p53 functions in the incorporation step in DNA base excision repair in mouse liver mitochondria


The tumor suppressor p53 protein stimulates nuclear base excision repair (BER) in vitro. In response to certain cellular stresses, p53 translocates to mitochondria, where it can trigger an apoptotic response. However, a potential role for p53 in modulating mitochondrial DNA repair has not yet been examined. In this study, we show that p53 also modulates mitochondrial BER. Uracil-initiated BER incorporation, which measures flux through the entire BER pathway, was lower in mitochondrial extracts from nonstressed p53 knockout mice than in wild type. The addition of recombinant p53 complemented the BER incorporation in p53 knockout extracts and stimulated BER in wt extracts. The activities of three major mitochondrial DNA glycosylases were similar in extracts from wild-type and knockout animals. Likewise, AP endonuclease activity was unaffected by the absence of p53. Gel shift experiments with recombinant p53 demonstrated that p53 did not bind to the uracil-containing substrate used in the repair assay. Polymerase γ gap-filing activity was less efficient in p53 knockout extracts, but it was complemented with the addition of recombinant p53. Thus, we conclude that p53 may participate in mtBER by stimulating the repair synthesis incorporation step.


The tumor suppressor protein p53 plays a central role in controlling cell fate after various stresses (Vogelstein et al., 2000). p53 protein is involved in many aspects of the cellular response to DNA damage and its expression and protein stability are both enhanced after damage. In addition to its primary cellular function as a transcription activator, p53 is directly involved in DNA repair, most notably nucleotide excision repair (Evans et al., 1993; Wang et al., 1995; Hanawalt, 2002). Recently, a direct participation of p53 in base excision repair (BER) has been reported. BER is the major DNA repair pathway for the processing of small base modifications, such as oxidative and alkylation damage. Offer et al. (1999) demonstrated that p53 expression augments apurinic/apyrimidinic (AP) site-induced DNA repair synthesis incorporation. This stimulation was found to be independent of the p53 transcriptional activity because a trans-activation-deficient p53 mutant was more efficient in stimulating BER than wild-type p53 (Offer et al., 2001). The mechanism by which p53 stimulates BER is still unclear. A direct interaction of p53 with APE 1 (Meira et al., 1997; Gaiddon et al., 1999) and with DNA polymerase (pol) β, and the stabilization of pol β binding to the AP site by p53 have been documented (Zhou et al., 2001). Moreover, Seo et al. (2002) demonstrated a deficient pol β expression in p53-deficent cell lines.

Recent results also suggest a role for p53 in controlling mitochondrial function and response to stress (Moll and Zaika, 2001). After various death stimuli, a fraction of the induced p53 protein translocates to mitochondria, where it elicits a series of responses that can ultimately lead to cell death (Marchenko et al., 2000; Dumont et al., 2003; Mihara et al., 2003). However, it is unknown whether p53 plays any role in DNA repair in mitochondria, and whether the protein is even present in the mitochondria in the absence of any stress.

Mammalian mitochondria are very proficient at removing small base modifications through the BER pathway (for a review, see Bohr et al., 2002). BER is initiated by a DNA glycosylase that specifically removes the damaged base, generating an abasic (AP) site. The AP site is processed by an AP endonuclease, or by the glycosylase-associated AP lyase activity. This activity introduces a single DNA strand break at the site of damage. After the processing of the ends and the insertion of a new nucleotide by DNA polymerase gamma (pol γ), the nick is sealed by DNA ligase, regenerating the parental DNA. Several of the enzymes involved in mitochondrial BER have been characterized. In many cases, the same gene encodes for both the mitochondrial and the nuclear enzyme, as it has been demonstrated for uracil DNA glycosylase (UDG) (Nilsen et al., 1997), oxoguanine DNA glycosylase (OGG1) (Souza-Pinto et al., 2001) and DNA ligase III (Lakshmipathy and Campbell, 1999). Although this pathway has been reconstituted in vitro (Stierum et al., 1999a), it is unclear how mitochondrial BER (mtBER) is regulated in vivo.

Since p53 modulates nuclear BER, and mtBER utilizes a similar set of proteins to that of nuclear BER, we asked whether basal p53 also modulates BER in mitochondria. To address this question, we compared DNA repair activities in mitochondrial and nuclear extracts obtained from unstressed mice with basal p53 levels and from heterozygous (p53−/+) and knockout (p53−/−) animals. Our results suggest that the glycosylase and AP-endonuclease steps of BER are not directly affected by p53 status. On the other hand, the repair synthesis incorporation step, which is catalysed in mitochondria by pol γ, was lower in p53−/− extracts. This defect was complemented by the addition of recombinant p53. Altogether, these results suggest that p53 modulates mtBER through the stimulation of the nucleotide incorporation step.


One major concern when investigating mitochondrial DNA repair is the possibility that the mitochondrial extracts are contaminated with nuclear DNA repair enzymes. In order to ascertain that our mouse liver mitochondrial extracts were free of nuclear contamination, we analysed for the presence of Lamin B2 in mitochondrial extracts, by Western blots. Lamin B2 was chosen because it is a very abundant nuclear protein, and thus, we would be able to detect even relatively small amounts of nuclear contamination in the mitochondrial fractions. Antibodies against cytochrome oxidase subunit IV (COX IV) were used as marker for mitochondrial content. Figure 1 shows a typical blot from nuclear (lanes 1–2) and mitochondrial (lanes 3–4) extracts from wild-type and p53−/− animals. The absence of the lamin B2 signal in the mitochondrial extracts demonstrates that our extracts were free of nuclear contamination.

Figure 1

Western blot analysis of mouse liver extracts: 10 μg of mouse liver nuclear (MLNE) or mitochondrial (MLME) extracts were separated by SDS–PAGE. The proteins were transferred to a PVDF membrane and blotted with monoclonal anti-Lamin B2 (NovoCastra) and monoclonal anti-cytochrome oxidase subunit IV (COX IV) (Molecular Probes)

A role for p53 in nuclear BER has been previously demonstrated using human and mouse cell lines. In order to investigate the involvement of p53 in mtBER, we isolated mitochondria and nuclei from livers of wild-type, p53 heterozygote (p53−/+) and p53 knockout (p53−/−) mice. Base excision repair was analysed as the incorporation of a radiolabeled dCTP nucleotide into an unlabeled 30-mer double-strand substrate containing a U/G base pair. This assay measure the flux through the entire BER pathway, since the incorporation of radioactivity into the full-length oligonucleotide requires the combined activities of a glycosylase, an AP-endonuclease, a DNA polymerase and a DNA ligase. The incubation of a uracil-containing unlabeled 30-mer oligonucleotide (U30) with mouse liver nuclear extracts (MLNE) resulted in the appearance of a 30-mer radioactive band, indicating the incorporation of 32P-dCTP in place of the uracil damage (Figure 2, panel a). A small amount of radioactivity was incorporated into the control oligonucleotide, most likely due to some unspecific nick-directed DNA synthesis. For quantification, the incorporation with control oligonucleotide was subtracted from the signals obtained with the uracil-containing oligonucleotide. The intensity of the 30-mer band increased in a linear fashion with the amount of protein added to the assays and with incubation time (not shown). In the MLNE from p53−/− mice, we found 40% less incorporation than in wt (Figure 2a), suggesting that p53 directly participates in BER in the mouse liver nuclei. BER incorporation in the p53−/+ heterozygote extracts was similar to wt, and these extracts thus contained sufficient p53 to support BER.

Figure 2

Uracil-initiated repair synthesis incorporation in mitochondrial and nuclear mouse liver extracts: Repair incorporation reactions were carried out with undamaged (C30, lanes 2–4) or uracil-containing (U30, lanes 5–7) 30-mer double-strand oligonucleotides, with 50 μg of mitochondrial (a) or nuclear (b) extracts for 4 or 2 h, respectively, at 32°C. The DNA was precipitated and resolved in a 20% PAGE and the band intensities were quantified. Relative incorporation was calculated relative to the incorporation obtained with wild-type extracts incubated with uracil-containing oligo. Radioactivity present in control lanes was subtracted from values obtained with damaged substrate. Results presented are the mean±standard deviation of three animals per genotype

The incubation of U30 oligonucleotides, but not C30, with mouse liver mitochondrial extracts (MLME) resulted in the appearance of a 12-mer radioactive band (Figure 2b). This band corresponds to the unligated intermediate generated after uracil removal and incorporation of the new nucleotide. We have previously observed that MLME, in our hands, have week DNA ligase activity and thus accumulate the unligated intermediate (Stuart et al., 2004). It is important to point out that the 12-mer intermediate is a ligatable species, since addition of T4DNA ligase to the reactions converts the 12-mer intermediate into the full-length 30-mer product (not shown). Interestingly, MLME from p53−/+ and p53−/− showed lower incorporation activity than wild-type extracts, by 10 and 21%, respectively. Although the differences were not statistically significant, the gradual decrease over heterozygous to homozygous does suggest a role for p53 in mitochondrial BER.

We then determined whether the decreased BER activity in extracts from p53−/− mice was related to the absence of the p53 protein or to an indirect effect of the p53 deficiency. We performed uracil-initiated BER assays with nuclear and mitochondrial extracts from wt and p53−/−, with or without the addition of recombinant p53 protein (Figure 3, panels a and b). In both cases, 0.5 μg of recombinant p53 fully complemented the deficient BER in the p53−/− extracts (see bar graphs). Additional p53, up to 1 μg, did not enhance BER activity further. The complementation of the uracil-initiated BER deficiency in mitochondrial extracts from p53−/− mice was dependent on the integrity of the p53 protein, since 0.5 μg of heat-denatured (5 min at 95°C, followed by 10 min in ice) failed to complement the defect (Figure 3c). Similarly, addition of excess BSA (1 μg) failed to stimulate mitochondrial BER.

Figure 3

p53 stimulation of BER in mitochondrial and nuclear mouse liver extracts: 50 μg of nuclear (a) or mitochondrial (b) extracts from wild type (lanes 1–4) or p53 knockout (lanes 5–8) were incubated with C30 (lanes 1 and 5) or U30 (lanes 2–4 and 6–8) oligonucleotides for 2 h at 32°C, in the absence or presence of 500 ng or 1 μg of recombinant p53. (c) 50 μg of wt (lanes 1 and 2) or p53−/− (lanes 3–6) mitochondrial extracts were incubated with C30 (lane 1) or U30 (lanes 3–6) oligonucleotides for 2 h at 32°C, with the following additions: 0.5 μg of recombinant p53 (lane 4), 0.5 μg of heat-denatured p53 (lane 5) or 1 μg of BSA (lane 6). The graphs present the mean±standard deviation of two animals per genotype, each assayed twice

Since p53 is a DNA-binding protein that efficiently binds some DNA mismatches (Foord et al., 1991; Lee et al., 1995; Degtyareva et al., 2001), we tested whether p53 could bind to our particular BER substrate, which contains a U/G mismatch, and thus might affect BER activity through binding to the DNA substrate. Recombinant full-length p53 was incubated with one of the following oligonucleotides: a 30-mer oligonucleotide containing the partial p53 consensus binding sequence (CATG) (p53C), the U30 oligonucleotide and oligonucleotides in which a damaged base (either uracil (p53U) or 8-oxoguanine (p53OG)) was introduced within the p53-binding site. Recombinant p53 efficiently shifted the p53C oligonucleotide but not the U30 (Figure 4, lanes 1 and 2), indicating that the repair substrate was not bound by p53 in vitro. The presence of a damaged base in the p53 consensus sequence diminished DNA binding by p53 (Figure 4, lanes 3–4). These results indicate that the complementation of uracil-initiated mtBER activity observed after the addition of recombinant p53 was not due to p53 binding to the substrate, but rather a direct effect of the protein in one or more enzymatic steps of the BER pathway.

Figure 4

Binding of p53 to the oligonucleotide substrates: 4 pmoles of recombinant p53 or BSA were incubated with 300 fmoles of 32P-labeled double-strand oligonucleotides (Table 1). The protein–DNA complexes were detected as a shift in the oligonucleotide's mobility in a 5% PAGE (a). (b) Quantification of % of shifted substrate. Results are average ± standard deviation of three separate experiments

The first step in BER is catalysed by a DNA glycosylase, which recognizes the damaged base. Each DNA glycosylase recognizes a defined set of modified bases, and thus these enzymes provide the specificity of the repair process (Krokan et al., 1997). In mitochondria, three major DNA glycosylases have been characterized: mtUDG, which catalyses the removal of uracil (Domena and Mosbaugh, 1985); mtOGG1, which removes oxidized purines, most notably 8-oxoG (Souza-Pinto et al., 2001); and mtNTH1, which recognizes oxidized pyrimidines (Karahalil et al., 2003). We measured the activities of each of these enzymes in mitochondrial and nuclear extracts from wt, p53−/+ and p53−/− mice to investigate whether p53 levels modulate activity of those enzymes. DNA glycosylase activity was measured as incision of a radiolabeled single lesion-containing double-strand oligonucleotide. To measure UDG activity, we used an oligonucleotide containing a single uracil (U30, Table 1). For OGG1 and endonuclease III homologue 1 (NTH1), substrates containing a single 8-oxoG or 5-OHC were used, respectively (OG30 and OC30, Table 1). In mitochondrial extracts, UDG activity is many-fold higher than OGG1 and NTH1. However, no differences in any of the three enzymes activities were observed in extracts from wt, p53−/+ or p53−/− animals (Figure 5, panel a). Similarly, UDG activity was higher than OGG1 and NTH1 in the nuclear extracts. No differences in the glycosylase activities were observed among the different genotypes in nuclear extracts either (Figure 5, panel b). These results suggest that p53 does not directly act in the glycosylase step of BER initiated by UDG, OGG1 or NTH1, since absence of p53 did not change activity of any of those enzymes.

Table 1 Oligonucleotides used in this study
Figure 5

DNA glycosylase activities in mitochondrial and nuclear mouse liver extracts: 100 μg of mitochondrial (MLME) (a) or 10 μg of nuclear (MLNE) (b) extracts from p53 wild-type (wt), heterozygous (+/−) or knockout (−/−) mice were incubated with 1 ng of oligonucleotides containing a single uracil, 8-oxoG or 5-OHC for 6 h at 32°C. The samples were resolved as described. Percent incision was calculated as the amount of radioactivity in the product band over the total radioactivity in the lane. Results presented are the mean±standard deviation of three animals per genotype. Each mitochondrial preparation was assayed twice

The removal of the damage base by the DNA glycosylase yields the generation of an AP site. AP sites are mostly processed by an AP endonuclease, although some glycosylases, such as OGG1 and NTH1, posses an associated AP lyase activity that can also catalyse the cleavage of the phosphate backbone. This step leads to the formation of a single-strand break and creates the substrate for the incorporation step, in which the new nucleotide is added to the DNA. Thus, we tested whether AP endonuclease activity varied in mitochondrial extracts from wt, p53−/+ and p53−/− mice. AP endonuclease activity was monitored as incision of an oligonucleotide containing a tetrahydrofuran AP analog, which is a model substrate for AP endonuclease cleavage but is resistant to β-elimination catalysed by AP lyases (Wilson et al., 1995). Mitochondrial extracts from animals of the three different genotypes have very similar THF incision activity (Figure 6), indicating that basal mitochondrial AP endonuclease does not change with p53 status.

Figure 6

AP endonuclease activity in mouse liver mitochondrial extracts: 1 μg of mitochondrial extracts from p53 wild-type (wt), heterozygous (+/−) or knockout (−/−) mice were incubated with 100 fmoles of THF oligonucleotide for 15 min at 32°C. The samples were resolved as described. Percent incision was calculated as the amount of radioactivity in the product band over the total radioactivity in the lane. Results presented are the mean±standard deviation of three animals per genotype. Each mitochondrial preparation was assayed twice

The results above suggested that the incorporation step might be the target for p53 modulation of mitochondrial BER. We thus investigated pol γ activity in mitochondrial extracts from wt and p53−/− mice. Pol γ activity was measured using a gap-filling assay, in which a 34-mer substrate (Table 1) containing a one-nucleotide gap is incubated with the extracts in presence of a radioactivity nucleotide. Polymerase activity is measured as the amount of radioactivity incorporated in the 5′ nucleotide. Incubation of the GAP oligonucleotide with increasing amounts of wt mitochondrial extracts yields a dose-dependent increase in the amount of radioactivity incorporated (Figure 7a). Incubation of the substrate with mitochondrial extracts from p53−/− mice results in less incorporation of radioactivity compared to the wt, suggesting that pol γ activity is lower in mitochondria from p53 knockout mice. To ascertain that the decreased pol γ activity was related to the absence of p53 itself, and not to a secondary effect of p53 knockout, we added recombinant p53 to the gap filing assays (Figure 7b). In all, 0.25 μg of p53 fully complemented the defect in pol γ gap-filling activity. Interestingly, addition of more p53 stimulated pol γ even further, both in the p53−/− and in wt mitochondrial extracts (not shown). Together, these results suggest a direct action of p53 on the nucleotide incorporation activity of pol γ.

Figure 7

Pol γ activity stimulation by p53: increasing amounts of mitochondrial extracts (0.25–5 μg) from wild-type (wt) or p53 knockout (p53−/−) mice were incubated with 1 pmol of a 1 nucleotide gap oligonucleotide, in the presence of 32P-dCTP, and gap-filing activity was quantified as the amount of radioactivity incorporated, relative to 0.25 μg of MLME-wt (a). Alternatively, 5 μg of MLME from wt or p53−/− mice were incubated with 1 pmol of gap substrate alone or with increasing concentrations of recombinant p53 (b). Gap-filing activity was quantified as above, relative to incorporation with 5 μg of MLME-wt in the absence of added p53. Results presented are average±SEM of three and two separate experiments


The various roles of p53 in the cellular responses to DNA damage are still not clear. Although not required for BER (Sobol et al., 2003), p53 has been shown to stimulate nuclear BER in vitro (Zhou et al., 2001), in cells in culture (Offer et al., 1999) and now in mouse liver (this work). Ours is the first study on the direct role of p53 on mitochondrial BER. Mitochondrial BER resembles nuclear BER, and many of the repair enzymes in both compartments are encoded by the same genes. Thus, we hypothesized that p53 modulates mitochondrial BER in a similar fashion as it does nuclear BER. To test this hypothesis, we compared BER activities in liver mitochondrial and nuclear extract from wild-type, p53−/+ and p53−/− mice.

We found that uracil-initiated BER activity is lower in mitochondria and nuclei from p53 knockout mice. The BER impairment is related to the absence of the p53 protein itself, as supported by two lines of experimental evidence. Firstly, addition of recombinant p53 restored BER activity to wild-type levels both in mitochondrial and nuclear extracts. Secondly, the p53 complementation depended on the integrity of the protein, since heat-denatured p53 or excess BSA did not complement the BER defect in mitochondrial extracts from the p53 knockout animals. Taken together, these results suggest that p53 functions in one or more steps in the BER pathway in mitochondria. Our observations with mouse liver nuclear extracts are also in agreement with the previously published observation that p53-defficient mouse cell lines have decreased AP-initiated BER activity (Offer et al., 2001).

We investigated whether the role of p53 in mtBER was at the DNA glycosylase step. DNA glycosylases initiate BER by recognizing and removing the damaged base. However, we found that p53 levels did not affect the activities of the three major glycosylases so far characterized in mitochondria: mtUDG, mtOGG1 and mtNTH1. Similarly, UDG, OGG1 and NTH1 activities in nuclear extracts from p53−/+ and p53−/− were no different from wt extracts. Although OGG1 and NTH1 activities were significantly lower than that of UDG, particularly in mitochondrial extracts, those levels of incision activity were still sufficient to detect any possible defects in absence of p53. We have previously observed significant changes in OGG1 activity with age, even with the low incision levels observed in in vitro assays (Souza-Pinto et al., 1999, Souza-Pinto et al., 2001). These results demonstrate that the glycosylase step is not directly modulated by p53 in mitochondria or nuclei, under basal, nonstressed conditions. Recently, Zurer et al. (2004) demonstrated that 3-mehyladenine DNA glycosylase was modulated by p53, but this modulation depended on genotoxic stress. Thus, it is possible that p53 stimulation of the glycosylase step represents only a small component of the p53 modulation of BER, under specific stress conditions.

The second step in BER is catalysed by an AP-endonuclease. APE1, the major human AP endonuclease, has been shown to interact with and activate p53 (for a review, see (Evans et al., 2000)). However, AP endonuclease activity, measured as incision of a THF-containing substrate, did not differ in mitochondrial extracts from p53 heterozygous and knockout mice. These results suggest that, under nonstress conditions, p53 does not modulate the AP endonuclease step of BER in mitochondria. These results are in agreement with the observations reported by Seo et al. (2002), who found no changes in AP endonuclease activity in p53-null human whole-cell extracts.

Taken together, our results suggest that basal levels of p53 may function in the nucleotide incorporation/termini-processing step in mitochondrial BER. This would be consistent with our observations that uracil-initiated BER activity is lower in p53−/− mitochondrial extracts, but that mtUDG and AP endonuclease activities are unchanged. In the nuclei, both incorporation and termini processing activities are catalysed by pol β (Allinson et al., 2001), while in mitochondria they are catalysed by pol γ (Longley et al., 1998; Stierum et al., 1999b). The observation that p53 directly interacts with pol β in vitro (Zhou et al., 2001), together with reported lower levels of pol β in p53-null human cells (Seo et al., 2002) suggest that the polymerase step is likely the step in which p53 exerts it modulator effect in the nucleus. Our results suggest that a similar mechanism may be in effect in mitochondria. We found that pol γ nucleotide incorporation activity is lower in mitochondrial extracts from p53-deficient mice. Thus, it is very likely that the lower activity of uracil-initiated BER incorporation in the p53KO extracts is caused by the slower nucleotide incorporation activity of pol γ in the absence of p53. Additional support for this conclusion comes from the observation that recombinant p53 not only complemented the lower pol γ activity in p53−/−, but further stimulated pol γ, in a dose-dependent manner. These results suggest a direct effect of p53 on pol γ enzymatic activity, and also indicate that the deficient BER activity in mitochondria from p53 knockout mice is due exclusively to the lack of p53 itself. Although a direct interaction between p53 and pol γ has not yet been demonstrated, pol γ is stimulated by PCNA (Taguchi et al., 1995), which directly interacts with p53.

In conclusion, our results indicate that p53 may play a role in modulating mitochondrial DNA repair. The observations that addition of p53 stimulates repair synthesis incorporation and pol γ activity suggests that p53 may be one component of a stress response pathway that involves upregulation of mitochondrial repair, and thus plays a direct role in maintaining mitochondrial DNA stability. In support of this notion, Habano et al. (2000) found a direct association between alterations in the p53 gene and microsatellite instability in both the nuclear and the mitochondrial genomes. We (Hofseth et al., 2003) and others (Yamada and Farber, 2002) have shown that an adaptive imbalance in BER enzymes can cause microsatellite instability, including during ulcerative colitis, a chronic inflammatory disease. In addition, nitric oxide, which is produced in mitochondria at high levels, can increase 3-methyl Adenine glycosylase activity and cause deamination of cytosine by nitrosative stress to form uracil (Krokan et al., 2002). Thus, the appropriate coordination of BER enzymes, and its modulation by p53, may be fundamental in maintaining genomic stability in mitochondria after genotoxic and oxidative stress.

Materials and methods


Protease inhibitors were from Roche. Isotopes were from NEN Life Science Products. G25 spin columns and Percoll were from Pharmacia. T4 polynucleotide kinase and T4 DNA ligase were from Stratagene. Recombinant p53 protein was obtained from Santa Cruz Biotech or kindly supplied by Dr Kent Soe, Institute of Molecular Biotechnology, Germany. All other reagents were ACS grade.


Wild-type, p53 heterozygous and knockout mice, in a C57Bl background, were obtained from the National Cancer Institute, NIH (Frederick, MD, USA) and killed within 2 days of their arrival at the Gerontology Research Center (GRC). All experiments were approved by the GRC Animal Care and Use Committee (ACUC) and performed in accordance with the Guidelines for the Use and Care of Laboratory Animals (NIH Publication 85–23).


Oligonucleotides used in this study were synthesized by Midland Certified Reagents Co. (Midland, TX, USA), with exception of OC30 and THF, which were kindly provided by John Essigman, MIT, Cambridge, MA, USA. The sequence of each oligonucleotide is presented in Table 1. For incision reactions and EMSA assays, the substrates were 5′-32P-labeled by incubating with γ32P-ATP in the presence of T4 Polynucleotide kinase. After the kinase reaction, unincorporated 32P-ATP was separated from the oligonucleotides with a G25 desalting column. The labeled oligonucleotides were then annealed to the complementary strand (fivefold excess) in the presence of 100 mM KCl, by briefly incubating at 90°C and allowing them to slowly cool down to room temperature. For repair synthesis incorporation reactions, unlabeled substrates were annealed as above.

Isolation of liver mitochondria and preparation of extracts

Mouse liver mitochondria (MLM) and nuclei (MLN) were isolated using a combination of differential centrifugation and Percoll gradient centrifugation, as previously described (Souza-Pinto et al., 1999). After isolation, the mitochondrial fraction was suspended in 20 mM HEPES (pH 7.6), 1 mM EDTA, 5% glycerol, 0.015% Triton X-100, 5 mM dithiothreitol and 300 mM KCl with the following protease inhibitors added just before use: 1 μg/ml aprotinin, 1 μg/ml pepstatin A, 1 μg/ml chymostatin A, 2 μg/ml leupeptin, 2 μ M benzamidine HCl, 1 mM phenylmethylsulfonyl fluoride and 1 μ M E-64. Mitochondria were solubilized by adding sterile 10% Triton X-100 to a final concentration of 0.5% and clarified by centrifugation for 1 h at 130 000 g. The resulting supernatant was concentrated in Centricon 10 micro-concentrators (Amicon) and the buffer exchanged to 20 mM HEPES (pH 7.6), 1 mM EDTA, 5 mM DTT, 20% glycerol, 100 mM KCl and protease inhibitors. Nuclear extracts were prepared from the pellets obtained after the first low-speed spin of the crude homogenates, using the same procedure described above. All extracts were then aliquoted in small volumes and kept frozen at −80°C until use. Protein concentration was determined using the Lowry method, with BSA as standard.

Oligonucleotide incision assays

UDG activity was measured in a 20 μl reaction containing 70 mM HEPES-KOH (pH 7.5), 1 mM EDTA, 1 mM DTT, 75 mM NaCl, 0.05% BSA, 10 μg (mitochondria) or 1 μg (nuclear) extracts and 90 fmol of 32P-labeled double-strand uracil-containing oligonucleotide (U30) (Table 1) for 1 h at 32°C. Reactions were terminated by addition of 1 μl each 5 mg/ml proteinase K and 10% SDS, and incubation for 15 min at 55°C. To each tube, 20 μl formamide loading buffer (90% formamide, 10 mM EDTA, 1 mg/ml xylene cyanol FF and 1 mg/ml bromophenol blue) was added and the samples heated at 90°C for 10 min.

OGG1 activity was measured in a reaction (20 μl) containing 40 mM HEPES (pH 7.6), 5 mM EDTA, 1 mM DTT, 75 mM KCl, 10% glycerol, 100 μg (mitochondria) or 10 μg (nuclear) extracts and 88.7 fmol of 32P-labeled double-strand 8-oxoG-containing oligonucleotide (OG30) (Table 1) for 6 h at 32°C.The reactions were terminated as described above.

Endonuclease III homologue (NTH1) activity was determined in the same reaction conditions as for OGG1, with the exception that 50 μg (mitochondria) and 5 μg (nuclear) extracts were used, and the substrate was a 5-OHC-containing-oligonucleotide (OC30), incubated at 32°C for 4 h.

AP endonuclease activitiy was determined by incubating 1 μg mitochondrial protein with 100 fmol of a 32P-labeled tetrahydrofuran-containing duplexed 16-mer (Table 1) for 15 min at 32°C, in 20 μl of reaction buffer containing 50 mM HEPES-KOH (pH 7.5), 50 mM KCl, 100 μg/ml BSA, 10 mM MgCl2, 10% glycerol and 0.05% Triton X-100. Reactions were terminated by the addition of 20 μl formamide loading buffer and heating at 90°C for 10 min.

For all incision assays, substrate and product DNAs were resolved by electrophoresis at 15 W for 1 h 10 min on 20% polyacrylamide gels containing 7 M urea. Bands were visualized by PhosphorImager and analysed using ImageQuant™ (Molecular Dynamics). Incision activity was determined as the intensity of product bands relative to the total radioactivity in each lane.

Repair synthesis incorporation assay

Repair synthesis reactions (50 μl) contained 40 mM HEPES (pH 7.6), 0.1 mM EDTA, 5 mM MgCl2, 0.2 mg/ml BSA, 50 mM KCl, 1 mM DTT, 25 mM phosphocreatine, 50 μg/ml Creatine Phosphokinase, 2 mM ATP, 20 μ M of each dATP, dTTP, dGTP and 2 μ M of dCTP, 4 μCi α32PdCTP, 3% glycerol, 1 mM NAD+, 300 ng of double-strand U-containing oligonucleotide and the specified amounts of extracts. The reactions were incubated at 32°C for 2–4 h, with or without the addition of 1 U of T4-DNA ligase. The DNA was extracted with equal volumes of phenol : chlorophorm : iso-amyl alcohol (25 : 24 : 1), chlorophorm:iso-amyl alcohol (24 : 1) and precipitated by the addition of 1/5 vol of 11 M NH3-acetate and 2.5 vol of cold ethanol. The DNA was dried and resuspended in formamide loading dye and loaded onto a 20% acrylamide/7 M urea gel. The gels were resolved and visualized as described earlier.

Gap-filling assay

Pol γ activity was measured with a one-nucleotide gap filing assay, in a 10 μl reaction containing 40 mM HEPES (pH 7.6), 0.1 mM EDTA, 5 mM MgCl2, 0.2 mg/ml BSA, 50 mM KCl, 1 mM DTT, 25 mM phosphocreatine, 50 μg/ml Creatine Phosphokinase, 2 mM ATP, 4 μ M dCTP, 4 μCi α32PdCTP, 3% glycerol, 1 μg poly dI·dC, 1 pmol double-strand 1 nucleotide gap 34-mer oligonucleotide (GAP, Table 1) substrate and the specified amounts of extracts. The reactions were incubated at 37°C for 1 h and terminated by addition of 1 μl each 5 mg/ml proteinase K and 10% SDS, and incubation for 15 min at 55°C. The products were ethanol-precipitated, dried, suspended in formamide loading dye and resolved in a 20% polyacrylamide/7 M urea gel, as previously described.

Electro-mobility shift assay (EMSA)

In total, 300 fmol of each oligonucleotide (Table 1) were incubated with 4 pmol of recombinant p53 in a 20 μl reaction containing: 25 mM HEPES pH 7.5, 15% glycerol, 50 mM NaCl, 1 mM DTT and 10 μM EDTA. The reactions were incubated at 30°C for 30 min and loaded onto a 5% polyacrylamide gel containing 5% glycerol. The gels were resolved at 200 V for 2 h, at 4°C. The bands were visualized using PhosphoImager and the percentage of shifted oligonucleotides was calculated using the ImageQuant software.

Western blot analysis

Proteins were separated on 12% Tris-glycine gels (Invitrogen); amounts of mitochondrial or nuclear extracts loaded are specified in the figure legends. Transfer to PVDF membranes (Invitrogen) was carried out by electroblotting in transfer buffer (12 mM Tris, pH 8.3; 96 mM glycine, 20% methanol) for 2 h at 30 V. The membrane was blocked overnight at 4°C in 5% nonfat dry milk (BioRad) in TBST (20 mM Tris-HCl pH 7.2, 137 mM NaCl, 0.1 % Tween-20). Fresh milk-TBST was added with one of the following primary antibodies: mouse monoclonal anti-lamin B2 (Novocastra); mouse monoclonal anti-cytochrome oxidase IV (Molecular Probes); or goat polyclonal anti-p53 (N-19) (Santa Cruz). For detection, membranes were incubated with HRP-coupled secondary antibodies (Santa Cruz) and the bands visualized using the ECL+Plus® (Amersham-Pharmacia Biotech) kit.

Statistical analysis

The results are reported as mean±standard deviation of at least three different mitochondrial and nuclear preparations, each assayed twice. The differences among genotypes were analysed by the Student's t-test, and a P<0.05 was considered statistically significant.





base excision repair


mouse liver mitochondrial extract


mouse liver nuclear extract


mitochondrial base excision repair


mitochondrial DNA


endonuclease III homologue 1


oxoguanine DNA glycosylase

Pol γ:

DNA polymerase gamma




uracil DNA glycosylase


  1. Allinson SL, Dianova II and Dianov GL . (2001). EMBO J., 20, 6919–6926.

  2. Bohr VA, Stevnsner T and Souza-Pinto NC . (2002). Gene, 286, 127–134.

  3. Degtyareva N, Subramanian D and Griffith JD . (2001). J. Biol. Chem., 276, 8778–8784.

  4. Domena JD and Mosbaugh DW . (1985). Biochemistry, 24, 7320–7328.

  5. Dumont P, Leu JI, Della III PA, George DL and Murphy M . (2003). Nat. Genet., 33, 357–365.

  6. Evans AR, Limp-Foster M and Kelley MR . (2000). Mutat. Res., 461, 83–108.

  7. Evans MK, Taffe BG, Harris CC and Bohr VA . (1993). Cancer Res, 53, 5377–5381.

  8. Foord OS, Bhattacharya P, Reich Z and Rotter V . (1991). Nucleic Acids Res., 19, 5191–5198.

  9. Gaiddon C, Moorthy NC and Prives C . (1999). EMBO J., 18, 5609–5621.

  10. Habano W, Sugai T, Nakamura SI, Uesugi N, Yoshida T and Sasou S . (2000). Gastroenterology, 118, 835–841.

  11. Hanawalt PC . (2002). Oncogene, 21, 8949–8956.

  12. Hofseth LJ, Khan MA, Ambrose M, Nikolayeva O, Welliver MX, Kartalou M, Hussain SP, Roth RB, Zhou X, Mechanic L, Zurer I, Rotter V, Samson LD and Harris CC . (2003). J. Clin. Invest., 112, 1887–1894.

  13. Karahalil B, Souza-Pinto NC, Parsons JL, Elder RH and Bohr VA . (2003). J. Biol. Chem., 278, 33701–33707.

  14. Krokan HE, Drablos F and Slupphaug G . (2002). Oncogene, 21, 8935–8948.

  15. Krokan HE, Standal R and Slupphaug G . (1997). Biochem. J., 325, 1–16.

  16. Lakshmipathy U and Campbell C . (1999). Mol. Cell. Biol., 19, 3869–3876.

  17. Lee S, Elenbaas B, Levine A and Griffith J . (1995). Cell, 81, 1013–1020.

  18. Longley MJ, Prasad R, Srivastava DK, Wilson SH and Copeland WC . (1998). Proc. Natl. Acad. Sci. USA, 95, 12244–12248.

  19. Marchenko ND, Zaika A and Moll UM . (2000). J. Biol. Chem., 275, 16202–16212.

  20. Meira LB, Cheo DL, Hammer RE, Burns DK, Reis A and Friedberg EC . (1997). Nat. Genet., 17, 145.

  21. Mihara M, Erster S, Zaika A, Petrenko O, Chittenden T, Pancoska P and Moll UM . (2003). Mol. Cell, 11, 577–590.

  22. Moll UM and Zaika A . (2001). FEBS Lett., 493, 65–69.

  23. Nilsen H, Otterlei M, Haug T, Solum K, Nagelhus TA, Skorpen F and Krokan HE . (1997). Nucleic Acids Res., 21, 2579–2584.

  24. Offer H, Milyavsky M, Erez N, Matas D, Zurer I, Harris CC and Rotter V . (2001). Oncogene, 20, 581–589.

  25. Offer H, Wolkowicz R, Matas D, Blumenstein S, Livneh Z and Rotter V . (1999). FEBS Lett., 450, 197–204.

  26. Seo YR, Fishel ML, Amundson S, Kelley MR and Smith ML . (2002). Oncogene, 21, 731–737.

  27. Sobol RW, Kartalou M, Almeida KH, Joyce DF, Engelward BP, Horton JK, Prasad R, Samson LD and Wilson SH . (2003). J. Biol. Chem., 278, 39951–39959.

  28. Souza-Pinto NC, Croteau DL, Hudson EK, Hansford RG and Bohr VA . (1999). Nucleic Acids Res., 27, 1935–1942.

  29. Souza-Pinto NC, Eide L, Hogue BA, Thybo T, Stevnsner T, Seeberg E, Klungland A and Bohr VA . (2001). Cancer Res., 61, 5378–5381.

  30. Stierum RH, Dianov GL and Bohr VA . (1999a). Nucleic Acids Res., 27, 3712–3719.

  31. Stierum RH, Dianov GL and Bohr VA . (1999b). Nucleic Acids Res., 27, 3712–3719.

  32. Stuart JA, Karahalil B, Hogue BA, Souza-Pinto NC and Bohr VA . (2004). FASEB J., 18, 595–597.

  33. Taguchi T, Ogihara M, Maekawa T, Hanaoka F and Tanno M . (1995). Biochem. Biophys. Res. Commun., 216, 715–722.

  34. Vogelstein B, Lane D and Levine AJ . (2000). Nature, 408, 307–310.

  35. Wang XW, Yeh H, Schaeffer L, Roy R, Moncollin V, Egly JM, Wang Z, Freidberg EC, Evans MK and Taffe BG . (1995). Nat. Genet., 10, 188–195.

  36. Wilson III DM, Takeshita M, Grollman AP and Demple B . (1995). J. Biol. Chem., 270, 16002–16007.

  37. Yamada NA and Farber RA . (2002). Cancer Res., 62, 6061–6064.

  38. Zhou J, Ahn J, Wilson SH and Prives C . (2001). EMBO J., 20, 914–923.

  39. Zurer I, Hofseth LJ, Cohen Y, Xu-Welliver M, Hussain SP, Harris CC and Rotter V . (2004). Carcinogenesis, 25, 11–19.

Download references


We thank Drs Cayetano Von Kobbe and Syed Imam for the critical reading of this manuscript. We also thank Dr Draginja Djurickovic (NCIFCRF) for transferring the animals.

Author information



Corresponding author

Correspondence to Vilhelm A Bohr.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

de Souza-Pinto, N., Harris, C. & Bohr, V. p53 functions in the incorporation step in DNA base excision repair in mouse liver mitochondria. Oncogene 23, 6559–6568 (2004). https://doi.org/10.1038/sj.onc.1207874

Download citation


  • p53
  • mitochondrial DNA
  • DNA repair
  • BER
  • uracil

Further reading