The molybdenum cofactor (Moco) is a redox active prosthetic group found in the active site of Moco-dependent enzymes (Mo-enzymes). As Moco and its intermediates are highly sensitive towards oxidative damage, these are believed to be permanently protein bound during synthesis and upon maturation. As a major component of the plant Moco transfer and storage system, proteins have been identified that are capable of Moco binding and release but do not possess Moco-dependent enzymatic activities. The first protein found to possess these properties was the Moco carrier protein (MCP) from the green alga Chlamydomonas reinhardtii. Here, we describe the identification and biochemical characterisation of the Volvox carteri (V. carteri) MCP and, for the first time, employ a comparative analysis to elucidate the principles behind MCP Moco binding. Doing so identified a sequence region of low homology amongst the existing MCPs, which we showed to be essential for Moco binding to V. carteri MCP.
The molybdenum cofactor (Moco) biosynthesis pathway is evolutionarily conserved and comprises a number of enzymatically catalysed reactions (e.g. see ). Within this pathway, GTP is initially converted into cyclic pyranopterin monophosphate (cPMP), which is the most stable intermediate . cPMP is subsequently converted into molybdopterin (MPT), which is substrate of the molybdenum insertase (Mo-insertase ). MPT binds to the Mo-insertase G-domain where it is converted into adenylated MPT (MPT-AMP ). MPT-AMP is subsequently used as substrate for the Mo-insertase E-domain-catalysed molybdate insertion reaction . This step of the Moco biosynthesis pathway is the last step that remains to be understood mechanistically, and recent work [6–8] has already laid the basis for the elucidation of the metal insertion reaction. Upon maturation, the highly oxidation-sensitive Moco must be allocated to and inserted into respective user enzymes. The most likely candidate proteins for the cellular Moco transfer system have been best characterised for the alga Chlamydomonas reinhardtii (C. reinhardtii [9–13]) and the higher plant Arabidopsis thaliana (A. thaliana, [14,15]). While Arabidopsis was found to harbour numerous proteins involved in Moco binding and transfer , C. reinhardtii possesses only two such proteins with different Moco binding properties . Biochemical studies using recombinant proteins have identified the C. reinhardtii Moco carrier protein (MCP) and the A. thaliana Moco binding proteins (MoBP) to possess varying Moco binding properties, which were attributed to their proposed cellular functions, i.e. Moco allocation (MoBPs) and Moco storage (MCP) respectively [13,14,16]. After Moco synthesis and transfer, the cofactor must be inserted into the Moco-dependent enzymes (Mo-enzymes). In prokaryotes, a protein complex comprising enzymes involved in the final steps of Moco maturation is assumed to donate Moco to the Mo-enzymes  and contrary to eukaryotes, various enzyme-specific Moco chaperons have been identified here (for an overview, see ). In marked contrast, as yet, no Mo-enzyme specific chaperone has been identified in a eukaryote.
In studies using fully defined in vitro systems, the recombinant Moco-free Neurospora crassa (N. crassa) nitrate reductase (NR)  and A. thaliana sulphite oxidase proteins  revealed that Moco insertion into eukaryotic Mo-enzymes does not mandatorily require a Moco chaperon. However, at least for the N. crassa NR unknown factor(s) may be required to obtain NR with stoichiometric amounts of bound Moco . Interactions of the Mo-insertase and the Moco transfer system but also the user enzymes have been described [14,15], suggesting that during  but also downstream  synthesis, Moco and its metabolites require protected transport. In this work, we describe the identification and biochemical characterisation of the Volvox carteri (V. carteri) MCP, revealing that reversible Moco binding essentially depends on a species-specific highly flexible loop that covers the accepted MCP Moco binding site.
Identification of C. reinhardtii MCP homologous proteins
Cloning of V. carteri MCP
The coding sequence of V. carteri MCP (XP_002954772.1) was synthesised and subcloned via BamHI/SpeI into a modified pQE80 (Quiagen, Germany) vector which allowed the expression of a C-terminal StrepII-tag® fusion protein. Generation of the vector was carried out by replacing the His6 tag encoding sequence and the multiple cloning site of the pQE80 vector for the synthesised DNA sequence (5′-GAATTCATTAAAGAGGAGAAATTAACTATGTCTAGAGGATCCAGATCTCACGTGACTAGTGGCGGTTCCGCCTGGTCCCACCCCCAGTTCGAGAAGTAATGAAGCG CTAAGCTT-3′) by using the restriction sites EcoRI and HindIII. For generation of the V. carteri MCP low homology region (LHR) exchange variant, the following DNA sequence was synthesised: (5′-GAATTCATTAAAGAGGAGAAATTAACTATGTCTAGAGGATCCATGGCTCCACGCAAACCGATCATTGGCGTTATGGGACCAGGCGAACAGGCTACGCCTACGGATGTGG AACTGGCCACCGAACTGGGCAAGCAGATTGCGAGCCATGACTGGATTCTGCTCACAGGTGGGCGTAGTCTGGGCGTGATGGATGCGGCATGC-3′) and used to replace a DNA stretch within the V. carteri MCP expression construct by using the restriction sites EcoRI and SpeI. Doing so resulted in the exchange of the V. carteri MCP LHR region (spanning amino acids 15–26 for the Rippkaea orientalis (R. orientalis) MCP LHR region (spanning amino acids 13–19, ).
Expression and purification of V. carteri MCP
For recombinant protein production of V. carteri MCP and its derivative MCPRo LHR, Escherichia coli strains TP1000 , RK5206  or RK5204 [24,25] were used. Expression was carried out in LB medium containing 100 µg/mL ampicillin and at 25°C. When detailed as such, 1 mM sodium molybdate was used as a supplement to the media. MCP expression was induced with 50 µM isopropyl β-d-1-thiogalactopyranoside when the optical density of the culture reached 0.1 at 600 nm. Subsequently, cells were allowed to grow aerobically for 20 h. Upon harvesting, cell lysis was performed using an Avestin Emulsiflex C5 homogenizer. A subsequent centrifugation step (24,400 × g, 60 min, 4°C) was used to clear the lysate, which was treated with DNaseI (Roche) prior to being loaded on to Strep-Tactin® Superflow® high capacity resin (IBA). All following affinity purification steps were carried out according to the manufacturer (IBA) description and at 4°C. Purity of the eluted proteins was routinely documented by SDS/PAGE analysis. 12% SDS/PAGE gels were TGX-stained and processed using ImageLab 6.0 with a ChemiDoc XRS+ System (Bio-Rad). Pure protein fractions were concentrated (Vivaspin 6, Sartorius AG) and used directly for the subsequent biochemical characterisation. For the determination of its oligomeric state, V. carteri MCP was subjected to gel filtration experiments with a Superdex 200 10/300 Increase column (GE Healthcare) in a buffer containing 100 mM Tris-HCl pH 8.0, 150 mM NaCl, and 5% (v/v) glycerol. The column was calibrated with a High Molecular Weight calibration kit (GE Healthcare).
Quantification of protein concentrations
Protein concentration determination was carried out using the Pierce™ BCA Protein-Assay (Thermo Scientific). Absorption at 562 nm was measured using a Thermo Scientific™ Multiskan™ GO Microplate Spectrophotometer.
Thermal-shift experiments using recombinant V. carteri MCP
Thermal shift (Thermo Fluor) assays were employed to determine the thermal stability of V. carteri MCP. Using a CFX96/C1000 qPCR Instrument (Bio-Rad) melt curves were generated with SYPRO® Orange Protein Gel Stain (Sigma) and analysed using the CFX Manager Software V 3.0 (Bio-Rad). Melting temperatures were determined for the specified buffer conditions. To obtain MCP samples with different Moco saturations, an MCP protein preparation possessing a Moco occupancy of 0.24 ± 0.03 was diluted with corresponding amounts of Moco/MPT-free MCP.
V. carteri MCP Moco stabilisation
Pure recombinant V. carteri MCP protein obtained by expression and purification from E. coli strain TP1000 (Moco/MPT saturation ∼ 0.21) was incubated at 4, 22 and 60°C for 160 min. At defined time intervals, protein equivalent to 2–20 pmol of Moco/MPT (inferred from T0 min) was oxidised (n=3) to allow its quantification via the stable oxidation product FormA. Protein concentration was routinely determined for every sample at the end of the timed experiment (T160 min).
Quantification of V. carteri MCP-bound Moco/MPT
Quantification of V. carteri MCP and MCPRo LHR-bound Moco/MPT was carried out via FormA-based HPLC analysis, essentially as described in  and using synthetic FormA as calibration standard . A total of 0.25–2.5 µg of pure protein were oxidised to quantify Moco/MPT amounts of the TP1000 purified protein. In the case of Moco-free protein (obtained from strain RK5204), up to 50 µg were oxidised to document that no trace amounts of Moco/MPT were present.
UV VIS spectroscopy
For UV VIS spectroscopy, recombinant V. carteri MCP was adjusted to 600 µM in purification buffer (100 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM EDTA, 5% (v/v) glycerol). Spectra were recorded using a PerkinElmer UV/VIS Spectrometer Lambda 25 between 380 and 620 nm.
Inductively coupled plasma mass spectrometry
Molybdenum content in purified protein samples was quantified with an Agilent 7700 Series ICP-MS (Agilent Technologies) using a standard calibration curve of inorganic molybdenum (Fluka). The molybdenum calibration for the inductively coupled plasma mass spectrometry (ICP-MS) was in the range of 0.1–20 µg/l. Samples were mixed with rhodium (Rh(NO3)3) as an internal standard. Gathered data were processed using MassHunter work station software. Protein samples with concentrations of 0.8–1.5 mmol/l were used for analysis. Samples were diluted 1:1 in 69% Suprapur® nitric acid (Merck), after an overnight incubation at room temperature they were diluted with Milli-Q water prior to ICP-MS measurements.
The construction of a V. carteri MCP model was carried out with the one-to-one threading option of the Phyre2 modelling server  with the structures of the two confirmed MCPs from C. reinhardtii (, PDB: 2IZ6, 74.6% sequence identity with V. carteri MCP) and R. orientalis (, PDB: 6Y01, 51.5% sequence identity with V. carteri MCP) serving as templates. Phyre2 was run using a local alignment method and with secondary structure scoring enabled using a secondary structure weighing of 0.1.
Small-angle X-ray scattering
Small-angle X-ray scattering (SAXS) experiments were conducted on beamline BM29 of the European Synchrotron Radiation Facility, Grenoble, France . A concentration of 3.9 g/l of gel filtration purified V. carteri MCP were loaded into a quartz capillary and subjected to irradiation with X-rays of 12,500 eV at room temperature. The scattering intensities were recorded within a momentum transfer range of 1.0 × 10−2 to 5.0 × 10−1 Å−1 using a PILATUS 1 M hybrid photon-counting detector. A total of ten scattering images were collected with an exposure time of 0.5 s per image. The scattering data were processed with the ATSAS suite . We created oligomeric models of V. carteri MCP by arranging the Phyre2 model that was based on the C. reinhardtii MCP structure similar to the arrangements found in the C. reinhardtii MCP crystal structure (, PDB: 2IZ6) resulting in a V. carteri MCP monomer A, a dimer AB, a dimer AD and a tetramer ABCD. We finally calculated theoretical scattering curves of these oligomeric models and compared them with the experimental scattering curve with the program CRYSOL .
Identification of C. reinhardtii MCP homologous proteins
Hitherto a single MCP (from the green alga C. reinhardtii) has been characterised biochemically in depth [9–13], which documented its function for reversible Moco (Figure 1B) binding. Amongst eukaryotes, pBLAST  searches revealed three other potential MCP homologues, all within the order Chlamydomonadales. Here potential MCP homologues have been identified in the algae Chlamydomonas eustigma, Haematococcus lacustris, Gonium pectorale and V. carteri (Figure 1A). Amongst these, the putative MCP from V. carteri was chosen for subsequent characterisation as a rapid phylogenetic analysis  revealed this to be the closest relative to C. reinhardtii MCP.
Identification of the V. carteri MCP
The oligomeric state of V. carteri MCP
The oligomeric state of V. carteri MCP was determined by gel filtration and SAXS. In gel filtration, V. carteri MCP migrates at a retention volume of ∼13 ml, which correlates with an apparent molecular weight of ∼80 kDa (Figure 2A). Considering the monomeric molecular weight of the V. carteri MCP construct that was used in the present study (∼19.5 kDa), the gel filtration strongly suggests a V. carteri MCP tetramer in solution, agreeing with the tetrameric states that were published for C. reinhardtii MCP  and R. orientalis MCP . SAXS confirmed that notion of a V. carteri MCP tetramer. SAXS further revealed that the arrangement of the V. carteri MCP tetramer obeys D2 symmetry, as a V. carteri MCP tetramer model based on the structure of the C. reinhardtii MCP D2 tetramer shows the best agreement with the experimental SAXS curve derived from V. carteri MCP in solution (Figure 2B).
Oligomeric state of the V. carteri MCP
Biochemical characterisation of V. carteri MCP
Recombinant expression and purification of the putative MCP from V. carteri was carried out in the Moco accumulating E. coli strain TP1000 , the MPT accumulating E. coli strain RK5206  and the Moco/MPT free E. coli strain RK5204 (Figure 3A,B [24,25]). This yielded recombinant protein preparations of high purity as documented by SDS/PAGE analysis (Figure 3A). UV-VIS spectroscopic data for the recombinant putative MCP from V. carteri (Figure 4A) document overall comparable spectral properties of the putative MCP from V. carteri and the described MCP from C. reinhardtii . Consistent with previous reports, we have identified an absorption band at ∼420 nm (23,800 cm−1) that we tentatively assign as a dithiolene → Mo charge transfer transition (Figure 4A [13,34–37]). Electronic absorption data for [(bdt)MoO3]2−  (bdt = benzene-1,2-dithiol) and [(bdt)MoO2(OSiPh3)]1− , which are small molecule Moco analogues, indicate that there are no electronic transitions at wavelengths above 325 nm (30,800 cm−1) for [(bdt)MoO3]2−. Thus, the V. carteri MCP absorption data suggest that at least one of the oxo ligands is likely to be protonated. For example, the di-oxo compound [(bdt)MoO2(OSiPh3)]1− possesses lower energy electronic absorption bands at ∼395 nm (25,300 cm−1) and ∼495 nm (20,200 cm−1) . The protonation of a Moco oxo ligand is likely to be extremely important in providing a labile, presumably aqua, ligand that can dissociate and allow for amino acid coordination to Mo when the cofactor is inserted into apo-proteins.
Purification of the recombinant V. carteri MCP from different E. coli expression strains
Biochemical characterisation of the V. carteri MCP
Quantitative biochemistry of recombinant V. carteri MCP
We proceeded to quantify the amount of protein bound Moco/MPT via HPLC-based FormA analysis, which revealed a Moco/MPT binding stoichiometry of ∼0.21 ± 0.01 for recombinant V. carteri MCP (Figure 4B). The metal-free Moco precursor MPT was not co-purified with recombinant V. carteri MCP as documented by quantitative FormA analysis of V. carteri MCP derived from the Moco free but MPT accumulating strain RK5206 (Figure 4B). To give additional evidence for V. carteri MCP Moco selectivity, we carried out an ICP-MS analysis. In accordance with our hypothesis, Mo amounts above the experimental/methodical error range defined where only detected in V. carteri MCP preparations derived from strain TP1000 (Figure 4B and Supplementary Figure S1). Therefore, we concluded, that V. carteri MCP binds Moco but not MPT. As observed previously for C. reinhardtii MCP , when no molybdate was supplemented to the expression medium no Moco was detected bound to MCP (Figure 4B).
V. carteri MCP Moco stabilising properties
The stabilisation of Moco by V. carteri MCP was surveyed. To this end, the protein purified from E. coli TP1000  and hence saturated with Moco (∼0.21 ± 0.01, Figure 4B) was incubated at different temperatures (4, 22 and 60°C) for 160 min. Determination of the protein concentration revealed that, in accordance with thermal shift assay results (Supplementary Figure S2 and Table S1), there was no loss in the MCP amounts over the course of 160 min regardless of the incubation temperature. Furthermore, FormA quantification revealed that there was barely any loss in Moco amounts at 4°C and 22°C whereas at 60°C V. carteri MCP-bound Moco diminished to 0.00 over the course of 160 min (Figure 5A).
Moco stabilising properties of the V. carteri MCP
Functional complementation of the N. crassa NR using recombinant V. carteri MCP
To provide evidence that V. carteri MCP is fully capable of donating Moco to user enzymes, we carried out nit-1 assay  to qualitatively detect physiologically active Moco. This revealed that V. carteri MCP is fully capable of donating Moco to the Moco-free NR present in N. crassa nit-1 strain crude cell extracts  as documented by the fact that Moco-dependent enzymatic activity is reconstituted successfully (Figure 5B). Furthermore, the successful reconstitution even without the addition of molybdate proves that significant amounts of physiologically active Moco is bound to the protein.
Moco binding to V. carteri MCP
The amino acid sequence comparison (Figure 1A) of eukaryotic MCPs revealed a region of low sequence homology (LHR) located in the N-terminal part of the protein sequence. Results of docking experiments published for C. reinhardtii MCP  as well as a bacterial MCP from R. orientalis  indicate that Moco binding probably occurs in a predominantly positively charged crevice that is located in proximity to the LHR. We therefore investigated the LHR’s role by creating a chimeric V. carteri MCP, which had its endogenous LHR switched for its counterpart from R. orientalis MCP (Figure 6A). Earlier reports indicated that recombinant R. orientalis MCP purified from E. coli (Figure 6B) has a significantly lower saturation with Moco  than what we report here for V. carteri MCP. Accordingly, compared with wildtype V. carteri MCP (Figure 6C), the chimera showed a significantly reduced Moco saturation (∼0.03, Figure 6C), that was comparable with the saturation reported for R. orientalis MCP .
Characterisation of the chimeric V. carteri MCP variant Vc MCPRo LHR
Hitherto, the cellular Moco transfer system has been fundamentally characterised in the higher plant A. thaliana [14,15] and in the green alga C. reinhardtii [9–13]. Progress in understanding how Moco is transferred from source to sink involved the characterisation of in vivo protein–protein interactions  which were complemented by the biochemical characterisation of MCP and MoBP proteins respectively [13,14], though as yet it is not known how Moco binding and transfer to the cognate user enzymes are implemented on the structural level. Progress in genome sequencing and annotation has allowed us to identify, as of yet, uncharacterised eukaryotic C. reinhardtii MCP homologues, from which one identified in the alga V. carteri was characterised within the present study.
As observed for C. reinhardtii MCP  also V. carteri MCP was found to bind Moco selectively over the metal-free Moco precursor MPT which was shown by the HPLC-based detection of the pterin backbone as well as by the ICP-MS based detection of the Moco molybdenum component. Consistent results were obtained from the nit-1 based reconstitution and the Moco stabilisation assay and hence we conclude that V. carteri MCP possesses overall comparable characteristics to C. reinhardtii MCP . V. carteri MCP is the second eukaryotic MCP described and given the fact, that MCPs share overall high levels of sequence and structural similarities (this work, [13,22]), we assume that these have also a conserved Moco binding site. To verify this hypothesis for V. carteri MCP, we carried out one-to-one threading using the Phyre2 server  to create two models based on the structural templates C. reinhardtii MCP  and R. orientalis MCP (the first structurally resolved prokaryotic MCP ) respectively. Our analysis showed that both models have a surface charge distribution featuring a positively charged area in a crevice, which was already earlier suggested to be the C. reinhardtii MCP Moco binding site  (Figure 7). Further we identified a flexible loop of low sequence homology located in close proximity to the putative Moco binding site which we found to be of crucial importance for the V. carteri MCP Moco binding capability. As a possible explanation for the low sequence conservation between the loops from various MCPs, we suggest that these are important for the adaption of the MCPs to the existent cellular contexts in the respective host organisms. Consistently, we suggest that this loop may possess a sequence independent lid function.
Proposed Moco binding site of V. carteri MCP
Our combined data hence confirm that the crevice is the Moco binding site common for pro- and eukaryotic MCPs and the LHR loop is essential for Moco binding, probably because it acts as a lid for the Moco binding site.
The data and material presented in this manuscript are available from the corresponding author on reasonable request.
The authors declare that there are no competing interests associated with the manuscript.
This work was supported by the Deutsche Forschungsgemeinschaft [grant number GRK 2223/1 (to Tobias Kruse)]; and the National Institutes of Health [grant number GM-057378 (to Martin L. Kirk)].
Thomas W. Hercher: Acquisition of data, data analysis and interpretation, drafting the article. Joern Krausze: Acquisition of data analysis and interpretation, drafting the article. Jing Yang: Data interpretation and spectroscopic computations. Martin L. Kirk: Data interpretation, drafting the article. Tobias Kruse: Conception and design, analysis and interpretation of data, drafting the article, final approval of the version to be published.
We thank Prof. Dr. Ralf R. Mendel (TU Braunschweig, Institute of Plant Biology) for his continuous support, for carefully reading the manuscript and for stimulating discussions. We also thank Dr. Phillip Ringel (TU Braunschweig, Institute of Plant Biology) for his initial contributions to the project. We thank Franziska Holtkotte (TU Braunschweig, Institute of Plant Biology) for excellent technical assistance.
cyclic pyranopterin monophosphate
inductively coupled plasma mass spectrometry
low homology region
Moco carrier protein
Moco binding protein
small-angle X-ray scattering