Acetylation of putative arylamine and alkylaniline carcinogens in immortalized human fibroblasts transfected with rapid and slow acetylator N-acetyltransferase 2 haplotypes

Abstract

Exposure to alkylanilines found in tobacco smoke and indoor air is associated with risk of bladder cancer. Genetic factors significantly influence the metabolism of arylamine carcinogens and the toxicological outcomes that result from exposure. We utilized nucleotide excision repair (NER)-deficient immortalized human fibroblasts to examine the effects of human N-acetyltransferase 1 (NAT1), CYP1A2, and common rapid (NAT2*4) and slow (NAT2*5B or NAT2*7B) acetylator human N-acetyltransferase 2 (NAT2) haplotypes on environmental arylamine and alkylaniline metabolism. We constructed SV40-transformed human fibroblast cells that stably express human NAT2 alleles (NAT2*4, NAT2*5B, or NAT2*7B) and human CYP1A2. Human NAT1 and NAT2 apparent kinetic constants were determined following recombinant expression of human NAT1 and NAT2 in yeast for the arylamines benzidine, 4-aminobiphenyl (ABP), and 2-aminofluorene (2-AF), and the alkylanilines 2,5-dimethylaniline (DMA), 3,4-DMA, 3,5-DMA, 2-6-DMA, and 3-ethylaniline (EA) compared with those of the prototype NAT1-selective substrate p-aminobenzoic acid and NAT2-selective substrate sulfamethazine. Benzidine, 3,4-DMA, and 2-AF were preferential human NAT1 substrates, while 3,5-DMA, 2,5-DMA, 3-EA, and ABP were preferential human NAT2 substrates. Neither recombinant human NAT1 or NAT2 catalyzed the N-acetylation of 2,6-DMA. Among the alkylanilines, N-acetylation of 3,5-DMA was substantially higher in human fibroblasts stably expressing NAT2*4 versus NAT2*5B and NAT2*7B. The results provide important insight into the role of the NAT2 acetylator polymorphism (in the presence of competing NAT1 and CYP1A2-catalyzed N-acetylation and N-hydroxylation) on the metabolism of putative alkyaniline carcinogens. The N-acetylation of two alkylanilines associated with urinary bladder cancer (3-EA and 3,5-DMA) was modified by NAT2 acetylator polymorphism.

Keywords: 3,5-dimethylaniline; 3-ethylaniline; Alkylanilines; Arylamine N-acetyltransferase 2; N-acetylation polymorphism.

Fig. 1.

Construction of SV40-transformed human fibroblast cells that stably express human CYP1A2 and NAT2 haplotypes (NAT2*4, NAT2*5B, or NAT2*7B).

Fig. 2.

CYP1A2 EROD activity in SV40-transformed GM4429 cells. Each bar represents mean ± SEM for 3 experiments determining 7-ethoxyresorufin-O-deethylase (EROD) activity (pmoles/min/1x10⁶ cells) in SV40-transformed GM4429 cells stably transfected with human CYP1A2 and NAT2 haplotypes. No significant difference in EROD activity was observed among transfected cells using one-way ANOVA analysis (p > 0.05) EROD activity in non-transfected cells was below the limit of detection (<0.05 pmoles/min/1x10⁶ cells) in null (non-transfected cells).

Fig. 3.

PABA N-acetyltransferase activity and NAT1 mRNA in SV40 transformed GM4429 cells stably transfected with human CYP1A2 and NAT2 haplotypes. No significant difference in relative NAT1 mRNA levels was observed among non-transfected and transfected cells using one-way ANOVA (p > 0.05) No significant difference in PABA N-acetylation was observed among all GM4429 cells lines using one-way ANOVA (p > 0.05). Each bar represents mean ± SEM for 3 experiments. NAT1 genotype was NAT1*4/*4.

Fig. 4.

SMZ N-acetyltransferase activity and NAT2 mRNA in SV40 transformed GM4429 cells stably transfected with human CYP1A2 and NAT2 haplotypes. Each bar represents mean ± SEM for 3 experiments determining SMZ NAT activity and NAT2 mRNA levels in SV40-transformed GM4429 cells stably transfected with human CYP1A2 and human NAT2 alleles. No significant difference in relative NAT2 mRNA levels was observed among transfected cells using one-way ANOVA (p > 0.05). Levels in null and CYP1A2 only transfected cells had non-detectable levels of NAT2 mRNA and SMZ N-acetyltransferase activity (<0.15 nmoles/min/mg protein). Lysates collected from cells transfected with NAT2*5B and NAT2*7B showed significantly lower SMZ N-acetylation activity when compared to cell extracts transfected with the reference allele NAT2*4 using one-way ANOVA followed by Bonferroni posttest (p<0.001).

Fig. 5.

N-acetylation of benzidine (specificity constant 9.95), 3,4-DMA (specificity constant 2.93) 2-aminofluorene (specificity constant 2.71) 3,5-DMA (specificity constant 0.01), 2,5-DMA (specificity constant 0.05), and 3-EA (specificity constant 0.13) in cell lysates of immortalized human fibroblasts stably transfected with human CYP1A2 and NAT2 haplotypes. Each bar represents mean ± SEM for 3 experiments. N-acetylation activity among the untransfected and transfected cells were tested by one way ANOVA and the significance of any differences for each substrate is noted on the figures. With respect to differences between cells following Tukey’s multiple comparisons testing, 3,5-DMA N-acetylation in non-transfected cells and cells transfected with NAT2*5B and NAT2*7B was significantly lower than cells expressing the reference allele NAT2*4 (p < 0.05). 2,5-DMA, and 3-EA N-acetylation in non-transfected cells was significantly lower than cells expressing the reference allele NAT2*4 (p < 0.05). N-acetylation of 2-AF was not significantly (p > 0.05) lower in the untransfected compared to transfected cells.

Fig. 6.

ABP N-acetyltransferase and N-hydroxy-ABP O-acetyltransferase activities in immortalized human fibroblasts transfected with human CYP1A2 and NAT2 haplotypes. Each bar represents mean ± SEM for 3 experiments. No significant differences were observed among cell lines following one-way ANOVA analysis (p > 0.05).