Visual Overview
Abstract
N-acetyltransferase 1 (NAT1) is an enzyme that metabolizes carcinogens, which suggests a potential role in breast carcinogenesis. High NAT1 expression in breast tumors is associated with estrogen receptor α (ERα+) and the luminal subtype. We report that NAT1 mRNA transcript, protein, and enzyme activity were higher in human breast tumors with high expression of ERα/ESR1 compared with normal breast tissue. There was a strong correlation between NATb promoter and NAT1 protein expression/enzyme activity. High NAT1 expression in tumors was not the result of adipocytes, as evidenced by low perilipin (PLIN) expression. ESR1, NAT1, and XBP1 expression were associated in tumor biopsies. Direct regulation of NAT1 transcription by estradiol (E2) was investigated in ERα (+) MCF-7 and T47D breast cancer cells. E2 did not increase NAT1 transcript expression but increased progesterone receptor expression in a dose-dependent manner. Likewise, NAT1 transcript levels were not increased by dihydrotestosterone (DHT) or 5α-androstane-3β, (3β-adiol) 17β-diol. Dithiothreitol increased levels of the activated, spliced XBP1 in ERα (+) MCF-7 and T47D breast cancer cells but did not affect NAT1 or ESR1 expression. We conclude that NAT1 expression is not directly regulated by E2, DHT, 3β-adiol, or dithiothreitol despite high NAT1 and ESR1 expression in luminal A breast cancer cells, suggesting that ESR1, XBP1, and NAT1 expression may share a common transcriptional network arising from the luminal epithelium associated with better survival in breast cancer. Clusters of high-expression genes, including NAT1, in breast tumors might serve as potential targets for novel therapeutic drug development.
Introduction
N-acetyltransferase 1 (NAT1) genetic polymorphism is associated with breast cancer risk and mortality (Zheng et al., 1999; Millikan, 2000; Andres et al., 2015). Recent meta-analyses suggest the associations of NAT1 genetic polymorphisms with pancreatic (Zhang et al., 2015) and urinary bladder (Dhaini et al., 2017) cancers. Congenic rats with higher arylamine N-acetyltransferase 2 activity (homologous to human NAT1) exhibit greater carcinogen-induced mammary tumor susceptibility (Stepp et al., 2017). Overexpression of NAT1 enhances breast cancer cell proliferation and was associated with a gene signature of epithelial-to-mesenchymal transition in breast tumors (Adam et al., 2003; Savci-Heijink et al., 2016). Small molecule inhibition of NAT1 decreases proliferation and invasiveness (Tiang et al., 2010), and both genetic and small-molecule inhibition of NAT1 reduce anchorage-independent growth (Stepp et al., 2018) in human breast cancer cells. Thus, the role of NAT1 in breast carcinogenesis or progression remains unclear and requires further investigation.
Human NAT1 is located in chromosome 8p 21.3. NAT1 has two alternative promoters (NATa and NATb) and the NAT1 open reading frame is contained within the 3′ exon. The NATb promoter is located 11.8 kb upstream of the NAT1 coding exon (Husain et al., 2004, 2007), whereas an alternative NATa promoter is located 51.5 kb upstream of the NAT1 coding exon (Barker et al., 2006). NATb is the major promoter for NAT1 and is responsible for ubiquitous NAT1 expression (Husain et al., 2004). However, NATb lacks a TATA box, and thus correlates with low-to-moderate NAT1 expression (Husain et al., 2004). SP1 detected in the minimal NATb region is expressed ubiquitously and is probably responsible for wide NATb activity in every tissue throughout the human body (Sadrieh et al., 1996; Windmill et al., 2000). NATa is most active in specific tissues such as kidney, liver, lung, and trachea (Barker et al., 2006). In human breast tissues, NAT1 expression is detectable at the mRNA, protein, and enzyme activity levels (https://www.gtexportal.org/home/gene/NAT1), but whether expression is driven by NATb or NATa remains unknown.
NAT1 was reported to be overexpressed in breast tumors, and high NAT1 expression was mainly found in luminal tumors characterized by high estrogen receptor (ERα/ESR1) expression (Perou et al., 2000; Sorlie et al., 2001; Adam et al., 2003). Along with other gene signatures, including ESR1 and progesterone receptor (PGR), high NAT1 expression in luminal-type breast carcinomas is associated with better prognosis and longer relapse-free survival (Perou et al., 2000; Sorlie et al., 2001). Immunohistochemical staining showed NAT1 levels correlated with ERα positivity and that expression of NAT1 in ERα (−) breast cancer was low (Tozlu et al., 2006). Further, NAT1 expression was higher in breast cancer cell lines compared with other cancer cell lines (Ross et al., 2000). NAT1 expression was also significantly associated with increased overall survival and with disease-free survival in breast cancer patients who had received tamoxifen as an adjuvant therapy (Endo et al., 2013). Functionally, overexpression of NAT1 increased the proliferative capacity of the HB4a normal breast epithelium-derived cells in vitro, possible by upregulating the methionine salvage pathway (Adam et al., 2003; Witham et al., 2017). Interestingly, the correlation between NAT1 and ERα observed in primary tumors is lost in metastatic tumors from the same patients (Butcher and Minchin, 2012). Despite these observations, the mechanism for the correlation between NAT1 and ERα expression in luminal breast cancer is not yet clear.
Analysis of published microarray studies revealed a positive correlation between overexpression of NAT1 and ER positivity (Wakefield et al., 2008). One study showed a 2-fold increase in NAT1 expression in estradiol (E2)-treated MCF-7 cells (Finlin et al., 2001). In androgen receptor (AR)-positive 22Rv1 and LNCaP prostate cancer cells, NAT1 expression was induced by the synthetic androgen R1881, although the increased NAT1 transcription was not direct but was mediated instead by a secondary effect of upregulating the expression of heat shock factor (HSF1) that bound to NAT1 promoter to increase transcription (Butcher and Minchin, 2010). Interestingly, NAT1 was associated with AR expression as well as Trefoil factor 3, apolipoprotein D, and activated leukocyte cell adhesion molecule in a subset of ERα (−)/PGR (−) human breast tumors, but no study has examined androgen regulation of NAT1 expression in breast cancer (Doane et al., 2006).
The association between ERα and NAT1 expression described above was reviewed (Butcher and Minchin, 2012). We investigated NAT1 expression in matched tumor and normal human breast tissue specimens and in ERα (+) MCF-7 and T47D breast cancer cell lines to assess the regulation of NAT1 expression by E2 and selective ER modulators.
Material and Methods
Patients and Human Samples.
De-identified breast tumor and normal tissues of newly diagnosed breast cancer patients were obtained from Louisville Repository Tissue Network and University of Louisville Department of Surgery. The samples were collected and stored in liquid nitrogen immediately after surgery. The ER (ERα) status was clinically provided from immunohistochemical staining analyses. All procedures were performed under the approval of University of Louisville Institutional Review Board.
RNA Preparation and Quantitative Real-Time Reverse Transcription–Polymerase Chain Reaction.
Total RNA was extracted from frozen breast tissue specimens using RNeasy kit (Qiagen, Valencia, CA) following the manufacturer’s instructions. The purity of extracted RNA was determined by measuring the absorbance at 280/260 nm. Complementary DNA was synthesized from 1 μg RNA using the Superscript III Reverse Transcriptase kit (Invitrogen/Thermo Fisher Scientific, Carlsbad, CA). Total NAT1, NATb, and NATa mRNA transcripts were measured by quantitative real-time reverse transcription–polymerase chain reaction (RT-PCR) using a probe and a forward primer designed in the 79-bp exon region, and the reverse primer in the protein coding exon as shown in Table 1 (Husain et al., 2004; Barker et al., 2006). ERα (ESR1, Hs00174860_m1), perilipin (PLIN, Hs00160173_m1), progesterone receptor (PGR, Hs00172183_m1), and X-box binding protein 1 (XBP1, Hs00964359_m1) were measured using ABI predesigned primers and probes (Applied Biosystems, Foster City, CA). 18S rRNA (Applied Biosystems) was used as a reference to normalize the RT-PCR results.
Primers and probes for quantitative real-time RT-PCR
Measurement of NAT1 N-Acetyltransferase Activity in Tumor and Normal Breast Tissue Samples.
Breast tumor and normal tissues were homogenized in 20 mM sodium phosphate, pH 7.4, containing EDTA (1 mM), dithiothreitol (DTT: (1 mM) and protease inhibitors (1 μg/ml), phenylmethylsulfonyl fluoride (100 μM), and pepstatin (0.75 μM) using a Tissue Tearor (Biospec Products, Bartlesville, OK), and centrifuged at 100,000g for 1 hour at 4°C. Total cellular protein was measured using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA). NAT1 enzyme activity in human mammary tissues was detected using p-aminobenzoic acid as substrate. Reactions containing suitable diluted cell cytosol protein, p-aminobenzoic acid (100 μM), and 1 mM acetyl-coenzyme A were incubated at 37°C for 30 minutes. Reactions were stopped by adding acetic acid. Reactants and products were separated by high-performance liquid chromatography (Beckman-Coulter, Fullerton, CA) and quantified by their absorbance at 280 nm (Hein et al., 2006).
Culture of the Human Breast Cancer Cell Lines.
MCF-7 and T47D cells were obtained from ATCC (Manassas, VA). MCF-7 cells were cultured in modified Iscove’s modified Eagle’s medium (IMEM) (Invitrogen) supplemented with 10% fetal bovine serum (HyClone, Logan, UT). T47D cells were cultured in RPMI 1640 medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS; HyClone). Penicillin (100 IU/ml) and streptomycin (100 μg/ml) were added to complete medium just prior to use. Cell culture media was routinely changed every 3–4 days. For experiments, cells were grown in phenol red-free IMEM (MCF-7) or RPMI 1640 (T47D) supplemented with 5% FBS charcoal/dextran-stripped to remove hormones.
Hormone Treatment of the Human Breast Cancer Cells.
T47D and MCF-7 cells were grown in phenol red-free medium containing 5% charcoal/dextran-stripped FBS for 48 or 72 hours before each experiment. Different concentrations of E2 (Sigma-Aldrich, St. Louis, MO) ranging from 10−10 to 10−7 M were tested. In additional experiments, cells were treated with 10 nM E2, 100 nM/1 μM 4-hydroxytamoxifen (4-OHT; Sigma-Aldrich), or 100 nM ICI 182,780 (fulvestrant; Tocris, Ellisville, MO) alone or in combination for 6 or 24 hours. In other experiments, cells were treated with dihydrotestosterone (DHT), 5α-androstane-3β , 17β-diol (3β-adiol), both from Sigma-Aldrich. The selective AR modulator/antiandrogen bicalutamide (Casodex) was generously provided by AstraZeneca (Macclesfield, UK) (Teng et al., 2013). RNA was extracted as described above and quantitative real-time RT-PCR was performed to determine NAT1 and PGR mRNA expression.
Dithiothreitol Treatment of the Human Breast Cancer Cells.
MCF-7 cells were treated with 5 mM dithiothreitol (DTT; Sigma-Aldrich) for 4 hours when the cells had grown to 70–80% confluence in six-well plates. Cells were collected and RNA was extracted using RNeasy kit (Qiagen). cDNA was synthesized using Superscript III Reverse Transcriptase kit (Invitrogen). PCR was performed under the following conditions: 94°C for 3 minutes, followed by 30 cycles of 94°C for 30 seconds, 62°C for 1 minute, 72°C for 1.5 minutes, and then 72°C for 5 minutes. The XBP1 forward primer sequence was 5′-GAAGCCAAGGGGAATGAAGTGAGG-3′ and the reverse primer sequence was 5′-CATGGGGAGGAGATGTTCTGGAGGGG-3′.
Analysis of ESR1, FOXA1, GATA3, NAT1, and XBP1 RNA Expression between ERα (+) and ERα (−) Breast Tumor Samples.
Publically available data from The Cancer Genome Atlas (Cancer Genome Atlas Research Network et al., 2013) for the breast invasive carcinoma cohort were downloaded and analyzed (N = 1043) using FirebrowseR (Deng et al., 2017), an R client to the Broad Institute’s RESTful Firehose Pipeline. Differences in RNA expression of ESR1, FOXA1, GATA3, NAT1, and XBP1 between primary solid breast tumor samples stratified by ERα status (+ or −) were evaluated statistically by Mann-Whitney for each gene. RNA expression values are reported in RSEM from the Cancer Genome Atlas Research Network (RNA-Seq by Expectation-Maximization) (Li and Dewey, 2011).
Results
ESR1 Expression in ERα (+) and ERα (−) Breast Tumor Samples.
ESR1 mRNA was measured in 21 breast tumors and mRNA measurement in ERα (+) MCF-7 cells was used as a reference (Fig. 1). As expected, ESR1 expression differed significantly (P < 0.0025) between ERα (+) versus ERα (−) breast tumors with higher expression in ERα (+) tumors. For subsequent figures, the labeling of ERα status was consistently determined the same way.
ESR1 mRNA expression in ERα (+) and ERα (−) human breast tumors. Quantitative real-time RT-PCR was used to measure ESR1 mRNA expression that is shown for individual breast tumors and was normalized relative to expression in MCF-7 cells, i.e., set to 1, far right. Breast tumors listed on the abscissa are grouped into ERα (+) and ERα (−) on the basis of immunohistochemical staining information. Inserted graph: Each bar represents the mean ± S.E.M. of ESR1 mRNA in ERα (+) (n = 16) and ERα (−) (n = 5) breast tumor samples.
NAT1 mRNA Expression Correlates with NAT1 Enzymatic Activity in the Breast Samples.
NAT1 expression (both mRNA and catalytic activity for p-aminobenzoic acid N-acetylation) was measured in the breast tumor and normal breast tissues. Both NAT1 mRNA (Fig. 2A) and N-acetyltransferase activity (Fig. 2B) were significantly higher in ERα (+) breast tumors compared with ERα (−) tumors, normal breast tissues, and MCF-7 cells. However, as observed in the tumor sample no. 3, there were exceptions to high NAT1 mRNA and N-acetyltransferase activity. We also observed that breast tumor tissue from patient no. 9 had high NAT1 mRNA (Fig. 3A) but not high NAT1 N-acetyltransferase activity, a result that might be a consequence of NAT1 degradation, or other mechanisms resulting in reduced enzyme activity. The Pearson correlation coefficient between NAT1 mRNA and enzyme activity in 19 breast tumors was 0.709 (p < 0.0001).
NAT1 mRNA levels and N-acetyltransferase activity in breast tumor/normal breast samples. (A) Quantitative real-time RT-PCR was used to measure NAT1 mRNA expression in the same breast tumor samples as in Fig. 1. (B) NAT1 N-acetyltransferase activity was measured in the tumor samples using p-aminobenzoic acid as a substrate. No enzyme activity data for sample 1. The y-axis was divided into two segments to accommodate the wide range of values. Each bar represents the mean ± S.E.M. of NAT1 mRNA expression in ERα (+) (n = 16) and ERα (−) (n = 5) and NAT1 N-acetyltransferase activity in ERα (+) (n = 15) and ERα (−) (n = 5) breast tumor samples.
(A) Perilipin (PLIN) mRNA expression in the breast tumor/normal tissue samples. PLIN mRNA was measured by quantitative real-time RT-PCR in breast tumor and normal breast tissues and normalized to expression in MCF-7 cells. The tumor/breast sample numbers correspond to those in Figs. 1 and 2. The y-axis was divided into three segments owing to the low level of PLIN expression in tumor in comparison with normal breast tissue. (B) Each bar is the mean ± S.E.M. of PLIN mRNA expression in breast normal (n = 21) and tumor tissues (n = 21).
Perilipin Expression in ERα (+) and ERα (−) Breast Tumor Samples.
Adipose cells account for 90% of the total mass of the normal breast tissue (Nishidate et al., 2004). Perilipin (PLIN) is expressed exclusively in adipocytes in mammary gland (Saito-Hisaminato et al., 2002). To evaluate the relative purity of the breast tumor/tissue samples in our study, PLIN expression was measured in the same set of matched tumor/normal samples. Breast tumor samples had very low PLIN mRNA expression; in contrast, matched normal samples had much higher PLIN mRNA expression confirming that NAT1 expression was contributed primarily by breast tumor cells in our tumor samples (Fig. 3).
NATa and NATb Promoter Expression in the Breast Samples.
To better evaluate if the high NAT1 expression in these breast tumors was at the transcriptional level, the promoter utilization was examined. NAT1 is transcribed from a major promoter, NATb, and an alternative promoter, NATa, resulting in mRNAs with distinct 5′-untranslated regions (UTR) (Millner et al., 2012). Quantitative real time RT-PCR was used to measure NATa, NATb, and NAT1 mRNA expression in breast tumors. NATa was not detected or was used at very low levels in these breast tumors (data not shown). As shown in Fig. 4, NATb mRNA expression was high and correlated with NAT1 mRNA expression (Fig. 2A) in breast tumors (R2 = 0.60, P < 0.0001). Thus, we conclude that the high level of NAT1 transcripts is primarily from the major NATb promoter. ERα (+) breast tumors had a significantly higher NATb mRNA expression than the ERα (−) breast tumors (P = 0.0275).
NATb utilization in breast samples. Quantitative real-time RT-PCR was used to measure NATb expression in 21 breast tumors and expression in MCF-7 cells was used as a reference control. The tumor/breast sample numbers correspond to those in Figs. 1–3. The y-axis was broken into two segments. Inserted graph: Each bar represents the mean ± S.E.M. of NATb mRNA in ERα (+) (n = 16) and ERα (−) (n = 5) breast tumor samples.
NAT1 Expression Is Not E2-Regulated in ERα (+) Breast Cancer Cells.
Because NAT1 mRNA expression (Fig. 2A) correlated with ERα expression (Fig. 1), we determined if E2 stimulated NAT1 expression in ERα (+) breast cancer cell lines. As a positive control, we observed that E2 increased PGR mRNA expression in a concentration-dependent manner in MCF-7 cells; however, NAT1 expression was not increased by E2 (Fig. 5A). The active metabolite of tamoxifen, 4-OHT, inhibited E2-induced PGR expression but did not affect NAT1 expression (Fig. 5, B and C). These data indicate that E2 does not directly stimulate NAT1 transcription in these breast cancer cells.
NAT1 and PGR expression in MCF-7 cells. MCF-7 cells were incubated in “serum starvation” medium (5% charcoal-stripped FBS) for 48 hours prior to treatment with vehicle control (DMSO) or the indicated concentrations of E2 for 6 hours (A); E2 alone or in combination with 4-OHT for 6 hours (B); or with E2, DHT, 3β-adiol, 4-OHT, or bicalutamide, alone or together, as indicated, for 24 hours (C). NAT1 and PGR mRNA expression were measured by quantitative real-time RT-PCR using 18S as a reference gene. Values of NAT1 and PGR expression were normalized to vehicle (DMSO)-treated cells and are the mean ± S.E.M. for three separate experiments. *P < 0.05 vs. DMSO control. #P < 0.05 vs. the same treatment without antagonist.
Because NAT1 was associated with AR expression in a subset of ERα (−)/PGR(−) human breast tumors (Doane et al., 2006), we determined if dihydrotestosterone (DHT) increased NAT1 expression in AR(+) MCF-7 cells (Fig. 5C). For these experiments, cells were treated for 24 hours. DHT did not increase NAT1 but did increase PGR expression, although the increase was small compared with E2 and may be attributable to metabolism of DHT to E2 in MCF-7 cells (Maggiolini et al., 2001). In postmenopausal women, DHT can be metabolized to 5α-androstane-3β , 17β-diol (3β-adiol) which binds and activates ERα and ERβ (Maggiolini et al., 1999; Weihua et al., 2002; Sikora et al., 2009; Sharma et al., 2012). 3β-Adiol did not stimulate NAT1 expression, but significantly increased PGR expression in an ER-dependent manner, since the stimulation was inhibited by concomitant treatment of the cells with 4-OHT but not the AR antagonist bicalutamide. Together, these data indicate that NAT1 expression is not regulated by either estrogens or DHT in MCF-7 cells.
Similarly, in ERα (+) T47D breast cancer cell line, PGR expression was significantly (p = 0.0002) increased by E2 in a concentration-dependent manner (Fig. 6A) and this activation was inhibited by selective ER modulator 4-OHT and the pure ER antagonist ICI 182,780 (fulvestrant) (Fig. 6B). We also tested the effect of E2 and 4-OHT in ERα (−)/PGR(−) MDA-MB-231 triple-negative breast cancer cells and observed no effect of either treatment on NAT1 or PGR expression in these cells (data not shown). Taken together, our results indicate that E2 does not stimulate NAT1 mRNA expression in either ERα (+) breast cancer cell line or in MDA-MB-231 triple-negative breast cancer cells.
NAT1 and PGR expression in T47D cells. T47D cells were incubated in “serum starvation” medium (5% charcoal-stripped FBS) for 48 hours prior to treatment with vehicle control (DMSO) or the indicated concentrations of E2 for 6 hours. NAT1 expression increased significantly (P = 0.0002) with respect to E2 concentration (A); E2 alone or in combination with 4-OHT or ICI 182,780 (ICI) for 6 hours (B). NAT1 and PGR mRNA expression were measured by quantitative real-time RT-PCR using 18S as a reference gene. Values of NAT1 and PGR expression were normalized to vehicle (DMSO)-treated cells and are the mean ± S.E.M. for three separate experiments. *P < 0.05 vs. DMSO control. #P < 0.05 vs. the same treatment without antagonist.
XBP1 Expression in Breast Tumors and NAT1 Expression Regulation by DTT.
The alternatively spliced transcription factor XBP1 was more highly expressed in ERα (+) than ERα (−) breast tumors and was reported to be associated with NAT1 expression (Finlin et al., 2001). We examined XBP1 expression and evaluated the correlation between XBP1 and NAT1 expression in the breast tumor samples. The primer and probe used for quantitative real time RT-PCR measurements was designed across exon 1 and 2 and not in the exon 4 and 5 region, in which 26 bp was removed to become the spliced and activated XBP1 (Lemin et al., 2007). Accordingly, both unspliced and spliced XBP1 mRNA expression were measured in this quantitative RT-PCR assay. In agreement with the previous report (Finlin et al., 2001), we confirmed in a separate set of breast tumors that XBP1 mRNA was significantly (p = 0.0026) higher in ERα (+) than ERα (−) breast tumors (Fig. 7).
XBP1 mRNA expression in breast tumor samples. XBP1 mRNA expression was measured using quantitative real-time RT-PCR in 20 breast tumors with the indicated ERα expression status. The tumor/breast sample numbers correspond to those in Figs. 1–4. XBP1 mRNA expression in MCF-7 was used as a reference control. Inserted graph: Each bar represents the mean ± S.E.M. of XBP1 mRNA in ERα (+) (n = 16) and ERα (−) (n = 5) breast tumor samples.
DTT activates the uncoupled protein response (UPR) and stimulates XBP1 splicing (Lemin et al., 2007). We used agarose gel electrophoresis to separate the XBP1 spliced and unspliced products. DTT increased the amount of activated spliced XBP1 in ERα (+) MCF-7 and T47D cells (Fig. 8). Despite the correlation of ESR1, XBP1, and NAT1 expression in ERα (+) breast tumors (Doane et al., 2006), no significant (P > 0.05) effect of DTT on NAT1 or ESR1 (data not shown) mRNA expression was observed (Fig. 8).
DTT stimulates XBP1 splicing in ERα (+) breast cancer cell lines T47D and MCF-7. The cells were treated with 5 mM DTT for 4 hours. Products from the RT-PCR reaction were separated by 1% agarose gel electrophoresis and the bands visualized by ethidium bromide staining. The unspliced XBP1 PCR product is 125 bp and the spliced XBP1 product is 99 bp.
ESR1, FOXA1, GATA3, NAT1, and XBP1 RNA Expression in ERα (+) and ERα(−) Breast Tumor Samples.
ESR1, FOXA1, GATA3, NAT1, and XBP1 mRNA expression was significantly higher (P < 0.0001 for all) in ERα (+) primary solid breast tumor samples (n = 806) compared with ERα (−) primary solid breast tumor samples (n = 237) (Fig. 9).
RNA expression of ESR1, FOXA1, GATA3, NAT1, and XBP1 was increased in estrogen receptor positive primary solid breast tumor samples. Publically available data from The Cancer Genome Atlas for the breast invasive carcinoma cohort was downloaded and analyzed (N = 1043). Samples were stratified by estrogen receptor status and differences in RNA expression were evaluated by Mann-Whitney for each gene. In the boxplots the solid black line represents the median, the upper hinge represents the 75th quartile, the lower hinge represents the 25th quartile, the upper whisker represents the largest observation less than or equal to upper hinge + 1.5 * interquartile range (IQR), the lower whisker represents the smallest observation greater than or equal to the lower hinge – 1.5 * IQR, and outliers are plotted as open circles. White boxplots show ER positive samples (n = 237) and gray boxplots show ER negative samples (n = 806). N = negative, P = positive, ***P < 0.001.
Discussion
We observed that NAT1 expression is higher in ERα (+) breast tumor tissues than in the associated normal breast tissue and that NAT1 is not highly expressed in ERα (−) breast tumors, consistent with previous studies (Dolled-Filhart et al., 2006; Ring et al., 2006; Endo et al., 2013). We also observed that NAT1 mRNA correlates positively with NAT1 enzyme activity in ERα (+) breast tumor tissues. Although patients nos. 14, 15, and 16 were defined as ERα (+), on the basis of immunohistochemistry, the ESR1 mRNA levels in these tumors was less than the ERα (+) MCF-7 breast cancer cell line, suggesting the possibility of RNA degradation. Additional explanations for this lack of correlation between ESR1 mRNA and protein expression in these three tumor samples could be owing to protein stability in the presence of ligand(s) that stabilize ERα protein, ERα phosphorylation, or false positive information on ER status (Lupien et al., 2007). Interestingly, we note that NAT1 mRNA level and enzyme activity were also lower in these patients’ breast tumors compared with other tumors classified as ERα (+), instead resembling the ERα (−) breast tumors examined. These data support the suggestion that NAT1 is expressed in association with ESR1 in breast tumors.
Alternative promoter usage is one of the mechanisms by which gene expression is differentially regulated in tissues (Chen et al., 2005). NATb is active in all tissues tested, whereas NATa is active in certain tissues, including kidney, lung, trachea, and liver, that have the highest exposure potential to environmental chemicals (Barker et al., 2006). The NATa promoter is activated in the ERα (+) ZR-75-1 breast cancer cell line, which has relatively high NAT1 activity and mRNA levels (Yan et al., 2007). In our study, we clearly demonstrated that the NATb, not the NATa, promoter was active and responsible for the high NAT1 expression in ERα (+) breast tumors. Thus, we conclude that alternative promoter usage was not responsible for NAT1 overexpression in breast tumors.
Despite the strong correlation between ESR1 and NAT1 expression in breast tumors, our studies indicated that there was no direct effect of E2 or 4-OHT on NAT1 expression in either MCF-7 or T47D luminal breast cancer cells in vitro. Previous studies reported that NAT1 expression was increased ∼2-fold by 10 nM E2 and significantly reduced by 6 μM tamoxifen in MCF-7 cells (Finlin et al., 2001). In this study, NAT1 gene expression was measured after 48 hours of tamoxifen treatment; thus, NAT1 expression might have been downregulated by some cross-signaling pathway in addition to, or instead of, a direct effect on ERα transactivation of NAT1 expression, because transcriptional regulation usually occurs within 24 hours (Heintz et al., 1983). Although the synthetic androgen R1881 increased NAT1 expression by an unknown secondary mechanism in AR (+) 22Rv1 and LNCaP prostate cancer cells (Butcher et al., 2007; Butcher and Minchin, 2010), we observed no effect of DHT on NAT1 expression in MCF-7 cells. The difference in androgen regulation of NAT1 expression between the breast and prostate cancer cell lines may be attributed to tissue-specific differences in additional transcription factors and coregulators needed to induce NAT1 expression or to differences in chromatin architecture and epigenetic signaling, including micro-RNA and long noncoding RNA expression.
Likewise to NAT1, GATA3 is another gene highly expressed in breast tumors that is positively correlated with ERα expression (Licata et al., 2010). Like NAT1, GATA3 is not directly regulated by E2 or ERα, indicating that a set of genes expressed at high levels along with ESR1, including GATA3 and NAT1, might be involved in a complex regulatory pathway during breast cancer development (Hoch et al., 1999). Overexpression of the luminal/ESR1(+) cluster genes GATA3 and NAT1 is characteristic and specific for breast cancer, because there is no similar pattern of high luminal/ESR1(+) gene expression in other cancer types. Theodorou et al. (2013) show that ERα-chromatin binding in MCF-7 cells is mediated by GATA3 and FOXA1 binding to the cis-regulatory elements that drive transcription of the ESR1 target genes prior to ERα binding; thus acting as pioneer factors. However, in the absence of GATA3, chromatin accessibility at potential ERα -binding elements may be altered, resulting in a rewired ERα-binding profile and expression program. Their data suggests GATA3 is one of the central components of the ERα complex that determines the binding potential and transcriptional targets in breast cancer cells. Thus, the correlation observed between ESR1 and GATA3 is not because the ESR1 product (ERα) interacts with GATA3 to induce its expression, but rather because GATA3 interacts with ERα DNA binding sites to shape enhancer accessibility that leads to downstream effects on other genes not influenced by ERα.
XBP1 is a transcription factor associated with endoplasmic reticulum stress and the unfolded protein response when cells are exposed to hypoxia, low glucose, or low pH that is upregulated in endocrine-resistant breast cancer (Koong et al., 2006). Tumor cells are often hypoxic and it is reasonable to hypothesize that XBP1 activation is involved in the regulation of genes overexpressed in tumors compared with normal tissues (Koong et al., 2006). XBP1 binds to and activates ERα in a ligand-independent manner to initiate downstream gene expression, such as the antiapoptotic gene BCL2 and several other genes associated with the control of breast cancer cell cycle and apoptosis (Gomez et al., 2007). XBP1 as an estrogen-responsive gene was clearly induced ∼6- to 8-fold by E2 and downregulated by tamoxifen in MCF-7 cells (Finlin et al., 2001; Wang et al., 2004). Here we validated previous reports that XBP1 was more highly expressed in ERα (+) breast tumors (Bertucci et al., 2005; Tozlu et al., 2006; Lien et al., 2007; Guedj et al., 2012). Also, XBP1 can bind and activate ERα to initiate transcription of gene expression in a ligand-independent manner in breast cancer cells (Ding et al., 2003). Here we demonstrated that DTT induced the unfolded protein response in MCF-7 cells, as evidenced by increased amounts of spliced (active) XBP1, but XBP1 activation did not increase NAT1 expression in MCF-7 cells, despite XBP1 binding sites in the upstream NATb promoter (Butcher et al., 2007). We conclude that XBP1 does not regulate NAT1 expression in MCF-7 cells.
Evidence has shown that breast tumors with high expression of genes, including NAT1, probably arise from the luminal epithelium in the mammary gland duct (Dressman et al., 2001; Usary et al., 2004; Farmer et al., 2005; Badve et al., 2007). ESR1, NAT1, GATA3, and FOXA1 are primarily expressed in luminal epithelial cells in the lumen of the duct and not in the myoepithelial cells that compose the outer layer (Tong and Hotamisligil, 2007). Interestingly, miR-1290 was identified to be low in ERα (+) breast tumors and demonstrated to decrease the expression of FOXA1 and NAT1, suggesting a mechanism of NAT1 upregulation in ERα (+) breast cancer (Endo et al., 2013). Subsequent studies showed miR-1290 targets sites on the NAT1 3′-UTR (Endo et al., 2014). Targeting these genes might interrupt the tumor proliferation pathway and offers guidance for target screening and development of novel breast cancer drugs (Endo et al., 2011; Sim et al., 2014). However, there are no reports of E2 regulation of miR-1290 expression (Klinge, 2012; Muluhngwi and Klinge, 2015, 2017). Although miR-1290 is not included in the BreastMark miRNA survival analysis database (http://glados.ucd.ie/BreastMark/index.html), analysis of all breast tumors in the Kaplan-Meier Plotter showed high miR-1290 was associated with lower overall survival (http://kmplot.com/analysis/index.php?p=service&cancer=breast_mirna). Since miR-1290 reduced NAT1 expression (Endo et al., 2013), this observation is consistent with a positive association of ESR1 and NAT1 in breast tumors.
In summary, NAT1 mRNA transcript levels, protein, and enzyme activity were all higher in human breast tumors with high expression of ERα/ESR1 compared with normal breast tissue. High NAT1 expression in tumor cells was not attributable to adipocytes as evidenced by low PLIN mRNA expression in breast tumor samples. E2, DHT, and 3β-adiol did not increase NAT1 transcript levels in ERα (+) breast cancer cells. DTT increased levels of the activated spliced XBP1 level in these cells, but did not increase NAT1 expression. SP1 binds the NATb promoter and is responsible for basal NAT1 transcription (Sadrieh et al., 1996; Windmill et al., 2000). Interestingly, E2 increases SP1 transcription in MCF-7 breast cancer cells [Nuclear Receptor Signaling Atlas (NURSA) Transcriptomine, www.nursa.org] and in U-2 OS osteosarcoma cells (Hu et al., 2017). Thus, we speculate that E2-ERα upregulation of SP1 expression may indirectly increase NAT1 transcription, although further studies are needed to explore this suggestion. We have assembled the data presented and reviewed here in a model (Fig. 10). Overall, the lack of direct regulation of NAT1 expression by E2 or DTT despite high ESR1 and XBP1 expression suggest that ESR1, XBP1, and NAT1 expression share a common transcriptional network in breast cancer that remains to be fully elucidated. Clusters of high-expression genes, including NAT1, in breast tumors might serve as potential targets for development of novel drugs and therapeutic interventions for breast cancer patients.
Model of NAT1 transcriptional regulation and correlation with ESR1 (ERα), XBP1, FOXA1, and GATA3 expression in ERα+ breast tumors and in MCF-7 cells. Known regulators of NAT1 transcription acting through the NATb major promoter are SP1 and glucocorticoid receptor (GR). The AR-mediated increase in heat shock factor 1 (HSF1) was reported to be responsible for androgen-induced NAT1 transcription in 22Rv1 and LNCaP prostate cancer cells (Butcher et al., 2007). Although miR-1290 was demonstrated to target NAT1 and decrease NAT1 transcript and protein levels, whether E2-ERα inhibits miR-1290 expression in breast cancer cells or tumors is not yet known.
Acknowledgments
The authors thank Katherine Bourcy, David Barker, and Amar Singh for advice and technical support.
Authorship Contributions
Participated in research design: Zhang, States, Klinge.
Conducted experiments: Zhang, Carlisle, Doll.
Contributed new reagents or analytic tools: Zhang, Carlisle, Martin, Klinge.
Performed data analysis: Zhang, Carlisle, Klinge.
Wrote or contributed to the writing of the manuscript: Zhang, Carlisle, Klinge, States, Hein.
Footnotes
- Received December 7, 2017.
- Accepted January 12, 2018.
↵1 Current affiliation: Sanofi US, Bridgewater, New Jersey.
Supported by National Institutes of Health United States Public Health Service Grants [R01-CA-034627, R01-DK-053220, and T32-ES-011564].
Abbreviations
- 3β-adiol
- 5α-androstane-3β, 17β-diol
- AR
- androgen receptor
- DHT
- dihydrotestosterone
- DTT
- dithiothreitol
- E2
- estradiol
- ERα
- estrogen receptor α
- FBS
- fetal bovine serum
- NAT1
- N-acetyltransferase 1
- 4-OHT
- 4-hydroxytamoxifen
- PGR
- progesterone receptor
- PLIN
- perilipin
- RISC
- RNA-induced silencing complex
- RSEM
- RNA-Seq by expectation maximization
- XBP1
- X-box binding protein 1
- Copyright © 2018 by The American Society for Pharmacology and Experimental Therapeutics
References
In this issue
NAT1 Expression in Breast Cancer Cells
