Analysis of flavin oxidation and electron-transfer inhibition in Plasmodium falciparum dihydroorotate dehydrogenase

Abstract

Plasmodium falciparum dihydroorotate dehydrogenase (pfDHODH) is a flavin-dependent mitochondrial enzyme that provides the only route to pyrimidine biosynthesis in the parasite. Clinically significant inhibitors of human DHODH (e.g., A77 1726) bind to a pocket on the opposite face of the flavin cofactor from dihydroorotate (DHO). This pocket demonstrates considerable sequence variability, which has allowed species-specific inhibitors of the malarial enzyme to be identified. Ubiquinone (CoQ), the physiological oxidant in the reaction, has been postulated to bind this site despite a lack of structural evidence. To more clearly define the residues involved in CoQ binding and catalysis, we undertook site-directed mutagenesis of seven residues in the structurally defined A77 1726 binding site, which we term the species-selective inhibitor site. Mutation of several of these residues (H185, F188, and F227) to Ala substantially decreased the affinity of pfDHODH-specific inhibitors (40-240-fold). In contrast, only a modest increase in the Kmapp for CoQ was observed, although mutation of Y528 in particular caused a substantial reduction in kcat (40-100-fold decrease). Pre-steady-state kinetic analysis by single wavelength stopped-flow spectroscopy showed that the mutations had no effect on the rate of the DHO-dependent reductive half-reaction, but most reduced the rate of the CoQ-dependent flavin oxidation step (3-20-fold decrease), while not significantly altering the Kdox for CoQ. As with the mutants, inhibitors that bind this site block the CoQ-dependent oxidative half-reaction without affecting the DHO-dependent step. These results identify residues involved in inhibitor binding and electron transfer to CoQ. Importantly, the data provide compelling evidence that the binding sites for CoQ and species-selective site inhibitors do not overlap, and they suggest instead that inhibitors act either by blocking the electron path between flavin and CoQ or by stabilizing a conformation that excludes CoQ binding.

Figure 1

DHODH inhibitors.

Figure 2. Species-selective inhibitor binding site of pfDHODH

A stereo cartoon drawing of the enzyme backbone bound to A77 1726 is displayed. The backbone is colored gray, oxygen is red, nitrogen blue, fluorine cyan, and phosphate orange. Carbon atoms of detergent molecules (pentaethylene glycol monooctyl ether) are gray, the inhibitor A77 1726 and product orotate are colored magenta, FMN is yellow, and residues within 4 Å of the inhibitor chosen for mutation are displayed and colored green. Residue F188 is an alanine and residue I272 is a valine in the human enzyme. All other residues displayed are conserved between the two species.

Figure 3. Inhibition of wild-type pfDHODH steady-state reaction by DCPMNB using different electron acceptors

The steady-state reaction was allowed to proceed in the presence of dissolved oxygen (~300 μM) (open circles) or K₃Fe(CN)₆ (100 μM) (open squares), CoQ₁ (100 μM) (closed circles), or CoQ_D (100 μM) (closed squares). DCPMNB inhibition data for CoQ₁ and CoQ_D yield an IC₅₀ of 48 and 66 nM, respectively. DCPMNB (100 μM) inhibits oxygen- and K₃Fe(CN)₆-dependent activity by 12% and 30% respectively.

Figure 4. Absorbance spectra of wild-type pfDHODH

Oxidized pfDHODH (10 μM) in the presence of buffer containing dissolved oxygen (solid line). pfDHODH was mixed with DHO (500 μM) and the spectra of the reduced enzyme was recorded immediately (dotted line). The reaction with DHO and dissolved oxygen was allowed to proceed for several minutes to completion and the absorbance spectrum of re-oxidized enzyme, in the presence of product orotate, was recorded (dashed line).

Figure 5. Pre-steady-state kinetic analysis of the wild-type pfDHODH reductive half-reaction

(a) Absorbance traces (closed circles) for pfDHODH (final concentration 20 μM) after rapid mixing with DHO (final concentrations 62, 125, 250, 500, 1000 μM) at 4°C. Data were fitted to Equation 2 using double exponentials (solid curves) to obtain k_obs values. The residual plot for the fit are displayed above the graph. (b) The DHO concentration dependence of k_obs,1 and k_obs,2 (open circles and open squares, respectively). The k_obs,1 for the first observed kinetic step was fitted to Equation 3 (K_d,red = 230 ± 70 μM; k_1,red = 350 ± 30 s⁻¹).

Figure 6. Pre-steady-state kinetic analysis of the wild-type pfDHODH oxidative half-reaction

(a) Absorbance traces (closed circles) are displayed for enzyme (15 μM enzyme pre-reduced with 10 μM DHO, final concentrations) after rapid mixing with CoQ₁ (final concentrations 31, 62, 125, 250, 500 μM) at 4°C. Data were fitted to Equation 2 using a single exponential (solid curve). The residual plot for the fit are displayed above the graph. (b) CoQ₁ concentration dependence of the k_obs,1 (open circles). The k_obs for the observed kinetic step was fitted to the Equation 3 to determine the kinetic parameters (K_d,ox = 67 ± 5 μM; k_1,ox = 67 ± 2 s⁻¹).

Figure 7. Pre-steady-state kinetic analysis of wild-type pfDHODH in the presence of inhibitors

Reductive half-reaction: (a) Absorbance traces of enzyme (20 μM) alone (trace 1) or pre-mixed with DCPMNB (50 μM) (trace 2), A77 1726 (1 mM) (trace 3), or OA (1 mM) (trace 4) and then mixed with DHO (125 μM). All concentrations are after mixing. Oxidative half-reaction: (b) Absorbance traces of enzyme (15 μM pre-reduced with 10 μM DHO) alone (trace 1) or pre-mixed with DCPMNB (50 μM) (trace 2) or A77 1726 (1 mM) (trace 3) and then mixed with CoQ₁ (100 μM). All concentrations are after mixing. Experiments were performed at 4°C.

Scheme 1

Scheme 2

Scheme 3