cytochrome P450 reductase 2 (ATR2) - Wiley Online Library

Structure of the Arabidopsis thaliana NADPH-cytochrome P450 reductase 2 (ATR2) provides insight into its function Guoqi Niu1, Shun Zhao2,3, Lei Wang1, Wei Dong2,3, Lin Liu2 and Yikun He1 1 College of Life Sciences, Capital Normal University, Beijing, China 2 Key Laboratory of Photobiology, CAS Center for Excellence in Molecular Plant Sciences, Institute of Botany, Chinese Academy of Sciences, Beijing, China 3 University of Chinese Academy of Sciences, Beijing, China

Keywords conformational change; crystallography; electron transfer; flavoprotein; reduction/ oxidation Correspondence Y. He, College of Life Sciences, Capital Normal University, 105 Xisanhuan North Road, Beijing 100048, China Fax: +86-10-68903089 Tel: +86-10-68902345 E-mail: [email protected] or L. Liu, Key Laboratory of Photobiology, CAS Center for Excellence in Molecular Plant Sciences, Institute of Botany, Chinese Academy of Sciences, 20 Nanxincun, Xiangshan, Beijing 100093, China Fax/Tel: +86-10-62836483 E-mail: [email protected] (Received 4 November 2016, revised 10 January 2017, accepted 16 January 2017)

Members of the cytochrome P450 family catalyze a variety of mono-oxygenase reactions, and for the eukaryotic membrane-bound members, NADPH is typically used as the reducing agent. The flavoprotein NADPH-cytochrome P450 reductase (CPR) enables electron transfer from NADPH to cytochrome P450 via its flavin cofactors. ATR2 is one of the two authentic CPR genes in the genome of the model plant Arabidopsis thaliana, and its product has been physiologically and kinetically char structure of Arabidopsis thaliana acterized. Here, we report the 2.3 A NADPH-cytochrome P450 reductase 2 (ATR2) and find that the position of the two flavin cofactors differs from that of other known CPR structures. Mutation of residues related to possible interflavin electron transfer retains the reductase activity of ATR2, which suggests a direct electron transfer pathway between the flavins. In contrast, mutation of a single residue (R708) mediating interdomain interaction abolishes this activity. Because this residue is only conserved in plant CPRs, we speculate a plantspecific working mechanism as observed in ATR2. Database Atomic coordinates and structure factors of ATR2 are available in the Protein Data Bank under the accession code 5GXU.

doi:10.1111/febs.14017

Introduction NADPH-cytochrome P450 reductase (CPR), a prototypic diflavin reductase, supplies electrons from the reduced pyridine nucleotide of NADPH to the monooxygenase cytochrome P450 as well as heme oxygenase and the electron transport hemoprotein cytochrome b5 [1,2]. CPR is suggested as a fusion of two ancestral proteins, an FMN-containing flavodoxin and an FAD-containing reductase, the latter

represented by ferredoxin-NAD(P)+ reductase (FNR) and NAD(P)H-cytochrome b5 reductase (CYB5R) [3,4]. Homologs of CPR include the reductase portion of nitric oxide synthase [5], the flavocytochrome P450 from Bacillus megaterium (BM3) [6], the bacterial sulfite reductase [7], the mammalian methionine synthase reductase [8], and the novel reductase 1 cytoplasmic protein NR1 [9].

Abbreviations ATR, Arabidopsis thaliana NADPH-cytochrome P450 reductase; BM3, Bacillus megaterium flavocytochrome P450 BM3; CPR, cytochrome P450 reductase; CYB5R, NAD(P)H-cytochrome b5 reductase; FNR, ferredoxin-NAD(P)+ reductase; hCPR, human CPR; IPTG, isopropyl b-Dthiogalactoside; MBP, maltose-binding protein; rCPR, rat liver CPR; TEV, tobacco etch virus; yCPR, yeast CPR.

754

The FEBS Journal 284 (2017) 754–765 ª 2017 Federation of European Biochemical Societies

G. Niu et al.

Cytochrome P450 reductase is divided into four domains: an N-terminal FMN-binding domain, a connecting domain, an FAD-binding domain, and a C-terminal NADPH-binding domain [10]. The consecutive FAD-binding and NADPH-binding domains are also referred to as the FNR-like domain. Biochemical and crystallographic analysis confirmed the functional and structural similarity of the FMN-binding domain to flavodoxin, and of the FNR-like domain to FNR or CYB5R [11]. The connecting domain, necessary to maintain the orientation of the FMN-binding domain and the FNR-like domain, possesses well-defined secondary structures and shares no similarity with other known proteins. The CPR-catalyzed reduction includes multistep electron transfer, first from reduced NADPH to FAD, then to FMN, and finally to the ultimate electron acceptor [12]. The first reported CPR structure, that of rat liver CPR (rCPR) [10], and the structures of yeast CPR (yCPR) and of human CPR (hCPR) [13,14], show a ‘closed’ state in which internal electron transfer is facilitated by the proximity of the cofactors [15,16]. The orientation of FMN and FAD is compatible with a rapid electron transfer route. A hinge region links the N-terminal FMN-binding domain to the rest of CPR. Mutant rCPR with a fourresidue deletion in the hinge region gets locked into an ‘open’ state, in which the FMN-binding domain extends away from the FNR-like domain [17]. This mutant rCPR forms a stable complex with a heme-containing heme oxygenase [18]. A yeast-human hybrid CPR, whose FMN-binding domain is from yCPR and the rest sequence is from hCPR, is also observed in a widely open conformation [19]. An engineered disulfide linkage that bridges the FMN-binding and FAD-binding domains exhibits a significant decrease in the rate of electron transfer [20]. These findings indicate that the FMN-binding domain undergoes a large conformational change during the catalytic process, and that the ‘closed-open’ state transition is necessary for CPR’s activity [21]. In contrast to the massive P450 genes within plant genome, only a limited number of CPR genes exist [22‒24]. The model plant A. thaliana has two authentic CPRs, A. thaliana NADPH-cytochrome P450 reductase 1 (ATR1) and ATR2 [25‒27], and one hypothetic ATR3. ATR3 encodes a diflavin reductase essential for embryogenesis and its sequence is close to that of NR1 [28]. ATR2 shares ~ 40% identity with rCPR and yCPR, and ~ 30% identity with the reductase portion of BM3 (Fig. 1). Kinetic analysis of electron flux in CPRs reveals that the rate-determining steps in plant and mammalian enzymes are different [29]. However,

Structure of plant NADPH-cytochrome P450 reductase

the structural basis for these differences remains unclear. Here, we report the crystal structure of ATR2 at resolution. The FMN-binding domain is located 2.3 A in close contact with the connecting domain and the NADPH-binding domain, in a position between the closed and open states. The distance between the flavin and biochemical rings of FMN and FAD is 23 A, assay indicated that the electron transfer in ATR2 is direct between FAD and FMN. The interaction between the FMN-binding domain and the NADPHbinding domain is mediated by a salt bridge of E186 with R708. R708 adopts alternative conformations, which could act as a gating residue that plays critical roles in the closed-open cycle. Notably, R708 is only conserved in plant and algae CPRs but not in the mammalian or yeast ones. This suggests a specific regulation of CPR for the plants.

Results Overall structure resoThe structure of ATR2 was determined at 2.3 A lution (Table 1). Two forms of ATR2 exist in the crystal. Form A (chain A) has the electron density for all four domains except a 43-residue fragment of the hinge region; Form B (chain B) has the electron density for three domains, but lacks that for the FMNbinding domain (Fig. 2A). The structure of ATR2 is described based on Form A unless noted otherwise. The overall structure is similar to the known structures of CPR in a closed state (Fig. 2B), including rCPR [10] and yCPR [13], as well as the reductase portion of BM3 [30]. This suggests that ATR2 is probably in the closed state where the internal electron transfer is facilitated. Remarkably, the FMN-binding domain in Form A rotates away from the other three domains compared with the known CPRs in the closed state, resulting in a ‘half-closed’ state (Fig. 2C). Cofactors FMN and FAD The electron densities for the two electron carriers FMN and FAD are well defined (Fig. 3A). FMN is bound in a similar way as in other CPRs, as well as the bacterial FMN-containing flavodoxin [31] (Fig. 3B, C). The carbonyl oxygen atom of G169 in the vicinity of the flavin ring points away and adopts an ‘O-down’ conformation (Fig. 3D). Y168 stacks with the flavin ring in a tilted way, and forms hydrogen bond with the phosphate moiety of FMN. These two structural characteristics indicate that the FMN observed here is


755


G. Niu et al.

Fig. 1. Sequence alignment. The amino-acid sequences of CPR from four species are aligned, including ATR2 from Arabidopsis thaliana, rCPR from Rattus norvegicus, yCPR from Saccharomyces cerevisiae, and the reductase portion of BM3. The FMN-binding residues are indicated by solid circles; the FAD-binding residues are indicated by solid triangles; the NADPH-binding residues as observed for rCPR are indicated by solid squares. The four domains are indicated by lines above the sequences and colored as follows: the FMN-binding domain (residues: 103–262) is in bright blue, the connecting domain (residues: 357–485) is in cyan, the FAD-binding domain (residues: 304–355, 486–549) is in bright green, and the NADPH-binding domain (residues: 550–711) is in pink. The figure was generated with ESPript (http:// espript.ibcp.fr/ESPript/ESPript/).

in an oxidized state as recently shown in the high-resolution structure of rCPR [32]. The cofactor FAD is in an extended conformation (Fig. 3E). The flavin mononucleotide moiety is anchored by hydrogen bonds with the backbone amide or carbonyl groups of F490, S492, T507, A509, T524, and C525, with the hydroxyl groups of Y491 and S526, and with two water molecules. The isoalloxazine ring is stabilized by a parallel p–p interaction with the indole ring of the C-terminal conserved tryptophan (W711 in ATR2), and an almost perpendicular p–p interaction with the phenyl ring of Y491 as in other CPR structures. W711 blocks the interaction between the nicotinamide ring of NADPH and the flavin ring of FAD, which suggests that ATR2 is not in the state of receiving hydride from NADPH as revealed by the structural studies of rCPR-NADPH and FNR-NADPH complex [33,34]. The adenine moiety of FAD is bound by multiple interactions (Fig. 3E). The phosphate group is saltbridged with the side chains of R489 and H521. The

756

former is conserved in all CPRs and the latter is only conserved in plant’s CPRs. The adenine ring forms three hydrogen bonds, with the backbone carbonyl groups of L414 and Q458, and with the hydroxyl group of S416. These three residues are all from the connecting domain. This mode of adenine binding is different from those in rCPR, hCPR, and BM3 (Fig. 3F). The side chain of a tyrosine (Y478 in rCPR and Y481 in hCPR) stabilizes the adenine ring and therefore FAD is bound exclusively by the FAD-binding domain. A second FMN-binding site is located between the FMN-binding and connecting domains of yCPR, where the adenine ring of FAD is positioned between those in ATR2 and in rCPR/hCPR [13]. Interflavin electron transfer The edge-to-edge distance between FAD and FMN is (Fig. 4A), which is much longer than that in 10.6 A rCPR or yCPR. The distance between the N5 atoms of


G. Niu et al.


Table 1. Data collection and refinement statistics. Data collection Space group Resolution ( A)* Unit cell dimensions a, b, c ( A) a, b, c (°) No. of measured reflections No. of unique reflections Redundancy Completeness (%) I/rI Rmergea Refinement statistics Resolution ( A) Rworkb/Rfreec(%) Number of molecules Number of atoms Protein Ligand Water Average B value ( A2) Protein Ligand Water Wilson B factor R.m.s deviations Bond lengths ( A) Bond angles (°) Ramachandran plot Most favored (%) Additional allowed (%) Disallowed (%)

P1 50–2.30 (2.38–2.30) 55.3, 61.8, 88.3 97.8, 100.5, 90.3 92 112 48 501 1.9 (2.0) 96.7 (97.3) 10.9 (2.0) 0.073 (0.452) 30.6–2.3 20.0/24.9 2 7268 137 375 26.0 21.3 25.3 31.73 0.006 1.013 96.7 3.2 0.1

*Highest resolution shell is shown in parenthesis. Rmerge = ΣhklΣi|Ii(hkl ) |/ΣhklΣiIi(hkl ), where Ii(hkl ) is the i th observation of reflection hkl and is the weighted intensity for all observations i of reflection hkl. b Rwork = Σ||Fo| |Fc||/Σ|Fo|, where Fo and Fc are the observed and calculated structure factors, respectively. c Rfree is the cross-validated R-factor computed for a test set of 5% of the reflections, which were omitted during refinement. a

Resithe two flavin cofactors in ATR2 is about 23 A. dues including Q112, Y168, D175, D666, and K668 are positioned between FMN and FAD (Fig. 4B). Among them, Q112 and Y168 directly participate in FMN binding, and D175 and K668 are located midway between two flavins. D666 is important in maintaining the conformation of the FAD-binding site (Fig. 4C). Direct electron transfer can happen in a dis [35]. Whether the interflavin electron tance < 14 A transferring is direct as in other CPRs [10,13,14] needs to be tested. We first measured the in vitro reductive activity of ATR2 by using cytochrome c as electron acceptors (Fig. 4D,E). The results showed that ATR2 can

reduce cytochrome c in the presence of NADPH. To test whether any of the afore-mentioned residues between FMN and FAD are involved in possible electron relay, we mutated each residue to alanine and tested the activity of each mutant protein (Fig. 4F,G). Q112A, Y168A, and D666A mutant proteins lost about half activity, and D175A and K668A showed no difference from the wild-type protein. We constructed double mutants for Q112A, Y168A, and D175A and the triple mutant. Only one double mutant Q112A D666A was obtained, while the other mutants resulted in inclusion bodies during expression. The double mutant retained about one-third of the activity of the wild-type. Thus, none of these five residues is critical for the reductive activity, which suggests a direct electron transfer between these two cofactors. When cytochrome b5 was used as an electron acceptor, the activity of each mutant protein shows no significant difference from the wild-type ATR2 (Fig. 4H,K). This is consistent with the speculation that the reduction in cytochrome b5 is conducted by the FNR-like domain without involvement of the FMN-binding domain. Interdomain interactions The FMN-binding domain is in close contact with the connecting and NADPH-binding domains (Fig. 5A,B). Three pairs of interactions occur: the carbonyl groups of D239 and T113, and the side chain of E118, with the side-chain of K447, the main-chain amide of K419, and the side-chain of K418, respectively. This interface 2 as calculated by PISA proburies a total area of 265 A gram [36]. This is approximately one-third of that in 2). rCPR (791 A The interactions between FMN-binding and NADPH-binding domains include one salt bridge between R708 and E186. The buried surface area is 2, which is approximately twice of that in rCPR 409 A 2). R708 from the NADPH-binding domain (237 A adopts alternative conformations in Form A (Fig. 5A, B), and is the only residue that directly interacts with FMN-binding domain. To test the role of R708, we generated two mutants: R708E and R708L. The former mutant could abolish the salt bridge with E186, and the latter one is replaced by a neutral residue. We then measured the reductive activity of these two mutant proteins. The activity for cytochrome c reduction is totally lost, while the activity for cytochrome b5 is mostly retained (Fig. 5C). These results suggest that R708 is critical to bind FMN-binding domain, and thus facilitates the electron transfer from FAD to FMN.


757


G. Niu et al.

Fig. 2. ATR2 structure and its comparison with rCPR, yCPR, and BM3. (A) Ribbon diagrams of ATR2 in Form A (left) and Form B (right). For Form A, the four domains are colored as in Fig. 1. The cofactor FMN and FAD are shown as sticks and balls (yellow). For Form B, the connecting domain is in dark cyan, the FAD-binding domain is in dark green, and the NADPH-binding domain is in dark red. (B) Ribbon diagrams of rCPR (PDB 1AMO) (magenta), yCPR (PDB 2BF4) (orange), and BM3 (PDB 4DQL) (purple). The cofactor FMN, FAD, and NADP are shown as sticks and balls. (C) Superimposition of ATR2 with rCPR.

Discussion NADPH-binding site Arabidopsis thaliana NADPH-cytochrome P450 reductase 2 cannot form a stable complex with NADPH, and therefore we did not obtain the ATR2-NADPH complex in crystal. To get a clue of the NADPH-binding site, we modeled an NADPH molecule into the ATR2 structure by comparison with known homologs with bound NADPH (Fig. 6A,B). NADPH fits well with ATR2 in the ATR2-NADPH model, as most of the NADPH-binding residues are conserved among CPRs (Fig. 1). Similar to the cases of CPR bound with NADPH [10,14], the nicotinamide moiety in ATR2NADPH model is not stabilized by any interaction and is disordered. Upon NADPH binding to the NADPH-binding domain, the nicotinamide is stabilized by a stacking

758

interaction with the flavin ring of FAD [33]. Alternatively, the nicotinamide moiety of NADPH is too flexible to be observed when the flavin ring of FAD stacks with the indole ring of the C-terminal conserved tryptophan, W677 in rCPR [10]. In the ATR2 structure, the flavin ring of FAD is stacked by W711. Previous kinetics study has revealed that electron transferring from NADPH to FAD is much faster in plant CPR than that in hCPR [29], which could be a reason that NADPH cannot be ‘captured’ by ATR2 in this study. Rate-limiting step The distance between FAD and FMN is much longer in ATR2 than in rat, human, and yCPRs (Fig. 4A). However, the electron transfer in ATR2 is still proved to be direct (Fig. 4F,G,J, and K). This speculation is consistent with the kinetic assays [29]. In plant CPRs, interflavin electron transfer is slower than NADPH


G. Niu et al.


Fig. 3. Cofactor interactions. (A) Defined electron density of the cofactors FMN and FAD. The 2|Fo| |Fc| electron density map contoured at 1 r is in gray mesh. (B) The environment of FMN cofactor. The surrounding residues are colored according to each domain. Hydrogen bonds are shown in the dashed line with the water oxygen atoms in the red sphere. (C) Superimposition of FMN-binding domains from ATR2, rCPR, yCPR, BM3 (PDB 1BVY), and flavodoxin (PDB 4HEQ) (gray). (D) Details of the FMN-binding site in ATR2. The 2|Fo| |Fc| electron density map in the vicinity of the G169 is contoured at 1.5 r level. (E) The environment of FAD cofactor. The surrounding residues are colored according to each domain. Hydrogen bonds are shown in dashed line with the water oxygen atoms in red sphere. (F) Superimposition of FAD-binding domain from ATR2, rCPR, yCPR, and hCPR (PDB 3QE2) (violet). The side chain of a conserved tyrosine (Tyr478 of rCPR, Tyr861 of BM3, and Tyr481 of hCPR) is shown.

hydride transfer; in hCPR, transfer of a hydride ion from NADPH to FAD is tightly coupled with subsequent FAD to FMN electron transfer. The rate of the transfer of electrons from FMN to cytochrome c in all CPRs is rapid [29]. All these results suggest that the rate-limiting step for plant CPR is the interflavin electron transfer, and for hCPR/rCPR/yCPR is the electron transfer from NADPH to FAD. Conformational change in the FMN-binding domain The FMN-binding domain needs to undergo a large conformational change during each catalytic cycle [1,15,16,21]. When the FMN-binding and FAD-binding domains are fixed with an engineered disulfide bond, rCPR exhibited a significant loss of activity [20]. When a tetrapeptide sequence TGEE in the hinge region that links the FMN domain to the rest of rCPR was deleted, the protein was kept in an extended conformation and the electron transfer from FAD to FMN was defective [17]. Interestingly,

this extended rCPR is capable to bind to its redox partner [18]. Small-angle X-ray scattering of the oxidized and NADPH-reduced CPRs showed that they are in different shape [37]. The closed-open state transition should be a process coupled with the catalyzing cycle. In the known ‘closed-state’ CPR structures including the ATR2 structure reported here, the FMN-binding domain interacts directly with the connecting and NADPH-binding domains. These interactions are mostly salt-bridges and hydrogen bonds. This is consistent with the observation using small-angle X-ray scattering and nuclear magnetic resonance spectroscopy, which reveals that CPRs exist in a salt- and pH-dependent rapid equilibrium between a rigid/closed state and a highly flexible/ open state [38]. A model of plant-specific arginine switch R708 is on the interface between the N-terminal FMN-binding domain and the C-terminal NADPHbinding domain (Fig. 7A). Two conformations are


759


G. Niu et al.

Fig. 4. Interflavin electron transfer. (A) Superimposition of the flavin cofactors with the position of FAD fixed. The color scheme is as follows: cofactors of ATR2 is in yellow, of rCPR in magenta, of yCPR in orange, and of hCPR in violet. (B) The residues between the two flavins of ATR2. (C) The environment of D666. (D) The absorption spectra of three in vitro ATR2 activity experiments. The presence or absence of a component (8 lM cytochrome c, 0.4 nM ATR2, 100 lM NADPH) is indicated by + or , respectively. The liquid phase was spectroscopically monitored. (E) Progress curve of the ATR2-catalyzed cytochrome c reaction. The reaction was recorded every 10 s at 550 nm, and three replicates were conducted. (F) Kinetic progress curve of the wild-type and mutant ATR2 for cytochrome c reduction. (G) Assays for the critical residues involved in ATR2 catalysis of cytochrome c reduction in 60 s. (H) The absorption spectra of three in vitro ATR2 activity of cytochrome b5 experiments. (I) Progress curve of the ATR2-catalyzed cytochrome b5 reaction at 423 nm. (J) Kinetic progress curve of the wild-type and mutant ATR2 for cytochrome b5 reduction. (K) Assays for the critical residues involved in ATR2 catalysis cytochrome b5 in 60 s. The data in E, F, G, I, J, and K are presented as the means SD of three independent experiments.

seen for this arginine in Form A. Its mutation to glutamate or leucine abolishes the reductase activity of ATR2 for cytochrome c. This arginine is conserved from the green alga Chlamydomonas reinhardtii to higher plants, and it may act as a switch between the closed-open states by mediating the interdomain interaction. In the P450 BM3 system, the counterpart of this arginine is a lysine (Fig. 7B). The bacterial BM3 system is a self-sufficient mono-oxygenase with fused cytochrome P450 and CPR. Whether this reflects an evolutionary link between plant CPRs and bacterial self-sufficient P450 system awaits further analysis. 760

Experimental procedures Protein expression and purification The ATR2 gene (At4g30210) without the signal and transmembrane sequence (residues: 1‒72) was amplified by PCR and inserted into a modified pCWori+ vector. The maltosebinding protein (MBP) and the His6 tag-coding sequences from pETMALc-H vector [39] were ligated between the 50 NdeI and 30 XbaI sites of pCWori+ vector, which was a kind gift from D.J. Stuehr (Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH, USA). The ATR2 sequence (residues: 73‒711) was ligated between the


G. Niu et al.


Fig. 5. Interdomain interactions. (A) The two interfaces. The surface of ATR2 is colored according to the scheme described in the Fig. 1 (left), residues involved in interactions between the FMN-binding domain and the connecting domain (left), and residues involved in interactions between the FMN-binding domain and the NADPH-binding domain (right). (B) The 2| Fo| |Fc| electron density map of E186 and R708 contoured at 1 r level. (C) Assays for the critical residues involved in ATR2 catalysis of cytochrome c reduction or cytochrome b5 reduction in 1 min. The data are presented as the means SD of three independent experiments.

Fig. 6. NADPH-binding sites. (A) Superimposition of the NADPH-binding site from ATR2, rCPR, yCPR, and BM3. The side chain of residues involved in interactions with NADP are shown in sticks and balls, and the residues of ATR2 are labeled. (B) Overall shape of the NADPHbinding pocket.

50 XhoI and 30 XbaI sites of the modified pCWori+ vector. The resulting vector contained an MBP-His6 tag followed by a tobacco etch virus (TEV) protease cleavage site and the ATR2 sequence. To express the recombinant ATR2 protein, the vector was transformed into Escherichia coli BL21 cells. The cells were grown at 37 °C till the culture reached an optical density of around 0.8 at 600 nm, and then grown at 25 °C for 30 min before addition of 0.5 mM (final) IPTG. The cells were grown at 25 °C for 35–40 h, and then harvested by centrifugation. The cell pellets were resuspended in buffer A (200 mM NaCl, 20 mM Tris-HCl, pH 7.5) supplemented with 5 mM imidazole, sonicated in an ice bath, and then the cell debris was removed by centrifugation. The cleared lysate was loaded onto a Ni2+nitrilotriacetic acid (QIAGEN, Shanghai, China) agarose column equilibrated with buffer A, and then washed with buffer A supplemented with 20 mM imidazole. The recombinant MBP-His6-tagged ATR2 was eluted with 200 mM

imidazole in buffer A, and then concentrated by ultrafiltration using a Centriprep 50 filter (Amicon, Beverly, MA, USA) to a volume of 2 mL. The concentrate was loaded onto a HiLoad 16/60 Superdex 200 column (GE Healthcare, Piscataway, NJ, USA) equilibrated and eluted with buffer A. The fractions corresponding to MBP-His6-tagged ATR2 were collected, pooled, and mixed with TEV protease overnight at 4 °C. The cleaved MBP-His6 tag and the His6-tagged TEV protease were removed by passing the mixture through a Ni-nitrilotriacetic acid agarose column. The flow-through was concentrated to a volume of 2 mL and applied onto a HiLoad 16/60 Superdex 200 column equilibrated and eluted with buffer A. The fractions corresponding to ATR2 were collected and analyzed by SDS/ PAGE. The cytochrome b5 gene (At5g48810) without the disorder and transmembrane sequence (residues: 83‒140) was amplified by PCR and inserted into a pET-28a(+) vector


761


G. Niu et al.

Fig. 7. Structure comparison and sequence alignment. (A) Superimposition of ATR2 of chains A and B. The side chains of R708 are shown. (B) Sequence alignment. The amino-acid sequences can be found at http://www.p450.kvl.dk/. At: Arabidopsis thaliana, Gm: Glycine max, Ph: Petunia hybrid, Zm: Zea mays, Os: Oryza sativa, Tcu: Taxus cuspidate, Tch: Taxus chinensis, Sm: Selaginella moellendorffii, Pp: Physcomitrella patens, Cre: Chlamydomonas reinhardtii, Hs: Homo sapiens, Rn: Rattus norvegicus, Sc: Saccharomyces cerevisiae, Bm: Bacillus megaterium. The figure was generated with ESPript.

(Novagen, Darmstadt, Germany) between Nde I and Xho I sites to produce a fusion protein containing an N-terminal His6 tag. The vector was transformed into E. coli BL21 (DE3) cells. The cells were grown at 37 °C till the culture reached an optical density of around 0.8 at 600 nm, and then grown at 16 °C for 30 min before addition of 0.2 mM (final) IPTG. The cells were grown at 16 °C for 16 h, and then harvested by centrifugation. The cell pellets were resuspended in buffer A supplemented with 20 mM imidazole, sonicated in an ice bath, and then the cell debris was removed by centrifugation. The cleared lysate was loaded onto a Ni-nitrilotriacetic acid agarose column equilibrated with buffer A, and then washed with buffer A supplemented with 20 mM imidazole. The recombinant cytochrome b5 was eluted with 200 mM imidazole in buffer A, and then concentrated by ultrafiltration to a volume of 2 mL. The concentrate was loaded onto a HiLoad 16/60 Superdex 200 column equilibrated and eluted with buffer A supplemented with 2 mM DL-Dithiothreitol (Amresco, Solon, OH, USA). The fractions corresponding to cytochrome b5 were pooled and analyzed by SDS/PAGE, then concentrated, and frozen at 80 °C for further activity assay.

Crystallization and data collection The purified ATR2 was concentrated by ultrafiltration to 15 mgmL1. For crystallization, 1 lL ATR2 was mixed with 1 lL reservoir solution and equilibrated against a 200-lL reservoir by sitting drop method at 16 °C. The 762

crystals were grown in 21% (w/v) polyethylene glycol monomethyl ether 2000 and 0.1 M MES, pH 6.5. The crystals were transferred step by step into drops of the reservoir solution supplemented with 5%, 10%, and 15% (v/v) glycerol before being flash-cooled in liquid nitrogen. The diffraction data were collected at 100 K at the Shanghai Synchrotron Radiation Facility beamline BL17U. Diffraction data were indexed, integrated, and scaled using DENZO and SCALEPACK within the HKL2000 package (HKL Research, Inc., Charlottesville, VA, USA).

Structure determination and refinement The ATR2 structure was determined by molecular replacement using the program Phaser [40] in the CCP4 suite [41]. The coordinates of rat liver microsomal NADPH-CPR (PDB entry 1AMO) [10] were used as template. The resulting model was rebuilt using the PHENIX AutoBuild wizard [42]. Manual correction was performed with Coot [43] according to |Fo| |Fc| and 2|Fo| |Fc| maps, and further refinement was carried out with phenix.refine [44,45]. The quality of the refined structure was evaluated by MolProbity [46]. The results demonstrated that 96.7% of the residues were in the most favored regions of the Ramachandran plot, 3.2% were in the additional allowed regions, and none were in the disallowed regions. Coordinates and structure factors have been deposited in the PDB under the accession code 5GXU. All structure figures were odinger, LLC, New York, NY, prepared with PYMOL (Schr€ USA).


G. Niu et al.

Site-directed mutagenesis Arabidopsis thaliana NADPH-cytochrome P450 reductase 2 single mutants were generated with the Fast Mutagenesis System kit (TransGen Biotech, Beijing, China) using pCWori+MBP-His6-ATR2 as a template. The double and triple mutants were constructed using single and double mutants as a template, respectively. All mutants were sequenced to confirm the desired mutations. The procedure for purification of the mutant proteins was the same as that of the wild-type.

Reductase activity assay The ATR2 activity is assayed using both cytochrome b5 and cytochrome c (Sigma-Aldrich, St. Louis, MO, USA) as electron acceptors. The assay is performed in 300 lL buffer of 20 mM Tris-HCl, pH 7.5, and 100 lM NADPH in a quartz cuvette, with either 9.5 lM cytochrome b5 or 8.08 lM cytochrome c. The reduction in cytochrome b5 is started by adding ATR2 to a final concentration of 113 nM, then scanned between 350 and 600 nm after 270 s or recorded every 10 s at 423 nm. The reduction in cytochrome c is started by adding 0.4 nM ATR2, then scanned between 350 and 600 nm after 270 s or recorded every 10 s at 550 nm.

Acknowledgements We thank Ming-Zhu Wang at the Institute of Biophysics of the Chinese Academy of Sciences and the staff at Shanghai Synchrotron Radiation Facility for technical support during data collection. This work was supported by the National Natural Science Foundation of China (31670794 and 31530006) and the Strategic Priority Research Program of Chinese Academy of Sciences (XDB17030300).

Author contributions GN, LW, and WD performed the experiments; GN and SZ analyzed the data; GN, LL, and YH designed the study and wrote the paper.

Conflicts of interest The authors declared no conflicts of interest.

References 1 Iyanagi T, Xia C & Kim JJ (2012) NADPH-cytochrome P450 oxidoreductase: prototypic member of the diflavin reductase family. Arch Biochem Biophys 528, 72–89. 2 Guengerich FP (2005) Reduction of cytochrome b5 by NADPH-cytochrome P450 reductase. Arch Biochem Biophys 440, 204–211.


3 Porter TD & Kasper CB (1985) Coding nucleotide sequence of rat NADPH-cytochrome P-450 oxidoreductase cDNA and identification of flavinbinding domains. Proc Natl Acad Sci USA 82, 973–977. 4 Porter TD & Kasper CB (1986) NADPH-cytochrome P-450 oxidoreductase: flavin mononucleotide and flavin adenine dinucleotide domains evolved from different flavoproteins. Biochemistry 25, 1682–1687. 5 Schmidt HH, Smith RM, Nakane M & Murad F (1992) Ca2+/calmodulin-dependent NO synthase type I: a biopteroflavoprotein with Ca2+/calmodulin-independent diaphorase and reductase activities. Biochemistry 31, 3243–3249. 6 Sevrioukova IF, Li H, Zhang H, Peterson JA & Poulos TL (1999) Structure of a cytochrome P450-redox partner electron-transfer complex. Proc Natl Acad Sci USA 96, 1863–1868. 7 Eschenbrenner M, Coves J & Fontecave M (1995) NADPH-sulfite reductase flavoprotein from Escherichia coli: contribution to the flavin content and subunit interaction. FEBS Lett 374, 82–84. 8 Olteanu H & Banerjee R (2001) Human methionine synthase reductase, a soluble P-450 reductase-like dual flavoprotein, is sufficient for NADPH-dependent methionine synthase activation. J Biol Chem 276, 35558–35563. 9 Paine MJ, Garner AP, Powell D, Sibbald J, Sales M, Pratt N, Smith T, Tew DG & Wolf CR (2000) Cloning and characterization of a novel human dual flavin reductase. J Biol Chem 275, 1471–1478. 10 Wang M, Roberts DL, Paschke R, Shea TM, Masters BS & Kim JJ (1997) Three-dimensional structure of NADPH-cytochrome P450 reductase: prototype for FMN- and FAD-containing enzymes. Proc Natl Acad Sci USA 94, 8411–8416. 11 Smith GC, Tew DG & Wolf CR (1994) Dissection of NADPH-cytochrome P450 oxidoreductase into distinct functional domains. Proc Natl Acad Sci USA 91, 8710– 8714. 12 Munro AW, Noble MA, Robledo L, Daff SN & Chapman SK (2001) Determination of the redox properties of human NADPH-cytochrome P450 reductase. Biochemistry 40, 1956–1963. 13 Lamb DC, Kim Y, Yermalitskaya LV, Yermalitsky VN, Lepesheva GI, Kelly SL, Waterman MR & Podust LM (2006) A second FMN binding site in yeast NADPH-cytochrome P450 reductase suggests a mechanism of electron transfer by diflavin reductases. Structure 14, 51–61. 14 Xia C, Panda SP, Marohnic CC, Martasek P, Masters BS & Kim JJ (2011) Structural basis for human NADPH-cytochrome P450 oxidoreductase deficiency. Proc Natl Acad Sci USA 108, 13486–13491. 15 Laursen T, Jensen K & Møller BL (2011) Conformational changes of the NADPH-dependent


763


16

17

18

19

20

21

22

23

24

25

26

27

764

G. Niu et al.

cytochrome P450 reductase in the course of electron transfer to cytochromes P450. Biochim Biophys Acta 1814, 132–138. Aigrain L, Fatemi F, Frances O, Lescop E & Truan G (2012) Dynamic control of electron transfers in diflavin reductases. Int J Mol Sci 13, 15012–15041. Hamdane D, Xia C, Im SC, Zhang H, Kim JJ & Waskell L (2009) Structure and function of an NADPH-cytochrome P450 oxidoreductase in an open conformation capable of reducing cytochrome P450. J Biol Chem 284, 11374–11384. Sugishima M, Sato H, Higashimoto Y, Harada J, Wada K, Fukuyama K & Noguchi M (2014) Structural basis for the electron transfer from an open form of NADPH-cytochrome P450 oxidoreductase to heme oxygenase. Proc Natl Acad Sci USA 111, 2524–2529. Aigrain L, Pompon D, Morera S & Truan G (2009) Structure of the open conformation of a functional chimeric NADPH cytochrome P450 reductase. EMBO Rep 10, 742–747. Xia C, Hamdane D, Shen AL, Choi V, Kasper CB, Pearl NM, Zhang H, Im SC, Waskell L & Kim JJ (2011) Conformational changes of NADPH-cytochrome P450 oxidoreductase are essential for catalysis and cofactor binding. J Biol Chem 286, 16246–16260. Pudney CR, Heyes DJ, Khara B, Hay S, Rigby SE & Scrutton NS (2012) Kinetic and spectroscopic probes of motions and catalysis in the cytochrome P450 reductase family of enzymes. FEBS J 279, 1534–1544. Jensen K & Møller BL (2010) Plant NADPHcytochrome P450 oxidoreductases. Phytochemistry 71, 132–141. Mizutani M & Ohta D (2010) Diversification of P450 genes during land plant evolution. Annu Rev Plant Biol 61, 291–315. Bak S, Beisson F, Bishop G, Hamberger B, H€ ofer R, Paquette S & Werck-Reichhart D (2011) Cytochromes p450. Arabidopsis Book 9, e0144. Urban P, Mignotte C, Kazmaier M, Delorme F & Pompon D (1997) Cloning, yeast expression, and characterization of the coupling of two distantly related Arabidopsis thaliana NADPH-cytochrome P450 reductases with P450 CYP73A5. J Biol Chem 272, 19176–19186. Mizutani M & Ohta D (1998) Two isoforms of NADPH:cytochrome P450 reductase in Arabidopsis thaliana. Gene structure, heterologous expression in insect cells, and differential regulation. Plant Physiol 116, 357–367. Sundin L, Vanholme R, Geerinck J, Goeminne G, H€ ofer R, Kim H, Ralph J & Boerjan W (2014) Mutation of the inducible Arabidopsis thaliana cytochrome P450 reductase2 alters lignin composition and improves saccharification. Plant Physiol 166, 1956– 1971.

28 Varadarajan J, Guilleminot J, Saint-Jore-Dupas C, Piegu B, Chaboute ME, Gomord V, Coolbaugh RC, Devic M & Delorme V (2010) ATR3 encodes a diflavin reductase essential for Arabidopsis embryo development. New Phytol 187, 67–82. 29 Whitelaw DA, Tonkin R, Meints CE & Wolthers KR (2015) Kinetic analysis of electron flux in cytochrome P450 reductases reveals differences in rate-determining steps in plant and mammalian enzymes. Arch Biochem Biophys 584, 107–115. 30 Joyce MG, Ekanem IS, Roitel O, Dunford AJ, Neeli R, Girvan HM, Baker GJ, Curtis RA, Munro AW & Leys D (2012) The crystal structure of the FAD/NADPHbinding domain of flavocytochrome P450 BM3. FEBS J 279, 1694–1706. 31 Hsieh YC, Chia TS, Fun HK & Chen CJ (2013) Crystal structure of dimeric flavodoxin from Desulfovibrio gigas suggests a potential binding region for the electron-transferring partner. Int J Mol Sci 14, 1667–1683. 32 Rwere F, Xia C, Im S, Haque MM, Stuehr DJ, Waskell L & Kim JJ (2016) Mutants of cytochrome P450 reductase lacking either Gly-141 or Gly-143 destabilize its FMN semiquinone. J Biol Chem 291, 14639–14661. 33 Hubbard PA, Shen AL, Paschke R, Kasper CB & Kim JJ (2001) NADPH-cytochrome P450 oxidoreductase. Structural basis for hydride and electron transfer. J Biol Chem 276, 29163–29170. 34 Tejero J, Perez-Dorado I, Maya C, Martınez-J ulvez M, Sanz-Aparicio J, G omez-Moreno C, Hermoso JA & Medina M (2005) C-terminal tyrosine of ferredoxin-NADP+ reductase in hydride transfer processes with NAD(P)+/H. Biochemistry 44, 13477–13490. 35 Page CC, Moser CC, Chen X & Dutton PL (1999) Natural engineering principles of electron tunnelling in biological oxidation-reduction. Nature 402, 47–52. 36 Krissinel E & Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372, 774–797. 37 Ellis J, Gutierrez A, Barsukov IL, Huang WC, Grossmann JG & Roberts GC (2009) Domain motion in cytochrome P450 reductase: conformational equilibria revealed by NMR and small-angle x-ray scattering. J Biol Chem 284, 36628–36637. 38 Frances O, Fatemi F, Pompon D, Guittet E, Sizun C, Perez J, Lescop E & Truan G (2015) A well-balanced preexisting equilibrium governs electron flux efficiency of a multidomain diflavin reductase. Biophys J 108, 1527–1536. 39 Pryor KD & Leiting B (1997) High-level expression of soluble protein in Escherichia coli using a His6-tag and maltose-binding-protein double-affinity fusion system. Protein Expr Purif 10, 309–319.


G. Niu et al.

40 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC & Read RJ (2007) Phaser crystallographic software. J Appl Crystallogr 40, 658– 674. 41 Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie AG, McCoy A et al. (2011) Overview of the CCP4 suite and current developments. Acta Crystallogr D Biol Crystallogr 67, 235–242. 42 Terwilliger TC, Grosse-Kunstleve RW, Afonine PV, Moriarty NW, Zwart PH, Hung LW, Read RJ & Adams PD (2008) Iterative model building, structure refinement and density modification with the PHENIX AutoBuild wizard. Acta Crystallogr D Biol Crystallogr 64, 61–69. 43 Emsley P, Lohkamp B, Scott WG & Cowtan K (2010) Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, 486–501.


44 Adams PD, Afonine PV, Bunk oczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW et al. (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr 66, 213–221. 45 Afonine PV, Grosse-Kunstleve RW, Echols N, Headd JJ, Moriarty NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH & Adams PD (2012) Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallogr D Biol Crystallogr 68, 352–367. 46 Chen VB, Arendall WB 3rd, Headd JJ, Keedy DA, Immormino RM, Kapral GJ, Murray LW, Richardson JS & Richardson DC (2010) MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D Biol Crystallogr 66, 12–21.


765