Ferredoxin-NADP+ Oxidoreductase and Flavodoxin

Ferredoxin-NADP+ Oxidoreductase and Flavodoxin: Comparison

Please note this is a comparison between Version 1 by Takashi Iyanagi and Version 2 by Vivi Li.

Distinct isoforms of FAD-containing ferredoxin-NADP⁺ oxidoreductase (FNR) and ferredoxin (Fd) are involved in photosynthetic and non-photosynthetic electron transfer systems. The FNR (FAD)-Fd [2Fe-2S] redox pair complex switches between one- and two-electron transfer reactions in steps involving FAD semiquinone intermediates. In cyanobacteria and some algae, one-electron carrier Fd serves as a substitute for low-potential FMN-containing flavodoxin (Fld) during growth under low-iron conditions. This complex evolves into the covalent FNR (FAD)-Fld (FMN) pair, which participates in a wide variety of NAD(P)H-dependent metabolic pathways as an electron donor, including bacterial sulfite reductase, cytochrome P450 BM3, plant or mammalian cytochrome P450 reductase and nitric oxide synthase isoforms. These electron transfer systems share the conserved Ser-Glu/Asp pair in the active site of the FAD module. In addition to physiological electron acceptors, the NAD(P)H-dependent diflavin reductase family catalyzes a one-electron reduction of artificial electron acceptors such as quinone-containing anticancer drugs. Conversely, NAD(P)H: quinone oxidoreductase (NQO1), which shares a Fld-like active site, functions as a typical two-electron transfer antioxidant enzyme, and the NQO1 and UDP-glucuronosyltransfease/sulfotransferase pairs function as an antioxidant detoxification system.

ferredoxin-NADP+ oxidoreductase (FNR)
ferredoxin (Fd)
flavodoxin (Fld)
diflavin reductase family
catalytic cycle
electron transfer
redox potentials
evolutionary aspects

1. Introduction

Land plant and cyanobacterium ferredoxin (flavodoxin)-NADP⁺ oxidoreductases (FNRs) catalyze the reversible electron transfer that occurs in the photosynthetic reaction (formation of NADPH) and non-photosynthetic reactions (NADPH-dependent redox metabolic pathways) (Equation (1)) ^{[1][2][3][4][5][6]}[1,2,3,4,5,6]. The forward and reverse reactions of Equation (1) are catalyzed by distinct FNR and ferredoxin (Fd) isoforms [7]:

2Fd (Fe

²⁺

) + NADP

⁺

+ H

⁺

⇌ 2Fd (Fe

³⁺) + NADPH

) + NADPH

(1)

In cyanobacteria and some algae, FMN-containing flavodoxin (Fld) acts as a one-electron carrier instead of ferredoxin (Fd) under iron-limiting conditions [8]. Meanwhile, the FNR- and Fld-like domains are present in eukaryotic NAD(P)H-dependent enzymes, including cytochrome P450 reductase (cyt P450 reductase) ^[9][10][9,10] and nitric oxide synthase (NOS) isoforms [11]. The catalytic domains of these enzymes share a high sequence similarity and structure with FNR and Fld, where the FAD-FMN redox pair donates electrons to final electron acceptors during sequential one-electron transfer reactions [12]. NADH-cytochrome b₅ reductase (cyt b₅ reductase) shares structural similarities with plant FNR, despite its amino acid sequence similarities being lower [13]. The cyt b₅ reductase-cyt b₅ complex pair donates an electron to metal-containing proteins [14]. Cyt P450 reductase (FAD-FMN) and the cyt b₅ reductase (FAD)-cyt b₅ complex exhibit diverse functions in regard to the electron acceptors.

In addition to plant FNR, cyt P450 reductase and NOS isoforms catalyze a one-electron reduction of artificial electron acceptors such as quinone derivatives, where the resulting quinone radical reacts with molecular oxygen to form the superoxide radical ^{[15][16][17][18][19]}[15,16,17,18,19], while NAD(P)H: quinone oxidoreductase (NQO1) bears a common flavodoxin-like topology in its active site [20] and catalyzes the two-electron reduction of anticancer quinone derivatives; the resulting hydroquinone form is primarily conjugated by UDP-glucuronosyltransferase (UGT) and sulfotransferase (SULT) [21]. Thus, the NQO1-UGT/SULT pair systems function as potent antioxidant enzyme systems.

2. Structure and Properties of the FNR, Fd and Fld

The three-dimensional structure of FNR from spinach reveals two subdomains, the NADPH-binding domain and the FAD-binding domain ^[1][5][1,5]. FNR can transfer electrons to both Fd [2Fe-2S] and Fld (FMN), where an interaction with Fd or Fld occurs in the same structural region of the FAD-binding domain ^[1][22][1,22]. Both Fd and Fld act as one-electron acceptors or donors, and Fd is replaced with Fld under low-iron conditions ^[8][23][8,23]. The rate of electron transfer among these proteins is controlled by several factors, including the relative orientation and distance between protein–protein interactions and the differences in the redox potentials between the protein-bound donor and acceptor redox cofactors ^{[1][3][23][24][25][26]}[1,3,23,24,25,26]. Sinohara et al. [1] reported the crystal structures of the RFNR-RFd (Fd III) (R for root) and LFNR-LFd (Fd I) (L for leaf) complexes, and this provides a structural basis for reversing the redox pathway. In the LFNR-Fd complexes, the distance between the [2Fe-2S] cluster of Fd and the dimethylbenzene edge of the FAD ring is ~5.5–6.0 Å ^[1][24][1,24], while the RFNR-Fd complexes can utilize the different sides of the [2Fe-2S] cluster for intermolecular electron transfer. The modeling of the FNR-Fld complexes indicates that the distance between the two flavin rings is ~4.1 Å and that no intervening residues are present between the two cofactors, thus making possible direct one-electron transfer [25]. Taken together, these observations suggest that the complexes between oxidized FNR and oxidized Fd are relatively stable, as the complexes are stabilized by a salt bridge of FNR33Lys-Fd60Asp between FNR and Fd [26]. However, this stability is decreased by NADP(H) binding, and unstable complexes can promote the electron transfer rate during the catalytic cycle [27]. FNR interacts with Fd or Fld in photosynthetic and non-photosynthetic reactions. The isoalloxazine ring of FAD in the leaf and root FNR is sandwiched between two aromatic residues (Tyr314 and Tyr95), where the phenol ring of carboxyl(C)-terminal Tyr314 shields the face of the isoalloxazine ring of FAD ^[5][28][5,28]. Thus, C-terminal Tyr314 must move away for productive hydride transfer from reduced FAD to NADP⁺, and Ser96 interacts with the N5 atom of the isoalloxazine ring of FAD, where Ser96 forms a hydrogen bond with Glu312. Additionally, C-terminal Tyr314 and the Ser96-Glu312 pair modulates the affinity for NADP⁺, the stabilization of FAD semiquinone and the rate of electron transfer [29]. Therefore, Tyr314 and the Ser96-Glu312 pair are involved in the modulating of the flavin redox properties ^[30][31][30,31]. The formation of the reduced FAD-NADP⁺ (FADH⁻-NADP⁺) complex with the charge transfer bands of 500–750 nm proceeds via an FAD_ox-NADPH charge-transfer species [32]. In the presence of excess NADPH (>10-fold), bound NADP⁺ is replaced by NADPH: FADH⁻-NADP⁺ + NADPH

⇌

FADH⁻-NADPH + NADP⁺, where the FADH⁻- NADPH complex does not exhibit significant charge transfer bands. This could be caused by a decrease in π−π stacking interactions between reduced FAD and NADPH. The value of E_ox/red measured using dithionite as a reductant is −377 mV [32], which is fitted to a Nernstian n = 2 curve. However, the two-electron reduction potential in the presence of NADP⁺ is divided into E_ox/sq (−306 mV) and E_sq/red (−386 mV) (Table 1), suggesting that the redox potentials of FNR are regulated by the NADP⁺/NADPH ratio.

Table 1. Midpoint reduction potentials of FAD and FMN domains.

Enzyme Specicies	E_ox/sq	(mV)	E_sq/red	(mV)	pH

] and the reductase domain (FAD-FMN) of NOS isoforms ^[41][42][41,42] are different from those of plant FNR (FAD) [34] and Anabaena flavodoxin (FMN) [43], but the one-electron redox potentials of FAD and FMN always satisfy E_ox/sq > E_sq/red (Ref. [14] and Table 1). The kinetic parameters for Fd reduction by FNR using the NADPH-dependent cyt c reductase assay (NADPH → FNR → Fd/Fld → cyt c) were reported [29]. The values for pea FNR are in the range of those of the Anabaena enzyme, with a slightly larger k_cat and a smaller K_m (Table 2) [29], while Kimata-Ariga and co-workers recently reported the tissue-specific parameters, where maize RFNR has a higher K_m value for Fd I than for Fd III [1]. In addition, the orientation of the RFNR-Fd complex remarkably varies from that of the LFNR-Fd complex [1], suggesting that the root and leaf complexes utilize the different sides of the [2Fe-2S] cluster for the intermolecular electron transfer, which might lead to the evolutional switch between photosynthetic and heterotrophic assimilation.

Table 2.

Kinetic parameter of

Pea

and

Anabaena

FNR in NADPH-dependent cytochrome

reductase activity.

FNR Species	$K_{c a t}^{Fd} (s^{- 1})$	$K_{m}^{Fd}$	(µM)	$K_{c a t}^{Fd} / K_{m}^{Fd}$	(µM s	⁻¹	)	$K_{c a t}^{Fld} (s^{- 1})$	$K_{m}^{Fld}$	(µM)	$K_{c a t}^{Fld} / K_{m}^{Fld}$
FNR (FAD)
Pea	FNR	139	6.5	21.3	30.6	16.7	1.8
Spinach (FAD)	−350	−335	7
Spinach (FAD)	Anabaena	FNR	200	11	18.2	23.3	33	0.7	−402	−358	8
Spinach (FAD)	−306	−386	8
(in the presence of NADP				⁺				)
Anabaena	(FAD)	−385	−371	8
Anabaena	Flavodoxin (FMN)	−212	−436	7
Cyt P450 reductase (FAD-FMN)
Rabbit (FAD)	−290	−365	7
Rabbit (FMN)	−110	−270	7
Human (FAD)	−283	−382	7
Human (FMN)	−56	−274	7
nNOS reductase domain (FAD-FMN)
Rat (FAD)	−260	−280	7.6
Rat (FMN)	−50	−276	7.6
iNOSreductase domain (FAD-FMN)
Human (FAD)	−240	−270	7
Human (FMN)	−105	−245	7

The redox potentials for LFd I and RFd III are −401 mV and −321 mV, respectively. [33]. Thus, the one- and two-electron redox potentials of FNR are important factors that control the electron transfer rate (Scheme 1). For the spinach LFNR, the E_m (FAD/FADH⁻) value is −343 mV. Regarding the one-electron redox potential, the oxidized-semiquinone couple (E_ox/sq) is −350 mV, and the semiquinone-fully reduced couple (E_sq/red) is −335 mV at pH 7.0 [34] (Table 1). Thus, the electron transfer from Fd I to NADP⁺ is thermodynamically favorable. On the other hand, the electron transfer from NADPH to Fd III is relatively favorable. In both systems, tissue-specific FNR mediates reversible electron transfer reactions (Equation (1) and Scheme 1) [33].

Scheme 1.

Switching between one-electron and two-electron transfer reactions in the FNR-Fd I/Fld and FNR-Fd III/Fld systems. H

⁻

T, hydride transfer.

Fld is a small electron carrier that participates in low-redox-potential electron transfer pathways, which are classified into three groups based on the presence of tryptophan (Trp) and tyrosine (Tyr) near the isoalloxazine ring of FMN [35]. In all Flds, the semiquinone formation constant Ks value is larger than unity, indicating that the E_ox/sq value is always more positive than the E_sq/red value [36]. The E_ox/sq value is associated with the proton-coupled one-electron reduction from FMN to FMNH^•, while E_sq/red is associated with the one-electron reduction from FMNH^• to FMNH⁻. The E_ox/sq value is shifted from ~−240 mV to less-than-negative values, while E_sq/red is shifted from ~−170 mV to ~−540 mV, depending on the pH and the species of Fld. The neutral semiquinone, FMNH^•, is stabilized by the hydrogen bond of the carbonyl and amide groups of the protein backbone with the N5 atom of the isoalloxazine ring of the flavins. Its reactivity is lower than that of the fully reduced form, where the FMNH^•/FMNH⁻ couple acts as a one-electron carrier [36]. Flds possess a unique structure, and the loop structures in the FMN environment play an important role in electron transfer reactions [37]. The FMN binding sites in the bacterial Fld and eukaryotic diflavin reductase family possess similar loop structures and are sandwiched between two aromatic amino acid residues. Rwere et al. [38] reported that the length and sequence of the “140 s” FMN binding loop of cyt P450 reductase functions as a key determinant of its redox potential and activity with cyt P450s. The one-electron redox potentials of cyt P450 reductase (FAD-FMN) ^[39][40][39,40

3. Catalytic Cycle of the FNR-Fd and FNR-Fld Systems

3.1. Catalytic Cycle of the FNR-Fd System

As already mentioned, the distinct isoforms of FNR and Fd are expressed in the different tissues, where LFNR and LFd I are expressed in photosynthetic tissue, while RFNR and RFd III are expressed in non-photosynthetic tissues. FNR mediates a switching between one-electron and two-electron transfer reactions (Scheme 1). Thus, the tissue-specific expression of the FNR and Fd isoforms indicates that different FNR and Fd isoforms participate in the different catalytic electron transfer processes (see Figure 1A,B). LFNR catalyzes the reduction of NADP⁺ to NADPH during the photosynthetic process in plants and cyanobacteria, where the two electrons from photosystem I are transferred from FADH⁻ to NADP⁺ in a process defined by: photosystem I (PSI) → Fd I [2Fe-2S] (−401 mV) → LFNR (FAD) → NADP⁺ (−320 mV). In non-photosynthetic reactions, NADPH generated by the photosynthetic reaction and the oxidative pentose phosphate cycle donates electrons to RFd: NADPH (−320 mV) → RFNR (FAD) → Fd III (−321 mV) [33]. Thus, the photosynthetic process is more favorable than non-photosynthetic reactions. Both mechanisms are controlled by several factors: the redox potentials, the association/dissociation of complexes, and the distance between Fd [2Fe-2S] and FNR (FAD) in protein–protein interactions.

In photosystem I, the electron transfer sequence of the physiological reaction center is {P700* → A_o → A₁ → FeS-X → [FeS-A → FeS-B]} → Fd I, in which Fd I accepts an electron from the [FeS-B] site (−580 mV), and the resulting reduced Fd I (−401 mV) binds to the oxidized FNR (FAD_ox) [44], as shown in Figure 1A. Thus, Fd I functions as a mobile electron carrier. On the other hand, non-photosynthetic reactions begin from NADPH binding to the FAD_ox-Fd III_ox complex: NADPH → RFNR (FAD)-RFd III, as shown in Figure 1B.

In 1984, Batie and Kamin [45] proposed the catalytic cycle of the LFNR and LFd I system (Figure 1A). This cycle involves basic electron transfer reactions, including one-electron transfer (1e⁻T), proton-coupled one-electron transfer (PC1e⁻T) and two-electron hydride transfer (H⁻T) reactions. The catalytic cycle includes steps 1–7, where the catalytic cycle begins from LFNR (FAD_ox) (step 1). The oxidized Fd I accepts an electron from Photosystem I, and the resulting Fd I_red binds to the LFNR (FAD_ox)-NADP⁺ complex (step 2). One electron is then transferred from Fd I_red to LFNR (FAD_ox). In step 3, the one-electron transfer from reduced Fd I to oxidized FAD is coupled with proton transfer (PC1e⁻T). Electron transfer in this step is significantly decreased in D₂O solution. In step 4, the Fd I_ox of the LFNR (FADH^•)-NADP⁺ complex is replaced by reduced Fd I_red, thus resulting in the Fd I_red-LFNR (FADH^•)-NADP⁺ complex. In step 5, the Fd_ox-LFNR(FADH⁻)-NADP⁺ complex is formed via a one-electron transfer reaction (1e⁻T) from Fd I_red to LFNR (FADH^•). In step 6, NADP⁺ accepts a hydride from FADH⁻ to form NADPH. In the final step, both Fd I_ox and NADPH are released; a new cycle then begins. This catalytic cycle is also supported by kinetic studies [46].

In 2003, Carrillo and Ceccarelli [47] proposed the “open questions” in regard to the Batie-Kamin catalytic model [45] on the basis of kinetics and binding experiments on spinach FNR. Currently, new data in regard to these open questions can be found (see Refs ^{[1][48][49][50]}[1,48,49,50]).

In non-photosynthetic reactions (metabolic redox pathways) (Figure 1B), the reaction begins with RFNR (FAD_ox). In the first step, NADPH binds to the FAD_ox-Fd III_ox complex. In step 2, the complex NADP⁺-FADH⁻-Fd III_ox is formed via the hydride transfer (H⁻T) reaction, and this complex is followed by the formation of the NADP⁺-FNR (FADH^•)-Fd III_red complex via one-electron transfer (1e⁻T) from FADH⁻ to Fd III_ox (step 3). In step 4, Fd III_red is released from the complex, and Fd III_ox binds. In step 5, the NADP⁺-FNR (FAD_ox)-Fd III_red complex is formed via PC1e⁻T. In the final step, Fd III_red and NADP⁺ are released. Thus, reduced Fd III is formed in steps 3 and 5.

Figure 1. Proposed catalytic cycles of the LFNR (A) and RFNR (B) systems. (A) Photosystem I → LFd I (2S-2Fe) → LFNR (FAD) → NADP⁺ [45] and (B) NADPH→RFNR (FAD) → RFd III→metabolic pathways. H⁻T, hydride transfer; 1e⁻T, one-electron transfer; PC1e⁻T, proton-coupled one-electron transfer.

3.2. Catalytic Cycle of the FNR-Fld Systems

Fd can be replaced with FMN-containing Fld, where the FNR-Fd pair is replaced with the FNR-Fld pair in the catalytic cycle (Figure 2A). The FMN semiquinone (FMNH^•) is highly stable. Thus, Fld functions as a one-electron carrier (FMNH⁻

⇌

FMNH^• + e⁻). In the photosynthetic process, the catalytic cycle (Figure 2A) is similar to that presented in Figure 1A.

Figure 2. Proposed catalytic cycle of the FNR-Fld systems. (A) Photosystem I → Fld (FMN) → FNR (FAD) → NADP⁺ and (B) NADPH → FNR (FAD) → Fld (FMN) → metabolic redox pathways. In (B), the overall reaction is based on the catalytic cycle.

In the first step, NADP⁺ binds to LFNR (FAD_ox). In step 2, reduced Fld (FMNH⁻) binds to FNR (FAD_ox)-NADP⁺, and in step 3, the diflavin radical intermediate FMNH^•-FADH^•-NADP⁺ is formed via a PC1e⁻T reaction. Such intermediates are observed in the Anabaena PSI/Fld system [51]. In step 4, Fld (FMNH^•) releases, and Fld (FMNH⁻) binds. In step 5, FMNH^•-FADH⁻-NADP⁺ is formed during a 1e⁻T reaction, and in step 6, FMNH^•-FAD_ox-NADPH is formed during an H⁻T reaction. In the final step, Fld (FMNH^•) and NADPH are released.

The catalytic cycle of the non-photosystem includes the priming reaction (Figure 2B). The catalytic cycle begins from the complex NADP⁺-FAD_ox-FMNH^• and includes the following sequences: (6) → (7) → (8) → (9) → (4) → (5), where the catalytic cycle shuttles among the 1-3-2-1 electron reduced states. The two electrons from steps 5 and 9 are transferred to the metabolic redox pathways. This overall catalytic cycle is very similar to that of cyt P450 reductase (see Figure 3), where the diflavin radical intermediate, FADH^•-FMNH^•, is formed through the catalytic cycle (step 9).

Figure 3. Proposed catalytic cycle of the eukaryotic-membrane-bound cyt P450 reductase-cyt P450 system. In this figure, the overall reaction is based on the catalytic cycle. Note that cyt P450R indicates cyt P450 reductase. H⁻T, hydride transfer; 1e⁻T, one-electron transfer; PC1e⁻ T, proton-coupled one-electron transfer.

Recently, Utschig et al. [52] proposed a new approach for the electron transfer from photosystem I to Fd I/Fld in which Fd I and Fld are covalently bound by a ruthenium photosensitizer (RuPS) instead of photosystem I. The RuPS is activated by light, whereby an electron is transferred from activated RuPS*: RuPS*-Fd I → FNR and RuPS*-Fld → FNR, respectively. In these cases, an electron transfer occurs within the Fd I/Fld and FNR complexes. In both cases, FNR semiquinone intermediates are observed using electron-spin resonance. In the final step, oxidized NADP⁺ accepts hydride from FADH⁻. This report suggests that photosystem I could sequentially donate one electron to the oxidized Fd I/Fld-oxidized FNR complexes. The electron transfer mechanisms are different from those of Figure 1A and Figure 2A, but this new approach could provide an excellent model system for photosystem I.

4. Structure and Properties of Diflavin Reductase Family

The electron transfer cascade catalyzed by FNR is arranged as presented in Scheme 2. FNR selects a one-electron carrier as a partner of electron transfer systems in which three components, iron sulfur protein ferredoxin, FMN-containing flavodoxin and heme-containing cytochromes, function as a one-electron carrier. Thus, at least three different electron transfer systems might be diversified during the processes of evolution.

Scheme 2.

The roles of FNR (FAD) and one-electron carriers in the NAD(P)H-dependent electron transfer systems.

The eukaryotic FAD- and FMN-containing diflavin reductase family members are fusion enzymes in which the structure of the FAD/NADPH-binding C-terminal domain is structurally homologous to that of plant FNR, while the FMN-binding N-terminal domain is similar to that of bacterial FMN-containing flavodoxin (Fld). Furthermore, cyt P450 reductase ^[40][53][40,53], methionine synthase (MS) reductase (MS reductase) ^[54][55][54,55], novel reductase 1 (NR1) [56], sulfite reductase (SiR) [57] and P450BM3 [58] are members of the diflavin oxidoreductase family, where the distinct FAD and FMN domains are connected by an additional connecting domain (CD) and flexible hinge (H) [59].

The catalytic cycle of cyt P450 reductase as a prototypic member of the diflavin reductase family is very similar to that of the catalytic cycle of the non-photosystem FNR-Fld system (Figure 3). The catalytic cycle of the membrane-bound type is closely related to that presented in Figure 2B. However, the intermolecular electron transfer between FAD and FMN occurs in the closed form, while its open form donates electrons to cyt P450s in a stepwise manner [12]. On one hand, cyt P450 catalyzes the consumption of one molecule of oxygen/molecule of substrate (RH); one atom of this oxygen molecule inserts into the product (ROH), and the other undergoes two equivalents of reduction: RH + 2e⁻ + O₂ + 2H⁺ → ROH + H₂O. On the other hand, methionine synthase (MS) reductase [55] and novel reductase 1 (NR1) [56] are soluble forms that lack the N-terminal anchors domain. For these enzymes, the catalytic cycle is similar to that of the membrane-bound form [14].

In contrast, the reductase domain of NOS isoforms contains additional regulatory elements within the C-terminal reductase domain that control the electron transfer through Ca²⁺-dependent calmodulin (CaM) binding ^[12][14][60][12,14,60]. The mechanism of FAD reduction by NADPH is closely related to that of mammalian cyt P450 reductase. However, the neuronal (nNOS) and endotherial (eNOS) isoforms are activated by Ca²⁺/CaM binding [60]. On the other hand, the inducible (iNOS) isoform tightly binds Ca²⁺/CaM, and iNOS activity is independent of intracellular Ca²⁺ concentrations. Thus, the iNOS isoform is likely to have functions and mechanism similar to those of cyt 450 reductase [60].