Similar frequency and signature of untargeted substitutions induced by abasic site analog under reduced human APE1 conditions

Abstract

Abasic sites are formed in cells by various factors including environmental mutagens and considered to be involved in cancer initiation, promotion, and progression. A chemically stable abasic site analog (tetrahydrofuran-type analog, THF) induces untargeted base substitutions as well as targeted substitution and large deletion mutations in human cells. The untargeted substitutions may be initiated by the cleavage of the DNA strand bearing THF by the human apurinic/apyrimidinic endonuclease 1 (APE1) protein, the major repair enzyme for THF and abasic sites. To examine the effects of lower APE1 levels, the protein was knocked down by siRNA in human U2OS cells. A plasmid containing a single THF modification outside the supF gene was introduced into the knockdown cells, and the untargeted substitution mutations in the reporter gene were analyzed. Unexpectedly, the knockdown had no evident impact on their frequency and spectrum. The G bases of 5′-GpA-3′ dinucleotides on the modified strand were quite frequently substituted, with and without the APE1 knockdown. These results suggested that the DNA strand cleavage by APE1 is not essential for the THF-induced untargeted base substitutions.

INTRODUCTION

The complementary pairing of nucleobases is the foundation of the expression, transmission, and maintenance of genetic information. However, the base moiety is vulnerable and can be chemically modified by endogenous and environmental mutagens. In addition, the sugar and the phosphodiester backbone are also attacked and damaged. DNA/RNA lesions are considered to lead to mutagenesis, carcinogenesis, and cell death (Kamiya, 2003; Loeb and Harris, 2008; Ichikawa et al., 2008; Ikehata and Ono, 2011; Cao et al., 2014; Suzuki and Kamiya, 2017).

Chemically unstable base lesions are released from nucleic acids, resulting in abasic (apurinic/apyrimidinic, AP) site formation (Loeb and Preston,1986). Moreover, most of the damaged bases in DNA are removed by DNA repair enzymes. The base excision repair pathway involves the cleavage of the N-glycosyl bond by DNA glycosylases (Kamiya, 2020). The DNA glycosylases with a strong β-lyase activity conduct both the base removal and strand cleavage reactions. Meanwhile, abasic sites are formed when DNA glycosylases either lack or possess weak β-lyase activity. A single mammalian cell is estimated to contain tens of thousands of abasic sites, and thus they constitute one type of the major DNA lesions (Lindahl, 1993; Nakamura and Swenberg, 1999).

Due to their noncoding nature, abasic sites induce the “misincorporation” of 2′-deoxyribonucleoside 5′-triphosphates. In mammalian cells, true abasic sites and a chemically stable analog (tetrahydrofuran-type analog, THF) cause various base substitution mutations (Kamiya et al., 1992; Gentil et al., 1992; Kamiya et al., 1993; Cabral Neto et al., 1994). In addition to the mutations at the lesion site, untargeted substitutions and large deletions are also observed (Chen et al., 2014; Suzuki et al., 2018). Thus, abasic site formation represents a serious threat to DNA integrity.

Chen et al. (2014) reported that the presence of U:G and T:G pairs downstream of the supF gene caused base substitutions within the gene. The mutations occurred at the G bases of 5′-GpA-3′ dinucleotides, the complementary sequences of 5′-TpC-3′, which is the preferred substrate for the apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like (APOBEC) family of cytosine deaminases. The authors suggested that the formation of abasic sites after U and T removal by monofunctional DNA glycosylases, the subsequent cleavage of the DNA strand by apurinic/apyrimidinic endonuclease 1 (APE1, Ref-1, HAP, APEX: the major endonuclease for abasic sites), and the “hijacking” by mismatch repair (MMR) proteins resulted in the exposure of single-stranded (ss) DNA (corresponding to the strand with the G bases of U:G and T:G pairs). The conversions of C on the ss DNA to U conducted by APOBEC deaminase(s) would produce the base substitutions at the G locations. The authors showed that the APE1 knockdown seemed to reduce the supF mutant frequency for T:G and significantly decreased the frequency for U:G. However, they did not report the effects of the APE1 knockdown on the untargeted mutations induced by the abasic site (analog).

In this study, we knocked down APE1 in human U2OS cells and transfected plasmid DNA containing THF downstream of the supF gene into the knockdown cells. The knockdown had minimal effects on both the frequency and signature of the mutations found in the gene. These results suggest that APE1 does not play an important role in untargeted mutation induction.

MATERIALS AND METHODS

Nucleic acids

The oligodeoxyribonucleotides (ODNs) used for the plasmid construction (Supplementary Table 1) were purified by high-performance liquid chromatography, as described previously (Kamiya and Kasai, 1997). The double-stranded shuttle plasmid bearing THF outside the supF gene was constructed from the ss forms of pZ189-T_E107K/D402E (formerly pZ189-107K/402E), as described previously (Suzuki et al., 2011, 2018). After annealing the THF-containing ODN (the 5′-terminus was phosphorylated) to the ss DNA, the primer was extended by T4 DNA polymerase and the nick was sealed by T4 DNA ligase. Dam DNA methyltransferase was added into the reaction mixture to methylate the A bases of the extended strand. The control plasmid DNA containing G was prepared by the same procedures.

The siRNA against APE1 (Stealth RNAi, Thermo Fisher Scientific, Waltham, MA, USA) is shown in Supplementary Table 1, and the Stealth RNAi siRNA Negative Control Med GC (Thermo Fisher Scientific) was used as the control RNA.

Mutation analysis

Human U2OS cells (ATCC HTB-96) were treated with the control RNA or the siRNA, and the plasmid DNA containing THF was transfected into the knockdown cells. These treatments were conducted as described previously (Suzuki et al., 2018; Kamiya et al., 2018). The supF mutant frequency was determined using Escherichia coli KS40/pOF105, provided by Professor Tatsuo Nunoshiba of the International Christian University (Obata et al., 1998; Satou et al., 2009).

Western blotting

The APE1 expression was analyzed as described previously (Kamiya et al., 2018). The rabbit polyclonal antibody to APE1 was purchased from GeneTex (catalog number GTX110558, Hsinchu, Taiwan).

Statistical analysis

Statistical significance was examined by the Student’s t-test.

RESULTS

Minimal effects of APE1 knockdown on THF-induced mutation frequency

The plasmid DNA bearing THF was prepared by annealing of a 5′-phosphorylated THF-ODN to circular ss supF DNA and subsequent DNA polymerase and ligase reactions, and introduced into human U2OS cells with decreased amounts of APE1. The analog was incorporated into the position 9 “bases” downstream of the supF gene. Note that substitution and single-base (or small) deletion mutations at this position alone are undetectable, since they do not inactivate the gene. Western blotting was performed to confirm the siRNA-mediated knockdown of APE1 (Fig. 1). The transfection of plasmid DNA was conducted with cationic liposomes at 24 hr after siRNA treatment, and the replicated plasmid DNA was extracted from the transfected cells at 48 hr after transfection for the determination of the supF mutant frequency.

Fig. 1

Knockdown of APE1 by siRNAs. The siRNA was introduced into human U2OS cells, and total protein was extracted at 24, 48, and 72 hr after siRNA introduction. APE1 expression was analyzed by western blotting.

The APE1 knockdown did not affect the mutant frequency when the control plasmid DNA was introduced (Fig. 2). The knockdown appeared to increase the frequency upon transfection of the THF-bearing plasmid DNA (8.2 × 10^-3 versus 1.04 × 10^-2), but this difference was statistically insignificant.

Fig. 2

The supF mutant frequency in U2OS cells transfected with plasmid DNA bearing THF outside the supF gene. Transfection experiments were performed seven times. Data are expressed as the means + S.E. N. S. represents no significance.

Supplementary Tables 2 and 3 show the mutations detected in colonies on selection plates. Untargeted base substitution and large deletion mutations were the major mutations for the THF-plasmid DNA, both with and without the knockdown (Table 1). Due to the multiple substitutions in single colonies, the percentages of substitution exceeded 100% for the THF-plasmid DNA. In addition to the mutant frequencies, the mutation spectra were also similar in the control and knockdown cells.

Table 1. Spectra of mutations detected in the supF gene^a.

		control			si-APE1
		G	THF		G	THF
transition
	A:T -> G:C	0 (0)	1 (1)		1 (2)	0 (0)
	G:C -> A:T	12 (19)	61 (48)		12 (19)	68 (54)
transversion
	A:T -> T:A	4 (6)	0 (0)		1 (2)	3 (2)
	A:T -> C:G	0 (0)	1 (1)		5 (8)	0 (0)
	G:C -> T:A	25 (40)	40 (32)		9 (14)	38 (30)
	G:C -> C:G	18 (29)	52 (41)		17 (27)	51 (40)
small insertion (1-2 bp)		2 (3)	1 (1)		1 (2)	3 (2)
large insertion (> 2 bp)		7 (11)	8 (6)		2 (3)	11 (9)
small deletion (1-2 bp)		0 (0)	1 (1)		0 (0)	1 (1)
large deletion (> 2 bp)		28 (44)	54 (43)		30 (48)	54 (43)
others		0 (0)	0 (0)		0 (0)	0 (0)
total mutations		96	219		78	230
total colonies analyzed		63 (100)	126 (100)		63 (100)	127 (100)
aAll data are represented as cases found (%).

Analysis of untargeted base substitution mutations

We next calculated the frequencies of the untargeted base substitutions induced by THF, focusing on those at G:C pairs. According to the proposed mechanism, the deaminase(s) would catalyze the conversion reaction for the C bases on the DNA strand that was complementary to the THF-containing strand of the transfected plasmid (Chen et al., 2014). Given that the strand cleavage by APE1 is essential for the exposure of ss DNA and the subsequent C to U conversion by APOBEC deaminase(s), we expected that the frequency would be reduced by the APE1 knockdown.

The total numbers of substitutions at the G:C pairs were 153 and 157 among the 126 and 127 colonies analyzed for the control and knockdown cells, respectively. Thus, their frequencies were 1.0 × 10^-2 and 1.3 × 10^-2, respectively. Moreover, most of the substitutions were found in the 5′-GpA-3′ dinucleotides: 119 and 130 mutations for the control and knockdown cells, respectively (Table 2). Therefore, the mutation frequencies at the 5′-GpA-3′ sites were calculated to be 7.7 × 10^-3 in the control cells and 1.1 × 10^-2 in the knockdown cells. The APE1 knockdown slightly increased, rather than decreased, the untargeted substitution frequency at the 5′-GpA-3′ sites. These results suggested that the strand cleavage by APE1 is not essential for the untargeted substitutions.

Table 2. The 3′-neighboring bases of substitutions at G sites^a.

	control			si-APE1
	G	THF		G	THF
GA	16	119		8	130
GG	6	15		4	18
GC	2	5		0	0
GT	0	3		1	2

^aSubstitutions in 5′-GN-3′ sites on the upper strand are shown.

Analysis of large deletion mutations

We finally analyzed the large deletion mutations found in the THF-plasmid DNA isolated from the supF mutant colonies. The supF mutant frequencies were multiplied by the ratios of large deletions, and were 3.5 × 10^-3 and 4.9 × 10^-3 for the control and knockdown cells, respectively. Again, the APE1 knockdown slightly promoted the deletion mutations.

We previously reported that the deleted portions and junctions showed little, if any, sequence homology (Suzuki et al., 2018). Likewise, we found no similarity in these sequences, suggesting that most of the deletions were generated by one or more homology-independent event(s) (data not shown).

DISCUSSION

The members of the APOBEC family of cytosine deaminases are considered to inactivate viral genomes (nucleic acids) and promote immunoglobulin class-switching plus somatic hypermutations, suggesting that the enzymes play beneficial roles (Green and Weitzman, 2019). However, the evidence that the deaminases elicit mutagenesis and carcinogenesis shows their dark side for organisms (Chan et al., 2015; Green and Weitzman, 2019). Thus, it is important to understand the relationships between mutations and the endogenous mutagens, the APOBEC family of cytosine deaminases. In this study, we focused on THF, the chemically stable analog of an abasic site that has been used in most biochemistry and molecular biology studies, and APE1 involved in the base excision repair.

As shown in Fig. 2 and discussed above, APE1 seemed to be unessential for the untargeted mutations induced by THF. This was in contrast to the report by Chen et al. (2014) showing that the APE1 knockdown lowered the untargeted mutations in the cases of DNAs containing a U:G or T:G pair. Since the U and T bases are substrates of monofunctional DNA glycosylases, the true abasic sites would be formed. APE1 is reportedly the major repair enzyme that cleaves the DNA strand containing the abasic site and THF (Demple et al., 1991; Robson and Hickson, 1991; Seki et al., 1992; Xanthoudakis et al., 1992; Wilson et al., 1995). Thus, our initial hypothesis was that the untargeted mutations induced by THF would be decreased when APE1 was knocked down. However, this expectation was not proved, as shown in this study.

The signature of the untargeted base substitutions observed in this study was those at the 5′-GpA-3′ dinucleotides complementary to the preferred substrate of APOBEC deaminases, in agreement with the report by Chen et al. Thus, the question raised is which enzyme(s) triggers the strand cleavage of the THF-bearing strand in cells. In the case of THF, the lyase activity of bifunctional DNA glycosylases fails to incise the DNA strand. Human cells contain a second AP endonuclease, APE2 (APEX2) (Hadi and Wilson, 2000; Ide et al., 2003). This protein may incise the THF-strand in the knockdown cells. However, this possibility seems remote, since the AP endonuclease activity of the protein is weak (Burkovics et al., 2006).

Previously, Kitsera et al. (2019) reported that THF is a substrate for human nucleotide excision repair (NER). In NER, the lesion-containing strand is nicked at the 5′- and 3′-sides of the lesion (Kusakabe et al., 2019). The MMR proteins may “hijack” the dually nicked or gapped intermediate. Thus, the NER proteins could compensate for the less efficient cleavage due to the knocked-down APE1.

Moreover, MMR may directly act on the THF-containing strand. Glaab et al. (1999) reported that methyl methanesulfonate, which forms unstable methylated bases, induced more mutations in human MMR-deficient cancer cell lines than MMR-proficient cell lines. The authors suggested that MMR may repair the true abasic sites generated from the unstable bases. If THF is also a substrate for MMR, then the canonical function of MMR would facilitate the untargeted substitution mutations.

We previously demonstrated that the untargeted mutations by THF are dependent on the direction of replication. The mutations occur more frequently when THF is on the lagging strand template than on the leading strand template (Suzuki et al., 2019). Thus, the untargeted mutations induced by THF may have been generated by one or more mechanisms that differ, at least in part, from that proposed by Chen et al. (2014).

The discrepancy between the studies by Chen et al. and us could be derived from difference in DNA lesions, T/U:G and THF:C, respectively. As described above, the T/U bases opposite G are possibly removed by the respective specific monofunctional DNA glycosylases to leave the true abasic sites opposite G. We placed THF opposite C on the assumption that modified G bases are processed by DNA glycosylases. Thus, the bases in the opposite position in addition to the chemical structures are different in the two studies. Moreover, difference in cell lines used should be considered: uterine cervical carcinoma-derived HeLa (Chen et al.) and bone osteosarcoma-derived U2OS cells (this study). Protein expressions depend on cell types and culture conditions. These differences could affect recognition by proteins involved in the untargeted mutations and lead to the contrasting consequences of APE1 knockdown in the two studies.

Interestingly, the supF mutant frequency and the frequencies of substitutions at the G:C pairs, ones at the G bases of 5′-GpA-3′ sites, and large deletions seem higher in the knockdown cells than in the control cells. We observed that the growth rate of the knockdown cells was lower than that of the control cells (data not shown). Given the fact that the APE1 knockout causes early embryonic lethality in mice (Xanthoudakis et al., 1996), this protein has highly important role(s) in cell growth and the knockdown had impacts on U2OS cells. For example, unrepaired THF might induce replication stress and DNA damage response(s). Consequently, various types of mutations might be increased in the stressed cells. In addition, the slow growth rate indicates that decrease of APE1 by siRNA affected cellular function(s) in U2OS cells.

Abasic sites, produced in ss DNA by the actions of APOBEC deaminases followed by uracil DNA glycosylase, can induce all types of mutations at G:C pairs, unlike the U bases that mainly induce G:C→A:T mutations. Since G:C→T:A and G:C→C:G mutations were induced at 5′-GpA-3′ motifs in the THF-bearing plasmid, abasic sites would be generated in the ss DNA region. APE1 acts on abasic sites not only in double-stranded DNA but also ss DNA in vitro (Marenstein et al., 2004). However, the APE1 knockdown had minimal effects on the mutation spectra. This result suggested that in cultured cells, APE1 might not act on abasic sites in ss DNA.

The abasic lesion is a major form of DNA damage and induces various types of mutations in mammalian cells (Kamiya et al., 1992; Gentil et al., 1992; Kamiya et al., 1993; Cabral Neto et al., 1994). Understanding its roles in mutagenesis, carcinogenesis, and cell death requires quantitative analyses, in addition to investigations of its mutational properties. Probes for the abasic site have been developed, and their use in extensive analyses would provide important information about this type of DNA damage (Ide et al., 1993; Kojima et al., 2009).

In conclusion, the APE1 knockdown had negligible impacts on the untargeted substitution and large deletion mutations caused by THF. Since the untargeted substitutions occurred at the G bases of 5′-GpA-3′ sequences, the involvement of the APOBEC-family of cytosine deaminases is still expected. The reasons for the unchanged mutation frequency and signature must be examined in further studies. Recently, we found that untargeted base substitutions induced by 8-hydroxyguanine, another major DNA lesion, were formed at the G bases of 5′-GpA-3′ dinucleotides (Suzuki et al., unpublished results). Thus, more detailed studies are required to understand the molecular mechanisms of the untargeted mutagenesis by DNA lesions.

ACKNOWLEDGMENTS

This work was supported in part by the Japan Society for the Promotion of Science (JSPS) KAKENHI grant numbers JP 16H02956 and JP 19H04278 (HK), and JP 17K12824 (TS).

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

•  Burkovics,  P.,  Szukacsov,  V.,  Unk,  I. and  Haracska,  L. (2006): Human Ape2 protein has a 3′-5′ exonuclease activity that acts preferentially on mismatched base pairs. Nucleic Acids Res., 34, 2508-2515.
•  Cabral Neto,  J.B.,  Cabral,  R.E.,  Margot,  A.,  Le Page,  F.,  Sarasin,  A. and  Gentil,  A. (1994): Coding properties of a unique apurinic/apyrimidinic site replicated in mammalian cells. J. Mol. Biol., 240, 416-420.
•  Cao,  L.,  Kawai,  H.,  Sasatani,  M.,  Iizuka,  D.,  Masuda,  Y.,  Inaba,  T.,  Suzuki,  K.,  Ootsuyama,  A.,  Umata,  T.,  Kamiya,  K. and  Suzuki,  F. (2014): A novel ATM/TP53/p21-mediated checkpoint only activated by chronic γ-irradiation. PLoS One, 9, e104279.
•  Chan,  K.,  Roberts,  S.A.,  Klimczak,  L.J.,  Sterling,  J.F.,  Saini,  N.,  Malc,  E.P.,  Kim,  J.,  Kwiatkowski,  D.J.,  Fargo,  D.C.,  Mieczkowski,  P.A.,  Getz,  G. and  Gordenin,  D.A. (2015): An APOBEC3A hypermutation signature is distinguishable from the signature of background mutagenesis by APOBEC3B in human cancers. Nat. Genet., 47, 1067-1072.
•  Chen,  J.,  Miller,  B.F. and  Furano,  A.V. (2014): Repair of naturally occurring mismatches can induce mutations in flanking DNA. eLife, 3, e02001.
•  Demple,  B.,  Herman,  T. and  Chen,  D.S. (1991): Cloning and expression of APE, the cDNA encoding the major human apurinic endonuclease: definition of a family of DNA repair enzymes. Proc. Natl. Acad. Sci. USA, 88, 11450-11454.
•  Gentil,  A.,  Cabral-Neto,  J.B.,  Mariage-Samson,  R.,  Margot,  A.,  Imbach,  J.L.,  Rayner,  B. and  Sarasin,  A. (1992): Mutagenicity of a unique apurinic/apyrimidinic site in mammalian cells. J. Mol. Biol., 227, 981-984.
•  Glaab,  W.E.,  Tindall,  K.R. and  Skopek,  T.R. (1999): Specificity of mutations induced by methyl methanesulfonate in mismatch repair-deficient human cancer cell lines. Mutat. Res., 427, 67-78.
•  Green,  A.M. and  Weitzman,  M.D. (2019): The spectrum of APOBEC3 activity: from anti-viral agents to anti-cancer opportunities. DNA Repair (Amst.), 83, 102700.
•  Hadi,  M.Z. and  Wilson,  D.M. 3rd. (2000): Second human protein with homology to the Escherichia coli abasic endonuclease exonuclease III. Environ. Mol. Mutagen., 36, 312-324.
•  Ide,  H.,  Akamatsu,  K.,  Kimura,  Y.,  Michiue,  K.,  Makino,  K.,  Asaeda,  A.,  Takamori,  Y. and  Kubo,  K. (1993): Synthesis and damage specificity of a novel probe for the detection of abasic sites in DNA. Biochemistry, 32, 8276-8283.
•  Ide,  Y.,  Tsuchimoto,  D.,  Tominaga,  Y.,  Iwamoto,  Y. and  Nakabeppu,  Y. (2003): Characterization of the genomic structure and expression of the mouse Apex2 gene. Genomics, 81, 47-57.
•  Ichikawa,  J.,  Tsuchimoto,  D.,  Oka,  S.,  Ohno,  M.,  Furuichi,  M.,  Sakumi,  K. and  Nakabeppu,  Y. (2008): Oxidation of mitochondrial deoxynucleotide pools by exposure to sodium nitroprusside induces cell death. DNA Repair (Amst.), 7, 418-430.
•  Ikehata,  H. and  Ono,  T. (2011): The mechanisms of UV mutagenesis. J. Radiat. Res. (Tokyo), 52, 115-125.
•  Kamiya,  H. (2003): Mutagenic potentials of damaged nucleic acids produced by reactive oxygen/nitrogen species: approaches using synthetic oligonucleotides and nucleotides: survey and summary. Nucleic Acids Res., 31, 517-531.
•  Kamiya,  H. (2020): High-throughput analysis of DNA repair in microplates towards identification of inhibitors. Genes Environ., 42, 11.
•  Kamiya,  H. and  Kasai,  H. (1997): Substitution and deletion mutations induced by 2-hydroxyadenine in Escherichia coli: effects of sequence contexts in leading and lagging strands. Nucleic Acids Res., 25, 304-310.
•  Kamiya,  H.,  Suzuki,  M.,  Komatsu,  Y.,  Miura,  H.,  Kikuchi,  K.,  Sakaguchi,  T.,  Murata,  N.,  Masutani,  C.,  Hanaoka,  F. and  Ohtsuka,  E. (1992): An abasic site analogue activates a c-Ha-ras gene by a point mutation at modified and adjacent positions. Nucleic Acids Res., 20, 4409-4415.
•  Kamiya,  H.,  Suzuki,  M. and  Ohtsuka,  E. (1993): Mutation-spectrum of a true abasic site in codon 12 of a c-Ha-ras gene in mammalian cells. FEBS Lett., 328, 125-129.
•  Kamiya,  H.,  Makino,  T.,  Suzuki,  T.,  Kobayashi,  M. and  Matsuoka,  I. (2018): Mutations induced by 8-oxo-7,8-dihydroguanine in WRN- and DNA polymerase λ-double knockdown cells. Mutagenesis, 33, 301-310.
•  Kitsera,  N.,  Rodriguez-Alvarez,  M.,  Emmert,  S.,  Carell,  T. and  Khobta,  A. (2019): Nucleotide excision repair of abasic DNA lesions. Nucleic Acids Res., 47, 8537-8547.
•  Kojima,  N.,  Takebayashi,  T.,  Mikami,  A.,  Ohtsuka,  E. and  Komatsu,  Y. (2009): Construction of highly reactive probes for abasic site detection by introduction of an aromatic and a guanidine residue into an aminooxy group. J. Am. Chem. Soc., 131, 13208-13209.
•  Kusakabe,  M.,  Onishi,  Y.,  Tada,  H.,  Kurihara,  F.,  Kusao,  K.,  Furukawa,  M.,  Iwai,  S.,  Yokoi,  M.,  Sakai,  W. and  Sugasawa,  K. (2019): Mechanism and regulation of DNA damage recognition in nucleotide excision repair. Genes Environ., 41, 2.
•  Lindahl,  T. (1993): Instability and decay of the primary structure of DNA. Nature, 362, 709-715.
•  Loeb,  L.A. and  Harris,  C.C. (2008): Advances in chemical carcinogenesis: a historical review and prospective. Cancer Res., 68, 6863-6872.
•  Loeb,  L.A. and  Preston,  B.D. (1986): Mutagenesis by apurinic/apyrimidinic sites. Annu. Rev. Genet., 20, 201-230.
•  Marenstein,  D.R.,  Wilson,  D.M. 3rd and  Teebor,  G.W. (2004): Human AP endonuclease (APE1) demonstrates endonucleolytic activity against AP sites in single-stranded DNA. DNA Repair (Amst.), 3, 527-533.
•  Nakamura,  J. and  Swenberg,  J.A. (1999): Endogenous apurinic/apyrimidinic sites in genomic DNA of mammalian tissues. Cancer Res., 59, 2522-2526.
•  Obata,  F.,  Nunoshiba,  T.,  Hashimoto-Gotoh,  T. and  Yamamoto,  K. (1998): An improved system for selection of forward mutations in an Escherichia coli supF gene carried by plasmids. J. Radiat. Res. (Tokyo), 39, 263-270.
•  Robson,  C.N. and  Hickson,  I.D. (1991): Isolation of cDNA clones encoding a human apurinic/apyrimidinic endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III) mutants. Nucleic Acids Res., 19, 5519-5523.
•  Satou,  K.,  Hori,  M.,  Kawai,  K.,  Kasai,  H.,  Harashima,  H. and  Kamiya,  H. (2009): Involvement of specialized DNA polymerases in mutagenesis by 8-hydroxy-dGTP in human cells. DNA Repair (Amst.), 8, 637-642.
•  Seki,  S.,  Hatsushika,  M.,  Watanabe,  S.,  Akiyama,  K.,  Nagao,  K. and  Tsutsui,  K. (1992): cDNA cloning, sequencing, expression and possible domain structure of human APEX nuclease homologous to Escherichia coli exonuclease III. Biochim. Biophys. Acta, 1131, 287-299.
•  Suzuki,  T. and  Kamiya,  H. (2017): Mutations induced by 8-hydroxyguanine (8-oxo-7,8-dihydroguanine), a representative oxidized base, in mammalian cells. Genes Environ., 39, 2.
•  Suzuki,  T.,  Harashima,  H. and  Kamiya,  H. (2011): Unexpectedly weak impacts of decreased p53 and retinoblastoma protein levels on mutagenesis by 8-oxo-7,8-dihydroguanine (8-hydroxyguanine). Genes Environ., 33, 103-108.
•  Suzuki,  T.,  Katayama,  Y.,  Komatsu,  Y. and  Kamiya,  H. (2018): Analysis of large deletion mutations induced by abasic site analog in human cells. Genes Environ., 40, 24.
•  Suzuki,  T.,  Katayama,  Y.,  Komatsu,  Y. and  Kamiya,  H. (2019): Large deletions and untargeted substitutions induced by abasic site analog on leading versus lagging strand templates in human cells. Mutagenesis, 34, 421-429.
•  Wilson,  D.M. 3rd,  Takeshita,  M.,  Grollman,  A.P. and  Demple,  B. (1995): Incision activity of human apurinic endonuclease (Ape) at abasic site analogs in DNA. J. Biol. Chem., 270, 16002-16007.
•  Xanthoudakis,  S.,  Miao,  G.,  Wang,  F.,  Pan,  Y.-C. and  Curran,  T. (1992): Redox activation of Fos-Jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J., 11, 3323-3335.
•  Xanthoudakis,  S.,  Smeyne,  R.J.,  Wallace,  J.D. and  Curran,  T. (1996): The redox/DNA repair protein, Ref-1, is essential for early embryonic development in mice. Proc. Natl. Acad. Sci. USA, 93, 8919-8923.