Thioredoxin Reductase-Type Ferredoxin: NADP⁺ Oxidoreductase of Rhodopseudomonas palustris: Potentiometric Characteristics and Reactions with Nonphysiological Oxidants

Abstract

Rhodopseudomonas palustris ferredoxin:NADP⁺ oxidoreductase (RpFNR) belongs to a novel group of thioredoxin reductase-type FNRs with partly characterized redox properties. Based on the reactions of RpFNR with the 3-acetylpyridine adenine dinucleotide phosphate redox couple, we estimated the two-electron reduction midpoint potential of the FAD cofactor to be −0.285 V. 5-Deaza-FMN-sensitized photoreduction revealed −0.017 V separation of the redox potentials between the first and second electron transfer events. We examined the mechanism of oxidation of RpFNR by several different groups of nonphysiological electron acceptors. The k_cat/K_m values of quinones and aromatic N-oxides toward RpFNR increase with their single-electron reduction midpoint potential. The lower reactivity, mirroring their lower electron self-exchange rate, is also seen to have a similar trend for nitroaromatic compounds. A mixed single- and two-electron reduction was characteristic of quinones, with single-electron reduction accounting for 54% of the electron flux, whereas nitroaromatics were reduced exclusively via single-electron reduction. It is highly possible that the FADH· to FAD oxidation reaction is the rate-limiting step during the reoxidation of reduced FAD. The calculated electron transfer distances in the reaction with quinones and nitroaromatics were close to those of Anabaena and Plasmodium falciparum FNRs, thus demonstrating their similar “intrinsic” reactivity.

1. Introduction

Ferredoxin:NADP⁺ oxidoreductases (FNRs) transfer redox equivalents between NADP(H) and the low-redox-potential FeS protein ferredoxin (Fd), or flavodoxin, a low-molecular-weight flavin mononucleotide (FMN)–containing protein [1,2,3,4,5,6,7]. FNRs comprise separate flavin adenine dinucleotide (FAD)- and NADP(H)-binding domains. The stabilization of the neutral (blue) FAD semiquinone (FADH·) as the reaction intermediate takes place by transforming the two-electron (hydride) transfer into two single-electron transfer events [1,2,8]. The complex formation between FNR and Fd is frequently attributed to the electrostatic and hydrophobic interactions [2,6,9].

FNRs are found in a wide variety of organisms and are classified into several groups and subclasses, whose representatives differ in amino acid sequence, catalytic rate, specificity for NAD(P)(H), physiological functions, and direction of electron transfer [3,7,10]. Most recently, a distinctive subclass of thioredoxin reductase-type FNRs has been discovered, whose structure exhibits the low M_r thioredoxin reductase (TrxR) fold ([10], and references therein). This fold consists of two domains with Rossmann-like three-layer ββα sandwich folds that bind FAD and NADP(H). The typical representatives of this subclass are dimeric FNRs of the green sulfur bacterium Chlorobaculum tepidum [11,12], the heterotrophic Gram-positive bacterium Bacillus subtilis [13,14,15], and the photosynthetic purple nonsulfur bacterium Rhodopseudomonas palustris [15,16]. In these enzymes, the residues of the NADP(H)-binding domain are inserted between the two sections of the FAD-binding domain residues, and a hinge region connects the two domains (Figure 1). The rotation of the domains relative to each other may take place in catalysis, e.g., in B. subtilis FNR, NADP⁺ is bound ca. 15 Å away from the isoalloxazine ring of FAD, which is too distant for efficient hydride transfer [13]. The redox properties of Trx-type FNRs have been characterized partly, limited mostly by the studies of their reactions with NADP⁺/NADPH and Fds.

R. palustris TrxR-type FNR (RPA3954, RpFNR, EC 1.18.1.2) consists of the FAD-binding domain (residues 2–122 and 255–344, including the flexible C-terminal region, residues 319–344) containing a specific Tyr328 residue covering the re side of the isoalloxazine ring, and the NADP(H)-binding domain (residues 127–250). The reduction of RpFNR by NADPH and reoxidation by NADP⁺ proceeds in several phases, the fastest ones exceeding 500 s⁻¹, and involves the formation of several intermediate charge-transfer complexes [16]. On the other hand, RpFNR has low reactivity toward Fe₂S₂-type ferredoxin (RPA3956), whereas its reactivity toward Fe₄S₄-type Fds of R. palustris has not been reported [15].

In order to extend the understanding of the redox properties of RpFNR, we investigated its reactions with nonphysiological electron acceptors with different structures and single-electron reduction potentials (E¹₇). It is worth noting that R. palustris is capable of metabolizing aromatic compounds formed during plant degradation, which may involve FNR/Fd and cytochrome P-450-dependent redox systems [17,18]. Besides, some FNRs, such as the malaria parasite Plasmodium falciparum FNR, are a potential target for redox active drug candidates, quinones and nitroaromatic compounds [19,20]. In order to quantitatively analyze the obtained results, the redox potentials of RpFNR were also determined in this work.

2. Materials and Methods

2.1. Enzymes and Reagents

Recombinant R. palustris ferredoxin:NADP⁺ oxidoreductase was prepared as previously described [16]. Its concentration was determined spectrophotometrically according to ε₄₆₆ = 10.8 mM⁻¹ cm⁻¹ [16]. 2,4,6-Trinitrotoluene (TNT) and 2,4,6-trinitrophenyl-N-methylnitramine (tetryl), synthesized as described [21], and 3-amino-1,2,4-benzotriazine-1,4-dioxide (tirapazamine) and its 7-methyl- and 7-fluoro- derivatives, synthesized according to [22], were a generous gift from Dr. Jonas Šarlauskas (Institute of Biochemistry, Vilnius). 5-(1-Aziridinyl)-2,4-dinitrobenzamide (CB-1954) synthesized as described in [23], was a generous gift from Dr. Vanda Miškinienė (Institute of Biochemistry, Vilnius). The above compounds were characterized by their melting points and their ¹H-NMR, UV, and IR spectra. The purity of compounds determined using HPLC-MS (LCMS-2020, Shimadzu, Kyoto, Japan) was >98%. NADPH, 3-acetylpyridine adenine dinucleotide phosphate (AcPyP⁺), horse heart cytochrome c, superoxide dismutase, and other reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA) and used as received.

2.2. Photoreduction of RpFNR

RpFNR (16–17 µM) photoreduction was performed under anaerobic conditions in 0.02 M Hepes buffer, pH 7.0, using 5-deaza-FMN (0.125 µM) and EDTA (8 mM) as photosensitizers. Before protein introduction from a concentrated stock solution, the solution in a sealed spectrophotometer cuvette was flushed with O₂-free argon for 60 min. Here and in subsequent experiments, a Cary60 UV-Vis (Agilent Technologies, Santa Clara, CA, USA) or a PerkinElmer Lambda 25 UV–VIS spectrophotometer (PerkinElmer, Waltham, MA, USA) was used. Subsequently, the cell was irradiated for short periods at 20 °C with a 100 W incandescent lamp (Osram) at a distance of 20 cm; the progress of the reaction was followed spectrophotometrically for 1–1.5 h. The maximal amount of neutral semiquinone (E-FADH·) formed under irradiation was assumed to be defined by the inflection point of the A₆₀₀ vs. A₄₆₆ plot. The FAD semiquinone concentration was calculated using ε₆₀₀ = 5.0 mM⁻¹cm⁻¹ [24]. The separation between the two single-electron-transfer potentials (ΔE¹₇) was further calculated from the semiquinone formation constant K_s (Equations (1) and (2)):

[E-FADH]_max/[E-FAD]_tot = K_s^1/2/(2 + K_s^1/2),

(1)

ΔE¹₇ = E₇(E-FAD/E-FADH) − E₇(E-FADH / E-FADH⁻) = 0.059 V × log K_s,

(2)

where [E-FADH]_max is the maximal amount of formed semiquinone, E₇(E-FAD/E-FADH) is the potential of oxidized/semiquinone couple, E₇(E-FADH/E-FADH⁻) is the potential of semiquinone/reduced FAD couple, and [E-FAD]_tot is the total enzyme concentration [25].

2.3. Steady-State Kinetic Studies

The kinetic experiments were performed spectrophotometrically in 0.02 M Hepes + 1 mM EDTA buffer (pH 7.0), at 25 °C. The kinetic data were fitted to a parabolic expression in SigmaPlot (v. 11.0, SPSS Inc., Chicago, IL, USA) to yield the steady-state parameters of the reactions, catalytic constants (k_cat(app.)), and bimolecular rate constants (k_cat/K_m)) of the oxidants under fixed concentrations of NADPH. They are equal to the reciprocal intercepts and slopes of Lineweaver–Burk plots, [E]/v vs. 1/[oxidant], respectively, where [E] is the enzyme concentration, and v is the reaction rate. k_cat represents the number of molecules of NADPH oxidized by a single active center of the enzyme per second at saturated concentrations of both substrates. Kinetic parameters of steady-state reactions according to a “ping-pong” mechanism were calculated according to Equation (3):

$$\frac{v}{\left[\mathrm{E}\right]}=\frac{k_{\mathrm{cat} }\left[\mathrm{S}\right]\left[\mathrm{Q}\right]}{K_{\mathrm{m}\left(\mathrm{S}\right)}\left[\mathrm{Q}\right]+K_{\mathrm{m}\left(\mathrm{Q}\right) }\left[\mathrm{S}\right]+\left[\mathrm{S}\right]\left[\mathrm{Q}\right]} ,$$

(3)

where S stands for NADPH, and Q stands for the electron acceptor. The competitive inhibition constant (K_is) of NADP⁺ (I) vs. NADPH was calculated according to Equation (4):

$$\frac{v}{\left[\mathrm{E}\right]}=\frac{k_{\mathrm{cat}\left(\mathrm{app}\right) }\left[\mathrm{S}\right]}{K_{\mathrm{m}\left(\mathrm{S}\right)}\left(1+\frac{\left[\mathrm{I}\right]}{K_{\mathrm{is}}}\right)+\left[\mathrm{S}\right]} ,$$

(4)

and the noncompetitive inhibition constant (K_ii) of NADP⁺ vs. electron acceptor (Q) was calculated according to Equation (5):

$$\frac{v}{\left[\mathrm{E}\right]}=\frac{k_{\mathrm{cat}\left(\mathrm{app}\right) }\left[\mathrm{Q}\right]}{(K_{\mathrm{m}\left(\mathrm{Q}\right)}+\left[\mathrm{Q}\right])\left(1+\frac{\left[\mathrm{I}\right]}{K_{\mathrm{ii}}}\right)} .$$

(5)

The rates of enzymatic NADPH oxidation in the presence of quinones, nitroaromatic compounds, or tirapazamine derivatives were determined using the value ∆ε₃₄₀ = 6.2 mM⁻¹ cm⁻¹. The rates were corrected for the intrinsic NADPH-oxidase activity of RpFNR, 0.12 s⁻¹. In separate experiments, in which 50 µM cytochrome c was additionally added into the reaction mixture, its quinone- or nitroaromatic-mediated reduction was assessed using the value ∆ε₅₅₀ = 20 mM⁻¹cm⁻¹. The ferricyanide reduction rate was measured using the value ∆ε₄₂₀ = 1.03 mM⁻¹cm⁻¹. The rate of the RpFNR-catalyzed reduction of AcPyP⁺ by NADPH was determined using the value ∆ε₃₆₃ = 5.6 mM⁻¹ cm⁻¹ [26]. AcPyPH, the reduced form of AcPyP⁺, was prepared in situ by the reduction of AcPyP⁺ with 10 mM glucose-6-phosphate and 0.01 mg/mL glucose-6-phosphate dehydrogenase. AcPyPH concentration was determined according to ε₃₆₅ = 7.8 mM⁻¹cm^–1 [26]. The statistical analysis was performed using Statistica (version 4.3, Statsoft, Toronto, ON, Canada).

2.4. Presteady-State Kinetic Studies

The rapid kinetic studies of RpFNR were performed using a SX20 stopped-flow spectrophotometer (Applied Photophysics, Leatherhead, UK) under aerobic conditions. The enzyme reduction by NADPH and its reoxidation was monitored between 450 and 800 nm, as further described in the Results section. During turnover studies, RpFNR in the first syringe (4.0 µM after mixing) was mixed with the contents of the second syringe (50 µM NADPH and 250 µM tetramethyl-1,4-benzoquinone after mixing). The control experiments were performed in the absence of quinone. The reoxidation kinetics were analyzed by the method of Chance [27] according to Equation (6), where k_ox is the apparent first-order rate constant of enzyme reoxidation, [NADPH]₀ is the initial NADPH concentration, [E_red]_max is the maximal concentration of the reduced enzyme formed during the turnover, and t_1/2(off) is the time interval between the formation of the half-maximal amount of E_red and its decay to the half-maximal value:

k_ox = [NADPH]₀/([E_red]_max × t_1/2(off)).

(6)

3. Results

3.1. Determination of Redox Potentials of RpFNR

According to the best of our knowledge, the potentiometric characteristics of RpFNR were unavailable so far. In order to determine the standard redox potential (E⁰₇, potential of E-FAD/E-FADH⁻ redox couple) of RpFNR, we examined its reactions with the analogue of NADP(H), 3-acetylpyridine adenine dinucleotide phosphate, AcPyP(H) (E⁰₇ = −0.258 V). AcPyP⁺ was chosen instead of NADP⁺ because the reduction of NADP⁺ by RpFNR under steady-state conditions is problematic due to the lack of a suitable electron donor [15]. During the enzymatic reduction of AcPyP⁺ by NADPH, the maximum reaction rate was reached at 200 µM NADPH. In this reaction, k_cat = 53.3 ± 3.1 s⁻¹, and k_cat/K_m for AcPyP⁺ is estimated to be 2.27 ± 0.35 × 10⁶ M⁻¹s⁻¹ (Figure 2A).

In the reverse reaction using AcPyPH generated in situ and 1.0 mM ferricyanide as an electron acceptor, k_cat = 18.6 ± 0.7 s⁻¹, and k_cat/K_m for AcPyPH is calculated to be 6.0 ± 0.6 × 10⁵ M⁻¹s⁻¹ on the two-electron basis (Figure 2A). According to the Haldane relationship, the equilibrium constant (K) of the redox reaction with the AcPyP⁺/AcPyPH couple corresponds to the ratio of k_cat/K_m for AcPyPH and AcPyP⁺, respectively. According to the Nernst equation, the difference between the redox potentials of the reactants, ΔE⁰, equals 0.0295 V × log K. This provides the K value of 0.264 ± 0.067, and the E⁰₇ value for the enzyme of −0.276 ± 0.003 V, respectively.

During the photoreduction of RpFNR in the presence of 5-deaza-FMN and EDTA, the neutral FAD semiquinone (FADH·) with the characteristic absorbance at 550–650 nm is formed (Figure 2B). The amount of E-FADH· calculated using ε₆₀₀ = 5.0 mM⁻¹cm⁻¹ [24] and the data from the inset of Figure 2B is 26.5%. According to Equations (1) and (2), this gives K_s = 0.520, and ΔE¹₇ = −0.017 V, which corresponds to E₇ (E-FAD / E-FADH·) = −0.285 V and E₇(E-FADH·/ E-FADH⁻) = −0.268 V. However, a slight correction for these potentials is not ruled out, as the ε₆₀₀ of FADH of RpFNR is not definitely determined.

3.2. Steady-State Kinetics and Oxidant Substrate Specificity Studies of RpFNR

The previous kinetic studies of RpFNR were performed using the classical nonphysiological electron acceptor of FNRs, ferricyanide [16]. We preliminarily identified a representative of another group of compounds, juglone (5-hydroxy-1,4-naphthoquinone) as an efficient nonphysiological electron acceptor of RpFNR. A series of parallel lines obtained in Lineweaver–Burk plots at varied concentrations of NADPH and fixed concentrations of juglone are indicative of a “ping-pong” mechanism for the quinone reductase activity of RpFNR (Figure 3).

As deduced from Equation (3), the k_cat value for the juglone reduction at infinite NADPH concentration is equal to 157 ± 7.0 s⁻¹, and the values of the bimolecular rate constants (k_cat/K_m) for NADPH and juglone are equal to 8.7 ± 0.7 × 10⁶ M⁻¹s⁻¹ and 1.62 ± 0.15 × 10⁶ M⁻¹s⁻¹, respectively. The value of k_cat/K_m for NADPH is similar to that obtained previously, 5.5 × 10⁶ M⁻¹s⁻¹, using ferricyanide as the acceptor [16].

Next, we assessed the oxidant substrate specificity of RpFNR, examining its reactions with quinones (Q), nitroaromatic compounds (ArNO₂), and aromatic N-oxides (ArN→O), which comprise three distinct groups of electron acceptors characterized by single-electron reduction midpoint potentials (E¹₇) in the range of 0.09 to −0.494 V. These compounds were studied along with several single-electron acceptors such as ferricyanide, Fe(EDTA)⁻, and benzylviologen. The studied compounds included the explosives tetryl and 2,4,6-trinitrotoluene, antibacterial agents nitrofurantoin and nifuroxime, and anticancer agents CB-1954 and tirapazamine. The apparent maximal reduction rate constants, k_cat(app), of electron acceptors at 100 µM NADPH and their respective k_cat/K_m are given in

Table 1. Steady-state rate constants of the reduction of nonphysiological electron acceptors by 100 µM NADPH catalyzed by RpFNR. The E¹₇ values of compounds taken from [22,28,29].

No.	Compound	E17 (V)	kcat(app) (s−1)	kcat/Km (M−1s−1)
Quinones
1	1,4-Benzoquinone	0.090	130 ± 16	9.4 ± 0.8 × 105
2	2-CH3-1,4-benzoquinone	0.010	130 ± 12	5.6 ± 0.6 × 105
3	2,6-(CH3)2-1,4-benzoquinone	−0.070	52.1 ± 1.8	2.1 ± 0.13 × 105
4	5-OH-1,4-naphthoquinone	−0.090	138.5 ± 9.3	1.5 ± 0.23 × 106
5	5,8-(OH)2-1,4-naphthoquinone	−0.110	45.4 ± 3.4	3.5 ± 0.2 × 106
6	9,10-Phenanthrene quinone	−0.120	34.6 ± 2.4	2.0 ± 0.4 × 106
7	1,4-Naphthoquinone	−0.150	110 ± 13	2.3 ± 0.4 × 105
8	2-CH3-1,4-naphthoquinone	−0.200	21.6 ± 2.1	6.0 ± 0.8 × 104
9	(CH3)4-1,4-benzoquinone(duroquinone)	−0.260	4.07 ± 0.53	1.1 ± 0.1 × 104
10	9,10-Anthraquinone-2-sulphonate	−0.380	3.56 ± 0.33	1.0 ± 0.16 × 104
11	2-OH-1,4-naphthoquinone	−0.410	0.26 ± 0.03	3.1 ± 0.2 × 103
12	2-CH3-3-OH-1,4-naphthoquinone	−0.460	1.35 ± 0.13	4.8 ± 0.4 × 103
Nitroaromatic compounds
13	Tetryl	−0.191	5.69 ± 0.14	4.35 ± 0.30 × 104
14	N-methylpicramide	−0.225	1.93 ± 0.26	4.8 ± 0.6 × 103
15	2,4,6-Trinitrotoluene	−0.253	1.30 ± 0.13	2.43 ± 0.14 × 103
16	Nifuroxime	−0.255	4.40 ± 0.32	6.9 ± 0.4 × 103
17	Nitrofurantoin	−0.255	2.21 ± 0.12	5.1 ± 0.5 × 103
18	p-Dinitrobenzene	−0.257	2.21 ± 0.35	3.1 ± 0.2 × 103
19	o-Dinitrobenzene	−0.287	0.48 ± 0.07	1.28 ± 0.2 × 103
20	4-Nitrobenzaldehyde	−0.325	0.97 ± 0.13	2.38 ± 0.4 × 103
21	3,5-Dinitrobenzoic acid	−0.344	0.09 ± 0.01	2.91 ± 0.2 × 103
22	m-Dinitrobenzene	−0.348	0.42 ± 0.06	9.6 ± 0.7 × 102
23	4-Nitroacetophenone	−0.355	0.30 ± 0.05	8.0 ± 0.67 × 102
24	CB-1954	−0.385	0.52 ± 0.05	1.75 ± 0.14 × 103
25	4-Nitrobenzyl alcohol	−0.475	0.23 ± 0.03	2.50 ± 0.16 × 102
Aromatic N-oxides
26	7-F-tirapazamine	−0.400	1.20 ± 0.11	1.80 ± 0.31 × 103
27	Tirapazamine	−0.456	0.53 ± 0.04	9.41 ± 0.82 × 102
28	7-C2H5O-tirapazamine	−0.494	0.46 ± 0.03	4.91 ± 0.32 × 103
Single-electron acceptors
29	Ferricyanide a	0.410	394 ± 19	8.8 ± 1.0 × 106
30	Fe(EDTA)−	0.120	1.2 ± 0.1	2.4 ± 0.2 × 103
31	Benzylviologen	−0.354	19.6 ± 2.3	3.6 ± 0.3 × 104

^a On the single-electron basis.

.

The log k_cat/K_m values of nitroaromatics exhibit a linear although scattered dependence on their E¹₇ (Table 1 and Figure 4). In general, the log k_cat/K_m values of quinones, including the single-electron acceptor benzylviologen, and ArN→O, are higher than those of ArNO₂, and demonstrate a parabolic dependence on their E¹₇ values (Figure 4).

It is established that FNRs from spinach, Anabaena spp. and Plasmodium falciparum reduce quinones and nitroaromatics in a single-electron way [19,30,31]. For quinone reduction by NAD(P)H-oxidizing flavoenzymes, the single-electron flux is defined as a ratio of the rate of 1,4-benzoquinone-mediated reduction of the added cytochrome c to the doubled rate of 1,4-benzoquinone-mediated NAD(P)H enzymatic oxidation at pH < 7.2 [30]. This approach is based on the fast reduction of cytochrome c by 1,4-benzosemiquinone (k ~ 10⁶ M⁻¹s⁻¹), and its slow reduction by the hydroquinone form. We found that for enzymatic reduction of 50–100 µM 1,4-benzoquinone by 50–100 µM NADPH, the single-electron flux was equal to 54 ± 4.0% of the total flux. The assessment of the single-electron flux in the reduction of aromatic nitrocompounds can be based on the ArNO₂⁻·-mediated reduction of added cytochrome c. We found that, in the presence of 50 μM NADPH and 100 μM TNT or p-nitrobenzaldehyde, the rate of RpFNR-catalyzed reduction of added 50 μΜ cytochrome c was equal to 91 ± 2.0% and 97 ± 3.0% of the doubled NADPH oxidation rate, respectively. These reactions were inhibited by 100 U/mL superoxide dismutase by 49% and 37%, respectively, which reflects the rapid reoxidation of ArNO₂⁻· with O₂ and the participation of superoxide in the reduction of cytochrome c. Thus, one may conclude that RpFNR reduces ArNO₂ in a single-electron way.

Finally, we examined the inhibition of quinone reductase reaction of RpFNR by the reaction product NADP⁺. At a fixed juglone concentration (200 μM), NADP⁺ acted as a competitive inhibitor toward NADPH (Figure 5A) with K_is = 150 ± 10 µM, as deduced from Equation (4). In turn, at a fixed concentration of 100 µM NADPH, NADP⁺ acts as an apparently noncompetitive inhibitor toward juglone (Figure 5B) with K_ii = 1.7 ± 0.1 mM, as obtained using Equation (5).

3.3. RpFNR Oxidation under Multiple Turnover Conditions

The spectral changes of RpFNR-bound FAD during its multiple turnover under aerobic conditions in the presence of NADPH and tetramethyl-1,4-benzoquinone (duroquinone) provides insight into the reoxidation mechanism of the enzyme. Duroquinone does not possess absorbance at ≥460 nm; besides, its semiquinone form is rapidly reoxidized by oxygen [28]. Control experiments were performed without the addition of a quinone, and the initial fast phase of FAD reduction by NADPH observed at 460 nm was followed by a slow reoxidation by oxygen (Figure 6A). A transient increase in absorbance at 600 nm at the same time scale accompanies this process (Figure 6A). In the presence of quinone, the reoxidation of FADH⁻ and the disappearance of the 600 nm absorbing species are accelerated by more than one order of magnitude (Figure 6B).

The maximal ΔA₄₆₀ after the enzyme mixing with NADPH (Figure 6A) corresponds to 90% of RpFNR FAD absorbance decrease after the enzyme mixing with an excess NADPH under anaerobic conditions [15,16]. Assuming that the maximal concentration of the reduced enzyme form under aerobic conditions is 90% of total enzyme, for the reoxidation of RpFNR with oxygen (Figure 6A), using Equation (6) we obtain a k_ox = 0.11 ± 0.01 s⁻¹, which was close to the steady-state NADPH oxidase activity of RpFNR. In the presence of 250 µM tetramethyl-1,4-benzoquinone, we obtain a k_ox = 2.05 ± 0.07 s⁻¹, which is close to the steady-state reduction rate of this oxidant (Table 1).

In order to characterize the reaction intermediates absorbing at 600 nm (Figure 6), the measurements were performed at different wavelengths. The results show that the absorbance initially increased in the range of 525–750 nm with λ_max ~ 720 nm (Figure 7). Subsequently, the formation and decay of a secondary flat absorbance band with a maximum at 600–700 nm took place (Figure 7). One may note that the formation of the transient species is not caused by the interaction of isoalloxazine ring of FAD and quinone, because the analogous transient absorbance spectra were obtained during the reoxidation of RpFNR with oxygen (data not shown).

4. Discussion

Due to the large differences in the structure between plant-type and TrxR-type FNRs, the main focus of this work was to disclose the possible differences and similarities in their redox properties. In addition to the mechanistic aspects of studies of the reactions of RpFNR with redox active xenobiotics, an important aspect is that R. palustris is a model microorganism of anaerobic metabolism of organic compounds [18] in which RpFNR may be involved.

In this work, the redox potentials of a representative of TrxR-type FNR were identified for the first time. The E⁰₇ of RpFNR, −0.276 V (Figure 2A), is close to the redox potential of plant-type FNR from P. falciparum, −0.280 V [32], and less negative than that of spinach FNR, −0.342 V [8]. It is currently difficult to draw conclusions about the reasons for these differences. However, we can mention some factors that may lead to a relatively high FADH· stability, 26.5% at equilibrium, which is comparable to that of semiquinone of spinach FNR, 27% at pH 7.0 [8], and Anabaena FNR, 22% at pH 8.0 [33]. RpFNR lacks the specific Anabaena FNR ion pair Ser80-Glu301 [34] (Ser96-Glu312 in spinach FNR [35]), which forms a H-bond with isoalloxazine N5 and most significantly contributes to FADH· stability. However, based on structural data of other FNRs and flavodoxins, the stability of FADH· may be enhanced by the π–π interaction of isoalloxazine with Tyr328 (Tyr98 in Desulfovibrio vulgaris flavodoxin [36], Tyr303 in Anabaena PCC7119 FNR [37]), formation of H-bonds of isoalloxazine N3 with the carboxy oxygen of Asp56 (Glu59 of Clostridium beijerinckii flavodoxin [38]), and O2 with the amide nitrogen of Ile300 (Ile356 of adrenodoxin reductase [39]). The data obtained suggest that the likely direction of electron transfer catalyzed by RpFNR is the reduction of Fd at the expense of NADPH.

The “ping-pong” mechanism of quinone reduction by RpFNR (Figure 3), pointing to the occurrence of separate reductive and oxidative half-reactions, is common to other FNRs [19,31]. As in our previous studies of Anabaena and P. falciparum FNR [19,31], no strict oxidant specificity in the reduction of quinones, aromatic nitrocompounds, and N-oxides by RpFNR can be discerned from our data except for an increase in their log k_cat/K_m with E¹₇ (Figure 4). This suggests a possible applicability of the “outer sphere” electron transfer model [40]. According to this model, the bimolecular rate constant of the electron transfer between the reactants (k₁₂) is expressed as

k₁₂ = (k₁₁ × k₂₂ × K × f)^1/2,

(7)

where k₁₁ and k₂₂ are the electron self-exchange rate constants of the reactants, K is the equilibrium constant of the reaction (log K = ΔE¹/0.059 V), and f is expressed as

log f = (log K)²/4log (k₁₁ × k₂₂/Z²),

(8)

where Z is the frequency factor, 10¹¹ M⁻¹s⁻¹ [40]. According to Equations (7) and (8), in the reaction of the electron donor with a series of homologous oxidants (which display the same k₂₂), log k₁₂ will exhibit a parabolic (square) dependence on ΔE¹ with a slope 8.45 V⁻¹ at ΔE¹ = ± 0.15 V. Because k₂₂ = ~10⁶ M⁻¹s⁻¹, characteristic of nitroaromatics, is 100-fold lower than that of quinones and aromatic N-oxides, k₂₂ = ~10⁸ M⁻¹s⁻¹, the reactivity of ArNO₂ is about 10-fold lower when compared to quinones and ArN→O of similar E¹₇ values [41,42]. In this context, it can be noted that the reactivity of RpFNR in reactions with both quinones and ArNO₂, i.e., their k_cat/K_m, is close to that previously observed in reactions with plant-type PfFNR and Anabaena FNR [19,31]. On the other hand, for reasons not yet known, AcPyP⁺ is 10 times better at oxidizing RpFNR (Figure 2A) than PfFNR [19]. Since the k_cat of reactions vary considerably (Table 1), it can be suggested that the limiting stage of the catalytic cycle is the oxidative half-reaction. The data of the Figure 5A show that the dominant mechanism of inhibition of the reaction product NADP⁺ is its competition with NADPH for binding to the oxidized form of the enzyme. The noncompetitive NADP⁺ inhibition with respect to the oxidant (Figure 5B) is slightly different from the uncompetitive inhibition of analogous Anabaena and P. falciparum FNR-catalyzed reactions [19,31]. This is most likely related to the binding of NADP⁺ to the reduced form of the enzyme with low affinity [16].

According to previous studies, photoreduced Anabaena FNR was reoxidized by quinones in two steps, FADH⁻ → FADH·, and FADH· → FAD, with the rate-limiting FADH· oxidation step [31,43]. This was evidenced by a transient formation of FADH· with 600 nm absorption. The reoxidation of RpFNR should also involve single-electron transfer steps, since it reduces quinones in a predominantly single-electron way. Because the decay of the transient 600 nm absorption and the enzyme reoxidation monitored at 460 nm proceeds with a similar rate (Figure 6B), the FADH oxidation can be a rate-limiting step in quinone reduction by RpFNR. However, the absorption characteristics of RpFNR multiple turnover intermediates (Figure 7) differ from those of FADH· formed in the absence of NADP⁺ (Figure 2B), although over time they become more similar. Absorption above 700 nm is not characteristic of FADH [24] and indicates the parallel formation of other reaction intermediates. For example, FADH⁻–NADP⁺ charge-transfer complexes absorb up to 1000 nm [44]. However, this possibility is ruled out because the spectrum of the intermediates ends at 770–800 nm (Figure 7). In addition, the FADH⁻–NADP⁺ complexes possess ε ~1.0 mM⁻¹cm⁻¹ at 610–725 nm [45]; thus, they would give about three times less increase in absorption than we see in Figure 6. Most likely, the data in Figure 7 reflect the formation of FADH·–NADP(H) complexes observed in adrenodoxin reductase, which absorb at λ > 700 nm but are only partly characterized [46,47].

The catalytic cycle of RpFNR should be characterized by significant movement of the FAD- and NADP(H)-binding domains relative to each other, including the motions of flexible C-terminal region [10,13,16,48]. Hypothetically, this could lead to shielding and deshielding of the FAD isoalloxazine ring, potentially complicating the access of oxidants. The electron exchange rate constants (k₁₁) of metalloproteins can be used to estimate the electron transfer distance (R_p) during the reactions with inorganic complexes at infinite ionic strength, where the electrostatic interactions are absent [49]

R_p (Å) = 6.3 − 0.35 ln k₁₁.

(9)

We applied this approach to the analysis of reactions of FNRs and other single-electron transferring flavoenzymes with uncharged aromatic oxidants, quinones, and nitroaromatics and obtained R_p levels of 5.0 Å (Q) and 4.4 Å (ArNO₂) for Anabaena FNR, and 4.8–5.0 Å (Q) and 4.9–5.6 Å (ArNO₂) for P. falciparum FNR [19]. However, a systematic overestimation of the electron transfer distances in this case is possible, because the dimethylbenzene part of the FAD isoalloxazine ring of the above enzymes is partly exposed to the solvent [2,5,16]. Therefore, the obtained values may be of limited usefulness only in approximately assessing the “intrinsic” flavoenzyme reactivity. For the reactions of RpFNR with Q and ArNO₂, the approximate k₁₁ values may be obtained from the data of Figure 3 at ΔE¹₇ = 0, where k₁₂ = (k₁₁ × k₂₂)^1/2. At E¹₇ of the oxidant being equal to −0.285 V, the log k₁₁ values are equal to 1.36 ± 0.46 (quinones) and 1.04 ± 0.22 (nitroaromatics), which then gives R_p = 5.2 ± 0.4 Å and R_p = 5.4 ± 0.2 Å, respectively, according to Equation (9). It can be noted that the possible uncertainty of the E-FADH· potential in the range of 10–15 mV has almost no effect on the R_p value, changing it by only 0.1 Å. Thus, these R_ps are close to the R_p values for plant-type FNRs given above, indicating that possible steric interferences in the structure of RpFNR may not affect the low M_r oxidant reduction rate. These data will be valuable in our further studies focusing on the specific features of the interaction of RpFNR with its ferredoxin-type redox partners.

5. Conclusions

Despite structural differences, many of the redox properties of TrxR-type RpFNR redox are similar to those of plant-type FNR: (i) the standard redox potential of FAD and its neutral semiquinone stability, (ii) single-electron reduction of quinones and nitroaromatic compounds, their reactivity and its dependence on single-electron reduction potential, (iii) transient FAD semiquinone formation, and (iv) calculated electron transfer distances. The slight differences in the action of plant-type and RpFNR are manifested through the much faster reduction of AcPyP⁺ by RpFNR and the different mode of its inhibition by the reaction product NADP⁺. These data may be useful in further studies of the specific interaction of RpFNR with its ferredoxin-type redox partners.