Background
Breast cancer (BC) is the leading cause of cancer death in women worldwide. Chemotherapy is an important adjuvant for breast cancer treatment, but resistance is a major obstacle for chemotherapy in some patients. Combination treatment targeting molecules that contribute to chemoresistance is an important approach to overcome resistance and improve the efficacy of chemotherapy.
Nicotinamide
N-methyltransferase (NNMT), a phase II metabolizing enzyme, mainly catalyzes the methylation of nicotinamide into 1-methylnicotinamide (MNA) and other pyridines into pyridinium ions [
1], and it is involved in the biotransformation of many drugs and xenobiotics [
2]. In 1984, Seifert R was the first to confirm that alterations in NNMT activity are involved in the development and progression of carcinoma in vivo [
3]. A large number of subsequent studies demonstrated that NNMT is aberrantly expressed and associated with a poor prognosis in various cancers, such as colorectal cancer [
4], gastric cancer [
5,
6], hepatocellular carcinoma [
7], and lung cancer [
8]. Other studies have shown that NNMT affects the proliferative, migratory, invasive, and differentiation profiles of various cancers [
9‐
11]. Furthermore, NNMT overexpression has recently been found to be associated with chemotherapy resistance. After analyzing the correlation between the cancer-related genes and 99 anti-tumor drugs with known molecular mechanisms, Hsu et al. found that the NNMT expression level might be related to the sensitivity to chemotherapeutic drugs [
12]. Yu et al. reported that NNMT knockdown PANC-1 cells were much less resistant to rapamycin as well as glycolytic inhibitor 2-deoxyglucose, whereas NNMT-overexpressing cells showed the opposite effects [
11]. We also have previously reported that NNMT overexpression inhibits the activation of ASK1-p38 pathway via MNA production, which results in a decrease in the apoptosis induced by 5-fluorouracil (5-FU) to enhance resistance in colorectal cancer cells [
13]. These reports suggested that NNMT might be involved in the resistance to chemotherapy and thus serve as a potential target for combination therapy. Therefore, we investigated the role of NNMT in breast cancer chemotherapy, which might be beneficial for improving chemotherapeutic efficacy in breast cancer.
At the beginning of this study, we found that NNMT was upregulated in breast carcinomas of patients who were undergoing mastectomy by immunohistochemistry on tissue microarray (p < 0.001). After correlation with the clinicopathological characteristics of 82 patients with their chemotherapy efficacy record, NNMT overexpression was found to be associated with a shorter survival and reduced chemotherapy efficacy (p < 0.05). We then confirmed that NNMT overexpression significantly enhanced resistance to adriamycin (ADM) and paclitaxel (PTX) in BCs. Furthermore, we demonstrated that NNMT overexpression attenuated the apoptosis that was induced by ADM and PTX to enhance the resistance through SIRT1 protein stabilization in BCs.
Methods
Drugs and antibodies
Adriamycin (ADM) and 1-methylnicotinamide (MNA) were obtained from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO, USA), and paclitaxel (PTX) and the selectively SIRT1 inhibitor EX527 were obtained from Selleck Chemicals (SelleckChemicals, Houston, TX, USA). The anti-SIRT1, anti-β-actin, and anti-acetyl-p53 were all obtained from Cell Signaling Technology (CST, Beverly, MA, USA). The mouse anti-NNMT monoclonal antibody 1E7 was prepared through the hybridoma technique as previously described [
14].
Cell line
Human SK-BR-3 and MCF7 cell lines, which have low NNMT expression, and the MDA-MB-231 cell lines, which have high NNMT expression, were obtained from Cell Bank at the Chinese Academy of Sciences (Shanghai, China) and cultured in DMEM (Gibco, Grand Island, NY, USA). The authenticity of the three cell lines was verified using STR. All media were supplemented with 10% fetal bovine serum (Gibco, Long Island, NY, USA), 100 U/ml penicillin (Sigma-Aldrich, St. Louis, MO, USA), and 100 mg/ml streptomycin (Sigma-Aldrich, St. Louis, MO, USA), and the cells were maintained at 37 °C in a humidified 5% CO2 incubator.
Human tissue specimens and patient clinical information
This study was approved by the Human Research Ethics Committee of Sir Run Run Shaw Hospital (Hangzhou, China). As the initial treatment, total 165 treatment-naive patients with breast cancer received mastectomy from Oct 1, 2000, to Dec 31, 2006, 82 of whom had a chemotherapy efficacy record at Sir Run Run Shaw Hospital (Hangzhou, China) and were therefore included in this study. The diagnoses of breast cancer were confirmed by postoperative pathological results.
The clinical characteristics of 165 cancer patients were extracted from their medical record, including age, gender, tumor diameter, TNM stage, ER, HER-2, PR, Ki-67, and chemotherapy response judged by the revised RECIST guideline (version 1.1). Eighty-two patients with breast cancer received the chemotherapy mainly using CMF (cyclophosphamide + methotrexate + fluorouracil) and FEC-P (fluorouracil + epirubicin + cyclophosphamide + paclitaxel) regimens, which accounted for more than 90% of the patients. These patients were followed up, and the OS were calculated from the date of surgical treatment to the date of death or last follow-up.
Immunohistochemistry analysis (IHC) on paraffin-embedded tissue array
The tissue microarray block was cut into 4-μm sections, and immunohistochemical staining was performed. Briefly, the sections were first deparaffinized and hydrated. After antigen retrieval with 0.01 M citrate buffer (pH 6.0) and microwave heat induction, the sections were treated with 3% hydrogen peroxide for 10 min. NNMT were detected using mouse monoclonal anti-human NNMT antibody (dilution 1:10000). After secondary antibody staining, diaminobenzidine was used as the chromogen for 3 min, and then, the nuclei were counterstained with hematoxylin. Two pathologists without prior knowledge of the clinicopathological data evaluated the staining results independently.
The expression of NNMT was scored according to the intensity and percentage of positive cells. The staining intensity was scored as 0 (no staining), 1+ (weak staining), 2+ (moderate staining), or 3+ (intense staining). Then, the percentage of positive cells and the respective intensity scores were used to determine the final staining score. Therefore, the staining score had a minimum value of 0 and a maximum value of 300. A cutoff value of 120 was found to be statistically significant using the X-tile software program (
https://medicine.yale.edu/lab/rimm/research/software.aspx) [
15], which means a score from 0 to 119 is considered low expression (NNMT
l) but from 120 to 300 is high expression (NNMT
h).
NNMT plasmid transfection and stable cell strain selection
The pcDNA3.1/NNMT and pcDNA3.1/Vector plasmids have been successfully constructed and are described in our previous paper [
14]. SK-BR-3 and MCF7 cells were transfected with pcDNA3.1/NNMT or pcDNA3.1/Vector using Lipofectamine™ 3000, and then, the cells were grown in complete medium containing 800 mg/L geneticin (G418; Gibco, Grand Island, NY, USA) for 2 weeks. Single colonies were picked and placed in 96-well plates to proliferate separately, and they were evaluated for NNMT expression by real-time quantitative RT-PCR and Western blotting. SK-BR-3/NNMT-1, SK-BR-3/NNMT-2, and MCF7/NNMT-1, MCF7/NNMT-2 with stable NNMT overexpression, and SK-BR-3/Vector and MCF7/Vector controls were selected for further analysis.
Lentiviral NNMT shRNA infection into MDA-MB-231 cells
Lentiviral NNMT shRNA construction and infection of MDA-MB-231 cells were conducted as previously described [
16]. Briefly, MDA-MB-231 cells were seeded (3 × 10
5 cells/well) in six-well plates and incubated for 24 h. When the cells reached 30–50% confluence, lentivirus containing shRNAs (NNMT shRNA 1#, NNMT shRNA 2#, or shRNA NC; MOI = 10 for MDA-MB-231) was added. Ten hours after coculturing with lentivirus, the supernatant was replaced with fresh medium. Forty-eight hours after infection, the transduced cells were sorted using a BD FACS Aria II System (BD Biosciences, San Jose, CA, USA) to obtain the GFP-positive cell populations, and these populations were then subjected to functional assays. Cells infected with shRNA NC were used as the negative control.
CCK-8 assay to determine IC50
The CCK-8 assay was used to explore the IC50 of ADM and PTX in SK-BR-3, MCF7, and MDA-MB-231 cell models. One hundred microliters of cells (density 3 × 104/mL) was seeded into each well of a 96-well plate. After 24 h, 100 μL fresh medium containing different concentrations of ADM (0, 0.125, 0.25, 0.5, 1, 2, 4, 8, 16, and 32 μM) or PTX (0, 0.625, 1.25, 2.5, 5, 10, 20, 40, 80, and 160 nM) were added into each well. After another 48-h incubation with ADM or PTX, 10 μL CCK-8 (CCK-8, Dojindo Laboratories, Japan) solution was added to each well, and the cells were incubated for an additional 2 h at 37 °C. Finally, the absorbance value was read at 450 nm using an ELISA plate reader instrument (Bio-Rad, Model 680, Japan). The cells in the wells treated only with ddH2O served as the control group for each cell model treated with ADM. The cells in the wells treated only with DMSO served as the control group for each cell model treated with PTX. The inhibition rate (IR) was calculated by the following equation: [1 − (mean absorbance of drug wells/mean absorbance of control wells)] × 100%. ADM and PTX resistance was evaluated by calculating the IC50, which was determined as the concentration of the drug required when the IR was 50%.
Apoptosis analysis
Apoptosis was detected by flow cytometric analysis using a FITC-Annexin V/7-AAD Apoptosis Detection Kit (BD, CA, USA). Briefly, cells (1 × 105 cells/well) were seeded in a 12-well plate. After culturing for 48 h, the treated cells were harvested, incubated with FITC-Annexin V and 7-AAD for 30 min at room temperature in the dark, and immediately analyzed by flow cytometry (FACSCalibur flow cytometer, BD, CA, USA). Each experiment was conducted at least three times.
Western blot analysis
RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China) was used to extract cell proteins. A BCA Protein Assay Kit (Beyotime Biotechnology, Shanghai, China) was used to measure the protein concentrations. A 40-ug protein sample was subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to an Immobilon P Transfer Membrane (Millipore, Bedford, MA, USA). After regular blocking and washing, the membranes were incubated with primary antibodies overnight at 4 °C followed by incubation with HRP-conjugated secondary antibodies for 1 h at room temperature. Signals were visualized using enhanced chemiluminescence detection reagents (FD Bioscience, Hangzhou, China) and imaged using an Image Lab (BIO-RAD, Hercules, CA, USA). All the experiments were independent and were conducted at least three times. Protein quantification of the Western blotting results was achieved by densitometry using ImageJ software, normalization to β-actin, and then comparison to the control group, which was normalized as 1.
siRNA transfection
SIRT1 siRNAs were obtained from RiboBio (RiboBio Co, Guangzhou, China), dissolved in 20 μM stock solution with distilled water and stored at − 80 °C. According to the protocol, 2 × 105 cells were plated in 6-well plates. When the cell density reached 30%, the culture medium was replaced with fresh medium containing 30 nM SIRT1-specific siRNA in transfection reagent (RiboBio, Guangzhou, China) and cultured for another 72 h. The control siRNA contained a scrambled sequence that would not lead to the specific degradation of any known cellular mRNA.
RNA isolation and real-time quantitative RT-PCR
Real-time quantitative RT-PCR analysis was conducted using the SYBR Premix EX Taq™ RealTime PCR Detection System (TaKaRa Biotechnology, Dalian, China). According to the protocol, total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse-transcribed into cDNAs with the M-MLV Reverse Transcriptase kit (Promega, Madison, WI, USA). The sequences of the PCR primers that were used are listed in a previous paper [
16]. The experiments were run in an initial denaturation step of 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 34 s using an ABI PRISM 7500 Fast Real-Time PCR System. All experiments were independent and conducted at least three times. The results were calculated using the 2
-ΔΔCt method. The data were normalized to GAPDH and then compared to the control group, which was normalized to 1.
Cells were plated (SK-BR-3, 1000 cells/well; MCF7, 400 cells/well; MDA-MB-231, 1500 cells/well) in 60-mm dishes and treated with ADM or PTX or ddH2O or DMSO (ddH2O as a control for ADM and DMSO as a control for PTX) for 12 days. The cells were fixed with methanol and stained with Giemsa (Sigma, St. Louis, MO, USA) for 30 min. Colonies (foci > 100 μm) were counted, and the data were normalized to each control group that was treated only with ddH2O or DMSO, which was normalized to 100%. Experiments were repeated at least three times.
SIRT1 activity assay
The intracellular SIRT1 activities were measured using a SIRT1 deacetylase fluorometric reagent kit (SIRT1 Deacetylase Fluorometric Assay Kit, CycLex Co., Ltd., Japan). Briefly, 1 × 107 cells were harvested and resuspended in 1 mL of lysate buffer. The supernatant was discarded after 13,000g for 10 min at 4 °C. Then, nuclear protein was collected by centrifugation at 20,000g for 10 min after sonication. After the protein concentration was determined using the BCA Protein Assay Kit (Beyotime Biotech, Shanghai, China), the activity of SIRT1 in the nuclear protein fraction was measured according to the manufacturer’s instructions. Each experiment was conducted at least three times.
Detection and quantification of MNA by HPLC-UV
The HPLC-UV method for the separation and detection of MNA has been described in our previous paper [
17]. Briefly, HPLC-UV was performed using a Hewlett-Packard 1100 photodiode array detector (Waldbronn, Germany) incorporating a 250 × 4.6-mm-inner-diameter Agilent TC-C18 5-μm reversed-phase column. After the injection of 100 μL of cell supernatant, MNA was monitored by the absorbance at 265 nm. The level of MNA was calculated based on the calibration curve.
Statistical analysis
Statistical analysis was conducted using the SPSS 20.0 statistical software package (SPSS Inc., Chicago, IL). The Student’s test was used to determine the statistical significance of differences between comparison groups in vitro. Error bars represent the mean ± SEM. The relationships between NNMT expression and clinicopathological attributes were analyzed using Pearson’s χ2 test. Survival rates were calculated using the Kaplan-Meier method and compared using the log-rank test, and p < 0.05 was considered statistically significant.
Discussion
In the USA, breast cancer was recently reported as the most common cancer and the second most common cause of death among cancers in woman [
18]. According to the data reported by China’s National Cancer Registry in 2015, the incidence and mortality of breast cancer were both the highest among cancers in women and are still increasing in China [
19]. Recurrence after treatment failure caused by chemoresistance has introduced a great dilemma in breast cancer therapy. Therefore, the search for indicators to predict the efficacy of chemotherapy can improve the prognosis of breast cancer patients. At present, several biomarkers in tumor tissues have been used to predict the efficacy of chemotherapy drugs in breast cancer research and in the clinic. Recent studies have found that BRCA1/2, a breast cancer susceptibility gene, is closely related to triple-negative breast cancer (TNBC) and can repair DNA damage through homologous recombination, and its mutation can be used to predict the efficacy of PTX in TNBC chemotherapy [
20]. Current treatments for BRCA1/2 mutations have shown that treatment with DNA repair-related ribose polymerase (PARP-1) inhibitors or chemotherapy with platinum-based drugs can improve the efficacy and prognosis of patients [
21,
22]. A study has been conducted to detect miRNAs in tumor tissues and screened out a combination of four miRNAs (miR-30a, miR-9-3p, miR-770, and miR-143-5p) that can predict the efficacy of neoadjuvant chemotherapy in TNBC [
13]. However, more reliable biomarkers are still needed to predict the efficacy of chemotherapy drugs for all molecular subtype in breast cancer therapy.
Growing evidence shows that NNMT is aberrantly expressed in several cancers and is a promising prognostic predictor in some of cancers, such as pancreatic cancer and gastric carcinoma [
6,
23]. After evaluating NNMT expression and its clinical relevance in breast cancer, we found that NNMT expression was significantly higher in breast carcinoma than in paracancerous tissues and breast hyperplasias, which suggests that NNMT is also aberrantly expressed in breast cancer and might be a potential diagnostic biomarker for breast cancer. Furthermore, we found that NNMT overexpression was associated with a shorter survival and reduced chemotherapy efficacy in 82 patients who had a chemotherapy efficacy record. Classification of molecular subtypes in breast cancer is useful in the prediction of therapeutic response and prognosis. We also found that there was a significant difference in the efficacy of chemotherapy among four different molecular subtypes. Meanwhile, we analyzed the interaction of NNMT and molecular subtype on chemotherapy efficacy. In luminal B (42 samples) subtype, the patients with NNMT overexpression had a lower chemotherapy efficacy (
p = 0.01), while there was no significant difference in luminal A (20 samples), ERBB2 (13 samples), and basal-like (6 samples) subtypes (Additional file
4). For different molecular subtypes, patients received other different treatments along with chemotherapy, such as trastuzumab treatment for ERBB2 patients, which might also affect the efficacy of chemotherapy. However, the small sample size of this study may lead to potential limitations, and we still need to expand the sample size to verify this result.
We then investigated the effects of NNMT on the chemoresistance in breast cancer cells. Considering that breast cancer cell lines have different molecular phenotypes that might impact chemoresistance, we selected cell lines with different molecular phenotypes to study the effect of NNMT on chemotherapy in breast cancer cells. Therefore, the SK-BR-3 (ER-, Her2+) and MCF7 (ER+, Her-) cell lines, which both have low NNMT expression, and the MDA-MB-231 (ER-, Her2-) cell line, which has high NNMT expression, were selected for study. We choose ADM and PTX as our chemical drugs, because they were the most important chemotherapeutic drugs in the patient’s chemotherapy regimens.
NNMT methylates nicotinamide (NAM) to MNA using the universal methyl donor
S-adenosyl methionine (SAM) to produce
S-adenosyl homocysteine (SAH). NNMT has been reported as a metabolic regulator in adipocytes through global changes in histone methylation and increased NAD+ content [
24], which acts as a redox cofactor for more than 200 enzymatic reactions and serves as a cosubstrate for the sirtuins, which constitute a family of NAD+-dependent deacetylases. In addition, NAM has been reported as a reversible inhibitor of the sirtuins (namely, SIRT1–7). These reports indicated that NNMT may have effects on sirtuins. SIRT1, which is the most important sirtuin, was originally identified as a longevity gene. Recently, the oncogenic function of SIRT1 has also been reported in cancer, including colon and prostate cancer [
25,
26]. The findings of these studies suggested that once cancer cells acquire the ability to produce SIRT1, the presumed function of SIRT1 may promote the survival of carcinoma cells. SIRT1 regulates various cellular functions, including DNA repair, cell survival, and metabolism, via the deacetylation of target proteins such as histone and p53. Deacetylation of p53 plays an important role in downregulating p53 transcriptional activity and promoting cell survival following a stress response [
27]. These phenomena indicate that the SIRT1-p53 pathway regulates the apoptosis of cancer cells. Therefore, we hypothesized that NNMT promotes the ADM and PTX resistance in breast cancer by increasing the SIRT1 stabilization and activity. To test our hypothesis, we examined the levels of SIRT1 protein and mRNA and its target acetyl-p53 after overexpression and downregulation of NNMT in BCs. Our result showed that NNMT and its product MNA were not significantly altered at the SIRT1 mRNA level, but both increased the SIRT1 protein and activity levels and decreased the acetyl-p53 level in BCs. Moreover, the higher NNMT protein level in SK-BR-3/NNMT-2 cells than that in SK-BR-3/NNMT-1 cells represented higher IC50 value of ADM and PTX, higher SIRT1 protein and activity level, and lower acetyl-p53 protein level. These results suggested that NNMT could increase the cellular SIRT1 activity level in BCs through SIRT1 protein stabilization. Asaka et al. reported that SIRT1 overexpression enhanced resistance for cisplatin and paclitaxel in HHUA cells and the resistance was canceled by EX527 [
28]. To further verify our hypothesis, we utilized EX-527 and SIRT1-specific siRNA to inhibit the cellular SIRT1 activity in BC cells. SIRT1-specific siRNA showed a better inhibition efficiency of SIRT1 protein and activity than EX527. Consistent with the inhibition efficiency of SIRT1, the NNMT-related resistance in the cells treated with SIRT1-specific siRNA was reduced more than that in EX527-treated cells, which suggests that the effect of NNMT on ADM and PTX resistance was crippled by SIRT1 inhibition. Taken together, the result indicated that NNMT expression enhances the chemoresistance through SIRT1 stabilization and activity.
According to our study, NNMT has the potential to become a biomarker for diagnostic and chemotherapeutic efficacy predication in breast cancer. Moreover, NNMT could also play other extensive biological roles by regulating SIRT1. Hong et al. reported that increasing NNMT expression or MNA levels stabilizes SIRT1 protein to regulate hepatic nutrient metabolism [
29]. You et al. reported NNMT enhances the progression of prostate cancer by stabilizing SIRT1 [
30]. These results indicated that NNMT may be a potential therapeutic target not only in cancer but also in other diseases. However, the exact mechanism by which NNMT regulates the stability of SIRT1 protein requires further study.