Cultivable Methylobacterium Species Diversity in Rice Seeds Identified With

Journal of Bioscience and Bioengineering VOL. 123 No. 2, 190e196, 2017 www.elsevier.com/locate/jbiosc

Cultivable Methylobacterium species diversity in rice seeds identiﬁed with whole-cell matrix-assisted laser desorption/ionization time-of-ﬂight mass spectrometric analysis

Marie Okumura,1 Yoshiko Fujitani,1 Masahiko Maekawa,1 Jittima Charoenpanich,2 Hunja Murage,3 Kazuhide Kimbara,1,4 Nurettin Sahin,5 and Akio Tani1,*

Institute of Plant Science and Resources, Okayama University, 2-20-1 Chuo, Kurashiki, Okayama 710-0046, Japan,1 Department of Biochemistry, Faculty of Science, Burapha University, Bangsaen, Chonburi 20131, Thailand,2 Jomo Kenyatta University of Agriculture and Technology, P.O. Box 62000-00200, Nairobi, Kenya,3 Faculty of Engineering, Shizuoka University, 3-5-1 Johoku, Kita-ku, Hamamatsu 432-8561, Japan,4 and Egitim Fakultesi, Mugla Sitki Kocman University, 48170 Kotekli, Mugla, Turkey5

Received 17 August 2016; accepted 5 September 2016 Available online 6 October 2016 Methylobacterium species are methylotrophic bacteria that widely inhabit plant surfaces. In addition to studies on methylotrophs as model organisms, research has also been conducted on their mechanism of plant growth promotion as well as the speciesespecies specificity of plantemicrobe interaction. We employed whole-cell matrix-assisted laser desorption/ionization (MALDI) mass spectrometry (WC-MS) analysis, which enables the rapid and accurate identification of bacteria at the species level, to identify Methylobacterium isolates collected from the rice seeds of different cultivars harvested in Japan, Thailand, and Kenya. Rice seeds obtained from diverse geographical locations showed different communities of Methylobacterium species. We found that M. fujisawaense, M. aquaticum, M. platani, and M. radiotolerans are the most frequently isolated species, but none were isolated as common species from 18 seed samples due to the highly biased communities in some samples. These findings will contribute to the development of formulations containing selected species that promote rice growth, though it may be necessary to customize the formulations depending on the cultivars and farm conditions. Ó 2016, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Methylobacterium species; Rice seeds; Whole-cell matrix-assisted laser desorption/ionization-mass spectrometry]

Methylobacterium species are members of the gram-negative of harvest (9), which suggests that the inoculated strains had been Alphaproteobacteria class of bacteria. They are well-known facul- eliminated. Common groups of Methylobacterium, however, were tative methylotrophs that utilize methanol and other C1 com- isolated from the inoculated and non-inoculated plants, which pounds. Metagenomic analysis has demonstrated that they occur suggest that they are specialized to rice or to the environment. To as one of the major bacteria inhabiting the phyllosphere (1). realize improved rice growth using Methylobacterium and to un- Methanol is one of the major volatiles emitted from plants, with an derstand the mechanism of the speciesespecies interaction, it re- estimated annual emission of 100 Tg (2). The predomination of mains necessary to accumulate more data on the interaction Methylobacterium species is explained by the methanol present in speciﬁcity between them. the phyllosphere, the concentration of which is equivalent to Leaf-inhabiting Methylobacterium species were suggested to be 2.5e250 mM methanol added to agar medium in the case of seed-borne rather than from environmental sources (16). Arabidopsis thaliana (3). These species produce the phytohor- Romanovskaya et al. (8) showed that inoculation of mones auxin (4) and cytokinin (5,6), and 1-aminocyclopropane-1- M. mesophilicum into soil did not result in leaf colonization of Zea carboxylate deaminase that reduces ethylene levels in plants (4,7). mays. They suggested that colonization occurs by soil particle In addition, they show antagonistic activity toward phytopatho- transfer by air under natural conditions. Mizuno et al. (17,18) genic bacteria (8). Probably due to these traits, the inoculation of demonstrated that Methylobacterium isolates from red perilla Methylobacterium onto plants results in a growth promotion effect, plants grown in different sites in Japan belong solely to as reported for rice (9,10), canola (4,11), bryophytes (12,13),and M. fujisawaense that had been isolated from the seeds, and that the other plants (14,15). This ability could be applicable for agricultural seed-inoculated strain was recovered from the plant leaves grown purposes. axenically, thus supported the seed-borne hypothesis. In the case of We attempted to promote rice growth through the inoculation rhizobia, Rhizobium strains were shown to migrate from tobacco of Methylobacterium isolates. We found that the inoculated strains roots to the leaf surface even when their epiphytic migration was could not be recovered from the ﬁeld-grown rice plants at the time blocked, suggesting that they migrate through plant stems (19). Sinorhizobium strains were shown to migrate from roots to leaves, colonizing every part of rice plants (20). In addition to these * Corresponding author. Tel./fax: þ81 86 434 1228. investigations on individual species, recent metagenomic analyses E-mail address: [email protected] (A. Tani).

1389-1723/$ e see front matter Ó 2016, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2016.09.001 VOL. 123, 2017 METHYLOBACTERIUM SPECIES DIVERSITY IN RICE SEEDS 191

have revealed that phyllospheric communities initially mirrored airborne communities and then converged to a distinct composition in a greenhouse experiment (21). Since Methylobacterium species are widespread in the environment, and are found in soil, dust, air, hailstones, and rain, leaf sampling might be strongly

Remarks affected by environmental factors. It is thus important to analyze the species colonizing the seeds, since endophytic microbes in seeds can be maternally passed on to plant descendants and would Paenibacillus hunanensis Paenibacillus hunanensis Pseudomonas oryzihabitans become the ﬁrst colonizer of juvenile seedlings. In this study, we used rice seeds of various cultivars grown in completely different environments in Japan, Kenya, and Thailand

isolates for Methylobacterium isolation. In this way, we examined which group can be commonly isolated, and whether rice genotype and fertilization affect the community structure. For this purpose, we used whole-cell matrix-assisted laser-desorption/ionization-mass Number of spectrometry (WC-MS) analysis, which enables the rapid identi- fi Methylobacterium cation of unknown isolates at the species level (22). We have demonstrated the effectiveness of the method for isolates from non- rice leaves and barley (9). The method is only applicable to isolated bacteria, and metagenomics or specifically targeted ribosomal spacer analyses (23) are more suited for quantitative isolates analysis of microbial community composition. As we seek to utilize the isolates for agricultural purposes, the method is convenient and rapid enough to effectively classify unidentified Number of isolates. Methylobacterium MATERIALS AND METHODS

Plant materials and isolation of Methylobacterium species Thericeseed 393 383 samples used in this study are summarized in Table 1. Rice samples were used for isolates

Number of bacteria isolation as soon as possible after harvest. Chemical fertilization at the Institute of Plant Science and Resources (IPSR), Okayama University, Akita Prefectural University, and Kinki University was applied in the amount of 5 kg each of N, P, and K per 1000 m2. Ten rice seeds from each sample were coding washed extensively with sterile water. The seeds were then surface-sterilized by washing twice with 20 ml of 0.5% sodium hypochlorite and 0.005% Tween 20 for 1 min, followed by rinsing with sterile water ﬁve times. The rice Rice seeds used in this study. seeds were husked, and the resultant brown rice was put in sterile 0.85% NaCl, and vortexed vigorously for 30 s. The resultant solution was spread onto solidiﬁed mineral medium containing 0.5% methanol (13) as the sole carbon

TABLE 1. source and 50 mg/L cycloheximide. Pink colonies that appeared after 3e7 days at 28C were isolated and puriﬁed by streaking on plate media of the same composition. WC-MS analysis and 16S rRNA gene sequencing All isolates were subjected to WC-MS analysis as reported previously (22). The data were analyzed with MALDI BioTyper 3.0 software (Bruker Daltonics) to construct the main spectra projection (MSP) dendrogram based on spectra similarity with the default settings (9).The type strains of Methylobacterium species obtained from the culture collections were included in the analysis. Representative isolates from each cluster in the MSP dendrogram were subjected to 16S rRNA gene sequencing and phylogenetic analysis as reported previously (9,22). The 16S rRNA gene sequence similarities of an isolate against those of Methylobacterium type strains were determined using the EzTaxon server (24). For species assignment of an isolate, pairwise sequence similarity values and nearest phylogenetic neighbors were taken into consideration. Nucleotide accession numbers The DDBJ accession numbers for the 16S rRNA gene sequences reported in this paper are LC025976eLC026013. Nipponbare Japonica PaddyKhao Taheng Japonica Yes Paddy AK 20 No TT 20 46 46 0 0 RESULTS AND DISCUSSION

Isolates from Norin 18 During the 2009, 2012, and 2013 harvesting season, 147 Methylobacterium isolates were obtained from Oryza sativa cultivar Norin 18 seeds. The isolates obtained in University province 2009 were categorized into four species (M. gossipiicola, M. fujisawaense, M. platani, and M. aquaticum) as a result of WC-MS and 16S rRNA gene sequencing (Fig. S1). Discrimination of isolates

Country Location Variety Japonica/Indica Paddy/Upland Fertilization Sample belonging to M. fujisawaense, M. oryzae, and M. phyllosphaerae type strain clusters was difﬁcult, since their 16S rRNA genes share more than 99% identity and also their WC-MS spectra have high Harvest year 200920092012 Japan2012 Japan2013 Japan IPSR2013 Japan IPSR2009 Japan IPSR2009 Japan IPSR2012 Japan IPSR2012 Japan IPSR Norin 182012 Japan IPSR Norin 18 Japan IPSR Norin 182009 Japan Kinki University Norin 182009 Kinki University Norin Nipponbare 182009 Kenya Akita Prefectural Japonica Norin Nipponbare 182009 Kenya Japonica Nipponbare Bunyala2009 Kenya Japonica Nipponbare Taveta Paddy2009 Kenya Japonica Japonica Kaloleni Paddy Kenya Japonica Japonica Mwea Paddy2009 Thailand Japonica Japonica Paddy Hola PaddyTotal Chachoengsao Basmati Japonica 370 Paddy Paddy Thailand Paddy Japan 64 Paddy Lopburi Matuko province Yes Nyeusi Paddy No Khao Hom Mari Nerica Indica 4 Yes Indica JoF Yes No Sindano Bahari Japonica Yes JoN Yes Indicasimilarity. No12F Paddy Yes No R76 Paddy/Upland No Japonica No12N Paddy 14 R15 No13F Japonica 20 JiF 18 Paddy No13N Paddy No 20 JiN 20 Upland 40 16S 20 No 40 KK 20 No rRNA 20 No KB 14 No 20 Yes 18 TH gene 20 15 20 KT 40 KH 20 KM 8 40 sequence 20 42 20 10 6 14 15 similarity 0 8 0 0 35 0 0 9 0 value 14 0 4 0 0 less 0 than 0 0 7 1 0 2 7 1 2 192 OKUMURA ET AL. J. BIOSCI.BIOENG.,

M. radiotolerans M. extorquens M. bullatum M. persicinum M. komagatae M. thuringiense AK8 JiN1 (M. radiotolerans 99.861), 9 JiF1,2,3,4,6,10,15 R15-11,12,13,15,16,17,18,19,20 R76-4,8 (M. radiotolerans 99.861),13,15,16,18 No13N7,9,10,12,18,28,37,40 No13F1,2,6,7,11,12,15,19,21,25 (M. radiotolerans 99.93),28,34,38,40 No12N4,14,16 No12F2,3,4,5,7,9 (M. radiotolerans 99.861),11,13,16,20 M. radiotolerans KK1,2,3,7,10,11,12 (M. radiotolerans 99.861),14,15 KH1,KH2,KH4, KM5,KM12, KT5 TH3,4,8,14,17,18,32,33,35 (M. radiotolerans 99.931),39 TT1-8,9 (M. radiotolerans 99.79), 10-46

M. mesophilicum M. brachiatum M. dankookense M. aerolatum M. oxalidis M. variabile JoN7 M. fujisawaense 99.930 JoN3 M. oryzae M. Phyllosphaerae M. Fujisawaense JoN4,10,17,18 (M. fujisawaense 99.93) JiF5,7 (M. fujisawaense 99.928),12,16,17,18,20 JiN16 R15-1,14 R76-11,12,20 M. fujisawaense No12N5,6,7,9,10,11,12,19 No12F10,12,14 No13N12,4,5,6,8,11,13,14,15,16,20,24,27,31,33,35(M. oryzae 99.93),36 No13F3,10,13,17,24,29,30,31,32,35,37,39 AK-15,17 KB1, 4 (M. fujisawaense 99.93) KM2,3,4,6,7,8,9(M. fujisawaense 99.93),11,14

TH20 KK5 M. platani 98.379 M. zatmanii KH6 M. platani 98.165 M. longum M. tardum M. goesingense M. platani No12N4 No12N16 M. organophilum JiN2,4,5,7,20 JoF1,3,9,11 (M. platani 97.886) JoN5 (M. platani 98.096), 11,16 JiF8,9,13 (M. platani 98.096),14 (M. platani 98.32) No12N1,2,3,8,13, 15 (M. platani 98.096),17,18,20 No12F6,8, 19 M. platani No13N3,17,19(M. platani 98.11),17,21,22,23,25,26,29,30,39 No13F4,5,9,14,16,23,26,27,33,36 R76-3,5,7,9 ,10 (M. platani 97.74),14 (M. platani 98.06),17 (M. platani 98.874),19 R15-2,3,4,5,6,7,8,9,10 KB5,6,7 (M. platani 98.311) KK4,6,9,13 (M. platani 98.520) TH13,15,16,31,37 (M. platani 98.02)

M. marchanƟae M. gnaphalii M. brachythecii R76-1 M. salsuginis 99.722 M. salsuginis M. suomiense M. rhodinum M. iners M. thiocyanatum M. podarium M. rhodesianum M. aminovorans M. haplocladii TH1,2,5,6,7 (M. salsuginis 99.513), 9,10,11,21,22,23,24,25,27,28,29,30,34,36 KT4 M. salsuginis 99.444 KT2 M. salsuginis 99.791 KT6 M. salsuginis 99.583 M. salsuginis KT1 M. nodulans M. jeotgali M. isbiliense M. gossipiicola M. gregans M. hispanicum M. cerasƟi M. adhaesivum M. aquaƟcum JoF2 (M. gossipiicola 99.41),4,5,6,7,8 (M. gossipiicola 99.48),10,12 (M. gossipiicola 99.48),13,14 M. gossipiicola M. trifolii JiN15 (M. aquaƟcum 99.488) JoN1,2,6 (M. aquaƟcum 99.359),8,9,13,14 (M. aquaƟcum 99.359),15 No13N32 M. aquaƟcum 98.18 No13F8 (M. aquaƟcum 99.03), 18 KM10 (M. aquaƟcum 99.283), 13 KT7,9 (M. aquaƟcum 98.432),10 No12F1 (M. aquaƟcum 97.91),15,17,18(M. aquaƟcum 99.004) KM1 M. aquaƟcum 98.649 JiF11, JiN17 (M. aquaƟcum 98.43),19 M. tarhaniae M. aquaticum JoN12 M. aquaƟcum 97.582 R76-2 (M. aquaƟcum 98.719),6 No12N1,2,3,8,13, 15 (M. aquaƟcum 98.719),17,18,20 No12F6,8, 19 No13N3,17,21,22,23,25,26,29,30,39 No13F4,5,9,14,16,23,26,27,33,36 AK2,3,4,5,6,7,9,10,11,12,13,14,16,18,19,20 KB2 (M. aquaƟcum 98.51), 3,8 KT8

1000 800 600 400 200 0 Distance Level

FIG. 1. MSP dendrogram based on WC-MS analysis of the Methylobacterium isolates from all rice samples. The strains are colored in red (Japan), blue (Kenya), green (Thailand), and black (type strains) with 16S rRNA gene identity (%) to corresponding type strains. VOL. 123, 2017 METHYLOBACTERIUM SPECIES DIVERSITY IN RICE SEEDS 193

FIG. 2. Isolation frequency of Methylobacterium species from all rice samples. Numbers of isolates are indicated in parentheses. Sample coding refers to Table 1.

98.65e98.70% has been suggested as the cutoff for delineating community structure in 2009. The tendency of the increased fre- species without DNAeDNA hybridization in bacterial quency of M. radiotolerans and the decreased frequency of classiﬁcation (25). The isolates closest to the M. platani type M. fujisawaense seen in the No12 series was also seen in 2013 (No13 strain cluster showed less than 98.6% 16S rRNA gene identity, series). suggesting that the members of the group may represents a new species. Isolates from Nipponbare Forty Methylobacterium isolates Twenty isolates from each of the fertilized and non-fertilized from the O. sativa cultivar Nipponbare harvested in 2009 were Norin 18 rice seeds harvested in 2012 were analyzed (Fig. S2). categorized into four species (Fig. S4), namely, M. platani, The isolates were categorized into four species, M. platani, M. aquaticum, M. fujisawaense, and M. radiotolerans. Compared to M. aquaticum, M. radiotolerans, and M. fujisawaense. The difference the isolates from Norin 18 in 2009, fertilization did not seem to between the isolates from fertilized and non-fertilized Norin 18 affect the Methylobacterium community, but the low frequency of was the high frequency of M. platani and the low frequency of M. fujisawaense in the non-fertilized sample was observed. M. radiotolerans in the non-fertilized seeds. M. gossipiicola was not Interestingly, this tendency was opposite to the one seen in Norin found in the samples from 2012; instead, isolates belonging to 18 (No12 and No13 series). M. radiotolerans were found. Thus, the results obtained in 2009 We obtained Nipponbare seed samples grown in different were not reproduced in 2012. locations in Japan. Twenty isolates each from fertilized (coded as The identiﬁcation of the isolates from Norin 18 in 2013 is R76) and non-fertilized (R15) Nipponbare rice seeds harvested in shown in Fig. S3. Four species were found, M. platani, Wakayama prefecture in 2012 were analyzed (Fig. S5). Four M. radiotolerans, M. aquaticum,andM. oryzae (M. fujisawaense). groups were found; they belong to M. platani, M. radiotolerans, Interestingly, the isolates belong to the same species as that ob- M. fujisawaense,andM. aquaticum, except for one isolate tained in 2012, and there was no clear difference in species variety belonging to M. salsuginis (strain R76-1). There was no clear between different fertilization conditions. Thus, the Methyl- difference in the isolated species between R76 and R15, except obacterium community structure in the same rice species and for strain R76-1, suggesting that fertilization effect was little. conditions in different years was not constant and was, in fact, Twenty isolates from fertilized Nipponbare (coded as AK in 2012 rather variable. Common species such as M. platani, M. aquaticum, in Akita prefecture) were isolated, most of which belong to and M. fujisawaense were, however, found to be relatively stable M. aquaticum. The others belong to M. radiotolerans and species in Norin 18. M. fujisawaense, showing a fairly biased structure. The climate in Norin 18 is recognized as a low-input cultivar, which can grow Akita prefecture is colder than that in Okayama and Wakayama well with less fertilization compared to other cultivars. We specu- prefectures (about 5C difference in autumn); we speculate that lated that the bias might be due to the cultivar trait, but it was not this could be the reason for this bias, though the optimum reproduced in the following years. Therefore, other unknown growth temperatures of all the type strains of these species are environmental factors might have affected the Methylobacterium 25e30C (26e29) and there might be additional differences in 194 OKUMURA ET AL. J. BIOSCI.BIOENG.,

M. radiotolerans AB175640 TH35 M. radiotolerans 99.931 TT9 M. radiotolerans 99.79 KK12 M. radiotolerans 99.861 JiN1 M. radiotolerans 99.861 No13F25 M. radiotolerans 99.93 No12F9 M. radiotolerans 99.861 R76-8 M. radiotolerans 99.861 KM5 M. fujisawaense 99.717 99 M. longum FN868949.1 M. tardum AB252208 78 M. phyllostachyos EU912444 M. fujisawaense AB175634 No13N35 M. oryzae 99.93 JoN18 M. fujisawaense 99.93 JoN7 M. fujisawaense 99.93 JiF7 M. fujisawaense 99.928 KB4 M. fujisawaense 99.93 KM9 M. fujisawaense 99.93 M. phyllosphaerae EF126746 M. oryzae AY683045 M. pseudosasicola EU912439 AB175636 99 M. mesophilicum 74 M. brachiatum AB175649 M. komagatae AB252201 AB252202 92 M. persicinum 82 M. aerolatum EF174498 M. dankookense FJ155589 M. hispanicum AJ635304 99 M. gregans AB252200 M. organophilum AB175638 99 M. thuringiense FR847847 98 M. haplocladii AB698691 M. gnaphalii AB627071 99 M. brachythecii AB703239 93 M. marchanae FJ157976 M. bullatum GU983169 M. jeotgali DQ471331 M. trifolii FR847848 M. cerasi FR733885 97 JoF8 M. gossipiicola 99.48 JoF2 M. gossipiicola 99.41 74 JoF12 M. gossipiicola 99.48 97 M. adhaesivum AM040156 M. gossipiicola EU912445 M. goesingense AY364020 M. iners EF174497 M. oxalidis AB607860 EU860984 M. soli 77 75 M. thiocyanatum AB175646 M. populi AY251818 M. rhodesianum AB175643 M. zatmanii AB175647 M. extorquens AB175633 M. suomiense AB175645 M. aminovorans AB175629 M. rhodinum AB175644 72 TH7 M. salsuginis 99.513 99 KT4 M. salsuginis 99.444 KT2 M. salsuginis 99.791 99 R76-1 M. salsuginis 99.722 M. salsuginis EF015478 KT6 M. salsuginis 99.583 AF514774 M. podarium99 91 M. nodulans AF220763 M. isbiliense AJ888239 M. variabile AJ851087 M. platani EF426729 98 77 R76-9 M. aquacum 98.719 No12N15 M. aquacum 98.719 No12F1 M. aquacum 97.91 R76-2 M. aquacum 98.719 98 JiN17 M. aquacum 98.43 No12F18 M. aquacum 99.004 KB2 M. aquacum 98.51 M. tarhaniae JQ864432 98 KM10 M. aquacum 99.283 M. aquacum AJ635303 JoN6 M. aqucum 99.359 JoN14 M. aquacum 99.359 JiN15 M. aquacum 99.488 88 No13F8 M. aquacum 99.03 KT9 M. aquacum 99.423 KM1 M. aquacum 98.649 No13N32 M. aquacum 98.18 92 KK5 M. platani 98.379 KK13 M. platani 98.52 R76-17 M. platani 98.874 JiF14 M. platani 98.32 JoN12 M. platani 97.673 76 KH6 M. platani 98.165 R76-10 M. platani 97.94 90 JiF13 M. platani 98.096 TH37 M. platani 98.02 JoF11 M. platani 97.886 R76-14 M. platani 98.06 No13N19 M. platani 98.11 JoN5 M. platani 98.096 KB7 M. platani 98.311 R. palustris AB175650

0.01

FIG. 3. Phylogenetic analysis of the 16S rRNA gene sequences of the isolates and Methylobacterium type strains. The isolates are shown in red with identity to their closest type strains identified with EZtaxon. Rhodopseudomonas palustris (AB175650) is placed as the outgroup. The tree was constructed with the neighbor-joining method. Bootstrap values (>70%) expressed as percentages of 1000 replicates are shown at branch points. Bar, Tajima-Nei distance (units of the number of base substitutions per site). environmental factors. It is thus evident that the structure is not and M. radiotolerans. Although there seemed to be some determined solely by plant genotype when plants are grown in specificity between the plant cultivars and the bacterial species, different environments. due to the limited number of isolates, it was not possible to elucidate the relationship between the isolates and plant races fi Isolates from Kenyan rice From ve samples of different rice (Japonica or Indica) and the growth condition (paddy or upland). varieties harvested in Kenya, 53 isolates were obtained. The isolates It is of note that the isolated species from Kenyan rice samples were categorized into seven species (Fig. S6). Group 2 belongs to were also isolated from Japanese rice samples. Paenibacillus hunanensis, which is not a Methylobacterium species. The species have been isolated from rice seeds (30). We have not Isolates from Thai rice From samples of two different rice tested its growth on methanol in liquid culture, and its varieties from Thailand, 88 isolates were analyzed (Fig. S7). As non- methylotrophy is not yet known. The Methylobacterium isolates Methylobacterium isolates, seven isolates closest to Pseudomonas belong to M. salsuginis, M. platani, M. aquaticum, M. fujisawaense, oryzihabitans were isolated. This species is known to cause sepsis VOL. 123, 2017 METHYLOBACTERIUM SPECIES DIVERSITY IN RICE SEEDS 195

(31), but its methylotrophy and importance of rice seed inhabitation be important for intercellular colonization in plant cells (15), and are not known. The Methylobacterium isolates were grouped into comparative genomics showed important bacterial traits and M. radiotolerans, M. platani, and M. salsuginis. Interestingly, all of functions conserved in distinctive groups (33). Comparative geno- the isolates from Kao Tah Haeng (coded as TT, isolated in mics between leaf and seed isolates may shed light on the factors Chachoengsao province) belong to M. radiotolerans whereas those that drive species specificity. from Kao Hom Mari (TH, isolated in Lopburi province) belong to Conclusions In this survey we identified the most frequently the three species. Thus, the Methylobacterium community in the occurring four Methylobacterium species from rice seeds of former was quite biased, although the reason for this was unclear. various cultivars collected in different places in the world. We There is a climatic difference between the two sites; showed that the presence and frequency of Methylobacterium spp. Chachoengsao province is quite dry whereas Lopburi province is in seeds were affected by the cultivation and/or physiological warmer and humid. To investigate whether such climatic characteristics of the seeds and cultivars, as indicated previously difference plays a role in different or biased communities, it is (34).Eventhoughweusedavarietyofriceseedsamples,the necessary to plant the same cultivars in different places. isolates were restricted to these species. As there are 48 published names in the genus Methylobacterium, this restriction Summary of Methylobacterium species isolated from all rice suggests the niche-specific inhabitation or predomination of samples The analysis using all spectra of Methylobacterium specific Methylobacterium species. Their rice growth promoting isolates (i.e., excluding non-Methylobacterium isolates) shows that ability should be examined further. Whether they persist during most predominant species were M. radiotolerans, M. fujisawaense, different stages of rice growth before residing in the seeds again, M. platani,andM. aquaticum (Fig. 1). M. salsuginis and should also be further investigated. A mixture of these M. gossipiicola were found with minor frequency. The isolates representative species may aid in the development of a rice belonging to M. radiotolerans, M. fujisawaense, M. salsuginis,and biofertilizer, but tailor-made formulations may be necessary for M. gossipiicola groups showed high 16S rRNA gene identity to specific cultivars due to the fact that some had biased structures. those of their type strains, suggesting that these isolates belong Supplementary data to this article can be found online at http:// to their corresponding species. Many of the isolates belonging to dx.doi.org/10.1016/j.jbiosc.2016.09.001. M. platani and M. aquaticum groups showed less identity to their corresponding type strains (<98.6%), suggesting that they comprise possible new species. Their wider branching in the ACKNOWLEDGMENTS MSP dendrogram also suggested that there are several subgroups under these groups. Since further classification of The authors thank Dr. H. Takahashi (Akita Prefectural Univer- these isolates is out of the scope of this study, we tentatively sity) and Dr. T. Kato (Kinki University) for rice seed samples. This grouped them as M. platani and M. aquaticum. study was funded by Research for Promoting Technological Seeds Based on the identification, the species of the isolates in each from the Japan Science and Technology Agency (grant number 12- sample are summarized in Fig. 2. Overall, M. aquaticum, 032) and Grants-in-Aid for Scientific Research, Japan Society for the M. fujisawaense, M. radiotolerans, and M. platani were relatively Promotion of Science (grant number 23688012). MO, KK, and AT common in the rice samples used, and there was no ubiquitous conceived and designed the experiments. MO, YF, NS, and AT per- species that was found in all rice samples, due to some samples formed the experiments. MO, MM, JC, and HM collected rice seed with highly biased structures. samples. MO, NS, KK, and AT analyzed the data and wrote the The phylogenetic tree based on 16S rRNA gene sequences of the manuscript. isolates is shown in Fig. 3. The isolates belonging to M. platani type strain cluster comprise many putative new species with low References sequence identity (<98.6%). Further characterization and descrip- tion of these members are necessary to clarify their taxonomical 1. Delmotte, N., Knief, C., Chaffron, S., Innerebner, G., Roschitzki, B., positions. The phylogenetic analysis also supports the notion that Schlapbach, R., Mering, von, C., and Vorholt, J. A.: Community proteoge- the isolates are restricted to several species of Methylobacterium.In nomics reveals insights into the physiology of phyllosphere bacteria, Proc. Natl. e one of our previous studies (22), we found that M. extorquens and Acad. Sci. USA, 106, 16428 16433 (2009). 2. Galbally, I. E. and Kirstine, W.: The production of methanol by flowering M. radiotolerans were two of the most frequently isolated species plants and the global cycle of methanol, J. Atomos. Chem., 43, 195e229 (2002). from various plants, whereas M. platani and M. aquaticum were not. 3. Kawaguchi, K., Yurimoto, H., Oku, M., and Sakai, Y.: Yeast methylotrophy and Thus, the isolated species from rice seeds in this study are consid- autophagy in a methanol-oscillating environment on growing Arabidopsis ered to be more adapted to colonization in rice seeds. In another thaliana leaves, PLoS One, 6, e25257 (2011). study (9), we isolated M. tarhaniae, M. aerolatum, M. komagatae, and 4. Madhaiyan, M., Poonguzhali, S., and Sa, T.: Characterization of 1- fi aminocyclopropane-1-carboxylate (ACC) deaminase containing Methyl- M. suomiense species from eld-grown Nipponbare leaves, in obacterium oryzae and interactions with auxins and ACC regulation of ethylene addition to the species isolated in this study. These findings suggest in canola (Brassica campestris), Planta, 226, 867e876 (2007). that the leaves harbor more variety than seeds, and that there are 5. Lidstrom, M. E. and Chistoserdova, L.: Plants in the pink: cytokinin production unknown limiting factors that select microbial species in seeds. In by Methylobacterium, J. Bacteriol., 184, 1818 (2002). 6. Koenig, R. L., Morris, R. O., and Polacco, J. C.: tRNA is the source of low-level other words, there might be common characteristics for the seed trans-zeatin production in Methylobacterium spp. J. Bacteriol., 184, 1832e1842 species. We thought that one factor might be nutritional restriction (2002). in seeds, that is, the starch utilization capability. We found that all 7. Chinnadurai, C., Balachandar, D., and Sundaram, S. P.: Characterization of 1- isolates were negative in a-amylase, and that only the isolates aminocyclopropane-1-carboxylate deaminase producing methylobacteria from belonging to M. aquaticum were positive in b-amylase (data not phyllosphere of rice and their role in ethylene regulation, World J. Microbiol. Biotechnol., 25, 1403e1411 (2009). shown). We thus conclude that the common characteristic for these 8. Romanovskaya, V. A., Stolyar, S. M., Malashenko, Y. R., and Dodatko, T. N.: four species was not the starch utilization capability. The limiting The ways of plant colonization by Methylobacterium strains and properties of factor remains unknown, but similar organ-specific microbial these bacteria, Microbiology, 70, 221e227 (2001). community structures are also reported for grapevine (32). 9. Tani, A., Sahin, N., Fujitani, Y., Kato, A., Sato, K., and Kimbara, K.: Methyl- obacterium species promoting rice and barley growth and interaction speci- Recent studies on Methylobacterium-plant interaction have been ficity revealed with whole-cell matrix-assisted laser desorption/ionization- extensively done on the basis on their genome analyses, and time-of-flight mass spectrometry (MALDI-TOF/MS) analysis, PLoS One, 10, eukaryote-like genes encoded in Methylobacterium genome might e0129509 (2015). 196 OKUMURA ET AL. J. BIOSCI.BIOENG.,

10. Senthilkumar, M., Madhaiyan, M., Sundaram, S., and Kannaiyan, S.: Inter- Methylobacterium species using whole-cell MALDI-TOF/MS analysis, PLoS One, cellular colonization and growth promoting effects of Methylobacterium sp. 7, e40784 (2012). with plant-growth regulators on rice (Oryza sativa L. Cv CO-43), ISME J., 164, 23. Knief, C., Frances, L., Cantet, F., and Vorholt, J. A.: Cultivation-independent 92e104 (2009). characterization of Methylobacterium populations in the plant phyllosphere by 11. Munusamy, M., Selvaraj, P., Jeounghyun, R., and Tongmin, S.: Regulation of automated ribosomal intergenic spacer analysis, Appl. Environ. Microbiol., 74, ethylene levels in canola (Brassica campestris) by 1-aminocyclopropane-1- 2218e2228 (2008). carboxylate deaminase-containing Methylobacterium fujisawaense, Planta, 224, 24. Chun, J., Lee, J.-H., Jung, Y., Kim, M., Kim, S., Kim, B. K., and Lim, Y.-W.: 268e278 (2006). EzTaxon: a web-based tool for the identification of prokaryotes based on 16S 12. Hornschuh, M., Grotha, R., and Kutschera, U.: Epiphytic bacteria associated ribosomal RNA gene sequences, Int. J. Syst. Evol. Microbiol., 57, 2259e2261 with the bryophyte Funaria hygrometrica: effects of Methylobacterium strains (2007). on protonema development, Plant Biol., 4, 682e687 (2002). 25. Stackebrandt, E. and Goebel, B. M.: Taxonomic note: a place for DNA-DNA 13. Tani, A., Takai, Y., Suzukawa, I., Akita, M., Murase, H., and Kimbara, K.: reassociation and 16S rRNA sequence analysis in the present species definition Practical application of methanol-mediated mutualistic symbiosis between in bacteriology, Int. J. Syst. Bacteriol., 44, 846e849 (1994). Methylobacterium species and a roof greening moss, Racomitrium japonicum, 26. Gallego, V.: Methylobacterium hispanicum sp. nov. and Methylobacterium PLoS One, 7, e33800 (2012). aquaticum sp. nov., isolated from drinking water, Int. J. Syst. Evol. Microbiol., 14. Abanda-Nkpwatt, D., Müsch, M., Tschiersch, J., Boettner, M., and 55, 281e287 (2005). Schwab, W.: Molecular interaction between Methylobacterium extorquens and 27. Kang, Y. S., Kim, J., Shin, H. D., Nam, Y. D., Bae, J. W., Jeon, C. O., and Park, W.: seedlings: growth promotion, methanol consumption, and localization of the Methylobacterium platani sp. nov., isolated from a leaf of the tree Platanus methanol emission site, J. Exp. Bot., 57, 4025e4032 (2006). orientalis, Int. J. Syst. Evol. Microbiol., 57, 2849e2853 (2007). 15. Koskimäki, J. J., Pirttila, A. M., Ihantola, E.-L., Halonen, O., and Frank, A. C.: 28. Green, P. N., Bousfield, I. J., and Hood, D.: Three new Methylobacterium spe- The intracellular Scots pine shoot symbiont Methylobacterium extorquens cies: M. rhodesianum sp. nov., M. zatmanii sp. nov., and M. fujisawaense sp. nov. DSM13060 aggregates around the host nucleus and encodes eukaryote-like Int. J. Syst. Bacteriol., 38, 124e127 (1988). proteins, MBio, 6, e00039ee00115 (2015). 29. Green, P. N. and Bousfield, I. J.: Emendation of Methylobacterium Patt, Cole, 16. Holland, M. A. and Polacco, J. C.: Urease-null and hydrogenase-null pheno- and Hanson 1976; Methylobacterium rhodinum (Heumann 1962) comb. nov. types of a phylloplane bacterium reveal altered nickel metabolism in two corrig.; Methylobacterium radiotolerans (Ito and Iizuka 1971) comb. nov. corrig.; soybean mutants, Plant Physiol., 98, 942e948 (1992). and Methylobacterium mesophilicum (Austin and Goodfellow 1979) comb. nov. 17. Mizuno, M., Yurimoto, H., Yoshida, N., Iguchi, H., and Sakai, Y.: Distribution Int. J. Syst. Evol. Microbiol., 33, 875e877 (1983). of pink-pigmented facultative methylotrophs on leaves of vegetables, Biosci. 30. Liu, Y., Liu, L., Qiu, F., Schumann, P., Shi, Y., Zou, Y., Zhang, X., and Song, W.: Biotechnol. Biochem., 76, 578e580 (2012). Paenibacillus hunanensis sp. nov., isolated from rice seeds, Int. J. Syst. Evol. 18. Mizuno, M., Yurimoto, H., Iguchi, H., Tani, A., and Sakai, Y.: Dominant Microbiol., 60, 1266e1270 (2010). colonization and inheritance of Methylobacterium sp. strain OR01 on perilla 31. Lucas, K. G., Kiehn, T. E., Sobeck, K. A., Armstrong, D., and Brown, A. E.: plants, Biosci. Biotechnol. Biochem., 77, 1533e1538 (2013). Sepsis caused by Flavimonas oryzihabitans, Medicine (Baltimore), 73, 209e214 19. Ji, K.-X., Chi, F., Yang, M.-F., Shen, S.-H., Jing, Y.-X., Dazzo, F. B., and (1994). Cheng, H.-P.: Movement of rhizobia inside tobacco and lifestyle alternation 32. Zarraonaindia, I., Owens, S. M., Weisenhorn, P., West, K., Hampton- from endophytes to free-living rhizobia on leaves, J. Microbiol. Biotechnol., 20, Marcell, J., Lax, S., Bokulich, N. A., Mills, D. A., Martin, G., Taghavi, S., van der 238e244 (2010). Lelie, D., and Gilbert, J. A.: The soil microbiome influences grapevine- 20. Chi, F., Shen, S.-H., Cheng, H.-P., Jing, Y.-X., Yanni, Y. G., and Dazzo, F. B.: associated microbiota, MBio, 6, e02527ee02614 (2015). Ascending migration of endophytic rhizobia, from roots to leaves, inside rice 33. Minami, T., Anda, M., Mitsui, H., Sugawara, M., Kaneko, T., Sato, S., plants and assessment of benefits to rice growth physiology, Appl. Environ. Ikeda, S., Okubo, T., Tsurumaru, H., and Minamisawa, K.: Metagenomic Microbiol., 71, 7271e7278 (2005). analysis revealed methylamine and ureide utilization of soybean-associated 21. Maignien, L., DeForce, E. A., Chafee, M. E., Eren, A. M., and Simmons, S. L.: Methylobacterium, Microbes Environ., 31,268e278 (2016). Ecological succession and stochastic variation in the assembly of Arabidopsis 34. Hardoim, P. R., Hardoim, C. C. P., van Overbeek, L. S., and Van Elsas, J.-D.: thaliana phyllosphere communities, MBio, 5, e00682ee00713 (2014). Dynamics of seed-borne rice endophytes on early plant growth stages, PLoS 22. Tani, A., Sahin, N., Matsuyama, Y., Enomoto, T., Nishimura, N., Yokota, A., One, 7, e30438 (2012). and Kimbara, K.: High-throughput identification and screening of novel