Metagenomic Analysis Revealed Methylamine and Ureide Utilization of Soybean-Associated Methylobacterium

Abstract

Methylobacterium inhabits the phyllosphere of a large number of plants. We herein report the results of comparative metagenome analyses on methylobacterial communities of soybean plants grown in an experimental field in Tohoku University (Kashimadai, Miyagi, Japan). Methylobacterium was identified as the most dominant genus (33%) among bacteria inhabiting soybean stems. We classified plant-derived Methylobacterium species into Groups I, II, and III based on 16S rRNA gene sequences, and found that Group I members (phylogenetically close to M. extorquens) were dominant in soybean-associated Methylobacterium. By comparing 29 genomes, we found that all Group I members possessed a complete set of genes for the N-methylglutamate pathway for methylamine utilization, and genes for urea degradation (urea carboxylase, urea amidolyase, and conventional urease). Only Group I members and soybean methylobacterial isolates grew in a culture supplemented with methylamine as the sole carbon source. They utilized urea or allantoin (a urea-related compound in legumes) as the sole nitrogen source; however, group III also utilized these compounds. The utilization of allantoin may be crucial in soybean-bacterial interactions because allantoin is a transported form of fixed nitrogen in legume plants. Soybean-derived Group I strain AMS5 colonized the model legume Lotus japonicus well. A comparison among the 29 genomes of plant-derived and other strains suggested that several candidate genes are involved in plant colonization such as csgG (curli fimbriae). Genes for the N-methylglutamate pathway and curli fimbriae were more abundant in soybean microbiomes than in rice microbiomes in the field. Based on these results, we discuss the lifestyle of Methylobacterium in the legume phyllosphere.

Fig. 1

Neighbor-joining phylogenetic tree of 16S rRNA genes from Methylobacterium species. The scale bar represents the substitution number per site. The numbers at nodes are bootstrap values (%; values <50 are not shown). M1 to M8 are the OTUs described previously (2); the numbers in parentheses after each OTU name are the numbers of isolates assigned to the OTU. Strain sources are indicated after the strain names: A, air; F, food; H, human; P, plant; S, soil; W, water. *draft genome sequence published; **complete genome sequence published.

Fig. 2

Relative abundance of Groups I–III in different Methylobacterium populations revealed by a metagenome analysis. (A) Soybean metagenomes (SoyJp1 and SoyJp2) vs. rice metagenomes (RiceJp1 and RiceJp1). Metagenomes were obtained from plants grown at the same site. (B) The SoyJp1 dataset vs. a collection of Methylobacterium isolates. The SoyJp1 dataset was analyzed using the MG-RAST server (all sequence reads) (a), and on the basis of 16S rRNA (1,324 sequence reads) (b). The culture collection contained 126 isolates from soybean plants grown at the same site (c) (2).

Fig. 3

Distribution of orthologous genes within the genus Methylobacterium. (A) Flower plot: the numbers of genes shared by all strains of the same group or all strains in all three groups (center). (B) Venn diagram: genes shared by three plant-associated Group I strains; genes shared with other Group I members were excluded.

Fig. 4

Distribution of genes involved in methylamine and methanol metabolism among Methylobacterium species and the ability of Methylobacterium strains to use methylamine and methanol as sole carbon sources. (A) Percentage of strains in each group that have genes responsible for each step; pathway maps are shown next to the graphs. Abbreviations for enzymes: Fae, formaldehyde-activating enzyme; Fch, methenyl-H₄F cyclohydrolase; FDH, formate dehydrogenase; Fhc, formyltransferase/hydrolase complex; FtfL, formate-tetrahydrofolate ligase; Mch, methenyl-dH₄MPT cyclohydrolase; MDH, methanol dehydrogenase; MtdA and MtdB, methylene-tetrahydromethanopterin dehydrogenase. Abbreviations for compounds: CH₂=dH₄MPT, methylene-dH₄MPT; CH≡dH₄MPT, methenyl-dephospho H₄MPT; CHOdH₄MPT, formyl dH₄MPT; MFR, methanofuran; CHOMFR, formyl-MFR; H₄F, tetrahydrofolate; CHOH₄F, formyl-H₄F; CH≡H₄F, methenyl H₄F; dH₄MPT, dephosphotetrahydromethanopterin; GMA, γ-glutamylmethylamide; NMG, N-methylglutamate; α-KG, α-ketoglutarate. (B) Growth test without a carbon source or with methylamine or methanol as the sole carbon source for 7 d. 5, Methylobacterium sp. AMS5 (Group I); 1, M. extorquens AM1 (Group I); C, M. oryzae CBMB20 (Group III); J, M. radiotolerans JCM2831 (Group III).

Fig. 5

Relative abundance of genes involved in methanol and methylamine oxidation. The relative abundance of genes was normalized to the length of each gene. (A) Total reads that were divided by the number of metagenome reads. (B) Total reads that were divided by the proportion of methylobacteria in each of the four sets. Fae, formaldehyde-activating enzyme; GMAS, gamma-glutamylmethylamide synthetase; NMGD, N-methyl glutamate dehydrogenase; NMGS, N-methyl glutamate synthase; MaDH, methylamine dehydrogenase; MDH, methanol dehydrogenase; Mch, methenyl-dH₄MPT cyclohydrolase; H₄MPT, tetrahydromethanopterin.

Fig. 6

Urea utilization by Methylobacterium species. (A) Urea degradation pathways predicted from genome sequences and the average copy number of the relevant genes in each group. (B) Predicted allantoin degradation pathway (64). Accession numbers for the relevant genes are indicated. (C) Growth test without a nitrogen source or with ammonium, urea, or allantoin as the sole nitrogen source for 5 d. 5, Methylobacterium sp. AMS5; 1, M. extorquens AM1; C, M. oryzae CBMB20; and J, M. radiotolerans JCM2831. Strains isolated from soybean plants were marked by a green rectangle. Strains S9, S5, S2, and S1 were isolated from the stems of soybean plants and spotted at the upper section from left to right. Strains L21, L17, L14, and L13 were isolated from the leaves of soybean plants and spotted at the lower section from left to right. Strains B38, B33, B32, and B28 were isolated from rice plants and marked by a yellow rectangle and spotted from left to right.

Fig. 7

Methylobacterial colonization in Lotus japonicus visualized by GUS staining. Plants were inoculated with GusA-tagged Methylobacterium sp. AMS5 (A–C) or M. extorquens AM1 (D, E). (A) Cotyledons and the stem of a 10-d-old seedling. (B) Leaves of a 23-d-old plant. (C, E) Root nodules and (D) leaves of a 40-d-old plant. Scale bars indicate 1,000 μm (A) or 100 μm (B–E).