TH5427

The roles of human MTH1, MTH2 and MTH3 proteins in maintaining genome stability under oxidative stress

Kazunari Hashiguchia,b*, Michio Hayashia, Mutsuo Sekiguchib, and Keiko Umezua

Highlights

• MTH1-KO cells shows only a slight increase in mutant frequency.
• Both MTH1-KO and MTH2-KO are sensitive to hydrogen peroxide.
• MTH3 cannot substitute the E. coli MutT activity.
• MTH3 may possess another function besides the degradation of oxidized nucleotides.

ABSTRACT

The hydrolysis of nucleotides containing 8-oxo-7,8-dihydroguanine (8-oxoG) is important in the maintenance of genome stability. Human cells possess three types of proteins, MTH1 (NUDT1), MTH2 (NUDT15) and MTH3 (NUDT18), which have the potential to hydrolyze deoxyribonucleoside di- and triphosphates containing 8-oxoG to the monophosphate, the form of which is unusable for DNA synthesis. To elucidate the physiological roles of these enzymes, we constructed single knockout (KO) cell lines for each of the MTH1, MTH2 and MTH3 genes and MTH1 and MTH2-double KO cell lines from the human HeLa S3 line using CRISPR/Cas9. With the exception of MTH3-KO, all of the KO cell lines showed similar proliferation rates to the parental line, HeLa S3, indicating that the MTH1 and MTH2 functions are dispensable for cell growth. On the other hand, the MTH3-KO cells showed a significantly slower growth rate, suggesting that MTH3 has a definite role in cell growth in addition to the cleavage of 8-oxoG-containing deoxyribonucleotide. MTH1-KO, MTH2-KO and MTH1- MTH2-KO cells exhibited increased sensitivity to hydrogen peroxide, whereas MTH3-KO did not. MTH1-KO cells showed only a slight increase in mutant frequency in comparison to the parental HeLa S3 line. The overproduction of MTH1 and MTH2 suppressed the mutator phenotype of mutT-deficient E. coli cells, whereas the overproduction of MTH3 did not show such a suppressive effect. Our findings suggest that both MTH1 and MTH2 are involved in the maintaining genome stability in human cells against oxidative stress, while MTH3 may play some other role(s).

Keywords:
8‐oxoguanine (8‐oxoG) genomic instability MutT homolog
nucleoside/nucleotide metabolism Nudix
oxidative stress

1. Introduction

Significant amounts of reactive oxygen species (ROS) are produced as byproducts of oxygen utilization through respiration [1,2]. Although most of these radicals are eliminated by the actions of antioxidant systems, some remain and attack cellular constituents, including proteins, nucleic acids and lipids. Among the oxidized DNA bases thus produced, 8-oxo-7,8-dihydroguanine (8-oxoguanine, 8-oxoG) is particularly important with respect to the maintenance and transfer of genetic information [3,4]. 8-oxoG has the ability to pair with both adenine and cytosine and, thus, induces base-substitution mutations.
Through the action of ROS, the guanine-containing nucleotides in the DNA precursor pool are converted to the oxidized forms of nucleotides, such as 8-oxo-dGTP. Since DNA polymerases hardly discriminate oxidized nucleotides from normal ones, 8-oxoG can be misincorporated into DNA [5,6]. It is therefore important to eliminate oxidized guanine-containing nucleotides from the precursor pool. Furthermore, organisms possess enzymes to degrade 8-oxo-dGTP and 8-oxo-dGDP to 8-oxo-dGMP, a form that is unusable for DNA synthesis. Since guanylate kinase, which converts dGMP to dGDP, is unable to phosphorylate 8-oxo-dGMP, the 8-oxo-dGTPase and guanylate kinase collaborate to prevent the misincorporation of 8-oxoG into DNA [7,8]. MutT protein of Escherichia coli is the first of this type of enzyme [9], and its physiological function has been the most intensively investigated [10,11]. It has been shown that the spontaneous mutant frequency of E. coli mutants with a defective mutT gene is more than 1,000 times higher than that in wild-type cells [12].
8-OxoG-related mutagenesis may account for a considerable part of spontaneous mutagenesis in mammalian cells. At least three human enzymes with activities similar to that of the E. coli MutT have been found, among which MTH1 (MutT Homolog 1) has been studied the most extensively [13,14]. Since the expression of human MTH1 cDNA in E. coli mutT-deficient cells considerably suppresses the increase in the frequency of spontaneous mutations, MTH1 protein may have the same antimutagenic role, similar to MutT. A mouse line with a defective Mth1 gene, constructed by gene targeting, yielded a significantly increased number of tumors in various organs [15]. The spontaneous mutant frequency of the Mth1-/- cell lines, which were established from a gene-targeted mouse, was increased in comparison to that of Mth1+/+ cells.
The comparison of the sequences of MutT and MTH1 proteins revealed that these two proteins are similar in size and carry a highly conserved 23-amino acid sequence in their central regions [14]. This sequence was named the “MutT signature”; it was subsequently found that at least 22 human proteins were found to carry this sequence [16,17]. Since most of these proteins are capable of hydrolyzing compounds with a general structure of nucleoside diphosphate linked to another moiety X, this class of proteins was listed as Nudix proteins. MTH1 was the first of this category of enzymes, and was therefore designated ‘NUDT1’. Although the substrate specificities of most of Nudix proteins have been characterized, some remain uncharacterized; these include NUDT15 and NUDT18. Since the amino acid sequences of these two proteins resemble the sequence of MTH1 (NUDT1), we suspected that these proteins might be capable of acting on 8-oxoG-containing nucleotides. Subsequent studies revealed that NUDT15 and NUDT18 possess activities that cleave 8-oxoG-containing deoxyribonucleoside di- and/or triphosphates; these were named MTH2 and MTH3, respectively [18,19].
Another important aspect of MTH1 is its possible involvement in the progression of cancer. The expression of MTH1 mRNA in cancer was considerably higher than that in adjacent normal tissue [20], and the expression of MTH1 mRNA in advanced-stage tumors was significantly higher than that in early-stage tumors [21]. It was further demonstrated that—in comparison to normal cells—tumor cells were more susceptible to certain chemicals that bind to MTH1 [22,23], implying that the disruption of nucleotide homeostasis via the inhibition of MTH1 by these chemicals may be useful for interrupting tumor progression.
The existence of the three types of MutT-related enzymes in human cells raises a question as to which type would play a major physiological role in maintaining genomic stability. To answer this question, we developed cell lines in which one or two of these genes were defective and studied their properties under conditions of oxidative stress.

2. Materials and methods

2.1. Vector construction

SpCas9- and chimeric guide RNA-expression vector, pX330, was purchased from Addgene. The CRISPR DESIGN algorithm (Feng Zhang’s lab., MIT) was used to search for the CRISPR target sequences for the MTH1, MTH2 and MTH3 genes; their target sequences are shown in Supplementary Fig.1. The oligonucleotide sequences used for the construction of the pX330-based KO vector are shown in Supplementary Table 1. pX330-MTH1KO, which was used for the KO of the MTH1 gene, was a kind gift from T. Ishii (Fukuoka Dental College). For the construction of pX330-MTH2KO and pX330-MTH3KO, annealed oligodeoxyribonucleotides were ligated to pX330, which was linearized by BbsI (Thermo Fischer), according to the protocol provided by Feng Zhang’s laboratory. [24]. The E. coli expression vector, pCold I, was purchased from Takara. pQE-80L-MutT, pCold I-MTH1, pCold I-MTH2 and pQE-80L-MTH3 were a generous gift from R. Ito (Fukuoka Dental College) [18]. For the construction of the pCold I-MutT expression vector, the coding region of the mutT gene was amplified by a PCR using the pQE-80L-MutT plasmid as a template, and inserted into the NdeI/PstI site of pCold I. For the construction of pCold I-MTH3, the coding region of the MTH3 gene was amplified by a PCR using pQE-80L-MTH3 as a template, and inserted into the NdeI/XbaI site of pCold I. The construction of the expression vectors for catalytically inactive mutant proteins was performed according to the site-directed mutagenesis protocol (Stratagene). All of the PCR primers used in this study are listed in Supplementary Table 1.

2.2 Cell culture and transfection

The HeLa S3 cells, which were obtained from our laboratory stock, were maintained in Dulbecco’s modified-MEM (DMEM) containing 4.5 g/L glucose (Wako) supplemented with 10% fetal bovine serum (Sigma) in a humidified atmosphere at 37C and 5% CO2. The cells were routinely sub-cultured twice a week. Transfection was performed using Lipofectamine 2000 (Thermo Fisher) according to the manufacturer’s protocol.

2.3. Isolation of the knockout cell lines

The transfection of pX330-based CRISPR/Cas9 vector was performed as described below. HeLa S3 cells were plated onto 24-well plates at 5104 cells per well on the day before transfection. Two days after transfection, all cells were removed by trypsinization, and 100-200 cells were seeded onto a 100-mm dish. Colonies were grown for approximately 14 days, picked up, and resuspended in a 24-well plate containing 500 l of medium. Cells in sub-confluent growth were plated onto a 24-well plate at 1:10 dilution, and the remaining cells were collected and cell lysates were prepared for Western blotting to screen candidate KO clones.

2.4. The cell proliferation assay

KO cell lines and their parental cell line were plated on a 96-well plate at 1103 cells per well. At 2, 3, 4 and 5 days after plating, the cell proliferation rate was determined with a Cell Counting Kit-8 (CCK-8, Dojindo, Japan), according to the manufacturer’s protocol.

2.5. The hydrogen peroxide survival assay

Cells were plated on a 6-well plate at 1.5105 cells per well. Twenty-four hours after plating, the supernatant was removed and 2 ml of pre-warmed serum-free DMEM containing an appropriate concentration of hydrogen peroxide was added. Hydrogen peroxide treatment was performed at 37C for 30 min in a CO2 incubator. After treatment, the medium was completely removed and fresh complete DMEM was added, and the plate was placed back in the CO2 incubator. At 72 h after treatment, all cells were detached by trypsinization, and the number of viable cells was counted by the trypan blue dye exclusion method. A colony formation assay was performed as follows. Cells were plated on a 60-mm dish at 500 cells per dish. At 24 h after plating, the medium was removed and 5 ml of pre-warmed serum-free DMEM containing an appropriate concentration of hydrogen peroxide was added. The cells were incubated at 37C for 30 min in a CO2 incubator. After treatment, the medium was completely removed and fresh complete DMEM was added and the dishes were placed back in the CO2 incubator. At 14 days after treatment, the colonies were stained with crystal violet, and colonies consisting of more than 50 cells were counted.

2.6. Determination of the mutant frequency

HAT media supplement (50) Hybri-Max and HT media supplement (50) Hybri-Max were purchased from Sigma. 6-TG was obtained from Wako. For the measurement of the mutant frequency, five wild-type sub-lines were randomly isolated from HeLa S3 cells. To eliminate any pre-existing HPRT-defective mutant cells, 5 sub-lines of HeLa S3 cells and 4 lines of MTH1-KO cells were incubated in complete DMEM containing 1HAT (100 M hypoxanthine, 0.4 M aminopterin and 16 M thymidine) for 3 days and then in complete DMEM containing 1HT (100 M hypoxanthine and 16 M thymidine) for 3 days. Each culture was seeded into three independent dishes; three biological replicates were prepared per clone. The cells were incubated for 4 days to allow for the expression of the 6-TG-resistant characteristic. The cells were trypsinized and 2106 cells in complete DMEM were placed onto four 100-mm dishes containing 30 M 6-TG for 14 days. After staining with crystal violet, the number of resistant colonies containing >50 cells was determined. In parallel, cell suspension containing approximately 500 cells was incubated in 3100-mm dishes and the number of viable cells was determined. The mutant frequencies were calculated from 15 and 12 fractions of wild-type and MTH1-KO cells, respectively, and then the mean values were compared. The mutant frequencies were compared using Welch’s t-test.

2.7. Antibody preparation

Rabbit polyclonal primary antibody against MTH2 protein was prepared as described below. Full-length recombinant MTH2 protein was produced and purified according to the method of Takagi et al. [18]. Anti-sera containing anti-MTH2 antibodies was prepared by MBL (Japan). The antibodies were affinity-purified using purified MTH2 protein, and the eluted materials were dialyzed against 2PBS at 4C. Finally, the antibody solution was mixed with one volume of ice-cold glycerol, and stored at -20C.

2.8. Western blotting

The cell pellet that was used for Western blotting was stored at -80C. To prepare whole cell lysate, the cell pellet was lysed in a buffer (50 mM Tris-HCl pH7.4, 100 mM NaCl, 0.1 mM EDTA, 0.5% Triton X-100) supplemented with Protease Inhibitor Cocktail Set III (Wako), and was allowed to stand on ice for 5 min. Sonication was performed using a Bioruptor (UCD-250, BM Equipment) with an ice-cold water bath and a rotating 1.5-ml microtube unit for 10 cycles of 30 s ON and 30 s OFF on the HIGH setting. The protein concentration was measured using a Coomassie Plus (Bradford) Assay Kit (Thermo Fischer) with bovine serum albumin as a standard. An appropriately diluted protein sample was mixed with standard SDS-PAGE sample buffer containing dithiothreitol, and boiled at 95C for 5 min. The protein sample was separated on 5-20% or 15% SDS-PAGE gel (SuperSep Ace, Wako), and transferred to a PVDF membrane in a Trans-Blot Turbo (Biorad). The membrane was blocked with PBST buffer (PBS containing 0.05% Tween 20) containing 5% skim milk for 1 h at room temperature. The primary and secondary antibodies were diluted in CanGet Signal (Toyobo). After the washing out of the secondary antibody conjugated with horse radish peroxidase with PBST buffer, the membrane was soaked in Clarity Western ECL Substrate (Biorad) or ECL Prime (GE Healthcare), and the signal was detected using an LAS4000 apparatus (GE Healthcare). The primary antibodies included: rabbit polyclonal anti-MTH1(Novus, cat# NB100-109, 1:2,000 dilution), mouse monoclonal anti-NUDT18 (OriGene, clone 1B4, cat# TA503828, lot# A01, 1:3,000 dilution), rabbit polyclonal anti-NUDT18 (GeneTex, cat# GTX120944, lot# 40674, 1:2,000 dilution), mouse monoclonal anti--Actin (Wako, clone 2F3, cat# 013-24553, lot# SAQ2468, 1:3,000 dilution), and mouse monoclonal peroxidase-conjugated anti-6His (Wako, clone 9C11, cat#016-23183, lot# PDR5831, 1:3,000 dilution).

2.9. DNA sequencing of the MTH3 gene

To examine whether the MTH3 gene is disrupted by CRISPR/Cas9, DNA sequencing of the region around the CRISPR target sequence was performed. Genomic DNA was extracted from the cell pellet using a Gentra Puregene Cell kit (Qiagen), and stored at 4C until use. A 518-bp region surrounding the CRISPR target site was amplified by a PCR using a PrimeSTAR Max DNA polymerase (Takara) and a primer set, designated MTH3-KO-check-FW and -RV (Supplementary Table 1). The PCR conditions were as follows: 35 cycles of 98C for 10 s, 55C for 5 s and 72C for 5 s. The PCR was performed in a GeneAmp PCR System 9700 (Applied Biosystems) on MAX mode. The PCR products were directly sub-cloned into a SalI site of pUC19 vector using an In-Fusion HD cloning system (Takara). DH5 competent cells (Toyobo) were transformed with the products, and the plasmids were recovered from transformants using a NucleoSpin Plasmid EasyPure (Macherey-Nagel). Sixteen plasmids per cell line were sequenced.

2.10. The E. coli strains and the measurement of the mutant frequency

A parental strain BW25113 [F- - rrnB3 lacZ4787 hsdR514 (araBAD)567 (rhaBAD)568 rph-1] and a mutT-deficient derivative JW0097 [BW25113 mutT::KmR], which were constructed by the National BioResource Project E. coli (NBRP-E. coli), were obtained from the National Institute of Genetics (Mishima, Japan) [25]. Wild-type and mutT-deficient competent cells were transformed with appropriate expression vectors. To measure the mutant frequency, five of each of the single colonies were independently isolated and cultured overnight in LB medium containing 100 g/ml ampicillin in a 37C shaking incubator at 180 r.p.m. Overnight cultures were ten-fold concentrated or appropriately diluted with 0.9% NaCl, and 100 l of cell suspension was plated onto LB or LB containing 100 g/ml rifampicin plates, and incubated at 37C. At 20 h after incubation, the visible colonies were counted. Data were obtained from five independent assays, and the mean and standard deviation were calculated from three independent experiments. Welch’s t-test was used to compare the mutant frequencies. It should be noted that overnight cultures for MTH1 and MTH3 were supplemented with 0.1 mM IPTG while 0.01 mM IPTG was added to the MTH2 culture, in order to obtain similar protein levels for the three MTHs. To determine the protein level in each overnight culture, the collected cell pellet was resuspended in an appropriate volume of 1sample buffer, and cell concentrations were determined based on the number of colonies on the LB plate. The cell suspension was then boiled, sonicated and subjected to SDS-PAGE and Western blotting. After obtaining a Western blot, the membrane was stained with Bio-Safe Coomassie Stain (Biorad).

2.11. Phylogenetic Analysis

The amino acid sequences of the NUDT proteins were obtained from NCBI. The primary structure of human Nudix protein was analyzed using ClustalW (KEGG) with “Slow/accurate” pairwise alignment and default setting parameters, and a rooted phylogenetic tree was created with branch length clustering determined by the UPGMA method. The following EMBL-EBI UniProt IDs were used for alignment: NUDT1: P36639; NUDT2: P50583; NUDT3: O95989; NUDT4: Q9NZJ9; NUDT5: Q9UKK9; NUDT6: P53370; NUDT7: P0C024; NUDT8: Q8WV74; NUDT9: Q9BW91; NUDT10: Q8NFP7; NUDT11: Q96G61; NUDT12: Q9BQG2; NUDT13: Q86X67; NUDT14: O95848; NUDT15: Q9NV35; NUDT16: Q96DE0; NUDT17: P0C025; NUDT18: Q6ZVK8; NUDT19: A8MXV4; NUDT20: Q8IU60; NUDT21: O43809; NUDT22: Q9BRQ3.

3. Results

3.1. The characterization of MTH1- and MTH2-depleted cells

3.1.1. The isolation of MTH1-knockout (KO) Cells

The human HeLa S3 cell line possesses all three types of MTH proteins (MTH1, MTH2 and MTH3), the enzyme activities of which have been well characterized [18]. We first isolated MTH1-KO cell lines from HeLa S3 cells, using CRISPR/Cas9 technology [26–28]. The MTH1 gene consists of five exons; two transcripts with distinct start codons are generated [29,30]. Since the start codons for the major MTH1 protein (p18) and for a longer isoform (p22) are located in exon III, we constructed the target sequence just downstream of the second start codon (Fig. 1A). As mentioned in the Experimental Procedures section, the CRISPR-target sequence was determined using the CRISPR DESIGN algorithm (Feng Zhang’s lab., MIT) with attention paid to the off-target effects. In this target sequence, the quality score and maximum score of the off-target site for MTH1-KO were 79 and 0.9, respectively, suggesting that optimized on-target and minimized off-target effects can be expected with the selected target sequence.
The disruption of the MTH1 gene was verified by Western blotting. No band corresponding to the MTH1 protein was detected in four independently isolated cell lines (Fig. 1B). These MTH1-KO cell lines were designated MTH1-KO #6, #8, #16 and #21, and their proliferation rates were measured. Sub-confluent cells were trypsinized, and approximately 1,000 cells were re-seeded in a 96-well plate. Two days after plating, the number of viable cells was measured using a Cell counting kit-8, as mentioned in the Experimental Procedures section. As shown in Fig. 1C, the cell growth of the MTH1-KO lines did not differ from that in wild-type cells to a statistically significant extent. Thus, MTH1 protein is not required for cellular proliferation under normal growth conditions.

3.1.2. The isolation of MTH2-KO cells

The establishment of MTH2-KO cell lines was performed according to a similar procedure to that used for the isolation of MTH1-KO cells. As shown in Fig. 2A, the MTH2 gene consists of two exons and two isoforms of MTH2 protein are formed (as reviewed in the NCBI RefSeq). Since both isoforms use the same start codon for the initiation of translation, for the disruption of the MTH2 gene, we placed the CRISPR/Cas9-target sequence on the start codon. Using this strategy, cleavage would occur between the adenine and thymine residues within the start codon. The quality score and the maximum score of the off-target site were 97 and 0.6, respectively.
Disruption of the MTH2 gene was verified by Western blotting. As shown in Fig. 2B, we successfully isolated four MTH2-KO lines, designated MTH2-KO #23, #34, #38 and #41. These independently isolated clones grew well in normal medium, and their proliferation rates were slightly lower than the proliferation rate of wild-type cells (Fig. 2C). Thus, MTH2 is not required for cell proliferation. Western blotting of the MTH2-KO cell lines and their parental line (WT). (C) The growth rates of MTH2-KO cells. Parental cells (wild-type, WT) and KO cells are indicated as dashed and solid lines, respectively.

3.1.3. The construction of the MTH1- and MTH2-double KO cell lines

Since MTH1 and MTH2 exhibit somewhat overlapping substrate specificities, it is advisable to construct double-KO cell lines. For this purpose, the CRISPR/Cas9 vector used for MTH2 knockout was introduced to MTH1-KO cell lines. In three independent experiments, three double-KO cells were isolated from MTH1-KO clones #6, #8 and #16, respectively; these were designated MTH1- MTH2-KO #6-3, #8-23 and #16-8, respectively. All of the double-KO clones were devoid of both MTH1 and MTH2 proteins. Fig. 3A shows the results of Western blotting performed with the representative clones. The proliferation rates of MTH1- MTH2-KO cells were almost the same as the proliferation rate of the wild-type cell line, except clone #16-8 (Fig. 3B). This clone might have some additional mutation, and was eliminated from the further analysis. Thus, the absence of both MTH1 and MTH2 did not affect cell survival.

3.1.4. The sensitivity of MTH1- and MTH2-depleted cells to hydrogen peroxide

Both MTH1 and MTH2 proteins hydrolyze the triphosphate form of nucleotides containing 8-oxoG, whereas MTH3 only hydrolyzes the diphosphate form [18]. Thus, we first performed the characterization of MTH1- and/or MTH2-depleted cells. Various cellular metabolites, including hydrogen peroxide, are known to oxidize the guanine residue of nucleotides and DNA. In the next experiments, we determined the sensitivity levels of MTH1-KO, MTH2-KO and double-KO lines against externally supplied hydrogen peroxide (Fig. 4). Cell lines defective in either MTH1 or MTH2 and both MTH1 and MTH2 exhibited increased sensitivity to hydrogen peroxide in comparison to wild-type cells. This result implies that MTH1 and MTH2 may play a role in the maintenance of genomic stability under oxidative stress.

3.1.5. The mutant frequency of MTH1-KO cells

Based on the sensitivity to hydrogen peroxide, we wondered if MTH1 and MTH2 may have a role in reducing spontaneous and induced mutant frequencies. Although 6-thioguanine (6-TG) resistance is a marker that is commonly used for mutation analyses, cells defective in MTH2 exhibit an increased sensitivity to 6-TG [31,32]. Thus we only compared the frequency of 6-TG resistance mutations between MTH1-KO and its parental cell lines. In our pilot study, the rates of spontaneous and hydrogen peroxide-induced mutation in MTH1-KO cells were 1.2-fold and 1.8-fold higher, respectively, than those in wild-type cells (data not shown). Then, to obtain more reliable data, the mutant frequencies of four lines of MTH1-KO cells were compared with those of five sub-lines of HeLa S3 cells. The data are shown in Supplementary Table 2. As shown in Fig. 5, the median mutant frequencies of wild-type and MTH1-KO cells were 3.6410-6 and 5.0410-6, respectively. Although the frequency of mutations in MTH1-KO cells was 1.39-fold higher than that in wild-type cells, the difference was not statistically significant (P=0.55).

3.2. The characterization of MTH3-depleted cells

3.2.1. The isolation of MTH3-KO cells

The establishment of MTH3-KO cell lines was performed according to a similar method to that which was used for the isolation of MTH1-KO and MTH2-KO cells. As shown in Fig. 6A, the MTH3 gene consists of four exons, and two isoforms are registered in the NCBI Protein database (Accession # NP_079091.3 and XP_011542952.1). These isoforms would produce a 323-aa protein (35.5 kDa) and a 539-aa protein (58.7 kDa). To eliminate both forms, the CRISPR/Cas9 cleavage site was set downstream of the second start codon located in exon 2. The quality score and maximum score of the off-target site in this target sequence were 97 and 0.6, respectively.
In this way, we obtained two candidate clones, #40 and #48, but it was difficult to verify MTH3 gene knockout by Western blotting due to the appearance of many bands with two commercially available antibodies (Fig. 6B). We then determined the genomic sequences of these clones. Clone #40 had two mutations, a 1-bp insertion and an 86-bp deletion which would cause a frameshift (Fig. 6C and supplementary Fig. 1B). Clone #48 had different two mutations, a 1-bp insertion and a 47-bp deletion, which would also cause frameshift (Fig. 6C and supplementary Fig. 1C). Thus, clone #40 and #48 can be regarded as MTH3-KO cell lines. Unlike MTH1- and MTH2-KO cells, the growth rates of both MTH3-KO lines (#40 and #48) were significantly lower than the growth rate of wild-type cells (Fig. 6D). This implies that MTH3 may be required for optimal efficiency of cell proliferation.

3.2.2. The sensitivity of MTH3-KO cells to hydrogen peroxide

Next, we assessed the sensitivity of the wild-type and MTH3-KO lines to hydrogen peroxide based on the colony forming ability (Fig. 7). Unexpectedly, both MTH3-KO clones #40 and #48 were somewhat more resistant than wild-type cells (see Discussion).

3.3. The suppression of the MutT mutator by the overproduction of human proteins

The E. coli MutT protein possesses a potent activity for degrading oxidized forms of dGTP and dGDP [5], and thus the mutT-mutants yield a strong mutator phenotype. Taking advantage of this phenotype, we evaluated the anti-mutagenic activities of the human MTH1, MTH2 and MTH3 proteins in E. coli mutT- cells. For this, N-terminally his-tagged human proteins were overproduced in E. coli cells using the pCold I vector expression system. In the presence of 0.1 mM (for MTH1 and MTH3) or 0.01 mM (for MTH2) IPTG, the three proteins were expressed at equivalent levels (Fig. 8A). Under these conditions, the mutator activity of MTH1 was strongly suppressed, to a degree that was similar to that of MutT (P=0.0085 for MTH1 and MutT) (Fig. 8B). The effect of MTH2 was slightly lower that of MTH1; however, the suppression was evident (P=0.0067). The overproduction of MTH3 caused a very low level of suppression; the effect did not reach statistical significance (P=0.0222). Since this level of weak suppression might have been caused by the overproduction of exogenous protein in cells, we examined the effect of a catalytically inactive MTH3 mutant protein. MutT and MTH1 proteins possess a highly conserved 23-amino acid-sequence, called the NUDIX motif. Since it is known that the glutamic acid residue at position 57 of the MutT protein is important for its catalytic activity [33,34], we changed the corresponding glutamic acid residue at position 95 of MTH3 to alanine (E95A), and examined its suppression effect on the mutT mutator phenotype. As shown in Fig. 8C, no significant suppressive effect was observed with the overproduction of MTH3 E95A proteins. In this experiment, even wild-type MTH3 showed no significant suppressive activity. We therefore concluded that there was a distinct difference between MTH3 and MTH1 and MTH2, in that MTH3 cannot substitute the E. coli MutT activity. containing either of the human MTHs cDNAs were introduced into mutT-deficient lines and the proteins were overproduced to approximately the same level. B, The mutant frequency was calculated as the number of rifampicin-resistant colonies per 108 visible colonies. pCTRL is a pCold I vector without cDNA. Double asterisks (**) and a single asterisk (*) indicate statistical significance (P<0.01 and P<0.05, respectively) with mutT_pCTRL. The median values were obtained from five independent cultures in a single experiment; and the mean and standard deviation were calculated from three independent experiments. C, pCold I vector containing human wild-type or an E95A mutation containing MTH3 cDNA was introduced into mutT-deficient lines, and the mutant frequency was calculated as the number of rifampicin-resistant colonies per 108 visible colonies. n.s., not significant. 4. Discussion The elimination of oxidized guanine nucleotides from the DNA precursor pool is a prerequisite for accurate DNA replication. The E. coli MutT protein functions in this process by hydrolyzing 8-oxo-dGTP and 8-oxo-dGDP to 8-oxo-dGMP, an form that is unusable for DNA polymerase [5]. Human cells possess three MutT-related enzymes, MTH1, MTH2 and MTH3, which act toward 8-oxoG-containing deoxyribonucleotides and carry the 23-residue MutT-related sequence (Nudix box) [11]. However, the roles that the three human proteins play in maintaining genome stability have not been elucidated. To explore this problem, we isolated human cells that were deficient in one or two of these proteins and examined their characteristics. As shown in Figs. 1, 2, 3 and 6, HeLa S3-derived clones defective in each one of the three proteins and in MTH1 and MTH2 doubly-deficient clones were isolated. With the exception of the MTH3-KO cell lines, all of these clones grew well in ordinary culture media and thus, MTH1 and MTH2 proteins are dispensable for cell reproduction. The growth retardation observed in the MTH3-KO cell lines implies that MTH3 might possess another function besides the degradation of oxidized nucleotides. The sequence motif called the Nudix box is present in many proteins in prokaryotes, archaea and eukaryotes [16]. The analysis of the human genome sequence has identified 22 proteins with this motif, in which MTH1, MTH2 and MTH3 are listed as NUDT1, NUDT15 and NUDT18, respectively. Fig. 9A shows the phylogenetic tree of the 22 NUDT proteins, which are arranged based on the overall amino acid sequence homology. It is noteworthy that MTH1, MTH2 and MTH3, all of which act on 8-oxoG-containing deoxyribonucleotides in vitro, are located in close proximity to each other in this tree. However, MTH1 and MTH2 are in the same cluster, while MTH3 is not. Fig. 9B shows the sequence alignment of E. coli MutT and human MTH proteins, in which the conserved Nudix box was placed in the same position. MTH3 is approximately twice as large as the other human proteins and E. coli MutT, which suggests that MTH3 might be capable of exhibiting activity toward substances other than 8-oxoG-containing nucleotides. In this regard, it is noteworthy that MTH3-deficient cells exhibit significantly lower growth rates and insensitivity to hydrogen peroxide (Fig. 7), while cells deficient in MTH1 and MTH2 exhibited increased sensitivity to hydrogen peroxide in comparison to wild-type cells (Fig. 4). MTH3 may involve in some cellular process such as leading to cell death under oxidative stress. In spite of the fact that there was little increase in the frequency of spontaneous mutations in the isolated cell lines, mice with MTH1 gene deficiency develop numerous tumors in various organs [15]. In this study, the mutation rate of ES cells derived from MTH1-KO mice was significantly increased by 1.78-fold and 2.55-fold in two MTH1-KO lines [15]. In the present study, the spontaneous mutant frequency in human MTH1-KO cells was only slightly increased in comparison to wild-type cells (Fig. 5). Although MTH1 depletion only has a minor effect on genome stability in the short-term, over a longer period, this might be sufficient to promote tumorigenesis in vivo. The potential of human proteins in 8-oxoG-related mutagenesis can be examined by expressing cDNA for human enzymes in E. coli mutT-deficient cells. The overproduction of MTH1 and MTH2 almost completely suppressed the mutator phenotype of mutT-deficient cells (Fig. 8B), indicating that these two human proteins (as well as MutT) have the potential to prevent 8-oxoG-related mutagenesis. On the other hand, we found no evidence to suggest that MTH3 possessed this ability (Fig. 8B and C). This difference may be related to the intrinsic substrate specificities of these human enzymes; namely, that MTH1 and MTH2 can degrade 8-oxo-dGTP, which is the direct substrate of DNA polymerase. On the other hand, MTH3 cleaves 8-oxo-dGDP but not 8-oxo-dGTP (Fig. 9C). Thus, the human enzymes can be divided into two categories: one possesses anti-mutagenic capacity by degrading the mutagenic substrate for DNA synthesis; the other lacks this capability. It is presently unknown how MTH3 exerts its function in the cellular process. Further studies should be performed to resolve this problem. 5. Conclusions MTH1- and MTH2-KO cells showed increased sensitivity to hydrogen peroxide compared with wild-type HeLa S3 cells. MTH1-KO cells showed only a slight increase in spontaneous mutant frequency in comparison to the parental HeLa S3 line, suggesting that MTH1 depletion only has a minor effect on genome stability in the short-term. On the other hand, the MTH3-KO cells showed a significantly slower growth rate, unlike MTH1-KO and MTH2-KO cells. Unlike MTH1 and MTH2, MTH3 cannot substitute the E. coli MutT activity. Our findings suggest that MTH1 and MTH2 are involved in the maintaining genome stability, while MTH3 may play some other role(s). References [1] J.F. Turrens, Mitochondrial formation of reactive oxygen species., J. Physiol. 552 (2003) 335–344. [2] A. Starkov, K.B. Wallace, Yin and Yang of Mitochondrial ROS, Imperial College Press, London, 2004. [3] M.L. Wood, M. Dizdaroglu, E. Gajewski, J.M. Essigmann, Mechanistic studies of ionizing radiation and oxidative mutagenesis: genetic effects of a single 8-hydroxyguanine (7-hydro-8-oxoguanine) residue inserted at a unique site in a viral genome., Biochemistry. 29 (1990) 7024–7032. [4] M. Moriya, C. Ou, V. Bodepudi, F. Johnson, M. Takeshita, A.P. Grollman, Site-specific mutagenesis using a gapped duplex vector: A study of translesion synthesis past 8-oxodeoxyguanosine in E. coli, Mutat. Res. Repair. 254 (1991) 281–288. [5] H. Maki, M. Sekiguchi, MutT protein specifically hydrolyses a potent mutagenic substrate for DNA synthesis., Nature. 355 (1992) 273–275. [6] B.D. Freudenthal, W.A. Beard, L. Perera, D.D. Shock, T. Kim, T. Schlick, S.H. Wilson, Uncovering the polymerase-induced cytotoxicity of an oxidized nucleotide., Nature. 517 (2015) 635–639. [7] H. Hayakawa, A. Taketomi, K. Sakumi, M. Kuwano, M. Sekiguchi, Generation and elimination of 8-oxo-7,8-dihydro-2’-deoxyguanosine 5’-triphosphate, a mutagenic substrate for DNA synthesis, in human cells., Biochemistry. 34 (1995) 89–95. [8] T. Sekiguchi, R. Ito, H. Hayakawa, M. Sekiguchi, Elimination and utilization of oxidized guanine nucleotides in the synthesis of RNA and its precursors, J. Biol. Chem. 288 (2013) 8128–8135. [9] M. Akiyama, T. Horiuchi, M. Sekiguchi, Molecular cloning and nucleotide sequence of the mutT mutator of Escherichia coli that causes A:T to C:G transversion., Mol. Gen. Genet. 206 (1987) 9–16. [10] A. Sakai, M. Nakanishi, K. Yoshiyama, H. Maki, Impact of reactive oxygen species on spontaneous mutagenesis in Escherichia coli, Genes to Cells. 11 (2006) 767–778. [11] T. Tajiri, H. Maki, M. Sekiguchi, Functional cooperation of MutT, MutM and MutY proteins in preventing mutations caused by spontaneous oxidation of guanine nucleotide in Escherichia coli, Mutat. Res. Repair. 336 (1995) 257–267. [12] C.G. Cupples, J.H. Miller, A set of lacZ mutations in Escherichia coli that allow rapid detection of each of the six base substitutions., Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 5345–5349. [13] M. Furuichi, M.C. Yoshida, H. Oda, T. Tajiri, Y. Nakabeppu, T. Tsuzuki, M. Sekiguchi, Genomic structure and chromosome location of the human mutT homologue gene MTH1 encoding 8-oxo-dGTPase for prevention of A:T to C:G transversion., Genomics. 24 (1994) 485–490. [14] K. Sakumi, M. Furuichi, T. Tsuzuki, T. Kakuma, S.I. Kawabata, H. Maki, M. Sekiguchi, Cloning and expression of cDNA for a human enzyme that hydrolyzes 8-Oxo-dGTP, a mutagenic substrate for DNA synthesis, J. Biol. Chem. 268 (1993) 23524–23530. [15] T. Tsuzuki, A. Egashira, H. Igarashi, T. Iwakuma, Y. Nakatsuru, Y. Tominaga, H. Kawate, K. Nakao, K. Nakamura, F. Ide, S. Kura, Y. Nakabeppu, M. Katsuki, T. Ishikawa, M. Sekiguchi, Spontaneous tumorigenesis in mice defective in the MTH1 gene encoding 8-oxo-dGTPase., Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 11456–11461. [16] A.G. McLennan, The Nudix hydrolase superfamily, Cell. Mol. Life Sci. 63 (2006) 123–143. [17] M.J. Bessman, D.N. Frick, S.F. O’Handley, The MutT proteins or “Nudix” hydrolases, a family of versatile, widely distributed, “housecleaning” enzymes, J. Biol. Chem. 271 (1996) 25059–25062. [18] Y. Takagi, D. Setoyama, R. Ito, H. Kamiya, Y. Yamagata, M. Sekiguchi, Human MTH3 (NUDT18) protein hydrolyzes oxidized forms of guanosine and deoxyguanosine diphosphates: Comparison with MTH1 and MTH2, J. Biol. Chem. 287 (2012) 21541–21549. [19] J.P. Cai, T. Ishibashi, Y. Takagi, H. Hayakawa, M. Sekiguchi, Mouse MTH2 protein which prevents mutations caused by 8-oxoguanine nucleotides, Biochem. Biophys. Res. Commun. 305 (2003) 1073–1077. [20] K. Okamoto, S. Toyokuni, W.J. Kim, O. Ogawa, Y. Kakehi, S. Arao, H. Hiai, O. Yoshida, Overexpression of human mutT homologue gene messenger RNA in renal-cell carcinoma: Evidence of persistent TH5427 oxidative stress in cancer, Int. J. Cancer. 65 (1996) 437–441.
[21] C.H. Kennedy, R. Cueto, S.A. Belinsky, J.F. Lechner, W.A. Pryor, Overexpression of hMTH1 mRNA: A molecular marker of oxidative stress in lung cancer cells, FEBS Lett. 429 (1998) 17–20.
[22] H. Gad, T. Koolmeister, A.-S. Jemth, S. Eshtad, S.A. Jacques, C.E. Ström, L.M. Svensson, N. Schultz, T. Lundbäck, B.O. Einarsdottir, A. Saleh, C. Göktürk, P. Baranczewski, R. Svensson, R.P.-A. Berntsson, R. Gustafsson, K. Strömberg, K. Sanjiv, M.-C. Jacques-Cordonnier, M. Desroses, A.-L. Gustavsson, R. Olofsson, F. Johansson, E.J. Homan, O. Loseva, L. Bräutigam, L. Johansson, A. Höglund, A. Hagenkort, T. Pham, M. Altun, F.Z. Gaugaz, S. Vikingsson, B. Evers, M. Henriksson, K.S.A. Vallin, O.A. Wallner, L.G.J. Hammarström, E. Wiita, I. Almlöf, C. Kalderén, H. Axelsson, T. Djureinovic, J.C. Puigvert, M. Häggblad, F. Jeppsson, U. Martens, C. Lundin, B. Lundgren, I. Granelli, A.J. Jensen, P. Artursson, J.A. Nilsson, P. Stenmark, M. Scobie, U.W. Berglund, T. Helleday, MTH1 inhibition eradicates cancer by preventing sanitation of the dNTP pool., Nature. 508 (2014) 215–221.
[23] K.V.M. Huber, E. Salah, B. Radic, M. Gridling, J.M. Elkins, A. Stukalov, A.-S. Jemth, C. Göktürk, K. Sanjiv, K. Strömberg, T. Pham, U.W. Berglund, J. Colinge, K.L. Bennett, J.I. Loizou, T. Helleday, S. Knapp, G. Superti-Furga, Stereospecific targeting of MTH1 by (S)-crizotinib as an anticancer strategy., Nature. 508 (2014) 222–227.
[24] F.A. Ran, P.D. Hsu, J. Wright, V. Agarwala, D.A. Scott, F. Zhang, Genome engineering using the CRISPR-Cas9 system, Nat. Protoc. 8 (2013) 2281–2308.
[25] T. Baba, T. Ara, M. Hasegawa, Y. Takai, Y. Okumura, M. Baba, K.A. Datsenko, M. Tomita, B.L. Wanner, H. Mori, Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection, Mol. Syst. Biol. 2 (2006).
[26] L. Cong, F.A. Ran, D. Cox, S. Lin, R. Barretto, N. Habib, P.D. Hsu, X. Wu, W. Jiang, L.A. Marraffini, F. Zhang, Multiplex Genome Engineering Using CRISPR/Cas Systems, Science. 339 (2013) 819–823.
[27] P. Mali, L. Yang, K.M. Esvelt, J. Aach, M. Guell, J.E. DiCarlo, J.E. Norville, G.M. Church, RNA-Guided Human Genome Engineering via Cas9, Science. 339 (2013) 823–826.
[28] M. Jinek, A. East, A. Cheng, S. Lin, E. Ma, J. Doudna, RNA-programmed genome editing in human cells., Elife. 2 (2013) e00471.
[29] H. Oda, Y. Nakabeppu, M. Furuichi, M. Sekiguchi, Regulation of Expression of the Human MTH1 Gene Encoding 8-Oxo-dGTPase: ALTERNATIVE SPLICING OF TRANSCRIPTION PRODUCTS, J. Biol. Chem. 272 (1997) 17843–17850.
[30] H. Oda, A. Taketomi, R. Maruyama, R. Itoh, K. Nishioka, H. Yakushiji, T. Suzuki, M. Sekiguchi, Y. Nakabeppu, Multi-forms of human MTH1 polypeptides produced by alternative translation initiation and single nucleotide polymorphism, Nucleic Acids Res. 27 (1999) 4335–4343.
[31] M. Carter, A.-S. Jemth, A. Hagenkort, B.D.G. Page, R. Gustafsson, J.J. Griese, H. Gad, N.C.K. Valerie, M. Desroses, J. Boström, U. Warpman Berglund, T. Helleday, P. Stenmark, Crystal structure, biochemical and cellular activities demonstrate separate functions of MTH1 and MTH2., Nat. Commun. 6 (2015) 7871.
[32] N.C.K. Valerie, A. Hagenkort, B.D.G. Page, G. Masuyer, D. Rehling, M. Carter, L. Bevc, P. Herr, E. Homan, N.G. Sheppard, P. Stenmark, A.S. Jemth, T. Helleday, NUDT15 hydrolyzes 6-thio-deoxyGTP to mediate the anticancer efficacy of 6-thioguanine, Cancer Res. 76 (2016) 5501– 5511.
[33] Y. Fujii, H. Shimokawa, M. Sekiguchi, Y. Nakabeppu, Functional significance of the conserved residues for the 23-residue module among MTH1 and MutT family proteins., J. Biol. Chem. 274 (1999) 38251–38259.
[34] H. Shimokawa, Y. Fujii, M. Furuichi, M. Sekiguchi, Y. Nakabeppu, Functional significance of conserved residues in the phosphohydrolase module of Escherichia coli MutT protein., Nucleic Acids Res. 28 (2000) 3240–3249.