Introduction

Leguminous plants in the genus Lotononis and their associated root nodule bacteria are being studied because of their potential as well-adapted pasture legumes able to combat dryland salinity in southern Australian agricultural systems (Yates et al. 2007). The genus Lotononis is of mainly southern African origin, comprising some 150 species of herbs and small shrubs (Van Wyk 1991). Species in the Listia section are of particular interest, as they are perennial, stoloniferous and lack the poisonous metabolites found in some other species of Lotononis (Van Wyk and Verdoorn 1990). The Listia section includes L. angolensis, L. bainesii, L. listii, L. macrocarpa, L. marlothii, L. minima, L. solitudinis and L. subulata. Nodulation has been described for L. angolensis, L. bainesii and L. listii, which characteristically form collar nodules (Norris 1958; Yates et al. 2007).

The root nodule bacteria from L. bainesii were first described by Norris (1958), who reported that isolates from L. bainesii were red- or pink-pigmented and that the symbiosis was highly specific. These pigmented bacteria were subsequently characterized and identified as a species of Methylobacterium (Jaftha et al. 2002). Yates et al. (2007) further found isolates from L. bainesii, L. listii and L. solitudinis to be pink-pigmented, highly effective, most closely related to Methylobacterium nodulans (with >97% similarity of the 16S rRNA gene sequence) and to form a cross-inoculation group. The non-pigmented M. nodulans that specifically nodulates Senegalese Crotalaria spp. (Sy et al. 2001) is the only other Methylobacterium species so far reported to nodulate legumes.

Free-living methylobacteria are found in a variety of habitats, such as soil, dust, and fresh water (Green 1992). Methylobacteria are also ubiquitous in the plant phyllosphere and rhizosphere (Trotsenko et al. 2001). They promote the germination or growth of soybeans, rice and other plants, probably because of their ability to synthesise auxins, cytokinins, vitamin B12 and other plant growth-promoting substances (Basile et al. 1985; Holland and Polacco 1994; Ivanova et al. 2000; Trotsenko et al. 2001; Madhaiyan et al. 2004; Abanda-Nkpwatt et al. 2006; Ryu et al. 2006). The closeness of the association between plants and Methylobacterium spp. varies; epiphytes (Omer et al. 2004), endophytes (Van Aken et al. 2004) and nitrogen-fixing symbionts (Sy et al. 2001; Jaftha et al. 2002; Yates et al. 2007), have all been described.

Methylobacterium spp. are characterized by their ability to utilize methanol and other C1 compounds, as well as a variety of multicarbon substrates (Green 1992; Lidstrom 2006). Utilization of carbohydrates as a sole carbon source is variable and can be used to differentiate the various species (Green 1992). Methylotrophy in Methylobacterium spp. involves over 100 genes constituting a set of metabolic functional modules (Chistoserdova et al. 2003). In the model organism Methylobacterium extorquens AM1, such modules involve the primary oxidation of methanol or methylamine to formaldehyde, the oxidation of formaldehyde, and the assimilation of C1 products via the serine cycle (Chistoserdova et al. 2003; Lidstrom 2006). Methanol is oxidized by methanol dehydrogenase (MDH), a protein with an α2β2 tetramer structure, a pyrroloquinoline quinone (PQQ) cofactor and a calcium ion, essential for maintaining the PQQ in its active configuration, in the active site of each α-subunit (Anthony 1996; Goodwin and Anthony 1998). The genes encoding the MDH structural polypeptides, the specific cytochrome c electron acceptor, proteins essential for the insertion of the calcium ion, a regulatory protein and several proteins of unknown function are transcribed in a single operon, mxaFGIRSACKLDEHB (Chistoserdova et al. 2003).

The M. nodulans isolates from Crotalaria, like all previously described Methylobacterium species, can use methanol as a sole carbon source and contain a copy of mxaF, the gene that codes for the large subunit of MDH (Sy et al. 2001). Isolates from L. bainesii are also reported as able to grow in minimal media with methanol as a substrate (Jaftha et al. 2002). In contrast, our isolates from L. bainesii, L. listii and L. solitudinis appeared unable to grow on methanol in minimal media. The aim of this work was thus to determine the ability of these isolates to grow on or utilize a variety of C1 and other carbon substrates.

The genomes of five Methylobacterium spp., including the L. bainesii symbiont Methylobacterium sp. 4-46, have recently been sequenced and are available on the Integrated Microbial Genomes database of the Joint Genome Institute (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi) in either draft or finished form. Methylobacterium sp. 4-46 is closely related to xct9 on the basis of 16S rRNA gene homology (Darryl Fleischman, personal communication). We therefore performed BLASTP searches for sequences required for methylotrophy in these genomes, to determine the genetic basis for our isolates’ inability to grow on methanol.

Materials and methods

Bacterial strains, origins and cultural conditions

The bacterial strains (11 Lotononis isolates and four reference strains) used and their collection details are listed in Table  1. Root nodule bacteria were isolated (Yates et al. 2007) from L. bainesii, L. listii and L. solitudinis plants growing in eight sites across north-eastern South Africa, between latitudes 24°S and 30°S and form part of the Western Australian Soil Microbiology (WSM) collection, Murdoch University, Western Australia. Strain xct9 was isolated from a South African L. bainesii nodule. It is synonymous with CB376, the current commercial inoculant for L. bainesii in Australia (Ian Law, personal communication). These root nodule bacteria and the reference strains Sinorhizobium medicae WSM419 and Bradyrhizobium japonicum USDA6 were grown on half lupin agar (1/2 LA) medium (Howieson et al. 1988). All strains were stored in 1/2 LA plus 15% (v/v) glycerol broths at −80°C. M. nodulans ORS 2060 and the non-symbiont Methylobacterium organophilum DSM 760 (kindly supplied by Dr Catherine Boivin-Masson, INRA) were streaked onto agar plates of minimal mineral medium M72 (Belgian Co-ordinated Collection of Micro-organisms 1998) supplemented with 1% (v/v) methanol. They were then grown in broths of M72 plus 1% (v/v) methanol and stored as detailed above.

Table 1 List of strains used in this study
Full size table

Acidification or alkalinization of 1/2 LA medium

The ability of the isolates to acidify or alkalinize 1/2 LA medium was tested on unbuffered 1/2 LA agar plates, adjusted to pH 7.0 and containing 5 ml l−1 of Universal range pH indicator (Vogel 1962). The isolates and reference strains were streaked onto the 1/2 LA plus universal indicator plates and colour change recorded after 7-day incubation at 28°C. Strains were scored as acidifying if the medium turned yellow (pH 6.0), alkalinizing if the medium turned blue/green (pH 8) or strongly alkalinizing if the medium turned blue (pH 9.0).

Growth on sole carbon substrates

General growth procedures

All isolates and reference strains were streaked from −80°C stocks onto fresh 1/2 LA plates, except for ORS 2060, which was streaked onto M72 agar containing 1% (v/v) methanol. All media were adjusted to pH 7.0. All plates were incubated at 28°C for 7 days. Glassware used to grow cultures (McCartney bottles and conical flasks) was soaked in a 10% (v/v) hydrochloric acid solution for at least 24 h prior to use and then rinsed twice in reverse osmosis deionized water. All broth cultures were grown at 28°C with shaking (200 rpm). Lids of McCartney bottles were wrapped with parafilm prior to incubation to prevent contamination. Optical densities were read on a Hitachi U-1100 spectrophotometer.

Growth on methanol and multicarbon substrates

Isolates were tested for growth on arabinose, glucose, galactose, mannitol, succinate, glutamate and methanol as sole carbon sources. Cells were inoculated into 5 ml broths of M72 medium, supplemented with sodium pyruvate (10 mM), yeast extract (0.5 g l−1) and vitamins (thiamine HCl, 1.0 mg l−1; pantothenic acid, 1.0 mg l−1 and biotin, 20 μg l−1) and grown for 40 h to an optical density at 600 nm (OD600) between 0.6 and 0.9. The cultures were centrifuged (20,800g for 30 s), washed twice with 0.89% (w/v) saline, resuspended in M72 medium containing vitamins (M72v) and devoid of carbon source, then added to duplicate 5 ml broths of M72v and one of the carbon substrates to a final OD600 of 0.05. The concentration of all carbon substrates was 20 mM, except for methanol, where the concentration was 1% (v/v) (approximately 260 mM). Several strains were also grown in broths with 50 mM methanol. The methanol and the stock solutions of the other carbon substrates (adjusted to pH 7.0 where necessary) were filter sterilized (0.2 μm filter) and added to the autoclaved M72v medium prior to inoculation. Inoculated culture media were incubated for 10 days before a visual assessment was made. Growth on the carbon substrate was assessed as being no growth (OD600 was the same as for the minus carbon substrate control), poor (0.1 < OD600 < 0.2), moderate (0.2 < OD600 < 0.5) or abundant (OD600 > 1.0). Two negative controls were used—an uninoculated control containing M72v medium and various carbon sources and a control of M72v devoid of carbon substrate, but containing bacterial inoculant.

Growth on C1 sole carbon substrates

Isolates were examined for growth on methanol (0.2%, v/v), methylamine (0.1, 0.2 or 0.5%, v/v), formaldehyde (0.5 or 1.0 mM) and formate (30 mM) as sole carbon sources in JMM medium (O’Hara et al. 1989) devoid of galactose and arabinose and with NH4Cl (10 mM) replacing glutamate as a nitrogen source. Bacteria capable of utilizing C1 carbon sources have previously been shown to grow on these substrates at these concentrations, while formaldehyde is toxic above 1 mM (Marx et al. 2003; Miller et al. 2005). JMM medium with succinate (20 mM) as a sole carbon source served as a positive control. Isolates were also examined for growth in medium containing succinate (20 mM) plus methanol (1.0%, v/v). Stock solutions of methylamine and other carbon sources were adjusted to pH 7.0 with HCl if required. Inoculum was prepared and grown as described above, but with JMM replacing M72v medium.

Growth of WSM2799 on formate

Cells were inoculated into broths of JMM medium containing succinate (20 mM) and NH4Cl (10 mM) as carbon and nitrogen sources, and grown for 40 h to an OD600 of 0.6. The cultures were centrifuged and washed as described, resuspended in JMM medium (containing either formate or succinate) and added to duplicate 250 ml conical flasks containing pre-warmed 50 ml JMM broths with either succinate (20 mM) or formate (30 mM) as sole carbon source to give an initial OD600 of 0.05. Duplicate samples were taken from each flask at regular intervals for OD600 readings.

Biochemical assays for utilization of methanol

Cells of xct9, USDA6 and ORS 2060 were inoculated into broths of M72 medium supplemented with sodium pyruvate (10 mM), yeast extract (0.5 g l−1) and vitamins. The cultures were grown for 40 h to an OD600 of between 0.6 and 0.9, centrifuged and washed as described and resuspended in M72v medium, then added to duplicate 100 ml conical flasks containing 20 ml of M72v medium and either 25 or 100 mM methanol, to a final OD of 0.05. These and duplicate uninoculated controls were incubated for 70 h. Aliquots (1 ml) were taken from each flask at 0, 22, 40 and 70 h, growth measured as OD600 and the supernatant stored at −20°C after centrifugation (20,800g for 30 s).

Sulfuric acid (0.5 M) was added to the supernatant samples (10 μl from the cultures in 100 mM methanol, 40 μl from the 25 mM methanol cultures), to make a total volume of 1 ml. For ORS 2060, the 40 and 70-h 100 mM methanol samples were also of 40 μl. The samples were oxidized with potassium permanganate, excess permanganate removed with sodium arsenite (Wood and Siddiqui 1971), and formaldehyde concentration determined by measuring OD412 after reaction with Nash’s reagent (Nash 1953).

The concentration of methanol in the succinate (20 mM) plus methanol (1% v/v) media 10 days after inoculation was determined in the same way.

PCR amplification of mxaF

PCR amplification of mxaF was performed using the primers f1003—5′-GCG GCA CCA ACT GGG GCT GGT-3′ and r1561—5′-GGG CAG CAT GAA GGG CTC CC-3′. Primers were obtained from GeneWorks Pty Ltd.

Whole cell DNA templates were prepared from isolates and reference strains. Cells were suspended in PCR-grade water to an OD600 of approximately 10. An initial PCR amplification to optimize the magnesium chloride concentration of the PCR reaction mix resulted in subsequent reactions using 1.5 mM MgCl2. The total volume of the PCR reaction mix was 20 μl, consisting of 4 μl of 5× PCR polymerization buffer (Fisher Biotec), 9.8 μl of PCR-grade water, 4 μl of 7.5 mM MgCl2, 0.5 μl of each 50 μM primer, 0.2 μl of Taq DNA polymerase (5 U μl−1) (Invitrogen) and 1 μl of DNA template. The PCR conditions for the thermocycler were an initial 4 min at 94°C; 31 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 1 min; then 1 cycle of 94°C for 1 min, 55°C for 1 min and 72°C for 5 min. The samples were then subjected to gel electrophoresis, or storage at −20°C.

The PCR amplification product was visualized after gel electrophoresis using a 1% (w/v) agarose in TAE gel submerged in TAE running buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.0) run at 80 V for about 2 h with loading dye added to each sample prior to electrophoresis. A 1 kb DNA ladder (Promega) was used as a marker. The gels were stained in 0.5 μg ml−1 ethidium bromide for 40 min, destained in deionized water and visualized under UV light. Images of the gels were captured using the Gel Doc 2000 (BioRad) system.

Comparative genomics of Methylobacterium spp.

Homologs of genes required for methylotrophy were identified using BLASTP searches of sequences from the well-studied facultative methylotroph M. extorquens AM1 (Chistoserdova et al. 2003) against the five sequenced Methylobacterium spp. [M. chloromethanicum CM4, M. extorquens PA1, M. nodulans ORS 2060, M. populi BJ001 and Methylobacterium sp. (Lotononis) 4-46] and the S. medicae WSM419 genomes deposited in the Integrated Microbial Genomes database of the Joint Genome Institute.

Results

Acidification or alkalinization of 1/2 LA media

The Lotononis isolates all alkalinized or strongly alkalinized the media, consistent with organic acid utilization. USDA6 and ORS 2060 also alkalinized the media, while WSM419 caused acidification.

Growth on sole carbon substrates

Isolates from L. bainesii, L. listii and L. solitudinis were highly selective in their utilization of carbon sources (Table 2). None was able to utilize arabinose, galactose, glucose or mannitol as sole carbon sources. All the isolates from Lotononis spp. grew well on succinate and glutamate, but none grew on methanol as a sole carbon source.

Table 2 Visual rating of growth of strains on various sole carbon sources after 10 days incubation
Full size table

Growth on C1 sole carbon substrates

Neither the Lotononis isolates nor WSM419 grew on methanol, methylamine or formaldehyde at any of the concentrations supplied in the media (Table 3). ORS 2060 grew on methanol, but not on methylamine or formaldehyde. Growth of the Lotononis isolates on formate was variable. WSM2603, WSM2666, WSM2678, WSM2693, WSM3032, WSM3034 and WSM3035 did not grow, while WSM2598, WSM2660 and xct9 grew poorly and WSM2799 grew moderately well. WSM419 grew poorly and ORS 2060 grew moderately well on formate. To determine whether methanol had an inhibitory effect on growth, all strains were also grown in broths containing both succinate and methanol as carbon sources (Table 3). All of the Lotononis isolates and WSM419, as well as ORS 2060, grew abundantly in the succinate plus methanol medium.

Table 3 Visual rating of growth of strains on various sole C1 carbon sources after 10 days incubation
Full size table

Growth of WSM2799 on formate

The growth rate of logarithmic phase cells of WSM2799 growing on formate was compared with that of succinate-grown cells by measuring OD600. The mean generation time (MGT) for WSM2799 grown on formate was 24 h, compared with 5.5 h MGT when grown on succinate.

Biochemical assays for utilization of methanol

Utilization by xct9, ORS2060 and USDA6 over 70 h

Three strains (xct9, ORS 2060 and USDA6) were chosen for determination of methanol utilization at two different methanol concentrations (25 and 100 mM). The L. bainesii isolate xct9 was included as Jaftha et al. (2002) reported that it grew on M72 medium containing methanol as a sole carbon source. M. nodulans ORS 2060 and B. japonicum USDA6 served as positive and negative controls, respectively. Over a period of 70 h and at both 25 and 100 mM of methanol, ORS 2060 was the only strain for which the concentration of methanol in the medium notably decreased (Fig. 1). The methanol concentrations in the cultures of xct9 and USDA6 showed slight decreases equal to those in uninoculated media. Over the same period of time, and at both 25 and 100 mM methanol, the OD600 of the ORS 2060 culture increased, whereas the OD600 of the xct9 and USDA6 cultures remained unchanged.

Fig. 1
figure 1

Non-utilization of methanol by the L. bainesii isolate xct9. Methylobacterium nodulans strain ORS 2060 was used as a positive control. An uninoculated control was also used. Growth of xct9 (filled squares), growth of ORS 2060 (filled triangles) and OD600 of the uninoculated control (filled circles) are shown. MeOH concentrations in supernatants of cultures of xct9 (open squares) and ORS 2060 (open triangles) and of the uninoculated control (open circles) are also shown

Full size image

Utilization of methanol in succinate plus methanol media

The concentration of methanol in inoculated medium containing both succinate and methanol (Table 3) was determined for all strains after 10-day incubation. In media inoculated with the Lotononis isolates or with WSM419, the methanol concentration was similar to that of the uninoculated control, but decreased by 50% in media inoculated with ORS 2060 (results not shown).

PCR amplification of mxaF

To evaluate the presence of methanol dehydrogenase genes in the Lotononis isolates, PCR amplification of mxaF, the gene that codes for the large subunit of methanol dehydrogenase, was performed on Lotononis isolates WSM2603, WSM2660, WSM2666, WSM2678, WSM2693, WSM2799, WSM3035 and xct9 and the methylotrophic strains ORS 2060 and DSM 760. The primers used have previously been shown to specifically amplify methylotrophic DNA (McDonald et al. 1995). The correct size 555-bp PCR product was obtained only for ORS 2060 and DSM 760. No amplification was obtained for any of the Lotononis isolates, or for the reaction devoid of template DNA.

Comparative genomics of Methylobacterium spp.

Several different homologs of the M. extorquens AM1 α-subunit of MDH, coded for by mxaF, were found in each of the genomes of the five sequenced methylobacteria and of S. medicae WSM419. However, only the genomes of the methylotrophic M. chloromethanicum CM4, M. extorquens PA1 and M. populi BJ001 had a mxaF homolog present in a mxaFGIRSACKLDEHB operon. The M. nodulans ORS 2060 mxaF appeared to be in a separate section of the genome, but this may be due to its genome sequence being still in a draft state. The remaining ORS 2060 mxa genes were present in an orthologous operon. These four mxaF homologs had a high similarity (>88% over the length of the protein). In contrast, mxaF homologs found in Methylobacterium sp. 4-46 and S. medicae WSM419 were not in an orthologous operon and had a similarity of only 50% or less. No homologs of more than 50% similarity could be found in Methylobacterium sp. 4-46 for the MDH β-subunit (mxaI), the specific cytochrome c electron acceptor (mxaG), proteins essential for Ca2+ insertion (mxaACKL) or for the protein products of mxaSH. Similarly, no homologs were found in WSM419 for the products of mxaGIACKLD.

Discussion

Methylobacterium species are described as able to grow on methanol, formaldehyde and formate (Green 1992; Lidstrom 2006). Sy et al. (2001) demonstrated that M. nodulans, isolated from root nodules of Crotalaria, was able both to fix nitrogen symbiotically and to utilize methanol as a sole carbon source. They also confirmed the presence in M. nodulans of mxaF, the gene that codes for the alpha subunit of methanol dehydrogenase. In contrast, our results from the carbon substrate utilization tests, the methanol assay and the mxaF PCR indicate that our isolates from L. bainesii, L. listii and L. solitudinis differ from other described Methylobacterium species (Green 1992; Gallego et al. 2006; Kato et al. 2008) in being unable to utilize or oxidize methanol. Kato et al. (2008) did, however, report that two of their Methylobacterium strains, isolated from freshwater, only grew weakly on methanol as a sole carbon source. Our findings also differ from those of Jaftha et al. (2002), who reported that their L. bainesii isolates grew in the presence of methanol as a sole carbon source. Consistent with our results, the L. bainesii isolate CB376 has previously been described as unable to utilize methanol (O’Brien and Murphy 1993), but was identified as a species of Rhizobium at the time. The ability of the Lotononis isolates to grow in media containing succinate and methanol (1% v/v) indicates that their inability to grow on methanol as a sole carbon substrate is not due to any toxic effects of methanol at this concentration.

In methylobacteria, methanol dehydrogenase is essential for the oxidation of methanol to formaldehyde. The negative result for the PCR amplification of the mxaF gene suggests that the Lotononis isolates in this study lack at least one of the genes required to synthesize methanol dehydrogenase. Searches for methylotrophy genes in the sequenced genome of Methylobacterium sp. 4-46, isolated from L. bainesii, indicate that the inability to utilize methanol may be due to the absence of the mxa operon, which is present in the other sequenced methylotrophic Methylobacterium genomes, and the products of which are required for the primary oxidation of methanol (Amaratunga et al. 1997; Chistoserdova et al. 2003). It is likely that the mxaF homologs with similarity <50% that are present in these genomes code for PQQ-dependent alcohol or glucose dehydrogenases that are not specific for methanol as a substrate.

Most species of Methylobacterium can only grow on a narrow range of carbohydrates (Green 1992). The growth of the L. bainesii, L. listii and L. solitudinis isolates on other sole carbon sources is consistent with the substrate utilization patterns of other Methylobacterium species and with results obtained from other L. bainesii isolates (Jaftha et al. 2002). None of these isolates grew on the sugars or sugar alcohol supplied. Their ability to grow on succinate and to alkalinize 1/2 LA plus universal indicator media also suggests that they preferentially utilize organic acids, including dicarboxylic acids, which are the usual form of carbon supplied to nitrogen-fixing bacteroids within legume nodules (Lodwig and Poole 2003). M. nodulans ORS2060, in contrast, appears to be able to use methanol within the Crotalaria podocarpa nodule (Jourand et al. 2005).

Plants produce methanol as a by-product of pectin metabolism, with high pectin methyl esterase activity correlating with areas of rapid growth, such as seen in seedlings (Obendorf et al. 1990). Studies have suggested that the ability of Methylobacterium spp. to utilize this methanol confers an advantage in plant colonization (Corpe and Rheem 1989; Sy et al. 2005) and contributes to the effectiveness of the specific C. podocarpa/M. nodulans symbiosis (Jourand et al. 2005). Yates et al. (2007) have shown that the symbiosis between L. bainesii, L. listii and L. solitudinis and their nodulating methylobacteria is a highly effective one. Our results suggest that the inability to utilize methanol is not deleterious to effective colonization or symbiosis in the association between these methylobacteria and Lotononis and factors other than methylotrophy must be implicated in the specificity of the symbiosis.