Oxidative DNA damage induced by 3,4-diOH-TAM, a tamoxifen metabolite, in relation to endometrial carcinogenesis

Abstract

Tamoxifen (TAM) is a non-steroidal anti-estrogen used widely in the treatment of breast cancer. However, long-term use of TAM increases the risk of endometrial cancer. To clarify the molecular mechanism of carcinogenesis, we examined DNA damage by TAM and its metabolites, 4-hydroxytamoxifen (4-OH-TAM), 3-hydroxytamoxifen (3-OH-TAM) and 3,4-dihydroxytamoxifen (3,4-diOH-TAM). 3,4-DiOH-TAM caused oxidative damage to ³²P-5’-end-labeled DNA fragments in the presence of Cu(II) and NADH, while TAM, 4-OH-TAM and 3-OH-TAM did not. The HOMO level of 3,4-diOH-TAM was calculated to be the highest among TAM metabolites. 3,4-DiOH-TAM induced DNA cleavage at cytosine and guanine residues of the ACG sequence complementary to codon 273, a well-known hotspot in the p53 gene. Catalase and bathocuproine inhibited DNA damage induced by 3,4-diOH-TAM in the presence of Cu(II) and NADH, suggesting the involvement of H₂O₂ and Cu(I). 3,4-DiOH-TAM with Cu(II) induced 8-oxodG formation in calf thymus DNA, which was markedly enhanced in the presence of NADH. Oxidative DNA damage induced by a TAM metabolite 3,4-diOH-TAM may contribute to the initiation of endometrial carcinogenesis. Furthermore, the reported proliferative effect of TAM on endometrial cells is involved in the promotion stage of tumor development. Taken together, it is suggested that both mechanisms may play roles in endometrial carcinogenesis.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-32037-3.

Introduction

Tamoxifen (TAM) is a non-steroidal anti-estrogen used widely in the treatment and chemoprevention of breast cancer. TAM has been regularly used in clinical therapy for breast cancer since 1971¹. In 1999, the U.S. Food and Drug Administration approved TAM for primary prevention of breast cancer². In the US and UK, it is also administered for prophylactic chemoprevention in women identified as high risk for breast cancer and appears to reduce disease incidence by 50% in pre- or post-menopausal cohorts^3,4. TAM, available as a generic drug worldwide now, cost effectiveness, has become the golden standard treatment of premenopausal breast cancer⁵.

TAM has been referred to as a pro-drug that requires activation to exert anti-breast cancer effects⁶. TAM is metabolized by hepatic cytochrome P450s (CYPs) to produce its major primary metabolites in the plasma. 4-Hydroxytamoxifen (4-OH-TAM) and 3-hydroxytamoxifen (3-OH-TAM) is principally formed by the CYP2D6 and CYP3A5 enzymes, respectively⁶. Human hepatic microsomal cytochrome P450 enzymes metabolize 4-OH-TAM and 3-OH-TAM to produce the further oxidized metabolite, 3,4-dihydroxytamoxifen (3,4-diOH-TAM)^7,8. TAM and its metabolites interact with the estrogen receptor and this is generally considered to be the mechanism by which its pharmacological effects are mediated⁹. However, TAM has anti-estrogenic effects in breast tissue, but estrogen-like effects on the endometrium and other gynecological organs, and dose-dependently increases endometrial cell proliferation¹⁰. TAM treatment has the serious side effect of increasing the incidence of endometrial cancer in breast cancer patients^11,12. For this reason, TAM has been classified as a Group 1 carcinogen by IARC (International Agency for Research on Cancer)¹³. On the other hand, TAM has been successfully used as treatment in advanced or recurrent endometrial cancer¹⁴. The detailed mechanism of TAM action of side effects, especially endometrial carcinogenesis, remains unclear.

A step-wise process of carcinogenesis begins with genotoxicity. TAM-DNA adducts have been detected in the endometrial tissues of breast cancer patients treated with TAM, using³²P-postlabeling analysis¹⁵, accelerator mass spectrometry¹⁶, and TAM-DNA chemiluminescence immunoassay¹⁷. Some suggest that a low level of TAM-DNA adducts in the human uterus is unlikely to be involved with endometrial cancer development¹⁶, and others suggest that the presence of TAM-DNA adducts in the human uterus leads to genotoxicity and endometrial tumor induction^15,17. Thus, the importance of genotoxicity as a major pathway for TAM-induced endometrial cancer is still unclear¹⁸. In our previous study, we investigated the role of oxidative DNA damage in the carcinogenicity of compounds generally linked to DNA adduct formation¹⁹. Based on this, we also investigated the potential involvement of oxidative DNA damage in TAM-induced carcinogenesis.

To clarify the other genotoxicity rather than previously-reported DNA adduct formation, we examined oxidative DNA damage including 8-oxo-7, 8-dihydro-2’-deoxyguanosine (8-oxodG) by TAM and its metabolites in the presence of an endogenous reductant and metal ion. In addition, we calculated the highest occupied molecular orbital (HOMO) energy on TAM and its metabolites, which can inflict oxidative DNA damage. The HOMO represents the highest-energy orbital from which an electron can be donated.

Materials and methods

Materials

3,4-DiOH-TAM was synthesized according to the reported method⁸. The purity of the synthesized 3,4-diOH-TAM was confirmed to be greater than 95% based on the 1H NMR spectrum. Restriction enzymes (Sma I, Eco RI, BssHII, Apa I, and Sty I) and proteinase K were purchased from Roche Molecular Biochemicals (Mannheim, Germany). Restriction enzymes (Hind III, Ava I, and Xba I) and T₄ polynucleotide kinase were purchased from New England Biolabs (Beverly, MA). [γ-³²P] ATP (222 TBq/mmol) was obtained from New England Nuclear (Boston, MA). 4-OH-TAM (Z-isomer > 98%), superoxide dismutase (SOD, 3,000 units/mg from bovine erythrocytes) and catalase (45,000 units/mg from bovine liver) were purchased from Sigma Chemical Co. (St. Louis, MO). 3-OH-TAM, also known as dorolox, was from Toronto Research Center (North York, Canada). ß-Nicotinamide adenine dinucleotide disodium salt (reduced form) (NADH) was purchased from Kohjin Co. (Tokyo, Japan). Diethylenetriamine-N, N,N’,N’’,N”-pentaacetic acid (DTPA) and bathocuproinedisulfonic acid were obtained from Dojin Chemicals Co. (Kumamoto, Japan).

DFT calculation of TAM and its metabolites

Density functional theory (DFT) calculations were carried out to investigate the most stable conformation and the distribution of HOMO. The calculations were performed using the ωB97X-D exchange-correlation functional with the 6-31G(d) basis set, as implemented in the Spartan 24 software package (Wavefunction Inc., Irvine, CA, USA).

Preparation of ³²P-5’-end-labeled DNA fragments

Exon-containing DNA fragments obtained from the human p53 tumor suppressor gene²⁰ were prepared, as described previously²¹. A 5’-end-labeled 650 bp fragment (Hind III*13972 - Eco RI*14621) was obtained by dephosphorylation with calf intestine phosphatase and rephosphorylation with [γ-³²P] ATP and T₄ polynucleotide kinase (*, ³²P -label). The 650 bp fragment was further digested with Apa I to obtain a singly labeled 443 bp fragment (Apa I 14179 - Eco RI*14621) and a 211 bp fragment (Hind III*13972 - Apa I 14182). The fragment was prepared from plasmid pbcNI, which carries a 6.6-kb Bam HI chromosomal DNA segment containing the c-Ha-ras-1 protooncogene²². A singly labeled 341 bp fragment (Xba I 1906-Ava I*2246) was obtained according to the method described previously²³. Nucleotide numbering starts with the BamHI site²².

Detection of DNA damage by TAM and its metabolites in the presence of NADH and Cu(II)

A standard reaction mixture (in a 1.5 ml Eppendorf microtube) contained the 3,4-diOH-TAM, CuCl₂, NADH, ³²P -5’-end labeled DNA fragments and calf thymus DNA (5–10 µM per base) in 200 µl of 10 mM sodium phosphate buffer (pH 7.8) containing 5 µM DTPA. After incubation at 37 °C for 1 h, the DNA fragments were heated at 90 °C in 1 M piperidine for 20 min, where indicated. The DNA fragments were then electrophoresed on an 8% polyacrylamide-8 M urea gel (14 × 16 cm), and the autoradiogram was obtained by exposing an X-ray film (Fujifilm Corp., Tokyo, Japan) to the gel, as previously described²⁴. In certain experiments, the DNA was treated with 10 units of formamidopyrimidine-DNA glycosylase (Fpg, 20,000 units/mg from Escherichia coli) from Trevigen Inc. (Gaithersburg, MD) in 10 µl of reaction buffer (10 mM HEPES-KOH (pH 7.4), 100 mM KCl, 10 mM EDTA and 0.1 mg/ml BSA) at 37 °C for 2 h. The preferred cleavage sites were determined by direct comparison of the labeled, cleaved oligonucleotides with a standard 5’-end-labeled Maxam-Gilbert sequencing reaction²⁵ (LKB 2010 Macrophor, LKB Pharmacia Biotechnology Inc.). The relative amounts of oligonucleotides from the treated DNA fragments were measured with a laser densitometer (LKB 2222 UltroScan XL, LKB Pharmacia Biotechnology Inc. NJ, USA).

Analysis of 8-oxodG formation in calf thymus DNA by 3,4-diOH-TAM in the presence of NADH and Cu(II)

DNA fragments (100 µM per base) from calf thymus DNA were incubated with 3,4-diOH-TAM, CuCl₂, and NADH at 37 °C for the indicated times. 0.1 mM DTPA was added to stop the reaction. After ethanol precipitation, DNA was digested to the nucleosides with nuclease P₁ and calf intestine phosphatase, and then analyzed by HPLC-ECD, as described previously²⁶.

Analysis of NADH consumption by UV-visible spectra measurements

UV-visible spectra were measured with a UV-visible spectrometer (UV-2500PC, Shimadzu, Kyoto). The spectra of the mixtures were measured repeatedly at 37 °C for the indicated durations. To analyze the redox reaction with NADH, the change of absorption at 340 nm was measured. The reaction mixture contained 100 µM NADH, 3,4-diOH-TAM and CuCl₂ in 10 mM sodium phosphate buffer (pH 7.8) containing 5 µM DTPA. The decrease in absorbance of NADH at 340 nm (ε = 6.22 × 10³ M^− 1cm^− 1) was due to its oxidization to NAD⁺.

Results

DFT calculation of TAM and its metabolites, and their damage to ³²P-labeled DNA fragments

Figure 1A shows the DFT-optimized structures and HOMO energies of the TAM and its metabolites. Molecular orbital calculations revealed that tamoxifen itself has a HOMO energy of − 7.12 eV. Introduction of a single hydroxyl group increased the HOMO level to − 7.11 eV for 3-OH-TAM and − 7.01 eV for 4-OH-TAM, indicating that mono-hydroxylation enhances electron-donating potential relative to the parent compound. Subsequent oxidation to generate the catechol derivative, 3,4-diOH-TAM, further raised the HOMO level to − 6.97 eV, the highest among the series. Thus, sequential oxidation from tamoxifen to mono-hydroxylated and then catechol-type metabolites systematically increases the electron-donating capacity.

Fig. 1.

DFT-optimized structures with HOMO values of TAM and its metabolites and the damage to ³²P-5’-end-labeled DNA fragments. (A) DFT-optimized three-dimensional chemical structures and HOMO value (eV). (B) The reaction mixture contained the ³²P-5’-end-labeled 443 bp DNA fragment, 20 µM/base calf thymus DNA, 5 µM TAM, 4-OH-TAM, 3-OH-TAM or 3,4-diOH-TAM, 20 µM CuCl₂, and 200 µM NADH in 10 mM phosphate buffer (pH 7.8) containing 5 µM DTPA. The mixtures were incubated at 37 °C for 1 h. The DNA fragments were treated with 1 M piperidine at 90 °C for 20 min, and then electrophoresed on an 8% polyacrylamide/8 M urea gel. The autoradiogram was obtained by exposing an X-ray film to the gel.

Figure 1B shows an autoradiogram of DNA fragments treated with TAM and its metabolites. The autoradiogram detected the oligonucleotides, which were generated both from DNA cleavage at base modification yielded by piperidine and DNA strand break. TAM, 4-OH-TAM and 3-OH-TAM did not cause DNA damage even in the presence of Cu(II) and NADH under the conditions used. 3,4-DiOH-TAM caused slight DNA damage in the presence of Cu(II). When NADH was added, Cu(II)-mediated DNA damage by 3,4-diOH-TAM was dramatically enhanced.

DNA damage induced by 3,4-diOH-TAM in the presence of Cu(II) and NADH

The intensity of DNA damage increased with successive concentrations of 3,4-diOH-TAM (Fig. 2A). 3,4-DiOH-TAM/Cu(II)/NADH caused a little strand breakage, as detected with no treatment (Fig. 2B, left). An increase in the number of oligonucleotides following piperidine treatment suggested that 3,4-diOH-TAM caused not only strand breakage but also base modification/liberation (Fig. 2B, middle). Fpg treatment also increased oligonucleotides, indicating the formation of 8-oxodG and other oxidized bases (Fig. 2B, right).

Fig. 2.

DNA damage induced by 3,4-diOH-TAM in the presence of Cu(II) and NADH. The reaction mixture contained the³²P-5’-end-labeled 211 bp DNA fragment, 20 µM/base calf thymus DNA, the indicated concentrations of 3,4-diOH-TAM, 20 µM CuCl₂, and 200 µM NADH in 10 mM phosphate buffer (pH 7.8) containing 5 µM DTPA. After incubation at 37 °C for 1 h, the DNA fragments were treated with (A) hot piperidine, (B) no treatment, hot piperidine or Fpg protein, and electrophoresed on a polyacrylamide gel.

Effects of scavengers and bathocuproine on DNA damage induced by 3,4-diOH-TAM in the presence of Cu(II) and NADH

To clarify the reactive species for DNA damage, we performed experiments with radical scavengers. Figure 3 shows the effects of scavengers and bathocuproine, a Cu(I)-specific chelator, on DNA damage induced by 3,4-diOH-TAM in the presence of Cu(II) and NADH. Inhibition of DNA damage by catalase and bathocuproine suggests the involvement of hydrogen peroxide (H₂O₂) and Cu(I). Methional reduced the amount of DNA damage, although other typical hydroxyl radical (•OH) scavengers, ethanol, mannitol and sodium formate did not decrease the damage. In contrast, SOD slightly enhanced DNA damage.

Fig. 3.

Effects of ROS scavengers and bathocuproine on 3,4-diOH-TAM-induced DNA damage. The reaction mixture contained the³²P-5’-end-labeled 341 bp DNA fragment, 20 µM/base calf thymus DNA, 5 µM 3,4-diOH-TAM, 20 µM CuCl₂, and 200 µM NADH in 10 mM phosphate buffer (pH 7.8) containing 5 µM DTPA. Scavenger or bathocuproine was added as follows: 5% ethanol, 0.1 M mannitol, 0.1 M sodium formate, 0.1 M methional, 30 units/ml catalase, 30 units/ml SOD, and 50 µM bathocuproine. The mixtures were incubated at 37 °C for 1 h. The DNA fragments were treated with 1 M piperidine at 90 °C for 20 min, and then analyzed as described above.

Site specificity of DNA damage induced by 3,4-diOH-TAM/Cu(II)/NADH

To examine the site specificity of DNA damage, an autoradiogram was obtained and scanned with a laser densitometer to measure the relative intensity of DNA cleavage in the human p53 tumor suppressor gene. 3,4-DiOH-TAM/Cu(II)/NADH induced piperidine-labile sites (Fig. 4, upper) preferentially at T and C residues, and Fpg-sensitive sites (Fig. 4, lower) at G and C residues of the p53 gene. Fpg protein catalyzes the excision of piperidine-resistant 8-oxodG and other oxidized piperidine-labile G residues, and also mediates cleavage of oxidized products of C such as 5-hydroxyC²⁷. 3,4-DiOH-TAM induced DNA cleavage at piperidine-labile C and Fpg-sensitive G residues of the ACG sequence complementary to codon 273, a well-known hotspot^28,29 in the p53 gene. Several studies reported that 8-oxodG causes DNA misreplication, which can lead to mutation, particularly G to T transversion^30,31. Relevantly, p53 codon 273 mutation was reported in a human cell line derived from endometrial carcinoma³².

Fig. 4.

Site specificity of 3,4-diOH-TAM-induced DNA damage. To clarify the site specificity of DNA damage, the reaction mixture contained the³²P-5’-end labeled 443 bp (Apa I 14179 - Eco RI*14621) DNA fragment, 20 µM/base calf thymus DNA, 5 µM 3,4-diOH-TAM, 20 µM CuCl₂, and 200 µM NADH in 10 mM phosphate buffer (pH 7.8) containing 5 µM DTPA. After incubation for 1 h at 37 °C, the DNA fragments were treated with piperidine (upper panel) or Fpg protein (lower panel) and electrophoresed by the method described in MATERIALS AND METHODS. The relative amounts of DNA fragments were measured by scanning the autoradiogram with a laser densitometer. The horizontal axis shows the nucleotide number of the human p53 tumor suppressor gene, and underscoring shows the complementary sequence to codon 273 (nucleotide numbers 14486–14488).

Formation of 8-oxodG in calf thymus DNA by 3,4-diOH-TAM in the presence of Cu(II) and NADH

We examined Cu(II)-mediated 8-oxodG formation in calf thymus DNA treated with 3,4-diOH-TAM in the presence and absence of NADH, using HPLC-ECD (Fig. 5A). The metabolite caused oxidative DNA damage in the presence of Cu(II), and the addition of NADH significantly increased 8-oxodG formation.

Fig. 5.

Formation of 8-oxodG in calf thymus DNA treated with 3,4-diOH-TAM in the presence of Cu(II) and NADH, and NADH consumption. (A) The reaction mixture contained 100 µM/base calf thymus DNA, the indicated concentrations of 3,4-diOH-TAM, 20 µM CuCl₂, and no or 200 µM NADH in 4 mM phosphate buffer (pH 7.8) containing 5 µM DTPA. After incubation at 37 °C for 1 h, DNA was precipitated in ethanol and enzymatically digested into nucleosides. The levels of 8-oxodG were quantified by HPLC-ECD. (B) The reaction mixture containing the indicated concentrations of 3,4-diOH-TAM, 100 µM NADH and 20 µM CuCl₂ was measured every 10 min for 60 min at 37 °C.

Cu(II)-mediated NADH consumption by 3,4-diOH-TAM

The spectral changes of NADH were measured to clarify the effect of NADH oxidation on the reaction mixture of 3,4-diOH-TAM and Cu(II). Figure 5B shows that the Cu(II)-mediated redox reaction of 3,4-diOH-TAM with 100 µM NADH was observed as a decreasing absorbance at 340 nm of NADH (reduced form) for 60 min. The higher concentration of 3,4-diOH-TAM oxidized a greater amount of NADH. When Cu(II) was omitted, only a little NADH oxidation occurred (data not shown).

Discussion

TAM undergoes oxidative metabolism at either the 3- or 4-position of the aromatic ring to yield 3-OH- or 4-OH-TAM by the P450s, such as CYP2D6 enzyme⁶. Further oxidation generates the catechol-type metabolite 3,4-diOH-TAM by the hepatic microsomal cytochrome P450s^7,8. The present study demonstrated that 3,4-diOH-TAM has the ability to cause oxidative DNA damage in the presence of Cu(II) and NADH, in addition to TAM-DNA adduct formation, which is generally accepted as a mechanism of TAM genotoxicity. On the other hand, TAM, 4-OH-TAM and 3-OH-TAM did not cause DNA damage even in the presence of Cu(II) and NADH. The introduction of the catechol moiety fundamentally alters the redox behavior of TAM, enabling it to activate molecular oxygen. In general, phenolic compounds can reduce and activate O₂ through redox cycling with Cu(I)/Cu(II). Metal ions play an important role in this process. The ortho-dihydroxy arrangement of catechol confers enhanced acidity compared to monophenols due to intramolecular hydrogen bonding and resonance stabilization, facilitating the generation of the corresponding phenoxide anions. It is reported that upon deprotonation of both hydroxyl groups, the two adjacent oxygen atoms act as a bidentate ligand, forming a stable chelate complex with metal ions³³. This chelation is coupled with an electronic advantage: hydroxyl substitution progressively elevates the HOMO level. Such an elevated HOMO facilitates reduction of Cu(II) to Cu(I), thereby accelerating redox cycling.

Umemoto et al. ³⁴ reported that DNA adduct formation in rat liver induced by TAM is not observed with toremifene, a chlorinated analogue of TAM. Toremifene, which shares the stilbene scaffold with tamoxifen, is also likely to undergo aromatic ring hydroxylation to a catechol derivative (3,4-diOH metabolite), which could potentially lead to oxidative DNA damage. Due to the structural differences in the side chain (e.g., the chlorine atom on the ethyl group), toremifene’s capacity to form genotoxic DNA adducts is significantly lower than that of tamoxifen. Consequently, it is predicted that the oxidative DNA damage pathway may become relatively dominant in the bioactivation of toremifene in vivo, potentially contributing more significantly to its overall toxicological profile compared to tamoxifen. The catechol form of toremifene metabolites may also cause oxidative DNA damage, but not the formation of DNA adducts. This mechanistic comparison, particularly the predicted prominence of the oxidative pathway for TAM and toremifene will be a critical area for future investigation.

The response to TAM has a high degree of inter-individual variability, which is mainly due to genetic variants in P450 CYP genes. CYP3A and CYP2D6 play major roles in the metabolic steps leading to 3,4-diOH-TAM formation⁶. Genetic polymorphisms that alter their activities may therefore change individual exposure to this oxidative metabolite. Reduced-function alleles could limit its formation, whereas higher-activity variants may increase oxidative DNA damage and thereby influence susceptibility to TAM-related endometrial carcinogenesis.

A possible mechanism for the oxidative DNA damage by 3,4-diOH-TAM can be envisioned on the basis of our results (Fig. 6). Since the HOMO was predominantly localized on catechol moiety, dihydroxy form of the TAM metabolite, 3,4-diOH-TAM, can be autoxidized to semiquinone radical, and further to quinone form together with the reduction of Cu(II) to Cu(I). Generation of O₂^•− would then occur, coupled with the autoxidation of the metabolite. Thereafter, O₂^•− is dismutated to generate H₂O₂. Inhibitory effects of catalase and bathocuproine on DNA damage by the metabolite suggest that H₂O₂ and Cu(I) participate in DNA damage. Methional attenuated DNA damage, suggesting the involvement of reactive species such as Cu(I)-hydroperoxo complexes, since methional scavenges not only the •OH but also crypto-OH radicals^35,36. The addition of NADH efficiently enhanced Cu(II)-mediated oxidative DNA damage by 3,4-diOH-TAM. An endogenous reductant, NADH, reduces quinone form and semiquinone radical to dihydroxy form, resulting in enhanced generation of reactive oxygen species and DNA damage through the redox cycle. Several studies have indicated that NADH may react nonenzymatically with some xenobiotics and mediate their reduction^37,38. The experimental conditions for the oxidative DNA damage by 3,4-diOH-TAM in the presence of Cu(II) (20 µM) and NADH (200 µM), are physiologically relevant concentrations^39–41. The cycling of redox reactions would cause the DNA damage with excessive generation of reactive oxygen species by the low concentrations of 3,4-diOH-TAM. Accordingly, the metabolic conversion of TAM into its catechol derivative not only confers redox properties but also implicates this metabolite in pro-oxidant activity with potential genotoxic consequences. As a limitation, there have been no reports of 3,4-diOH-TAM being detected in vivo. The 3,4-diOH form (catechol form) in the stilbene skeleton of 3,4-diOH-TAM is known to be extremely unstable, and therefore there is no data on the amount of 3,4-diOH-TAM in vivo. Further study is needed to measure the concentration of the TAM metabolite in vivo.

Fig. 6.

A possible mechanism of oxidative DNA damage induced by a tamoxifen metabolite 3,4-diOH-TAM in relation to endometrial carcinogenesis.

Numerous studies have established an increased incidence of endometrial cancer among woman taking TAM. It has been reported that TAM forms 4-hydroxylated metabolite, which then become quinone methides⁴² and form DNA adducts in vivo⁴³. The present study has shown that 3,4-diOH-TAM has the ability to cause oxidative DNA damage. Cancer cells have mechanisms to protect themselves from oxidative stress and have developed adaptation strategies including the up-regulation of antioxidants, such as glutathione (GSH)^44,45. Our previous study showed that breast cancer cells were more resistant to oxidative stress than normal mammary epithelial cells, since the GSH level was 2-fold higher⁴⁶. It is reasonable to assume that the TAM metabolite induces oxidative DNA damage in normal endometrial cells more easily than in breast cancer cells. Therefore, we conclude that 3,4-diOH-TAM-induced oxidative DNA damage may play a role in carcinogenicity of TAM, in addition to the previously-reported DNA adduct.

Carcinogenesis is usually viewed as a step-wise process that begins with genotoxicity (initiation), followed by enhanced cell proliferation (promotion). Exposure to estrogen and elevated circulating estrogen levels increase the risk of breast and endometrial cancer in animals and humans^47,48. Estrogens promote carcinogenesis by estrogen receptor-mediated cell proliferation, which results in an increased risk of genomic mutations during DNA replication⁴³. One postulated carcinogenic mechanism for the female reproductive organs is that estrogen acts as a hormone stimulating cell proliferation. TAM and its metabolites interact with the estrogen receptor, acting as the antagonist in mammary tissues. This is generally considered to be the mechanism by which its pharmacological effects on breast cancer are mediated⁴⁹. On the other hand, it is reported that TAM enhanced the proliferation of endometrial cells in vitro¹⁰ and in vivo^50,51. Molecular mechanisms underlying TAM-induced proliferation of endometrial cells are proposed to involve several pathways, including estrogen receptor (ER)-mediated proliferative effects via PI3K activation^51,52, G-protein coupled estrogen receptor 1 (GPER1)-mediated signaling^14,53, in which TAMs act as ER agonists. In conclusion, although TAM acts effectively as an anti-breast cancer drug, the oxidative DNA damage induced by a TAM metabolite shown in this study may be tumor-initiating, and coupled with the cell proliferation (tumor promotion) induced by TAM and its metabolites as revealed in the references, may contribute to the development of endometrial carcinogenesis.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

TAM

Tamoxifen

4-OH-TAM

4-Hydroxy-tamoxifen

3-OH-TAM

3-Hydroxytamoxifen

3,4-diOH-TAM

3,4-Dihydroxytamoxifen

8-oxodG

8-Oxo-7, 8-dihydro-2’-deoxyguanosine

HOMO

Highest occupied molecular orbital

Fpg

Formamidopyrimidine-DNA glycosylase

Author contributions

SK and MM conceived and designed the study. YM and MM wrote a paper. KM, MM and KF performed experiments and data analysis. SO and YM performed data curation. SK and MM reviewed and critically edited the paper. All authors have read and approved the final manuscript.

Funding

This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (23K24589 for MM).

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.