Research Article

Two-Component Flavin-Dependent Riboflavin Monooxygenase Degrades Riboflavin in Devosia riboflavina

Hiroshi Kanazawa, Ryosuke Shigemoto, Yukie Kawasaki, Ken-Ichi Oinuma, Akira Nakamura, Shunsuke Masuo, Naoki Takaya
William W. Metcalf, Editor
DOI: 10.1128/JB.00022-18

ABSTRACT

The actinobacterium Microbacterium maritypicum splits riboflavin (vitamin B2) into lumichrome and d-ribose. However, such degradation by other bacteria and the involvement of a two-component flavin-dependent monooxygenase (FMO) in the reaction remain unknown. Here we investigated the mechanism of riboflavin degradation by the riboflavin-assimilating alphaproteobacterium Devosia riboflavina (formerly Pseudomonas riboflavina). We found that adding riboflavin to bacterial cultures induced riboflavin-degrading activity and a protein of the FMO family that had 67% amino acid identity with the predicted riboflavin hydrolase (RcaE) of M. maritypicum MF109. The D. riboflavina genome clustered genes encoding the predicted FMO, flavin reductase (FR), ribokinase, and flavokinase, and riboflavin induced their expression. This finding suggests that these genes constitute a mechanism for utilizing riboflavin as a carbon source. Recombinant FMO (rFMO) protein of D. riboflavina oxidized riboflavin in the presence of reduced flavin mononucleotide (FMN) provided by recombinant FR (rFR), oxidized FMN and NADH, and produced stoichiometric amounts of lumichrome and d-ribose. Further investigation of the enzymatic properties of D. riboflavina rFMO indicated that rFMO-rFR coupling accompanied O2 consumption and the generation of enzyme-bound hydroperoxy-FMN, which are characteristic of two-component FMOs. These results suggest that D. riboflavina FMO is involved in hydroperoxy-FMN-dependent mechanisms to oxygenize riboflavin and a riboflavin monooxygenase is necessary for the initial step of riboflavin degradation.

IMPORTANCE Whether bacteria utilize either a monooxygenase or a hydrolase for riboflavin degradation has remained obscure. The present study found that a novel riboflavin monooxygenase, not riboflavin hydrolase, facilitated this process in D. riboflavina. The riboflavin monooxygenase gene was clustered with flavin reductase, flavokinase, and ribokinase genes, and riboflavin induced their expression and riboflavin-degrading activity. The gene cluster is uniquely distributed in Devosia species and actinobacteria, which have exploited an environmental niche by developing adaptive mechanisms for riboflavin utilization.

INTRODUCTION

Vitamins are essential compounds for human activity, since they function as cofactors or precursors of enzymes that are involved in essential metabolic pathways and physiological processes. Some vitamins are dietary supplements that have received considerable focus from the perspectives of nutrition and health sciences, in terms of preventing deficiency-related diseases (1–6). Plants and microorganisms naturally produce vitamins, which are classified into groups A, B, C, D, E, and K according to their chemical and physiological properties. Extensive studies over several decades have revealed the biosynthetic mechanisms of most vitamins (2, 7), whereas their environmental degradation, which is an indispensable cue for molecular homeostasis on the planet, has received less attention than their biosynthesis and physiological functions. For example, the group of water-soluble B vitamins includes thiamine, riboflavin, niacin, pantothenic acid, pyridoxine, folate, and cobalamins, which are diverse chemicals with a range of physiological functions (2), yet many of their degradative enzymes and encoding genes remain controversial.

Riboflavin (vitamin B2) is a yellow, water-soluble isoalloxazine with a covalently bound ribitol moiety at the N10 atom, and it is a precursor of intracellular flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD), both of which are cofactors of cellular redox enzymes and other components of respiratory chains (8). Riboflavin kinase phosphorylates riboflavin to FMN, which is then adenylated to FAD (8). A riboflavin deficiency (ariboflavinosis) damages mucous membranes and leads to neurodegeneration, cancer, and cardiovascular diseases in association with other vitamin deficiencies (8). Plants (9), fungi (10), and bacteria all synthesize riboflavin using GTP as a precursor (10). The mechanisms of Bacillaceae bacteria include the generation of isoalloxazine derived from a guanine base [5-amino-6-ribityl-amino-2,4(1H,3H)-pyrimidinedione], which are metabolized by lumazine and riboflavin synthases to yield riboflavin (10). As a salvage mechanism, many bacterial species take up riboflavin to produce FMN and FAD via flavokinase (FK) and FMN/FAD synthase. Since these cofactors, together with riboflavin, are major metabolites produced in amounts of >0.2 mM by Escherichia coli cells fed glucose (11), this mechanism appears to be an economically significant salvage process in which these cofactors are recycled.

In contrast to extensive studies of riboflavin biosynthesis, little is known about riboflavin biodegradation, although it is known that riboflavin is abiotically photodegraded to lumichrome (6,7-dimethylisoalloxadine) when exposed to relatively low levels of light (12) The riboflavin-degrading soil alphaproteobacterium Devosia riboflavina (formerly known as Pseudomonas riboflavina) was originally isolated in 1944 (13) and was reclassified in 1996 by Nakagawa et al. (14). The degradation product of riboflavin is lumichrome. A study of the cell-free bacterial enzyme indicated that another degradation product is ribitol (15); therefore, the proposed initial reaction in riboflavin degradation is thought to be catalysis by riboflavin hydrolase, although this particular enzyme has not been purified to homogeneity and the gene involved in this process has not been clarified. Another riboflavin degradation mechanism, in which d-ribose remains bound to the heterocycle throughout the early stages of alloxazine degradation, has been proposed for this bacterium (16). The molecular mechanism of riboflavin degradation has remained obscure for 70 years. However, the actinobacterium Microbacterium maritypicum G10 was recently isolated from a riboflavin production plant, and the rca gene cluster for bacterial riboflavin catabolism was identified, although nucleotide sequences were not published (17). The RcaE protein encoded by the gene cluster is thought to be riboflavin hydrolase, but the reaction between recombinant RcaE and riboflavin produced only d-ribose in addition to lumichrome, indicating that this is not a simple hydrolase reaction that produces ribitol (15). Sequence analysis has predicted rca gene clusters in other microbacterial genomes, in which the predicted RcaE protein is an N5,N10-methylene tetrahydromethanopterin reductase and not a hydrolase. This suggests that oxidative cleavage of riboflavin by the enzyme would be responsible for generating lumichrome and d-ribose, which would be a monooxygenase reaction that does not involve a water molecule as a catalytic substrate, unlike hydrolase reactions. However, this notion awaits further investigation.

The present study revisited the D. riboflavina mechanism involved in riboflavin degradation and identified flavin-dependent monooxygenase (FMO), a riboflavin monooxygenase with 67% amino acid identity with the predicted riboflavin hydrolase RcaE of M. maritypicum MF109. Enzymatic studies of recombinant FMO (rFMO) and flavin reductase (FR) indicated that they constitute a two-component monooxygenase system that utilizes riboflavin, NADH, and O2 as substrates to produce lumichrome and d-ribose. The electron mediator FMN generates FMO-bound hydroperoxy-FMN (FMN-OOH), which is typical of two-component FMO. These results showed that D. riboflavina FMO oxygenizes riboflavin in the initial step of riboflavin degradation. Furthermore, genes encoding FMO, FR, ribokinase (RK), and FK, which are clustered in the D. riboflavina genome, are inducible by riboflavin, suggesting that they are involved in mechanisms that allow the alphaproteobacterium to use riboflavin as a carbon source.

RESULTS

Degradation of riboflavin and related compounds by D. riboflavina.We cultured Devosia riboflavina strain JCM 21244 (strain IFO13584) (originally isolated as Pseudomonas riboflavina in 1944 [13]) in flasks containing minimal medium with riboflavin as the sole carbon source. The cultures degraded riboflavin and produced lumichrome, which accumulated as a yellowish-white precipitate in the flasks (Fig. 1A). Lumichrome production was confirmed by liquid chromatography-mass spectrometry (LC-MS), which detected [M+H]+ at m/z 243.2; this matched the molecular mass of lumichrome, which is 242.2 (see Fig. S1 in the supplemental material). Cell growth monitored by counting of viable cells increased during the culture period, along with lumichrome production. The culture was protected from light, and control experiments generated <0.1 mM lumichrome in the absence of cells, indicating that D. riboflavina catabolized riboflavin, as described previously (13, 15). We cultured D. riboflavina JCM 21224 in minimal medium containing other carbon sources, instead of riboflavin. The results indicated that adding d-ribose as the carbon source generated the same amount of bacterial cells as riboflavin (Fig. 1B). Lumichrome, FMN, FAD, and ribitol supported slight increases in bacterial cells, but the numbers did not significantly differ from cultures without an added carbon source, indicating that the inoculated preculture media supported growth and D. riboflavin did not assimilate any of those compounds. Flavin mononucleotide did not serve as a carbon source, probably because of deficient uptake by the bacteria, but this concept awaits confirmation. These results agreed with the notion that D. riboflavina degraded riboflavin to lumichrome and d-ribose, the latter of which is a bacterial source of carbon and energy. Notably, this notion could revise the proposal by Yanagita and Foster that bacterial riboflavin is degraded to ribitol (15), but it is consistent with the proposal that the M. maritypicum G10 mechanism generates d-ribose (17). We then further investigated the mechanism of riboflavin degradation by D. riboflavina.

FIG 1
FIG 1

Degradation of riboflavin by D. riboflavina JCM 21244. (A) Degradation of riboflavin by D. riboflavina JCM 21244 cultured at 30°C in minimal medium, in the dark. Viable cells (●) were monitored by counting colonies on plates. Riboflavin (■) and lumichrome (▲) levels were measured by HPLC. (B) D. riboflavina cultured at 30°C in minimal medium containing 5 mM (each) riboflavin (●), d-ribose (■), lumichrome (▲), ribitol (○), FMN (□), FAD (♢), or none (△) as a carbon source. Typical data are presented. Standard errors for the measurements are within the symbols. (C) Riboflavin degradation by supernatant (Sup) and pellet (Pellet) fractions of D. riboflavina cell extracts cultured in the presence (+) or absence (−) of riboflavin. Data are means ± standard errors. *, P < 0.01. (D) Image showing the pellet fraction in the presence (+) or absence (−) of riboflavin on the SDS-PAGE gel. Arrowhead, protein band generated by adding riboflavin to the culture. M, molecular size markers.

Identification of riboflavin-degrading activity and FMO.Devosia riboflavina was cultured in minimal medium with or without riboflavin as a carbon source, and then supernatants and pellets of cell extracts were prepared. The soluble fractions had minimal activity for riboflavin-dependent lumichrome production (1.0 ± 0.4 nmol min−1 mg−1), regardless of the presence or absence of riboflavin (Fig. 1C). The pellet fractions of cells cultured with and without riboflavin had higher (14 ± 1 nmol min−1 mg−1) and minimal activities, respectively. These findings indicated that the riboflavin-degrading activity is associated with the pellet fraction and is inducible by riboflavin in the culture medium. Notably, FMN did not induce activity. Images of SDS-PAGE gels revealed that a protein band migrating at 45 kDa was generated by the pellet fractions of cells cultured with, but not without, riboflavin (Fig. 1D). Peptide mass fingerprinting of tryptic fragments of this protein and database searches found a hypothetical protein (GenBank accession number WP_035082397.1) from the same bacterial strain (18) that matched with a score of 97, an E value of 0.022, and coverage of 34%. The calculated molecular mass of the hypothetical protein (molecular mass of 49,985 Da) agreed with that estimated from the protein band on the gel image.

Similarity searches showed that the protein is a luciferase-like monooxygenase (LLM) class FMO that is similar (67% identity at the amino acid level) to the predicted RcaE protein of M. maritypicum MF109 (GenBank accession number WP_021201082.1). The D. riboflavina FMO gene clustered with genes in the D. riboflavina genome encoding predicted FR (RcaB; GenBank accession number WP_084587600.1), FK (RcaA; GenBank accession number WP_084587592.1), and RK (RcaD) (GenBank accession number WP_035082393.1), which had 49%, 47%, and 38% amino acid sequence identities, respectively, with their M. maritypicum MF109 counterparts (Fig. 2A). These clusters also involved genes for PadR-type transcription regulator (TR) (RcaC; GenBank accession number WP_035082396.1) and ABC-type transporters (RcaF, RcaG, and RcaI; GenBank accession numbers WP_084587593.1, WP_051960572.1, and WP_051960573.1, respectively). The transporter proteins of D. riboflavina are highly similar to their M. maritypicum MF109 counterparts (74%, 74%, and 78% identity, respectively), suggesting that their encoding genes are orthologs. The D. riboflavina transporter proteins are distantly related to the only characterized ABC-type riboflavin transporter, spirochete RfuABCD (<35%), and they lack the riboflavin-binding subunit RfuA (19). The bacterial genomes locate these genes in different orders, and the genes could not form single operons, since the rcaC genes are carried on the DNA strand complementary to the other genes within the clusters. The amounts of D. riboflavina gene transcripts quantified by PCR indicated that the addition of riboflavin to the culture upregulated the expression of these genes (Fig. 2B), whereas neither FMN nor FAD upregulated expression. We measured FMO activity using coupling assays with recombinant FR (rFR) pellet fractions of the cell extracts of D. riboflavina, as shown below. The activities were 9.1 ± 1.4 and 0.05 ± 0.01 nmol min−1 g−1 when the bacteria were cultured with and without riboflavin, respectively. This finding agreed with the transcriptional regulation of rcaE and indicated that riboflavin induces FMO. These results suggested that the FMO gene cluster is involved in riboflavin degradation by D. riboflavina.

FIG 2
FIG 2

D. riboflavina gene cluster for riboflavin degradation. (A) Schematic illustration of gene clusters encoding FMO (rcaE; GenBank accession number WP_035082397.1), FR (rcaB; GenBank accession number WP_084587600.1), FK (rcaA; GenBank accession number WP_084587592.1), RK (rcaD; GenBank accession number WP_035082393.1), PadR-family transcription regulator (TR) (rcaC; GenBank accession number WP_035082396.1), and ABC transporters (rcaF, rcaG, and rcaI; GenBank accession numbers WP_084587593.1, WP_051960572.1, and WP_051960573.1, respectively). (B) Transcript levels determined by quantitative PCR after addition of riboflavin to cultures. Values are relative to the level of gyrA transcripts at 0 h, which was taken as 1. Data are means of three experiments. Bars represent standard errors. *, P < 0.05; †, P < 0.01; ‡, P < 0.1; §, P < 0.2.

Steady-state kinetics of recombinant FK and recombinant FR.We investigated the enzymatic properties of FK and FR to understand the role of the proteins encoded by the FMO gene cluster. Recombinant FK (rFK) and rFR produced using an E. coli expression system were purified to apparent homogeneity (Fig. 3A). Reactions of rFK, riboflavin, and ATP at 37°C converted riboflavin to FMN (Fig. 3B). Steady-state kinetics showed an apparent Km value for riboflavin of 57 ± 1 μM and an apparent rate constant (kcat) of 3.5 ± 0.1 min−1; both values were comparable to those for the known FK (20). These findings indicate that the predicted protein is a D. riboflavina FK. Recombinant FR did not absorb in the visible region. The typical FMN spectrum has absorption peaks at 371 nm and 444 nm. Addition of saturating amounts of rFR to FMN altered this spectrum to one with absorption peaks at 378 and 455 nm, with a shoulder at 483 nm, which is considered to indicate the FMN-bound form of rFR (Fig. 3C). The typical absorption spectrum of FMN (Fig. 3D, solid line) disappeared when NADH was added in the presence of rFR (Fig. 3D, dashed-dotted line), indicating that NADH reduced FMN. The specific activity of FMN-dependent NADH oxidation by rFR was 51 μmol min−1 mg−1. Steady-state kinetics for NADH oxidation in the presence of various flavins indicated that the apparent Km value for FMN was 6.0 ± 0.3 μM, which was 15-fold lower than that for FAD (Table 1). This Km value for FMN was comparable to that of known bacterial FR, which ranges between 1 and 32 μM (21, 22). Little riboflavin-dependent NADH oxidation occurred at riboflavin concentrations of up to 0.2 mM, indicating that rFR specifically reduced FMN. These results indicated that the D. riboflavina FR is an NADH:FMN oxidoreductase.

FIG 3
FIG 3

Preparation and characterization of recombinant FMO, FR, and FK. (A) Images of SDS-PAGE gels. Lane M, molecular size markers; lane 1, rFMO (1 μg); lane 2, rFR (1 μg); lane 3, rFK (0.2 μg). (B) Recombinant FK phosphorylation of riboflavin. Standard reaction mixtures containing 85 μg ml−1 rFK, 0.1 mM riboflavin (RF), and 5 mM ATP were analyzed by HPLC. (C) Absorption spectra of 10 μM FMN in 50 mM potassium phosphate (pH 7.0) with (solid line) or without (dashed line) 200 μM rFR. (D) Reduction of FMN by rFR. Absorption spectra of 50 μM FMN in 50 mM potassium phosphate buffer (pH 7.0) (solid line) or in buffer with added 0.1 μM rFR (dashed line) and after further addition of 0.2 mM NADH (dashed-dotted line) were recorded.

View this table:
TABLE 1

Steady-state kinetics of rFR with various substratesa

Coupled rFMO and rFR degradation of riboflavin to lumichrome.Bacterial FMO and FR often constitute a two-component FMO system in which FR supplies FMO with reduced FMN or FAD for monooxygenase reactions (23, 24). Evaluating the simple FMO reaction using reduced FMN and O2 is difficult due to the speed and nonenzymatic nature of the reaction; therefore, we coupled the rFMO reaction with FR. In reactions of NADH, FMN, and riboflavin in the presence of rFR and rFMO (at a ratio of 1:10) under aerobic conditions, riboflavin was degraded and lumichrome was produced (Fig. 4A, FMO/FR). Reactions of NADH, FMN, and riboflavin without rFMO generated background amounts of lumichrome, similar to those generated in the absence of both enzymes, indicating that rFMO is required to generate lumichrome (Fig. 4A, FR). Reactions without rFR yielded lower but significant amounts of lumichrome (Fig. 4A, FMO); this was probably due to the nonenzymatic reduction of FMN by NADH. Reaction mixtures lacking either riboflavin or FMN produced little lumichrome, indicating that riboflavin but not FMN was degraded to lumichrome and that FMN functions as an electron mediator for the reaction (Fig. 4B, −riboflavin and −FMN). After the reaction, 42 ± 3 μM riboflavin was degraded and 35 ± 4 μM lumichrome was produced, indicating a stoichiometric reaction. The specific activity of the coupling reaction was measured at a fixed concentration of rFR (Fig. 4C). The coupling reaction was dependent on a rFMO concentration of 30 nmol min−1 mg−1 below a rFMO/rFR ratio of 25:1. This concentration is far less than that required for the FMN-dependent NADH oxidation activity of rFR (51 μmol min−1 mg−1), indicating that rFMO limits the coupling reaction. The activity increased linearly in the presence of riboflavin concentrations of up to ∼0.2 mM when rFMO and rFR, at a ratio of 25:1, were included in the reaction (Fig. 4D). This indicated that the apparent Km value of the coupling reaction for riboflavin was >0.2 mM. These results showed that coupled rFMO and rFR degrade riboflavin to lumichrome.

FIG 4
FIG 4

Coupled rFMO and rFR degradation of riboflavin to lumichrome (LC). (A) Coupling reaction of rFMO and rFR analyzed by HPLC. Reaction mixtures contained 5 mM NADH, 20 μM FMN, and 0.2 mM riboflavin in 50 mM potassium phosphate (pH 7.0) and included the following enzymes: 0.5 μM rFR (FR), 5 μM rFMO (FMO), 0.5 μM rFR, and 5 μM rFMO (FMO/FR). (B) Coupling reaction of 0.5 μM rFR and 5 μM rFMO without riboflavin (−riboflavin), FMN (−FMN), or oxygen (−O2), analyzed by HPLC. (C) Dependence of activity of coupled rFMO and rFR on rFMO. Reactions proceeded with 2 mM NADH, 0.2 μM FMN, 0.2 mM riboflavin, 0.4 μM rFR, and various concentrations of rFMO in 50 mM potassium phosphate (pH 7.0). (D) Coupling reaction of 0.5 μM rFR and 5 μM rFMO with 0 to 0.2 mM riboflavin.

Monooxygenation of riboflavin to lumichrome and d-ribose.We investigated the reaction by D. riboflavina rFMO in detail because the M. maritypicum RcaE protein, which is an oxygen-dependent hydrolase, is unlikely because the M. maritypicum RcaE product is d-ribose (17), which would not be a simple hydrolase product. The native FMN spectrum has absorption peaks at 371 nm and 444 nm. The rFMO has no absorption peak in the visible region, and addition of excess rFMO changed the peaks of FMN to 366 and 445 nm, with a shoulder at 468 nm (Fig. 5A). These results showed that rFMO can bind FMN. Addition of NADH and catalytic amounts of rFR decreased the absorption peak at 445 nm for the FMN-bound rFMO and formed a spectral species that absorbed with a peak at 353 nm (Fig. 5B), which is typical of FMN-OOH (25). This spectrum disappeared upon the addition of riboflavin, which agrees with the notion that FMN-OOH is a reaction intermediate of rFMO. Turnover of the coupling reaction consumed oxygen at a rate of 580 nmol min−1 mg−1 rFMO and was dependent on rFMO (Fig. 5C). The reaction produced little lumichrome under anaerobic conditions (Fig. 4B, −O2), indicating a requirement for oxygen. After the complete consumption of 0.25 mM oxygen, gas chromatography-mass spectrometry (GC-MS) detected 20 ± 7 μM d-ribose in the reaction, indicating that the reaction products were d-ribose (Fig. 5D) and lumichrome (Fig. 4A). The amount of d-ribose generated was comparable to that generated when riboflavin was degraded and lumichrome was produced (42 ± 3 and 35 ± 4 μM, respectively), and these values accounted for 8% to 16% of the consumed oxygen, suggesting that the monooxygenase reaction is uncoupled in the presence of oxygen, like other FMOs that oxidize lysine (26) and nitrilotriacetate (27). These results indicated that D. riboflavina rFMO is a riboflavin monooxygenase that functions with FR and is responsible for riboflavin degradation in D. riboflavina.

FIG 5
FIG 5

Devosia riboflavina FMO as a riboflavin monooxygenase. (A) Binding of FMN and rFMO. The UV-visible absorption spectrum of FMN (10 μM) in 50 mM potassium phosphate (pH 7.0) (dashed line) was changed with the addition of 400 μM rFMO (solid line). (B) Generation of FMN-OOH species of rFMO. Absorption spectra of 50 μM FMN in 50 mM potassium phosphate (pH 7.0) containing 25 μM rFMO and a catalytic amount (0.25 μM) of rFR, without (dashed line) or with (solid line) added 0.2 mM NADH, were recorded. (C) Oxygen consumption by coupling reaction mixtures containing 5 mM NADH, 20 μM FMN, 0.2 mM riboflavin, and 0.1 μM rFR in 50 mM potassium phosphate (pH 7.0), with (●) or without (○) 10 μM rFMO. (D) Mass spectrometric identification of the reaction product as d-ribose.

Distribution of FMO and gene clusters for riboflavin degradation.Xu et al. showed that riboflavin-degrading FMOs are distributed in specific bacteria, including Microbacteria and related Actinomycetales bacteria (17). Since D. riboflavina FMO has only 67% amino acid identity with M. maritypicum FMO (GenBank accession number WP_021201082.1), we revisited the phylogenetic analysis of those proteins. BLASTP searches of amino acid sequence databases showed that D. riboflavina FMO is a member of the nitrilotriacetate monooxygenase family with the TIGRFAM accession number TIGR03860. Phylogenetic analysis of the top 80 most diverse members of this family indicated that they are grouped into N5,N10-methylene tetrahydromethanopterin reductases, xenobiotic compound monooxygenases, nitrilotriacetate monooxygenases, and other proteins in the LLM class FMO without known function (Fig. 6A; also see Fig. S2 in the supplemental material). Devosia riboflavina FMO is located in the same branch as the latter family members. We analyzed this phylogeny by adding proteins in the NCBI database with >40% amino acid identity with D. riboflavina FMO, and we showed that D. riboflavina FMO constitutes a group together with predicted LLM class FMOs from other Devosia species, actinobacteria (Microbacterium, Herbiconiux, Leifsonia, Kitasatospora, and Streptomyces species), Paenibacillus spp., and other bacteria (Fig. S2). This group contains the M. maritypicum FMO and branches from the clade of N5,N10-methylene tetrahydromethanopterin reductases (Fig. 6A). These results indicated that these FMOs have diversified from the known proteins and constitute a novel group within the FMO family of riboflavin monooxygenases.

FIG 6
FIG 6

Distribution of FMOs among bacteria. (A) Phylogeny of bacterial FMO-like proteins. Protein sequences were aligned using MEGA5.1 software (36), and the phylogenetic tree was constructed using the maximum likelihood method. The tree was rooted to outgroup LUXA2_PHOLU (bacterial luciferase α-chain). (B) Proposed model for D. riboflavina catabolism of riboflavin (top) and the reaction catalyzed by riboflavin monooxygenase (coupled FMO and FR) (bottom).

We investigated the distribution of bacterial gene clusters for riboflavin degradation using Patric 3.4.15 software (28). Searching for gene clusters with orthologs for both D. riboflavina FMO and FR found that Devosia sp. strain Root413D1, Devosia epidermidihirudinis E84, and Devosia sp. strain H5989 genomes encode all of the genes identified in the D. riboflavina JCM13427 cluster for riboflavin degradation, and these bacteria have the same order and orientation of these genes (Fig. S3). The results of the analysis also found that M. maritypicum, Microbacterium paraoxydans DSM 15019, Herbiconiux sp. strain YR403, and Leifsonia sp. strain Root 4 organize essentially the same gene cluster. No bacteria other than actinomycetes and Devosia spp. contain a complete set of the cluster genes, whereas some actinomycetes (marine actinobacterium PHSC20C1 and Leifsonia aquatic H1aii) lack some of the genes for RK and FK, respectively (Fig. S3). These results indicated that the gene cluster is distributed among Devosia spp. and actinomycetes.

DISCUSSION

Figure 6B shows that a key two-component FMO system consisting of FMO and FR oxygenizes riboflavin in the D. riboflavina mechanism of riboflavin degradation. The reaction requires NADH for FR to reduce FMN, and then the reduced FMN becomes a substrate of FMO. In the presence of reduced FMN and oxygen, FMO generates the intermediate FMN-OOH (Fig. 5B), which decays upon the addition of riboflavin and is a key oxidizing species for the reaction (25). The two-component reaction consumes oxygen (Fig. 5C), indicating that the system includes a monooxygenase (23, 24) and not an oxygen-dependent hydrolase, as suggested previously (17). The finding that the reaction produces stoichiometric amounts of d-ribose and lumichrome agrees with enzymatic oxygenation in the system, since simple chemical processes should hydrolyze riboflavin to ribitol instead of d-ribose. The reaction properties are essentially similar to those of two-component FMOs that hydroxylate other compounds (23, 24, 29), and the reaction equation is as follows: riboflavin + O2 + NADH + H+ → lumichrome + d-ribose + NAD+ + H2O. The detailed redox chemistry of the reaction presented in Fig. 6B indicated the mechanism of bacterial riboflavin degradation; the process is facilitated by a novel riboflavin monooxygenase, not riboflavin hydrolase.

The present study found that addition of riboflavin to the culture medium of D. riboflavina induced FMO and FR gene expression as a cell-free FMO activity, which had not been identified previously in riboflavin-degrading bacteria. The mechanism is considered to be that of riboflavin utilization as a carbon source, since D. riboflavina grows when riboflavin is the sole source of carbon (Fig. 1). We also found that the genes for RK and FK are clustered with those for FMO and FR and that riboflavin induced the expression of both genes. RK is ubiquitous among bacteria; it phosphorylates d-ribose to d-ribose-5-phosphate, which is subsequently metabolized in the pentose phosphate pathway (30), and it is likely involved in utilizing the d-ribose residue of riboflavin as a source of carbon and energy (Fig. 6B). The D. riboflavina FK (rcaA product) phosphorylates riboflavin to FMN (Fig. 3B) and could supply FMN to be a substrate of FR. The D. riboflavina genome does not predict an extra FK gene but predicts a gene for bifunctional riboflavin kinase/FAD synthetase (GenBank accession number WP_084587657.1), which phosphorylates riboflavin to FMN. Elucidation of the roles of these enzymes in the riboflavin degradation mechanism will require further investigation.

The present study uncovered the genes involved in riboflavin degradation by D. riboflavina. Bacterial genes for riboflavin degradation clustered in the actinobacterium M. maritypicum genome have only recently been characterized (17). Although no specific gene was characterized, Microbacterium sp. strain TPU 3518 degrades riboflavin to lumichrome (31, 32). The present study of D. riboflavina is the first to reveal the molecular mechanism of proteobacterial riboflavin degradation. We initially searched gene clusters for riboflavin degradation, and we found that they are mainly distributed in actinomycetes, including Microbacterium species (Fig. 6A; also see Fig. S3). Their distribution among proteobacteria is restricted to Devosia species. These findings showed that the mechanism is shared among these evolutionarily distantly related bacterial genera. Considering the ubiquitous distribution of riboflavin in the biosphere, this relatively narrow distribution of genes for riboflavin utilization is notable. The bacteria Microbacterium and Devosia have exploited an ecological niche by developing mechanisms of riboflavin utilization.

The riboflavin-degrading bacteria found to date produce lumichrome as a degradation product, and no other characterized bacteria are known to metabolize lumichrome. This raises a new question regarding the degradation mechanisms and ecological roles of bacteria. The study of interactions among riboflavin- and lumichrome-degrading bacteria represents an intriguing opportunity to understand bacterial catabolic diversity and riboflavin homeostasis in nature.

MATERIALS AND METHODS

Strains, plasmids, cultures, and media.Devosia riboflavina JCM 21224 (IFO13584, ATCC 9526; RIKEN Bioresource Center, Wako, Japan) was precultured in Luria-Bertani (LB) medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl) at 30°C for 24 h. A portion (1 ml) was inoculated into 100 ml of minimal medium (10 mM NH4Cl, 10 mM potassium phosphate, 0.03% MgSO4, 0.05% KCl, 10 mM riboflavin, 0.2% trace element solution [33] [pH 6.5]) and rotary cultured at 120 rpm and 30°C for 8 to 12 h. Cultures were protected from light to prevent riboflavin photodegradation. Riboflavin was replaced with lumichrome, d-ribose, ribitol, and FMN (5 mM each) unless otherwise stated. Escherichia coli JM109 and BL21(DE3) for plasmid construction and recombinant protein production were cultured in LB medium.

Determination of riboflavin and lumichrome levels.Culture broth diluted 10-fold with 80 mM KOH to solubilize cells and to precipitate lumichrome was separated by centrifugation at 204,000 × g for 5 min at 4°C to remove the pellet. Riboflavin and lumichrome levels in the soluble fraction were then analyzed by high-performance liquid chromatography (HPLC) using an Agilent 1260 Infinity system (Agilent Technologies, Santa Clara, CA) equipped with a Purospher STAR RP-18 end-capped column (Merck & Co. Inc., Kenilworth, NJ), with monitoring of absorption at 254 nm. The mobile phase was composed of sodium acetate (pH 4.5) and methanol (1:1 [vol/vol]), with a flow rate of 0.8 ml min−1. Analysis by LC-MS proceeded using an LCMS8030 system (Shimadzu Corp., Kyoto, Japan) and the same column and mobile phase.

Determination of riboflavin-inducible protein.Devosia riboflavina JCM 21224 was cultured at 30°C for 8 h in minimal medium or in the same medium containing 0.02% tryptone and 0.01% yeast extract instead of riboflavin. Cells were harvested, sonicated, and separated into supernatant and pellet fractions by centrifugation at 8,000 × g for 5 min at 4°C. The fractions were incubated at 30°C for 1 h in 50 mM potassium phosphate buffer (pH 7.0) containing 50 μM riboflavin, and then lumichrome levels were quantified by HPLC. The fractions were boiled for 3 min in 75 mM Tris-HCl sample buffer containing 10% glycerol, 2% SDS, 5% β-mercaptoethanol, and 0.25 mg liter−1 bromophenol blue, and then proteins were resolved by SDS-PAGE (34). Small portions of gels with protein bands were excised and bleached in 0.1 M ammonium bicarbonate-50% acetonitrile, and then the proteins were digested with trypsin. Peptide mass fingerprints were analyzed using an AB Sciex TOF/TOF 5800 matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) system (Sciex, Framingham, MA). Mass ion fragments were obtained in the MS mode, and encoding genes were identified using the MASCOT program (Matrix Science Inc., Boston, MA).

Quantitative PCR.Devosia riboflavina was cultured at 30°C for 8 h in minimal medium containing 5% LB medium, and then 50 μM riboflavin was added to the cultures. After additional 0.5, 1, and 2 h of incubation at 30°C, the cells were collected, and total RNA was prepared using the RNAprotect bacteria reagent (Qiagen Benelux B.V., Venlo, Netherlands). First-strand cDNA was synthesized using total RNA (1 μg) and the QuantiTect reverse transcription kit (Qiagen) and then was quantified by PCR using iQ SYBR green supermix (Bio-Rad Laboratories, Hercules, CA) and the Thermal Cycler Dice real-time system 2 (TaKaRa, Kyoto, Japan). Amounts of transcripts were normalized against those of the gyrA gene (GenBank accession number WP_035083147.1). Table 2 shows the gene-specific primers. Relative expression levels were calculated.

View this table:
TABLE 2

Primers used in this study

Preparation of recombinant FMO, FR, and FK proteins.Fragments of DNA were amplified by PCR using the primer sets listed in Table 2, digested with BamHI and EcoRI, cloned into pRSFDuet-1 (Novagen, Madison, WI), and introduced into E. coli BL21(DE3). Transformants were cultured at 15°C for 18 h in LB medium (2 liters), at 100 rpm. Cells were collected by centrifugation at 6,500 × g for 15 min, suspended in 20 mM sodium phosphate (pH 7.4) containing 20 mM imidazole, and sonicated. The pellet fraction was removed by centrifugation at 8,000 × g for 30 min, and then the sonicate was filtered through 0.22-μm Millex-GV filters (Merck) and purified by metal-chelate chromatography using HiTrap FF crude columns (GE Healthcare, Chicago, IL). The columns were washed with 20 mM sodium phosphate (pH 7.4), and then proteins were eluted with 20 mM sodium phosphate (pH 7.4) containing 0.1, 0.15, or 0.3 M imidazole.

Enzyme assays.Reactions of rFR were monitored in 50 mM potassium phosphate, 20 μM FMN, 5 mM NADH (pH 7.0), with 1.8 μg ml−1 rFR by measuring absorbance at 340 nm using a DU800 spectrophotometer (Beckman Coulter Inc., Brea, CA) at 25°C. The molar extinction coefficient of NADH was set at ε340 of 6.22 mM cm−1. Typical reactions involving rFMO-rFR coupling proceeded at 37°C in 50 mM potassium phosphate (pH 7.0) containing 20 μM FMN, 5 mM NADH, 0.2 mM riboflavin, 4 μM rFMO, and 0.4 μM rFR. Reaction solutions were assessed by HPLC as described above. To create an anaerobic environment, air in the reaction tubes was purged with dinitrogen gas, to decrease the oxygen concentration to <3 μM, and then the tubes were sealed with double butyl rubber stoppers. FK activity was analyzed in 50 mM potassium phosphate (pH 7.0) containing 5 mM ATP, 5 mM MgSO4, and 0.1 mM riboflavin, and reactions proceeded at 37°C for 15 min. The amounts of FMN and riboflavin produced were determined by HPLC.

Determination of d-ribose levels.d-Ribose was trimethylsilylated using trimethylsilyl iodide (TMSI) (type H; GL Science, Tokyo, Japan) and analyzed by GC-MS using a QP2010 gas chromatograph-mass spectrometer (Shimadzu, Kyoto, Japan) equipped with a DB-5 capillary column (30 m by 0.32 mm; J & W Scientific Inc., Folsom, CA). The injection and ion-source temperatures were 250°C and 200°C, respectively, and the carrier was helium gas (1 ml min−1). The oven temperature was maintained at 120°C for 2 min, increased to 180°C at a rate of 6°C min−1, maintained for 1 min, increased to 300°C at a rate of 10°C min−1, and then held for 1 min. We identified d-ribose by comparing mass spectra and GC elution times with those for commercial d-ribose.

Phylogenetic analysis.Protein sequences were obtained from TIGR (35) and NCBI databases. Amino acid sequences were aligned using MEGA5.1 software (36), and a phylogenetic tree was constructed using the maximum likelihood method. Gene clusters with orthologs for both D. riboflavina FMO and FR were searched for using Patric 3.4.15 software (28).

Other methods.UV-visible absorption spectra were recorded using the DU800 spectrophotometer. Oxygen concentrations were measured at 25°C in air-tight vessels using a model 5300 biological oxygen monitor with a Clark-type O2 electrode (YSI Inc., Yellow Springs, OH).

ACKNOWLEDGMENTS

We thank Norma Foster for critical reading of the manuscript.

This study was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Culture, and Sports of Japan.

We have no conflicts of interest to declare.

H.K. designed experiments and analyzed the data; Y.K., K.-I.O., and A.N. conducted molecular biology experiments; R.S. and S.M. conducted mass spectrometry experiments and phylogenetic analyses; H.K. and N.T. designed the figures and wrote the paper.

FOOTNOTES

    • Received 12 January 2018.
    • Accepted 28 March 2018.
    • Accepted manuscript posted online 2 April 2018.
  • Address correspondence to Naoki Takaya, takaya.naoki.ge{at}u.tsukuba.ac.jp.

  • Citation Kanazawa H, Shigemoto R, Kawasaki Y, Oinuma K-I, Nakamura A, Masuo S, Takaya N. 2018. Two-component flavin-dependent riboflavin monooxygenase degrades riboflavin in Devosia riboflavina. J Bacteriol 200:e00022-18. https://doi.org/10.1128/JB.00022-18.

  • Supplemental material for this article may be found at https://doi.org/10.1128/JB.00022-18.

REFERENCES

  1. 1.↵
    1. Underwood BA
    . 2004. Vitamin A deficiency disorders: international efforts to control a preventable “pox.” J Nutr 134:231S–236S. doi:10.1093/jn/134.1.231S.
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    1. Roje S
    . 2007. Vitamin B biosynthesis in plants. Phytochemistry 68:1904–1921. doi:10.1016/j.phytochem.2007.03.038.
    OpenUrlCrossRefPubMedWeb of Science
  3. 3.↵
    1. Figueroa-Méndez R,
    2. Rivas-Arancibia S
    . 2015. Vitamin C in health and disease: its role in the metabolism of cells and redox state in the brain. Front Physiol 6:397.
    OpenUrl
  4. 4.↵
    1. Holick MF,
    2. Chen TC
    . 2008. Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr 87:1080S–1086S. doi:10.1093/ajcn/87.4.1080S.
    OpenUrlAbstract/FREE Full Text
  5. 5.↵
    1. Dror DK,
    2. Allen LH
    . 2011. Vitamin E deficiency in developing countries. Food Nutr Bull 32:124–143. doi:10.1177/156482651103200206.
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. DiNicolantonio JJ,
    2. Bhutani J,
    3. O'Keefe JH
    . 2015. The health benefits of vitamin K. Open Heart 2:e000300. doi:10.1136/openhrt-2015-000300.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    1. Miret JA,
    2. Munné-Bosch S
    . 2014. Plant amino acid-derived vitamins: biosynthesis and function. Amino Acids 46:809–824. doi:10.1007/s00726-013-1653-3.
    OpenUrlCrossRef
  8. 8.↵
    1. Powers HJ
    . 2003. Riboflavin (vitamin B-2) and health. Am J Clin Nutr 77:1352–1360. doi:10.1093/ajcn/77.6.1352.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Herz S,
    2. Eberhardt S,
    3. Bacher A
    . 2000. Biosynthesis of riboflavin in plants: the ribA gene of Arabidopsis thaliana specifies a bifunctional GTP cyclohydrolase II/3,4-dihydroxy-2-butanone 4-phosphate synthase. Phytochemistry 53:723–731. doi:10.1016/S0031-9422(00)00013-3.
    OpenUrlCrossRefPubMedWeb of Science
  10. 10.↵
    1. Schwechheimer SK,
    2. Park EY,
    3. Revuelta JL,
    4. Becker J,
    5. Wittmann C
    . 2016. Biotechnology of riboflavin. Appl Microbiol Biotechnol 100:2107–2119. doi:10.1007/s00253-015-7256-z.
    OpenUrlCrossRef
  11. 11.↵
    1. Bennett BD,
    2. Kimball EH,
    3. Gao M,
    4. Osterhout R,
    5. Van Dien SJ,
    6. Rabinowitz JD
    . 2009. Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli. Nat Chem Biol 5:593–599. doi:10.1038/nchembio.186.
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.↵
    1. Ahmad I,
    2. Fasihullah Q,
    3. Vaid FHM
    . 2004. A study of simultaneous photolysis and photoaddition reactions of riboflavin in aqueous solution. J Photochem Photobiol B 75:13–20. doi:10.1016/j.jphotobiol.2004.04.001.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Foster JW
    . 1944. Microbiological aspects of riboflavin: I. Introduction. II. Bacterial oxidation of riboflavin to lumichrome. J Bacteriol 47:27–41.
    OpenUrlFREE Full Text
  14. 14.↵
    1. Nakagawa Y,
    2. Sakane T,
    3. Yokota A
    . 1996. Transfer of “Pseudomonas riboflavina” (Foster 1944), a Gram-negative, motile rod with long-chain 3-hydroxy fatty acids, to Devosia riboflavina gen. nov., sp. nov., nom. rev. Int J Syst Bacteriol 46:16–22. doi:10.1099/00207713-46-1-16.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Yanagita T,
    2. Foster JW
    . 1956. A bacterial riboflavin hydrolase. J Biol Chem 221:593–607.
    OpenUrlFREE Full Text
  16. 16.↵
    1. Harkness DR,
    2. Stadtman ER
    . 1965. Bacterial degradation of riboflavin. VI. Enzymatic conversion of riboflavin to 1-ribityl-2,3-diketo-1,2,3,4-tetrahydro-6,7-dimethylauinoxaline, urea, and carbon dioxide. J Biol Chem 240:4089–4096.
    OpenUrlFREE Full Text
  17. 17.↵
    1. Xu H,
    2. Chakrabarty Y,
    3. Philmus B,
    4. Mehta AP,
    5. Bhandari D,
    6. Hohmann H,
    7. Begley TP
    . 2016. Identification of the first riboflavin catabolic gene cluster isolated from Microbacterium maritypicum G10. J Biol Chem 291:23506–23515. doi:10.1074/jbc.M116.729871.
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    1. Hassan YI,
    2. Lepp D,
    3. He J,
    4. Zhou T
    . 2014. Draft genome sequences of Devosia sp. strain 17-2-E-8 and Devosia riboflavina strain IFO13584. Genome Announc 2:e00994-14. doi:10.1128/genomeA.00994-14.
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    1. Deka RK,
    2. Brautigam CA,
    3. Biddy BA,
    4. Liu WZ,
    5. Norgard MV
    . 2013. Evidence for an ABC-type riboflavin transporter system in pathogenic Spirochetes. mBio 4:e00615-12. doi:10.1128/mBio.00615-12.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Serrano A,
    2. Frago S,
    3. Velazquez-Campoy A,
    4. Medina M
    . 2012. Role of key residues at the flavin mononucleotide (FMN):adenylyltransferase catalytic site of the bifunctional riboflavin kinase/flavin adenine dinucleotide (FAD) synthetase from Corynebacterium ammoniagenes. Int J Mol Sci 13:14492–14517. doi:10.3390/ijms131114492.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Filisetti L,
    2. Fontecave M,
    3. Nivière V
    . 2003. Mechanism and substrate specificity of the flavin reductase ActVB from Streptomyces coelicolor. J Biol Chem 278:296–303. doi:10.1074/jbc.M209689200.
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    1. Eschenbrenner M,
    2. Covès J,
    3. Fontecave M
    . 1995. The flavin reductase activity of the flavoprotein component of sulfite reductase from Escherichia coli: a new model for the protein structure. J Biol Chem 270:20550–20555. doi:10.1074/jbc.270.35.20550.
    OpenUrlAbstract/FREE Full Text
  23. 23.↵
    1. Ellis HR
    . 2010. The FMN-dependent two-component monooxygenase systems. Arch Biochem Biophys 497:1–12. doi:10.1016/j.abb.2010.02.007.
    OpenUrlCrossRefPubMedWeb of Science
  24. 24.↵
    1. van Berkel WJH,
    2. Kamerbeek NM,
    3. Fraaij MW
    . 2006. Flavoprotein monooxygenases, a diverse class of oxidative biocatalysts. J Biotechnol 124:670–689. doi:10.1016/j.jbiotec.2006.03.044.
    OpenUrlCrossRefPubMedWeb of Science
  25. 25.↵
    1. Valton J,
    2. Fontecave M,
    3. Douki T,
    4. Kendrew SG,
    5. Nivière V
    . 2006. An aromatic hydroxylation reaction catalyzed by a two-component FMN-dependent monooxygenase: the ActVA-ActVB system from Streptomyces coelicolor. J Biol Chem 281:27–35. doi:10.1074/jbc.M506146200.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    1. Robinson RM,
    2. Rodriguez PJ,
    3. Sobrado P
    . 2014. Mechanistic studies on the flavin-dependent N6-lysine monooxygenase MbsG reveal an unusual control for catalysis. Arch Biochem Biophys 550-551:58–66. doi:10.1016/j.abb.2014.04.006.
    OpenUrlCrossRef
  27. 27.↵
    1. Uetz T,
    2. Schneider R,
    3. Snozzi M,
    4. Egli T
    . 1992. Purification and characterization of a two-component monooxygenase that hydroxylates nitrilotriacetate from “Chelatobacter” strain ATCC 29600. J Bacteriol 174:1179–1188. doi:10.1128/jb.174.4.1179-1188.1992.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    1. Wattam AR,
    2. Davis JJ,
    3. Assaf R,
    4. Boisvert S,
    5. Brettin T,
    6. Bun C,
    7. Conrad N,
    8. Dietrich EM,
    9. Disz T,
    10. Gabbard JL,
    11. Gerdes S,
    12. Henry CS,
    13. Kenyon RW,
    14. Machi D,
    15. Mao C,
    16. Nordberg EK,
    17. Olsen GJ,
    18. Murphy-Olson D,
    19. Olson R,
    20. Overbeek R,
    21. Parrello B,
    22. Pusch GD,
    23. Shukla M,
    24. Vonstein V,
    25. Warren A,
    26. Xia F,
    27. Yoo H,
    28. Stevens RL
    . 2017. Improvements to PATRIC, the all-bacterial bioinformatics database and analysis resource center. Nucleic Acids Res 45:D535–D542. doi:10.1093/nar/gkw1017.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Arunachalam U,
    2. Massey V,
    3. Vaidyanathan CS
    . 1992. p-Hydroxyphenylacetate-3-hydroxylase: a two-protein component enzyme. J Biol Chem 267:25848–25855.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Lopilato JE,
    2. Garwin JL,
    3. Emr SD,
    4. Silhavy TJ,
    5. Beckwith JR
    . 1984. d-Ribose metabolism in Escherichia coli K-12: genetics, regulation, and transport. J Bacteriol 158:665–673.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Yamamoto K,
    2. Asano Y
    . 2017. Genome sequence of Microbacterium sp. strain TPU 3598, a lumichrome producer. Genome Announc 5:e00204-17. doi:10.1128/genomeA.00204-17.
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    1. Yamamoto K,
    2. Asano Y
    . 2015. Efficient production of lumichrome by Microbacterium sp. strain TPU 3598. Appl Environ Microbiol 81:7360–7367. doi:10.1128/AEM.02166-15.
    OpenUrlAbstract/FREE Full Text
  33. 33.↵
    1. Hutner SH,
    2. Provasoli L,
    3. Schatz A,
    4. Haskins CP
    . 1950. Some approaches to the study of the role of metals in the metabolism of microorganisms. Proc Am Philos Soc 94:152–170.
    OpenUrl
  34. 34.↵
    1. Laemmli UK
    . 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680. doi:10.1038/227680a0.
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    1. Haft DH,
    2. Selengut JD,
    3. Richter RA,
    4. Harkins D,
    5. Basu MK,
    6. Beck E
    . 2013. TIGRFAMs and Genome Properties in 2013. Nucleic Acids Res 41:D387–D395. doi:10.1093/nar/gks1234.
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    1. Hall BG
    . 2013. Building phylogenetic trees from molecular data with MEGA. Mol Biol Evol 30:1229–1235. doi:10.1093/molbev/mst012.
    OpenUrlCrossRefPubMedWeb of Science
Back to top
Download PDF
Citation Tools
Print

Alerts
Email

KEYWORDS

vitamin B2
lumichrome
isoalloxazine
ribitol
proteobacterium
riboflavin

Related Articles

Cited By...