Apurinic/Apyrimidinic Endonuclease (APE/REF-1) Haploinsufficient Mice Display Tissue-specific Differences in DNA Polymerase (cid:1) -Dependent Base Excision Repair*

Apurinic/apyrimidinic (AP) endonuclease (APE) is a multifunctional protein possessing both DNA repair and redox regulatory activities. In base excision repair (BER), APE is responsible for processing spontaneous, chemical, or monofunctional DNA glycosylase-initiated AP sites via its 5 (cid:1) -endonuclease activity and 3 (cid:1) -“end-trimming” activity when processing residues produced as a consequence of bifunctional DNA glycosylases. In this study, we have fully characterized a mammalian model of APE haploinsufficiency by using a mouse containing a heterozygous gene-targeted deletion of the APE gene ( Apex (cid:2) / (cid:3) ). Our data indicate that Apex (cid:2) / (cid:3) mice are indeed APE-haploinsufficient, as exhibited by a 40– 50% reduction ( p < 0.05) in APE mRNA, protein, and 5 (cid:1) -endonuclease activity in all tissues studied. Based on gene dosage, we expected to see a concomitant reduction

tegrity and in the regulation of gene expression. After the initial discovery in Escherichia coli (1), APE was purified from calf thymus DNA and extensively characterized as an endonuclease that cleaves the backbone of double-stranded DNA containing AP sites (2,3). APE homologues were subsequently identified and characterized in many organisms, including yeast as APN1 (4), mice as Apex (5,6), and humans as HAP1 (7). In addition to its major 5Ј-endonuclease activity, APE also expresses minor 3Ј-phosphodiesterase, 3Ј-phosphatase, and 3Ј 3 5Ј-exonuclease activities (8), the biological significance of which is controversial (9). Independent of its discovery as a DNA repair protein, APE was also characterized as REF-1, for redox factor-1, a redox activator of cellular transcription factors (10 -12). Although the molecular detail of APE redox activity is still unclear (13), the discovery of APE as a regulator of transcriptional activity may underscore the importance of its involvement in cellular stress-response pathways.
APE is the primary enzyme responsible for recognition and incision of non-coding AP sites in DNA arising as a consequence of spontaneous, chemical, or DNA glycosylase-mediated hydrolysis of the N-glycosyl bond initiated by the base excision repair (BER) pathway. These lesions are particularly common, arising at the rate of ϳ50,000 -200,000 AP sites per cell per day under normal physiological conditions (14,15). Unrepaired AP sites may threaten genomic stability by serving as blocks to DNA replication (16), by stalling RNA polymerase II during transcription (17), or by promoting topoisomerase II-mediated double strand breaks (18). Of the repair pathways available to a cell (19,20), BER is the main pathway responsible for repairing AP sites in DNA. Initiation of BER is made possible by recognition of a damaged base by either a monofunctional or bifunctional DNA glycosylase, in addition to AP site recognition by APE. In monofunctional glycosylase-initiated BER (MFG-BER), a damaged or improper base is recognized and removed by enzymatic hydrolysis of the N-glycosyl bond resulting in the formation of an AP site. AP sites serve as a substrate for APE, which incises the DNA backbone immediately 5Ј to the AP site via its 5Ј-endonuclease activity, producing a single strand break with a normal 3Ј-hydroxyl group and an abnormal 5Јdeoxyribose 5-phosphate (dRP) residue (21). DNA polymerase ␤ (␤-pol) inserts a new base followed by the coupled ␤-pol-mediated excision of the abnormal 5Ј-dRP, removal of which has been shown to be rate-limiting (22). In bifunctional glycosylaseinitiated BER, the damaged base is recognized and removed by a damage-specific DNA glycosylase followed by incision of the DNA backbone by the associated AP lyase activity, yielding a normal 5Ј-terminal deoxynucleoside-5Ј-phosphate residue and an abnormal 3Ј-terminal ␣,␤-unsaturated aldehyde residue that must be processed prior to repair completion (21). The removal of the abnormal 3Ј-blocking lesion by APE 3Ј-phosphodiesterase activity is believed to be rate-limiting (23); however, this rate-limiting role is controversial (24 -26). After APE recognition of AP sites, BER may proceed by one of two pathways: (i) ␤-pol-mediated single nucleotide insertion, similar to MFG-BER, or (ii) Ͼ1 nucleotide strand-displacement synthesis, required to process modified (i.e. reduced, oxidized) AP sites and involves components of the DNA replication machinery (27). Completion of BER requires the nick sealing activity of DNA ligase complexes (28).
A review of structural studies has proposed a model of BER requiring highly intimate, yet transient protein-protein interactions among BER enzymes to ensure proper damage repair, with APE being at the center of activity (29). For example, x-ray cross-complementing protein (XRCC1), a protein with no known enzymatic activity, functions as both a scaffold protein and modulator of BER via functional and physical interaction with APE, bridging the incision and nick-sealing steps of BER (30). APE has been shown also to interact with ␤-pol, recruiting it to the incised AP site and enhancing its rate-limiting dRPase activity (31); this activity is also believed to be involved in the repair of oxidative base lesions (32). The functional importance of ␤-pol in oxidative damage repair may be due to the interaction of APE with bifunctional DNA glycosylases responsible for recognizing and removing these lesions. For example, APE has been shown to stimulate 8-oxoguanine DNA glycosylase (OGG1) turnover and enhance its glycosylase activity while minimizing its associated AP lyase activity (33,34), with XRCC1 accelerating this process (35), thus eliminating a potentially rate-limiting step of APE and potentiating MFG-BER. Similar results have also been obtained with other bifunctional DNA glycosylases such as endonuclease III (hNTH1), responsible for recognition and removal of ring-saturated pyrimidines (36). These studies, in addition to a recent mathematical model of BER throughput (37), suggest a preference for ␤-pol-mediated MFG-BER in vivo.
The objective of the research described in this study is to characterize in more detail the phenotype of an APE heterozygous knockout (Apex ϩ/Ϫ ) mouse reported previously (38), and to address the effect of APE haploinsufficiency on BER capacity. It is important to note that homozygous deletion of the APE gene (Apex Ϫ/Ϫ ) is embryonic lethal, but heterozygous mice survive and are fertile (38 -40). Because the sequences encoding both the DNA repair and redox regulatory activities of APE are disrupted in Apex Ϫ/Ϫ mice, it is unclear whether one or both of these activities are necessary for embryogenesis. Although the role of APE in the redox activation of p53 and other cellular transcription factors suggests its importance in signal transduction pathways, the embryonic lethality observed for APE and three other BER genes (␤-pol, DNA ligase I, and XRCC1) suggests a critical role for BER during embryogenesis (41). Recent studies have implicated a role for p53 in the regulation of the BER pathway (42); therefore, it is inviting to suggest that APE repair activity in general, and perhaps APE redox regulatory activity in particular, is the reason for the embryonic lethality observed when APE is deficient. Here we present evidence that half the gene dosage of APE results in tissuespecific alterations in MFG-BER and suggest that APE redox activity, as opposed to repair per se, potentiates this phenotypic effect. Data obtained from Apex ϩ/Ϫ mice may have relevant transnational implications because APE variants have been identified in the human population (43), and variants in DNA repair have been associated with increased risk for disease such as cancer (44,45).

Animals
The experiments were performed on young (3-6 months) male C57BL/6-specific pathogen-free mice in accordance with the National Institutes of Health guidelines for the use and care of laboratory animals. The Wayne State University Animal Investigation Committee approved the animal protocol. Mice were maintained on a 12-h light/ dark cycle and fed a standard lab diet and water ad libitum. The mice were sacrificed by cervical dislocation, and the organs to be studied were flash-frozen in liquid nitrogen and stored at Ϫ70°C for later enzyme studies and Western blot analyses. Tissues for total RNA isolation and RT-PCR analysis were immediately homogenized in TRIzol® Reagent (Invitrogen) according to the manufacturer's protocol.
The APE heterozygous knockout (Apex ϩ/Ϫ ) mice were developed in Friedberg's laboratory as described previously (38). In order to obtain the requisite number of animals for the study, the Apex ϩ/Ϫ mice were inbred and maintained on a 12-h light/dark cycle and fed a standard lab diet and water ad libitum. The mice appeared normal, were fertile, and there was no retardation in food intake, weight gain, or growth rate; however, it was observed that pups were not produced in expected Mendelian ratios, with heterozygote births predominating, similar to previous observations (39). A three-primer PCR strategy was employed to genotype the animals generated by Apex ϩ/Ϫ intercrosses as described previously (38).

mRNA Expression Analysis
The level of APE mRNA was measured using RT-PCR analysis using AccessQuick TM RT-PCR System (Promega, Madison, WI) according to the manufacturer's protocol. Total cellular RNA was isolated from selected tissues using TRIzol® Reagent (Invitrogen), and RNA concentration was determined by measuring UV absorption at 260/280 nm. Oligonucleotide primers specific for the mouse Apex gene (forward, 5Ј-CT-CAAGATATGCTCCTGGAA-3Ј; reverse, 5Ј-GGTATTCCAGTCTTACC-AGA-3Ј) were designed using GeneFisher Interactive Primer Design Tool (Bielefeld, Germany) and synthesized by Sigma-Genosys (The Woodlands, TX). RT-PCR thermal cycling conditions were as follows: 48°C for 45 min, 1 cycle; 95°C for 2 min, 1 cycle; 95°C for 1 min, 52°C for 1 min, 70°C for 2 min, 22 cycles; and 70°C for 5 min, 1 cycle. The 350-bp RT-PCR product was stained with ethidium bromide and analyzed on a 2% agarose gel. Intensity of the bands was detected and quantified using a ChemiImager TM system (AlphaInnotech, San Leandro, CA) and expressed as the integrated density value (I.D.V) per g of RNA used per reaction. Data were normalized based on the amount of ␤-actin present in each sample.

Isolation of Crude Nuclear Extract
Isolation of crude nuclear extract was accomplished using Sigma CelLytic TM NuCLEAR TM extraction kit (Sigma). All samples and tubes were handled and chilled on ice, and all solutions were made fresh according to the manufacturer's protocol. Resultant nuclear extracts were dialyzed against 1 liter of dialysis buffer (20 mM Tris-HCl, pH 8.0; 100 mM KCl; 10 mM NaS 2 O 5 ; 0.1 mM DTT; 0.1 mM phenylmethylsulfonyl fluoride; 1 g/ml pepstatin A) for 4 -6 h at 4°C using Slide-A-Lyzer® minidialysis units (Pierce). Protein concentrations were determined using Protein Assay Kit I (Bio-Rad).

Protein Expression Analysis
Western analysis was performed using 200 g of crude nuclear extract isolated from selected tissues according to standard protocol. Protein levels were determined using manufacturer recommended dilutions of monoclonal antisera developed against mouse APE/REF-1 (Novus Biologicals, Littleton, CO), and monoclonal antisera developed against rat ␤-pol (Ab-1 Clone 18S, NeoMarkers, Fremont, CA). The bands were detected and quantified using a ChemiImager TM system (AlphaInnotech, San Leandro, CA) after incubation in SuperSignal® West Pico Chemiluminescent Substrate (Pierce). Data are expressed as the I.D.V. of the band per g of protein loaded.

APE Endonuclease Activity Assay
The 5Ј-endonuclease activity of APE was analyzed using a quantitative in vitro assay that measures the incision of a 26-mer duplex oligonucleotide substrate containing a synthetic tetrahydrofuran (F) AP site (upper strand, 5Ј-AATTCACCGGTACCFTCTAGAATTCG-3Ј; lower strand, 5Ј-CGAATTCTAGAGGGTACCGGTGAATT-3Ј) as described previously (46). Briefly, 2.5 pmol of radio-end-labeled and purified duplex AP DNA substrate was incubated with 100 ng of crude nuclear extract-selected tissues in a 10-l reaction mixture containing 50 mM Hepes, pH 7.5; 50 mM KCl; 10 mM MgCl 2 ; 2 mM DTT; 1 g/ml BSA; and 0.05% Triton X-100. The reaction mixtures were incubated for 15 min at 37°C and stopped by the addition of 50 mM EDTA. Assay products (5 l) were added to 15 l of loading dye (95% formamide, 5% glycerol, 10 mg of xylene cyanol, 10 mg of bromphenol blue) and heated at 95°C for 5 min. Aliquots (3 l) were run on a 15% denaturing 19:1 acrylamide/ bisacrylamide gel (SequaGel® Sequencing System, National Diagnostics, Atlanta, GA) at 55°C, soaked in fixing solution (10% glacial acetic acid; 10% methanol), wrapped in Saran wrap, and exposed to a Molecular Imaging Screen (Bio-Rad). Endonuclease activity (presence of a 14-mer band) was visualized and quantified using a Molecular Imager® System (Bio-Rad) by calculating the relative amount of the 14-mer oligo product with the unreacted 26-mer substrate (product/(product ϩ substrate)). Data are expressed as machine counts per ng of protein.

Base Excision Repair Assay
The principle of this assay is to measure MFG-BER activity. Radio end-labeled and purified 30-bp oligonucleotides (upper strand, 5Ј-ATA-TACCGCGGUCGGCCGATCAAGCTTATTdd-3Ј; lower strand, 3Ј-ddT-ATATGGCGCCGGCCGGCTAGTTCGAATAA-5Ј) containing a G:U mismatch and an HpaII restriction site (CCGG) were incubated in a reaction mixture (100 mM Tris-HCl, pH 7.5, 5 mM MgCl 2 ; 1 mM DTT, 0.1 mM EDTA, 2 mM ATP, 0.5 mM NAD, 20 M dNTPs, 5 mM diTrisphosphocreatine, 10 units of creatine phosphokinase) with 50 g of crude nuclear extract isolated from selected tissues. The reaction mixtures were incubated for 30 min at 37°C, followed by 5 min at 95°C to stop the reaction. The duplex oligonucleotides were allowed to reanneal for 1 h at room temperature before being briefly centrifuged to pellet the denatured proteins. Repair of the G:U mismatch to a correct G:C base pair was determined via treatment of the duplex oligonucleotide with 20 units of HpaII (Promega, Madison, WI) for 1 h at 37°C and analysis by electrophoresis on a 20% denaturing 19:1 acrylamide/bisacrylamide gel (SequaGel® Sequencing System, National Diagnostics, Atlanta, GA). Repair activity (presence of a 16-mer band) was visualized and quantified using a Molecular Imager® System (Bio-Rad) by calculating the ratio of the 16-mer product with the 30-mer substrate (product/ substrate). Data are expressed as machine counts per g of protein.

Analysis of AP Site DNA Binding
The DNA binding ability of nuclear extracts isolated from selected tissues was determined using electrophoretic mobility shift assays (EM-SAs). Nuclear extracts (20 g) were incubated with 4ϫ reaction buffer (final concentrations: 50 mM Hepes, pH 8.0; 100 mM NaCl; 10 mM EDTA; 1 mM DTT; 9.5% glycerol v/v) plus 1 g of BSA and 2 g of poly(dI-dC) for 5 min at room temperature. A radio-end-labeled and purified oligonucleotide probe containing an AP site (0.0125 pmol) was added to the reaction mix and incubated for 30 min at room temperature. Negative controls (all components except nuclear extract) were included in all experiments. In competitive assays, 100ϫ molar excess of unlabeled DNA was added to the reaction mixture. The protein-DNA complex was resolved on a 5% non-denaturing polyacrylamide gel in 0.5ϫ TBE buffer. Reaction products were visualized and quantified using a Molecular Imager® System (Bio-Rad).

DNA Extraction
TEMPO Extraction of DNA-DNA for the aldehyde-reactive probe slot blot (ASB) assay was extracted according to the method described by Hofer and Moller (47) with some modifications. This method minimizes artifactual DNA damage by using 20 mM TEMPO in all solutions and reagents and by minimizing heat treatment of DNA. Briefly, tissue was homogenized in ice-cold PBS and centrifuged (2000 ϫ g at 4°C for 5 min). Resultant supernatant was decanted and pellet resuspended in lysis buffer (Applied Biosystems, Foster City, CA). Proteinase K (30 units, Ambion, Austin, TX) was added, and samples were incubated overnight at 4°C, followed by phenol/chloroform and Sevag (chloroform/ isoamyl alcohol, 24:1) extraction. Extracted DNA was precipitated using 7.5% 4 M NaCl and 2 volumes of 100% cold ethanol and centrifuged (2000 ϫ g at 4°C for 5 min). Resultant pellet was washed in 70% ethanol and resuspended in ice-cold PBS and rehydrated at 4°C. The samples were incubated with RNase A (2 g) and RNase T1 (1,000 units, Ambion, Austin, TX) for 30 min at 37°C, and resultant DNA was cold ethanol-precipitated, resuspended in deionized water at 4°C, and stored at Ϫ70°C.
Gravity Tip Extraction of DNA-DNA for the random oligonucleotide primed synthesis (ROPS) assay was isolated using Qiagen (Valencia, CA) gravity tip columns as described in the manufacturer's protocol. This method generates large fragments of DNA (up to 150-kb) while minimizing shearing.

DNA Damage Analysis
ASB-The ASB assay was carried out as described previously (14) with slight modifications. TEMPO-extracted DNA (8 g) was incubated in 30 l of PBS with 2 mM aldehyde-reactive probe (ARP) (Dojindo Laboratories, Kumamoto, Japan) at 37°C for 10 min. DNA was cold ethanol-precipitated (as described above) and resuspended in 1ϫ TE buffer overnight at 4°C. DNA was heat-denatured at 100°C for 5 min, quickly chilled on ice, and mixed with equal amount of 2 M ammonium FIG. 1. Expression of the Apex gene in liver of Apex ؉/؉ and Apex ؉/؊ mice. A, the level of APE mRNA was measured using RT-PCR. The 350-bp cDNA product corresponding to the level of APE mRNA from each liver sample was detected and quantified using a ChemiImager TM system (AlphaInnotech, San Leandro, CA) and expressed as the I.D.V. of the band per g of RNA used per reaction. Data were normalized based on the amount of ␤-actin present in each sample. B, the level of the 37-kDa APE protein in 200 g of nuclear extract was determined by Western blot analysis using monoclonal antisera developed against mouse APE. Data are expressed as the I.D.V. of the band per g of protein loaded. C, the 5Ј-endonuclease activity of APE was analyzed using a 32 P-end-labeled 26-mer duplex AP-DNA substrate (2.5 pmol) incubated with 100 ng of crude nuclear extract from liver of Apex ϩ/ϩ and Apex ϩ/Ϫ mice. Endonuclease activity (presence of a 14-bp fragment) was visualized and quantified using a Molecular Imager® System by calculating the relative amount of the 14-bp product with the unreacted 26-bp substrate (product/(product ϩ substrate)). Data are expressed as machine counts per ng of protein. All values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group. *, value significantly different from Apex ϩ/ϩ mice at p Ͻ 0.05.
acetate. DNA was immobilized on a nitrocellulose membrane (Schleicher & Schuell) by using a Invitrogen Filtration Manifold system. The membrane was washed in 5ϫ SSC for 15 min at 37°C and then baked under vacuum at 80°C for 30 min. The dried membrane was incubated in a hybridization buffer (final concentrations: 20 mM Tris, pH 7.5; 0.1 M NaCl; 1 mM EDTA; 0.5% casein w/v; 0.25% BSA w/v; 0.1% Tween 20 v/v) for 30 min at room temperature. The membrane was then incubated in the same hybridization buffer containing 100 l of streptavidin-conjugated horseradish peroxidase (BioGenex, San Ramon, CA) at room temperature for 45 min. Following incubation in horseradish peroxidase, the membrane was washed three times for 5 min each at 37°C in TBS, pH 7.5 (final concentrations: 0.26 M NaCl; 1 mM EDTA; 20 mM Tris, pH 7.5; 0.1% Tween 20). The membrane was incubated in ECL (Pierce) for 5 min at room temperature and visualized using a Chemi-Imager TM system (AlphaInnotech, San Leandro, CA). Standards containing known amounts of AP sites (14) were used to determine the relative level of AP sites in liver DNA. Data are expressed as number of AP sites per 10 6 nucleotides.
ROPS-The relative number of 3Ј-OH group-containing DNA strand breaks was quantified using a Klenow(exo Ϫ ) incorporation assay based on the ability of Klenow to initiate DNA synthesis from 3Ј-OH ends of single strand DNA (48). Gravity tip extracted DNA (0.25 g) was heat-denatured at 100°C for 5 min and added to the Klenow reaction buffer (0.5 mM dTTP, 0.5 mM dGTP, and 0.5 mM dATP; 0.33 M dCTP, 1 l of Klenow(exo Ϫ )) with 10ϫ Klenow buffer per manufacturer's protocol (New England Biolabs, Beverly, MA) and 5 Ci of [␣-32 P]ATP (3000 Ci/mmol, PerkinElmer Life Sciences). Reaction mixtures were incubated at 16°C for 30 min, and the reaction was stopped with the addition of 25 l of 12.5 mM EDTA, pH 8.0. Samples (5 l) were spotted onto scored and numbered Whatman DE81 paper and allowed to air dry. The spotted chromatography paper was washed five times for 5 min in 0.5 M Na 2 HPO 4 (dibasic) followed by a brief rinse in deionized water two times and then allowed to air-dry. Paper was cut and placed into scintillation vials with 2.5 ml of ScintiVerse mixture (Fisher). Incorporation of [␣-32 P]dCTP was quantified using a Packard scintillation counter.

Statistical Analysis
Statistical significance between means was determined using analysis of variance followed by the Fisher's least significant difference test where appropriate (49). A p value less than 0.05 was considered statistically significant.

RESULTS
In order to elucidate the molecular effects of gene-targeted disruption of the mouse APE gene (Apex), our laboratory has characterized in detail a transgenic "knockout" mouse containing a heterozygous deletion of the APE gene (Apex ϩ/Ϫ ). As reported previously, the Apex Ϫ/Ϫ mice are embryonic lethal, whereas the Apex ϩ/Ϫ animals are fertile, appear normal, and exhibit reduced APE mRNA and APE protein levels in mouse embryonic fibroblasts and brain cells as compared with their Apex ϩ/ϩ counterparts (38). In a series of experiments, we were able to confirm whether APE heterozygosity would cause these mice to exhibit haploinsufficiency with respect to APE in brain, liver, and testes. In order to determine whether loss of a functional allele of APE would result in reduced expression of this gene, we have quantified the expression of APE via RT-PCR analysis. By using total RNA isolated from liver, we observed a 40 -50% decrease in APE mRNA (Fig. 1A) in Apex ϩ/Ϫ mice as compared with their normal, wild-type (Apex ϩ/ϩ ) counterparts. Because differences in APE mRNA may not necessarily reflect differences in APE protein levels, we also measured APE protein levels using Western blot analysis. We show a corresponding 40 -50% decrease in APE protein in liver (Fig. 1B). To confirm that changes in APE expression (i.e. changes in mRNA and protein levels) result in changes in APE enzymatic activity, we measured APE 5Ј-endonuclease activity using a 26-bp oligonucleotide substrate containing a synthetic tetrahydrofuran (F) AP site. As described previously (46), this in vitro 5Ј-endonuclease assay allows us to quantify APE enzymatic cleavage of the 26-bp oligonucleotide substrate by analyzing the resultant 14-bp product on a denaturing sequencing gel. Data in Fig.  1C show that APE activity was reduced ϳ40% in liver obtained from Apex ϩ/Ϫ mice. Additionally, APE protein and activity were also reduced ϳ40 -50% in testes and brain (Fig. 2, A and B).
Our data demonstrate that half the gene dosage for Apex results in APE haploinsufficiency. Because APE is one of many enzymes involved in the BER process, we were interested in determining the effect of APE haploinsufficiency on BER capacity. We employed an enzymatic activity assay to measure the ability of various tissues from Apex ϩ/ϩ and Apex ϩ/Ϫ animals to repair a uracil-containing synthetic oligonucleotide. In this assay, a double-stranded oligonucleotide containing a G:U mismatch is incubated with crude nuclear extract isolated from various tissues of Apex ϩ/ϩ and Apex ϩ/Ϫ mice. If repair occurs, the G:U mismatch is converted into a correct G:C pair, creating a HpaII restriction site. Thus, upon digestion with HpaII, repair capacity can be quantified, i.e. the 12-bp band indicates repair, and the 30-bp band indicates lack of repair. We find that liver nuclear extracts isolated from Apex ϩ/Ϫ mice resulted in a statistically significant decrease (35%, p Ͻ 0.05) in BER activity as compared with the level of BER activity observed in liver of Apex ϩ/ϩ mice (Fig. 3A).
During BER processing, ␤-pol removes the dRP residue after AP endonuclease cleaves DNA at AP sites and inserts a new nucleotide in the resulting gap. Because the dRPase activity of ␤-pol is the rate-determining step in BER (22), it is interesting to know whether the diminished BER activity in liver tissue of Apex ϩ/Ϫ mice is driven by down-regulation in ␤-pol protein/ activity, i.e. to determine whether observed changes in BER activity correlate to similar changes in ␤-pol. We show corre-sponding 50% decrease in ␤-pol protein levels in the liver extract of Apex ϩ/Ϫ mice (Fig. 3B), suggesting that the reduced BER activity in liver of Apex ϩ/Ϫ mice is due to reduced ␤-pol level and/or activity.
It is of interest to determine whether reduced BER in the liver of Apex ϩ/Ϫ mice is due to reduced APE endonuclease activity. Thus, in an enrichment in vitro BER assay, we determined whether purified recombinant APE protein (Novus Biologicals, Littleton, CO) would reverse the deficiency in BER activity observed in the liver of Apex ϩ/Ϫ mice. Crude nuclear extracts isolated from liver of Apex ϩ/Ϫ mice were enriched with 100 ng of human APE/REF-1 recombinant protein, and the BER activity was quantified. The enrichment of the nuclear extracts with the recombinant APE protein failed to restore diminished BER activity in liver extract of Apex ϩ/Ϫ mice (Fig.  4A), whereas APE endonuclease activity was fully restored FIG. 3. APE-haploinsufficient mice display tissue-specific BER capacity. The experiments were conducted using crude nuclear extracts as described under "Experimental Procedures." A, an autoradiograph of a sequence gel indicating repair activity as visualized by the appearance of a 12-bp fragment. The relative level of base excision repair in selected tissues of mice was quantified using a Bio-Rad Molecular Imager® System, and data were normalized based on the amount of protein used in each reaction. Data are expressed as machine counts per g of protein.
Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group. M, 10-bp DNA ladder; Ϫ, G:U mismatch oligonucleotide incubated in the absence of nuclear extract and treated with HpaII restriction endonuclease; *, value significantly different from Apex ϩ/ϩ mice at p Ͻ 0.05. B, the relative level of the 39-kDa ␤-pol protein in 200 g of nuclear extract was determined by Western blot analysis using monoclonal antisera developed against mouse ␤-pol. The bands were detected and quantified using an AlphaInnotech ChemiImager TM system. Data are expressed as the I.D.V. of the band per g of protein loaded. Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group. *, value significantly different from Apex ϩ/ϩ mice at p Ͻ 0.05. (Fig. 4B). These data demonstrate that the down-regulation in BER activity in liver of Apex ϩ/Ϫ mice is not due to reduced APE endonuclease activity in the extract.
In order to determine directly whether APE-haploinsufficient mice display altered levels of DNA damage, we have measured the presence of AP sites and single strand breaks in isolated liver DNA from Apex ϩ/Ϫ mice as compared with liver DNA from Apex ϩ/ϩ mice. By using a highly sensitive quantitative assay that measures the number of aldehydic DNA lesions corresponding to AP sites, we observed no significant difference in the level of AP sites in Apex ϩ/Ϫ mice as compared with Apex ϩ/ϩ mice (Fig. 5A). Similarly, by using the ROPS assay to measure the number of single strand breaks in liver DNA, we observed no significant difference in single strand breaks in Apex ϩ/Ϫ as compared with Apex ϩ/ϩ (Fig. 5B); thus, the lack of accumulation of DNA damage in Apex ϩ/Ϫ mice suggest that APE haploinsufficiency in liver does not cause an accumulation of genotoxic DNA repair intermediate products in young mice.
We have demonstrated that BER activity is reduced in liver of Apex ϩ/Ϫ mice, and that this reduction does not appear to be due to loss of APE endonuclease activity. In addition to the characterization of APE haploinsufficiency on BER in liver, we wanted to investigate whether other tissues exhibited a similar effect. We initially compared the relative level of APE protein in selected tissues of wild type (Apex ϩ/ϩ ), and we show that in testes the APE protein level was severalfold higher than in both liver and brain (Fig. 6). We chose to analyze brain and testes because these two tissues are highly susceptible to the deleterious consequences of high levels of oxidative damage, which is recognized and repaired by BER. Based on gene dosage, we expected to see a concomitant reduction in BER activity in nuclear extracts isolated from both brain and testes of Apex ϩ/Ϫ mice as compared with nuclear extracts obtained from Apex ϩ/ϩ mice. However, the data presented in Fig. 3A depicts tissue-specific alterations in BER activity. By using the monofunctional glycosylase-initiated BER assay (i.e. G:U mismatch repair), nuclear extracts isolated from the brain of Apex ϩ/Ϫ mice did not show any significant difference in repair activity as compared with nuclear extracts isolated from the brain of Apex ϩ/ϩ mice, although a downward trend was observed. However, when measuring BER activity in testes, Apex ϩ/Ϫ mice exhibited a significant increase (55%, p Ͻ 0.05) in BER. Accordingly, observed changes in BER correlated to similar changes in ␤-pol, i.e. using Western blot analysis we observe a 38% decrease (p Ͻ 0.05) in brain and a 23% increase (p Ͻ 0.05) in testes of Apex ϩ/Ϫ mice as compared with Apex ϩ/ϩ mice (Fig. 3B).
The BER pathway is initiated by the recognition and binding of BER proteins to damaged DNA, e.g. altered bases and/or AP sites, forming a repair complex. Therefore, it is of interest to determine whether Apex ϩ/Ϫ mice display any reduced activity in the recognition of DNA lesions, specifically binding to an AP site in vitro. By using EMSAs, we quantified the ability of nuclear extracts obtained from brain, liver, and testes to bind to a radio-end-labeled oligonucleotide containing an AP site. As shown in Fig. 7, the nuclear extracts isolated from Apex ϩ/Ϫ mice exhibit tissue-specific binding to AP sites in a manner corresponding to the alterations in BER activity observed above. In other words, a significant reduction in AP site binding activity was observed in both liver and brain tissue, where BER/␤-pol was shown to be reduced, whereas AP site binding activity was induced in testes, where BER/␤-pol is up-regulated. Thus, these data indicate that loss of APE may affect BER activity by altering the formation of a BER repair complex on DNA containing AP lesions. DISCUSSION In this study, we have characterized the effect of APE haploinsufficiency on mammalian BER using mice containing a heterozygous gene-targeted deletion of APE (Apex ϩ/Ϫ ). We report data demonstrating that decreased gene-dosage of APE results in tissue-specific differences in the ability of cells/tissues to carry out MFG-BER in vitro. Specifically, BER was shown to decrease in liver and increase in testes, whereas a

FIG. 4. Enrichment with recombinant APE protein does not affect BER activity in liver of Apex ؉/؊ mice.
The experiments were conducted using crude nuclear extracts and extracts enriched with 100 ng of recombinant APE protein as described under "Experimental Procedures." A, the relative level of BER activity in liver tissue of Apex ϩ/Ϫ mice was visualized and quantified using a Bio-Rad Molecular Imager® System, and data were normalized based on the amount of protein used in each reaction. The BER data are expressed as machine counts per g of protein. B, APE activity was visualized and quantified using a Bio-Rad Molecular Imager® System by calculating the relative amounts of the 14-bp product with the unreacted 26-bp substrate (product/(product ϩ substrate)). The APE data are expressed as machine counts per ng of protein. Ϫ, negative control. decreased trend in brain was observed. Studies characterizing the effect of reduced APE have shown increased sensitivity to oxidative stress in cells (50,51) and in mice (38), supporting our recent observation that BER is responsible for the repair of oxidative stress (52). Similarly, explanted homozygous Apex Ϫ/Ϫ blastocysts displayed increased sensitivity to ionizing radiation (39), whereas neuronal cells from Apex ϩ/Ϫ mice were sensitive to menadione (38). However, it is currently unclear whether enhanced sensitivity to stress is strictly due to loss of APE endonuclease activity, loss of APE redox regulatory activity, or both.
The tissue-specific differences in BER observed in response to APE haploinsufficiency may be due to differences in the relative level of BER. For example, we have shown previously (53) that testes have the highest relative level of MFG-BER as compared with liver and brain. Additionally, Intano et al. (54) have also shown that MFG-BER is highest in mixed germ cells, followed by liver, and then brain. With respect to the relative tissue levels of APE, its importance for cellular survivability suggests that its expression in cells would be ubiquitous. However, Tan et al. (55) have shown that although APE mRNA was ubiquitously expressed in rats, the levels varied significantly among tissues with the highest level observed in testes and comparably low levels in both liver and brain. Although tissuespecific differences in the level of mRNA do not necessarily imply tissue-specific protein levels, we have shown that although APE protein was detectable in all tissues studied, testes displayed a significantly higher level than both liver and brain. Furthermore, tissue-specific variations in cellular and subcellular localization of APE have been observed in humans, with spermatocytes and hepatocytes showing unusually high levels of cytoplasmic staining in addition to expected nuclear localization (56,57).
A question that remains is why a 40 -50% reduction in the expression and endonuclease activity of APE would cause BER to respond differently among various tissue types. It has been clearly demonstrated and widely accepted that ␤-pol is the rate-limiting enzyme in MFG-BER (22). Our laboratory has also demonstrated that ␤-pol expression and activity strongly correlate with MFG-BER under a variety of environmental stimuli (52,53,58,59). A recent study (37) examining kinetic data of various BER enzymes has also suggested that ␤-polmediated MFG-BER may be the predominant BER pathway in vivo. As such, the finding that decreased levels of APE in livers of APE-haploinsufficient mice resulted in decreased MFG-BER was surprising. However, the current dogma of BER requires that BER proteins function in a highly coordinated series of complex protein-protein and protein-DNA interactions to minimize the formation of toxic DNA repair intermediates. For example, APE has been shown to interact with ␤-pol, facilitating its loading onto AP DNA, and stimulates its rate-determining dRP excision activity (31). Here we present data demonstrating that reduced APE results in decreased AP site DNA binding activity, which may consequently affect BER activity by altering the formation of a repair complex on AP lesions. FIG. 5. Analysis of AP sites and single strand breaks in liver DNA of Apex ؉/؉ and Apex ؉/؊ mice. A, liver DNA was isolated using the TEMPO extraction method, and the number of AP sites was measured using the ASB assay as described under "Experimental Procedures." Data are expressed as the I.D.V. of the band per g of DNA loaded. Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group. ADL, aldehydic DNA lesions, which correspond to the level of AP sites. B, liver DNA was isolated using gravity tip extraction columns, and the number of single strand breaks was measured using the ROPS assay as described under "Experimental Procedures." Data are expressed as machine counts per min corresponding to the level of [␣-32 P]dCTP incorporation as quantified by a Packard scintillation counter. Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group.

FIG. 6. Relative levels of APE protein in various tissues of
Apex ؉/؉ mice. Crude nuclear extracts were isolated from brain, liver, and testes using Sigma CelLytic TM NuCLEAR TM extraction kit as described under "Experimental Procedures." The relative level of APE protein in 200 g of nuclear extract was determined by Western blot analysis. The bands were detected and quantified using an Alpha-Innotech ChemiImager TM system. Data are expressed as the I.D.V. of the band per g of protein loaded. Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group.
Furthermore, stabilization of ␤-pol to AP DNA complexes by p53 has been shown to require the presence of APE (60). Although the evidence suggests that a decrease in BER in liver of APE-haploinsufficient mice may be due to decreased protein-DNA stabilization when APE is deficient, in testes a similar decrease in APE protein levels resulted in an increase in AP DNA binding activity as well as an increase in BER. Thus, the ability of a BER repair complex to bind to AP sites and initiate BER activity appears to be regulated by ␤-pol and not APE.
Because BER declined in liver in response to APE haploinsufficiency, we expected that enrichment of the BER reaction mixtures with recombinant APE would fully restore BER activity. However, APE enrichment did not replenish BER activity, whereas APE endonuclease activity was fully restored. This finding provides convincing evidence that the mechanism behind the decline in BER in response to APE haploinsufficiency is not due to a decrease in APE endonuclease activity and provides further support that ␤-pol is indeed rate-determining, i.e. BER activity is tightly correlated with ␤-pol protein level/activity. Further support for the role of ␤-pol, and not APE, in determining BER capacity is shown by data demonstrating a corresponding up-regulation of ␤-pol and BER in testes of Apex ϩ/Ϫ mice, where APE is haploinsufficient.
When examining DNA damage in liver, decreased BER did not result in increased single strand breaks or AP sites, suggesting that in the absence of stress, liver of Apex ϩ/Ϫ mice may still possess the capacity to protect against DNA damage repaired by BER. This finding is in support of our recent study that BER deficiency due to ␤-pol haploinsufficiency does not significantly increase DNA damage in young mice, whereas environmental stress and/or age promoted higher levels of DNA strand breaks (52,58). In line with these studies, we suggest that environmental stress could potentially exacerbate an APE-haploinsufficient phenotype by rendering Apex ϩ/Ϫ tissues susceptible to DNA damage. This notion is supported by Meira et al. (38), reporting that neuronal cells from Apex ϩ/Ϫ mice were sensitive to menadione. In addition, preliminary data from our laboratory suggest that whereas basal levels of NF-B activation are not altered in response to APE haploinsufficiency, Apex ϩ/Ϫ mice exposed to 2-nitropropane-induced oxidative stress were unable to stimulate activation of NF-B as efficiently as wild-type mice (data not shown).
In testes, we demonstrate that APE haploinsufficiency upregulates BER activity, with corresponding increases in the level of ␤-pol protein and the level of binding to AP site DNA. The tissue-specific stimulation of damage recognition and BER activity in testes incurred by the loss of APE in Apex ϩ/Ϫ mice may be the result of increased susceptibility to the background levels of stress that are present within the cellular milieu of rapidly dividing cells. Paradoxically, a delicate balance between oxidant production and oxidant removal must be present in testes, because reactive oxygen species have been shown to play an important role in normal fertilization (61). We have established previously that oxidative stress is capable of inducing the ␤-pol-mediated BER pathway in various tissues (52). Furthermore, in response to an active state of spermatogenesis, there is a constant level of oxidative stress in testes (62,63), and unrepaired oxidative damage has been shown to induce abnormal sperm with reduced fertility (64). However, recent reports (65) have indicated that nuclear DNA of human spermatozoa is relatively resistant to damage induced by H 2 O 2 and iron treatment in vitro. This particular paradox may be due to the high relative level of BER found in testes (53,54,66) and the ability of APE to function both as a BER enzyme and as a redox regulatory molecule. Because BER may function independently of APE (67), it is likely that under circumstances of FIG. 7. APE-haploinsufficient mice display tissue-specific differences in binding to DNA containing AP sites. The experiments were conducted using crude nuclear extracts and radio-end-labeled oligonucleotide containing a synthetic tetrahydrofuran AP site as described under "Experimental Procedures." The relative level of AP site DNA binding in various tissues of Apex ϩ/ϩ and Apex ϩ/Ϫ mice were determined using EMSA. The level of bandshift was determined using a Bio-Rad Molecular Imager® System, and data were normalized based on the amount of protein used in each reaction. Data are expressed as machine counts per g of protein.
Values represent an average (ϮS.E.) for data obtained from at least 5 animals in each group. Comp., nuclear extracts incubated in the presence of 100ϫ molar excess of unlabeled AP site DNA to confirm binding specificity; n.s., nonspecific binding; *, value significantly different from Apex ϩ/ϩ mice at p Ͻ 0.05. constant stress due to high levels of reactive oxygen species, APE may be partitioned to function as a redox protein to stimulate transcription factors involved in various cellular stress-response pathways. Nonetheless, these data provide further evidence that APE endonuclease activity is not rate-determining in MFG-BER and strongly suggest that the involvement of APE as a signaling molecule in the BER response to DNA damage may help maintain oxidative homeostasis in testes.
Based on the data presented herein, it is inviting to suggest that tissue-specific differences in BER capacity observed in APE-haploinsufficient cells/tissues is not due to a simple alteration in APE endonuclease activity but perhaps is due to an alteration in the redox function of APE, as well as to differences in the susceptibility of specific tissues to background levels of stress. We propose that a decrease in APE redox activation of p53 may be the mechanism by which BER displays tissuespecific activity. Data in the literature support this premise by showing the following: p53 is involved in the cellular response to stress (68); p53 activation is APE-dependent (69,70); and p53 may regulate BER activity, perhaps through alteration in the expression of ␤-pol (42,60). Because APE is known to promote the activation of p53 in a redox-dependent and -independent manner (69,71), it is likely that the decrease in BER observed in liver of Apex ϩ/Ϫ mice may be due to a decrease in the activation of p53. In support, down-regulation of APE by antisense RNA has been shown to reduce dramatically p53 activation (70), and preliminary data show a reduction in activation of p53 in liver of Apex ϩ/Ϫ mice (data not shown). Thus, reduced APE redox activity may render a cell/tissue more susceptible to environmental stress by altering cellular responses such as BER. As such, APE appears to form a unique link between BER and the cellular signaling mechanisms responsible for the DNA damage response.