) at 37°C in 5% CO₂.

Retroviral Vector Construction

The coding region of the yeast APN1 was amplified by polymerase chain reaction (PCR) using yeast genomic DNA as template. The 5' and 3' primers containing restriction sites are 5'-GGC CGTCGACATGCCTTCGACACCTAGCTTTG-3' and 5'-GG CCGAATTCTTATTCTTTCTTAGTCTTCCTC-3', respectively. PCR was performed in 10 mM Tris-HCL, 1.5 mM MgCl, 50 mM KCl, pH 8.3, with Tfl DNA polymerase (Promega, Madison, WI). Each of the 30 cycles of PCR was performed at 95°C for 30 s, 55°C for 1 min, and 72°C for 2 min. PCR products were purified and sequenced to verify that no mutations were introduced into the APN1 coding region during PCR amplification. After several subclonings, an APN1 fragment was obtained with both sides flanked by the EcoRI restriction sites.

An improved murine stem cell virus (MSCV)-based bicistronic retroviral vector, MIEG3, expressing an enhanced green fluorescence protein (EGFP), was constructed as previously described (18). Both APN1 cDNA fragment and linearized MIEG3 vector were obtained by restriction digestion with EcoRI (New England Biolabs, Beverly, MA) and gel-purified using GENECLEAN II kit (BIO 101, Vista, CA). APN1 cDNA was ligated into linearized MIEG3 using T4 DNA ligase (New England Biolabs). The ligation product was introduced into DH5 competent cells (Life Technologies, Grand Island, NY). Plasmid DNA of APN1-containing MIEG3 (MIEG3-APN1) was purified from the confirmed colonies using Qiagen Plasmid Maxi Kit (Qiagen, Chatsworth, CA).

Eight micrograms of purified plasmid DNA of MIEG3 or MIEG3-APN1 were mixed with LipofectAMINE transfection reagent (Life Technologies). Phoenix-Ampho cells (American Type Culture Collection) were transfected by MIEG3 or MIEG3-APN1 viral vector using a manufacture-provided protocol. Viral supernatant was collected 48-72 h after transfection. The retroviral titer was determined by fluorescent-activated cell sorter (FACS) analysis using a previously reported method (19). The titers of MIEG3 and MIEG3-APN1 viral supernatant were 2 × 10⁵ and 1 × 10⁵ colony-forming units (cfu)/ml, respectively.

Retroviral Transduction

A549 cells (1 × 10⁵) were plated in a 6-well culture dish (Falcon, Becton Dickinson Labware, Franklin Lakes, NJ) and cultured overnight. The following day, 2 ml of viral supernatant of either MIEG3 or MIEG3-APN1 were added to cell culture. After 4 h of incubation at 37°C, 2 ml of 10% DMEM were added and the infection was continued overnight. Two days after infection, transduced cells were sorted using FACStar^Plus (Becton Dickinson, San Jose, CA) as EGFP-expressing cells. MIEG3- and MIEG3-APN1-transduced A549 cell populations were cloned using the method of limiting dilution, in which cells were diluted to ~ 1 cell/ well on a 96-well plate. Clones were expanded when a single colony was observed in each well. EGFP expression levels in transduced cell populations and clones were determined by measuring relative intensities (represented by the values of geometric means of EGFP) of fluorescence of 10,000 cells. Data of flow cytometry were analyzed by FACStar^Plus-installed software, Cell Quest 3.1 (Becton Dickinson). Four MIEG3-APN1-transduced clones representing various levels of EGFP expression, and 1 MIEG3-transduced clone expressing a high level of EGFP, were selected for clonal analysis. Selected clones were also visualized by fluorescent microscopy using Olympus BX60 System Microscope (Olympus, Tokyo, Japan)

Antibody Production

The APN1 protein was produced by overexpressing APN1 in a pGEX-glutathione-S-transferase (GST) Escherischia coli system (Pharmacia, Upsula, Sweden) as previously described for the production of human AP endonuclease 1 (Ape1) protein (20). The purified APN1 protein served as antigen to immunize rabbits and generate APN1 antisera (HTI, Ramona, CA). For purifying crude antisera, APN1 protein was resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (10%) gel and transferred to polyvinylidene fluoride microporous membrane (Immobilon-P Transfer membrane; Millipore, Bedford, MA). The membrane strip bound with APN1 protein was incubated with the crude antisera overnight at 4°C. After washing, the antibody was eluted off the strip with ImmunoPure Ag/Ab elution buffer (Pierce, Rockford, IL). The specificity of purified APN1 antibody was confirmed using recombinant APN1 protein and lysates of cells overexpressing APN1 in Western blot analysis.

Northern Blot Analysis

Total RNA was isolated from transduced cells (~ 5 × 10⁶) by RNeasy Mini Kit (Qiagene, Valencia, CA). A quantity of 6 µg of total RNA was resolved on formaldehyde denaturing agarose gel. Subsequently, separated RNA was transferred to Hybond-N nylon membrane (Amersham Pharmacia Biotech, Buckinghamshire, UK). APN1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) DNA probes were made by random oligonucleotide priming using Megaprime DNA Labeling System (Amersham Pharmacia Biotech, Piscataway, NJ). The activity of probes was > 1 × 10⁸ dpm/µg. Hybridization was performed overnight at 60°C. After low, medium, and high stringency washing, exposure time for autoradiagraphy was typically 2-3 d. Densitometry was performed on the hybridized membranes using Kodak Digital Science 1D software (Kodak, New Haven, CT) and the intensities of APN1 bands were normalized to the intensities of GAPDH bands.

Western Blot Analysis

A quantity of 10 µg of cell lysate was resolved on SDS-PAGE (10%) gel using Mini-PROTEAN II Cell (Bio-Rad Laboratories, Hercules, CA). Gels were then electroblotted to Immobilon-P Transfer membrane (Millipore), in Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad Laboratories). Using BM Chemiluminescence Western Blotting Kit (Roche Molecular Biochemicals, Indianapolis, IN), blots were developed using antiyeast APN1 or antihuman Ape1 antibodies (Novus Biologicals, Littleton, CO). Densitometry was performed on the blots using Kodak Digital Science 1D software (Kodak) and the intensities of APN1 bands were normalized to the intensities of Ape1 bands.

AP Endonuclease Activity Assay

In this assay, the substrate was a 26-mer oligonucleotide containing an abasic analog at position 15, where the normal base was substituted by a tetrahydrofuran (THF). The substrate sequence was 5'-AATTCACCGGTACCFTCTAGAATTCG-3', where F = THF-containing deoxynucleotide. As previously described (21), THF-containing oligonucleotide was labeled at the 5'-end with -[³²P]ATP in a polynucleotide kinase-catalyzed reaction. The labeled oligonucleotide was hybridized to its complementary strand by heating the mixture to 95°C, then gradually cooling to room temperature.

Cell-free extracts were obtained from ~ 3 × 10⁶ cells of transduced A549 cells. Each reaction mixture contained cell-free extract expressed as 0.0625, 0.125, 0.25, 0.5, 1, or 2 µg of protein standard, 2 pMol/L labeled THF-containing oligonucleotide, 50 mM Hepes (pH 7.5), 50 mM KCl, 20 mM ethylenediamine tetraacetic acid (EDTA), 1 µg/ml bovine serum albumin, and 0.05% Triton X-100. To specifically assay APN1 activity as opposed to human Ape1 activity, this assay was conducted in the absence of Mg²⁺ (17, 22). Purified APN1 and Ape1 protein were used as controls. The reaction was conducted at 37°C for 15 min. Reaction products were separated by 20% PAGE containing 7 M urea. The autoradiography derived directly from PAGE gel was analyzed using Kodak Digital Science 1D software (Kodak). The percentage of cleavage was calculated as the density of the cleaved band (14mer) divided by the sum of the density of the cleaved band and that of the uncleaved band (26mer).

To determine whether bleomycin has a direct toxic effect on APN1 activity, cell-free extracts were obtained from three clones, APN1-1, APN1-6, and APN1-12, aliquoted, and incubated with 0, 1, 4, 20, 100, and 1,000 µg/ml bleomycin (1.8 U/mg; Hande Tech Development, Houston, TX) for 1 h at 37°C. AP endonuclease activity assay was performed using 1 µg of cell-free extract from each bleomycin treatment.

Single Cell Gel Electrophoresis/Comet Assay

Bleomycin sulfate solutions in 10% DMEM were freshly prepared for each experiment. Transduced A549 cells were seeded onto 6-well cell culture dishes (Falcon Becton Dickinson Labware) at a density of 5 × 10⁵ cells/well. The following day, cells were incubated with bleomycin at concentrations of 0, 20, 50, 100, and 200 µg/ml for 1 h at 37°C. CometAssay kit (Trevigen, Gaithersburg, MD) was used to assess for DNA damage. Following a manufacturer recommended protocol, cells were first detached by scraping and resuspended in Ca⁺⁺ and Mg⁺⁺-free phosphate buffered saline (Trevigen) at a concentration of 3 × 10⁵ cells/ml. The cell suspension was mixed with liquefied agarose at a 1:10 ratio. Immediately, a small aliquot of the mixture was transferred on to a provided slide. After solidification and cell-lysis at 4°C, all slides were treated with alkali solution (0.3 M NaOH, 1 mM EDTA) for 20 min to unwind the double-stranded DNA. Subsequently, slides were placed in an electrophoresis apparatus and a voltage of 1 volt/cm distance between the two electrodes was applied for 10 min to show the comet tails. After staining with provided fluorescent staining solution, samples were examined and photographed by fluorescent microscopy using Olympus BX60 System Microscope (Olympus) and Paxit software (MIS, Franklin Park, IL). During data analysis, tail length was defined as the distance between the leading edge of the nucleus and the end of the tail. In each experiment, ~ 80 determinations were made for each sample using Adobe Photoshop software (Adobe Systems Incorporated, San Jose, CA). Data represent three separate experiments.

Colony-Forming Assay

A549 cells were seeded on 6-well cell culture dishes (Falcon Becton Dickinson Labware) at a density of 2 × 10⁵ cells/well and cultured overnight. Transduced A549 populations were incubated with 0, 0.8, 4, 20, and 100 µg/ml bleomycin for 1 h at 37°C. Transduced A549 clones were incubated with 0, 4, and 20 µg/ml bleomycin in subsequent experiments. After washing and diluting, cells were replated in triplicate into 100-mm cell culture dishes (Falcon, Becton Dickinson Labware). After 14 d, cell colonies were stained with 1.5% methylene blue and counted. The number of colonies surviving in the presence of bleomycin was expressed as % survival with 100% survival defined as the number of colonies surviving in the absence of bleomycin. Data represent three separate experiments.

Statistics

Pearson Product Moment Correlation was adopted to analyze the relationship between the intensities of Northern/Western blots and the intensities of fluorescence obtained from FACS analysis among selected clones. ANOVA was employed to analyze the differences among the APN1 expression levels of selected clones and among the tail lengths and survival rates of transduced A549 cells. Significance was accepted as P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of EGFP

Retroviral infection by itself did not alter the rate of replication of A549 cells. In multiple FACS analyses, 78.5 ± 4.8% and 71.5 ± 3.2% of MIEG3- and MIEG3-APN1- transduced A549 populations were EGFP-positive, respectively. All clones derived from transduced cells consist of > 90% EGFP-positive cells. Among MIEG3-APN1- transduced clones, four clones identified as APN1-1, APN1-6, APN1-8, and APN1-12 were selected to represent different levels of EGFP expression. APN1-1 and APN1-12 showed the highest levels of EGFP expression with mean fluorescence intensities of 233.7 ± 68.1 and 267.0 ± 74.7, respectively. Next, APN1-6 had a mean fluorescence intensity of 144.0 ± 13.4, while APN1-8 exhibited the lowest EGFP expression with mean fluorescence intensity of 61.7 ± 9.0 (P < 0.05, comparisons of APN1-8 with APN1-1 and APN1-12). Among control MIEG3-transduced clones, MIEG3-3, the clone with the highest EGFP expression with mean fluorescence intensity of 348.5 ± 65.1, was selected for clonal analysis. The mean fluorescence intensity of untransduced A549 cells was 8.3 ± 5.2. EGFP expression in selected clones was also verified by fluorescent microscopy. The observed intensities of fluorescence were consistent with FACS analysis (data not shown). The persistence of EGFP expression in both cell populations and clones was demonstrated for up to 2 mo using FACS analysis. There was no significant decrease of fluorescence intensity during the period (data not shown).

Expression of APN1

APN1 expression in transduced A549 populations was determined for both APN1 mRNA and protein. While the MIEG3-APN1-transduced population revealed APN1 expression, the MIEG3-transduced population did not show any detectable evidence of APN1 (data not shown). Clones APN1-1, APN1-6, APN1-8, APN1-12, and MIEG3-3 were also assayed for APN1 mRNA (Figures 1A and 1C) and protein expression (Figures 1B and 1D). GAPDH and human Ape1 were used as internal controls of Northern and Western analyses, respectively (Figures 1A and 1B). Clone MIEG3-3 demonstrated neither mRNA nor protein expression of APN1 (Figures 1A and 1B). In Northern blot, APN1-12 showed the highest APN1 mRNA level followed by APN1-1 and APN1-6, while APN1-8 expressed the lowest level of APN1 mRNA (Figure 1C). In Western blot, the purified antibody recognized the APN1 protein at a molecular weight of 40 kD as expected (Figure 1B). Clones APN1-1 and APN1-12 showed the highest APN1 protein level followed by APN1-6, while APN1-8 expressed the lowest level of APN1 protein. Both mRNA and protein expression levels of APN1 were positively correlated to the expression levels of EGFP represented by fluorescence intensities from FACS analysis with correlation coefficients of 0.88 and 0.83, respectively (Figures 1E and 1F). The persistence of APN1 expression in selected clones was demonstrated for up to 2 mo by Western blot. There was no significant decrease in protein expression among the clones with the exception of APN1-8, which showed ~ 23% reduction at 2 mo (data not shown).

View larger version (45K): [in this window] [in a new window]   Figure 1.   Expression of APN1 in transduced A549 clones. The expression of APN1 was determined by Northern blot for mRNA expression (A, C, and E) and by Western blot for protein expression (B, D, and F ). Each clone is represented as: (1) MIEG3-3, (2) APN1-1, (3) APN1-6, (4) APN10-8, and (5) APN1-12. Representative experiments (n = 3) are shown for: (A) Northern blot (hybridized with probes for APN1 and GAPDH) and (B) Western blot (developed with polyclonal antibodies for yeast APN1 and human Ape1). C and D represent densitometry data expressed as means ± SD (n = 3). Significant differences exist among the expression levels of clones, P < 0.05. E and F represent the intensities of either Northern or Western blot versus the intensities of EGFP fluorescence (r = 0.88 and 0.83, respectively).

Activity of APN1

APN1 activity in transduced A549 cells was specifically determined using the THF-oligonucleotide in reaction buffer that was Mg²⁺-deficient and also contained 20 mM EDTA (17, 22). As a positive control, purified APN1 protein (23 ng) cleaved nearly 100% of substrate (Figure 2A). As a control to determine the relative contribution of the endogenous Ape1 to the cleavage of THF-oligonucleotide, human Ape1 was used in a parallel reaction. Ape1 did not cleave a significant amount of THF-containing oligonucleotide even using concentration as high as 367 ng under the same conditions (Figure 2A). Cell-free extract of MIEG3-APN1-transduced A549 population (equivalent to 0.5 µg protein standard) degraded nearly 100% THF-containing substrate (data not shown). In contrast, MIEG3-transduced A549 population revealed no AP endonuclease activity even using a 10-fold higher concentration of the cell-free extract (data not shown). To quantify the APN1 activity in selected clones, different amounts of cell-free extract from each clone were applied to reveal a spectrum of activity for each clone (Figures 2A and 2B). Clone APN1-12 exhibited the highest level of APN1 activity, i.e., ~ 0.125 µg cell-free extract was sufficient to cleave 50% of 2 pMol/L THF-containing oligonucleotide (Figure 2B). Clones APN1-1 and APN1-6 required 0.5 µg to reveal a 50% cleavage. For clone APN1-8, APN1 activity was the lowest, requiring 1 µg of cell-free extract to yield 50% incision. As expected, the MIEG3-3 clone did not produce any detectable cleavage of substrate (Figure 2A).

View larger version (30K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 2.   AP endonuclease activity of transduced A549 clones. (A) Different amounts of cell-free extract obtained from clones, APN1-1, APN1-6, APN1-8, and APN1-12, reacted with 2 pMol/L -[32P]dATP-labeled tetrahydrofuran-containing 26-mer oligonucleotide and produced a 14-mer oligonucleotide. The numbers on the top of each lane represent the micrograms of cell-free extract applied in each reaction. The same reaction was also performed using 2 µg of cell-free extract from clone MIEG3-3 (a), 23 ng purified APN1 protein (b), and 367 ng purified human Ape1 protein (c). (B) Quantification of AP endonuclease activity of each clone. % Cleavage = density of 14-mer/(density of 14-mer + density of 26-mer) × 100%.

The toxicity of bleomycin on APN1 activity was also evaluated. Cells expressing APN1 were incubated with various concentrations of bleomycin at 0, 1, 4, 20, 100, and 1,000 µg/ml. One microgram of cell-free extract of APN1-expressing clones produced similar amounts of cleavage of THF-containing oligonucleotide in AP endonuclease activity assay (data not shown), indicating there was no evidence that bleomycin demonstrated any direct toxicity to APN1.

APN1 and Bleomycin-Induced DNA Damage

The single cell gel electrophoresis or Comet assay was employed to detect AP sites and DNA strand breaks induced by bleomycin in transduced A549 cell populations. Bleomycin at 0, 20, 50, 100, and 200 µg/ml caused a concentration-dependent increase in DNA damage in both MIEG3- and MIEG3-APN1-transduced cells (Figure 3). The MIEG3-transduced cells in the presence of 20, 50, and 100 µg/ml bleomycin showed comet tail lengths of 30.5 ± 25.9, 55.4 ± 25.4, and 57.2 ± 33.0 µm, respectively, whereas the MIEG3- APN1-transduced population treated with the same concentrations of bleomycin showed significantly reduced tail lengths of 10.8 ± 13.1, 14.8 ± 14.4, and 16.5 ± 19.0 µm, respectively (Figure 3, P < 0.001, all comparisons). The DNA damage was similar in both cell populations at 200 µg/ml bleomycin.

View larger version (23K): [in this window] [in a new window]   Figure 3.   Protective effect of APN1 against bleomycin-induced DNA damage. MIEG3-transduced and MIEG3-APN1-transduced A549 populations were incubated with bleomycin and DNA damage was quantified by the Comet assay. Data represent three separate experiments. In each experiment, ~ 80 cells of each sample were measured for tail length, which was defined as the distance between the leading edge of the nucleus and the end of the tail. The tail lengths of APN1-transduced cells were significantly shorter than those of MIEG3-transduced cells at 20, 50, and 100 µg bleomycin /ml (*P < 0.001).

APN1 and Bleomycin Cytotoxicity

A series of colony-forming assays was performed to examine the protective effect of APN1 from bleomycin toxicity in transduced A549 cells. Bleomycin caused a decrease in cell survival in a concentration-dependent manner in both cell populations and clones. The MIEG3-transduced A549 cell population in the presence of bleomycin at 0.8, 4, 20, and 100 µg/ml had % cell survival of 97.3 ± 19.1, 46.2 ± 2.4, 17.1 ± 11.6, and 2.4 ± 3.3, respectively (Figure 4A). In contrast, the population of A549 cells expressing APN1, in the presence of the same concentrations of bleomycin, had % cell survival of 96.2 ± 19.1, 82.9 ± 21.5, 48.0 ± 21.4, and 4.0 ± 3.1, respectively (P < 0.01, for the differences between the two cell populations at the presence of 4 or 20 µg/ml bleomycin, Figure 4A).

View larger version (29K): [in this window] [in a new window]   View larger version (37K): [in this window] [in a new window]   Figure 4.   Protective effect of APN1 against bleomycin cytotoxicity. (A) The effect of APN1 on survival of MIEG3- and MIEG3-APN1-transduced A549 populations was measured using a colony-forming assay. (B) The effect of APN1 was also shown for survival of transduced clones, MIEG3-3, APN1-1, APN1-6, APN1-8, and APN1-12. The data represent means ± SD (n = 3). The survival of APN1-transduced cells was significantly higher than that of MIEG3-transduced cells (*P < 0.05).

On the basis of these results, 4 or 20 µg/ml bleomycin was adopted in subsequent experiments using transduced A549 cell clones. When 4 µg/ml bleomycin was incubated with the cells, clone MIEG3-3 exhibited 43.4 ± 12.4% survival, whereas APN1-1, APN1-6, APN1-8, and APN1-12 displayed survival of 60.1 ± 10.4%, 83.5 ± 9.0%, 72.3 ± 19.3%, and 58.1 ± 8.4%, respectively (Figure 4B). Among these clones, APN1-1, APN1-6, and APN1-8 were significantly protected from bleomycin toxicity (P < 0.05). A similar result occurred using 20 µg/ml bleomycin. However, at this concentration, APN1-1 and APN1-12 did not achieve significant protection from bleomycin.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Bleomycin is a cancer chemotherapeutic drug that causes damage by inducing AP sites and strand breaks in the DNA of target cells (6, 7). APN1 is a yeast DNA repair protein with both AP endonuclease and 3'-diesterase activities, and with the potential to reverse DNA damage by drugs such as bleomycin (8). Expression of yeast APN1 in A549 lung epithelial cells was verified by analysis of APN1 mRNA, protein, and activity levels. A549 cells expressing APN1 were resistant to bleomycin toxicity, suggesting that strategies to enhance the activities of specific DNA repair proteins in lung cells might reduce bleomycin-induced DNA damage and cell injury.

On the basis of the expression of an EGFP gene, transduction with the retroviral vector MIEG3 allowed separation of transduced cells from untransduced cells using either FACS analysis or fluorescent microscopy (18). In the current study, MIEG3 yielded high efficiency and stability of both target gene and reporter gene expression in A549 cells. The levels of EGFP were positively correlated with expression levels of APN1; thus EGFP served as a good marker for APN1 expression in this system.

A previous investigation has shown that expression of APN1 restores the resistance to oxidants and alkylating agents in repair-deficient E. coli and removes 3'-blocking fragments in bleomycin-treated bacterial DNA (8). In our study, the APN1-expressing A549 cells demonstrated less DNA damage as measured by the Comet assay, which detects both bleomycin-induced AP sites and strand breaks. The results of the clonogenic survival assays performed on both cell populations and clones revealed that APN1 also corrects bleomycin-induced cytotoxicity to the lung cells. Bleomycin may also cause pulmonary toxicity by upregulating DNA damage-inducible proteins or directly inducing lipid peroxidation (23, 24); however, enhanced repair of DNA damage is likely the primary reason for the improved survival of lung cells in the current system.

There are two major AP endonuclease families, endonuclease IV and exonuclease III (17). Endonuclease IV family is represented by E. coli endonuclease IV and yeast APN1. Exonuclease III family is represented by E. coli exonuclease III and human Ape1. The AP endonuclease activity and 3'-repair diesterase activity (targeting 3'-blocking fragments at strand breaks) differ between the families. For example, the 3'-diesterase activity of E. coli endonuclease IV is 5-fold more active in repairing bleomycin-induced damage than that of exonuclease III (25). This may explain the phenomenon that endonuclease IV (homologous to APN1) mutants of E. coli are more sensitive to bleomycin than exonuclease III (homologous to Ape1) mutants (26). In addition, the AP endonuclease and 3'-diesterase activity of endonuclease IV (homologous to APN1) are comparable (14), whereas the 3'-diesterase activity of Ape1 is up to 200-fold lower than its AP endonuclease activity (9, 14, 27). Since bleomcyin-induced AP sites and strand breaks with 3'-blocking fragments are formed in comparable quantities (7), APN1 might be a better candidate to repair bleomycin-induced DNA lesions.

Although most APN1 expressing clones were protected from bleomycin toxicity, the level of APN1 expression or activity did not correlate with the level of protection. Clones APN1-1 and APN1-12 with the relatively higher expression of APN1 showed relatively lower levels of protection. It is possible that APN1 activity does not represent the rate-limiting step in correcting the bleomycin-mediated DNA damage. High levels of AP endonuclease activity may exceed the repair capacity of downstream enzymes, i.e., -pol and DNA ligase. The downstream enzymes might be rate-limiting instead of AP endonuclease in this DNA repair cascade. In fact, there is evidence demonstrating -pol is the rate-limiting step of base excision repair (28). Also, yeast APN1 might not interact as efficiently with human -pol and DNA ligase (29). Moreover, overexpressing APN1 might introduce genetic instability (30). Thus, solely increasing AP endonuclease activity might not necessarily produce enhanced survival. On the other hand, clone APN1-8 with the lowest level of APN1 exhibited an intermediate level of protection, whereas clone APN1-6 with the intermediate APN1 expression showed the highest protection. It is possible that there is an optimal intracellular concentration of endonuclease activity that will provide maximum protection from bleomycin toxicity and not adversely affect cell growth.

Most drug-induced injury of the lung represents both direct toxicity from the drug and secondary toxicity associated with a vigorous host response (6). Bleomycin pulmonary toxicity in vivo is more complicated than in vitro studies. It is associated with an intense inflammatory or immune reaction in the lung (23, 31). In addition, other factors, such as the level of bleomycin hydrolase, have been shown to influence bleomycin toxicity (23, 32). The relationship between bleomycin-induced acute cell injury and inflammation is unclear. For example, it is not known if the inflammatory or immune response would be ameliorated in bleomycin pulmonary toxicity if the DNA damage to alveolar epithelial cells were reduced.

Bleomycin can be directly toxic to certain cellular enzymes, such as O⁶-alkylguanine-DNA alkyltransferase, an important enzyme involved in direct reversal repair of DNA damage (33). In our study, however, there was no inhibition of APN1 activity due to bleomycin treatment. This suggests that protection by APN1 will persist in bleomycin-rich environments.