Identification and characterization of potent, selective, and efficacious inhibitors of human arylamine N-acetyltransferase 1

Abstract

Arylamine N-acetyltransferase 1 (NAT1) plays a pivotal role in the metabolism of carcinogens and is a drug target for cancer prevention and/or treatment. A protein-ligand virtual screening of 2 million chemicals were ranked for predicted binding affinity towards the inhibition of human NAT1. Sixty of the five hundred top-ranked compounds were tested experimentally for inhibition of recombinant human NAT1 and N-acetyltransferase 2 (NAT2). The most promising compound 9,10-dihydro-9,10-dioxo-1,2-anthracenediyl diethyl ester (compound 10) was found to be a potent and selective NAT1 inhibitor with an in vitro IC₅₀ of 0.75 μM. Two structural analogs of this compound were selective but less potent for inhibition of NAT1 whereas a third structural analog 1,2-dihydroxyanthraquinone (a compound 10 hydrolysis product also known as Alizarin) showed comparable potency and efficacy for human NAT1 inhibition. Compound 10 inhibited N-acetylation of the arylamine carcinogen 4-aminobiphenyl (ABP) both in vitro and in DNA repair- deficient Chinese hamster ovary (CHO) cells in situ stably expressing human NAT1 and CYP1A1. Compound 10 and Alizarin effectively inhibited NAT1 in cryopreserved human hepatocytes whereas inhibition of NAT2 was not observed. Compound 10 caused concentration-dependent reductions in DNA adduct formation and DNA double strand breaks following metabolism of aromatic amine carcinogens beta-naphthylamine and/or ABP in CHO cells. Compound 10 inhibited proliferation and invasion in human breast cancer cells and showed selectivity towards tumorigenic versus non-tumorigenic cells. In conclusion, our study identifies potent, selective, and efficacious inhibitors of human NAT1. Alizarin’s ability to inhibit NAT1 could reduce breast cancer metastasis particularly to bone.

1. INTRODUCTION

Human arylamine N-acetyltransferase 1 (NAT1) and 2 (NAT2) differ in tissue distribution and substrate selectivity [1,2]. NAT1 is expressed in most human tissues whereas NAT2 expression is restricted primarily to the liver and gastrointestinal tract [3,4]. Prototype selective substrates are p-aminobenzoic acid (PABA) and sulfamethazine (SMZ) for NAT1 and NAT2, respectively [5]. NAT1*10 is the most common NAT1 allele or haplotype and is associated with elevated catalytic activity [6]. Individuals possessing NAT1*10 have been reported with increased risk for cancers of the urinary bladder [7–8], lung [9,10], colon/rectum [11,12], breast [13,14], prostate [15,16], and pancreas [17,18] as well as non-Hodgkin lymphoma [19,20]. Although there are also many published studies which did not report this association, particularly for urinary bladder cancer [21–23], different results reported in the literature [24] may be confounded by differences in risk factors, including the identity of carcinogens metabolized by NAT1 and their level of exposure.

Environmental and occupational arylamine carcinogens are ubiquitous and over 10% of all known or suspected human carcinogens are either an arylamine or metabolite [25]. 4-Aminobiphenyl (ABP) and beta-napthylamine (BNA) are listed as Group 1 human carcinogens [26]. Arylamine carcinogens require N-hydroxylation by cytochrome P450 followed by O-acetylation catalyzed by arylamine N-acetyltransferases such as NAT1 to form arylnitrenium ions which bind covalently to nucleophilic groups such as DNA. DNA adducts which are not repaired lead to mutagenesis and initiate cancer [1, 26].

Both human NAT1 and NAT2 catalyze the N- acetylation of ABP and BNA [5] and the O-acetylation of N-hydroxy-4-aminobiphenyl (N-OH-ABP) [27,28]. Using recombinant constructs expressing the major transcript of human NAT1, our laboratory compared NAT1*10 relative to NAT1*4 reference allele in DNA repair-deficient Chinese hamster ovary (CHO) cells subjected to transient or stable transfection of NAT1 and CYP1A1 [29]. ABP showed higher DNA adducts and mutants than in cells transfected with NAT1*10 compared to NAT1*4 consistent with increased risk for cancers associated with arylamine exposure in individuals possessing NAT1*10 referenced above.

Our laboratory has previously employed molecular modeling to explore the effects of genetic polymorphisms on NAT protein structure [30,31]. In this study, in silico screening of 2 million compounds was completed to assess virtual binding to the active site of NAT1 and NAT2. The top 500 ranked compounds for both were examined and sixty were tested experimentally for inhibition of recombinant human NAT1. A subset of the most potent and selective (for NAT1 versus NAT2) inhibitors were characterized for their inhibition of NAT1 and/or NAT2 in cryopreserved human hepatocytes, DNA repair-deficient CHO cells, and human breast cancer cell lines.

2. MATERIALS AND METHODS

2.1. In Silico Screening for Small Molecule Inhibitors

High resolution crystal structures of human NAT1 (PDBID: 2PQT) and NAT2 (PBDID: 2PFR) have previously been described [32]. Compounds that interacted with NAT1 and/or NAT2 were identified by in silico screening of a 2007 ZINC drug-like library containing 2 million compounds [33]. The computational screening program Surflex-Dock 2.3 [34] was used to perform protein-ligand docking studies. The top five hundred scored and ranked compounds were identified, based on their approximate active site binding affinity to NAT1 and/or NAT2. and one hundred and fifty compounds were selected for potential purchase.

2.2. Initial Testing of N-Acetyltransferase Inhibitors

Sixty of the highest-ranked compounds following the in silico screening described above were further tested for their ability to inhibit recombinant human NAT1 and NAT2 stably expressed in yeast (Schizosaccharomyces pombe) as previously described [35,36]. NAT1 and NAT2 reactions were conducted with 300 μM of the prototype arylamine substrates p-aminobenzoic acid (PABA) or sulfamethazine, (SMZ) respectively in the presence of 1 mM acetyl coenzyme A (AcCoA) and 0.1 mg/ml compound inhibitor or vehicle control. Substrate and N-acetylated product were separated and measured by high performance liquid chromatography (HPLC) as previously described [35,36].

2.3. Further validation of NAT1 inhibitors

Further validation was performed for 9,10-dihydro-9,10-dioxo-1,2-anthracenediyl diethyl ester (compound 10) and three structural analogs of compound 10 available commercially: 1-(9,10-dihydro-8-hydroxy-9,10-dioxo-1-anthracenyl)4-ethyl ester (analog 1), 9,10-dihydro-9,10-dioxo-1,8-anthracenediyl diethyl ester (analog 2) and 1,2-dihydroxyanthraquinone (an hydrolysis project of compound 10 also known as Alizarin; analog 3). Recombinant human NAT1 and NAT2 assays were conducted with PABA or SMZ concentrations of 10 μM, 100 μM AcCoA and inhibitor concentrations from 0–1000 μM.

2.4. Compound 10 inhibition of N-acetylation in Chinese hamster ovary cells.

UV5-Chinese hamster ovary (CHO) cells lack nucleotide excision repair due to a mutation in the XPD (ERCC2) gene [37] rendering them hypersensitive to bulky adduct mutagens. Nucleotide excision repair deficient- UV5-Chinese hamster ovary (CHO) cells (CRL-1865), previously transfected with human CYP1A1 and NAT1*4, were cultured as described [27–29]. To test for inhibition of NAT1 in situ, 1 × 10⁶ cells were seeded in 6-well plates and allowed to attach overnight and then incubated with 300 μM PABA and various concentrations (0–1000 μM) of Compound 10 for 24 hr. The culture medium was collected and N–acetyl-PABA was quantified by HPLC as described above. Additional experiments were conducted to further investigate the inhibitory properties of compound 10 in CHO cells expressing human NAT1 in situ. The CHO cells were plated in 6-well tissue culture plates 24 hr prior to treatment. On the day of treatment media was removed and replaced with α-MEM supplemented with 10 μM PABA or ABP and various concentrations of compound 10. Two hrs after the addition of substrate and inhibitor a small amount of media was removed from the cells and PABA or ABP N-acetylation measured by HPLC.

2.5. Compound 10 inhibition of N-acetylation in human breast cancer cells

MDA-MB-231 (HTB-26) and MCF-7 (HTB-22) human breast adenocarcinoma cells were obtained from American Type Culture Collection (ATCC). Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM), Lonza, Walkersville, MD) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 100 units/mL penicillin (Lonza), and 100 μg/mL streptomycin (Lonza). Non-tumorigenic human breast epithelial cells (MCF-10A) were a gift from Mrs. Sheila Thomas (University of Louisville). Cells were cultured in Mammary Epithelium Basal Medium (MEBM, Lonza, Walkersville, MD) supplemented with bovine pituitary extract (2 mL / 500 mL MEBM, Lonza), epidermal growth factor (0.5 mL / 500 mL MEBM, Lonza), insulin (0.5 mL / 500 mL MEBM, Lonza), transferrin (0.5 mL / 500 mL MEBM, Lonza), hydrocortisone (0.5 mL / 500 mL MEBM, Lonza), epinephrine (0.5 mL, 500 mL MEMB Lonza), and gentamicin (0.5 mL / 500 mL MEBM, Lonza). Cells were maintained at 37°C in 5% CO_2. MDA-MB-231 and MCF-7 cells were trypsinized with 0.25% Trypsin-EDTA (Gibco). MCF-10A cells were trypsinized with 1X TrypLE^™ Express (Gibco). To prepare cell lysates, cells were washed with cold PBS and resuspended in homogenization buffer (20 mM NaPO₄ at pH7.4, 1 mM EDTA, 2 μg/mL aprotinin, 0.1mM phenylmethanesulfonyl fluoride, 2 mM pepstatin A, 1 mM dithiothreitol, and 0.2% Triton X-100) on ice. Homogenates were centrifuged at 15,000 × g for 10 min at 4°C. Cells were grown to confluence and treated with media supplemented with 10 μM ABP or PABA with varying concentrations of Compound 10 (0 – 1000 μM) for 4hr at 37°C. The culture media were collected and 1/10 volume of 1M acetic acid was added. The mixture was centrifuged at 13 000 g for 10 min. The supernatant was analyzed for N–acetyl-PABA or ABP by HPLC as described above.

2.6. Measurement of N-OH-ABP O-acetyltransferase activity

N-OH-ABP O-acetyltransferase activity was evaluated in MDA-MB-231 cell lysate using reverse phase HPLC as described previously [38]. Reaction mixtures containing 100 μg total protein, 1 mg/mL deoxyguanosine (dG), 100 μM N-OH-ABP, 1 mM AcCoA, and compound 10 (0–100 μM) were incubated at 37°C for 10 min.

2.7. DNA isolation and dG-C8-ABP quantification

The CHO cells stably expressing human NAT1 and CYP1A1 [27–29] were incubated in alpha-MEM media containing 12.5, 40, and 200 μM ABP and various concentrations of compound 10 (0–250 μM). After cells reached approximately 90% confluency, cells were harvested. The cells were centrifuged for 5 min at 13, 000g, and the cell pellet was resuspended in homogenization buffer, 0.1 volume of 10% SDS, and 0.1 volume of Proteinase K and incubated overnight at 37°C. The DNA was extracted and quantitated by absorbance spectrophotometry and the dG-C8-ABP adducts were separated and quantitated by mass spectrometry as previously described [39].

2.8. Inhibition of NAT1 and NAT2 in cryopreserved human hepatocytes

The ability of compound 10 and its structural analog Alizarin to inhibit the N-acetylation of PABA and/or SMZ was investigated in cell cultures of cryopreserved human hepatocytes obtained from BioIVT and cultured as previously described [40]. Five human samples (all of intermediate NAT2 acetylator genotype) were incubated with 50 μM PABA or SMZ for 24 hr and various concentrations of compound 10. PABA or SMZ and their N-acetylated metabolites were separated and quantified by HPLC as described above. In order to compare compound 10 and Alizarin to inhibit human NAT1-catalyzed N-acetylation, media was removed and replaced with media containing compound 10 or Alizarin (1–200 μM). After 1 hr of incubation with inhibitor, PABA was added to a final concentration of 10 μM and the N-acetylation of PABA was quantified by HPLC.

2.9. Compound 10 inhibition of BNA and ABP-induced DNA damage in CHO cells

DNA damage was assessed by a γH2AX in-cell western staining protocol using slight modifications of methods previously described [41]. Cells were grown with selective agents in 10-cm plates and 1×10⁵ cells were plated into black/clear bottom 96-well plates (Corning, Corning, NY, USA) and allowed to attach overnight. The next morning media was removed (cell debris and non-adherent cells were washed away and removed) and attached cells washed with PBS and replaced with fresh pre-warmed media containing 0, 50, 100, or 150 μM compound 10. Cells were incubated for 24 hr, after which BNA or ABP (800 μM) was added and incubated for 24 hr. Media were removed and γH2AX in-cell western staining protocol was performed as follows; cells were fixed to the plate using 3.7% formaldehyde and incubated at room temperature for 20 min. Then, the cells were permeabilized by washing five times with 0.1% Triton X-100 in TBS. After permeabilization, the cells were blocked using FISH Gelatin Blocking Agent (Biotium, Fremont, CA, USA) diluted in TBS for 90 min at room temperature with constant agitation. Primary antibody anti-phospho-histone H2AX (Millipore-Sigma, Burlington, MA, USA) was diluted to 2 μg/mL and added to the cells and then incubated overnight at 4°C. The next morning cells were washed with 0.1% Tween 20 in TBS for 5 min, five times. Secondary antibody IRDye^® 800CW goat anti-mouse IgG (H+L) (LI-COR, Lincoln, NE, USA) was used at a 1:1200 dilution and DNA dye RedDot^™ 2 diluted to 1X (Biotium, Fremont, CA, USA) to normalize for DNA content. Cells were incubated with this combination for 60 min and washed again with the Tween 20 solution as previously described. DNA and the γH2AX were simultaneously visualized using an Odyssey CLx imaging system (LI-COR, Lincoln, NE, USA) with the 680 nm fluorophore (red) and the 800 nm fluorophore (green). Relative fluorescent units for γH2AX per cell (as determined by γH2AX divided by DNA content) were divided by untreated cells.

2.10. Michaelis-Menten kinetic analyses of compound 10 inhibition of NAT1

To further investigate the type of inhibition by compound 10, PABA (10–810 μM) was incubated with compound 10 (0, 50 and 100 μM) in the cell culture media of CHO cells stably expressing human CYP1A1 and NAT1. After 2 hr treatment with various concentrations of compound 10, NAT1 inhibition was determined by quantifying N-acetyl-PABA by HPLC. The resulting data was analyzed by Eadie-Hofstee and Lineweaver-Burke plots.

2.11. Reversibility of compound 10 inhibition of human NAT1

To determine if compound 10 inhibition of NAT1 was reversible or irreversible, primary hepatocytes were plated and after 24 hr, media was removed and cells washed before adding fresh media containing compound 10 (0, 100 or 200 μM). After 1 hr compound 10 was removed from some wells and replaced with fresh media before adding PABA (10 μM) to all wells and incubating for an addition 5 hr to measure remaining NAT1 activity, by quantitating N-acetylated-PABA, as described above.

2.12. Effect of compound 10 on human breast cancer cell viability and proliferation

The effect of compound 10 on cell viability and cell proliferation was assayed using the AlamarBlue (Invitrogen, Eugene, OR) fluorescence assay. MDA-MB-231 and MCF-10A cells were seeded in 96-well plates and incubated for 24 hr. The cells were then treated with varying concentrations of compound 10 in their respective culture medium. AlamarBlue was added 1 hr prior to the desired time point in an amount equal to 1/10 of the culture volume, and the plates were further incubated at 37°C. The amount of AlamarBlue reduction was measured using a fluorometric plate reader (excitation 570 nm; emission 580 nm).

2.13. Effect of compound 10 on human breast cancer cell invasion

To determine the effects of compound 10 on invasive potential of MDA-MD-231 and MCF-7 cells, we utilized the CytoSelectTM 24-well cell invasion assay, colorimetric format (Cell Biolabs Inc., San Diego, CA). Cells (300,000) were seeded and allowed to invade towards culture media containing 10% FBS for 24 hr in the presence of various concentrations (0, 31.25, 62.5, and 125 μM) of compound 10. Cell invasion was plotted as the percentage of controls (DMSO vehicle).

2.14. Statistical Analyses

Statistical significance for differences was assessed using one-way analysis of variance (ANOVA) followed by Tukey multiple comparison test and Student’s t test for single comparisons. IC₅₀ values were determined by nonlinear regression. Kinetic analysis of compound 10 inhibition was assessed by Eadie-Hofstee and Lineweaver-Burk analyses. p < 0.05 was considered significant. All analyses were conducted with GraphPad Prism software (San Diego, CA).

3. RESULTS

3.1. Screening of chemicals to inhibit expression of human recombinant NAT1 and NAT2.

A protein-ligand virtual screening approach of docking approximately 2 million chemicals generated a list of 150 compounds ranked for their approximate binding affinity towards the inhibition of NAT1 or NAT2 (Fig. 1). Sixty of the top-ranked and commercially available compounds were purchased and tested for their ability to inhibit recombinant human NAT1 and NAT2 expressed in yeast. This was determined using the substrates PABA, selective for NAT1, or SMZ, selective for NAT2. Inhibitor specificity for NAT1 was determined by dividing the percent NAT1 inhibition by the percent NAT2 inhibition. The results obtained for these sixty compounds are shown in Supplemental Table 1.

Fig. 1.

Flow chart illustrating experimental scheme for the identification and characterization of potent, selective, and efficacious inhibitors of human NAT1.

3.2. Compound inhibition of human recombinant NAT1 and NAT2

Compound 10 (9,10-dihydro-9,10-dioxo-1,2-anthracenediyl diethyl ester was selected for determination of IC₅₀ values towards both recombinant human NAT1 and NAT2. Compound 10 was a potent and selective NAT1 inhibitor towards PABA with in an in vitro IC₅₀ of 0.75 ± 0.2 μM which was over 100-fold lower for NAT1 versus the IC₅₀ towards SMZ for NAT2 (82.2 ± 12.9 μM) （Fig. 2a). Compound 10 also was a potent and selective NAT1 inhibitor towards the N-acetylation of ABP with in an in vitro IC₅₀ of 0.996 ± 0.2 μM that was over 50-fold lower for NAT1 versus the IC₅₀ for NAT2 (41.3 ± 12.9 μM). Compound 10 exhibited an in situ IC₅₀ of 118 ± 5 μM and 81.6 ± 4.4 μM for inhibition of human NAT1 expressed in CHO cells using PABA and ABP as substrates respectively (Fig. 2b).

Fig. 2.

Compound 10 and Alizarin inhibition of N-acetylation in vitro and in situ. (a): Percent ABP N-acetyltransferase activity determined in vitro using 10 μM ABP, 100 μM AcCoA, and Compound 10 (0 – 200 μM) in yeast lysates that express recombinant human NAT1 or NAT2. Each point represents mean ± SD for 3 experiments. IC₅₀ is 0.996 ± 0.2 μM for NAT1 and 41.3 ± 12.9 μM for NAT2. (b): Compound 10 inhibition of PABA and ABP N-acetylation determined in situ in UV5 DNA repair-deficient CHO cells that express human NAT1 and CYP1A1. Each point represent mean ± SD for 3 experiments. (c): Percent N-acetyltransferase activity remaining in vitro using 10 μM PABA (NAT1) or SMZ (NAT2), 100 μM AcCoA, and Alizarin (0 – 200 μM) in yeast lysates that express recombinant human NAT1 or NAT2. Each point represents mean ± SD. In vitro IC₅₀ was 0.886 ± 0.90 μM for NAT1 and 98.6 ± 4.4 for NAT2. (d): Compound 10 and Alizarin inhibition of PABA N-acetylation determined in situ in human hepatocytes. Each point represents mean ± SD, N=3. In situ IC₅₀ for compound 10 and Alizarin in human hepatocytes was 75.5 ± 0.3 and 70.0 ±1.4 μM respectively.

Three available structural analogs of Compound 10 were tested for their ability to inhibit NAT1. As shown in Table 1, analogs 1 and 2 exhibited in vitro IC₅₀ values of 171 ± 4.7 and 94 ± 6.8 μM towards recombinant human NAT1, showing they were over 100-fold less potent than compound 10. No inhibition of human NAT2 was exhibited with analog 1 or analog 2 at concentrations up to 2 mM. In contrast, analog 3 (Alizarin) was a potent and selective NAT1 inhibitor with in an in vitro IC₅₀ of 0.886 ± 0.0895 μM which was over 100-fold lower for NAT1 versus NAT2. IC₅₀ values were determined in situ in CHO cells that express human NAT1. Compound 10 exhibited an in situ IC₅₀ of 118 μM for inhibition of human NAT1 expressed in CHO cells. Alizarin showed comparable potency towards inhibiting human NAT1 in human hepatocytes（Fig. 2c–d). The in situ IC₅₀ was 75.5 ± 0.3 μM for compound 10 and 70.0 ± 1.4 μM for Alizarin. Compound 10 inhibited the N-acetylation of both PABA and ABP in MDA-MB-231 (Fig. 3a) and MCF-7 (Fig. 3b) breast cancer cells.

Table 1.

IC₅₀ values of compound 10 and three structural analogs. IC₅₀ values listed as Mean ± SEM (n=3) were determined in vitro for recombinant human NAT1 and NAT2 recombinantly expressed in yeast.

Compound	Structure	IC50 for NAT1 (μM)	IC50 for NAT2 (μM)	IC50 NAT2/IC50 NAT1
10		0.75 ± 0.2	82.2 ± 12.9	110
Analog 1		171 ± 4.7	>2000	>11.7
Analog 2		94 ± 6.8	>2000	>21.3
Analog 3 (Alizarin)		0.886 ± 0.0895	98.6 ± 9.4	113

Fig. 3.

Inhibition of N-acetylation capacity in human breast adenocarcinoma MDA-MB-231 (a) and MCF-7 (b) cells that express endogenous NAT1. These cells were treated with media supplemented with 10 μM ABP or PABA with varying concentrations of compound 10 (0 – 250 μM). Each point represents mean ± SEM (n=3).

3.3. Compound 10 inhibits DNA adducts and DNA damage mediated via human NAT1

The ability of compound 10 to inhibit aromatic amine carcinogen-induced DNA damage and formation of DNA adducts was assessed in UV5 DNA repair-deficient CHO cells expressing NAT1 and CYP1A1. Aromatic amine carcinogens such as ABP and BNA induce DNA damage and DNA adducts by N-hydroxylation (catalyzed by CYP1A1) followed by O-acetylation of the N-hydroxy metabolite to the highly electrophilic metabolite (catalyzed by NAT1) which binds to DNA. Compound 10 caused significant (p<0.0001) concentration-dependent reductions in DNA double strand breaks resulting from BNA and ABP concentrations as low as 50 μM（Fig. 4a–b）. Compound 10 also decreased ABP-DNA adducts (via O-acetylation) in human breast cancer cells in vitro with an IC₅₀ of 1.03 ± 0.24 μM and in UV5 DNA repair-deficient CHO cells in situ (Fig. 4c–d）.

Fig. 4.

Compound 10 inhibition of BNA (a) and ABP (b)-induced DNA double-strand breaks in CHO cells expressing human NAT1 and CYP1A1. Inhibition of BNA or ABP-induced double strand breaks following incubation of CHO cells with 0–150 μM compound 10 for 24 hr, followed by 800 μM BNA or ABP for 24 hr. Each bar illustrates the mean ± SEM percent inhibition in DNA double strand breaks relative to vehicle control (n=3). BNA- and ABP-induced DNA double strand breaks differed significantly (**** p<0.0001) with respect to compound 10 across the entire range of concentrations tested. (c) Compound 10 inhibition of N-OH-ABP O-acetyltransferase activity in human breast adenocarcinoma cells (MDA-MB-231) that express endogenous NAT1. Each point represents mean ± SEM (n=3). IC₅₀ = 1.03 ± 0.24 μM. (d) Percent ABP-induced dG-C8-ABP adduct formation in UV5 DNA repair-deficient CHO cells stably expressing CYP1A1 and NAT1. ABP-induced dG-C8-ABP adducts/10⁸ nucleosides were quantitated by mass spectrometry after cells were treated with compound 10 for 24 hr. Adduct formation was plotted as the percentage of untreated cells (control). Each point represents the mean ± SEM (n=3).

3.4. Compound 10 inhibition of human NAT1 is non-competitive and reversible

Kinetic analyses were conducted to further assess inhibition of PABA N-acetylation by compound 10. Eadie-Hofstee (Fig. 5a) and Lineweaver-Burke (Fig. 5b) plots suggest compound 10 inhibition to be noncompetitive or irreversible. Compound 10 showed selective inhibition towards the N-acetylation of PABA (NAT1-selective substrate) versus SMZ (NAT2-selective substrate) in cryopreserved human hepatocytes which express both human NAT1 and NAT2. Compound 10 effectively inhibited the N-acetylation of PABA with an IC₅₀ of 158 ± 22 μM whereas no inhibition of SMZ N-acetylation was observed at compound 10 concentrations up to 200 μM（Fig. 6a). To assess whether or not compound 10 inhibition of human NAT1 was reversible, primary hepatocytes were plated and after 24 hr, media was removed and cells washed before adding fresh media containing compound 10 (0, 100 or 200 μM). After 1 hr, compound 10 was removed from some wells and replaced with fresh media before adding PABA (10 μM) to all wells and incubating for an addition 5 hr to measure PABA N-acetylation. The inhibition of PABA N-acetylation in cryopreserved human hepatocytes by compound 10 was reversible (Fig. 6b).

Fig. 5.

Eadie-Hofstee (a) and Lineweaver-Burke (b) plots of Compound 10 inhibition of human NAT1. CHO cells stably expressing human NAT1 were treated with multiple concentrations of PABA (10–810 μM) were incubated with (0 (circles), 50 (squares) and 100 (triangles) μM compound 10 in the cell culture media of CHO cells plated in 6-well tissue culture plates. After 2 hr treatment with various concentrations of Compound 10, NAT1 inhibition was determined by quantifying N-acetyl-PABA by HPLC. Velocity (V) is in units of nmole/min. Each point represents mean ± SD (n=3).

Fig. 6.

(a) Compound 10 inhibition of PABA and SMZ N-acetylation in situ in cryopreserved human hepatocytes. Each point represent mean ± SEM for 5 individual hepatocyte samples, all with intermediate NAT2 acetylator genotypes. (b) Reversibility of compound 10 binding to NAT1. Data expressed as Mean ± SD (N=2). Significant differences are indicated **, p<0.01; ****, p<0.0001.

3.5. Compound 10 inhibits cell proliferation and invasion in human breast cancer cell lines

We examined the effect of NAT1 inhibition by compound 10 on cell proliferation. Initially, we tested the effect of compound 10 on cell viability to address any concern that a decrease in NAT1 activity was caused by compound 10 toxicity. Cell viability was measured by the AlamarBlue fluorescence assay. This nontoxic dye detects functional oxidative phosphorylation to determine the number of viable adherent and floating cells which is indicated by fluorescence intensity. We evaluated compound 10 toxicity in MDAMB-231 cultured in medium treated with various concentrations of compound 10 for 4, 24, 48, 72, and 144 hr. Fluorescence intensity was measured at time points 4 – 48 hr to determine effects of compound 10 on cell viability. These time points correspond to the periods at which NAT1 activity inhibition experiments were conducted (4 hr) and are within doubling times for these cells. Compound 10 did not decrease cell viability of MDA-MB-231 breast cancer cells at these time points since fluorescence intensity increases at these time points indicate that compound 10 did not affect cell viability (Fig. 7a). A dose-dependent decrease in cell proliferation was observed in MDA-MB-231 breast cancer cells at 72 and 144 hr determined by the decrease in fluorescence intensity which is indicative of a decrease in cell number (Fig. 7a).

Fig. 7.

Compound 10 effect on cell viability and cell proliferation using AlamarBlue® fluorescence assay in (a) tumorigenic MDA-MB-231 and (b) non-metastatic MCF-10A cells. Cells were plated and treated with compound 10 at various concentrations (0– 250 μM) for up to 144 hr. Each point represents the mean ± SEM (n=4). Inhibition of cell invasion by compound 10 in MDA-MB-231 cells (c) and MCF-7 (d) cells. 300,000 cells were seeded and allowed to invade towards culture media containing 10% FBS for 24 hr in the presence of various concentrations (0, 31.25, 62.5, and 125 μM) of compound 10. Each point represents the mean ± SEM (n=3). (e) Representative photographs from cell invasion assay of MDA-MB-231 cells at compound 10 concentrations of 0 and 125 μM.

One major concern when developing a therapeutic inhibitor is the effect the compound will have on non-tumorigenic cells at later time points. Therefore, we also tested the ability of compound 10 to decrease cell viability and proliferation in a non-tumorigenic cell line, MCF-10A. a similar pattern was observed in the MCF-10A non-tumorigenic cells (Fig. 7b) as in the MDA-MB-231 tumorigenic cells (Fig. 7a). Cell viability of MCF-10A cells was not affected as indicated by fluorescence intensity at early time points. However, as observed in MDA-MB-231 cells, a decrease in cell proliferation was observed in the non-tumorigenic cell line. A comparison between the tumorigenic and non-tumorigenic cell lines was established to determine selectivity of compound 10 for tumorigenic cells. This was achieved by calculating the corresponding IC₅₀ for cell proliferation at 72 hr in MCF-10A and MDA-MB-231 which was 113 and 32 μM respectively. We also calculated the in situ therapeutic index = IC₅₀ non-tumorigenic cells/ IC₅₀ tumorigenic cells. The therapeutic index for compound 10 was 3.4, consistent with moderately low toxicity towards non-tumorigenic human breast epithelial cells.

Cell invasion is considered one of the hallmarks of progression in metastatic cancers. Cell migration is an important facilitator of cell invasion because once cells gain the ability to migrate to distal organs and tissues, cell invasion can ensue. Compound 10 treatment resulted in dose-dependent inhibition of MDA-MB-231 (Fig. 7 c,e) and MCF-7 (Fig. 7d) cancer cell migration.

4. DISCUSSION

As previously reviewed [42,43], potential NAT1 inhibitors as a cancer target include numerous polyphenolic chemicals thought to be chemopreventative [44] as well as chemotherapeutic agents such as cisplatin [45] and disulfiram [46]. High-throughput screening studies have resulted in the identification of small molecules that specifically bound to NAT enzymes [47,48]. Additional NAT1 inhibitors include rhodanine derivatives (Z)-5-(2-hydroxybenzylidene)-2-thioxothiazolidin-4-one with an IC₅₀ of 101 μM [47] and (5E)-[5-(4-hydroxy-3,5-diiodobenzylidene)-2-thioxo-1,3-thiazolidin-4-one with an IC₅₀ of 0.732 μM which was 25-fold more selective for NAT1 than NAT2 [48]. Both of these rhodanine NAT1 inhibitors decreased anchorage-independent growth and invasiveness in MDA-MB-231 breast cancer cells [49,50].

Independently of the studies described above, we utilized a protein-ligand docking algorithm of approximately 2 million chemicals which generated a list of 150 compounds ranked for their approximate binding affinity towards the inhibition of arylamine NAT1 or NAT2. Sixty of the top-ranked compounds were tested for their ability to inhibit recombinant human NAT1 and NAT2. Compound 10 was the most potent and selective NAT1 inhibitor with in an in vitro IC₅₀ of 0.75 μM which was over 100-fold lower for NAT1 versus NAT2. Two available structural analogs (analogs 1 & 2) of compound 10 were selective for inhibition of NAT1 versus NAT2 but were less potent than compound 10. In contrast, the third structural analog (Alizarin) showed comparable potency and selectivity to compound 10.

Both NAT1 and NAT2 possess a functional Cys–His–Asp catalytic triad wherein Cys68 is acetylated by AcCoA followed by transfer of the acetyl group to substrate [32]. Compound 10 inhibition of NAT1 was reversible which is inconsistent with covalent binding. The NAT1 catalytic pocket is about 40% smaller than that of NAT2 and substrate selectivity is strongly influenced by the three key active site loop residues PHE125, ARG127 and TYR129 [51]. No steric hindrance appears to hinder compound 10 or Alizarin from binding to the NAT1 active site and selectivity towards NAT1 may be because NAT2 possesses serine groups at amino acids 127 and 129 [51].

Compound 10 inhibited N-acetylation of the arylamine carcinogen ABP both in vitro and in situ in CHO cells stably expressing human NAT1 and CYP1A1. Compound 10 effectively inhibited the N-acetylation of PABA (catalyzed by NAT1) in cryopreserved human hepatocytes whereas inhibition of SMZ N-acetylation (catalyzed by NAT2) was not observed at compound 10 concentrations up to 200 μM. Analog 3 (Alizarin) showed comparable potency and efficacy to compound 10 for inhibition of NAT1 in human hepatocytes.

Since compound 10 can hydrolyze to 1,2-dihydroxyanthraquinone (otherwise known as Alizarin), we also tested this structural analog for inhibition of NAT1 in vitro and in situ. Hydrolysis of compound 10 to Alizarin in the cell-based experiments could account for the differences in potency observed for compound 10 in vitro versus in situ. However, Alizarin and compound 10 showed comparable potency and efficacy for the inhibition of human NAT1 both in vitro and in situ, suggesting that the difference in compound 10 in vitro and in situ potency is not the result of compound 10 hydrolysis. Similarly, the ester component is not responsible for the relative specificity of NAT1 versus NAT2 inhibition since compound 10 and Alizarin exhibited comparable selectivity. The observation that compound 10 inhibition of human NAT1 was reversible may explain in part the higher IC₅₀ values observed in situ. The in situ measurements necessitated incubation times of up to 24 hr and preliminary studies in our laboratory have shown markedly lower in situ IC₅₀ values for compound 10 when incubation times were shorter. Differences also were observed in situ between different human hepatocyte samples as illustrated by the 2-fold difference in NAT1 IC₅₀ determined in human hepatocytes.

Compound 10 caused significant (p<0.0001) concentration-dependent reductions in DNA double strand breaks following metabolism of BNA or ABP in DNA repair-deficient CHO cells expressing human NAT1 and CYP1A1. It also caused concentration -dependent inhibition of ABP DNA adducts. Compound 10 inhibited metabolic activation of N-OH-ABP (via O-acetylation) catalyzed by MDA-MB-231 breast cancer cell lysates with an IC₅₀ of 1.03 ± 0.24 μM. Thus, compound 10 may be useful in reducing DNA adduct formation and damage following exposure to arylamine carcinogens such as BNA and ABP.

Compound 10 inhibited both proliferation and invasion in human breast cancer cells and showed selectivity towards tumorigenic (MDA-MB-231) versus non-tumorigenic cells. These findings support previous studies reporting that treatment of breast cancer cells with small molecule inhibitors of NAT1 leads to decreased invasive ability, proliferation, and anchorage-independent colony formation [49,50]. Other reports have reported the effects of NAT1 knockdown on breast cancer [52] and colon cancer [53] cells and NAT1 knockout in breast cancer cells [54]. A previous study found that Alizarin decreased proliferation, anchorage-independent colony formation, and tumorigenesis of MDA-MB-231 breast cancer cells [55]. A recent report showed that NAT1 promoted bone metastasis and bone destruction in patients with luminal breast cancer, which could be reversed by an unidentified NAT1 inhibitor [56]. Alizarin’s ability to inhibit NAT1 could reduce breast cancer metastasis particularly to bone.

In conclusion, our study identifies potent, selective, and efficacious inhibitors of human NAT1. Compound 10 was shown to reduce DNA adducts and damage resulting from exposure to the arylamine carcinogens BNA and ABP. Further studies are necessary to investigate potency and efficacy of both compound 10 and Alizarin in breast cancer metastasis.

ABBREVIATIONS

NAT1

Arylamine N-acetyltransferase 1

NAT2

Arylamine N-acetyltransferase 2

PABA

p-aminobenzoic acid

SMZ

sulfamethazine

BNA

beta-naphthylamine

ABP

4-aminobiphenyl

CHO

Chinese hamster ovary

AcCoA

acetyl coenzyme A

HPLC

high performance liquid chromatography

dG

deoxyguanosine

N-OH-ABP

N-hydroxy-4-aminobiphenyl

Availability of data and material

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