Methylobacterium populi sp. nov., a novel aerobic, pink-pigmented, facultatively methylotrophic, methane-utilizing bacterium isolated from poplar trees (Populus deltoides×nigra DN34)

Species of the genus Methylobacterium are strictly aerobic, facultatively methylotrophic, Gram-negative, rod-shaped bacteria that are able to grow on one-carbon compounds (e.g. methanol or methylamine), as well as on a variety of C₂, C₃ and C₄ substrates (Green, 1992). Only the type species, Methylobacterium organophilum, has been shown to use methane as the sole source of carbon and energy (Patt et al., 1976). The genus Methylobacterium belongs to the α2 subclass of the Proteobacteria and currently consists of 14 species with validly published names (Heumann, 1962; Ito & Iizuka, 1971; Kouno & Ozaki, 1975; Patt et al., 1976; Rock et al., 1976; Austin & Goodfellow, 1979; Green & Bousfield, 1983; Urakami & Komagata, 1984; Bousfield & Green, 1985; Green et al., 1988; Urakami et al., 1993; Wood et al., 1998; Doronina et al., 2000, 2002; McDonald et al., 2001; Sy et al., 2001). Members of the genus Methylobacterium are distributed in a wide variety of natural and man-made environments, including soil, air, dust, fresh- and marine water and sediments, water supplies, bathrooms, air-conditioning systems and masonry (Hiraishi et al., 1995; Trotsenko et al., 2001). Some species have been described as opportunistic human pathogens (Truant et al., 1998; Hornei et al., 1999). In addition, methylotrophic bacteria are frequently associated with terrestrial and aquatic plants, where they colonize roots and leaf surfaces (Austin et al., 1978; Yoshimura, 1982; Corpe & Rheem, 1989; Trotsenko et al., 2001; Lidstrom & Chistoserdova, 2002). The association of Methylobacterium species with plants seems to rely on a symbiotic relationship between the bacterium and the plant host. Plants produce methanol (representing nearly 50 % of the total volatile atmospheric organic carbon), which is toxic and is used by Methylobacterium species as the sole source of carbon and energy (i.e. the methanol cycle; Trotsenko et al., 2001). In response, Methylobacterium species produce phytohormones (cytokinins and auxins), which are known to stimulate plant growth (Ivanova et al., 2001; Koenig et al., 2002), fix atmospheric nitrogen (Sy et al., 2001) or help plants to fight pathogens (Holland & Polacco, 1994). Bacteria are often pink to red, due to the presence of carotenoids, and are referred to as pink-pigmented facultative methylotrophs. Members of the genus Methylobacterium are highly resistant to dehydration, freezing, chlorine, UV and ionizing radiation and elevated temperatures (Trotsenko et al., 2001). Methylobacterium species are known to metabolize a range of toxic organic chemicals, such as methyl chloride (McDonald et al., 2001), methyl bromide (Goodwin et al., 2001), methyl iodide (Schaefer & Oremland, 1999), dichloromethane (Doronina et al., 2000), ethylated sulfur-containing compounds (de Zwart et al., 1996), methylated amines (Trotsenko et al., 2001), methyl tert-butyl ether (Mo et al., 1997) and cyanate and thiocyanate (Wood et al., 1998). In this paper, the formal taxonomic description of a novel Methylobacterium strain, BJ001^T, isolated from poplar tissues and able to use methane as the sole source of carbon and energy, is reported.

Methylobacterium sp. strain BJ001^T was isolated from poplar plantlets and from tissue cultures (Populus deltoides×nigra DN34) that were developed initially from surface-sterilized explants and maintained under axenic conditions (Van Aken & Schnoor, 2002). Images of tissue cultures containing strain BJ001^T are available in Supplementary Fig. A in IJSEM Online. Poplar plantlets were surface-sterilized before bacterial isolation. The isolated bacterium was maintained routinely on Luria–Bertani (LB) medium, supplemented with fructose (5 g l⁻¹), or on selective methanol mineral medium (Green, 1992). Morphological properties were studied according to general protocols (Gerhardt et al., 1994). Scanning electron microscope (SEM) observations were performed on fixed material that was prepared for routine examination by glutaraldehyde fixation, osmium tetroxide post-fixation and graded ethanol dehydration (Bozzola & Russell, 1998). Samples were critical point-dried, mounted on stubs, sputter-coated with gold/palladium and visualized by using a Hitachi S-4000 SEM equipped with a field-emission electron source. Carbon-source utilization tests were performed by using a standard protocol described by Green & Bousfield (1982). Cellular fatty acids were extracted according to Bligh & Dyer (1959), purified on Sephadex beads (Amersham Biosciences) and analysed by GC-mass spectroscopy (Gerhardt et al., 1994). DNA G+C content was determined by HPLC analysis of individual nucleosides, resulting from DNA hydrolysis and dephosphorylation (Mesbah et al., 1989). DNA manipulations were carried out according to standard protocols (Ausubel et al., 1999; Sambrook & Russell, 2000). 16S and 16S–23S intergenic spacer (IGS) rDNA analyses were performed by PCR amplification using the following primers: 27f (positions 11–27 of bacterial 16S rDNA, Escherichia coli numbering), 1522r (positions 1492–1522), 926f (positions 901–926) and 115r/23S (positions 97–115 of bacterial 23S rDNA, E. coli numbering) (Hurek et al., 1997; Tan et al., 2001). PCR conditions were as described by Tan et al. (2001). PCR products were cloned in a pGEM vector (Promega) and submitted to the University of Iowa DNA Core (Iowa City, IA, USA) for sequencing. Determined rDNA and reference sequences from GenBank were aligned by using clustal_w multiple alignment and bioedit (version 5.0.9) software. The tree topology was inferred by the parsimony method (heuristic search) using paup (version 4.0) software (Sinauer Associates). DNA–DNA hybridization was carried out according to Doronina et al. (2002), using a method based on that of Denhardt (1966). Unlabelled, denatured DNA of Methylobacterium species was immobilized on nitrocellulose membranes (Bio-Rad) (Ausubel et al., 1999; Sambrook & Russell, 2000). Reference DNA from strain BJ001^T was labelled with deoxy[1′,2′,5-³H]cytidine 5′-triphosphate (57 Ci mmol⁻¹=2·1×10⁶ MBq mmol⁻¹) by using a Nick Translation kit (N5500; Amersham Biosciences). After labelling, reference DNA exhibited a specific activity of 1·47×10⁵ Bq μg⁻¹ and a mean size of 400–600 bp (as determined on 2·0 % agarose gel). Labelled DNA was denatured at 100 °C for 10 min. Hybridization was performed according to Denhardt (1966), with a ratio of labelled to immobilized DNA of 1 : 100 (Doronina et al., 2002). Low-stringency (pre)-hybridization solution consisted of 30 % (v/v) formamide, 1× SSC, 5× Denhardt's solution, 1·0 % SDS and 100 μg denatured salmon sperm DNA ml⁻¹. Radioactivity retained on the membrane was determined by liquid scintillation counting.

The bacterium isolated from poplar tissues, strain BJ001^T, was identified by phylogenetic analysis as a member of the genus Methylobacterium (Fig. 1⇓). Members of this genus are known to colonize the rhizosphere and phyllosphere of a variety of plant species (Austin et al., 1978; Yoshimura, 1982; Corpe & Rheem, 1989; Holland & Polacco, 1994; Trotsenko et al., 2001). However, this is the first report of an endophytic association with a poplar tree (Populus sp.). The isolate had the following characteristics of the genus Methylobacterium (Green, 1992). Cells were rod-shaped (0·8–1·0×1·0–10·0 μm), frequently branched and occurred singly or in rosettes (Fig. 2⇓). They exhibited polar growth and were motile by a single polar or lateral flagellum. Photomicrographs of strain BJ001^T are available in Supplementary Figs B and C in IJSEM Online. Cells stained Gram-negative and colonies were pink to red. Cells were strictly aerobic and catalase- and oxidase-positive (Gerhardt et al., 1994). Due to the chemotaxonomic homogeneity of the genus Methylobacterium, phylogenetic analyses constitute a critical tool for species identification (Green & Bousfield, 1982; Doronina et al., 2002). According to 16S rDNA sequences, the closest relatives of strain BJ001^T were Methylobacterium thiocyanatum, Methylobacterium extorquens, Methylobacterium zatmanii and Methylobacterium rhodesianum, with 99·3, 99·1, 98·6 and 98·5 % 16S rDNA sequence similarity, respectively, corresponding to the interspecies separation level of the genus Methylobacterium (94·2–99·4 %; Doronina et al., 2002). On the basis of 16S–23S IGS rDNA sequences, strain BJ001^T shared 78·7–82·1 % similarity with M. extorquens (GenBank accession nos AF293375 and AF338180) and 66·5 % similarity with the type species, M. organophilum (GenBank accession no. AF338181). The 16S and 16S–23S IGS rDNA sequence of strain BJ001^T and sequence similarity matrices are available in Supplementary Tables A and B in IJSEM Online. Levels of DNA relatedness between strain BJ001^T and its closest relatives, as determined by DNA–DNA hybridization, were 15–59 %, which indicates that strain BJ001^T can be separated from other members of the genus Methylobacterium (Table 1⇓). Phenotypic differences between Methylobacterium species are limited and often rely on utilization of carbon and energy sources (Green, 1992). Like other members of the genus, strain BJ001^T grew on C₁ substrates, such as methanol, methylamine, formate and formaldehyde. In addition, strain BJ001^T utilized methane, an ability that is shared with only one other species of the genus, M. organophilum (Patt et al., 1976). Strain BJ001^T may play an important ecological role by consuming methane, the greenhouse effect of which is 20 times higher than that of carbon dioxide (Trotsenko et al., 2001). Strain BJ001^T differed from its closest relatives (i.e. M. thiocyanatum, M. extorquens, M. zatmanii and M. rhodesianum) in several other carbon-source utilization features (summarized in Table 2⇓). M. thiocyanatum grows on glucose, arabinose, glutamate, citrate, cyanate and thiocyanate, M. zatmanii grows on trimethylamine and M. rhodesianum grows on dimethylamine, none of which support growth of strain BJ001^T. On the other hand, M. extorquens does not use fructose, M. zatmanii does not use betaine and M. rhodesianum does not use tartrate, all of which are substrates for strain BJ001^T (Rock et al., 1976; Urakami & Komagata, 1984; Green, 1992; Wood et al., 1998). Fructose, the first hexose synthesized by plant photosynthesis, was by far the best carbon substrate for strain BJ001^T, which may be related to its association with poplar trees. Tables of carbon- and nitrogen-source utilization and enzymic reactions of BJ001^T are available as Supplementary Tables D, E and F in IJSEM Online.

Strain BJ001^T has been shown to mineralize the toxic explosives hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and octahydro-1,3,5,7-tertanitro-1,3,5,7-tetrazocine (HMX), which contaminate numerous military sites in the USA, to CO₂ (data not shown). Whereas this extends the list of organic pollutants that are potentially biodegraded by Methylobacterium species, this particular ability has not been tested among other members of the genus and cannot be regarded as a diagnostic test. According to their widespread distribution in a diversity of environments, Methylobacterium species may be suited for natural attenuation or in situ bioremediation (including phytoremediation) of polluted sites.

On the basis of its 16S and 16S–23S IGS rDNA sequence similarity data, DNA–DNA hybridization values, carbon-source utilization pattern (including the use of methane) and endophytic association with poplar trees, strain BJ001^T is proposed as the type strain of a novel Methylobacterium species, with the name Methylobacterium populi sp. nov.