The molybdenum cofactor (Moco) is a redox active prosthetic group, essentially required for numerous enzyme-catalyzed two electron transfer reactions. Moco is synthesized by an evolutionarily old and highly conserved multistep pathway. In the last step of Moco biosynthesis, the molybdenum center is inserted into the final Moco precursor adenylated molybdopterin (MPT-AMP). This unique and yet poorly characterized maturation reaction finally yields physiologically active Moco. In the model plant Arabidopsis, the two domain enzyme, Cnx1, is required for Moco formation. Recently, a genetic screen identified novel Arabidopsis cnx1 mutant plant lines each harboring a single amino acid exchange in the N-terminal Cnx1E domain. Biochemical characterization of the respective recombinant Cnx1E variants revealed two different amino acid exchanges (S197F and G175D) that impair Cnx1E dimerization, thus linking Cnx1E oligomerization to Cnx1 functionality. Analysis of the Cnx1E structure identified Cnx1E active site-bound molybdate and magnesium ions, which allowed to fine-map the Cnx1E MPT-AMP-binding site.

Introduction

In the cell, molybdenum exists as molybdate [1], which is redox inactive unless complexed by a low-molecular-mass molecule called molybdopterin (MPT, [2]). The latter constitutes the organic part of the molybdenum cofactor (Moco), which is found in the active site of all eukaryotic molybdenum enzymes (Mo-enzymes).

Moco is synthesized by a highly conserved multistep pathway (Figure 1) whereby initially guanosine triphosphate (GTP) is converted into the first stable Moco biosynthesis intermediate, cyclic pyranopterin monophosphate (cPMP). Its synthesis involves the radical-initiated, S-adenosylmethionine-dependent formation of the cyclic nucleotide (8S)-3′,8-cyclo-7,8-dihydroguanosine 5′-triphosphate (3′,8-cH2 GTP) followed by subsequent complex rearrangement and ring-closing reactions [3].

Biosynthesis of the molybdenum cofactor.

Figure 1.
Biosynthesis of the molybdenum cofactor.

General scheme of the Moco biosynthesis pathway. With the exception of 3′,8-cH2 GTP (a precursor of cPMP), known and characterized intermediates of the Moco biosynthesis pathway are shown sequentially. GTP, guanosine triphosphate; MPT, molybdopterin; MPT-AMP, adenylated molybdopterin. The domain structure of Arabidopsis molybdenum insertase Cnx1 is shown and catalyzed reactions are indicated.

Figure 1.
Biosynthesis of the molybdenum cofactor.

General scheme of the Moco biosynthesis pathway. With the exception of 3′,8-cH2 GTP (a precursor of cPMP), known and characterized intermediates of the Moco biosynthesis pathway are shown sequentially. GTP, guanosine triphosphate; MPT, molybdopterin; MPT-AMP, adenylated molybdopterin. The domain structure of Arabidopsis molybdenum insertase Cnx1 is shown and catalyzed reactions are indicated.

In the following, cPMP is converted into MPT. The necessary reaction steps for its formation are catalyzed by the MPT synthase complex, which was identified to introduce the MPT dithiolene motif stepwise [2]. The final Moco precursor is adenylated MPT (MPT-AMP, [4]), which is the substrate for molybdate insertion [2]. In the cell, MPT-AMP is immediately converted into physiologically active Moco. Two functional domains are required to catalyze MPT adenylation and subsequent molybdate insertion. In prokaryotes, both are expressed separately [5], whereas in eukaryotes — with the sole known exception of the green alga Chlamydomonas reinhardtii [6] — both domains are fused together. As a convention, these domains are named E- and G-domain, respectively [7]. The latter was shown to catalyze the formation of MPT-AMP [4], whereas the E-domain was identified to hydrolyze MPT-AMP and to insert molybdate into the MPT dithiolene motif [8]. However, when compared with all other reactions underlying cPMP and MPT formation, the mechanism(s) of Moco formation is only poorly characterized as yet. Structural and biochemical work, mostly carried out using recombinant Escherichia coli and Arabidopsis thaliana enzymes as representative mechanistic models, revealed the basis of the catalyzed reactions [817]. MPT-AMP, which was the latest Moco precursor to be identified, was found in complex with an MPT-AMP accumulating (plant) Cnx1G variant [4]. However, to our knowledge, all previous attempts to create a substrate (MPT-AMP and molybdate) accumulating E-domain variant failed, and correspondingly, the proposed Cnx1E reaction mechanism is purely hypothetical.

In the present study, the biochemical characterization of yet unknown plant Mo-insertase mutant lines led to the identification of hitherto unknown Cnx1 functional relevant residues. Elucidation of their possible contribution to Cnx1E active site chemistry was possible through solving the atomic structure of Cnx1E in complex with active site-bound molybdate and magnesium ions. As a major finding, we identified Cnx1E dimer-dependent activity being linked to dimer-dependent cofactor binding, thus explaining for the first time why Cnx1E is a functional dimer.

Experimental/materials and methods

A. thaliana lines used in the present study

Lines used in the present study are listed in Supplementary Table S1.

Cultivation of plants

For cultivation, one-half of MS medium [18] containing 0.8–1% (w/v) agar was used. After 3 weeks, pre-grown plants were transferred to flower soil. Plants were cultivated for 4 weeks under long-day conditions at 22°C.

RNA extraction and cDNA preparation of cnx1 mutant plants

Leaves of the XD lines (Supplementary Table S1) were chosen for RNA extraction using the NucleoSpin RNA Plant Kit (Macherey and Nagel) according to the manufacturer's instructions. cDNA was synthesized from 4 to 6 µg of total RNA using M-MLV-Reverse Transcriptase (Promega), and cnx1 cDNAs were amplified using Phusion® High-Fidelity DNA Polymerase (NEB) and appropriate primers. Subcloning was carried out using the CloneJET™ PCR Cloning Kit (Thermo Scientific™) according to the manufacturer's instructions. Sequencing of the subcloned cDNA fragments confirmed the presence of the mutations.

Metabolite extraction from A. thaliana leaves

For metabolite extraction, Arabidopsis leaves were ground in liquid nitrogen. Subsequently, the metabolites were extracted by adding ice-cold extraction buffer [50 mM HEPES/KOH (pH 7.2), 150 mM NaCl and 1% (v/v) Triton X-100, containing Roche complete EDTA-free protease inhibitor] and vigorous mixing. The extraction was performed at 4°C for 20 min. For clarification, the crude extract was centrifuged for 20 min at 21 000 × g and 4°C. Subsequently, the supernatant was transferred to a fresh and cold reaction tube and was used for further analysis.

Detection and quantification of Moco metabolites

Moco and its metal-free precursors, including cPMP, MPT and MPT-AMP, were detected through conversion to their stable oxidation products FormA-dephospho (MPT, MPT-AMP and Moco [16,19]) and CompoundZ (cPMP) [20], respectively. Quantification was performed based on the method described previously [21]. For metabolite quantification, 500 µg of protein (in plant crude extract) or 10 µg of recombinant protein was used.

Mo enzyme activity assay

For the determination of xanthine dehydrogenase (XDH) and aldehyde oxidase (AO) activities, Arabidopsis leaves were squeezed at 4°C in two volumes of extraction buffer [100 mM potassium phosphate (pH 7.5) and 1 mM EDTA], sonicated and centrifuged. AO and XDH enzyme activities were measured by an in-gel activity assay based on a method described previously [22,23]. For the determination of nitrate reductase (NR) and sulfite oxidase (SO) activity, 100 mg of leaf material was ground in liquid nitrogen. The enzyme activity assays were performed as described previously [24,25].

Cloning of cnx1E and variants

Wild-type (wt) cnx1E and the variants g108d, g175d and s197f were cloned by PCR using the respective, sequenced (see above) cnx1 full-length cDNAs as templates. Consistent with work published previously [7], we defined the first 453 codons of At5g20990 [26] as the region encoding Cnx1E. Wt cnx1E and the variants g108d, g175d and s197f were amplified by PCR using the Phusion® High-Fidelity DNA Polymerase (NEB) and appropriate primers (Cnx1E_BamHI_for 5′-cacacaggatccatggaaggtcaaggttgttg-3′ and Cnx1E_SpeI_rev 5′-cctcattatcagaaccaggctctactagtcacacac-3′). Subcloning of the PCR products was carried out using the CloneJET™ PCR Cloning Kit according to the manufacturer's instructions. After confirmation of sequence correctness, cnx1E and its variants characterized in the present study were subcloned into expression vector pGPlus.

Cnx1G s583a sequence

Sequencing confirmed that the published construct comprised nucleotides 1378–1884 of the annotated cnx1 coding sequence (At5g20990, [26]).

Cloning of the cnx1E co-expression vector pGPlus

For expression of wt cnx1E and its variants, a cnx1G co-expression vector was used, which is based on the pQE80L vector (Qiagen). Therefore, the existing multiple cloning site and the His6-Tag encoding sequence (5′-GAATTCATTAAAGAGGAGAAATTAACTATGAGAGGATCGCATCACCATCACCATCACGG ATCCGCATGCGAGCTCGGTACCCCGGGTCGACCTGCAGCCAAGCTT-3′) were replaced by a synthetic DNA fragment using restriction sites EcoRI and HindIII (Supplementary Table S2A). The cnx1G fragment was amplified by the PCR from a cnx1 full-length cDNA template (see above). Cloning was performed following a two-step overlap extension PCR strategy using Phusion® High-Fidelity DNA Polymerase (New England Biolabs) and appropriate primers (Supplementary Table S2B). Thus, obtained pGPlus vector contained the cnx1G wt coding sequence (At5g20990, nucleotides 1384–2013, [26]) controlled by a second ribosome-binding site, allowing the co-expression of untagged cnx1G and cnx1E strepII-tag fusion constructs.

Expression and purification of recombinant proteins

For expression, E. coli strains RK5204 [27,28] and RK5206 [27] were used. Expression of Cnx1E variants was carried out in 2YT medium [29] containing 50 µg/ml ampicillin at 22°C. For expression in RK5206 cells, 1 mM sodium molybdate was added at A600 = 0.6. At A600 = 1.0, cnx1E expression was induced with 1 mM isopropyl 1-thio-β-d-galactopyranoside, and cells were allowed to grow aerobically for 20 h. Cell lysis was performed using a French pressure cell, followed by a sonication step (2 × 5 min on ice). Cell debris was removed by centrifugation at 45 000 × g (60 min, 4°C). For affinity purification, we used Strep-Tactin® Superflow® high-capacity resin (IBA) according to the manufacturer's description. If not stated otherwise, all protein concentrations were determined with the Bradford assay (Roti-Quant; Roth) using bovine serum albumin as a concentration standard.

In vitro loading of MPT-AMP on Cnx1E variants

For MPT-AMP loading on Cnx1E, bacterial cells (E. coli strain RK5204) expressing Cnx1E were lysed and proteins were extracted as described above. Then, the lysate was loaded on Strep-Tactin® Superflow® high-capacity resin (IBA) and the resin was washed with five column volumes of washing buffer [100 mM Tris/HCl (pH 8.0), 150 mM NaCl, 5 mM EDTA and 10% (v/v) glycerol]. For in vitro loading, the immobilized Cnx1E was incubated with Cnx1G-bound MPT-AMP in the presence of 5 mM sodium molybdate for 30 min at 4°C. As MPT-AMP source Cnx1G S583A 6 × His was used [4]. Afterwards, the resin was washed with 10 column volumes of washing buffer, and the StrepII-tag fusion protein was eluted in elution buffer [100 mM Tris/HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 5 mM d-desthiobiotin and 10% (v/v) glycerol].

Circular dichroism spectroscopy

For circular dichroism (CD) data collection, all protein samples were buffered in 50 mM sodium phosphate (pH 7.2), 200 mM Na2SO4 and 5% (v/v) glycerol, and their concentration was adjusted to an amide bond concentration of 1 mM. CD spectra were recorded at 4°C with a Jasco J-815 CD spectrometer (Jasco, Inc.) in a quartz cuvette of 0.5 mm path length. Spectra were accumulated over 10 repeats in a wavelength range from 190 to 260 nm and in 1 nm increments. For baseline correction, the spectrum of the protein buffer was subtracted from the protein spectra. Finally, the CD data were converted into units of mean molar residue ellipticity (deg cm2 dmol−1) and plotted vs. wavelength.

HPLC-based MPT-AMP hydrolysis assay

For the determination of MPT-AMP hydrolysis activity, 100 pmol of Cnx1E-bound MPT-AMP was used. The reaction was performed in 250 µl of reaction buffer [50 mM sodium phosphate (pH 7.2), 200 mM NaCl, 5 mM GSH and 2% (v/v) glycerol]. Hydrolysis was initiated by the addition of 5 mM MgCl2. A control reaction was performed in the presence of 5 mM EDTA (instead of MgCl2). After 0 and 15 min incubation at 24°C, the reaction was stopped by mixing 100 µl of the hydrolysis reaction with 350 µl of oxidation solution [300 µl of 0.1 M Tris/HCl (pH 7.2) + 50 µl of acidic iodine (1% (w/v) I2/2% (w/v) KI in 1 M HCl]. MPT-AMP was quantified as described above. Recombinant Cnx1E consumes MPT-AMP approximately linear for at least 25 min after starting the reaction.

Inductively coupled plasma mass spectrometry

The molybdenum content of Cnx1E variants was quantified with Agilent 7700 Series inductively coupled plasma mass spectrometry (ICP-MS; Agilent Technologies) according to a standard calibration curve of inorganic molybdenum (Fluka). The detection limit of ICP-MS for molybdenum ranged from 1 to 20 µg/l. Protein solutions and standards were mixed automatically with rhodium as an internal standard. All values were corrected for the molybdenum content of control samples consisting of buffer alone. All data were collected and processed using the MassHunter work station software.

Crystallization, data collection and structure determination of Cnx1E

Before crystallization, the protein buffer was changed to 0.02 M Tris/HCl (pH 8.0), 0.15 M NaCl and 10% (v/v) glycerol via a PD-10 desalting column (GE Healthcare), and the protein was concentrated to ∼7 g/l. Initial crystallization conditions were found by carrying out sitting drop vapor diffusion experiments at 20°C using commercial screens. After optimization of initial crystallization conditions, best crystals of apo-Cnx1E grew from 0.05 M Li2SO4, 0.1 M HEPES/KOH (pH 7.5), 15% (v/v) PEG 600 and 10% (v/v) glycerol, whereas best crystals of Cnx1E loaded with molybdate and Mg2+ grew in the presence of 5 mM sodium molybdate and 5 mM MgCl2 from 0.2 M Li2SO4, 0.1 M HEPES/KOH (pH 7.3) and 50% (v/v) PEG 300. Prior to data collection, crystals were flash-cooled in liquid nitrogen and kept at 100 K throughout the crystallographic experiments. For apo-Cnx1E crystals, data were collected in-house with copper Kα radiation (λ = 1.5418 Å) generated by a Rigaku MicroMax 007HF rotating anode X-ray generator. The data set was processed with XDS [30], and the phase problem was solved by molecular replacement with Molrep [31] using a monomer of MoeA (PDB: 1FC5) as the search model [32]. The initial structure was improved by alternating steps of refinement with phenix.refine [33] and manual rebuilding in Coot [34]. For crystals of Cnx1E complexed with molybdate and Mg2+, data collection was carried out on beamline X06DA (Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland) at a wavelength of 1.7710 Å. Data processing revealed a significant mismatch of Friedel pair intensities due to anomalous scattering, which we attributed to molybdenum's L-I absorption edge. The corresponding anomalous signal, however, was too weak to be used for experimental phasing. As a consequence, phases for this data set were transplanted from the isomorphous apo-Cnx1E structure. We found two blobs of electron density ∼14 Å apart that were absent from the apo-Cnx1E structure and that were too strong to be explained by ordered water molecules. SHELXC [35] and ANODE [36] were used to analyze the anomalous signal intrinsic to the data set by combining the Bijvoet differences with model-derived phases. ANODE localized a few anomalous peaks, the strongest of which corresponded to an averaged anomalous density of ∼22σ and coincided with one of the two unexplained electron density blobs, whereas the density corresponding to the weaker peaks did not exceed 7σ and coincided with the sulfur atoms of methionine and cysteine residues. Its strength and the absence of other suitable anomalous scatterers from the crystallization condition allowed us to link the strongest anomalous peak unambiguously to the site of molybdate binding. The second blob of unexplained electron density did not coincide with the position of an anomalous scatterer and was interpreted as a not fully resolved magnesium–water complex due to its size and intensity. The structure was finalized by manual building in Coot and refinement with phenix.refine, accounting for the anomalous signal arising from the presence of molybdate by performing anomalous group refinement. The files containing the structure factors and the coordinates of the final structures were deposited with the RCSB PDB ([37], www.rcsb.org) with accession numbers 5G2R and 5G2S. The complete data collection and refinement statistics are presented in Table 1.

Table 1
Data collection and refinement statistics

The values in parentheses account for the shell of highest resolution. RMSD, root mean square deviation.

Cnx1ECnx1E + MoO4
Wavelength (Å) 1.5418 1.7710 
Resolution range (Å) 50.0–2.45 (2.54–2.45) 50.0–2.84 (2.94–2.84) 
Space group I222 
Unit cell dimensions a, b, c (Å) 65.9, 122.4, 131.5 
Number of reflections 
 Total 89 826 (8507) 166 246 (15 036) 
 Unique 19 065 (1808) 24 170 (2356) 
Completeness 0.96 (0.92) 0.99 (0.98) 
Mean I/σ(I16.3 (2.0) 21.7 (2.0) 
Wilson B (Å251.7 83.4 
CC1/2 0.999 (0.898) 1.0 (0.827) 
Rwork 0.2334 (0.3747) 0.2247 (0.3914) 
Rfree 0.2545 (0.3831) 0.2429 (0.4251) 
Number of non-hydrogen atoms 
 Total 3115 2967 
 In protein molecules 3043 2945 
 In ligands 13 12 
Number of amino acid residues 
 Total 470 
 Ordered 419 414 
RMSD bond lengths (Å) 
 Bond lengths (Å) 0.002 0.003 
 Angles (°) 0.45 0.57 
Ramachandran plot 
 Favored (%) 96.6 93.9 
 Disallowed 0.24 1.2 
Average B factor (Å260.4 97.3 
Cnx1ECnx1E + MoO4
Wavelength (Å) 1.5418 1.7710 
Resolution range (Å) 50.0–2.45 (2.54–2.45) 50.0–2.84 (2.94–2.84) 
Space group I222 
Unit cell dimensions a, b, c (Å) 65.9, 122.4, 131.5 
Number of reflections 
 Total 89 826 (8507) 166 246 (15 036) 
 Unique 19 065 (1808) 24 170 (2356) 
Completeness 0.96 (0.92) 0.99 (0.98) 
Mean I/σ(I16.3 (2.0) 21.7 (2.0) 
Wilson B (Å251.7 83.4 
CC1/2 0.999 (0.898) 1.0 (0.827) 
Rwork 0.2334 (0.3747) 0.2247 (0.3914) 
Rfree 0.2545 (0.3831) 0.2429 (0.4251) 
Number of non-hydrogen atoms 
 Total 3115 2967 
 In protein molecules 3043 2945 
 In ligands 13 12 
Number of amino acid residues 
 Total 470 
 Ordered 419 414 
RMSD bond lengths (Å) 
 Bond lengths (Å) 0.002 0.003 
 Angles (°) 0.45 0.57 
Ramachandran plot 
 Favored (%) 96.6 93.9 
 Disallowed 0.24 1.2 
Average B factor (Å260.4 97.3 

Analytical ultracentrifugation

Sedimentation velocity experiments were carried out in a Beckman Coulter ProteomeLab XL-I analytical ultracentrifuge using an An-50 Ti rotor at 50 000 rpm and 4°C. Concentration profiles were measured with the manufacturer's data acquisition software ProteomeLab XL-I Version 6.0 (Firmware 5.7) using the absorption scanning optics at 280 nm. For examination of cofactor binding to different oligomeric states of Cnx1E, absorbance data were acquired at 276 nm and 392 nm. Analysis was performed in a concentration range from 1.9 to 44.6 µM of Cnx1E variants in 3 or 12 mm double-sector centerpieces filled with 100 or 400 µl of sample, respectively. As a buffer, 0.15 M NaCl supplemented with either 50 mM sodium citrate (pH 5.4) or 50 mM Tris/HCl (pH 8.0) and 1 mM EDTA was used. Prior to analytical ultracentrifugation (AUC), Cnx1E and its variants were buffer-exchanged on a Superdex increase 10/300 column (GE Healthcare) into the respective buffers. To ensure that oligomerization equilibrium was reached before the start of centrifugation, samples were incubated for 24 h at 4°C after dilution of the stock solutions. Samples containing hydrolyzable cofactors were incubated only for 4–5 h at 4°C after dilution. For data analysis, a model for diffusion-deconvoluted differential sedimentation coefficient distributions [continuous c(s) distributions] implemented in the program SEDFIT [38] was used. Buffer densities and viscosities, partial-specific volumes and extinction coefficients at 280 nm were calculated from buffer or amino acid composition, respectively, by the program SEDNTERP [39] and were used to correct the experimental s-values to s20,w. Protein concentrations were determined spectrophotometrically using an extinction coefficient at A280 = 26 930 M−1 cm−1 for Cnx1E wt and the variants analyzed. The calculated concentrations are given in monomers throughout the text. Contributions of bound cofactors to absorption at 280 nm and partial-specific volumes of cofactor-loaded Cnx1E variants were not taken into account, since saturation was rather low.

Small-angle X-ray scattering analysis of Cnx1E

For small-angle X-ray scattering (SAXS) experiments, purified proteins were buffer-exchanged on a Superdex increase 10/300 column (GE Healthcare) to 20 mM HEPES/KOH (pH 7.4), 150 mM NaCl, 1 mM EDTA and 2% (v/v) glycerol. SAXS data were recorded on beamline BM29 [40] of the European Synchrotron Radiation Facility in Grenoble, France. The protein solution (12.5, 6.25, 3.13 and 1.65 g/l) samples were loaded to a quartz capillary mounted in vacuum by an automated sample changer and exposed to an X-ray energy of 12.5 keV. To reduce radiation damage, the protein solution was kept under steady flow while being irradiated. For each sample, 10 frames with an exposure time of 1 s per frame were recorded. The data were subsequently processed using programs of the ATSAS software package [41]. Scattering functions were derived from the frames and checked for radiation damage; those without were averaged and normalized, and the buffer's contribution to scattering was subtracted. Subsequently, scattering curves from 12.50 and 3.13 g/l solutions were merged to maximize scattering power at high resolution and to minimize the effect of protein aggregation at low resolution. Then, the radius of gyration Rg and the pair distance distribution function p(r) were computed from the scattering functions. The comparison of the experimental scattering curves with the Cnx1E monomer and dimer structures was carried out with CRYSOL [42]. In addition, 20 low-resolution models of Cnx1E based on the SAXS data were calculated with the program DAMMIF [43] and used for the construction of a representative ab initio model by using the programs DAMAVER and SUPCOMB [44]. The models had a mean normalized spatial discrepancy of 1.535 and a standard deviation of NSD (σNSD) of 0.315. No model showed an NSD of >2.165 ; hence, no model had to be excluded from the calculation of the ab initio model.

Molecular docking

We used AutoDock Vina [45] to find the possible position of MPT-AMP in the active site cleft of the protein. To avoid artificial clashes, the side chains of four lysyl residues Lys290, Lys294, Lys297 and Lys66′ were allowed to rotate freely. The size of the search box was chosen such that the MPT moiety could freely adopt new conformations, but the AMP moiety remained in place. The best conformation found by Vina was selected for energy minimization with AMBER [46]. To further improve the orientation of the ligand, a short molecular dynamics calculation at 300 K was carried out, followed by a simulated annealing comprising a fast heat-up to 2000 K within 4 ps, a cool-down to 300 K in 4 ps and a relaxation step at 300 K for 15 ps.

Results

Identification and characterization of cnx1 mutant plants

In the past, plants harboring mutations in the Moco biosynthesis pathway were identified by their complete loss of Moco-dependent NR activity, causing an easily selectable chlorate tolerance [47]. In 2005, a novel sirtinol-based screening method was established [48] that allowed for the identification of plant mutants with both entirely lacking and reduced Moco biosynthesis activity. Potential Arabidopsis (ecotype Landsberg erecta, Ler) Moco biosynthesis mutant plant lines coming from the sirtinol screen carried out by the Zhao laboratory [48] were analyzed by Kahlfeldt [49], and we identified three lines (XD302, XD324 and XD462) being impaired in the last step of Moco biosynthesis. Each of these mutant lines was identified to harbor a single-point mutation in the E-domain encoding part of the cnx1 gene causing the following single amino acid substitutions: G108D (XD302), S197F (XD324) and G175D (XD462), respectively, (Figure 2A–C). All measurable Mo-enzyme activities were reduced in the three XD lines (Figure 3A–D), thus resembling the pleiotropic phenotype well known for the last step of Moco biosynthesis mutants of various species [5,50,51]. Moco metabolite analysis revealed that, compared with the wt strain (Ler), XD lines 302 and 462 showed a slightly elevated cPMP amount (∼127 and 122%, respectively), whereas the cPMP amount of line XD324 was ∼70% of the wt. Consistent with the fact that cnx1 is mutated in these lines, the Moco/MPT amount was reduced in all lines (∼41%, XD lines 302, 324 and ∼84%, XD line 462; Figure 3B).

Identification of A. thaliana cnx1 mutant plants.

Figure 2.
Identification of A. thaliana cnx1 mutant plants.

(A) Schematic representation of the A. thaliana Cnx1 enzyme domain structure. The first and last residues of the domains are indicated. (B) Partial sequence comparison of A. thaliana (At), Rattus norvegicus (Rn), Aspergillus nidulans (An) and E. coli (Ec) Cnx1E homologs. (A and B) Asterisks mark the identified, Cnx1 functional relevant amino acid residues. (B) The position of each identified functional relevant amino acid residue of A. thaliana Cnx1 is given above each asterisk. (C) Overview of XD mutant plant lines, identified cnx1 mutations and resulting Cnx1 amino acid exchanges.

Figure 2.
Identification of A. thaliana cnx1 mutant plants.

(A) Schematic representation of the A. thaliana Cnx1 enzyme domain structure. The first and last residues of the domains are indicated. (B) Partial sequence comparison of A. thaliana (At), Rattus norvegicus (Rn), Aspergillus nidulans (An) and E. coli (Ec) Cnx1E homologs. (A and B) Asterisks mark the identified, Cnx1 functional relevant amino acid residues. (B) The position of each identified functional relevant amino acid residue of A. thaliana Cnx1 is given above each asterisk. (C) Overview of XD mutant plant lines, identified cnx1 mutations and resulting Cnx1 amino acid exchanges.

Biochemical characterization of A. thaliana cnx1 mutant plants.

Figure 3.
Biochemical characterization of A. thaliana cnx1 mutant plants.

(A) NR and SO activities of A. thaliana (Ler) wt and cnx1 mutant plants (XD lines). (B) Moco/molybdopterin (MPT) and cPMP content of A. thaliana wt plant (Ler) and XD lines XD302, XD324 and XD462. (C) XDH and (D) AO activities of A. thaliana wt (Ler) and cnx1 mutant plants (XD lines). For analysis, at least six plants obtained from two different cultivations were used. For NR and SO activity measurements, three technical replicates were recorded. The bars represent the standard deviation. One representative gel for AO and XDH activities is shown.

Figure 3.
Biochemical characterization of A. thaliana cnx1 mutant plants.

(A) NR and SO activities of A. thaliana (Ler) wt and cnx1 mutant plants (XD lines). (B) Moco/molybdopterin (MPT) and cPMP content of A. thaliana wt plant (Ler) and XD lines XD302, XD324 and XD462. (C) XDH and (D) AO activities of A. thaliana wt (Ler) and cnx1 mutant plants (XD lines). For analysis, at least six plants obtained from two different cultivations were used. For NR and SO activity measurements, three technical replicates were recorded. The bars represent the standard deviation. One representative gel for AO and XDH activities is shown.

Characterization of recombinant Cnx1E variants

In a subsequent set of experiments, we characterized recombinant Cnx1E variants G108D, S197F and G175D for MPT-AMP and molybdate binding. Therefore, all Cnx1E variants and the wt protein were recombinantly produced using E. coli strain RK5206 as host (Figure 4C). HPLC-based FormA analysis revealed that a significant MPT-AMP amount was co-purified with Cnx1E variant G108D (binding stoichiometry ∼0.14; Figure 4A and Supplementary Figure S1). Notably, the amount of co-purified MPT-AMP varied markedly among individual Cnx1E G108D, S197F and G175D preparations (Supplementary Figure S1). For variant S197F, we identified an MPT-AMP-binding stoichiometry of ∼0.1, whereas the MPT-AMP-binding stoichiometry of Cnx1E variant G175D (Figure 4A) and recombinant Cnx1E wt protein was found to be ∼0.03 and ∼0.02, respectively, thus being close to the detection limit of the HPLC system (Figure 4A). Since Cnx1E is known to bind both MPT-AMP and molybdate [8], we next carried out ICP-MS-based Mo quantification. It revealed a molybdate-binding stoichiometry of ∼0.13 (Cnx1E wt), ∼0.10 (Cnx1E G108D) and ∼0.07 (Cnx1E S197F, G175D) (Figure 4A).

Biochemical characterization of Cnx1E variants.

Figure 4.
Biochemical characterization of Cnx1E variants.

(A) Recombinant Cnx1E wt protein and variants G108D, S197F and G175D were analyzed for Moco/MPT, Mo and MPT-AMP content upon expression and purification in E. coli strain RK5206 [27]. Analysis of the recombinant proteins and in vitro MPT-AMP transfer reactions were carried out as described in the Materials and Methods section. Data sets recorded for one of three analyzed protein preparations are shown. The bars represent the standard deviation resulting from three technical replicates. For comparison, the mean values [derived from the shown data set and data sets of two other different protein preparations (Supplementary Figure S1)] are shown. (B) Comparison of Cnx1E wt MPT-AMP hydrolysis activity with Cnx1E variants G108D S197F and G175D. For analysis, 2 µM of Cnx1E was used, which corresponds to ∼90 pmol MPT-AMP (Cnx1E wt), ∼110 pmol MPT-AMP (Cnx1E G108D), ∼125 pmol MPT-AMP (Cnx1E S197F) and ∼130 pmol MPT-AMP (G175D). The bars represent the standard error of the mean resulting from three full replicates. (C) PAGE analyses of recombinant Cnx1E variants Cnx1E G108D, Cnx1E S197F and Cnx1E G175D. A 10 µg quantity of each recombinant protein was loaded onto the PA-gel, which was subjected to Coomassie blue staining after SDS–PAGE electrophoresis.

Figure 4.
Biochemical characterization of Cnx1E variants.

(A) Recombinant Cnx1E wt protein and variants G108D, S197F and G175D were analyzed for Moco/MPT, Mo and MPT-AMP content upon expression and purification in E. coli strain RK5206 [27]. Analysis of the recombinant proteins and in vitro MPT-AMP transfer reactions were carried out as described in the Materials and Methods section. Data sets recorded for one of three analyzed protein preparations are shown. The bars represent the standard deviation resulting from three technical replicates. For comparison, the mean values [derived from the shown data set and data sets of two other different protein preparations (Supplementary Figure S1)] are shown. (B) Comparison of Cnx1E wt MPT-AMP hydrolysis activity with Cnx1E variants G108D S197F and G175D. For analysis, 2 µM of Cnx1E was used, which corresponds to ∼90 pmol MPT-AMP (Cnx1E wt), ∼110 pmol MPT-AMP (Cnx1E G108D), ∼125 pmol MPT-AMP (Cnx1E S197F) and ∼130 pmol MPT-AMP (G175D). The bars represent the standard error of the mean resulting from three full replicates. (C) PAGE analyses of recombinant Cnx1E variants Cnx1E G108D, Cnx1E S197F and Cnx1E G175D. A 10 µg quantity of each recombinant protein was loaded onto the PA-gel, which was subjected to Coomassie blue staining after SDS–PAGE electrophoresis.

To trace back Cnx1E functionality, we next quantified the MPT-AMP hydrolytic activity of Cnx1E variants G108D, G175D and S197F. MPT-AMP hydrolysis activity is routinely tested in a fully defined in vitro assay containing the MPT-AMP-free Cnx1E domain and an excess of the MPT-AMP loaded Cnx1G domain [8]. However, since the hydrolysis assay depends on the in vitro transfer of MPT-AMP from the Cnx1G domain to the Cnx1E domain, we initially characterized the in vitro MPT-AMP-binding properties of all three Cnx1E variants created. To our surprise, Cnx1E variant G108D was a poor in vitro MPT-AMP binder, whereas Cnx1E wt and variants S197F and G175D showed overall comparable binding properties (Figure 4A). Since this renders variant Cnx1E G108D unsuitable for a routine MPT-AMP hydrolysis assay, we chose to compare the hydrolysis activities of the Cnx1E variants when loaded with MPT-AMP but in the absence of Cnx1G as MPT-AMP donor in the in vitro assay. Therefore, we carried out in vitro MPT-AMP transfer to Cnx1E variants S197F and G175D as well as to the wt protein, resulting in an MPT-AMP-binding stoichiometry of ∼0.2 (wt) to ∼0.25 (S197F, G175D, Figure 4A). However, we noted that among the individual in vitro transfer reactions carried out, the obtained Cnx1E MPT-AMP-binding stoichiometries varied remarkably. To exclude any influence of the single amino acid exchanges on the overall Cnx1E folding, we carried out CD spectroscopy, revealing no measurable influence of the single exchanges on Cnx1E folding (Supplementary Figure S2). Consequently, the varying MPT-AMP-binding stoichiometries quantified were traced back to the use of different Cnx1G S583A protein preparations and hence varying Cnx1G MPT-AMP saturations. Cnx1E was shown to bind MPT-AMP only in the presence of molybdate or other oxo-anions [8], which we experimentally accounted for by using an excess of molybdate in the transfer reactions. Hydrolysis activities of thus obtained in vitro loaded Cnx1E variants were directly compared with those of the Cnx1E G108D variant co-purified with ∼0.2 molecules MPT-AMP per molecule Cnx1E (Figure 4A and Supplementary Figure S1A). Doing so revealed that all of the variants are capable of Mg2+-induced MPT-AMP hydrolysis (Figure 4B) with, however, different activities.

We concluded that either reduced MPT-AMP hydrolysis capacity or impaired molybdate binding/activation [8] is causal for the observed biochemical properties of the variants tested. To disclose either of them, we co-crystallized recombinant Cnx1E wt protein with a single subdomain III-bound molybdate ion per monomer (Figure 5A,C, Table 1 and Supplementary Figure S4) and thus identified Cnx1 residues Gly108, Gly175 and Ser197 as located far from the molybdate-binding site. Therefore, we deduce that impaired MPT-AMP hydrolysis activity must be the reason for reduced Mo-enzyme activities and Moco content of the XD mutant plant lines (Figure 3).

Structure of Cnx1E.

Figure 5.
Structure of Cnx1E.

(A) The structure of the Cnx1E dimer shown in cartoon representation. Cnx1E crystallized with one monomer in the asymmetric unit and the physiological dimer is formed through crystallographic symmetry. The second monomer is shown in paler colors to indicate that it is not part of the asymmetric unit. Subdomains I–IV are colored as follows: domain I, blue; domain II, yellow; domain III, green; domain IV, maroon. The co-crystallized molybdate and magnesium ions are shown in sphere representation. The stability of the dimer was confirmed by SAXS and the derived envelope is shown in light yellow (for scattering curves, see Supplementary Figure S5). The positions of residues Gly108, Gly175 and Ser197 within monomer A are illustrated by arrows. (B and C) Superimposition of Cnx1E (PDB entry 5G2S) and GephE (PDB entry 5ERU) monomers, shown in cartoon representation. Cnx1E is colored as in (A). The superimposed GephE structure is shown outlined. The RMSD value calculated is 0.958 Å.

Figure 5.
Structure of Cnx1E.

(A) The structure of the Cnx1E dimer shown in cartoon representation. Cnx1E crystallized with one monomer in the asymmetric unit and the physiological dimer is formed through crystallographic symmetry. The second monomer is shown in paler colors to indicate that it is not part of the asymmetric unit. Subdomains I–IV are colored as follows: domain I, blue; domain II, yellow; domain III, green; domain IV, maroon. The co-crystallized molybdate and magnesium ions are shown in sphere representation. The stability of the dimer was confirmed by SAXS and the derived envelope is shown in light yellow (for scattering curves, see Supplementary Figure S5). The positions of residues Gly108, Gly175 and Ser197 within monomer A are illustrated by arrows. (B and C) Superimposition of Cnx1E (PDB entry 5G2S) and GephE (PDB entry 5ERU) monomers, shown in cartoon representation. Cnx1E is colored as in (A). The superimposed GephE structure is shown outlined. The RMSD value calculated is 0.958 Å.

The molecular function of Cnx1 residues Gly108, Gly175 and Ser197

The Cnx1E crystal structure revealed the positions of residues Gly108, Gly175 and Ser197. Residues Gly108 (subdomain II) and Gly175 (subdomain I, Figure 6) were identified as being part of the Cnx1E dimerization interface (Figure 5A and Supplementary Figure S4). Therefore, we reasoned that Cnx1E variants G108D and G175D are impaired in dimerization which is causal for the observed biochemical phenotypes of the respective mutant plant lines (XD 302 and XD462, Figure 3) and recombinant proteins (Figure 4). To provide evidence for this notion, we carried out AUC experiments. For comparison, Cnx1E variant S197F was included into the analysis.

Topology diagram of the Cnx1E monomer.

Figure 6.
Topology diagram of the Cnx1E monomer.

Helices are shown as cylinders; strands are shown as arrows. The first and the last amino acids of the secondary structure elements are indicated with numbers according to the A. thaliana Cnx1 sequence (At5g20990, [26]). Disordered regions are shown as broken lines. Black arrows with asterisks indicate the boundaries of Cnx1E proper with the surplus amino acids corresponding to expression tags. Red arrows with asterisks indicate functionally relevant amino acids identified in mutant plant lines XD302, XD324 and XD462. Molybdate and the magnesium–water complex are shown as stylized spheres with their binding sites indicated by gray arrows. The water molecules that are absent from the octahedral magnesium–water complex are shown not filled. Subdomains I–IV are colored as in Figure 5.

Figure 6.
Topology diagram of the Cnx1E monomer.

Helices are shown as cylinders; strands are shown as arrows. The first and the last amino acids of the secondary structure elements are indicated with numbers according to the A. thaliana Cnx1 sequence (At5g20990, [26]). Disordered regions are shown as broken lines. Black arrows with asterisks indicate the boundaries of Cnx1E proper with the surplus amino acids corresponding to expression tags. Red arrows with asterisks indicate functionally relevant amino acids identified in mutant plant lines XD302, XD324 and XD462. Molybdate and the magnesium–water complex are shown as stylized spheres with their binding sites indicated by gray arrows. The water molecules that are absent from the octahedral magnesium–water complex are shown not filled. Subdomains I–IV are colored as in Figure 5.

In an initial set of experiments, we aimed for the identification of the oligomerization state of the Cnx1E wt protein. Therefore, 1.9 to 44.6 μM of the protein were subjected to sedimentation velocity analysis in AUC. In Tris/HCl buffer (pH 8.0), wt Cnx1E sedimented, independent of protein concentration, predominantly as a single species with an s20,w of 5.1 S (Supplementary Figure S6A). From the continuous c(s) distribution model [38], a molar mass of ∼90 kg/mol was determined from sedimentation coefficient and diffusion broadening of the sedimenting boundary. Since the molar mass of the monomer as calculated from amino acid composition is 49.7 kg/mol, it is indicated that Cnx1E exists as a dimer in solution. Comparison of s20,w of the Cnx1E dimer to the s-value calculated for an unhydrated, spherical protein of the same mass yielded a frictional ratio of 1.44. Since, for a spherical, hydrated protein, a frictional ratio of 1.1–1.2 is expected [52], Cnx1E deviates substantially from the shape of a sphere. This indicates that Cnx1E is either elongated or contains unfolded regions, which is consistent with the results from the crystal structure and SAXS analysis (Figure 5 and Supplementary Figure S5). When the G175D mutant was analyzed in the same buffer, the protein showed substantial aggregation. Therefore, the buffer conditions were optimized by Thermofluor analysis and the highest stability was obtained in a citrate buffer at a pH of 5.4. Whereas wt Cnx1E formed again stable dimers under these conditions (Supplementary Figure S6B), G175D showed a concentration-dependent dissociation into monomers (Supplementary Figure S6C). At 1.9 µM G175D, >80% of the protein was monomeric. From the concentration dependency of the monomer–dimer equilibrium, a KD in the order of 10 µM can be roughly estimated. For the S197F variant, a clear tendency to dissociate into monomers is also evident. In addition, this mutant forms higher aggregates (Supplementary Figure S6D). In contrast, the Cnx1E G108D variant showed no tendency to dissociate, but a slight tendency to form aggregates at higher protein concentrations (Supplementary Figure S6E).

We conclude that the exchange of Cnx1E Gly175 for aspartic acid indeed impairs Cnx1E oligomerization, a notion that is supported by analysis of the dimerization interface. Variant S197F was found to be impaired in oligomerization, too. Although Ser197 is no immediate part of the dimerization interface, it is located just below that part of the interface that is contributed by subdomain III. Substitution of Ser197 by a bulky phenylalanine may easily disturb the dimerization interface, giving a likely explanation for the impaired oligomerization properties of variant S197F. Also, Cnx1E variant G108D displayed other oligomerization properties than expected. This is best explained by the fact that the Gly108 opposing part of the other monomer shows a depression that can accommodate an aspartic acid side chain (G108D), thus preventing an impact of this substitution on Cnx1E dimerization. How to explain the reduced Moco synthesis activities of Cnx1E variants G108D, S197F and G175D?

Recently, the structure of the mammalian Cnx1E homolog GephE was solved in complex with subdomain III-bound adenosine, which led to the conclusion that Cnx1E subdomain III binds MPT-AMP [53]. Since GephE subdomain III of one monomer is in close proximity to subdomains I and II of the other monomer, we reasoned that potentially also the bound substrate/educt interacts with both monomers of the dimer. Careful inspection of the available structural data (PDB entries 5ERR, 5ERS, 5ERT, 5ERU and 5ERV) revealed that in GephE, no directional interactions exist between adenosine bound by subdomain III of monomer A and monomer B. However, since the MPT moiety of MPT-AMP was not present in these GephE structures, we suggest that, in fact, the MPT moiety is sandwiched between both E-domain monomers, thus explaining why the dimerization-impaired Cnx1E variants G175D and S197F both display reduced Moco synthesis activities. To provide evidence for this, we checked whether or not the different oligomeric states of Cnx1E variants G175D and S197F are capable of binding cofactor to a comparable extent. Therefore, MPT-AMP loaded proteins were subjected to sedimentation velocity experiments at two different detection wavelengths. On the one hand, signal was measured at 276 nm, the absorbance maximum of the aromatic amino acids of Cnx1E, and on the other hand, at 392 nm where Cnx1E-bound MPT-AMP shows an absorbance maximum. For technical reasons, it is impossible to obtain Cnx1E preparations that exclusively contain protein-bound MPT-AMP, but are free of Moco/MPT. Therefore, also Cnx1E-bound Moco/MPT contributes to absorption at 392 nm. When the c(s) distributions at the two wavelengths are compared (Figure 7), it can be clearly seen that the ratio of the signals depends on the oligomerization state of the G175D and S197F mutants that are partially dissociated into monomers. It is evident that the monomeric form binds the cofactor with a significantly lower affinity than the dimer (Figure 7B,C). Since dimerization does not contribute to the binding of the AMP moiety (see above), we conclude that, in fact, binding of the MPT moiety requires dimeric Cnx1E.

Cofactor binding is significantly reduced by dissociation of Cnx1E dimers.

Figure 7.
Cofactor binding is significantly reduced by dissociation of Cnx1E dimers.

Sedimentation velocity analyses of 37 µM Cnx1E wt (A), 28 µM Cnx1E G175D (B) and 32 µM Cnx1E S197F (C) loaded with MPT-AMP at 276 nm (dashed line) and 392 nm (solid line) in sodium citrate buffer. The ratio of absorbance at 392 and 276 nm was 1:18, 1:25 and 1:22 for wt, G175D and S197F, respectively. Therefore, c(s) distributions were normalized to an area of 1 for better comparison. Since the signals at 392 nm were low, c(s) distributions at this wavelength are significantly broadened. It can be clearly seen that monomeric Cnx1E binds the cofactor with a reduced affinity. (D) Total absorbance of monomeric and dimeric Cnx1E G175D at 276 and 392 nm was calculated from c(s) distributions in B. For ease of comparison, relative signals in percent (%) are given for each species. One representative result of three experiments is shown.

Figure 7.
Cofactor binding is significantly reduced by dissociation of Cnx1E dimers.

Sedimentation velocity analyses of 37 µM Cnx1E wt (A), 28 µM Cnx1E G175D (B) and 32 µM Cnx1E S197F (C) loaded with MPT-AMP at 276 nm (dashed line) and 392 nm (solid line) in sodium citrate buffer. The ratio of absorbance at 392 and 276 nm was 1:18, 1:25 and 1:22 for wt, G175D and S197F, respectively. Therefore, c(s) distributions were normalized to an area of 1 for better comparison. Since the signals at 392 nm were low, c(s) distributions at this wavelength are significantly broadened. It can be clearly seen that monomeric Cnx1E binds the cofactor with a reduced affinity. (D) Total absorbance of monomeric and dimeric Cnx1E G175D at 276 and 392 nm was calculated from c(s) distributions in B. For ease of comparison, relative signals in percent (%) are given for each species. One representative result of three experiments is shown.

In the case of Cnx1E G175D, only two baseline-separated peaks are observed that can be assumed to represent monomer and dimer. Therefore, comparison of the total signals monitored for each species at 276 and 392 nm allows an estimation of the different extents of loading with cofactor. Whereas >40% of the protein is monomeric, this fraction contributes to <20% of the signal at 392 nm (Figure 7D). Hence we estimate that the cofactor saturation of the monomer is ∼2.9-fold lower than that of the dimer.

The molecular function of Cnx1 residue Gly108

We finally deciphered the molecular reason for the reduced Moco synthesis activity of Cnx1E variant G108D (plant mutant line XD 302). In the Cnx1E wt structure, the surface depression opposing residue Gly108 (see above) was found to be occupied by a magnesium ion (Supplementary Figure S3). Analysis of the recently solved GephE structure in complex with ADP (PDB 5ERR) revealed that here the residue Gly446, which is corresponding to Cnx1E Gly108, is also located in close proximity to one of the two magnesium ions that are present per GephE monomer. Therefore, since, in the crystal structures of Cnx1E and GephE, a common magnesium ion-binding site was identified, we conclude that here the active site magnesium ion is bound. Finally, molecular modeling suggests a possible explanation for the reduced Moco synthesis activity of Cnx1E variant G108D, as it suggests that replacement of Gly108 by aspartic acid is potentially leading to interference with the active site-bound Mg2+ (Supplementary Figure S7), hence affecting MPT-AMP pyrophosphate bond cleavage.

Discussion

The final step of Moco biosynthesis requires the concerted action of both functional molybdenum insertase domains — catalyzing (i) MPT adenylation and (ii) subsequent Mg2+-dependent hydrolysis of MPT-AMP coupled with molybdate insertion [8,16]. This reaction yields physiologically active Moco, which can be inserted into Moco-free dioxo-molybdoenzymes without further maturation reactions required [8]. In the current working model, MPT adenylation is supposed to be required for MPT activation, which in turn shall be required for the formation of an (AMP) activated molybdate species [8]. The latter is suggested to be the insertion-competent form of molybdenum [8].

In our work, we identified the molybdate- and Mg2+-binding sites of the plant molybdenum insertase Cnx1E as being located in subdomains IV and III, respectively (Figure 6). The enzyme-bound molybdate was unambiguously identified by both the 2FoFc electron density and molybdenum's anomalous signal. The second strong peak of electron density observed in the Cnx1E active site was interpreted as a magnesium ion surrounded by an octahedral water solvation shell (Supplementary Figure S8). AUC experiments identified primarily dimeric Cnx1E to co-sediment with the bound MPT moiety, thus leading to the question of how Cnx1E Moco/MPT and MPT-AMP binding are best explained on the molecular level. Recently, the structures of Rattus norvegicus molybdenum insertase GephE in the presence and absence of ADP, AMP, molybdate or tungstate were published [53] (PDB entries 5ERQ, 5ERR, 5ERS, 5ERT, 5ERU and 5ERV). The authors identified the site accommodating the AMP moiety of the MPT-AMP substrate as being located in subdomain III. Superimposition of Cnx1E with these GephE structures revealed close similarity between Cnx1E and GephE (RMSD = 0.947–1.004 Å, Supplementary Figure S9), and thus ADP placed within the Cnx1E active site by superposition was used as an anchor for molecular docking experiments with MPT-AMP. Doing so revealed conformations of MPT-AMP bound to Cnx1E and GephE which are highly similar to each other but not similar to the MPT-AMP conformation found in Cnx1G (PDB entry: 1UUY; Figure 8). This is in line with the fact that GephE and Cnx1E catalyze the same reactions and hence most probably use MPT-AMP of a similar/identical conformation as a substrate for molybdate insertion, whereas Cnx1G catalyzes a much different reaction that yields MPT-AMP as a product. In contrast with the Cnx1E structure presented by us, the GephE–molybdate co-structure (5ERU) contains several partially occupied molybdenum sites, the strongest of which was attributed to the physiological molybdate-binding site. The single molybdate-binding site per Cnx1E monomer now confirms this notion ([53]; Figure 5B,C) and the absence of partially occupied molybdenum sites in our structure rules out any role thereof in Moco biosynthesis. Likewise, the Cnx1E structure possesses only a single Mg2+–water complex per monomer. In contrast, the GephE structure features two binding sites that are located in close proximity to each other (∼4 Å) and are occupied by different species of bivalent metal ions (5ERR: two Mg2+; 5ERS: one Mg2+; 5ERT: two Mn2+; 5ERU: one Mg2+, one Ca2+), which leaves some ambiguity about their catalytic relevance.

Cnx1E active site model.

Figure 8.
Cnx1E active site model.

Cnx1E [PDB entry 5G2S, (A)] and GephE [PDB entry 5ERU, (B)] active sites with bound molybdate, Mg2+ and in silico placed MPT-AMP. The molybdate anion and the magnesium cation are shown as spheres. MPT-AMP is shown as sticks. (A and B) The colors of the protein surface represent the electrostatic surface potential of Cnx1E and GephE from −10 (red) kBT/ec to +10 (blue) kBT/ec. The ADP from 5ERU was used as anchor for the modeling of MPT-AMP. The MPT moiety was allowed to rotate freely around the phosphate–phosphate bond. Energy minimization resulted in very similar MPT-AMP conformations for both Cnx1E (A) and GephE (B). Comparison of both MPT-AMP conformations (C and D) with the conformation found in Cnx1G (1UUY, represented by semi-transparent sticks with black outline) shows clear differences in the orientation of the pterin moiety.

Figure 8.
Cnx1E active site model.

Cnx1E [PDB entry 5G2S, (A)] and GephE [PDB entry 5ERU, (B)] active sites with bound molybdate, Mg2+ and in silico placed MPT-AMP. The molybdate anion and the magnesium cation are shown as spheres. MPT-AMP is shown as sticks. (A and B) The colors of the protein surface represent the electrostatic surface potential of Cnx1E and GephE from −10 (red) kBT/ec to +10 (blue) kBT/ec. The ADP from 5ERU was used as anchor for the modeling of MPT-AMP. The MPT moiety was allowed to rotate freely around the phosphate–phosphate bond. Energy minimization resulted in very similar MPT-AMP conformations for both Cnx1E (A) and GephE (B). Comparison of both MPT-AMP conformations (C and D) with the conformation found in Cnx1G (1UUY, represented by semi-transparent sticks with black outline) shows clear differences in the orientation of the pterin moiety.

Identification of the precise ADP position and Cnx1E-based confirmation of the GephE–molybdate-binding site allowed for reliable distance measurements. The GephE ADP α-phosphate is located far (>9 Å) apart from the active site molybdate and the same holds for our modeled Cnx1E MPT-AMP complex. In the current model of molybdate insertion, it is presumed that a molybdate oxygen atom promotes divalent metal cation (Me2+)-dependent MPT-AMP pyrophosphate bond cleavage, going in hand with the formation of an AMP-activated (i.e. an adenylated) molybdate species [8]. Identification of the GephE and Cnx1E-bound active site molybdate as being located far from the MPT-AMP (ADP) pyrophosphate bond renders this notion implausible. Careful inspection of the MPT-AMP conformations modeled for GephE and Cnx1E revealed that, in both cases, the MPT dithiolene motif is facing the active site molybdate with both dithiolene sulfur atoms directly pointing toward it (Figure 8). As a consequence of this conformation, the MPT part of MPT-AMP is sandwiched between the two monomers of the GephE and Cnx1E dimers, respectively. We reason that this conformation is crucial for molybdate insertion and that the dimer is in fact needed to stabilize it. This notion is in line with the data obtained for Cnx1E variants S197F and G175D, which show both impaired dimerization and reduced Moco synthesis activity in vivo (mutant plants) as well as in vitro (recombinant Cnx1E variants). Their residual Moco synthesis activity can be linked to the small fraction of dimeric protein still present under analysis conditions, as identified by the AUC. The fact that in vitro MPT-AMP loading to Cnx1E variants S197F and G175D can still be induced at high enzyme concentrations is in line with this finding. The exclusively dimeric Cnx1E variant G108D was co-purified with highest MPT-AMP amounts. As identified for Cnx1E variants S197F and G175D, variant G108D is also capable of MPT-AMP hydrolysis, however — as shown for the other variants — with reduced efficiency. This was identified as being likely due to the interference of Asp108 with the active site-bound magnesium ion (Supplementary Figure S7). Cnx1E was shown to catalyze Me2+-dependent MPT-AMP hydrolysis only in the presence of molybdate [8], which we traced back to a single molybdate ion-binding site per Cnx1E monomer. The latter was occupied by a sulfate anion (PDB 5G2R) in the absence of molybdate in the crystallization buffer. In the presence of molybdate, even a 40-fold molar excess of sulfate did not prevent molybdate binding in the crystal structure (PDB 5G2S), which is fully consistent with previous reports that identified even a ∼160-fold excess of sulfate over molybdate as being insufficient for replacing enzyme-bound molybdate [8]. Therefore, we conclude that this site has a much higher affinity for molybdate than for other bioavailable oxo-anions. This is in line with the fact that — in the absence of molybdate — and in a fully defined in vitro system for sulfate is compatible with the transfer of MPT-AMP from Cnx1G to Cnx1E [8], which hints at both oxo-anions binding to the same site.

Residues Ser400 and Arg369 were identified as being required for molybdate co-ordination. Both residues are located in the Cnx1E subdomain IV (Figure 6), thus attributing for the first time a function to this Cnx1E subdomain. These residues are conserved among molybdenum insertases from various species (Supplementary Figure S10), thus documenting their importance for molybdenum insertases' catalytic activity. Identifying residues Ser400 and Arg369 as novel active site residues and considering the postulated Cnx1E-bound MPT-AMP conformation, the current model of AMP-activated molybdate insertion requires revision. It is long known that, in the presence of high molybdate concentrations, the MPT dithiolene motif can be complexed with molybdate in a non-enzymatic fashion yielding physiologically active Moco [15,27,50,51,5456]. Given that, we suggest that Cnx1E compensates for the low cellular molybdate concentrations, by providing binding capacity for both substrates, i.e. molybdate and MPT-AMP. In our model, AMP fulfills an anchoring function, required to position MPT in place between both monomers of the dimer and directly opposite to the enzyme-bound molybdate. We suggest that hydrolysis of MPT-AMP initiates MPT complexation with molybdate, finally yielding physiologically active Moco, which subsequently can be transferred to cellular Moco acceptors [8,5759]. A similar molybdenum insertion mechanism may underlie Fe-Moco formation required in the process of nitrogen fixation [60,61].

Abbreviations

     
  • AO

    aldehyde oxidase

  •  
  • AUC

    analytical ultracentrifugation

  •  
  • CD

    circular dichroism

  •  
  • cPMP

    cyclic pyranopterin monophosphate

  •  
  • GTP

    guanosine triphosphate

  •  
  • ICP-MS

    inductively coupled plasma mass spectrometry

  •  
  • Me2+

    divalent metal cation

  •  
  • Mo

    molybdenum

  •  
  • Moco

    molybdenum cofactor

  •  
  • MPT

    molybdopterin

  •  
  • MPT-AMP

    adenylated MPT

  •  
  • NR

    nitrate reductase

  •  
  • NSD

    normalized spatial discrepancy

  •  
  • RBS

    ribosome-binding site

  •  
  • RMSD

    root mean square deviation

  •  
  • SAXS

    small-angle X-ray scattering

  •  
  • SO

    sulfite oxidase

  •  
  • XDH

    xanthine dehydrogenase

  •  
  • wt

    wild type.

Author Contribution

J.K. and U.C. contributed toward conception and design, acquisition of data, analysis and interpretation of data, and drafting of the article. C.P. contributed to conception and design, acquisition of data, and analysis and interpretation of data. J.R., S.S. and D.S. participated in acquisition of data, and in analysis and interpretation of data. D.W.H. gave his final approval of the version to be published. R.R.M. contributed to drafting of the article and gave his final approval of the version to be published. T.K. contributed to conception and design analysis and interpretation of data and drafting of the article, and gave his final approval of the version to be published.

Funding

This work was financed by grants of the Deutsche Forschungsgemeinschaft to R.R.M. [FOR1220, TP1] and to D.W.H. [FOR1220, TP3].

Acknowledgments

We thank Yunde Zhao (UCSanDiego) for providing the plant lines characterized in the present study. We also thank Lidia Litz (Hannover Medical School), and Luise Fricke and Franziska Holtkotte (Braunschweig University of Technology) for excellent technical assistance. We thank Adelina Calean (Braunschweig University of Technology) for excellent support with ICP-MS analysis.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Mendel
,
R.R.
(
2013
)
The molybdenum cofactor
.
J. Biol. Chem.
288
,
13165
13172
doi:
2
Mendel
,
R.R.
and
Kruse
,
T.
(
2012
)
Cell biology of molybdenum in plants and humans
.
Biochim. Biophys. Acta Mol. Cell Res.
1823
,
1568
1579
doi:
3
Hover
,
B.M.
,
Loksztejn
,
A.
,
Ribeiro
,
A.A.
and
Yokoyama
,
K.
(
2013
)
Identification of a cyclic nucleotide as a cryptic intermediate in molybdenum cofactor biosynthesis
.
J. Am. Chem. Soc.
135
,
7019
7032
doi:
4
Kuper
,
J.
,
Llamas
,
A.
,
Hecht
,
H.-J.
,
Mendel
,
R.R.
and
Schwarz
,
G.
(
2004
)
Structure of the molybdopterin-bound Cnx1G domain links molybdenum and copper metabolism
.
Nature
430
,
803
806
doi:
5
Leimkühler
,
S.
,
Wuebbens
,
M.M.
and
Rajagopalan
,
K.V.
(
2011
)
The history of the discovery of the molybdenum cofactor and novel aspects of its biosynthesis in bacteria
.
Coord. Chem. Rev.
255
,
1129
1144
doi:
6
Llamas
,
A.
,
Tejada-Jimenez
,
M.
,
Gonzalez-Ballester
,
D.
,
Higuera
,
J.J.
,
Schwarz
,
G.
,
Galvan
,
A.
et al
(
2007
)
Chlamydomonas reinhardtii CNX1E reconstitutes molybdenum cofactor biosynthesis in Escherichia coli mutants
.
Eukaryot. Cell
6
,
1063
1067
doi:
7
Stallmeyer
,
B.
,
Nerlich
,
A.
,
Schiemann
,
J.
,
Brinkmann
,
H.
and
Mendel
,
R.R.
(
1995
)
Molybdenum co-factor biosynthesis: the Arabidopsis thaliana cDNA cnx1 encodes a multifunctional two-domain protein homologous to a mammalian neuroprotein, the insect protein Cinnamon and three Escherichia coli proteins
.
Plant J.
8
,
751
762
doi:
8
Llamas
,
A.
,
Otte
,
T.
,
Multhaup
,
G.
,
Mendel
,
R.R.
and
Schwarz
,
G.
(
2006
)
The mechanism of nucleotide-assisted molybdenum insertion into molybdopterin: a novel route towards metal cofactor assembly
.
J. Biol. Chem.
281
,
18343
18350
doi:
9
Stallmeyer
,
B.
,
Schwarz
,
G.
,
Schulze
,
J.
,
Nerlich
,
A.
,
Reiss
,
J.
,
Kirsch
,
J.
et al
(
1999
)
The neurotransmitter receptor-anchoring protein gephyrin reconstitutes molybdenum cofactor biosynthesis in bacteria, plants, and mammalian cells
.
Proc. Natl Acad. Sci. U.S.A.
96
,
1333
1338
doi:
10
Schwarz
,
G.
,
Schulze
,
J.
,
Bittner
,
F.
,
Eilers
,
T.
,
Kuper
,
J.
,
Bollmann
,
G.
et al
(
2000
)
The molybdenum cofactor biosynthetic protein Cnx1 complements molybdate-repairable mutants, transfers molybdenum to the metal binding pterin, and is associated with the cytoskeleton
.
Plant Cell
12
,
2455
2472
doi:
11
Kuper
,
J.
,
Palmer
,
T.
,
Mendel
,
R.R.
and
Schwarz
,
G.
(
2000
)
Mutations in the molybdenum cofactor biosynthetic protein Cnx1G from Arabidopsis thaliana define functions for molybdopterin binding, Mo-insertion and molybdenum cofactor stabilization
.
Proc. Natl Acad. Sci. U.S.A.
97
,
6475
6480
doi:
12
Liu
,
M.T.W.
,
Wuebbens
,
M.M.
,
Rajagopalan
,
K.V.
and
Schindelin
,
H.
(
2000
)
Crystal structure of the Gephyrin-related molybdenum cofactor biosynthesis protein MogA from Escherichia coli
.
J. Biol. Chem.
275
,
1814
1822
doi:
13
Xiang
,
S.
,
Nichols
,
J.
,
Rajagopalan
,
K.V.
and
Schindelin
,
H.
(
2001
)
The crystal structure of Escherichia coli MoeA and Its relationship to the multifunctional protein Gephyrin
.
Structure
9
,
299
310
doi:
14
Kuper
,
J.
,
Winking
,
J.
,
Hecht
,
H.-J.
,
Mendel
,
R.R.
and
Schwarz
,
G.
(
2003
)
The active site of the molybdenum cofactor biosynthetic protein domain Cnx1G
.
Arch. Biochem. Biophys.
411
,
36
46
doi:
15
Heck
,
I.S.
,
Schrag
,
J.D.
,
Sloan
,
J.
,
Millar
,
L.J.
,
Kanan
,
G.
,
Kinghorn
,
J.R.
et al
(
2002
)
Mutational analysis of the gephyrin-related molybdenum cofactor biosynthetic gene cnxE from the lower eukaryote Aspergillus nidulans
.
Genetics
161
,
623
632
PMID:
[PubMed]
16
Llamas
,
A.
,
Mendel
,
R.R.
and
Schwarz
,
G.
(
2004
)
Synthesis of adenylated molybdopterin: an essential step for molybdenum insertion
.
J. Biol. Chem.
279
,
55241
55246
doi:
17
Nichols
,
J.D.
,
Xiang
,
S.
,
Schindelin
,
H.
and
Rajagopalan
,
K.V.
(
2007
)
Mutational analysis of Escherichia coli MoeA: two functional activities map to the active site cleft
.
Biochemistry
46
,
78
86
doi:
18
Murashige
,
T.
and
Skoog
,
F.
(
1962
)
A revised medium for rapid growth and Bio Assays with tobacco tissue cultures
.
Physiol. Plant.
15
,
473
497
doi:
19
Johnson
,
J.L.
,
Hainline
,
B.E.
,
Rajagopalan
,
K.V.
and
Arison
,
B.H.
(
1984
)
The pterin component of the molybdenum cofactor. Structural characterization of two fluorescent derivatives
.
J. Biol. Chem.
259
,
5414
5422
PMID:
[PubMed]
20
Johnson
,
J.L.
,
Wuebbens
,
M.M.
and
Rajagopalan
,
K.V.
(
1989
)
The structure of a molybdopterin precursor. Characterization of a stable, oxidized derivative
.
J. Biol. Chem.
264
,
13440
13447
PMID:
[PubMed]
21
Schwarz
,
G.
,
Boxer
,
D.H.
and
Mendel
,
R.R.
(
1997
)
Molybdenum cofactor biosynthesis: the plant protein Cnx1 binds molybdopterin with high affinity
.
J. Biol. Chem.
272
,
26811
26814
doi:
22
Koshiba
,
T.
,
Saito
,
E.
,
Ono
,
N.
,
Yamamoto
,
N.
and
Sato
,
M.
(
1996
)
Purification and properties of flavin- and molybdenum-containing aldehyde oxidase from coleoptyles of maize
.
Plant Physiol.
110
,
781
789
doi:
23
Hesberg
,
C.
,
Hansch
,
R.
,
Mendel
,
R.R.
and
Bittner
,
F.
(
2004
)
Tandem orientation of duplicated xanthine dehydrogenase genes from Arabidopsis thaliana: differential gene expression and enzyme activities
.
J. Biol. Chem.
279
,
13547
13554
doi:
24
Scheible
,
W.-R.
,
González-Fontes
,
A.
,
Morcuende
,
R.
,
Lauerer
,
M.
,
Geiger
,
M.
,
Glaab
,
J.
et al
(
1997
)
Tobacco mutants with a decreased number of functional nia genes compensate by modifying the diurnal regulation of transcription, post-translational modification and turnover of nitrate reductase
.
Planta
203
,
304
319
doi:
25
Lang
,
C.
,
Popko
,
J.
,
Wirtz
,
M.
,
Hell
,
R.
,
Herschbach
,
C.
,
Kreuzwieser
,
J.
et al
(
2007
)
Sulphite oxidase as key enzyme for protecting plants against sulphur dioxide
.
Plant Cell Environ.
30
,
447
455
doi:
26
Berardini
,
T.Z.
,
Reiser
,
L.
,
Li
,
D.
,
Mezheritsky
,
Y.
,
Muller
,
R.
,
Strait
,
E.
et al
(
2015
)
The Arabidopsis information resource: making and mining the ‘gold standard’ annotated reference plant genome
.
Genesis
53
,
474
485
doi:
27
Stewart
,
V.
and
MacGregor
,
C.H.
(
1982
)
Nitrate reductase in Escherichia coli K-12: involvement of chlC, chlE, and chlG loci
.
J. Bacteriol.
151
,
788
799
PMID:
[PubMed]
28
Rivers
,
S.L.
,
McNairn
,
E.
,
Blasco
,
F.
,
Giordano
,
G.
and
Boxer
,
D.H.
(
1993
)
Molecular genetic analysis of the moa operon of Escherichia coli K-12 required for molybdenum cofactor biosynthesis
.
Mol. Microbiol.
8
,
1071
1081
doi:
29
Sambrook
,
J.
and
Russel
,
D.W.
(
2001
)
Molecular Cloning: A Laboratory Manual
,
Cold Spring Harbour Laboratory Press
,
Cold Spring Harbour, New York
30
Kabsch
,
W.
(
2010
)
XDS. Acta Crystallogr. D Biol. Crystallogr.
66
,
125
132
doi:
31
Vagin
,
A.
and
Teplyakov
,
A.
(
1997
)
MOLREP: an automated program for molecular replacement
.
J. Appl. Cryst.
30
,
1022
1025
doi:
32
Schrag
,
J.D.
,
Huang
,
W.
,
Sivaraman
,
J.
,
Smith
,
C.
,
Plamondon
,
J.
,
Larocque
,
R.
et al
(
2001
)
The crystal structure of Escherichia coli MoeA, a protein from the molybdopterin synthesis pathway
.
J. Mol. Biol.
310
,
419
431
doi:
33
Adams
,
P.D.
,
Afonine
,
P.V.
,
Bunkóczi
,
G.
,
Chen
,
V.B.
,
Davis
,
I.W.
,
Echols
,
N.
et al
(
2010
)
PHENIX: a comprehensive Python-based system for macromolecular structure solution
.
Acta Crystallogr. Sect. D Biol. Crystallogr.
66
,
213
221
doi:
34
Emsley
,
P.
,
Lohkamp
,
B.
,
Scott
,
W.G.
and
Cowtan
,
K.
(
2010
)
Features and development of Coot
.
Acta Crystallogr. Sect. D Biol. Crystallogr.
66
,
486
501
doi:
35
Sheldrick
,
G.M.
(
2008
)
A short history of SHELX
.
Acta Crystallogr. Sect. A Found. Crystallogr.
64
,
112
122
doi:
36
Thorn
,
A.
and
Sheldrick
,
G.M.
(
2011
)
ANODE: anomalous and heavy-atom density calculation
.
J. Appl. Crystallogr.
44
,
1285
1287
doi:
37
Berman
,
H.M.
,
Westbrook
,
J.
,
Feng
,
Z.
,
Gilliland
,
G.
,
Bhat
,
T.N.
,
Weissig
,
H.
et al
(
2000
)
The protein data bank
.
Nucleic Acids Res.
28
,
235
242
doi:
38
Schuck
,
P.
(
2000
)
Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling
.
Biophys. J.
78
,
1606
1619
doi:
39
Laue
,
M.T.
,
Shah
,
B.D.
,
Rigdeway
,
T.M.
and
Pelletier
,
S.L.
(
1992
)
Computer-Aided Interpretation of Analytical Sedimentation Data For Proteins
, pp.
90
125
,
Analytical Ultracentrifugation in Biochemistry and Polymer Science
,
Cambridge, UK
40
Pernot
,
P.
,
Round
,
A.
,
Barrett
,
R.
,
De Maria Antolinos
,
A.
,
Gobbo
,
A.
,
Gordon
,
E.
et al
(
2013
)
Upgraded ESRF BM29 beamline for SAXS on macromolecules in solution
.
J. Synchrotron. Radiat.
20
,
660
664
doi:
41
Petoukhov
,
M.V.
,
Franke
,
D.
,
Shkumatov
,
A.V.
,
Tria
,
G.
,
Kikhney
,
A.G.
,
Gajda
,
M.
et al
(
2012
)
New developments in the ATSAS program package for small-angle scattering data analysis
.
J. Appl. Crystallogr.
45
,
342
350
doi:
42
Svergun
,
D.
,
Barberato
,
C.
and
Koch
,
M.H.J.
(
1995
)
CRYSOL — a program to evaluate X-ray solution scattering of biological macromolecules from atomic coordinates
.
J. Appl. Crystallogr.
28
,
768
773
doi:
43
Franke
,
D.
and
Svergun
,
D.I.
(
2009
)
DAMMIF, a program for rapid ab-initio shape determination in small-angle scattering
.
J. Appl. Crystallogr.
42
,
342
346
doi:
44
Volkov
,
V.V.
and
Svergun
,
D.I.
(
2003
)
Uniqueness of ab initio shape determination in small-angle scattering
.
J. Appl. Crystallogr.
36
,
860
864
doi:
45
Trott
,
O.
and
Olson
,
A.J.
(
2009
)
AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading
.
J. Comput. Chem.
31
,
455
461
doi:
46
Case
,
D.A.
,
Berryman
,
J.T.
,
Betz
,
R.M.
,
Cerutti
,
D.S.
,
Cheatham
, III,
T.E.
,
Darden
,
T.A.
et al
(
2015
)
AMBER 2015
,
University of California
,
San Francisco
47
Mendel
,
R.R.
(
1997
)
Molybdenum cofactor of higher plants: biosynthesis and molecular biology
.
Planta
203
,
399
405
doi:
48
Dai
,
X.
,
Hayashi
,
K.-i
,
Nozaki
,
H.
,
Cheng
,
Y.
and
Zhao
,
Y.
(
2005
)
Genetic and chemical analyses of the action mechanisms of sirtinol in Arabidopsis
.
Proc. Natl Acad. Sci. U.S.A.
102
,
3129
3134
doi:
49
Kahlfeldt
,
N.
(
2006
)
Isolierung und Charakterisierung von Molybdän-Cofaktor-Mutanten aus Arabidopsis thaliana
.
Diploma Thesis, Technische Universitaet Braunschweig
50
Mendel
,
R.R.
,
Alikulov
,
Z.A.
,
Lvov
,
N.P.
and
Müller
,
A.J.
(
1981
)
Presence of the molybdenum-cofactor in nitrate reductase-deficient mutant cell lines of Nicotiana tabacum
.
Mol. Gen. Genet.
181
,
395
399
doi:
51
Tomsett
,
A.B.
and
Garrett
,
R.H.
(
1980
)
The isolation and characterization of mutants defective in nitrate assimilation in Neurospora crassa
.
Genetics
95
,
649
660
PMID:
[PubMed]
52
Zhao
,
H.
,
Brautigam
,
C.A.
,
Ghirlando
,
R.
and
Schuck
,
P.
(
2013
)
Overview of current methods in sedimentation velocity and sedimentation equilibrium analytical ultracentrifugation
.
Curr. Protoc. Protein Sci.
Chapter 20, Unit 20. 12
doi:
53
Kasaragod
,
V.B.
and
Schindelin
,
H.
(
2016
)
Structural framework for metal incorporation during molybdenum cofactor biosynthesis
.
Structure
24
,
782
788
doi:
54
MacDonald
,
D.W.
and
Cove
,
D.J.
(
1972
)
Studies of temperature-sensitive cnx mutants in the fungus Aspergillus nidulans
.
Biochem. J.
127
,
19P
doi:
55
Falciani
,
F.
,
Terao
,
M.
,
Goldwurm
,
S.
,
Ronchi
,
A.
,
Gatti
,
A.
,
Minoia
,
C.
et al
(
1994
)
Molybdenum(VI) salts convert the xanthine oxidoreductase apoprotein into the active enzyme in mouse L929 fibroblastic cells
.
Biochem. J.
298
,
69
77
doi:
56
Probst
,
C.
,
Ringel
,
P.
,
Boysen
,
V.
,
Wirsing
,
L.
,
Alexander
,
M.M.
,
Mendel
,
R.R.
et al
(
2014
)
Genetic characterization of the Neurospora crassa molybdenum cofactor biosynthesis
.
Fungal Genet. Biol.
66
,
69
78
doi:
57
Kruse
,
T.
,
Gehl
,
C.
,
Geisler
,
M.
,
Lehrke
,
M.
,
Ringel
,
P.
,
Hallier
,
S.
et al
(
2010
)
Identification and biochemical characterization of molybdenum cofactor-binding proteins from Arabidopsis thaliana
.
J. Biol. Chem.
285
,
6623
6635
doi:
58
Fischer
,
K.
,
Llamas
,
A.
,
Tejada-Jimenez
,
M.
,
Schrader
,
N.
,
Kuper
,
J.
,
Ataya
,
F.S.
et al
(
2006
)
Function and structure of the molybdenum cofactor carrier protein from Chlamydomonas reinhardtii
.
J. Biol. Chem.
281
,
30186
30194
doi:
59
Ringel
,
P.
,
Krausze
,
J.
,
van den Heuvel
,
J.
,
Curth
,
U.
,
Pierik
,
A.J.
,
Herzog
,
S.
et al
(
2013
)
Biochemical characterization of molybdenum cofactor-free nitrate reductase from Neurospora crassa
.
J. Biol. Chem.
288
,
14657
14671
doi:
60
Hernandez
,
J.A.
,
Curatti
,
L.
,
Aznar
,
C.P.
,
Perova
,
Z.
,
Britt
,
R.D.
and
Rubio
,
L.M.
(
2008
)
Metal trafficking for nitrogen fixation: NifQ donates molybdenum to NifEN/NifH for the biosynthesis of the nitrogenase FeMo-cofactor
.
Proc. Natl Acad. Sci. U.S.A.
105
,
11679
11684
doi:
61
Hu
,
Y.
and
Ribbe
,
M.W.
(
2011
)
Biosynthesis of the metalloclusters of molybdenum nitrogenase
.
Microbiol. Mol. Biol. Rev.
75
,
664
677
doi:

Author notes

*

These authors contributed equally to this work.

Present address: MSD, Molenstraat 110, 5342 CC Oss, The Netherlands.