Targeting the substrate preference of a type I nitroreductase to develop antitrypanosomal quinone-based prodrugs

Abstract

Nitroheterocyclic prodrugs are used to treat infections caused by Trypanosoma cruzi and Trypanosoma brucei. A key component in selectivity involves a specific activation step mediated by a protein homologous with type I nitroreductases, enzymes found predominantly in prokaryotes. Using data from determinations based on flavin cofactor, oxygen-insensitive activity, substrate range, and inhibition profiles, we demonstrate that NTRs from T. cruzi and T. brucei display many characteristics of their bacterial counterparts. Intriguingly, both enzymes preferentially use NADH and quinones as the electron donor and acceptor, respectively, suggesting that they may function as NADH:ubiquinone oxidoreductases in the parasite mitochondrion. We exploited this preference to determine the trypanocidal activity of a library of aziridinyl benzoquinones against bloodstream-form T. brucei. Biochemical screens using recombinant NTR demonstrated that several quinones were effective substrates for the parasite enzyme, having K(cat)/K(m) values 2 orders of magnitude greater than those of nifurtimox and benznidazole. In tests against T. brucei, antiparasitic activity mirrored the biochemical data, with the most potent compounds generally being preferred enzyme substrates. Trypanocidal activity was shown to be NTR dependent, as parasites with elevated levels of this enzyme were hypersensitive to the aziridinyl agent. By unraveling the biochemical characteristics exhibited by the trypanosomal NTRs, we have shown that quinone-based compounds represent a class of trypanocidal compound.

Fig 1

Trypanosomal NTRs contain FMN as a cofactor. (A) Coomassie-stained SDS-PAGE gel (10%) containing purified, recombinant TbNTR (lane 2). Lane 1, size standards. (B) Absorption spectrum (300 to 550 nm) of purified TbNTR (2 mg) in 50 mM NaH₂PO. (C) Fluorescence spectra of FMN and FAD (both 25 μM) and of supernatant from boiled and purified recombinant TbNTR (0.45 μg) at pH 7.6 (solid line) and pH 2.2 (dashed line) with excitation λ at 450 nm and emission λ between 480 and 600 nm. All the fluorescence analyses were carried out in triplicate; the profiles are derived from the mean values. When TcNTR was used in place of TbNTR, similar absorption and fluorescence profiles were observed.

Fig 2

Site-directed mutagenesis of TcNTR. (A) Schematic of TcNTR from T. cruzi Silvio X10/6 showing the amino-terminal extension (residues 1 to 79), FMN binding motifs (residues 87 to 97 and 297 to 305), and substrate binding regions (residues 119 to 129) (44, 48, 65). Residues selected for mutagenesis are in bold, with arrows indicating the new amino acid substitution. Numbers refer to positions of amino acids in the TcNTR sequence (GenBank accession no. EFZ30370). (B) NTR activity was determined by following the reduction of nifurtimox (0 to 100 μM) at a fixed concentration of NADH (100 μM) in the presence of purified wild-type (WT) or mutant TcNTR protein (4 μg). All assays were carried out in triplicate. The apparent V_max for each protein was calculated, and the results are presented as a percentage of the wild-type activity ± standard deviation. (C) The flavin recovered from the boiled, clarified supernatants of purified wild-type or mutant TcNTR (0.45 μg) was identified as described for Fig. 1. The fluorescence of each supernatant was measured at pH 7.6 using excitation λ = 450 nm and emission λ = 535 nm. For each protein, the fluorescence value from three independent readings was taken, and the data are presented as a percentage of the wild-type value ± standard deviation.

Fig 3

Investigating the kinetic properties of trypanosomal type I nitroreductase toward quinones. (A) Proposed scheme for the reduction of quinone- and nitroaromatic-based substrates by NTRs using NADH as the electron donor; “red” represents the reduced and “oxid” the oxidized form of the compound. The interactions of the trypanosomal NTR with NADH (reaction I) and the substrate (reaction II) are indicated. (B) NTR activity was assayed by following the reduction of DCPIP at 600 nm. In reaction I, assays were carried out using different concentrations of NADH (0 to 100 μM) in the presence of DCPIP (0.5 [●], 1.0 [■], and 2 [▼] μM) and TcNTR (20 μg). In reaction II, NTR activity was assayed by following the reduction of various concentrations of DCPIP (0 to 2 μM) in the presence of NADH (20 [♦], 40 [▼], and 60 [▲] μM) and TcNTR (20 μg). When TbNTR was used in place of TcNTR, similar enzyme kinetic profiles were observed.

Fig 4

Inhibition of trypanosomal NTR quinone reductase activity by dicoumarol. (A) DCPIP (5 μM) reduction by TbNTR (20 μg) in the presence of NADH (100 μM) was readily inhibited by dicoumarol (0 to 200 nM). Inhibitory activity is expressed as a percentage compared to uninhibited control results. (B) NTR activity was assayed by following DCPIP (5 μM) reduction in the presence of NADH (0 to 100 μM) and 0 (●), 5 (■), 10 (▲), and 20 (▼) nM dicoumarol. All reactions were initiated by addition of TbNTR (20 μg), and activity (v) is expressed in micromoles of DCPIP reduced per minute per milligram.

Fig 5

Susceptibility of bloodstream-form T. brucei to aziridinyl benzoquinones. IC₅₀ values of the most potent aziridinyl benzoquinones on T. brucei cells induced to express elevated levels of TbNTR (black bars) compared to noninduced control values (white bars) are shown. Data represent means ± standard deviations (SD) of the results of four experiments, and the differences in susceptibility for DZQ (compound 1), MeDZQ (compound 2), RH1 (compound 3), TZQ (compound 5), and compound 6 were statistically significant (P < 0.01), as assessed by Student's t test.