MbnC Is Not Required for the Formation of the N-Terminal Oxazolone in the Methanobactin from Methylosinus trichosporium OB3b

ABSTRACT Methanobactins (MBs) are ribosomally synthesized and posttranslationally modified peptides (RiPPs) produced by methanotrophs for copper uptake. The posttranslational modification that defines MBs is the formation of two heterocyclic groups with associated thioamines from X-Cys dipeptide sequences. Both heterocyclic groups in the MB from Methylosinus trichosporium OB3b (MB-OB3b) are oxazolone groups. The precursor gene for MB-OB3b is mbnA, which is part of a gene cluster that contains both annotated and unannotated genes. One of those unannotated genes, mbnC, is found in all MB operons and, in conjunction with mbnB, is reported to be involved in the formation of both heterocyclic groups in all MBs. To determine the function of mbnC, a deletion mutation was constructed in M. trichosporium OB3b, and the MB produced from the ΔmbnC mutant was purified and structurally characterized by UV-visible absorption spectroscopy, mass spectrometry, and solution nuclear magnetic resonance (NMR) spectroscopy. MB-OB3b from the ΔmbnC mutant was missing the C-terminal Met and was also found to contain a Pro and a Cys in place of the pyrrolidinyl-oxazolone-thioamide group. These results demonstrate MbnC is required for the formation of the C-terminal pyrrolidinyl-oxazolone-thioamide group from the Pro-Cys dipeptide, but not for the formation of the N-terminal 3-methylbutanol-oxazolone-thioamide group from the N-terminal dipeptide Leu-Cys. IMPORTANCE A number of environmental and medical applications have been proposed for MBs, including bioremediation of toxic metals and nanoparticle formation, as well as the treatment of copper- and iron-related diseases. However, before MBs can be modified and optimized for any specific application, the biosynthetic pathway for MB production must be defined. The discovery that mbnC is involved in the formation of the C-terminal oxazolone group with associated thioamide but not for the formation of the N-terminal oxazolone group with associated thioamide in M. trichosporium OB3b suggests the enzymes responsible for posttranslational modification(s) of the two oxazolone groups are not identical.

The gene encoding the MB precursor peptide, mbnA (5,10), is found in a gene cluster that contains both genes of known function, such as mbnB (5,11), mbnN (9), and mbnT (12), as well as unannotated genes, such as mbnC (5,10,11,13,14). MbnB is a member of TIM barrel family as well as the DUF692 family of diiron enzymes (11,14). In heterologous expression studies in Escherichia coli, MbnBC was shown to catalyze a dioxygen-dependent four-electron oxidation of Pro-Cys in MbnA (11,14,15). The roles of MbnB and MbnC could not be separately determined as attempts to separately purify these gene products in E. coli failed (11). From these data, it has been argued that MbnBC must act in concert and by doing so create both heterocyclic groups in MBs (11). Such conclusions, however, appear to be premature for several reasons. First, the reported spectra (11) only show the presence of the C-terminal oxazolone group, not the N-terminal oxazolone group, as the 394-nm absorption maximum is missing. Second, the absorption maximum at 302 nm, diagnostic for the presence of the N-terminal oxazolone group, was absent (5,8,16). Third, no structural data were provided to support the presence of both oxazolone groups. To examine if MbnB and -C act in concert and are involved in the formation of both oxazolone groups in M. trichosporium OB3b, an MnbC deletion mutant (DmbnC) was constructed. The results show MbnC is required for the formation of the C-terminal oxazolone group, but not for the formation of the N-terminal oxazolone group.

RESULTS
Generation of the DmbnC mutant. The previously constructed DmbnAN strain, whereby the mbnABCMN genes were deleted using a sucrose counterselection technique (9), was back complemented with mbnABMN through selective amplification and ligation of mbnAB with mbnMN, deleting mbnC, and inserting this ligation product into pTJS140, creating pWG104 (Table 1). Successful removal of mbnC from this product was confirmed via sequencing (data not shown). The native s 70 -dependent promoter upstream of mbnA was also incorporated into pWG104, and expression of mbnABMN but not mbnC (from pWG104), as well as mbnPH (from the chromosome) was confirmed via reverse transcription-PCR (RT-PCR) (see Fig. S1 and S2 in the supplemental material).
UV-visible absorption and mass spectrometry of metal-free MB from M. trichosporium OB3b DmbnC. Comparison of the UV-visible absorption spectra of MB from M. trichosporium OB3b DmbnC to wild-type MB-OB3b suggested the of presence of the N-terminal oxazolone group, but the absence of C-terminal oxazolone ( Fig. 1: see Fig.  S3 in the supplemental material). The molecular mass of native, full-length MB-OB3b is 1,154 Da, and that of MB-OB3b lacking the C-terminal Met is 1,023 Da. It should be noted that both forms of MB-OB3b are present in most MB-OB3b preparations (2,5,17). The molecular mass of DMbnC was 1,024 as determined by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) (Fig. 2) (9), which was within 1 Da of the predicted molecular mass of MB-OB3b, in which only one oxazolone group was formed. Taken together, the UV-visible absorption spectra and molecular mass data suggest the DmbnC mutant lacked the C-terminal Met as well as the N-terminal oxazolone group with a 1-(N-[mercapto-(5-oxo-2-(3-methylbutanoyl)oxazol-(Z)-4-ylidene)methyl]-GSCYPCSC predicted structure (Fig. 3B). In contrast to wild-type MB-OB3b, the C-terminal Met was never observed in MbnC.
Chemical structure of metal-free DmbnC mutant as determined by NMR spectroscopy. Metal-free MB has multiple conformations, making structural studies of MBs via solution nuclear magnetic resonance (NMR) or crystallography difficult (see Fig. S4 in the supplemental material). In prior structural studies of MB, the addition of Cu 21 (which is bound and reduced to Cu 11 by native MB-OB3b) stabilizes MB-OB3b into one conformation, allowing for crystal formation and NMR characterization ( Fig. S4) (2-5, 8, 18). Our initial efforts to investigate the structure of the MB intermediate produced by the DmbnC strain via NMR were unsuccessful.  Fig. S5 in the supplemental material). In order to slow down the rate of exchange and reduce line broadening, we sampled various temperature and hydrostatic pressure conditions. We found that two-dimensional (2D) 1 H-15 N NMR spectra of the DmbnC mutant recorded at high pressure (300,000,000 Pa) and low temperature (265 K) (18,19) show significantly reduced line broadening and gave excellent spectra in the absence of copper (Fig. 4).
A series of NMR experiments were conducted on the DmbnC mutant, including homonuclear correlation spectroscopy (COSY), total correlation spectroscopy (TOCSY), rotating-frame   (Fig. 4). The 1D 15 N experiment showed a peak at 109 ppm that was absent from the 1 H-15 N HSQC spectra and was assigned to proline. However, the glycine nitrogen peak was especially broad, and could only be assigned with the 1 H-15 N HSQC. Finally, while the 1D 15 N experiment had several resonances around 180 ppm-likely due to hydrolysis and deprotonation-only one of them had a correlation with 1 H in the 1 H-15 N HSQC, indicating a single oxazolone group. The NMR results are consistent with the UV-visible absorption spectra and the ESI MS results, as well as with the structure shown in Fig. 3B.

DISCUSSION
Due to the variability in the core sequences of structurally characterized MBs, it is difficult to use mbnA to screen the potential ability of microbes to produce MB. Instead, mbnB and mbnC sequences are commonly used as they are found in all known mbn gene clusters (5,13). All known MBs contain two heterocyclic rings, with the N-terminal ring found to be either an oxazolone, pyrazinedione, or imidazolone ring, while the C-terminal ring was always found to be an oxazolone. Given these data, it could be presumed that MbnBC is involved in the formation of the C-terminal oxazolone group along with an associated thioamide, while the N-terminal oxazolone group is formed via a different process, such as the involvement of an aminotransferase, as concluded earlier (5,9,10,13).
Other researchers have attempted to elucidate the role of MbnB and MbnC in methanobactin maturation (11). These individuals were unable to separately heterologously express soluble protein from either MbnB or MbnC, but were able to coheterologously express MbnBC as a heterodimeric complex. In studies where the MbnA precursor polypeptide was incubated with this MbnBC complex, the authors conclude that MbnBC was involved in the formation of both oxazolone groups and the associated thioamides of MB-OB3b. It should be noted, however, that in this study, no structural evidence (i.e., solution NMR data) was provided to definitively show the presence of either ring: rather, such conclusions were largely based on mass spectral analyses of MbnA after incubation with the MbnBC complex. Further, the authors assumed that since their construct did not contain the N-terminal aminotransferase MbnN, the extended conjugation resulting from this reaction would result in both oxazolone groups having identical absorption maxima. The idea that the extended conjugation of the N-terminal oxazolone could be responsible for the bathochromic shift was first proposed as a possible reason for the 50-nm shift in the absorption maxima by Krentz et al. (5). Kenney et al. used this theory to bolster their claim that both oxazolone groups were present in the product from their heterologous system, with both oxazolone groups showing identical absorption spectra (11). The evidence to support this claim came from their M. trichosporium OB3b DmbnN strain. MbnN is responsible for the deamination of the N-terminal Leu in M. trichosporium OB3b, extending the conjugation one additional double bond. In this study, the authors claim they can stabilize the MB produced by the DmbnN strain by the addition of copper before purification. UVvisible absorption spectra of copper-containing DmbnN mutant suggest the possible presence of two oxazolone groups but no additional evidence was provided supporting this claim. This observation was surprising as the MB produced by the DmbnN strain in our laboratory showed similar UV-visible absorption spectra throughout the growth cycle, suggesting the absence of the N-terminal oxazolone group (see Fig. S7 in the supplemental material). In addition, the UV-visible absorption spectra, liquid chromatography (LC)-MS/MS, Fourier transform ion cyclotron resonance (FT-ICR) MS, amino acid analysis, number of thiol groups, copper-binding properties, and pattern of acid hydrolysis demonstrate the absence of the N-terminal oxazolone group in the DmbnN mutant (9).
Additional evidence that the bathochromic shift in MBs with two oxazolone groups is unlikely to solely arise from the addition of one double bond following deamination of the N-terminal amine comes from examination of the group I MB from Methylocystis parvus OBBP. Acid hydrolysis of the MB from M. parvus OBBP shows a similar hydrolysis pattern to that observed with the MB from M. trichosporium OB3b, demonstrating the presence of two oxazolone groups, with absorption maxima at 340 and 390 nm (see Fig. S8 in the supplemental material). However, both MB operons from M. parvus OBBP lack mbnN, and without deamination of the N-terminal Phe, the conjugation around the N-terminal oxazolone group would not be extended. It is possible that another aminotransferase in the M. parvus OBBP genome may catalyze deamination of the Nterminal Phe. However, this appears unlikely as deamination of the N-terminal amino acid has never been observed in structurally characterized MBs from operons lacking mbnN (3,5). The results suggest deamination of the N-terminal amino acid is not solely responsible for the 40-to 50-nm absorption maximum difference between oxazolone groups in MBs. The absence of either the N-terminal or C-terminal oxazolone group in a small (0.5 to 2%) fraction of most MB-OB3b preparations (Fig. S3) also questions the suggestion that the absorption maximum difference between the N-terminal and Cterminal oxazolone groups is due solely to extending the conjugation of an additional double bond introduced following the deamination reaction.
The results presented here confirm MbnC is required for the formation of the C-terminal oxazolone group (Fig. 5). However, the results presented here also demonstrate MbnC is not required for the formation of the N-terminal oxazolone group in M. trichosporium OB3b, suggesting the formation of the two hetercyclic groups with associated thioamides from XC dipeptides does not utilize the same enzyme(s). Future studies will determine if MbnB is involved in the formation of the N-terminal oxazolone, pyranzinedione, or imidazolone groups. Resolution of the pathway and enzymes responsible for the posttranslational modifications required for the synthesis of MB in methanotrophic bacteria will aid in the production of MB derivatives with pharmacological properties specific for different metal-related diseases (19-24) as well as for environmental applications (10,25).

MATERIALS AND METHODS
Bacterial strains, growth media, and culture conditions. Plasmid construction was accomplished using Escherichia coli strain TOP10 (Invitrogen, Carlsbad, CA) as described previously (9). Plasmids used and constructed during this study are shown in Table 1. The donor strain for conjugation of plasmids into Methylosinus trichosporium OB3b was E. coli S17-1 (26). E. coli strains were cultivated at 37°C in Luria broth medium (Dot Scientific, Burton, MI). Methanotrophic strains (i.e., M. trichosporium OB3b wild type, M. trichosporium OB3b DmbnAN, M. trichosporium OB3b DmbnC, Methylocystis sp. strain SB2, and Methylocystis parvus OBBP) were cultivated at 30°C on nitrate mineral salts (NMS) medium (27), either in 250-mL flasks with side-arms at 200 rpm or in a 15-L New Brunswick Bioflow 310 fermenter (Eppendorf, Hauppauge, NY), using methane as the sole carbon and energy source. Where necessary, filter-sterilized solutions of copper (as CuCl 2 ) and spectinomycin were added to culture media aseptically. A working concentration of 20 mg General DNA methods, transformation, and conjugation. DNA purification and plasmid extraction were performed using QIAquick and QIAprep kits from Qiagen following the manufacturer's instruction. DNA cloning, preparation of chemically competent cells, and plasmid transformation with E. coli were performed according to reference 28. Enzymes used for restriction digestion and ligation were purchased from New England Biolabs (Ipswich, MA). PCR of DNA for cloning purposes was accomplished using iProof high-fidelity polymerase (Bio-Rad, Hercules, CA). PCR for general purposes was Additional, yet to be identified genes may also be involved in the formation of oxazolone groups.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 1.8 MB.