Inhibition of Uracil DNA Glycosylase Alters Frequency and Spectrum of Action-at-a-Distance Mutations Induced by 8-Oxo-7,8-dihydroguanine

Abstract

The generation of DNA damage causes mutations and consequently cancer. Reactive oxygen species are important sources of DNA damage and some mutation signatures found in human cancers. 8-Oxo-7,8-dihydroguanine (G^O, 8-hydroxyguanine) is one of the most abundant oxidized bases and induces a G→T transversion mutation at the modified site. The damaged G base also causes untargeted base substitution mutations at the G bases of 5′-GpA-3′ dinucleotides (action-at-a-distance mutations) in human cells, and the cytosine deaminase apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3 (APOBEC3) is involved in the mutation process. The deaminated cytosine, i.e., uracil, bases are expected to be removed by uracil DNA glycosylase. Most of the substitution mutations at the G bases of 5′-GpA-3′ might be caused by abasic sites formed by the glycosylase. In this study, we expressed the uracil DNA glycosylase inhibitor from Bacillus subtilis bacteriophage PBS2 in human U2OS cells and examined the effects on the G^O-induced action-at-a-distance mutations. The inhibition of uracil DNA glycosylase increased the mutation frequency, and in particular, the frequency of G→A transitions. These results indicated that uracil DNA glycosylase, in addition to APOBEC3, is involved in the untargeted mutation process induced by G^O.

INTRODUCTION

Oxidation of DNA and its precursors results in mutations and consequently cancer formation.^1,2) The oxidation potential of the G base is the lowest among those of the four canonical DNA bases.^3,4) Thus, various damaged bases are formed from G by reactive oxygen species. One of the major damaged G bases is 8-oxo-7,8-dihydroguanine (G^O, also known as 8-hydroxyguanine).^5,6) The G^O base generated in DNA causes G:C→T:A (G→T) transversion mutations at the modified positions in mammalian cells.^7–12)

The G^O base also induces base substitution mutations at positions apart from the modified site in human cells.^13–15) Most of these untargeted mutations (hereafter “action-at-a-distance mutations”) are found at the G bases of 5′-GpA-3′ dinucleotides on the strand where G^O was located. The knockdown of 8-oxoguanine DNA glycosylase 1 (OGG1) decreases the mutations, although OGG1 is the major repair enzyme that catalyzes G^O removal in the base excision repair (BER) pathway (OGG1 paradox).¹⁶⁾ The G bases of 5′-GpA-3′ sequences are complementary to the C bases of 5′-TpC-3′, which the cytosine deaminase apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3 (APOBEC3) preferentially deaminates (except for APOBEC3G).¹⁷⁾ Indeed, the knockdown of APOBEC3B reduces the action-at-a-distance mutations.¹⁸⁾ Thus, OGG1 and APOBEC3 are involved in the action-at-a-distance mutations induced by G^O in human cells (Fig. 1).

Fig. 1. Possible Mechanism of Action-at-a-Distance Mutation Induced by G^O, Based on the Experimental Results^16,18) and the Proposal by Chen et al.²¹⁾

The G:C→Y:Z mutation was frequently found at 5′-GpA-3′/5′-TpC-3′ sites. Please note that this study aimed to examine the involvement of uracil DNA glycosylase (step 4).

OGG1 removes G^O by its DNA glycosylase activity, and the DNA strand containing an abasic site is then cleaved. This nick formation is expected to be conducted by AP endonuclease 1 (APE1)^19,20) (Fig. 1, step 1). The nicked DNA is then believed to be degraded, and the DNA strand complementary to the original G^O strand is exposed as a single-stranded (ss) gap²¹⁾ (Fig. 1, step 2). This exposure should accelerate the deamination, since APOBEC3 prefers C bases on ss DNA rather than double-stranded (ds) DNA²²⁾ (Fig. 1, step 3).

The deamination reaction of the C base produces the U base. Although the U base is one of the canonical bases in RNA, its presence in DNA corresponds to the damage that leads to mutations. The deamination reactions can spontaneously occur, besides the function of C deaminases such as APOBEC3.^6,23) Moreover, the incorporation of 2′-deoxyuridine 5′-triphosphate by DNA polymerases (pols) is another pathway for U accumulation. As in G^O removal, the U bases are removed by proteins in BER. In mammalian cells, two DNA glycosylases perform this function. UNG is the major uracil DNA glycosylase expressed in human cells.²⁴⁾ Two isoforms are produced from the human UNG gene, and the UNG1 and UNG2 proteins localize in the mitochondria and nucleus, respectively.²⁵⁾ UNG2 excises the U bases on ss and ds DNA in the order of U on ss DNA > U:G > U:A in vitro.^26,27) Single-strand specific monofunctional uracil DNA glycosylase 1 (SMUG1) is the other uracil DNA glycosylase.²⁸⁾ SMUG1 also prefers the U bases on ss DNA, followed by those paired with G and A.^29,30) Human UNG2 has a more than 100-fold higher k_cat/K_m value than human SMUG1 against U on ss oligodeoxyribonucleotides (ODNs) in vitro.²⁷⁾ Thus, UNG2 is the DNA glycosylase responsible for the removal of U on ss DNA.

Uracil DNA glycosylase inhibitor (UGI) from Bacillus subtilis bacteriophage PBS2 inhibits uracil DNA glycosylases from Escherichia coli and other species, including human.³¹⁾ In contrast, the human SMUG1 gene was cloned as the homolog of the Xenopus SMUG1 gene, which is not inhibited by UGI from B. subtilis bacteriophage PBS1 (identical to PBS2-UGI).^28,32) Indeed, the activities of human SMUG1 for oxidized thymine bases such as 5-formyluracil are not affected by UGI.²⁹⁾ Thus, UGI can suppress the UNG2 activity but not the SMUG1 activity in human cells.

In the pathway of action-at-a-distance mutations induced by G^O, the U bases are expected to be removed, at least in part, by uracil DNA glycosylase (Fig. 1, step 4). The products are abasic sites, since uracil DNA glycosylase is a monofunctional glycosylase enzyme. The exposed ss gap containing U and/or abasic site(s) is expected to be filled by DNA pols (Fig. 1, step 5). The abasic site is a non-informative lesion, and thus various mutations are induced.³³⁾ Accordingly, the activity of uracil DNA glycosylase in cells influences the frequency and spectrum of action-at-a-distance mutations.

In this study, we established human cells that stably express UGI in a doxycycline (dox)-dependent manner. The plasmid DNA containing a G^O base was transfected into the UGI-expressing cells and the induced mutations were analyzed. The UGI expression increased the frequency of action-at-a-distance mutations by the promotion of G→A transitions.

MATERIALS AND METHODS

Materials

The human osteosarcoma U2OS cell line (ATCC HTB-96) was purchased from American Type Culture Collection (Manassas, VA, U.S.A.). The E. coli CJ236 strain was provided by the National Institute of Genetics (Mishima, Shizuoka, Japan, National Bioresource Project E. coli strain). 5′-Phosphorylated ODNs containing G or G^O for plasmid construction were synthesized and purified by HPLC, as described previously¹⁴⁾ (Supplementary Table 1). ODNs used as PCR primers were purchased from Integrated DNA Technologies (Coralville, IA, U.S.A.) and Hokkaido System Sciences (Sapporo, Japan) (Supplementary Table 1). The gene fragment encoding UGI plus the SV40 nuclear localization signal (NLS) fused to the C-terminus, with humanized codons, was synthesized as gBlocks gene fragments (Integrated DNA Technologies). The pTetOne vector was purchased from Clontech (Palo Alto, CA, U.S.A.). The pSB189KL-BC12(D12) plasmid containing the supF gene was previously described¹⁵⁾ (Supplementary Fig. 1).

Establishment of Dox-Inducible UGI-Expression U2OS Cells

The right and left homologous arms to the targeted AAVS1 locus were amplified by PCR, using TK6 genomic DNA as the template and the primers AAVS HA-R Fw and Rv, and AAVS1 HA-L Fw and Rv, respectively. These right and left homologous arms were inserted into the Pci I and Aat II sites of pTetOne by HotFusion, respectively,³⁴⁾ and the resultant plasmid was named pTetOneHR. The puromycin resistance gene flanked by the piggyBac inverted terminal repeats was inserted into the Sal I site of pTetOneHR, and the resultant plasmid was named pTetOneHR-puro. The codon-humanized UGI-NLS gene fragment (Supplementary Fig. 2) was cloned into the Not I site of pTetOneHR-puro. The plasmid was co-transfected with a CRISPR-Cas9 plasmid that expresses HypaCas9³⁵⁾ and the sgRNA targeting the AAVS1 sequence (5ʹ-GGGGCCACTAGGGACAGGAT-3′) in U2OS cells. After puromycin selection, single clones were obtained by limiting dilution, and the knock-in of the TetOn-UGI expression cassette into the targeted AAVS1 locus was confirmed by PCR, using the primers AAVS1HR check Fw1 and CpGf-SV40pAn Rv for the left arm side, and PGKpA Fw1 and AAVS1HR check Rv1 for the right arm side. The established cell line was named U2OS TetOn-UGI.

Uracil DNA Glycosylase Assay Using U2OS TetOn-UGI Cell Extracts

U2OS TetOn-UGI cells (2.5 × 10⁵ cells/well in 6-well plates) were cultured for 72 h with or without dox at a final concentration of 50 ng/mL. The cells were rinsed twice with D-phosphate buffered saline (PBS)(−), and then lysed with 300 µL of CelLytic M (Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) containing Protease Inhibitor Cocktail for Use with Mammalian Cell and Tissue Extracts (Nacalai Tesque, Kyoto, Japan). The lysed cells were centrifuged for 15 min at 15000 × g and 4 °C. The supernatant was used for the uracil DNA glycosylase assay. The total protein concentration in the supernatant was quantified by Bradford assays using XL-Bradford (APRO Science, Tokushima, Japan) according to the supplier’s instructions.

Ss circular pSB189L-BsmBI (5521 bases) was prepared using JM109 (dut⁺, ung⁺) and CJ236 (dut⁻, ung⁻) as the E. coli hosts, as described previously.³⁶⁾ A total of 200 ng of ss circular pSB189L-BsmBI was incubated with 4 µg of each cell extract for 30 min at 37 °C in a final volume of 16 µL, in reaction buffer (50 mM N-(2-hydroxyethyl)piperazine-N′-2-ethanesulfonic acid (HEPES)-KOH, pH 7.8, 5 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT)). The reactions were analyzed by 1% agarose gel electrophoresis.

Construction and Transfection of the Plasmids Containing G/G^O

The pSB189KL-BC12(D12) plasmid containing G or G^O at position 176 of the supF gene was prepared as described previously.¹⁵⁾ Briefly, the 5ʹ-phosphorylated ODN with G or G^O was hybridized to the ss pSB189KL-BC12(D12) and then converted to the ds form by Phusion High-Fidelity DNA polymerase and Taq DNA ligase. The ds plasmid DNA was then treated with Dam methyltransferase and T5 exonuclease to generate the E. coli-specific methylation pattern and to digest the ds nicked and linear DNAs, respectively. The DNA was purified with a PureLink PCR Purification kit (Thermo Fisher Scientific, Waltham, MA, U.S.A.) and by ethanol precipitation.

The plasmids were transfected into U2OS TetOn-UGI cells as described previously, with slight modifications.¹⁵⁾ Briefly, U2OS TetOn-UGI cells (1.0 × 10⁵ cells/well in 12-well plates) were cultured for 24 h with or without dox (final concentration 50 ng/mL), and transfected by 200 ng of the plasmid DNA (59 fmol) with Lipofectamine 2000 (Thermo Fisher Scientific). After 48 h culture, the plasmid DNA was extracted from the cells and the unreplicated plasmid was digested with Dpn I, as described previously.³⁷⁾ The replicated plasmids (dissolved in 10 µL) were used for Differential DNA denaturation PCR (3D-PCR) and supF mutation analysis.

3D-PCR

The 3D-PCR was performed as described previously, with slight modifications.³⁸⁾ Primers used for 3D-PCR are shown in Supplementary Table 1. Both the first- and second-round PCRs were performed with KOD One PCR Master Mix (Toyobo, Osaka, Japan), according to the supplier’s instructions. The first-round PCR was performed using 0.5 µL of the replicated plasmid solutions as templates in 25 µL reactions. The cycling parameters were 94 °C for 2 min, followed by 25 cycles of 98 °C for 10 s, 50 °C for 5 s, and 68 °C for 1 s. The amplified PCR products were purified with an EconoSpin column for DNA (Ajinomoto Bio-Pharma, Osaka, Japan). The second-round PCR was performed using 10 ng of the first-round products as templates in 10 µL reactions. The cycling parameters were 25 cycles of 84–87.4 °C for 10 s, 55 °C for 5 s, and 72 °C for 1 s. The amplified DNAs were analyzed by 2% agarose gel electrophoresis.

supF Mutation Analysis

The indicator E. coli RF01 strain was transformed by the replicated plasmid to calculate the supF mutant frequency.³⁹⁾ The mutation spectra of the supF gene were analyzed by sequencing the plasmids obtained from the colonies on the selection plates.

Statistical Analysis

Statistically significant differences in the supF mutant frequencies and the substitution frequencies at 5′-GpA (F_GpA) between the G- and G^O-plasmids were examined by the Student’s t-test. The level of statistical significance was set at p < 0.05.

RESULTS

Human Cells Expressing UGI

We used the B. subtilis bacteriophage PBS2 UGI protein to suppress the UNG2 activity instead of RNA interference since small interfering RNA (siRNA) against UNG2 was frequently toxic for U2OS cells. The gene fragment encoding UGI plus the SV40 NLS, with humanized codons, was chemically synthesized (Supplementary Fig. 2). The Tet-On 3G transactivator, expressed from the human phosphoglycerate kinase (PGK) promoter, binds to the TRE3G promoter when dox, a tetracycline analog, is present, allowing the expression of UGI. These expression cassettes were integrated into the AAVS1 region, a safe harbor on chromosome 19, of U2OS cells (U2OS TetOn-UGI cells).⁴⁰⁾

The binding of UGI to UNG2 is practically irreversible.^41–43) To confirm that UGI inhibited UNG, cell extracts were prepared from U2OS TetOn-UGI cells cultured with or without dox. The ss DNAs obtained from JM109 (dut⁺, ung⁺) and CJ236 (dut⁻, ung⁻) E. coli strains were used as the control and U-containing DNAs, respectively. The DNAs were incubated with the extracts and analyzed by agarose gel electrophoresis. As shown in Fig. 2, the degradation of U-containing DNA was low in the case of the extract prepared from the dox-treated cells, suggesting that the expressed UGI inhibited UNG2 in the cells. The intensity of the DNA band corresponding to the full-length DNA was 5% relative to that of the untreated U-containing DNA in the case of the extract from dox-untreated cells. On the other hand, the relative band intensity was 67% under the “ + dox condition.” These results indicated that most of the uracil DNA glycosylase activity was inhibited by UGI.

Fig. 2. Confirmation of Uracil DNA Glycosylase Inhibition by UGI Expression

The control and U-containing ss DNAs (5521 bases) were incubated at 37 °C for 30 min with extracts from dox-untreated and -treated U2OS TetOn-UGI cells. The reaction mixtures were analyzed by 1% agarose gel electrophoresis. (a) Lane 1, 1 kbp ladder DNA size markers; lanes 2-4, control DNA; lanes 5-7, U-containing DNA; lanes 2 and 5, without extract; lanes 3 and 6, with extract from dox-untreated cells; lanes 4 and 7, with extract from dox-treated cells. The lengths of marker DNAs are also shown. (b) Graphical presentation of relative intensities of DNA bands corresponding to the full-length DNA. Data are expressed as the means + standard errors. Each dot represents the data obtained in a single experiment. ** p < 0.01. “N.S.” indicates not significant.

In addition, the plasmid DNAs (pSB189KL-BC12) containing G/G^O were transfected into the cells with or without dox, and the replicated DNAs were extracted (see the Materials and Methods section). The DNAs were then analyzed by 3D-PCR.³⁸⁾ The dox treatment was expected to increase G:C→A:T transitions and hence A:T pairs in the supF gene, since the removal of U formed from C was probably suppressed. The increase in A:T pairs would decrease the lowest denaturation temperature required for the dissociation of ds DNA during PCR. As shown in Supplementary Fig. 3, the dox addition decreased the lowest denaturation temperatures when the replicated plasmid DNAs were used as the templates. This result supported the proposal that the expression of UGI increases G:C→A:T transitions by inhibiting UNG2 (see Discussion).

Effects of UGI on supF Mutant Frequency

The G^O base was placed 14 bases downstream of the 3′-terminus of the supF gene, as described previously^14,15) (Supplementary Fig. 1). We refer to this position as “position 176” for convenience, although it is outside of the gene. The base substitution mutations at this position do not inactivate the gene and hence no supF mutants would be obtained, allowing us to preferentially select the mutants containing the action-at-a-distance mutations. The G^O-plasmid DNA was prepared by hybridization of the G^O-ODN to the ss circular DNA, extension of the ODN by DNA pol, ligation, and adenine methylation.¹⁵⁾ The DNA was transfected into U2OS TetOn-UGI cells by lipofection, and the replicated DNA was recovered after two days. The indicator E. coli strain RF01 was transformed by the recovered DNA, and supF mutants were selected.³⁹⁾ The supF gene expresses a suppressor tRNA that reads through an amber stop codon by incorporating tyrosine. Only supF mutants form colonies on the selection plates containing nalidixic acid and streptomycin. On the other hand, all bacterial cells form colonies on the titer plates.

The numbers of colonies on the titer agar plates, which semiquantitatively reflected the amounts of plasmid DNAs recovered from the transfected cells, were similar (data not shown). Thus, the UGI expression was not toxic and did not affect the plasmid replication efficiency.

As shown in Fig. 3, the supF mutant frequencies were higher for the G^O-plasmid than the control plasmid. The dox treatment seemed to increase the mutant frequency for the G^O-plasmid, although the difference was statistically insignificant.

Fig. 3. Effects of UGI Expression on the supF Mutant Frequency Induced by G^O

The control G-plasmid and the modified G^O-plasmid were introduced into the dox-untreated and -treated U2OS TetOn-UGI cells. Transfection experiments were performed four times. Data are expressed as the means + standard errors.

Altered Frequency and Spectrum of Action-at-a-Distance Mutations by UGI Expression

Next, we analyzed the mutations in the supF gene by Sanger sequencing (Supplementary Tables 2, 3, Tables 1, 2). The dox treatment enhanced the ratio of base substitution mutations for the G^O-plasmid. In particular, the G:C→A:T transition mutation was increased. When we focused on the substitution mutations of the C or G bases on the upper strand of the gene (the G^O strand of the transfected plasmid), the ratios of the C mutations at 5′-TpC-3′ and the G mutations at 5′-GpA-3′ were highest, in line with the APOBEC3 specificity (Table 3).

Table 1. Overall Mutation Spectra^a)

	G		GO
	−dox	+dox	−dox	+dox
Untargeted substitution
At A:T pair	3 (4)	9 (13)	1 (1)	0 (0)
At G:C pair	108 (154)	78 (111)	75 (107)	167 (239)
Targeted G:C→T:A	N.A.b)	N.A.b)	0 (0)	0 (0)
Small insertion (1–2 bp)	0 (0)	0 (0)	0 (0)	0 (0)
Large insertion (> 2 bp)	0 (0)	0 (0)	0 (0)	1 (1)
Small deletion (1–2 bp)	7 (10)	0 (0)	0 (0)	0 (0)
Large deletion (> 2 bp)	14 (20)	14 (20)	13 (19)	3 (4)
Rearrangement or complex	6 (9)	9 (13)	3 (4)	2 (3)
Unknown	3 (4)	13 (19)	14 (20)	10 (14)
Total mutations	141	123	106	183
Total colonies analyzed	70 (100)	70 (100)	70 (100)	70 (100)

a) All data are represented as cases found (%). b) Not applicable.

Table 2. Untargeted Base Substitution Mutations^a)

	G		GO
	−dox	+dox	−dox	+dox
Transition
A:T- > G:C	0 (0)	0 (0)	0 (0)	0 (0)
G:C- > A:T	21 (40)	25 (50)	18 (29)	83 (75)
Transversion
A:T- > T:A	2 (4)	4 (8)	1 (2)	0 (0)
A:T- > C:G	1 (2)	1 (2)	0 (0)	0 (0)
G:C- > T:A	11 (21)	11 (22)	9 (15)	7 (6)
G:C- > C:G	17 (33)	9 (18)	34 (55)	21 (19)
Total base substitutions	52 (100)	50 (100)	62 (100)	111 (100)

a) All data are represented as cases found (%). Barcode-identical colonies are excluded.

Table 3. Dinucleotide Signatures of Mutations at G and C^a)

	G		GO
	−dox	+dox	−dox	+dox
C mutations
AC	0 (0)	0 (0)	1 (2)	0 (0)
TC	10 (20)	15 (33)	4 (7)	20 (18)
GC	0 (0)	1 (2)	0 (0)	0 (0)
CC	5 (10)	0 (0)	1 (2)	2 (2)
Total C mutations	15 (31)	16 (36)	6 (10)	22 (20)
G mutations
GA	27 (55)	22 (49)	49 (80)	79 (71)
GT	3 (6)	2 (4)	0 (0)	0 (0)
GG	4 (8)	4 (9)	5 (8)	9 (8)
GC	0 (0)	1 (2)	1 (2)	1 (1)
Total G mutations	34 (69)	29 (64)	55 (90)	89 (80)
Total base substitutions at G:C sites	49 (100)	45 (100)	61 (100)	111 (100)

a) The sequence of the upper strand is shown. The percentages are shown in parentheses.

We calculated the product of the total supF mutant frequencies, the base substitution ratios, and the numbers of substitutions at 5′-GpA-3′ per plasmid DNA molecule containing substitution(s)¹⁴⁾ (Fig. 4). The obtained value, F_GpA, indexes the frequency of the action-at-a-distance mutations, and indicated that the dox treatment enhanced the action-at-a-distance mutations induced by the oxidized G base. In particular, the frequency of G→A transitions was promoted, whereas those of G→C and G→T transversions were similar (Table 4). The higher frequency of G→A transitions was the expected result, since the UNG2 inhibition would reduce the conversion of U to abasic site(s) and the incorporation of dATP opposite U by DNA pol(s) would be assumed.

Fig. 4. Effects of UGI Expression on the Frequency of Substitution Mutations at 5′-GpA-3′ Dinucleotides (F_GpA Value)

Data are expressed as the means + standard errors. * p < 0.05.

Table 4. Frequency of Substitution Mutations at 5′-GpA-3′ Sites

		G- > A	G- > T	G- > C
G	−dox	3.3 × 10−4	2.4 × 10−4	5.3 × 10−4
	+dox	3.8 × 10−4	1.2 × 10−4	3.5 × 10−4
GO	−dox	1.1 × 10−3	8.3 × 10−4	4.0 × 10−3
	+dox	9.0 × 10−3	1.1 × 10−3	3.7 × 10−3

The distribution of base substitutions at 5′-TpC-3′ and 5′-GpA-3′ sites is shown in Supplementary Fig. 4. Mutations at positions 5, 91, and 126 were frequently observed for the G^O-plasmid. In addition, mutations at other locations, such as positions 27 and 73, were increased upon dox treatment. The mutations at these positions frequently accompanied other mutations and predominantly included G→A transitions.

Interestingly, the dox treatment promoted substitutions, especially C→T transitions, at 5′-TpC-3′ sequences (Supplementary Fig. 4). This result would also be caused by the inhibition of UNG2 and the dATP incorporation opposite U.

DISCUSSION

This study aimed to examine whether uracil DNA glycosylase UNG2 is involved in the process of generating the action-at-a-distance mutations induced by G^O. The action-at-a-distance mutations were raised (indicated by the increase in the F_GpA value) and the G→A transition became the major substitution at 5′-GpA-3′ sites by UGI expression (Fig. 4, Table 4). In addition, novel mutational hotspots appeared (Supplementary Fig. 4). These results indicated that UNG2 actually participates the mutagenesis process.

We used the phage UGI protein to suppress the UNG2 activity, rather than RNA interference, since the treatment with siRNA against UNG2 was frequently toxic for U2OS cells. Indeed, the numbers of E. coli colonies indicated that the UGI expression was not toxic and the plasmid replication efficiencies were similar.

The UGI expression promoted the F_GpA value, and especially the frequency of G→A mutations (Fig. 4, Table 4). These results were expected since various nucleotides are incorporated opposite an abasic site, a “non-instructive lesion,” in mammalian cells.³³⁾ When dGTP is incorporated opposite an abasic site, no mutation are induced. In contrast, U can pair with A in the Watson-Crick manner and this U:A pair formation results in the G→A mutation induction.

However, unexpectedly, the frequencies of G→C and G→T transversions were maintained under the UGI expression. These transversions should be reduced upon the decrease of abasic sites and the increase of U if we assume that only dATP is incorporated opposite U. However, given that translesion synthesis (TLS) DNA pols participate in the ss gap filling and that the pols incorporate dTTP and dCTP opposite U, the occurrence of these mutations could be explained. In vitro, human DNA polι incorporates deoxyribonucleotides opposite U in the order of dTTP > dGTP > dATP > dCTP.⁴⁴⁾ The incorporation of dTTP is 4.3-fold more efficient than that of dATP, based on the k_cat/K_m values. In addition, human REV1 is a deoxycytidyl transferase and only incorporates dCTP opposite the template U.⁴⁵⁾ Thus, the G→C and G→T transversions were possibly induced by the TLS DNA pols, even under the assumption that UGI completely inhibited the total uracil glycosylase activity.

As shown in Supplementary Fig. 4, the distribution of the substitutions at the 5′-GpA-3′ sequences was non-uniform. Positions 5, 91, and 126 of the supF gene were the hotspots of action-at-a-distance mutations for the “G^O, –dox” group. In U2OS cells, APOBEC3B is the major APOBEC3 molecule responsible for the action-at-a-distance mutations. APOBEC3B prefers 5′-TCA-3′ and favors 5′-RTCA-3′ over 5′-YTCA-3′ (R = purine and Y = pyrimidine).^46,47) The sequences around positions 5, 91, and 126 are 5′-ATCAA-3′, 5′-GTCTG-3′, and 5′-TTCGA-3′, respectively (shown as those of the lower strand of the gene, which is assumed to be the deaminase substrate). Moreover, the recent biochemical analysis of the enzyme indicated that 5′-TpC-3′ in a loop region is an excellent substrate.⁴⁸⁾ Supplementary Fig. 5 shows the secondary structure of the lower strand of the gene (from positions −30 to 192) predicted by “DNA Folding Form” in the mFold web server (http://www.unafold.org/mfold/applications/dna-folding-form.php).⁴⁹⁾ From this viewpoint, position 5 follows the 5′-RTCA-3′ rule and is predicted to be located in a loop region. This is one of the reasons why this position is a mutational hotspot.

It is noteworthy that positions 27 (5′-ATCGG-3′), 73 (5′-CTCGG-3′), 86 (5′-CTCCC-3′), 118 (5′-TTCGA-3′), and 167 (5′-ATCTT-3′) joined the hotspots by the UNG2 inhibition (Supplementary Fig. 4). Under the assumption that the dox treatment does not affect the APOBEC3 expression and/or activity, this would occur due to the increased number of mutations per one mutant (see below). In particular, the mutations at positions 27, 86, and 167 were not found as sole mutations in single mutants (Supplementary Table 3). Positions 27 and 167 are located in the region corresponding to the pre-tRNA portion of the product and outside of the gene, respectively, and the mutations at these positions are principally unable to inactivate the gene.

Moreover, not all mutants in other regions can form bacterial colonies on the selection plates containing nalidixic acid and streptomycin.⁵⁰⁾ Although positions 27 and 73 and position 86 are predicted to be in loop and non-loop ss regions, respectively, ss DNA binding proteins such as RPA may alter the secondary structure. Thus, additional experiments are required to clarify the hotspot changes.

The UGI expression also affected the number of mutations at 5′-GpA-3′ sites per colony obtained by transfection of the G^O-plasmid. For the –dox and +dox groups, 20 among the 31 colonies and 15 among the 33 colonies carried a single substitution, respectively (Supplementary Table 4). In addition, two and one colonies had six and eleven substitutions, respectively, for the +dox group, although the maximum number of mutations was five for the –dox group. This may reflect the higher mutagenic property of U than an abasic site: the incorporation of dGTP by DNA pol(s) opposite an abasic site does not induce a mutation.

The 3D-PCR was conducted to confirm that UGI inhibited uracil DNA glycosylase (Supplementary Fig. 3). The decrease in the lowest denaturation temperatures indicated the increase in the A:T contents, supporting our expectation (the induction of G:C→A:T mutations caused by U). One may think that the F_GpA value for the G-plasmid was not enhanced and that this is contradictory to the decrease in the lowest denaturation temperature (Fig. 4). However, Table 3 and Supplementary Fig. 4 indicate that base substitutions were induced at 5′-TpC-3′ sites for the G-plasmid replicated in the dox-treated cells. The induced mutations explain the lowest denaturation temperature shift, at least in part, although we do not know the exact reason(s) for this phenomenon. In addition, antibiotic resistance was used for the supF mutant selection, and this may be another reason for the biased increase of substitutions at 5′-TpC-3′ sites of the G-plasmid.

In conclusion, the UGI expression, probably due to the inhibition of the major uracil DNA glycosylase UNG2, altered the spectrum of action-at-a-distance mutations in human cells. UNG2 would catalyze the U to abasic site conversion and consequently reduce the frequency of G→A transitions. This study indicates that abasic sites formed from U bases on putative ss DNA gaps are important for mutation spectrum determination. Further studies are in progress in our laboratory to reveal the detailed mechanism of the action-at-a-distance mutations.

Acknowledgments

This work was supported in part by the Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant Numbers: JP 19H04278, JP 22H03751, and JP 23K25005 to H.K., and JP 20K12181 to T.S.).

Author Contributions

H.K. wrote the first draft of the paper; T.S. edited the paper; T.S., S.Y., and H.K. designed the research; T.S. and S.Y. performed the research; T.S. and S.Y. analyzed the data. T.S. and H.K. wrote the final version of the paper.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation and disease. FASEB J., 17, 1195–1214 (2003).
• 2) Sekiguchi M, Tsuzuki T. Oxidative nucleotide damage: consequences and prevention. Oncogene, 21, 8895–8904 (2002).
• 3) Seidel CAM, Schulz A, Sauer MHM. Nucleobase-specific quenching of fluorescent dyes. 1. Nucleobase one-electron redox potentials and their correlation with static and dynamic quenching efficiencies. J. Phys. Chem., 100, 5541–5553 (1996).
• 4) Steenken S, Jovanovic SV. How easily oxidizable is DNA? One-electron reduction potentials of adenosine and guanosine radicals in aqueous solution. J. Am. Chem. Soc., 119, 617–618 (1997).
• 5) Kasai H, Nishimura S. Hydroxylation of deoxyguanosine at the C-8 position by ascorbic acid and other reducing agents. Nucleic Acids Res., 12, 2137–2145 (1984).
• 6) Lindahl T. Instability and decay of the primary structure of DNA. Nature, 362, 709–715 (1993).
• 7) Kamiya H, Miura K, Ishikawa H, Inoue H, Nishimura S, Ohtsuka E. c-Ha-ras containing 8-hydroxyguanine at codon 12 induces point mutations at the modified and adjacent positions. Cancer Res., 52, 3483–3485 (1992).
• 8) Moriya M. Single-stranded shuttle phagemid for mutagenesis studies in mammalian cells: 8-oxoguanine in DNA induces targeted G·C→T·A transversions in simian kidney cells. Proc. Natl. Acad. Sci. U.S.A., 90, 1122–1126 (1993).
• 9) Kamiya H, Murata-Kamiya N, Koizume S, Inoue H, Nishimura S, Ohtsuka E. 8-Hydroxyguanine (7,8-dihydro-8-oxoguanine) in hot spots of the c-Ha-ras gene: effects of sequence contexts on mutation spectra. Carcinogenesis, 16, 883–889 (1995).
• 10) Le Page F, Margot A, Grollman AP, Sarasin A, Gentil A. Mutagenicity of a unique 8-oxoguanine in a human Ha-ras sequence in mammalian cells. Carcinogenesis, 16, 2779–2784 (1995).
• 11) Tan X, Grollman AP, Shibutani S. Comparison of the mutagenic properties of 8-oxo-7,8-dihydro-2′-deoxyadenosine and 8-oxo-7,8-dihydro-2′-deoxyguanosine DNA lesions in mammalian cells. Carcinogenesis, 20, 2287–2292 (1999).
• 12) Suzuki T, Kamiya H. Mutations induced by 8-hydroxyguanine (8-oxo-7,8-dihydroguanine), a representative oxidized base, in mammalian cells. Genes Environ., 39, 2 (2017).
• 13) Kamiya H, Yamazaki D, Nakamura E, Makino T, Kobayashi M, Matsuoka I, Harashima H. Action-at-a-distance mutagenesis induced by oxidized guanine in Werner syndrome protein-reduced human cells. Chem. Res. Toxicol., 28, 621–628 (2015).
• 14) Suzuki T, Masuda H, Mori M, Ito R, Kamiya H. Action-at-a-distance mutations at 5′-GpA-3′ sites induced by oxidised guanine in WRN-knockdown cells. Mutagenesis, 36, 349–357 (2021).
• 15) Fukushima R, Suzuki T, Komatsu Y, Kamiya H. Biased distribution of action-at-a-distance mutations by 8-oxo-7,8-dihydroguanine. Mutat. Res. (Fundam. Mol. Mech. Mutagen.) , 825, 111794 (2022).
• 16) Suzuki T, Zaima Y, Fujikawa Y, Fukushima R, Kamiya H. Paradoxical role of the major DNA repair protein, OGG1, in action-at-a-distance mutation induction by 8-oxo-7,8-dihydroguanine. DNA Repair (Amst.), 111, 103276 (2022).
• 17) Pecori R, Di Giorgio S, Lorenzo JP, Papavasiliou FN. Functions and consequences of AID/APOBEC-mediated DNA and RNA deamination. Nat. Rev. Genet., 23, 505–518 (2022).
• 18) Fukushima R, Suzuki T, Kobayakawa A, Kamiya H. Action-at-a-distance mutations induced by 8-oxo-7,8-dihydroguanine are dependent on APOBEC3. Mutagenesis, 39, 24–31 (2024).
• 19) Hill JW, Hazra TK, Izumi T, Mitra S. Stimulation of human 8-oxoguanine-DNA glycosylase by AP endonuclease: potential coordination of the initial steps in base excision repair. Nucleic Acids Res., 29, 430–438 (2001).
• 20) Vidal AE, Hickson ID, Boiteux S, Radicella JP. Mechanism of stimulation of the DNA glycosylase activity of hOGG1 by the major human AP endonuclease: bypass of the AP lyase activity step. Nucleic Acids Res., 29, 1285–1292 (2001).
• 21) Chen J, Miller BF, Furano AV. Repair of naturally occurring mismatches can induce mutations in flanking DNA. eLife, 3, e02001 (2014).
• 22) Green AM, Weitzman MD. The spectrum of APOBEC3 activity: from anti-viral agents to anti-cancer opportunities. DNA Repair (Amst.), 83, 102700 (2019).
• 23) Mosbaugh DW, Bennett SE. Uracil-excision DNA repair. Prog. Nucleic Acid Res. Mol. Biol., 48, 315–370 (1994).
• 24) Aravind L, Koonin EV. The a/b fold uracil DNA glycosylases: a common origin with diverse fates. Genome Biol., 1, research0007.1 (2000).
• 25) Nilsen H, Otterlei M, Haug T, Solum K, Nagelhus TA, Skorpen F, Krokan HE. Nuclear and mitochondrial uracil-DNA glycosylases are generated by alternative splicing and transcription from different positions in the UNG gene. Nucleic Acids Res., 25, 750–755 (1997).
• 26) Parikh SS, Mol CD, Slupphaug G, Bharati S, Krokan HE, Tainer JA. Base excision repair initiation revealed by crystal structures and binding kinetics of human uracil-DNA glycosylase with DNA. EMBO J., 17, 5214–5226 (1998).
• 27) Kavli B, Sundheim O, Akbari M, Otterlei M, Nilsen H, Skorpen F, Aas PA, Hagen L, Krokan HE, Slupphaug G. hUNG2 is the major repair enzyme for removal of uracil from U:A matches, U:G mismatches, and U in single-stranded DNA, with hSMUG1 as a broad specificity backup. J. Biol. Chem., 277, 39926–39936 (2002).
• 28) Haushalter KA, Todd Stukenberg MW, Kirschner MW, Verdine GL. Identification of a new uracil-DNA glycosylase family by expression cloning using synthetic inhibitors. Curr. Biol., 9, 174–185 (1999).
• 29) Masaoka A, Matsubara M, Hasegawa R, Tanaka T, Kurisu S, Terato H, Ohyama Y, Karino N, Matsuda A, Ide H. Mammalian 5-formyluracil-DNA glycosylase. 2. Role of SMUG1 uracil-DNA glycosylase in repair of 5-formyluracil and other oxidized and deaminated base lesions. Biochemistry, 42, 5003–5012 (2003).
• 30) Darwanto A, Theruvathu JA, Sowers JL, Rogstad DK, Pascal T, Goddard W 3rd, Sowers LC. Mechanisms of base selection by human single-stranded selective monofunctional uracil-DNA glycosylase. J. Biol. Chem., 284, 15835–15846 (2009).
• 31) Wang Z, Mosbaugh DW. Uracil-DNA glycosylase inhibitor gene of bacteriophage PBS2 encodes a binding protein specific for uracil-DNA glycosylase. J. Biol. Chem., 264, 1163–1171 (1989).
• 32) Savva R, Pearl LH. Cloning and expression of the uracil–DNA glycosylase inhibitor (UGI) from bacteriophage PBS-1 and crystallization of a uracil–DNA glycosylase–UGI complex. Proteins, 22, 287–289 (1995).
• 33) Gentil A, Cabral-Neto JB, Mariage-Samson R, Margot A, Imbach JL, Rayner B, Sarasin A. Mutagenicity of a unique apurinic/apyrimidinic site in mammalian cells. J. Mol. Biol., 227, 981–984 (1992).
• 34) Fu C, Donovan WP, Shikapwashya-Hasser O, Ye X, Cole RH. Hot Fusion: an efficient method to clone multiple DNA fragments as well as inverted repeats without ligase. PLOS ONE, 9, e115318 (2014).
• 35) Chen JS, Dagdas YS, Kleinstiver BP, Welch MM, Sousa AA, Harrington LB, Sternberg SH, Joung JK, Yildiz A, Doudna JA. Enhanced proofreading governs CRISPR-Cas9 targeting accuracy. Nature, 550, 407–410 (2017).
• 36) Suzuki T, Kamiya H. Easily-controllable helper phage-free single-stranded phagemid production system. Genes Environ., 44, 25 (2022).
• 37) Stary A, Sarasin A. Simian virus 40 (SV40) large T antigen-dependent amplification of an Epstein–Barr virus-SV40 hybrid shuttle vector integrated into the human HeLa cell genome. J. Gen. Virol., 73, 1679–1685 (1992).
• 38) Suspène R, Henry M, Guillot S, Wain-Hobson S, Vartanian JP. Recovery of APOBEC3-edited human immunodeficiency virus G→A hypermutants by differential DNA denaturation PCR. J. Gen. Virol., 86, 125–129 (2005).
• 39) Fukushima R, Suzuki T, Kamiya H. New indicator Escherichia coli strain for rapid and accurate detection of supF mutations. Genes Environ., 42, 28 (2020).
• 40) Sadelain M, Papapetrou EP, Bushman FD. Safe harbours for the integration of new DNA in the human genome. Nat. Rev. Cancer, 12, 51–58 (2012).
• 41) Bennett SE, Mosbaugh DW. Characterization of the Escherichia coli uracil-DNA glycosylase·inhibitor protein complex. J. Biol. Chem., 267, 22512–22521 (1992).
• 42) Bennett SE, Schimerlik MI, Mosbaugh DW. Kinetics of the uracil-DNA glycosylase/inhibitor protein association. Ung interaction with Ugi, nucleic acids, and uracil compounds. J. Biol. Chem., 268, 26879–26885 (1993).
• 43) Assefa NG, Niiranen L, Johnson KA, Leiros HK, Smalås AO, Willassen NP, Moe E. Structural and biophysical analysis of interactions between cod and human uracil-DNA N-glycosylase (UNG) and UNG inhibitor (Ugi). Acta Crystallogr. D Biol. Crystallogr., 70, 2093–2100 (2014).
• 44) Vaisman A, Woodgate R. Unique misinsertion specificity of polι may decrease the mutagenic potential of deaminated cytosines. EMBO J., 20, 6520–6529 (2001).
• 45) Lin W, Xin H, Zhang Y, Wu X, Yuan F, Wang Z. The human REV1 gene codes for a DNA template-dependent dCMP transferase. Nucleic Acids Res., 27, 4468–4475 (1999).
• 46) Burns MB, Lackey L, Carpenter MA, Rathore A, Land AM, Leonard B, Refsland EW, Kotandeniya D, Tretyakova N, Nikas JB, Yee D, Temiz NA, Donohue DE, McDougle RM, Brown WL, Law EK, Harris RS. APOBEC3B is an enzymatic source of mutation in breast cancer. Nature, 494, 366–370 (2013).
• 47) Chan K, Roberts SA, Klimczak LJ, Sterling JF, Saini N, Malc EP, Kim J, Kwiatkowski DJ, Fargo DC, Mieczkowski PA, Getz G, Gordenin DA. An APOBEC3A hypermutation signature is distinguishable from the signature of background mutagenesis by APOBEC3B in human cancers. Nat. Genet., 47, 1067–1072 (2015).
• 48) Sanchez A, Ortega P, Sakhtemani R, Manjunath L, Oh S, Bournique E, Becker A, Kim K, Durfee C, Temiz NA, Chen XS, Harris RS, Lawrence MS, Buisson R. Mesoscale DNA features impact APOBEC3A and APOBEC3B deaminase activity and shape tumor mutational landscapes. Nat. Commun., 15, 2370 (2024).
• 49) Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res., 31, 3406–3415 (2003).
• 50) Kawai H, Iwata R, Ebi S, Sugihara R, Masuda S, Fujiwara C, Kimura S, Kamiya H. Development of a versatile high-throughput mutagenesis assay with multiplexed short read NGS using DNA-barcoded supF shuttle vector library amplified in E. coli. eLife, 11, e83780 (2022).