Polyamine Oxidase Expression Is Downregulated by 17β-Estradiol via Estrogen Receptor 2 in Human MCF-7 Breast Cancer Cells

Polyamine levels decrease with menopause; however, little is known about the mechanisms regulated by menopause. In this study, we found that among the genes involved in the polyamine pathway, polyamine oxidase (PAOX) mRNA levels were the most significantly reduced by treatment with 17β-estradiol in estrogen receptor (ESR)-positive MCF-7 breast cancer cells. Treatment with 17β-estradiol also reduced the PAOX protein levels. Treatment with selective ESR antagonists and knockdown of ESR members revealed that estrogen receptor 2 (ESR2; also known as ERβ) was responsible for the repression of PAOX by 17β-estradiol. A luciferase reporter assay showed that 17β-estradiol downregulates PAOX promoter activity and that 17β-estradiol-dependent PAOX repression disappeared after deletions (−3126/−2730 and −1271/−1099 regions) or mutations of activator protein 1 (AP-1) binding sites in the PAOX promoter. Chromatin immunoprecipitation analysis showed that ESR2 interacts with AP-1 bound to each of the two AP-1 binding sites. These results demonstrate that 17β-estradiol represses PAOX transcription by the interaction of ESR2 with AP-1 bound to the PAOX promoter. This suggests that estrogen deficiency may upregulate PAOX expression and decrease polyamine levels.


Introduction
Polyamines are polycationic alkylamines and include spermidine, spermine, and the precursor diamine putrescine. These small water-soluble molecules are fully protonated in cells owing to the presence of primary and secondary amino groups. They can easily bind to a range of negatively charged molecules, including RNA, DNA, and certain types of proteins and phospholipids. Thus, polyamines are involved in gene regulation, protein and nucleic acid synthesis, and DNA stability and play a role in many physiological processes, including cell proliferation and differentiation [1][2][3].
Polyamines are tightly controlled and regulated by enzymes involved in the anabolism and catabolism of polyamines [4,5]. Ornithine decarboxylase 1 (ODC1) catalyzes the first rate-limiting step of polyamine biosynthesis, which converts ornithine to putrescine. S-adenosylmethionine (SAM) decarboxylase 1 (AMD1) catalyzes the conversion of SAM into decarboxylated SAM. Putrescine promotes the conversion of the AMD1 proenzyme into the active enzyme and enhances AMD1 activity by binding to the allosteric site of AMD1. Spermidine synthase (SRM) converts putrescine into spermidine, and spermine synthase (SMS) converts spermidine into spermine. Spermidine/spermine N1acetyltransferase 1 (SAT1), polyamine oxidase (PAOX), and spermine oxidase (SMOX) are major enzymes that catalyze the backward conversion of polyamines. SAT1 is a ratelimiting enzyme of polyamine catabolism that catalyzes the N1-acetylation of spermidine

E2 Changes PAOX mRNA Levels Most among the Genes Involved in Polyamine Metabolism
The anabolic and catabolic enzymes involved in the polyamine metabolic pathway are shown ( Figure 1A). To detect changes in polyamine metabolic enzymes due to E2 treatment, the mRNA levels of polyamine metabolic enzymes were analyzed in MCF-7 cells treated with or without E2 by conventional PCR ( Figure 1B) and quantitative PCR ( Figure 1C). The data from the conventional PCR were consistent with the quantitative PCR data. The quantitative PCR revealed that the level of GREB1 mRNA, which is known to be induced by E2 [31], increased to 400.1 ± 72.2% in MCF-7 cells. E2 treatment in MCF-7 cells increased the mRNA levels of AMD1 (154.9 ± 21.6%), ODC1 (138.0 ± 12.4%), SRM (152.6 ± 22.6%), and SMOX (176.0 ± 15.1%) but decreased the mRNA levels of SMS (66.5 ± 12.7%), SAT1 (56.1 ± 10.5%), and PAOX (22.6 ± 9.4%).
as a major band of 56 kDa and a minor band of 35 kDa, suggesting isoform 1 (GenBank NP_690875.1) and isoform 2 (GenBank NP_997010.1), respectively (Figure 2A). Consistent with the E2-dependent decrease in PAOX mRNA levels, a significant decrease in levels of both PAOX polypeptides (16.9 ± 1.3% for the major 56-kda band) was observed after E2 treatment ( Figure 2B).  Among the genes encoding enzymes involved in polyamine metabolism, mRNA levels of PAOX showed the most significant change (4.3-fold decrease) after E2 treatment ( Figure 1C). Therefore, changes in PAOX expression at the protein level were investigated in MCF-7 cells with or without E2 treatment. As expected, GREB1 levels were significantly increased to 2708.2 ± 164.2% by E2 treatment (Figure 2). PAOX polypeptides were detected as a major band of 56 kDa and a minor band of 35 kDa, suggesting isoform 1 (GenBank NP_690875.1) and isoform 2 (GenBank NP_997010.1), respectively (Figure 2A). Consistent with the E2-dependent decrease in PAOX mRNA levels, a significant decrease in levels of both PAOX polypeptides (16.9 ± 1.3% for the major 56-kda band) was observed after E2 treatment ( Figure 2B).
ol. Sci. 2022, 23, x FOR PEER REVIEW Figure 2. PAOX expression is downregulated at the protein level upon were treated with or without E2 for 24 h. Expression levels of PAOX (56 k in cell lysates were evaluated by western blotting using corresponding quantitated using the ImageJ software (B). Data are shown as the mean ± GAPDH expression. ** p < 0.01 versus control.
To clearly define the estrogen receptor involved in the E2-dep of PAOX expression, we analyzed E2-mediated PAOX expressio ESR1 or ESR2 knockdown using siRNA. The transfection of ESR MCF-7 cells reduced the protein expression of ESR1 and ESR2, re ESR2 knockdown alleviated the decrease in PAOX induced by E Figure 2. PAOX expression is downregulated at the protein level upon E2 treatment. MCF-7 cells were treated with or without E2 for 24 h. Expression levels of PAOX (56 kDa), GREB1, and GAPDH in cell lysates were evaluated by western blotting using corresponding antibodies (A) and were quantitated using the ImageJ software (B). Data are shown as the mean ± S.D. (n = 3), normalized to GAPDH expression. ** p < 0.01 versus control.

E2 Decreases PAOX Expression via the ESR2
To understand the estrogen receptor signaling pathway involved in E2-mediated PAOX downregulation, we analyzed E2-mediated PAOX expression in MCF-7 cells in the presence of MPP (antagonist of ESR1 (also known as ERα)), PHTPP (antagonist of ESR2 (also known as ERβ)), G-15 (antagonist of G Protein-Coupled Estrogen Receptor 1 (also known as GPER1)), and ICI 182,780 (non-specific ESR antagonist). E2-mediated upregulation of GREB1, which is known to involve ESR1 [31], was blocked by MPP and ICI 182,780. However, the E2-dependent downregulation of PAOX was hampered by PHTPP and ICI 182,780 but was not affected by MPP and G-15 ( Figure 3A). E2-dependent decrease in PAOX mRNA levels was also blocked by PHTPP and ICI 182,780 but not MPP and G-15 ( Figure 3B). These results suggest that E2 reduces PAOX expression through ESR2 in MCF-7 cells.
To clearly define the estrogen receptor involved in the E2-dependent downregulation of PAOX expression, we analyzed E2-mediated PAOX expression in MCF-7 cells with ESR1 or ESR2 knockdown using siRNA. The transfection of ESR1 or ESR2 siRNA into MCF-7 cells reduced the protein expression of ESR1 and ESR2, respectively ( Figure 3C). ESR2 knockdown alleviated the decrease in PAOX induced by E2, while mock transfection or transfection with scrambled siRNA (control) or ESR1 siRNA did not alter the E2-mediated decrease in PAOX ( Figure 3C). As previously reported [32], ESR1 knockdown severely blocked the increase in GREB1 by E2. In addition, it was also confirmed that E2-mediated PAOX repression and GREB1 induction were inhibited by ESR2 knockdown and ESR1 knockdown at the mRNA level, respectively ( Figure 3D). We then analyzed whether the reduction in PAOX promoter activity by E2 was ESR2dependent by treatment with an ESR antagonist and siRNA in a PAOX promoter (−3126/−280)-driven reporter construct. As expected, the reduction in PAOX promoter activity by E2 was restored by PHTPP and ICI 182,780 but was not significantly altered by MPP and G-15 (Supplementary Figure S1A). In addition, transfection with ESR2 siRNA

E2 Reduces the Activity of the PAOX Promoter in an ESR2-Dependent Manner
To understand whether the decrease in PAOX mRNA and protein levels by E2 is due to the decrease in PAOX transcriptional activity, we established constructs in which firefly luciferase cDNA was fused to the PAOX promoter (nucleotide positions −3126 to −280 from the start codon), and changes in PAOX promoter activity by E2 treatment in MCF-7 cells were measured by luciferase activity. Luciferase activity is driven by the PAOX promoter without an enhancer (pGL3-Basic-PAOX promoter (−3126/−280)) and was reduced to 49.9 ± 5.7% by treatment with E2 ( Figure 4). Luciferase activity driven by the PAOX promoter in the presence of the SV40 enhancer (pGL3-Enhancer-PAOX promoter (−3126/−280)) was also reduced to 41.3 ± 1.6% by E2, a slightly larger difference in comparison to the absence of the SV40 enhancer. Therefore, it was confirmed that PAOX promoter activity is suppressed by E2. For subsequent analysis, we used a PAOX-promoterdriven luciferase construct containing the SV40 enhancer, where the effect of E2 can be better observed.
To confirm the importance of AP-1 sites in E2-induced PAOX repression, site-directed mutagenesis at each of the two AP-1 binding sites, and subsequently at both the sites in the PAOX promoter, was performed, and the promoter activity of the reporter We then analyzed whether the reduction in PAOX promoter activity by E2 was ESR2-dependent by treatment with an ESR antagonist and siRNA in a PAOX promoter (−3126/−280)-driven reporter construct. As expected, the reduction in PAOX promoter activity by E2 was restored by PHTPP and ICI 182,780 but was not significantly altered by MPP and G-15 (Supplementary Figure S1A). In addition, transfection with ESR2 siRNA reversed the decrease in PAOX promoter activity by E2, but transfection with ESR1 siRNA did not (Supplementary Figure S1B).

ESR2 Binds to the AP-1 Binding Sites of the PAOX Promoter
We showed that both ESR2 and AP-1 binding sites of the PAOX promoter are important for the E2-dependent repression of PAOX promoter activity. Therefore, the binding of c-JUN and c-FOS, constituting AP-1, and of ESR2 to the two AP-1 binding sites of the PAOX promoter was examined using chromatin immunoprecipitation (ChIP) analysis. ChIP reactions involving ESR1 and two putative half-ERE binding sites were used as negative controls. PCR products encompassing each of the two AP-1 binding sites were observed in all samples immunoprecipitated with the ESR2, c-JUN, or c-FOS antibodies in an E2-dependent manner but not in the sample immunoprecipitated with the ESR1 antibody ( Figure 6A). In addition, no PCR products for the two half-ERE binding sites were observed in samples precipitated with ESR1, ESR2, c-JUN, or c-FOS antibodies. Therefore, c-JUN, c-FOS, and ESR2 were found to specifically bind to each of the two AP-1 binding sites in the PAOX promoter. To confirm the importance of AP-1 sites in E2-induced PAOX repression, site-directed mutagenesis at each of the two AP-1 binding sites, and subsequently at both the sites in the PAOX promoter, was performed, and the promoter activity of the reporter constructs was then measured. Luciferase analysis of the constructs showed that the repression of the PAOX promoter by E2 was reduced when the distal or proximal AP-1 binding site was mutated. In addition, when both AP-1 binding sites were mutated, the PAOX promoter activity was no longer altered by E2 ( Figure 5B).

ESR2 Binds to the AP-1 Binding Sites of the PAOX Promoter
We showed that both ESR2 and AP-1 binding sites of the PAOX promoter are important for the E2-dependent repression of PAOX promoter activity. Therefore, the binding of c-JUN and c-FOS, constituting AP-1, and of ESR2 to the two AP-1 binding sites of the PAOX promoter was examined using chromatin immunoprecipitation (ChIP) analysis. ChIP reactions involving ESR1 and two putative half-ERE binding sites were used as negative controls. PCR products encompassing each of the two AP-1 binding sites were observed in all samples immunoprecipitated with the ESR2, c-JUN, or c-FOS antibodies in an E2dependent manner but not in the sample immunoprecipitated with the ESR1 antibody ( Figure 6A). In addition, no PCR products for the two half-ERE binding sites were observed in samples precipitated with ESR1, ESR2, c-JUN, or c-FOS antibodies. Therefore, c-JUN, c-FOS, and ESR2 were found to specifically bind to each of the two AP-1 binding sites in the PAOX promoter. To analyze whether ESR2 binds to the AP-1 (c-JUN and c-FOS) bound to each of the two AP-1 binding sites of the PAOX promoter in an E2-dependent manner, Re-Immunoprecipitation (Re-IP) analysis was performed by immunoprecipitation of chromatin extracts first with anti-ESR1 or anti-ESR2 antibodies and secondarily with c-JUN or c-FOS antibodies. PCR products for each of the two AP-1 binding sites were observed in immunoprecipitates with the ESR2 antibody and then the c-FOS or c-JUN antibody in an E2dependent manner ( Figure 6B). PCR products for AP-1 binding sites were not detected in any immunoprecipitates with the ESR1 antibody. Therefore, we concluded that ESR2 binds to AP-1 (c-JUN and c-FOS) and is bound to each of the two AP-1 binding sites of the PAOX promoter in an E2-dependent manner. Chromatin extracts were prepared from MCF-7 cells treated with or without E2, and immunoprecipitation of the extracts was performed using the indicated antibodies for ChIP assays (B). Chromatin extracts were immunoprecipitated using an anti-ESR1 antibody or anti-ESR2 antibody and then with control anti-IgG, anti-c-JUN, or anti-c-FOS antibody for Re-IP assays (C). PCR was performed with the above primer pairs, using DNA extracted from the immunoprecipitates as templates. PCR products were analyzed using 5% PAGE gels. . ESR2 interacts with AP-1, which binds to the AP-1 binding sites in the PAOX promoter in an E2-dependent manner. Primer pairs for PCR amplification of the AP-1 binding site and the half-ERE binding site in the PAOX promoter are indicated by arrows (A). The regions for PCR were −2896 to −2710 for the distal AP-1 site, −1271 to −1080 for the proximal AP-1 site, −2730 to −2477 for the distal putative half-ERE site, and −1100 to −1003 for the proximal putative half-ERE sites. Chromatin extracts were prepared from MCF-7 cells treated with or without E2, and immunoprecipitation of the extracts was performed using the indicated antibodies for ChIP assays (B). Chromatin extracts were immunoprecipitated using an anti-ESR1 antibody or anti-ESR2 antibody and then with control anti-IgG, anti-c-JUN, or anti-c-FOS antibody for Re-IP assays (C). PCR was performed with the above primer pairs, using DNA extracted from the immunoprecipitates as templates. PCR products were analyzed using 5% PAGE gels.
To analyze whether ESR2 binds to the AP-1 (c-JUN and c-FOS) bound to each of the two AP-1 binding sites of the PAOX promoter in an E2-dependent manner, Re-Immunoprecipitation (Re-IP) analysis was performed by immunoprecipitation of chromatin extracts first with anti-ESR1 or anti-ESR2 antibodies and secondarily with c-JUN or c-FOS antibodies. PCR products for each of the two AP-1 binding sites were observed in immunoprecipitates with the ESR2 antibody and then the c-FOS or c-JUN antibody in an E2-dependent manner ( Figure 6B). PCR products for AP-1 binding sites were not detected in any immunoprecipitates with the ESR1 antibody. Therefore, we concluded that ESR2 binds to AP-1 (c-JUN and c-FOS) and is bound to each of the two AP-1 binding sites of the PAOX promoter in an E2-dependent manner.

Discussion
Herein, we examined changes in the expression of polyamine metabolic enzymes at the RNA level in ER-positive MCF-7 breast cancer cells, with and without the influence of E2. We found that E2 upregulates the mRNA levels of anabolic enzymes of polyamine metabolism, including AMD1, ODC1, and SRM but not SMS, which catalyzes the conversion of spermidine to spermine, and downregulates the mRNA levels of catabolic enzymes such as SAT1, and PAOX but not SMOX, which catalyzes the direct conversion of spermine to spermidine. We also found that, among polyamine metabolic enzymes, PAOX was most significantly altered by E2 in MCF-7 cells at the mRNA level. Furthermore, PAOX was reduced by E2 at both the protein and mRNA levels. As PAOX converts N1acetylspermine and N1-acetylspermidine produced by SAT1 to spermidine or putrescine by oxidation, decreased PAOX expression results in increased polyamine and decreased H 2 O 2 levels [33,34].
Estrogen transmits signals through three ESRs, namely, ESR1, ESR2, and the G-proteincoupled estrogen receptor 1 (GPER1) [35][36][37]. Once E2 binds to ESR1 or ESR2 in the cytoplasm, the ESRs undergo a structural change that induces receptor dimerization [38]. These E2-ESR complexes translocate to the nucleus and play a role in gene regulation by binding to the estrogen response element (ERE)-binding sites or transcription factors such as AP-1 and SP-1 [39][40][41][42]. In contrast to the ESR1 and ESR2, the GPER1 shows characteristics of G-protein-coupled receptors that act via the activation of the heterotrimeric G protein [43]. The GPER1 is anchored in the plasma membrane, and unlike nuclear estrogen receptors, signaling pathways involving GPER1 activation occur through a variety of secondary messengers via G-protein activation [44].
ESRs can regulate the expression of target genes by directly binding to EREs identified in estrogen-responsive promoters [32]. In addition, ESRs can regulate gene expression through protein-protein interactions with other transcription factors, such as the SP-1 and AP-1 [39][40][41], and co-regulators that have chromatin remodeling functions [45,46]. Approximately one-third of estrogen target genes in humans do not have an ERE or an ERE-like region [39].
ESRs can activate or repress the transcription of target genes. Molecular mechanisms that increase target gene expression through E2-ESR have been studied more than those that repress expression [47]. However, because nearly 70% of E2-regulated genes are repressed in MCF-7 cells [48], estrogen-mediated repression of target genes appears to be more common than activation. In estrogen-stimulated MCF-7 cells, the 6024 ESR1 and 9702 ESR2 binding sites, where the ESR1, ESR2, or both ER subtypes can bind, were identified using ChIP-Seq analysis [49]. Based on these results, it is expected that ESR2 is involved in the regulation of more genes than ESR1, at least in MCF-7 cells.
We established that ESR2 is responsible for E2-dependent PAOX expression repression and reporter assays using antagonists and siRNAs; however, the mechanism by which the E2-ESR2 complex downregulates PAOX expression was unclear. When we deleted the two AP-1 binding sites of the PAOX promoter, PAOX promoter activity in the absence of E2 gradually increased, consistent with a previous report [50]. More importantly, the deletion or mutation of the two AP-1 binding sites of the PAOX promoter gradually leads to the loss of E2-dependent promoter activity repression, while the deletion of two putative half-ERE and SP-1 binding sites did not affect promoter activity. ChIP and Re-IP assays revealed that the ESR2 as well as the c-FOS, and the c-JUN bound to each of the two AP-1 binding sites in the PAOX promoter, but the ESR2 did not bind to either of the two putative half-ERE sites. We confirmed that the E2-ESR2 complex binds to AP-1 bound to the two AP-1 binding sites of the PAOX promoter, thereby repressing PAOX transcription in MCF-7 cells.
The molecular mechanism of target gene downregulation by E2-ESR2 has been reported previously. For example, TNFAIP1 is downregulated by the binding of the E2-ESR2 directly to the ERE site on its promoter in primary hippocampal cells [51]. In addition, ESR1 expression is repressed by the indirect binding of E2-ESR2 to SP-1 in MCF-7 cells [52]. As a novel finding, we showed that the E2-ESR2 complex downregulates the expression of PAOX by binding to the AP-1 bound to AP-1 binding sites. Therefore, to the best of our knowledge, our finding is the first example of E2 repressing a target gene through ESR2-AP-1 interaction, independent of ERE.
In this study, we found that E2 upregulated the expression of most anabolic enzymes and downregulated the expression of most catabolic enzymes involved in polyamine metabolism. In addition, we demonstrated that PAOX expression, which was most significantly altered by E2 in MCF-7 cells, was repressed by the binding of the E2-ESR2 complex to AP-1 bound to the two AP-1 binding sites of the PAOX promoter.
Our data suggest that an increase in estrogen level would lead to an increase in the expression of polyamine anabolic enzymes, including the AMD1, ODC1, and SRM, and a decrease in the expression of polyamine catabolic enzymes, including SAT1 and PAOX, which leads to an increase in polyamine levels. This finding is consistent with a previous report that polyamine levels were increased upon E2 treatment in MCF-7 cells [53]. This estrogen-dependent change in polyamine levels may help to understand the hormone therapy of breast cancer and side effects of menopause, from the perspective of polyamine. A selective estrogen receptor modulator (SERM), such as tamoxifen, can reduce the incidence of hormone-receptor-positive breast cancer and the likelihood of cancer recurrence [54]. SERM treatment reduces polyamine levels in MCF-7 cells [55]. It is well documented that polyamines play a role in tumorigenesis, at least by increasing cell proliferation and suppressing apoptosis [4]. For example, the ODC1 inhibitor α-difluoromethylornithine is known to prevent the growth-promoting effect of E2 [56,57]. Moreover, sensitivity to the growth-promoting effects of E2 was reduced by about a third in MCF-7 cells overexpressing ODC1 compared with vector-transfected cells [58]. These results suggest that elevated polyamine levels contribute to estrogen-induced oncogenic phenotypes in breast cancer cells. In addition, a decrease in estrogen during menopause leads to a decrease in polyamine levels. Polyamine reduction is a common occurrence in postmenopausal women and has many clinical implications [59][60][61][62]. For example, polyamines effectively prevent bone loss in ovariectomized mice [20]. In addition, polyamine reduction leads to increased H 2 O 2 and ROS levels, which, as previously mentioned, contribute to inflammatory diseases [63]. Therefore, we propose that preventing polyamine decline, for example by repressing or inhibiting PAOX, could be a potential treatment for alleviating postmenopausal symptoms.

RNA Isolation and PCR
The total RNA was isolated from MCF-7 cells using TRIZOL (Invitrogen, Thermo Fisher Scientific), as described previously [65]. The cDNA mixtures were synthesized from the total RNA by reverse transcription-PCR using oligo dT 15 primers and the AMV RT system (Promega, Madison, WI, USA) according to the manufacturer's instructions. Conventional PCR was performed in a final volume of 10 µL containing 1 pM each of the 5 primer and 3 primer, 0.2 mM dNTPs, 1× Taq PCR buffer, 50 U/mL Taq polymerase, and cDNAs synthesized from 0.1 µg total RNA. PCR was performed with 25-30 cycles; the cycling conditions were as follows: denaturation at 94 • C for 30 s, annealing at the appropriate annealing temperature (Supplementary Table S1) for 30 s, and extension at 72 • C for 30 s. The PCR products were detected by a 5% PAGE analysis and visualized by ethidium bromide staining. Real-time PCR was conducted using a QuantiTect SYBR Green PCR kit (Qiagen, Hilden, Germany) and QuantStudio 3 Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). The expression of the target genes was normalized to that of GAPDH expression.

Western Blotting
The cells were harvested using an SDS sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 0.1% bromophenol blue, and 10% glycerol) on ice for 15 min and centrifuged at 18,000 × g for 15 min after brief sonication. The cell lysate samples were boiled after the addition of 100 mM β-mercaptoethanol for 4 min and resolved by SDS-PAGE. Proteins in the gel were blotted onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). The blots were incubated with the indicated primary and secondary antibodies. Immunoreactive signals were detected using the Immobilon Western HRP substrate (Millipore, Billerica, MA, USA) and an LAS-3000 detector (Fujifilm, Tokyo, Japan) as previously described [66].

Prediction of Transcription Factor-Binding Site
The transcription factor-binding sites on the PAOX promoter were searched using CiiiDER, an integrated computational toolkit [67], and the transcription factor-binding site database from JASPAR (http://jaspar.cgb.ki.se/, accessed on 23 September 2021).

Cloning of the Human PAOX Promoter in Luciferase Reporter Vectors
To generate a luciferase reporter construct for the human PAOX promoter (GenBank accession no. NM_152911.4), a fragment of the PAOX promoter −3126/−280 (the first nucleotide upstream of the start codon is referred to as −1) was PCR-amplified using genomic DNA from human foreskin fibroblasts (as previously described [68]) as a template with PrimeSTAR ® GXL DNA polymerase (TaKaRa, Shiga-ken, Japan). The deletion products were PCR-amplified using the PAOX promoter (−3126/−280) as a template, using the indicated primer pairs (Supplementary Table S2). The PCR products were cloned into pGL3-Basic or pGL3-Enhancer luciferase reporter vectors digested with Acc65I and HindIII (Promega, Madison, WI, USA) using the In-Fusion Cloning system (Takara). Luciferase reporter constructs harboring the human PAOX promoter were confirmed to be error-free by DNA sequencing.

Dual-Luciferase Reporter Assay
The transfection of the luciferase constructs into MCF-7 cells was performed using Lipofectamine 2000. The MCF-7 cells were plated in 96-well plates at a density of 5 × 10 4 cells/well and incubated in 100 µL of DMEM supplemented with 10% FBS for 12 h. Promoterless pGL3-Basic (0.11 µg) or pGL3-Basic-PAOX promoter constructs (0.18 µg) encoding firefly luciferase driven by PAOX promoter, or promoterless pGL3-Enhancer (0.12 µg) or pGL3-Enhancer-PAOX promoter constructs (0.19 µg of pGL3-Enhancer-PAOX promoter(-3126/-280) or the amount corresponding to the size of each construct) and pRL-TK encoding Renilla luciferase driven by herpes simplex virus thymidine kinase promoter (0.02 µg; Promega, Madison, WI, USA) in 25 µL Opti-MEM (Gibco/Thermo Fisher Scientific) were incubated with Lipofectamine 2000 (0.75 µL) and diluted in 25 µL Opti-MEM for 30 min at room temperature. Cells in Opti-MEM were treated with this mixture for 5 h and then replaced with DMEM supplemented with 10% FBS for 12 h. The cells were then incubated with PRF-DMEM supplemented with 10% charcoal-stripped FBS for 24 h. Next, the cells were treated with or without E2 in PRF-DMEM supplemented with 10% charcoal-stripped FBS for 24 h. Luciferase activity was measured using the dual-luciferase reporter assay system (Promega), and the firefly luciferase activity in transfected cells was normalized to that of the Renilla luciferase.

Site-Directed Mutagenesis (SDM)
Mutations in the distal and proximal AP-1 sites in the PAOX promoter were introduced by overlap extension PCR mutagenesis using the PrimeSTAR ® GXL DNA polymerase. The sequences of the primer pairs used are listed in Supplementary Table S3. To generate a mutation in the distal AP-1 site, two primary PCR products with Primer −3126-F and mAP-1-D-R and with Primer mAP-1-D-F and −280-R using the pGL3-Enhencer-PAOX promoter (−3126/−280) as a template were obtained. Similarly, to generate a mutation in the proximal AP-1 site, two primary PCR products with Primer −3126-F and AP-1-P-R and with Primer mAP-1-P-F and −280-R using the pGL3-Enhencer-PAOX promoter (−3126/−280) as a template were obtained. In addition, to generate double mutations in the proximal and distal AP-1 sites, two primary PCR products with Primer −3126-F and AP-1-P-R and Primer mAP-1-P-F and −280-R using the pGL3-Enhencer-PAOX promoter (−3126/−280)-mAP-1-D as a template were obtained. Then, a second PCR was performed to create an overlap-extension PCR product using a mixture of PCR products for distal AP-1 site mutation or proximal AP-1 site mutation as a template, as well as Primer −3126 and −280. Overlap-extended PCR products harboring mutations in the distal or proximal AP-1 site or double mutations in the distal and proximal AP-1 sites were cloned into the pGL3-Enhancer vector cleaved with Acc65I and HindIII to generate the pGL3-Enhancer-PAOX promoter (−3126/−280)-mAP-1-D, -mAP-1-P, and -mAP-1-DP, respectively. Luciferase reporter constructs with mutations in AP-1 sites of the PAOX promoter were confirmed to be error-free by DNA sequencing.

ChIP and Re-IP
ChIP was performed as previously described [69], with minor modifications. The MCF-7 cells were treated with E2 or vehicle for 24 h and fixed with 1% formaldehyde at room temperature for 10 min. The reaction was stopped by adding 0.125 M glycine and washing twice after 5 min with ice-cold PBS. The cells were lysed for 10 min by adding a cell lysis/wash buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 5 mM EDTA, 0.5% NP-40, 1.0% Triton X-100) at 4 • C and a protease-inhibitor cocktail (Sigma-Aldrich, MO, USA). Then, cells were centrifuged at 1500 rpm at 4 • C for 5 min, and the pellet containing the chromatin fraction was lysed in a shearing buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.0) and protease inhibitor cocktail. The chromatin extracts containing genomic DNA were sonicated on ice to obtain an average fragment size of 500 bp. The supernatant was diluted 9 times with a diluted buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH 8.1, 167 mM NaCl) and subsequently immunoprecipitated using antibodies and Protein G-agarose (Invitrogen) at 4 • C. The immunoprecipitates were washed once with cold cell lysis/wash buffer and three times with cold ethanol. DNA purification followed by protein degradation was performed using Chelex 100 (Bio-Rad, Hercules, CA, USA), following the manufacturer's instructions. The PCR was performed in 30 cycles under the following conditions: denaturation at 94 • C for 30 s, annealing at an indicated annealing temperature (Supplementary Table S4) for 30 s, and extension at 72 • C for 30 s. The Re-IP assay was performed as previously described [70] with minor modifications. The immunoprecipitated antibody-protein-DNA complexes were washed twice with the cell lysis/wash buffer and eluted by incubating for 30 min at 37 • C in a Re-IP elution buffer (TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA), 2% SDS, 15 mM DTT). After centrifugation, the supernatant was diluted 20 times using a dilution buffer and individually re-immunoprecipitated with each secondary antibody at 4 • C. The immune complexes were incubated with protein G-agarose beads for 2 h at 4 • C. The DNA was purified from the secondary immune complex using Chelex 100. The PCR was performed using the extracted DNA as described above.

Statistical Analyses
All data are presented as a mean ± standard deviation of at least three independent experiments. Statistical significance was analyzed using the unpaired two-tailed Student's t-test. A p-value < 0.05 was considered statistically significant.

Conclusions
In this study, we showed that PAOX was significantly altered among polyamine metabolism-related enzymes upon E2 treatment in MCF-7 cells. We demonstrated that PAOX expression was downregulated by E2 via ESR2. The ESR2 binds to the AP-1 complex bound to two AP-1 binding sites in the PAOX promoter. The E2-ESR2-AP-1 complex binds to the PAOX promoter and downregulates PAOX expression. Based on our findings, we suggest that E2-dependent inhibition of PAOX is at least one cause of polyamine decline during menopause