Studies on Human arylamine N-acetyltransferases.

Abstract

Arylamine N-acetyltransferases (NATs, EC 2.3.1.5) are important phase II drug metabolism enzymes that are expressed in most human tissues. Humans express two NAT isozymes, NAT1 and NAT2, which share 81% sequence identity. NATs catalyze the AcCoA-dependent N-acetylation of arylamines to arylamides, which is a critical detoxification process and competes with cytochrome P450-catalyzed oxidation of arylamines to N-arylhydroxylamines. In this thesis, recombinant human NAT2 was successfully overexpressed and purified to homogeneity; the substrate specificities and molecular interactions of environmental arylamines with human NAT1 and NAT2 were characterized; the in vitro and intracellular inactivation of human NAT1 and NAT2 by the nitrosoarene and N-arylhydroxamic acid metabolites of toxic and carcinogenic arylamines was investigated. To investigate the substrate specificities and inhibition of human NATs, we developed a method to produce homogeneous human "wild type" NAT2 in milligram quantities. Human NAT2 was overexpressed as a L54F dihydrofolate reductase (DHFR) fusion protein linked with a TEV protease-cleavage linker. Chromatography with a methotrexate affinity column, followed by a DEAE column, afforded partial purification of the fusion protein. The fusion protein was digested with TEV protease, and human NAT2 was purified to homogeneity with a second DEAE column. A total of 2.8 mg of human rNAT2 from 2 L of cell culture was purified to homogeneity with this methodology. The kinetic specificity constants (kcat/Km) for N-acetylation of arylamine environmental contaminants were characterized for human NAT1 and NAT2. The dramatic effects of small alkyl substituents on the relative abilities of NAT1 and NAT2 to acetylate substituted anilines was reflected by the 1000-fold difference in the NAT1/NAT2 ratio of the specificity constants for monosubstituted and disubstituted alkylanilines containing methyl and ethyl ring substituents. A NMR-based model was used to interpret the interactions of binding site residues with the alkylanilines and to provide insight into how NATs achieve certain substrate specificities.

Arylamines and their N-hydroxylation products, N-arylhydroxylamines, can undergo oxidation to form nitrosoarenes in vivo. We investigated the inactivation of human NAT1 by the nitrosoarene metabolites of four environmental arylamines in vitro and in human cells. 4-Nitrosobiphenyl (4-NO-BP) and 2-nitrosofluorene (2-NO-F), which are nitroso metabolites of arylamines that are readily N-acetylated by NAT1, were found to be potent inactivators of human NAT1. 4-NO-BP is an affinity label for NAT1 (kinact/KI = 59,200 M-1s -1), whereas 2-NO-F inactivates NAT1 through an apparent bimolecular process (k2 = 34,500 M-1s -1). Glutathione (GSH) afforded only partial protection of NAT1 from inactivation by the two nitrosoarenes in vitro. Nitrosobenzene (NO-B) and 2-nitrosotoluene (2-NO-T), which are nitroso metabolites of arylamines that are less readily acetylated by NAT1, were much weaker inhibitors of NAT1. Treatment of HeLa cells with 4-NO-BP (10 uM) for 15 minutes and 60 minutes caused 39% and 58% losses of NAT1 activity, respectively, without causing a decrease in either glyceraldehyde phosphate dehydrogenase (GAPDH) or glutathione reductase (GR) activities. 2-NO-F was an even more effective inhibitor of HeLa cell NAT1 than 4-NO-BP. Tandem mass spectrometric analysis indicated that 4-NO-BP treatment of HeLa cells in which 3FLAG-NAT1 had been overexpressed resulted in a formation of (4-biphenyl)sulfinamide with the active site Cys68 of NAT1. This is consistent with the results obtained with recombinant NATs in vitro. It was also demonstrated that the nitrosoarene metabolites of arylamines that are efficiently N-acetylated by NAT2 are potent inactivators of NAT2 in vitro and in human cells. The second order rate constants for inactivation of NAT2 by 4-NO-BP and 2-NO-F were 80,400 M-1s -1 and 50,500 M-1s -1, respectively; the values for NO-B and 2-NO-T were 14 M-1s -1 and 16 M-1s -1. Treatment of HeLa cells with 4-NO-BP (5 uM) for 1 h caused a 23% reduction in NAT2 activity, and exposure to 2-NO-F (2.5 uM) for 1 h caused a 22% loss of NAT2 activity, without inhibiting GAPDH and GR activities. Therefore, HeLa intracellular NAT2 is less susceptible to the effects of the lower concentrations of the two nitrosoarenes than is NAT1. It is concluded that NAT1 and NAT2 are intracellular targets of the nitrosoarene metabolites of 4-aminobiphenyl and 2-aminofluorene. Low concentrations of nitrosoarenes may cause a loss of NAT1 and NAT2 activities and impair a key detoxification pathway. N-Arylhydroxamic acids, which are also potentially reactive metabolites of arylamines, have been shown to inactivate hamster NATs, in vitro and in vivo, and human recombinant NAT1. It was previously established that inactivation of NATs by N-hydroxy-4-acetylaminobiphenyl (N-OH-4-AABP) involves an initial NAT-mediated deacetylation to form N-OH-4-aminobiphenyl, which undergoes oxidation to the electrophilic 4-nitrosobiphenyl (4-NO-BP), followed by reaction with the nucleophilic active site Cys68 to form a sulfinamide. We hypothesized that the relative stabilities of the acetyl-enzyme intermediates influence the susceptibilities of NATs to inactivation by N-arylhydroxamic acids. The second order rate constant for inactivation of human NAT2 by N-OH-AAF was 459 M-1s-1, which was 8-fold greater than that for NAT1. Mass spectrometric analysis of both NATs after treatment with N-OH-AAF revealed that the principal adducts were sulfinamide conjugates of Cys68. Kinetic analysis revealed that the hydrolysis rate of acetyl-NAT2 was 4.7-fold greater than that of acetyl-NAT1. Thus, the more rapid inactivation of NAT2 was facilitated by the rapid hydrolysis of the Cys68 thioacetyl ester to free the Cys68 thiol group for reaction with 2-NO-F. The hypothesis was further supported by the results from inactivation of human NATs by N-OH-4-AABP. Treatment of HeLa cells with N-OH-4-AABP (50 uM) for 6 hours had no effect on intracellular NAT2 activity, but caused a 24% decrease in NAT1 activity. Incubation with N-OH-4-AABP (100 uM) for 6 h reduced NAT2 and NAT1 activities by 19% and 56%, respectively. Treatment of HeLa cells with 50 uM N-OH-AAF for 6 hours reduced NAT2 and NAT1 activities by 11% and 33%, respectively. Therefore, approximately 10-fold greater concentrations and longer incubation times were required for the N-arylhydroxamic acids to produce effects on intracellular NATs than were required for 2-NO-F and 4-NO-BP. HeLa NAT1 is more susceptible to the effects of the N-arylhydroxamic acids than NAT2.