Methanobactins (Mbns) are modified peptides that sequester copper (Cu) methanotrophs use to oxidize methane. Limited structural information is available for this class of natural products, as is an understanding of how cells are able to utilize Mbn-bound Cu. The crystal structure of Methylosinus sporium NR3K CuI -Mbn provides further information about the structural diversity of Mbns and the first insight into their Cu-release mechanism. Nitrogen ligands from oxazolone and pyrazinediol rings chelate CuI along with adjacent coordinating sulfurs from thioamides. In vitro solution data are consistent with a CuI -Mbn monomer as found for previously characterized Mbns. In the crystal structure, the N-terminal region has undergone a conformational change allowing the formation of a CuI2 -Mbn2 dimer with CuI sites bound by chelating units from adjacent chains. Such a structural alteration will facilitate CuI release from Mbns.
Methanobactins (Mbns) are modified peptides that sequester copper (Cu) methanotrophs use to oxidize methane. Limited structural information is available for this class of natural products, as is an understanding of how cells are able to utilize Mbn-bound Cu. The crystal structure of Methylosinus sporium NR3K CuI -Mbn provides further information about the structural diversity of Mbns and the first insight into their Cu-release mechanism. Nitrogen ligands from oxazolone and pyrazinediol rings chelate CuI along with adjacent coordinating sulfurs from thioamides. In vitro solution data are consistent with a CuI -Mbn monomer as found for previously characterized Mbns. In the crystal structure, the N-terminal region has undergone a conformational change allowing the formation of a CuI2 -Mbn2 dimer with CuI sites bound by chelating units from adjacent chains. Such a structural alteration will facilitate CuI release from Mbns.
Methanobactins (Mbns) are a rare example of copper (Cu)‐binding natural products, initially identified in methanotrophic bacteria (methanotrophs).1, 2, 3, 4 These organisms require large quantities of Cu to oxidize methane, a reaction of great environmental importance and biotechnological potential,5, 6, 7 using the membrane‐bound particulate methane monooxygenase (pMMO). Mbns are secreted under Cu‐limiting conditions to sequester this metal ion.1, 8, 9 The first Mbn crystallized was that from the model methanotrophMethylosinus trichosporium OB3b,10 and the structure was subsequently modified and improved (Figure 1 a).11, 12 The crystal structures of only two other Mbns, both from Methylocystis strains, are available,9 which exhibit differences to M. trichosporium OB3b Mbn (Figure 1). All Mbns characterized to date bind CuII and CuI, but have a strong preference for the latter with some of the highest known CuI affinities for biological sites.9, 12, 14, 15, 16, 17, 18 This avidity raises the important question of how Mbn‐associated CuI is released within cells.9
Figure 1
Crystal structures of M. trichosporium OB3b (a), Methylocystis hirsuta CSC1 (b), and Methylocystis strain M (c) CuI–Mbns.9, 12 CuI ions are shown as grey spheres and CuI‐ligand and key hydrogen bonds as dashed black and orange lines, respectively. Crystallized M. hirsuta CSC1 CuI–Mbn is missing three C‐terminal residues (Thr9, Asn10, and Gly11).9, 13 Differences between M. trichosporium OB3b and Methylocystis Mbns include a disulfide bond in the former, a Thr sidechain modified with a sulfate group in the Methylocystis Mbns and an N‐terminal pyrazinediol (pyraA) ring coordinating in these Mbns in place of the oxazolone (oxaA) in M. trichosporium OB3b Mbn (oxaB is conserved). The CuI sites are remarkably similar (see Table S1 in the Supporting Information), regardless of these alterations and the different overall folds of M. trichosporium OB3b and the Methylocystis Mbns.
Crystal structures of M. trichosporium OB3b (a), Methylocystis hirsuta CSC1 (b), and Methylocystis strain M (c) CuI–Mbns.9, 12 CuI ions are shown as grey spheres and CuI‐ligand and key hydrogen bonds as dashed black and orange lines, respectively. Crystallized M. hirsuta CSC1 CuI–Mbn is missing three C‐terminal residues (Thr9, Asn10, and Gly11).9, 13 Differences between M. trichosporium OB3b and Methylocystis Mbns include a disulfide bond in the former, a Thr sidechain modified with a sulfate group in the Methylocystis Mbns and an N‐terminal pyrazinediol (pyraA) ring coordinating in these Mbns in place of the oxazolone (oxaA) in M. trichosporium OB3b Mbn (oxaB is conserved). The CuI sites are remarkably similar (see Table S1 in the Supporting Information), regardless of these alterations and the different overall folds of M. trichosporium OB3b and the Methylocystis Mbns.The operon that includes the gene (mbnA) for the Mbn precursor peptide (leader and core sequences), was initially identified in M. trichosporium OB3b, and subsequently in other bacteria, some non‐methanotrophs.1, 2, 19, 20 Genes in this Mbn operon are expected to have roles essential to Mbn production and function, as already shown for MbnN21 and MbnT,22 which are involved in Mbn modification and import, respectively. Bioinformatics of the Mbn operon is limited by the unknown diversity of Mbns. Furthermore, many methanotrophs do not possess the identified Mbn operon, but do have pMMO (present in almost all methanotrophs), including the well‐studied organisms Methylococcus capsulatus (Bath), and Methylomicrobium album BG8. We attempted to isolate and purify Mbn‐like molecules from media in which these two methanotrophs were grown, but without success (Figure S1 in the Supporting Information). Whether alternative Cu‐sequestering molecules are present in methanotrophs that do not possess the Mbn‐operon remains to be established. The genome of Methylosinus sporium NR3K, isolated from Barro Colorado Island, Panama,23 is not available, but we have confirmed the presence of the Mbn operon (Figures S2 and S3 in the Supporting Information). MbnA from this methanotroph is similar to that of M. trichosporium OB3b, but the crystal structure of the Mbn has unusual features and provides the first insight into how Mbns could release Cu within cells.The UV/Vis and mass spectra (Figure 2 a–c and Table S2 in the Supporting Information) of the Mbn isolated from M. sporium NR3K, as well as its Cu‐binding properties (Figure 2 d–h and Figure S4 and Table S2 in the Supporting Information) are similar to those of previously characterized Mbns, and particularly that from M. trichosporium OB3b.1, 9, 12, 19, 24 The data are consistent with a CuI–Mbn monomer in solution (Figure 1).9, 12 Furthermore, they indicate a strong preference for CuI binding (Figure 2 d–h), confirmed by a CuI affinity of ≈4×1020
m
−1 (Figure S5 in the Supporting Information). However, the crystal structure of M. sporium NR3K CuI–Mbn (Figure 3 a, b) reveals surprising features. The molecule possesses two heterocycles, but the ring at the N‐terminus is a pyrazinediol/pyrazinedione (Figure 3 a and Figure S6 in the Supporting Information). Such a six‐membered ring has only been found to date in Methylocystis Mbns (Figure 1 b, c),9 with M. trichosporium OB3b Mbn, whose sequence is very similar to that of M. sporium NR3K Mbn (Figure 3 c), having two five‐membered oxazolones (Figure 1 a).
Figure 2
In vitro analysis of M. sporium NR3K Mbn. a) UV/Vis spectra of the as‐isolated apo‐ (black line) and Cu‐ (green line) forms. Mass spectra of these are shown in (b, negative ionization mode, [M−H]−) and (c, positive ionization mode, [M+Cu]+) respectively. d) Overlay of X‐band EPR spectra (20 K) of as‐isolated Cu–Mbn (2 mm) before (black line) and after (red line) treatment with 10 % nitric acid. The lack of a signal is consistent with the presence of diamagnetic CuI–Mbn, and the addition of nitric acid denatures the Mbn and oxidizes CuI to CuII. Also shown are UV/Vis spectral changes over time after the addition of ≈1 equivalent of CuII (e) and CuI (f) to the apo‐Mbn (22.5 μm). UV/Vis spectral changes upon the titration of up to 1 equivalent of CuII (g) and CuI (h) to the apo‐Mbn (22.5 and 24.0 μm respectively) are also included, with the red spectra being the end points of the titrations. The final products in (e) and (g) have UV/Vis spectra the same as that of CuI–Mbn (a, f and h) demonstrating that CuII is reduced to CuI, as observed for all previously characterized Mbns.1, 9, 12, 19, 24 The titration of up to 1 equivalent of CuI is very similar in the form (25 μm) with the disulfide‐bond reduced (i).
Figure 3
The structure of M. sporium NR3K CuI–Mbn. a) The CuI
2–Mbn2 dimer in the crystal structure made from two symmetry‐related monomers that are coloured green and slate, CuI ions are shown as grey spheres and CuI‐ligand and key hydrogen bonds as dashed black and orange lines, respectively. A close up of one of the CuI sites is shown in (b) with bond distances (Å) in red. c) Comparison of M. sporium NR3K and M. trichosporium OB3b MbnA amino‐acid sequences (see Supporting Figure S2 and S3) with conserved residues in the core peptides red.
In vitro analysis of M. sporium NR3K Mbn. a) UV/Vis spectra of the as‐isolated apo‐ (black line) and Cu‐ (green line) forms. Mass spectra of these are shown in (b, negative ionization mode, [M−H]−) and (c, positive ionization mode, [M+Cu]+) respectively. d) Overlay of X‐band EPR spectra (20 K) of as‐isolated Cu–Mbn (2 mm) before (black line) and after (red line) treatment with 10 % nitric acid. The lack of a signal is consistent with the presence of diamagnetic CuI–Mbn, and the addition of nitric acid denatures the Mbn and oxidizes CuI to CuII. Also shown are UV/Vis spectral changes over time after the addition of ≈1 equivalent of CuII (e) and CuI (f) to the apo‐Mbn (22.5 μm). UV/Vis spectral changes upon the titration of up to 1 equivalent of CuII (g) and CuI (h) to the apo‐Mbn (22.5 and 24.0 μm respectively) are also included, with the red spectra being the end points of the titrations. The final products in (e) and (g) have UV/Vis spectra the same as that of CuI–Mbn (a, f and h) demonstrating that CuII is reduced to CuI, as observed for all previously characterized Mbns.1, 9, 12, 19, 24 The titration of up to 1 equivalent of CuI is very similar in the form (25 μm) with the disulfide‐bond reduced (i).The structure of M. sporium NR3K CuI–Mbn. a) The CuI
2–Mbn2 dimer in the crystal structure made from two symmetry‐related monomers that are coloured green and slate, CuI ions are shown as grey spheres and CuI‐ligand and key hydrogen bonds as dashed black and orange lines, respectively. A close up of one of the CuI sites is shown in (b) with bond distances (Å) in red. c) Comparison of M. sporium NR3K and M. trichosporium OB3b MbnA amino‐acid sequences (see Supporting Figure S2 and S3) with conserved residues in the core peptides red.Most unexpectedly, the CuI–Mbn from M. sporium NR3K has formed a CuI
2–Mbn2 dimer in the crystal structure with intermolecular CuI sites (Figure 3 a, b). This arrangement is very different from the monomers observed in all other CuI–Mbn structures (Figure 1).9, 10, 12 The N‐terminus, containing two of the ligands, has formed an extended, almost linear, polypeptide chain (Figure 3 a), stabilized by hydrogen bonds between the backbone carbonyl and amide groups of Ser4 from adjacent molecules, which mimics a small section of anti‐parallel β‐strand. Adjacent to Ser4 in M. sporium NR3K Mbn is Cys5 that forms a disulfide bond with the C‐terminal Cys12 (Figure 3 a, c). The M. trichosporium OB3b Mbn also possesses a disulfide bond involving Cys5 (Figure 1 a),12 but in this case with the penultimate Cys10 residue (Figure 3 c). Disulfide bond cleavage has a limited influence on M. sporium NR3K apo‐Mbn and its CuI‐binding properties (Figure 2 i and Figure S5 and S7 in the Supporting Information), as observed previously for M. trichosporium OB3b Mbn,12 further highlighting the similarity of these Mbns in solution.The distorted tetrahedral S2N2 CuI sites found in the M. sporium NR3K CuI
2–Mbn2 dimer have two coordinating thioamides/enethiols that form chelating units with nitrogen atoms from adjacent heterocyclic rings (Figure 3 a, b). The C‐terminal nitrogen ligand is derived from the oxazolone group, as in all structurally characterized CuI–Mbns (Figure 1),9, 12 whilst the N‐terminal coordinating nitrogen is provided by the pyrazinediol ring (Figure 3 a, b), as in Methylocystis CuI–Mbns (Figure 1 b, c).9 There are second‐coordination sphere changes at these intermolecular CuI sites (Figure 3 a) compared to other CuI–Mbns (Figure 1). Regardless, these CuI sites (Figure 3 b) have remarkably similar structures to those in other CuI–Mbns (Table S1 in the Supporting Information).9, 12All Mbns investigated have very high affinities for CuI that are in the 1020 to 1021
m
−1 range.9, 12 The values for M. sporium NR3K and M. trichosporium OB3b Mbns are alike (see Figure S5 in the Supporting Information), as are their reduction potentials (E
m values of ≈640 mV for both, see Figure S8 in the Supporting Information). Therefore, their calculated12 affinities for CuII are also similar, albeit ≈8 orders of magnitude weaker than those for CuI. As we have pointed out previously,9, 12 such high CuI affinities imply that removing the metal must present a challenge to a cell. Oxidation of CuI–Mbns could facilitate release due to the lower affinities of Mbns for CuII, but the high E
m values will make conversion to the CuII‐forms difficult in a cell.9The high affinity ligand bathocuproine disulfonate (BCS) is used to measure how tightly Mbns bind CuI (for example, Figure S5 in the Supporting Information). This bulky 1,10‐phenanthroline derivative with methyl groups adjacent to the two coordinating nitrogen atoms is able to remove CuI from both Methylosinus and Methylocystis Mbns,9, 12 and equilibration is relatively fast (occurring in less than 1 min for M. sporium NR3K Mbn). Given the tight CuI affinities of Mbns, removal by BCS cannot take place via a dissociative mechanism (the unassisted off‐rate for CuI would be ≈10−12 s−1), and this large ligand must access the CuI site, with transfer involving a transient Mbn‐CuI‐BCS intermediate. The Atx1 family of cytosolic CuI metallochaperones (vide infra) have also been suggested to form transient Atx1–CuI–ligand intermediates through their CXXC Cu(I)‐binding motif with both BCS,25 and bicinchoninic acid.26 Formation of such intermediates for CuI–Mbns is less straightforward due to them having coordinatively saturated sites (Figure 1). The structure of the CuI
2–Mbn2 dimer provides the first indication of how this can occur. A conformational change at the N‐terminus of M. sporium NR3K CuI–Mbn results in the loss of the pyrazinediol and thioamide ligands (Figure 3 a). A structural change such as this could be triggered by a molecule that is able to coordinate CuI; BCS in affinity measurements or another Mbn molecule as seen at the high concentrations used for CuI–Mbn crystallization. CuI transfer occurs between Mbns in solution (including between M. trichosporium OB3b and Methylocystis Mbns),9 and this must involve Mbn–CuI–Mbn intermediates. Such a species has been stabilized in the crystal structure of the M. sporium NR3K CuI
2–Mbn2 dimer. A related mechanism facilitates CuI transfer from the high affinity,15, 16, 18 but in this case coordinatively unsaturated, sites of Atx1s to the metal‐binding domains of CuI‐transporting P‐type ATPases.27, 28, 29 Although humanAtx1 (ATOX1) is a monomer in solution with a two‐coordinate CuI site bound via its CXXC motif, in the crystal structure it forms a dimer with an intermolecular tetrahedral CuI‐S(Cys)4 site.28 This structure provided strong support for the suggested ligand‐exchange mechanism of CuI transfer,27 which was later shown to be correct.29The crystal structure of M. sporium NR3K CuI–Mbn demonstrates for the first time that this family of molecules can undergo a conformational change with the metal ion bound. This involves the region of the molecule prior to the first Cys residue of its disulfide bond. Most of the core MbnA peptides from Methylosinus strains (Figure S3 in the Supporting Information) exhibit sequence similarities that suggest they will release CuI via a related mechanism. For example, all have a Cys at either position four or five, and also as the penultimate or C‐terminal amino acid (in addition to the two Cys residues that are modified), which probably also form a disulfide bond (Figure 1 a and Figure 3 a).12 This could be important for minimising structural changes during CuI release. Alternatively, although disulfide bond cleavage does not have a very large effect on CuI affinity (lowers it by ≈2 orders of magnitude), it could assist CuI release by enabling a larger conformational change if required. A couple of Methylocystis Mbns are predicted to have a disulfide bond (Figure S3 in the Supporting Information), but most only have the two Cys residues that provide the sulfur ligands (Figure 1 b, c). Many of the Mbns from these strains also have comparable sequences, and those from M. hirsuta CSC1 and Methylocystis strain M have similar structures,9 with an altered overall fold compared to M. trichosporium OB3b Mbn (Figure 1). The conformational change they will undergo to facilitate CuI release may therefore be different to Methylosinus Mbns. The sulfate group on a modified Thr sidechain adjacent to the C‐terminal oxazolone group in Methylocystis Mbns (Figure 1 b, c), which can be removed,9 may play a role in this process. The presence of an N‐terminal coordinating pyrazinediol ring rather than an oxazolone could influence how CuI is released from Mbns. However, both provide a heterocyclic nitrogen ligand, are found in Methylosinus and Methylocystis Mbns and may be more important for recognition with an interacting partner.Mbns are secreted to sequester Cu and have therefore evolved to bind this metal ion tightly and to not release their cargo extracellularly. However, this makes removing Cu from Mbn within a cell thermodynamically and, considering the structures of monomeric CuI–Mbns (Figure 1), also kinetically unfavorable. In this study we present the crystal structure of a CuI
2–Mbn2 dimer whose formation is only possible due to a conformational change at the N‐terminus. This highlights how CuI release may be facilitated. Understanding this process is not only important for Cu accumulation in bacteria that secrete Mbns, but also potentially for their development and successful use in the treatment of diseases caused by Cu accumulation.30 The next challenge is to identify how conformational‐change‐assisted CuI release from Mbns occurs within cells.
Conflict of interest
The authors declare no conflict of interest.As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.SupplementaryClick here for additional data file.
Authors: R A Pufahl; C P Singer; K L Peariso; S J Lin; P J Schmidt; C J Fahrni; V C Culotta; J E Penner-Hahn; T V O'Halloran Journal: Science Date: 1997-10-31 Impact factor: 47.728
Authors: Josef Lichtmannegger; Christin Leitzinger; Ralf Wimmer; Sabine Schmitt; Sabine Schulz; Yaschar Kabiri; Carola Eberhagen; Tamara Rieder; Dirk Janik; Frauke Neff; Beate K Straub; Peter Schirmacher; Alan A DiSpirito; Nathan Bandow; Bipin S Baral; Andrew Flatley; Elisabeth Kremmer; Gerald Denk; Florian P Reiter; Simon Hohenester; Friedericke Eckardt-Schupp; Norbert A Dencher; Jerzy Adamski; Vanessa Sauer; Christoph Niemietz; Hartmut H J Schmidt; Uta Merle; Daniel Nils Gotthardt; Guido Kroemer; Karl Heinz Weiss; Hans Zischka Journal: J Clin Invest Date: 2016-06-20 Impact factor: 14.808
Authors: Alan A DiSpirito; Jeremy D Semrau; J Colin Murrell; Warren H Gallagher; Christopher Dennison; Stéphane Vuilleumier Journal: Microbiol Mol Biol Rev Date: 2016-03-16 Impact factor: 11.056
Authors: Abdelnasser El Ghazouani; Arnaud Baslé; Susan J Firbank; Charles W Knapp; Joe Gray; David W Graham; Christopher Dennison Journal: Inorg Chem Date: 2011-01-21 Impact factor: 5.165
Authors: Grace E Kenney; Anthony W Goering; Matthew O Ross; Caroline J DeHart; Paul M Thomas; Brian M Hoffman; Neil L Kelleher; Amy C Rosenzweig Journal: J Am Chem Soc Date: 2016-08-26 Impact factor: 15.419
Authors: Grace E Kenney; Laura M K Dassama; Anastasia C Manesis; Matthew O Ross; Siyu Chen; Brian M Hoffman; Amy C Rosenzweig Journal: J Biol Chem Date: 2019-09-11 Impact factor: 5.157
Authors: Yun Ji Park; Gerri M Roberts; Rana Montaser; Grace E Kenney; Paul M Thomas; Neil L Kelleher; Amy C Rosenzweig Journal: Biochemistry Date: 2021-09-12 Impact factor: 3.321
Authors: Anastasia C Manesis; Richard J Jodts; Brian M Hoffman; Amy C Rosenzweig Journal: Proc Natl Acad Sci U S A Date: 2021-06-08 Impact factor: 11.205
Authors: Obulisamy Parthiba Karthikeyan; Thomas J Smith; Shamsudeen Umar Dandare; Kamaludeen Sara Parwin; Heetasmin Singh; Hui Xin Loh; Mark R Cunningham; Paul Nicholas Williams; Tim Nichol; Avudainayagam Subramanian; Kumarasamy Ramasamy; Deepak Kumaresan Journal: Microbiome Date: 2021-07-06 Impact factor: 16.837
Authors: Yun Ji Park; Richard J Jodts; Jeffrey W Slater; Reyvin M Reyes; Valerie J Winton; Rana A Montaser; Paul M Thomas; William B Dowdle; Anahi Ruiz; Neil L Kelleher; J Martin Bollinger; Carsten Krebs; Brian M Hoffman; Amy C Rosenzweig Journal: Proc Natl Acad Sci U S A Date: 2022-03-23 Impact factor: 12.779