ARTICLE IN PRESS Soil Biology & Biochemistry 39 (2007) 1680–1688 www.elsevier.com/locate/soilbio Lotononis angolensis forms nitrogen fixing, lupinoid nodules with phylogenetically unique, fast-growing, pink-pigmented bacteria, which do not nodulate L. bainesii or L. listii R.J. Yatesa,b, J.G. Howiesona,b,, W.G. Reevea, K.G. Nandasenaa, I.J. Lawc, L. Bräua, J.K. Ardleya, H.M. Nistelbergera, D. Reala, G.W. O’Haraa a Centre for Rhizobium Studies, Murdoch University, Perth, WA 6150, Australia Department of Agriculture Western Australia, Baron-Hay Court, South Perth, WA 6151, Australia c ARC-Plant Protection Research Institute, Private Bag X134, Queenswood 0121, South Africa b Received 18 October 2006; received in revised form 5 January 2007; accepted 10 January 2007 Available online 1 March 2007 Abstract Root-nodule bacteria that nodulate the legume genus Lotononis are being investigated to develop new forage species for agriculture. Bacteria isolated from nodules of Lotononis angolensis were fast-growing, highly mucoid and pink-pigmented, and on the basis of 16S rRNA phylogeny o94% related to other genera in the Alphaproteobacteria. Root-nodule bacteria isolated from other Lotononis species (L. bainesii, L. solitudinis and L. listii) resembled the more common dry, slow-growing, pink-pigmented rhizobia previously described for L. bainesii. These isolates could be attributed to the Methylobacterium genus, although not to the type species Methylobacterium nodulans. Further differences were uncovered with nodulation studies revealing that nodule isolates from L. angolensis were effective at nitrogen fixation on their host plant, but could nodulate neither L. bainesii nor L. listii. Reciprocal tests showed isolates from L. bainesii, L. listii and L. solitudinis were incapable of nodulating L. angolensis effectively. Nodule morphology for L. bainesii, L. angolensis and L. listii was characteristically lupinoid, with little structural divergence between the species, and with nodules eventually enclosing the entire root. r 2007 Published by Elsevier Ltd. Keywords: Lotononis; Root-nodule bacteria; Nitrogen fixation; Methylobacterium; Phylogeny; Nodulation 1. Introduction The genus Lotononis contains approximately 150 species of herbs and small shrubs in the tribe Crotalarieae of the sub-family Fabaceae (Van Wyk, 1991). Their distribution is mainly in southern Africa with a few species extending elsewhere in Africa, the Mediterranean and central Asia (Van Wyk, 1991). Three Lotononis species of current agronomic interest, L. bainesii, L. angolensis and L. listii (previously Listia heterophylla) are taxonomically positioned in the section Listia, that overall contains eight species. The remaining species are rare and in some Corresponding author. Centre for Rhizobium Studies, Murdoch University, Perth WA 6150, Australia. E-mail address: jhowieso@murdoch.edu.au (J.G. Howieson). 0038-0717/$ - see front matter r 2007 Published by Elsevier Ltd. doi:10.1016/j.soilbio.2007.01.025 instances considered endangered; these include L. macrocarpa, L. marlothii, L. minima, L. solitudinis and L. subulata (Van Wyk, 1991). Norris (1958, 1959) first reported rhizobia from L. bainesii as ‘red in colour’ and speculated that other species from the genus Lotononis may also possess pigmented rhizobia. The red or pink colouration is due to intracellular carotenoid that, together with the high resistance of these bacteria to ultraviolet irradiation, is speculated to be of ecological advantage (Godfrey, 1972; Law, 1979). Nodulation studies in the 1960s recorded effective cross-nodulation between strains isolated from L. bainesii and L. listii, but not with L. macrocarpa (R. Date pers. comm.). Field observations from this era suggest that these strains may not have been effective in nitrogen fixation with L. angolensis, as this species was recorded as ‘‘unproductive’’ when inoculated with the ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 L. bainesii inoculant (K. Sandman, 1950, unpublished; ‘t Mannetje, 1967). Pink-pigmented isolates from L. bainesii nodules have been characterised as methylobacteria (Jaftha et al., 2002). The capacity for methylotrophic strains to nodulate was first described by Samba et al. (1999) for three Crotalaria species, and the name Methylobacterium nodulans was subsequently proposed (Sy et al., 2001). However, the classification of the microsymbionts for Lotononis is still in its infancy. Within southern Australia, the Lotononis species have recently been identified as perennial pasture legumes with the potential to reduce ground water recharge to assist in controlling dryland salinity (Yates et al., 2006). Whereas L. bainesii has previously been exploited in sub-tropical areas of Australia (cv. Miles; Bryan, 1961) and Uruguay (cv. INIA Glencoe; Real et al., 2005), it has not been evaluated in Mediterranean-type climates. Although it is an unusual step to evaluate sub-tropical legume species for their adaptation to a dry Mediterranean-type climate, the increasing frequency of summer rainfall events and the availability of water stored in the root-zone may support some hardy species through a summer dry period (Howieson et al., 2007). Diatloff (1977) showed that Lotononis rhizobia became well-established after introduction and were stable in acidic heath sands of Queensland. This complements our studies, which indicate that L. bainesii isolates have the ability to persist in acid and infertile sandy soils of Western Australia (Yates et al., 2006). With renewed interest in the agronomic potential of this legume genus, it is imperative that we fully understand the various root-nodulating organisms associated with it. Hence, this study aimed to identify and characterise rhizobia isolated from the nodules of L. angolensis, L. bainesii, L. listii and L. solitudinis. We also report on the nodule ultrastructure of these species. 1681 2. Material and methods 2.1. Bacterial micro-symbionts and legume hosts The strains of rhizobia investigated are listed in Table 1, together with their host plant and collection details. All strains originally isolated from Lotononis nodules were authenticated on their original hosts. Cultures were grown at 28 1C on modified 12 Lupin Agar (LA) (Howieson et al., 1988). The medium contained: (in g/L) mannitol (5.0); D-glucose (5.0); yeast extract (1.25); agar (18); and (in mM) MgSO4  7H2O (3.2); NaCl (1.7), CaCl2  2H2O (1.4); and (in mM) K2HPO4 (100); KH2PO4 (100); FeSO4 (18); Na2B4O7 (6); MnSO4  4H2O (12); ZnSO4  7H2O (0.76); CuSO4  5H2O (0.32) and Na2MoO4  2H2O (0.54). The pH was adjusted to 6.8 with 0.1 M NaOH. WSM3686 and WSM3674 were grown for 3 days prior to inoculation while the remaining strains were grown for 7 days. Colony growth rates and morphology of strains isolated from Lotononis spp. were compared with the type strain M. nodulans (ORS2060). The legume species and origin of seed used in glasshouse experiments are shown in Table 2. In addition to the three species of Lotononis (L. angolensis, L. bainesii and L. listii), Macroptilium atropurpureum was included because of its known promiscuous nature with a range of rhizobia (Thies et al., 1991; Perret et al., 2000; Yates et al., 2004). L. solitudinis could not be used in cross-inoculation and nodulation experiments because of the rarity of the species and consequent lack of seed. 2.2. General glasshouse procedures Four days before the commencement of experiments, seeds were surface sterilised by immersion in 70% (v/v) ethanol for 60 s and then 4% (w/v) NaHClO4 for 30 s, Table 1 Origin of root-nodule bacteria used in this study Strain a ORS2060 USDA6b WSM2598c WSM2693c WSM2799c WSM3032c WSM3674c WSM3686c CB376e XCT17f a Host plant Geographic origin Source (year isolated) Crotalaria podocarpa Glycine max Lotononis bainesii Lotononis listii Lotononis listii Lotononis solitudinis Lotononis angolensis Lotononis angolensis Lotononis bainesii Lotononis bainesii Senegal Japan South Africa South Africa South Africa South Africa Zambia Zambia Kenya South Africa Samba (1999) Aso (1929) Yates, Real and Law (2002) Yates, Real and Law (2002) Yates, Real and Law (2002) Yates, Van Wyk and Law (2005) Verboom (1963)d Verboom (1963)d Botha (1954) Law (1982) Isolate obtained from Dr. Catherine Boivin-Masson (LSTM), France. Isolate described by Samba et al. (1999). USDA ARS National Rhizobium Germplasm Collection, Beltsville, USA. c Western Australian Soil Microbiology (WSM) culture collection, Centre for Rhizobium Studies, Murdoch University, Perth, Australia. d Re-isolated from mixed cultures following passage through the host. e CSIRO culture collection, Brisbane, Australia (now resident at the University of Western Sydney). f Agricultural Research Council (ARC)—Plant Protection Research Institute culture collection, Pretoria, South Africa. b ARTICLE IN PRESS 1682 R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 Table 2 Acquisition details of legumes used in glasshouse experiments Legume species Accession/line no. Experiment Lotononis angolensis Lotononis angolensis Lotononis bainesii Lotononis bainesii Lotononis listii Macroptilium atropurpureum 8363a 8369a AusTRCF47575b cv. Milesb 2004CRSL69c cv. Siratrob 1 2 2 1 and 2 1 and 2 1 a ARC—LBD Animal Production Institute, Pretoria, South Africa. AusPGRIS (Australian Plant Genetic Resource Information Service), QDPI, Australia. c CRS—Centre for Rhizobium Studies, Legume breeding program, Murdoch University, Perth, Australia. b followed by rinsing in six changes of sterile distilled water. Seeds were germinated in the dark at 25 1C on the surface of water agar (1.5%, w/v). The base of the hypocotyl of each seedling was inoculated at planting with 1 mL of a cell suspension aseptically washed from two 12 LA plates with 30 mL sterile 1% (w/v) sucrose solution. Plants were grown for 56 days in a naturally-lit phytotron maintained at 22 1C. At harvest, roots were carefully excavated, washed and scored for the presence and colour of nodules (Glasshouse experiment 1). 2.2.1. Glasshouse experiment 1—nodulation studies The first nine strains of nodule bacteria listed in Table 1 were assessed for their capacity to nodulate four different legumes using a N-free closed vial (500 mL) growth system (Yates et al., 2004). The strains were inoculated separately onto L. angolensis (SA8363), L. bainesii (cv. Miles), L. listii (2004CRSL69) and M. atropurpureum (cv. Siratro). Two seedlings of the same species were placed into a vial then inoculated with a single strain. All treatments plus uninoculated controls were replicated four times and containers arranged in completely randomised blocks. Selected plants were transplanted into pots and grown to maturity for further study of nodule morphology. 2.2.2. Glasshouse experiment 2—nitrogen fixation studies The seven strains, which nodulated Lotononis in experiment 1, were further examined for their nitrogen-fixing ability in the axenic sand culture system previously described by Howieson et al. (1995). Briefly, this system is deficient in combined nitrogen necessary for plant growth, but adequate for all other nutrients essential for maximum plant growth, and hence the biomass achieved through symbiosis is a direct measure of nitrogen fixation. For assessment of nitrogen fixation by the symbiosis, shoot tops were harvested, dried at 70 1C for 5 days, and then weighed. Strains were inoculated onto four hosts: L. bainesii (cv. Miles), L. bainesii (AusTRCF47575), L. listii (2004CRSL69) and L. angolensis (SA8369) (Table 2). The experiment was a split-plot design, where the individual rhizobial strain formed the main treatment with two different Lotononis species as sub-treatments (two plants of each species per pot). All main treatments were replicated three times and the sub-treatments six times, with pots arranged in a randomised block structure. The experiment included uninoculated and supplied nitrogen (+N) controls, the latter receiving 10 mL of 0.25 M KNO3 per pot, weekly. To further validate that nodulation was by the inoculant strain, two nodules per host species per treatment were surface-sterilised, crushed and streaked onto 12 LA plates to re-isolate rhizobia which were then identified by PCR. The interactions between rhizobial strains and Lotononis hosts were compared statistically with a split-plot analysis of variance (two-way ANOVA) using GenStat 8s (Release 8.1, Lawes Agricultural Trust, Rothamsted Experimental Station). 2.3. Light microscopy Light microscopy was used to examine the inner structure of nodules by visualising sections embedded in Spurr’s resin (Spurr, 1969). Nodules were fixed overnight at 4 1C in 3% (v/v) gluteraldehyde in 25 mM phosphate buffer (pH 7.0). Fixed material was washed using three changes of phosphate buffer. The samples were dehydrated in a rotator using a series of acetone solutions (30%, 50%, 70%, 90% and 100%) at 4 1C, with two changes of each solution, each of 15 min duration. Dehydrated samples were infiltrated with Spurr’s resin mixed with acetone using an increasing succession of concentrations (5%, 10%, 15%, 20%, 30%, 40%, 50%, 70% and 90%). The material was left in each solution for a minimum of 2 h. Infiltrated material was transferred into 100% Spurr’s resin, left at room temperature for 1–2 h and then transferred into fresh 100% Spurr’s resin for 5–8 h at room temperature or left overnight at 4 1C. Finally, in order to obtain good polymerisation, the material was embedded in fresh Spurr’s resin at 60 1C for 24 h. For light microscopy, 1 mm sections were cut using a Reichert-Jung 2050 microtome with a glass knife. Sections were dried onto glass slides at 60 1C and stained with 1% (w/v) methylene blue and 1% (w/v) azur II in 1% (w/v) sodium tetraborate (Richardson et al., 1960) for 3–5 min at room temperature. Stained sections were rinsed in water then dried. The specimens were examined under an Olympus BX51 compound microscope and photographed with an Olympus DP70 digital camera. 2.4. Molecular fingerprinting Strains (Table 1) were fingerprinted by PCR using the ERIC primers described by Versalovic et al. (1991). All PCR amplification profiles were generated from cultures grown to stationary phase in 12 LA broth, then washed and re-suspended in sterile saline to an OD600 nm of 10.0. The PCR reaction conditions used have been previously described (Yates et al., 2005). ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 2.5. Phylogenetic analysis of the 16S rRNA genes An intragenic 1.4 kb fragment of the 16S rRNA gene was amplified and sequenced from WSM3686, WSM3674, WSM2598, CB376, WSM2693, WSM2799 and WSM3032 as described by Yanagi and Yamasato (1993). Bacterial cells were incorporated directly into the PCR master mix to provide template DNA. Cells from a culture freshly grown on 12 LA were concentrated in 0.89% (w/v) NaCl to an OD600 nm of 6.0. Each reaction mixture contained 1 mL of concentrated cells, 2.5 U of Tth Plus DNA polymerase (Biotech international Ltd.), 1.25 mM of each of the two primers, 20F and 1540R, 1.5 mM MgCl2, 1x PCR Polymerisation buffer (PB-1, Biotech international Ltd.) in a final volume of 100 mL. The amplified gene products were purified using QIAquickTM PCR purification kit (QIAGEN). Each purified product was incorporated into a separate Sanger dideoxy reaction containing the appropriate primer (20F, 420F, 800F, 1100F, 1540R, 1190R, 820R or 520R; Yanagi and Yamasato, 1993) using ddNTPs labelled with Big Dye 3.1 chemistry. The reactions were loaded into an Applied Biosystems model 377A automated sequencer to obtain sequence reads that were compiled and analysed using Genetool lite (version 1.0; Double Twist Inc., Oakland, CA, USA) to produce double stranded sequence for the 16S rRNA gene of each organism. The polished sequences were used to extract matching DNA sequences from the National Centre for Biotechnology Information databases using BLASTN (Altschul et al., 1990). The sequences were aligned using the CLUSTALW program using the Wisconsin package of the Genetics Computer Group (Madison, WI, USA). A phylogenetic tree was constructed through the neighbour-joining method (Saitou and Nei, 1987) using the Kimura 2 parameter distances (Kimura, 1980). Strain details and sequence accession numbers are given in Fig. 4. 3. Results 3.1. Colony morphology and growth The isolates from L. bainesii (CB376, WSM2598), L. listii (WSM2693, WSM2799) and L. solitudinis (WSM3032), all 1683 formed medium- to slow-growing, dry, small (1–2 mm) pinkpigmented colonies after 4–5 days incubation at 28 1C. However, their degree of pigmentation varied: WSM2598, CB376 and WSM2693 colonies were pink while WSM2799 and WSM3032 colonies were a more intense pink. In contrast, the strains isolated from L. angolensis (WSM3686 and WSM3674) formed mucoid, translucent colonies that were 3–4 mm in diameter after 1–2 days and that became light pink after 3 days. In comparison, the type strain for M. nodulans (ORS2060) formed medium to slow-growing, dry, small (1–2 mm) colonies that remained white after 4–5 days. 3.2. Glasshouse experiment 1—nodulation studies The nodulation characteristics of the nine strains are recorded in Table 3. Uninoculated controls remained nodule free. Type strains ORS2060 (M. nodulans) and USDA6 (Bradyrhizobium japonicum) were unable to nodulate any of the three species of Lotononis but did form white and pink nodules on M. atropurpureum, respectively. Strain CB376, which is the inoculant for L. bainesii in Australia, nodulated both these species and L. listii, but could not consistently nodulate L. angolensis (there were some plants not nodulated in some replicates). In the reciprocal test, WSM3686 and WSM3674 originally isolated from L. angolensis nodulated this host but neither L. bainesii nor L. listii. Strains WSM2598 and WSM2693 from L. bainesii, WSM2799 from L. listii and WSM3032 from L. solitudinis displayed the same nodulation pattern as CB376. M. atropurpureum was ineffectively nodulated by all seven isolates from Lotononis, although the two L. angolensis isolates were inconsistent in this respect (Table 3). 3.3. Glasshouse experiment 2—nitrogen fixation studies The two strains originating from L. angolensis (WSM3686 and WSM3674) produced effective nodules on this host, as evident from the significant differences in shoot dry weight between the uninoculated N-free controls and the inoculated plants (Fig. 1). As for Glasshouse experiment 1, neither strain nodulated L. bainesii nor L. listii and thus the top dry weights for these treatments Table 3 Nodulation compatibilities of rhizobial strains with Lotononis spp. and Macroptilium atropurpureum Strain Original host L. angolensis (SA8363) L. bainesii (cv. Miles) L. listii (2004CRSL69) M. atropurpureum (cv. Siratro) WSM3686 WSM3674 ORS2060 USDA6 WSM2598 WSM2693 WSM2799 WSM3032 CB376 L. angolensis L. angolensis C. podocarpa G. max L. bainesii L. listii L. listii L. solitudinis L. bainesii + +   7 7 7 7 7     + + + + +     + + + + + 7 7 + + + + + + + (P) (P) (W) (W) (W) (W) (W) (P) (P) (P) (P) (P) (P) (P) (P) (P) (P) +: All plants nodulated; 7: inconsistent nodulation; : no nodulation; P: pink inside nodules; W: white inside nodules. (W) (W) (W) (P) (W) (W) (W) (W) (W) ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 1684 0.24 Lotononis angolensis (SA8369) Lotononis bainesii (AusTRCF47575) Shoot dry weight (g/plant) 0.2 Lotononis bainesii (cv. Miles) Lotononis listii(2004CRSL69) 0.16 0.12 0.08 0.04 0 CB376 WSM2598 WSM2693 WSM2799 WSM3032 WSM3674 WSM3686 N+ N- Fig. 1. Total shoot dry weight (g/plant) produced by Lotononis angolensis (SA8369), Lotononis bainesii (cv. Miles and AusTRCF47575) and Lotononis listii (2004CRSL69) when inoculated separately with seven strains of root-nodule bacteria (Table 1). N, uninoculated nitrogen-free control; N+, nitrogen-fed control; ($), ineffective nodulation; (K), no nodulation. Absence of symbols indicates effective nodulation. LSD (po0.05 ¼ 0.07). were equivalent to the nitrogen-free controls, and were significantly less than the N-supplied controls (po0.05). All other strains in the experiment effectively nodulated and fixed nitrogen when associated with both L. bainesii and L. listii. There were, however, approximately two-fold differences in the shoot dry weight produced by these strains. WSM2598 and WSM2799 produced significantly greater shoot dry weights than CB376 when inoculated onto the AusTRCF47575 genotype of L. bainesii (po0.05). Of particular note was that WSM3032, an isolate originally from L. solitudinis, effectively nodulated both L. bainesii and L. listii. The set of strains, which nodulated L. bainesii, L. solitudinis and L. listii, produced many small, white and ineffective nodules on L. angolensis. (Fig. 2c), infected cells of the young nodule tissue were densely packed with symbiosomes. Although these symbiosomes appeared spherical in the sections, further examination with wet mounts revealed that the symbiosomes were normally oblong (data not shown). In more mature plants (18 months), the nodules gave the appearance of completely enveloping, or girdling, the root (Fig. 2d), to the point where there appeared to be equivalent amounts of nodular tissue as root tissue. A section from a mature, small root (Fig. 2e, f) indicated that the envelopment of the root by the nodule was a consequence of several nodule infection points fusing to give rise to coalesced nodules (Fig. 2g). The infected regions of these multiple infections were clearly revealed under light microscopy (Fig. 2h). 3.4. Description of nodules and nodule sections 3.5. PCR fingerprint analysis Root nodules formed by L. bainesii, L. angolensis and L. listii were all characteristically lupinoid or collaroid (Sprent, 2001). After emergence at multiple infection points, the young nodules developed from just below the upper hypocotyl to throughout the main tap root, in numerous quantities (Fig. 2a). The nodules were pink and rounded with a rough surface and seemed to subtend the top of the root in the manner of Lupinus and Arachis (Sprent, 2001), rather than to develop a discrete point of attachment. Smaller nodules were also evident on the major lateral roots. The nodules were morphologically similar for all three species. Transverse sections of these young nodules (Fig. 2b) revealed little structural divergence among the three species, with meristematic activity surrounding the infected zone in the outer cortex and developing laterally (arrowed). Under higher magnification The two strains isolated from L. angolensis (WSM3686 and WSM3674) shared markedly similar PCR banding patterns with ERIC primers (Fig. 3; lanes 2 and 3), whereas all other strains isolated from Lotononis sp. (CB376, WSM2598, WSM2693, WSM2799 and WSM3032), as well as ORS2060, were clearly distinguishable with the ERIC primer. 3.6. Phylogenetic analysis of the 16S rRNA genes The neighbour-joining tree (Fig. 4) shows the phylogenetic position of the strains isolated from L. bainesii, L. listii and L. solitudinis in relation to the genus Methylobacterium. The tree construction is based on approximately 1400 bp of the 16S rRNA gene and analysis of the tree reveals that the strains isolated from L. bainesii, ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 1685 Fig. 2. (a) Lupinoid, determinate, N2 fixing nodules displaying multiple infection points on Lotononis angolensis when inoculated with strain WSM3686. (b) Transverse section of a young nodule, with the infected zone in the outer cortex developing laterally (arrowed). Bar 200 mm. (c) Densely packed symbiosomes in the infected cells. Bar 100 mm. (d) A mature Lotononis bainesii plant inoculated with strain WSM2598. NB nodules completely enveloping the roots. (e, f, g) Fresh then stained nodule section from a mature, small root of L. bainesii showing coalesced, collar nodulation. (h) Transverse section of mature nodule displaying the infected zone. were strongly related to each other (499% identity). However, when a BLAST similarity search was performed, these strains showed o94% sequence similarity to the 16S rRNA gene sequence of all other published type strains available in GenBank (Table 4). Since these strains appeared to be distinct from the genus Methylobacterium they were not included in Fig. 4. The two strains WSM2693 and WSM2799 isolated from L. listii, and WSM2598 from L. bainesii had 99% 16S rRNA gene sequence similarity. Similarly, WSM3032 and CB376, which were isolated from L. bainesii shared 499% sequence similarity for the 16S rRNA gene. Interestingly, the above strains plus the strain from L. solitudinis (WSM3032) had o97.2% sequence similarity to XCT17 (Fig. 4, Table 4), which is reported to have been isolated from L. bainesii. Lane 1 2 3 4 5 6 7 8 9 10 Fig. 3. PCR products generated using ERIC primers with cells of WSM3686 (lane 2), WSM3674 (lane 3), CB376 (lane 4), WSM2598 (lane 5), WSM2693 (lane 6), WSM2799 (lane 7), WSM3032 (lane 8) and ORS2060 (lane 9). Promega 1 kb marker was used as a size standard (lanes 1 and 10) containing fragments of sizes 10, 8, 6, 5, 4, 3, 2.5, 2, 1.5, 1.0, 0.75, 0.5 and 0.25 kb. L. solitudinis and L. listii cluster in a group, which is distinct from the M. nodulans type strain, albeit closest to this strain (with 497% sequence similarity). The strains isolated from L. angolensis (WSM3686 and WSM3674) 4. Discussion Our study revealed that L. listii and L. solitudinis share micro-symbionts similar to the dry, slow-growing, pinkpigmented rhizobia previously described for L. bainesii (Norris, 1958; Jaftha et al., 2002). In contrast, L. angolensis nodulated with a clearly different organism: one that is mucoid, distinctly faster-growing and less-intensely pigmented. Whereas the L. angolensis isolates were effective on their own host plant, no nodulation occurred on L. bainesii and L. listii (Table 3, Fig. 1). Reciprocal tests ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 1686 42 M. aminovorans TH15 (AB175629)T 46 M. suomiense F20 (AY009404)T 23 M. thiocyanatum ATCC700647 (U58018)T 11 M. portugalicum RXM (AY009403)T M. chloromethanicum CM4 (AF198624)T 56 M. extorquens ATCC43645 (L20847)T 94 35 30 93 M. rhodesianum ATCC43882 (L20850)T M. zatmanii ATCC43883 (L20804)T M. dicloromethanicum DM4 (AF227128)T 99 M. rhodinum ATCC14821 (L20849)T M. organophilum JCM2833 (D32226)T M. mesophilicum ATCC29983 (D32225)T 84 100 99 100 M. fujisawaense ATCC43884 (AJ250801)T M. radiotolerans ATCC27329 (D32227)T M. aquaticum GR16 (AJ635303)T 84 M. variabile GR3 (AJ851087)T M. isbiliense AR24 (AJ888239)T M. nodulans ORS2060 (AF220763)T CB376 isolated from L. bainesii (AF467688) 52 WSM2693 isolated from L. listii (DQ838522) 52 99 WSM3032 isolated from L. solitudinis (DQ848137) 47 WSM2799 isolated from L. bainesii (DQ848136) WSM2598 isolated from L. bainesii (DQ838527) 67 92 XCT17 isolated from L. bainesii (AF467694) 0.005 Fig. 4. Phylogenetic tree showing relationships of nodule isolates from Lotononis spp. in relation to the Methylobacterium based upon aligned sequences of 16S rRNA sequences. Phylogenetic analyses were conducted using MEGA version 3.1 (Kumar et al., 2004). Kimura 2-parameter distances were derived from the aligned sequences (Kimura, 1980) and bootstrap (Felsenstein, 1985) was undertaken with 500 replicates to construct a consensus unrooted tree using the neighbour-joining method (Saitou and Nei, 1987). Database accession numbers are provided in parenthesis. showed that the ‘‘dry’’ isolates from L bainesii, L listii and L solitudinis were similarly incapable of nodulating L angolensis effectively, displaying inconsistent and ineffective nodulation. This would explain previous observations of poor species performance when L. bainesii inoculant was applied to L. angolensis (‘t Mannetje, 1967). Notably, within the Listia section, L. angolensis is the species most geographically separated, and this may explain why it acquires a different micro-symbiont to the others. Despite the apparent close botanical relatedness of the Lotononis species within the section Listia (Van Wyk, 1991), 16S rRNA gene sequencing suggests that the L. angolensis isolates are taxonomically distinct from those that nodulate L. bainesii, L. listii and L. solitudinis (which may be grouped), as well as from the M. nodulans strain (ORS2060) which was isolated from Crotalaria podocarpa (Fig. 4). Jaftha et al. (2002) considered the L. bainesii strains to be closely related to M. nodulans based upon analysis of a partial sequence of each 16S rRNA gene. M. nodulans is the only defined species in that genus currently known to form nodules on legumes. However, it is clear from our data that the type strain for M. nodulans differs considerably in its host range and 16S rRNA phylogeny from the suite of strains that nodulate L. bainesii, L. listii and L. solitudinis (Table 3, Fig. 4). The strains nodulating these species may therefore form a new species of root-nodule bacteria within the Methylobacteriaceae. A notable feature in constructing the phylogenetic tree was the fact that the strains nodulating L. angolensis did not cluster with other bacteria belonging to the family Methylobacteriaceae. Based on the polyphasic taxonomic approach, Gillis et al. (2005) has suggested that if the 16S rRNA gene sequence similarity between two strains exceeds 97%, then it can be assumed that these strains belong to the same genus. Our results have shown that the two strains isolated from L. angolensis have o94% sequence similarity to other genera within the alpha proteobacteriacea. It is likely, therefore, that the strains isolated from L. angolensis belong to a new genus of root nodulating organisms. However, further physiological, molecular and biochemical tests need to be completed before firm conclusions can be made in relation to all these strains. In particular, there has been recent attention drawn to a key role for methyloptrophy during symbiosis between ARTICLE IN PRESS R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 1687 Table 4 Sequence similarity of 16S rRNA for the root nodule isolates from Lotononis sp. and the type strains of Methylobacterium CB376 WSM2693 WSM3032 WSM2799 WSM2598 XCT17 M. aminovorans M. aquaticum M. chloromethanicum M. dicloromethanicum M. extorquens M. fujisawaense M. isbiliense M. mesophilicum M. nodulans M. organophilum M. portugalicum M. radiotolerans M. rhodesianum M. rhodinum M. suomiense M. thiocyanatum M. variabile M. zatmanii CB376 WSM2693 WSM3032 WSM2799 WSM2598 WSM3686 100 99.1 99.3 99.3 99.5 97.2 95.1 95.8 93.6 92.9 93.8 94.1 97.9 94.3 97.6 94.0 93.5 94.3 93.2 93.7 94.2 94.9 96.0 93.2 99.1 100 99.2 99.6 98.6 96.6 95.1 95.4 93.8 92.3 92.1 94.7 97.7 95.1 97.4 94.0 92.2 95.1 91.6 92.3 92.9 94.7 95.3 91.6 99.3 99.2 100 99.1 99.0 96.3 95.3 95.8 92.8 92.2 92.9 94.2 97.7 95.0 97.2 93.9 92.8 94.9 92.4 92.9 93.3 93.7 95.7 92.5 99.3 99.6 99.1 100 99.7 97.0 95.6 95.5 92.9 92.6 92.3 94.2 97.9 95.0 97.8 94.2 92.5 95.4 91.9 92.4 93.1 93.7 95.6 92.0 99.5 98.6 99.0 99.7 100 97.2 94.9 94.5 92.6 92.5 91.3 93.6 96.9 94.1 97.4 93.4 91.5 94.5 90.9 91.4 92.1 93.6 95.0 91.0 93.5 94.1 93.7 93.9 93.2 90.8 93.4 92.3 92.3 90.4 90.5 92.5 93.4 92.8 93.2 93.0 91.0 92.8 90.5 90.1 91.2 93.2 92.1 90.4 M. nodulans and C. podocarpa (Jourand et al., 2005). Our preliminary evidence suggests that few of the Lotononis strains we are examining are methylotrophic (unpublished data). Notwithstanding the marked differences between the two groups of Lotononis rhizobia, nodule morphology was very similar amongst the three species of Lotononis. This is consistent with the view that the host genetics primarily determine nodule morphology (Wood et al., 1985; Nandasena et al., 2004). In our observations over 16 months, the lupinoid-type nodules on the Lotononis spp. followed a gradual, possibly perennial, pattern of root encirclement. Closer analysis of the formation of new symbiotic tissue adjacent to the infected zone, and the rate of senescence of old tissue, is warranted to confirm this. This is the first report and description of nitrogen-fixing micro-symbionts for L. angolensis, L. solitudinis and L. listii, and is the prelude to further agronomic evaluation of these potentially useful perennial legumes. The observation that strains effective on L. bainesii differed in their capacity for nitrogen fixation on this host is also of significant agronomic value. Acknowledgements The authors would like to thank Regina Carr, Ertug Sezmis (Centre for Rhizobium Studies, Murdoch University) and Gordon Thompson (School of Biological Science, Murdoch University) for their skilled technical assistance. Authors are appreciative of Dr. Alison McInnes (University of Western Sydney) for assistance in sourcing the L. angolensis strains, Dr. Catherine Boivin-Masson (LIPM) for kindly providing strain ORS2060 and Prof. Ben-Erik van Wyk (University of Johannesburg) for assisting in the location of Lotononis species from the section Listia. References Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool. Journal of Molecular Biology 215, 403–410. Bryan, W.W., 1961. Lotononis bainesii Baker a legume for subtropical pastures. Australian Journal of Experimental Agriculture and Animal Husbandry 1, 4–10. Diatloff, A., 1977. Ecological studies of root-nodule bacteria introduced into field environments—6. Antigenic and symbiotic stability in Lotononis rhizobia over a 12-year period. Soil Biology & Biochemistry 9, 85–88. Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791. Gillis, M., Vandamme, P., De Vos, P., Swings, J., Kersters, K., 2005. Polyphasic taxonomy. In: Brenner, D.J., Krieg, N.R., Staley, J.T. (Eds.), Bergey’s Manual of Systematic Bacteriology, vol. 2. Springer, New York, USA, pp. 43–48. Godfrey, C.A., 1972. The carotenoid pigment and deoxyribonucleic acid base ratio of a Rhizobium which nodulates Lotononis bainesii Baker. Journal of General Microbiology 72, 399–402. Howieson, J.G., Ewing, M.A., D’Antuono, M.F., 1988. Selection for acid tolerance in Rhizobium meliloti. Plant and Soil 105, 179–188. Howieson, J.G., Loi, A., Carr, S.J., 1995. Biserrula pelecinus L.—a legume pasture species with potential for acid, duplex soils which is nodulated by a unique root-nodule bacteria. Australian Journal of Agricultural Research 46, 997–1009. Howieson, J.G., Yates, R.J., Foster, K., Real, D., Besier, B., 2007. Prospects for the future use of legumes. In: Dilworth, M.J., James, ARTICLE IN PRESS 1688 R.J. Yates et al. / Soil Biology & Biochemistry 39 (2007) 1680–1688 E.K., Sprent, J.I., Newton, W.E. (Eds.), Leguminous Nitrogen-Fixing Symbioses. Elsevier, London, in press. Jaftha, J.B., Strijdom, B.W., Steyn, P.L., 2002. Characterization of pigmented Methylobacterium bacteria which nodulate Lotononis bainesii. Systematic and Applied Microbiology 25, 440–449. Jourand, P., Renier, A., Rapior, S., Miana de Faria, S., Prin, Y., Galiana, A., Giraud, E., Dreyfus, B., 2005. Role of methylotrophy during symbiosis between Methylobacterium nodulans and Crotalaria podocarpa. Molecular Plant-Microbe Interactions 18, 1061–1068. Kimura, M., 1980. A simple model for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16, 111–120. Kumar, S., Tamura, K., Nei, M., 2004. MEGA3: integrated software for Molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5, 150–163. Law, I.J., 1979. Resistance of Rhizobium specific for Lotononis bainesii to ultraviolet irradiation. Soil Biology & Biochemistry 11, 87–88. Nandasena, K.G., O’Hara, G.W., Tiwari, R.P., Yates, R.J., Kishinevsky, D.B., Howieson, J.G., 2004. Symbiotic relationships and root nodule ultra-structure of the pasture legume Biserrula pelecinus L.—a new legume in agriculture. Soil Biology & Biochemistry 36, 1309–1318. Norris, D.O., 1958. A red strain of Rhizobium from Lotononis bainesii Baker. Australian Journal of Agricultural Research 9, 629–632. Norris, D.O., 1959. Legume bacteriology in the tropics. The Journal of the Australian Institute of Agricultural Science, 202–207. Perret, X., Staehelin, C., Broughton, W., 2000. Molecular basis of symbiotic promiscuity. Microbiology and Molecular Biology Reviews 64, 120–180. Real, D., Formoso, F., Martinez, A., Risso, I., Hugo, W., Rostan, C., Alzugaray, R., 2005. Pre-basic seed production of Lotononis bainesii Baker cv. ‘INIA Glencoe’. New Zealand Journal of Agricultural Research 48, 373–377. Richardson, K.C., Jarret, L., Finke, E.H., 1960. Embedding in epoxy resins for ultra-thin sectioning in electron microscopy. Strain Technology 35, 313. Samba, R.T., De Lajudie, P., Gillis, M., Neyra, M., Spencer-Baretto, M.M., Dreyfus, B., 1999. Diversity of rhizobia nodulating Crotalaria spp. from Senegal. Symbiosis 27, 259–268. Saitou, N., Nei, M., 1987. Reconstructing phylogenetic trees. Molecular Biology and Evolution 4, 406–425. Sprent, J.I., 2001. Nodulation in Legumes. Royal Botanic Gardens, Kew. Spurr, A.R., 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructure Reseach 26, 31–43. Sy, A., Giraud, E., Jourand, P., Garcia, N., Willems, A., De Lajudie, P., Prin, Y., Neyra, M., Gillis, M., Masson-Boivin, C., Dreyfus, B., 2001. Methylotrophic Methylobacterium bacteria nodulate and fix nitrogen in symbiosis with legumes. Journal of Bacteriology 183, 214–220. Thies, J.E., Bohlool, B.B., Singleton, P.W., 1991. Subgroups of the cowpea miscellany: symbiotic specificity within Bradyrhizobium spp. for Vigna unguiculata, Phaseolus lunatus, Arachis hypogaea, and Macroptilium atropurpureum. Applied and Environmental Microbiology 57, 1540–1545. ‘t Mannetje, L., 1967. Pasture improvement in the Eskdale district of south eastern Queensland. Tropical Grasslands 1, 9–19. Van Wyk, B-E., 1991. A Synopsis of the Genus Lotononis (Fabaceae: Crotalarieae). Rustica Press, Cape Town, South Africa. Versalovic, J., Koeuth, T., Lupski, J.R., 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Research 19, 6823–6831. Wood, S.M., Layzell, D.B., Newcomb, W., Pankhurst, C.E., 1985. A morphometric study of effective nodules induced by Rhizobium loti and Bradyrhizobium sp. (Lotus) on Lotus pedunculatus. Canadian Journal of Botany 63, 43–53. Yanagi, M., Yamasato, K., 1993. Phylogenetic analysis of the family Rhizobiaceae and related bacteria by sequencing of 16S rRNA gene using PCR and DNA sequencer. FEMS Microbiology Letters 107, 115–120. Yates, R.J., Nandasena, K.G., O’Hara, G.W., Howieson, J.G., 2004. Root-nodule bacteria from indigenous legumes in the north-west of Western Australia and their interaction with exotic legumes. Soil Biology & Biochemistry 36, 1319–1329. Yates, R.J., Howieson, J.G., Real, D., Reeve, W.G., Vivas-Marfisi, A., O’Hara, G.W., 2005. Evidence of selection for effective nodulation in the Trifolium spp. symbiosis with Rhizobium leguminosarum biovar trifolii. Australian Journal of Experimental Agriculture 45, 189–198. Yates, R.J., Real, D., Revell, C., Howieson, J.G., 2006. Development of perennial Lotononis sp. for Western Australian agriculture. In: Mercer, C. (Ed.), Proceedings of the 13th Australasian Plant Breeding Conference. NZ, pp. 1020–1028.