characterization of pigmented methylotrophic bacteria...

System. Appl. Microbiol. 25, 440–449 (2002)© Urban & Fischer Verlag http://www.urbanfischer.de/journals/sam

Characterization of Pigmented Methylotrophic Bacteriawhich Nodulate Lotononis bainesii

J. B. JAFTHA1, B. W. STRIJDOM2, and P. L. STEYN1

1 Department of Microbiology and Plant Pathology, University of Pretoria, South Africa2 ARC-Plant Protection Research Institute, Pretoria, South Africa

Received: June 17, 2002

Summary

Root nodule isolates from a shrubby legume, Lotononis bainesii, were characterized by 16S rRNA genesequencing and morphologically by substrate utilization patterns. The symbiotic genome of these isolateswas analysed by partial sequencing of the nifH gene. Based on the results of numerical taxonomy, theisolates formed a closely related cluster, showing no correspondence to any of the known rhizobial clus-ters. Analysis of nearly full-length 16S rDNA sequences demonstrated that these isolates were related toMethylobacterium nodulans (SY et al., 2001). In the absence of nifH sequence data for the genus Methy-lobacterium, the nifH phylogeny showed these isolates to be related to Azospirillum brasilense. The fac-ultative methylotrophic nature of these isolates was also demonstrated by their ability to grow in thepresence of methanol as a sole carbon source.

Key words: Lotononis bainesii – 16S rDNA sequencing – nifH gene – Methylobacterium

Introduction

Symbiotic nitrogen fixing bacteria, commonly referredto as rhizobia, are able to establish a symbiotic associationwith most leguminous plants. As a result of this symbioticassociation, specialised organs, called nodules, are inducedon the roots and stems of the host plant. Within such nod-ules atmospheric nitrogen is reduced to ammonia to thebenefit of the host plant. These nitrogen-fixing nodulatingrhizobia have been assigned to different genera within theα-subclass of the Proteobacteria and include: Rhizobium,Sinorhizobium, Mesorhizobium, Bradyrhizobium, Azorhi-zobium (for reviews see MARTINEZ-ROMERO, 1996; YOUNG

and HAUKKA, 1996; VAN BERKUM and EARDLY, 1998) andAllorhizobium (DE LAJUDIE et al., 1998). These rhizobialgenera are very diverse with some being phylogeneticallycloser related to other non-symbiotic genera, than to eachother (YOUNG, 1996; VAN BERKUM and EARDLY, 1998;YOUNG, 2001). Recently, SY et al. (2001) reported the exis-tence of an additional rhizobial branch involving bacteriaof the genus Methylobacterium. These rhizobia were iso-lated from Crotalaria species and were able to grow facul-tatively on methanol, a common trait for Methylobacteri-um species, but unique to the known rhizobial species.Analysis of the 16S rDNA gene, nodulation ability, as well

as amplification of the nodA gene confirmed these isolatesto be nodulating Methylobacterium species for which thename Methylobacterium nodulans was proposed. Methy-lobacterium species are usually isolated from water andleaf surfaces and are known as pink-pigmented facultativemethylotrophs (HOLLAND, 1997). However, M. nodulansdid not show any pigmentation (SY et al., 2001).

Lotononis species are herbs and shrubs of the subfami-ly Papilionoideae with more than 140 species commonlyoccurring under diverse climatological and geographicalconditions. Their distribution is mainly in southernAfrica but may extend to the Mediterranean, with a fewspecies in southern Europe and central Asia (VAN WYK,1991). Lotononis bainesii has proven its value as a pas-ture legume in regions in Australia. In addition to L.bainesii, other Lotononis species such as L. divaricata, L.tenella and L. laxa also have potential value as grazingplants since many are well adapted to the arid regions(SHEARING, 1994). In 1958 NORRIS described a pigmentednodulating strain obtained from the roots of Lotononisbainesii. The chemical structure of this pigment was sub-sequently determined by KLEINIG and BROUGHTON (1982)and found to be similar to that of Pseudomonas species.

0723-2020/02/25/03-440 $ 15.00/0

Methylobacterium nodulating Lotononis bainesii 441

DAGUTAT (unpublished results) compared the protein pro-files of a collection of bacterial isolates obtained from theroot nodules of Lotononis bainesii. These isolates formeda closely related cluster, clearly separated from the rhizo-bial reference strains. Initial partial sequence 16S rDNAsequencing performed as part of this study revealed thatthese isolates were indeed a group of unknown taxonom-ic status. However, with the report of the methylotrophicnodulating bacteria (SY et al., 2001) it became evidentthat these pigmented Lotononis isolates were also relatedto the genus Methylobacterium, showing high homologyto Methylobacterium nodulans.

The primary objective of this study was therefore tocharacterize nine isolates from the root nodules of Lotono-nis bainesii obtained from different localities in southernAfrica. Using nearly full length 16S rDNA, partial nifH se-quencing and substrate utilization patterns, it was possibleto show that these pigmented nodulating strains were fac-ultative methylotrophs, related to the genus Methylobac-terium, more specifically to M. nodulans.

Materials and Methods

Bacterial strains and growth conditionsStrains analysed (Table 1) in this study were obtained from

the root nodules of Lotononis bainesii and formed part of therhizobial collection of the ARC-Plant Protection Research Insti-tute (Private Bag X134, Pretoria 0001, South Africa). Referencecultures of the different rhizobial genera were obtained from thebacterial culture collection of the Laboratorium voor Microbi-ologie (LMG), State University Gent, Belgium and the UnitedStates Department of Agriculture (USDA), Soybean and AlfalfaResearch Laboratory, Maryland, USA. The bacteria were grownat 28 °C in yeast mannitol (YM) medium (VINCENT, 1970).Their ability to use methanol was also assessed by growth onmedium 72 (BELGIAN CO-ORDINATED COLLECTION OFMICROORGANISMS, 1998), which contained 10 ml methanolper liter medium.

Numerical taxonomySubstrate utilization patterns of the isolates were assessed

using the Gram-negative Biolog MicroPlate™ (Biolog, Hayward,California). These plates contained a panel of 95 different car-bon sources. Growth conditions and inoculation of the mi-croplates were performed as prescribed by the suppliers. The uti-lization of a specific carbon source was indicated by the develop-ment of a purple colour in the wells which was due to the pres-ence of a redox dye, tetrazolium violet. Such wells were scored asone (1), while negative wells were scored as zero (0). The Dicecoefficient (NEI and LI, 1979) within the Bionum computer pro-gramme (Applied Maths, Kortrijk, Belgium) was used to con-struct a distance matrix from this one (1)-zero (0) profiles. Thesedistance values were subsequently analysed to generate a treeusing the unweighted pair group method of arithmetic mean al-gorithm (UPGMA) in GelCompar 4.0 (Applied Maths, Kortrijk,Belgium). Substrate utilization patterns of the rhizobial referencestrains were obtained from a previous study (KRUGER, 1998).

Genomic DNA extractionBacterial cultures were grown in YM broth for 4–7 days and

genomic DNA extracted by the hexadecyltrimethyl ammoniumbromide (CTAB) method as described by WILSON (1994). The in-tegrity and concentration of the purified DNA samples was de-termined by agarose gel electrophoresis (SAMBROOK et al., 1989).

OligonucleotidesAmplification of the 16S rRNA gene was carried out using

the primers fD1 and rP2 (WEISBURG et al., 1991). To amplify a750 bp fragment of the nifH gene, two primers: nifH-F(5’CGGGAAGGGCGGAATCGGCAAG3’) and nifH-R(5’GCATGTCCTCGAGCTCWTCCAT3’) were designed bycomparing nifH sequence data of rhizobial reference strainswith the following GenBank accession numbers: K10620 (B.japonicum), J01781 (S. meliloti), Z95228 (Mesorhizobium sp.),M15942 (R. etli bv phaseoli), L16503 (Sinorhizobium sp.) andM55226 (R. leguminosarum bv. phaseoli). All primers were syn-thesized by Roche Molecular Biochemicals.

PCR amplification and sequencing of the 16S rRNA genesand partial nifH gene

Reaction and cycle conditions for the amplification of the16S rDNA gene were similar to those described by LAGUERRE etal. (1994). For sequencing two internal primers: 16SRNAII-Sand 16S RNAVII-S (KUHNERT et al., 1996) was used to obtainnearly full-length 16S rDNA sequence data. PCR reagents werepurchased from Southern Cross Biotechnologies. The amplifica-tion of the nifH gene was achieved using similar reaction condi-tions but the following cycle profile was used: an initial denatu-ration step at 95 °C for 3 min, 30 cycles of denaturation (94 °Cfor 1 min), annealing (37 °C for 30 sec) and extension (72 °C for1 min). This lower annealing temperature yielded non-specificproducts of amplification with all of the isolates. Therefore theband of the expected size, obtained from isolate xct7, was gel-purified using GlassMilk (Bio101) according to the manufactur-ers instructions and cloned using the Qiagen PCR cloning kit(Southern Cross Biotechnologies) prior to sequencing.

Sequencing reactions were performed using the ABI PrismBigDyeTM Terminator Cycle Sequencing Ready Reaction kit(Perkin Elmer Applied Biosystems) according to the manufactur-er’s instructions. Electrophoresis of the products of such se-quencing reactions were carried out on an Applied BiosystemsModel 377 automated sequencer. The quality of the sequencewas verified by sequencing both DNA strands. The nifH se-quence data was obtained by sequencing recombinant plasmidsusing the M13 forward and reverse primers (Promega). Theidentity of the nifH sequence was established by performing asimilarity search against the GenBank database (website:http://www.ncbi.nlm.nih.gov/BLAST).

Phylogenetic analysesThe ClustalX programme (THOMPSON et al., 1997) was used

to analyse the nucleic acid sequences. Additional sequence dataof related α-Proteobacteria was obtained from GenBank and ac-cession numbers are indicated in the relevant figures. The phylo-genetic trees were constructed from the distance matrices usingthe neighbour-joining method of SAITOU and NEI (1987). Thebootstrap method (FELSENSTEIN, 1985) was used in combinationwith the neighbour-joining method to estimate confidence levelsof the phylogenies. The phylogenetic trees were displayed usingNJplot (PERRIÈRE and GOUY, 1996).

The nifH sequence of isolate xct7 was also compared withnifH nucleotide sequences of other α- and γ-Proteobacteria. Adistance matrix expressing these genetic distances was generatedusing DNAdist of Phylip version 3.5c (FELSENSTEIN, 1989).

Pigment extraction and bacteriochlorophyll detectionIsolates were grown in YM broth for 7 days and pigments

extracted with cold acetone-methanol (7:2, v/v) as described byEVANS et al. (1990). A portion of the extracts was also acidifiedusing one drop of 4 N HCl. The absorption spectra of both acid-treated and untreated extracts were recorded between 400 and800 nm.

442 J. B. JAFTHA et al.

Results

Numerical taxonomy

Substrate utilization patterns of 95 different carbonsources were used to establish a numerical taxonomy forbacteria isolated from the root nodules of Lotononisbainesii. The results of the range of substrates utilised bythis group of symbionts are recorded in Table 2. Carbonsources not utilised by any of the isolates are not indicat-ed. These carbon sources were previously divided into 11categories by GARLAND and MILLS (1991) as indicated inTable 2. The highest number of carbon sources tested forincluded carboxylic acids, amino acids and carbohy-drates. Amongst the Lotononis isolates the full range ofcarboxylic acids assayed for was utilised, while only 12out 20 amino acids were used by the isolates. The isolateswere more specific with regard to their carbohydrates assole carbon sources since only 5 out of possible 28sources were used.

The metabolic fingerprints generated using the Biologsystem was subsequently used to generate a dendrogram(Fig. 1) expressing the phenotypic similarities among theLotononis isolates and rhizobial reference strains. Twomajor sections could be distinguished within this dendro-gram. The first contained members of the genera Rhizobi-um, Sinorhizobium, Mesorhizobium, Agrobacterium andAllorhizobium, while section 2 contained the members ofthe genus Bradyrhizobium and the Lotononis isolates. Theoverall similarity for groups within the sections is 63%and 67% for section 1 and 2 respectively. The Lotononisisolates were present in a cluster which showed no signifi-cant homology to any of the rhizobial reference strains. Anoverall similarity value of 90% was observed among iso-

Table 1. List of isolates analysed in this study.

Isolate/ Strain Host plant

xct7 Lotononis bainesiixct8 Lotononis bainesiixct9 Lotononis bainesiixct10 Lotononis bainesiixct12 Lotononis bainesiixct13 Lotononis bainesiixct14 Lotononis bainesiixct16 Lotononis bainesiixct17 Lotononis bainesiiRhizobium leguminosarum (LMG 6294) Lathyrus sp.Rhizobium. leguminosarum bv. trifolii(LMG 6119) Trifolium repensRhizobium galegae (USDA 4128T) Galega orientalisRhizobium tropici (USDA 9030T) Phaseolus vulgarisRhizobium etli (USDA 9032) Phaseolus vulgaris Bradyrhizobium japonicum (LMG 6138T) Glycine maxBradyrhizobium elkanii (LMG 6134T) Glycine maxBradyrhizobium sp. (LMG 8319) Macrotyloma

africanusAllorhizobium undicola (USDA 4903) Neptunia natansAllorhizobium undicola (USDA 4904) Neptunia natansSinorhizobium meliloti (LMG 6133T) Medicago sativaSinorhizobium fredii (LMG 6217T) Glycine maxSinorhizobium saheli (LMG 7837T) Sesbania

cannabinaMesorhizobium loti (LMG 6125T) Lotus corniculatusMesorhizobium huakuii (USDA 4778T) Astragalus sinicusAgrobacterium radiobacter (LMG 140T) NSAgrobacterium tumefaciens (LMG 187T) Lycopersicon

lycopersicumAgrobacterium rhizogenes (LMG 150T) NSAgrobacterium aggregatum (LMG 122T) NS

Fig. 1. Dendrogram showing thephenotypic similarities amongLotononis bainesii isolates as de-termined by substrate utilizationpatterns using the Biolog system.The UPGMA method was usedfor cluster analysis (SNEATH &SOKAL, 1973). The x-axis showsthe correlation between strains.S: Sinorhizobium; R: Rhizobium;Ag: Agrobacterium, A: Allorhizo-bium; M: Mesorhizobium, B:Bradyrhizobium. LMG strain6119: Rhizobium leguminosarumbv. trifolii.


Table 2. Oxidation patterns of the different carbon sources utilised by Lotononis bainesii isolates.

Carbon Sources Isolates––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––xct7 xct8 xct9 xct10 xct12 xct13 xct14 xct16 xct17

PolymersTween-40 + + + – – + + + –Tween-80 + + + – + + + + –CarbohydratesL-arabinose – + + + + + + + –L-fucose – – – – – + + – –A-D-glucose – – + – – + – – –sucrose + – – – – – – – –xylitol – – + – – – – – –Estersmethylpyruvate + + + + + + + + +mono-methylsuccinate + + + + + + + + +Carboxylic acidsacetic acid + + + + + + + + +cis-aconitic acid + + – + + + + + –citric acid – + + + + + + + –formic acid + – – – – + + + +D-galactonic acid lactone – + + + – + + + +D-galacturonic acid – – – – – + + + –D-gluconic acid + + + + + + + + +D–glucosaminic acid – – – + – – – + –D-glucoronic acid – + – – – – – – –A-/β-/γ-hydroxybutyric acid + + + + + + + + +itaconic acid – – – – – – – + –A-keto-butyric acid + + + + + + + + +A-keto-glutaric acid + + + + + + + + +A-keto-valeric acid + + + + + + + + +D, L lactic acid; malonic acid + + + + + + + + +propionic acid, quinic acid + + + + + + + + +D-saccharic acid, sebacic acid + + + + + + + + +succinic acid + + + + + + + + +Bromonated chemicalsbromo-succinic acid + + + + + + + + +Amidessuccinamic acid + + + + + + + + +alaninamide + + + + + + + + +Amino acidsL-Asparagine + + + + + + + + +D-alanine – – – – – + + – –L-alanyl-glycine – – + – – + – – –L-aspartic acid + + + + + + + + +L-glutamic acid + + + + + + + + +glycyl-L-glutamic acid – – + + + + + – –L-leucine – – – – – – – – +L-proline + – + + + + + + –L-pyroglutamic acid + – + – – + + + +D-serine; L-serine – – + – – – – – –L-threonine – – + – + + + – –Aromatic chemicalsurocanic acid – – – – – + + – –Amines2-amino-ethanol – – + – – – – – –Alcohols2,3-butanediol – – + – – – – – –

Carbon sources not utilised by any of the isolates are not listed. Those showing reaction indicated by positive (+) sign and no reac-tion by negative (–). Categories of substrates as determined by GARLAND & MILLS, (1991).


Fig. 2. Phylogenetic relationships of isolates nodulating Lotononis bainesii in comparison to rhizobial genera and other alpha-pro-teobacteria. This analysis was based on comparative sequence analysis of approximately 1200 bp fragment of the 16S rRNA gene.The tree was constructed using the Neighbour-joining method of SAITOU & NEI (1987). Horizontal branch lengths reflect the phyloge-netic distances, while vertical branch lengths are non-informative and set for clarity only. The scale bar indicates 2% nucleotide differ-ence and bootstrap values of some of the major branching points are shown. GenBank accession numbers are indicated in brackets.M: Mesorhizobium, S: Sinorhizobium, O: Ochrobactrum, R: Rhizobium, A: Agrobacterium, X: Xanthobacter, B: Bradyrhizobium.


lates xct14, xct13, xct17, xct8, xct10, xct12, xct7 andxct16. Isolate xct9 was separated from this group since itcould uniquely use the following substrates: xylitol, D-ser-ine, L-serine, 2-amino-ethanol and 2, 3-butanediol.

16S rDNA sequence analysis

Nearly full length (1180 bp) nucleotide sequence of the16S rRNA gene was determined for the nine isolates ob-tained from the root nodules of Lotononis bainesii. Thephylogenetic position of these isolates was inferred fromcomparative 16S sequence analysis and is indicated in

Fig. 2. Representatives of the other genera of the α-Pro-teobacteria were also included in this analysis. TheLotononis bainesii isolates were clearly distinct from theother known rhizobial genera and showed high sequencehomology to the Methylobacterium lineage of the α-Pro-teobacteria. The Lotononis isolates had almost identical16S rDNA sequences; with only xct17 showing approxi-mately 2.6% sequence difference from the other Lotononisisolates. Sequence similarities of the Lotononis group andother described Methylobacterium species was on average94%. The closest phylogenetic neighbor of the Lotononisisolates was Methylobacterium nodulans (SY et al., 2001),showing sequence similarity values of close to 98%.

Fig. 3. Phylogenetic tree, based on a560 bp fragment of the nifH gene, ex-pressing the relationship of Lotononisbainesii nodulating strain, xct7, toother rhizobial genera and other nitro-gen fixing strains of the alpha- andgamma-proteobacteria. The tree wasgenerated using the Neighbor-joiningmethod of SAITOU & NEI (1987). Hori-zontal branches are equal to the phylo-genetic distances of which the scale in-dicates 2%. Bootstrap values of someof the major branching points are indi-cated. GenBank accession numbers areindicated in brackets. The nifH se-quences of reference strains (marked*) were determined in a related study.Rhizobial isolates (indicated by #) ob-tained from indigenous legumes inSouth Africa were included as addi-tional references


Analysis of the nifH-gene sequences

A fragment of the nifH gene of isolate xct7 was clonedand sequenced. The similarity search verified that thecloned PCR product was a fragment of the nifH genewhich showed the highest homology with Azospirillumbrasilense nifH sequences. Most of the available nifH se-quence data, obtained from GenBank, were considerablyshorter fragments of sequence and therefore only a 560 bpfragment of the gene was used for comparative analysisand construction of a phylogenetic tree (Fig. 3). 16S rDNAsequence analyses have already indicated the Lotononisisolates to be related the genus Methylobacterium, howev-er, no nifH sequence data was available for members ofthis genus. In the absence of sufficient sequence data, thephylogenetic position of xct7, was assessed in the presenceof other genera of the α- and γ-Proteobacteria. The nifHsequences of rhizobial isolates (indicated by # in Fig. 3),obtained from the root nodules of various indigenousSouth African leguminous hosts species, were also includ-ed. These indigenous isolates maintained their generic af-filiations as determined by 16S rDNA sequence analyses(unpublished results). No close relationship was evidentbetween xct7 and any member of the known rhizobial gen-

era included in this analysis. Azospirillum brasilense andxct7 shared almost a 90% nifH sequence similarity.

The nifH sequence of xct7 was compared with a fewrepresentatives of the α- Proteobacteria which includedthe known rhizobial genera, Rhodobacter capsulatus andtype II methanotrophs: Methylocystis and Methylosinusspecies (AUMAN, et al. 2001). The following γ-Proteobac-teria were included: Pseudomonas stutzeri, Marichroma-tum purpuratum, Vibrio diazotrophicus and Klebsiellapneumoniae. Due to length variation among the differentsequences compared here, all sequences were shortenedand therefore only a 244 bp region of the nifH gene wasused in this comparison. These similarity values are indi-cated in Table 3. Similarity values within the rhizobialgenera ranged from 59% to 96%, while the type IImethanotrophs share at least 90% sequence similarity.Sequence similarity values between xct7 and the rhizobialgenera ranged from 72% to 83%. When comparing thexct7 nifH sequence with Methylosinus thrichosporium(type II methanotroph) and Marichromatum purpuratum(γ-Proteobacteria), similarity values of 84% and 81%were obtained respectively. In contrast, xct7 had a lowersimilarity value (72%) with B. japonicum and B. liaonin-gense.

Fig. 4. Absorption spectra of anacetone-methanol extract fromthe root nodule isolates ofLotononis bainesii. (a) Extract ofstrain xct14 showing the classi-cal three-peaked carotenoid spec-trum identical to that of Rhizobi-um BTAi1. The bacteriochloro-phyll a (Bchla) peak at around770 nm before (b) and after (c)the addition of a single drop of 4N HCl, which resulted in the re-placement of the Bchla peak withthe peak of bacteriopheophytin a(Bphea) at around 745 nm.


Pigment isolation

The results obtained here corresponded well with theabsorption spectra of the carotenoids and bacteriochloro-phyll a (Bchl a) of the pigmented Rhizobium strain, BTAi1 (EVANS et al. 1990), which nodulates the stem and rootsof Aeschynomene indica, and the photosyntheticBradyrhizobium strains described by LORQUIN et al.(1997). The pigment extracts of all Lotononis isolatesshowed main absorption bands around 400 to 550 nm(Figure 4a). Rescaling of the region between 700 and 800nm revealed the Bchl a peak around 760 nm (Fig. 4b).The acidification of the extracts resulted in a characteris-tic replacement of the Bchla band by the bacteriopheo-phytin (Bphea) band around 745 nm (Fig. 4c).

Discussion

In South Africa almost 100 different genera of legumi-nous plants are found, growing under diverse geographi-cal and climatological conditions (STRIJDOM, 1998). Inthis study isolates obtained from the root nodules ofLotononis bainesii were characterised in terms of pheno-

typic features, 16S rDNA and nifH phylogeny. Thesesymbionts are pink-pigmented and 16S rDNA sequencingproved them to be closely related to the non-pigmentedM. nodulans. These pink colorations were due to thepresence of carotenoids, similar to those found in thephotosynthetic Rhizobium-BTAi 1 and Bradyrhizobiumstrains. Since the ability of 16S rDNA sequences, to dis-criminate between these novel methylotrophic nodulatingstrains, is unknown, DNA:DNA homologies studies arenecessary to determine the species status of the Lotononisbainesii isolates.

The methylotrophic nature of the Lotononis bainesiiisolates was indicated by growth on medium 72 (resultsnot shown), with methanol as sole carbon source. How-ever, the maximum methanol tolerance values were notdetermined. Methylotrophy is not a common trait amongrhizobia since none of the rhizobial reference strains werecapable of utilising methanol.

Nitrogen fixing genes (nif genes) are found in manybacteria besides rhizobia (HAUKKA et al., 1998). Althoughit has been reported that the nifH phylogeny closely re-sembles that of the 16S rRNA gene (HENNECKE et al.,1985; UEDA et al., 1995), a report by EARDLY et al. (1992)

1. B

rady

rhiz

obiu

m ja

poni

cum

2. B

rady

rhiz

obiu

m li

aoni

ngen

se*

3. B

rady

rhiz

obiu

m e

lkan

ii

4. A

zorh

izob

ium

cau

linod

ans

5. R

hizo

bium

etl

i bv.

pha

seol

i

6. R

hizo

bium

legu

min

osar

um

7. M

esor

hizo

bium

loti

8. M

ethy

losi

nus

sp. L

W4

9. M

ethy

losi

nus

thri

chos

pori

um

10. M

ethy

losi

nus

LW3

11. M

ethy

locy

stis

LW

5a

12. A

zosp

irill

um b

rasi

lens

e

13. x

ct7

14. R

hodo

bact

er c

apsu

latu

s

15. P

seud

omon

as s

tutz

eri

16. M

aric

hrom

atum

pur

pura

tum

17. V

ibri

o di

azot

hrop

hicu

s

18. K

lebs

iella

pne

umon

iae

Table 3. Sequence identities, based on a 244 bp fragment of the nifH gene of isolate xct7 isolated from Lotononis bainesii and other relat-ed bacteria from the α- & γ-Proteobacteria. These identities were determined form a distance matrix created using DNAdist (Phylip ver-sion 3.5c). Members of the γ-Proteobacteria are shaded. The rest all belong to the α–group.

1. B. japonicum (K01620) 1002. B. liaoningense* 96 1003. B. elkanii* 87 87 1004. A. caulinodans (M16709) 77 77 84 1005. R. etli bv.phaseoli (M15942) 69 69 77 74 1006. R. leguminosarum (K00490) 60 59 65 66 79 1007. Mesorhizobium loti 66 66 69 73 76 65 1008. Methylosinus sp. LW4 (AF378721) 70 70 77 84 73 59 70 1009. M. thrichosporium (AF378724) 72 72 80 84 77 61 72 95 10010. Methylosinus LW3 (AF378720) 72 72 74 78 74 60 71 90 89 10011. Methylocystis LW5 (AF378719) 73 73 82 85 77 62 74 90 92 85 10012. Azospirillum brasilense (X51500) 74 74 80 80 83 73 79 83 85 81 83 10013. xct7 72 72 82 81 80 73 83 82 84 79 83 90 10014. Rhodobacter capsulatus (X07866) 60 59 70 70 76 68 73 73 75 75 74 84 76 10015. Pseudomonas stutzeri (AJ297529) 66 66 74 77 77 73 70 74 79 74 78 83 80 73 10016. M. purpuratum (AF059648) 68 68 71 75 75 70 71 79 81 80 77 81 81 71 87 10017. Vibrio diazothrophicus (AF111110) 56 56 60 65 58 56 57 60 60 60 64 57 55 58 73 67 10018. Klebsiella pneumoniae (J1740) 66 66 70 75 69 66 64 72 70 70 71 72 68 71 78 79 76 100

* Sequence determined in a related study (unpublished results)


presented evidence of phylogenetic discordance thatcould be due to the lateral transfer of nif genes. In this re-port it was difficult to determine the exact phylogeneticposition of xct7 based on nifH sequence due to a lack ofcorresponding sequence data for the Methylobacteriumgenus. However, the nifH sequence of xct7 was related tothat of Azospirillum brasilense, a related genus within theα- Proteobacteria. Therefore, more nifH sequence data ofthe genus Methylobacterium is needed to establish thetrue nifH phylogeny of the Lotononis isolates.

Recently, MOULIN, et al. (2001) described the isolationof a nodulating Burkholderia sp. (β-subclass of Pro-teobacteria) from the legume Aspalathus carnosa. Thisfinding showed the range of bacteria able to nodulatelegumes is more widespread than previously anticipated.Two novel features (nodulation by β-Proteobacteria andmethylotrophy) of legume symbiosis are now known. It ishowever, interesting to note that the plant genera (Crota-laria, Aspalathus and Lotononis) involved have the sametribal affiliation (Tribe Crotalarieae). Further investiga-tions of the symbionts associated with plant species with-in this tribe and other uninvestigated legumes are there-fore warranted. It should now be evident that our under-standing of the true bacterial diversity involved in legumesymbiosis is very limited and can only increase as morehost species are investigated.

AcknowledgementsThe research was made possible by funding obtained from

the National Research Foundation. We would also like to thankDr. I. J. Law for supplying the strains and Dr. M. Oosthuizen foruseful discussions during the preparation of this manuscript.

References

AUMAN, A. J., SPEAKE, C. C. and LINDSTORM, M. E.: nifH Se-quences and nitrogen fixation in TypeI and Type II methan-otrophs. Appl. Environ. Microbiol. 67, 4009–4016 (2001).

BELGIAN COORDINATED COLLECTION OF MICROOR-GANISMS/ LABORATORIUM VOOR MICROBIOLOGIE.Bacteria. Catalogue. University Gent, Ghent, Belgium 1998.

CHEN, W. X., YAN, G. H. and LI, J. L.: Numerical taxonomicstudy of fast-growing soybean rhizobia and a proposal thatRhizobium fredii be assigned to Sinorhizobium gen. nov. Int.J. Syst. Bacteriol. 38, 392–397 (1988).

DE LAJUDIE, P., LAURENT-FULELE, E., WILLEMS, A., TORCK, U.,COOMAN, R., COLLINS, M. D., KERSTERS, K., DREYFUS, B. andGILLIS, M.: Allorhizobium undicola gen. nov., sp. nov., nitro-gen-fixing bacteria that efficiently nodulate Neptunia natansin Senegal. Int. J. Syst. Bacteriol. 48, 1277–1290 (1998).

DREYFUS, B., ALAZARD, D. and DOMMWEGUES, Y. R.: Stem-nodu-lating rhizobia, pp. 161–169. In: Current perspectives in mi-crobial ecology (M. C. KLUG and C. E. REDDY, eds.) Washing-ton D.C, American Society for Microbiology 1984.

EARDLY, B. D., YOUNG, J. P. W., SELANDER, R. K. Phylogeneticposition of Rhizobium sp. Strain Or 191, a symbiont of bothMedicago sativa and Phaseolus vulgaris, based on partial se-quences of the 16S rRNA and nifH genes. Appl. Environ. Mi-crobiol. 58, 1809–1815 (1992).

EVANS, W. R., FLEISCHMAN, D. E., CALVERT, H. E., PYATI, P. V.,ALTER, G. M., SUBBA-RAO, N. S.:Bacteriochlorophyll andphotosynthetic reaction centers in Rhizobium strains BTAi1.Appl. Environ. Microbiol. 56, 3445–3449 (1990).

FELSENSTEIN, J.: Confidence limits on phylogenies: an approachusing the bootstrap. Evolution 39, 783–791 (1985).

FELSENSTEIN, J.: PHYLIP-phylogeny inference package (version3.2). Cladistics 5, 164–166 (1989).

GARLAND, J. L., MILLS, A. L.: Classification and characterisationof heterotrophic microbial communities on the basis of pat-terns of community-level sole-carbon source utilization. Appl.Environ. Microbiol. 57, 2.351–2359 (1991).

HENNECKE, H., KALUZA, K., THONY, B., FUHRMAN, M., LUDWIG,W., STACKELBRANT, E. Concurrent evolution of nitrogenasegenes and 16S rRNA in Rhizobium species and other nitro-gen-fixing bacteria. Arch. Microbiol. 142, 342–348 (1985).

HIRAISHI, A., FURUHATA, K., MATSUMOTO, A., KOIKE, K. A.,FUKUYAMA, M. and TABUCHI, K.: Phenotypic and genetic di-versity of chlorine-resistant Methylobacterium strains isolat-ed from various environments. Appl. Environ. Microbiol. 61,2099–2107 (1995).

HOLLAND, M. A.: Methylobacterium and plants. Recent Res.Dev. Plant Physiol. 1, 207–211 (1997).

HAUKKA, K., LINDSTORM, K., YOUNG, J. P. W. Three phylogeneticgroups of nodA and nifH genes in Sinorhizobium andMesorhizobium isolates from leguminous trees growing inAfrica and Latin America. Appl. Environ. Microbiol. 64,419–426 (1998).

JARVIS, B. D. W., VAN BERKUM, P., CHEN, W. X., NOUR, S. M.,FERNANDEZ, M. P., CLEYET-MAREL, J. C. and GILLIS, M.:Transfer of Rhizobium loti, Rhizobium huakuii, Rhizobiumciceri, Rhizobium mediterraneum, and Rhizobium tiansha-nense to Mesorhizobium gen. nov. Int. J. Syst. Bacteriol. 47,895–898 (1997).

JORDAN D. C.: Transfer of Rhizobium japonicum Buchanan1980 to Bradyrhizobium gen. nov., a genus of slow-growingroot nodule bacteria from leguminous plants. Int. J. Syst. Bac-teriol. 32, 136–139 (1982).

KLEINIG, H. and BROUGHTON, W. J.: Carotenoid pigments in ared strain of Rhizobium from Lotononis bainesii Baker. Arch.Microbiol. 133, 164 (1982).

KRUGER, J. T.: Generic identity of putative rhizobial isolates asdetermined by PCR-RFLP of 16S rDNA and selected pheno-typic properties. M. Sc Thesis, University of Pretoria, Preto-ria, South Africa (1998).

KUHNERT, P., CAPAUL, S. E., NICOLET, J., FREY, J.: Phylogeneticpositions of Clostridium chauvoei and Clostridium septicumbased on 16S rRNA gene sequences. Int. J. Syst. Bacteriol. 46,1174–1176 (1996).

LAGUERRE, G., ALLARD, M. R., REVOY, F., AMARGER, N. Rapididentification of rhizobia by restriction fragment length poly-morphism analysis of PCR-amplified 16S rRNA genes. Appl.Environ. Microbiol. 60, 56–63 (1994).

LORQUIN, J., MOLOUBA, F. DREYFUS, B. L. Identification of thecarotenoid pigment canthaxanthin from photosyntheticBradyrhizobium strains. Appl. Environ. Microbiol. 63,1151–1154 (1997).

MARTINEZ-ROMERO, E., CABALLERO-MELLADO, J. Rhizobiumphylogenies and bacteria genetic diversity. Crit. Rev. PlantSci. 15, 113–140 (1996).

MOULIN, L., MUNIVE, A., DREYFUS, B. and BOIVIN-MASSON, C.:Nodulation of legumes by members of the β- subclass of Pro-teobacteria. Nature 411, 948–950 (2001).

NEI, M., LI, W. H. Mathemitical model for studying genetic vari-ation in terms of restriction endonucleases. Proc. Natl. Acad.Sci. USA 76, 5269–5273 (1979).

NORRIS, D. O.: A red strain of Rhizobium from Lotononisbainesii Baker. Aust. J. Agric. Res. 9, 629–632 (1958).

PATT, T. E., COLE, G. E. and HANSON, R. S.: Methylobacterium,a new genus of facultatively methylotrophic bacteria. Int. J.Bacteriol. 26, 226–229 (1976).


PERRIÈRE, G., and GOUY, M.: WWW-Query: an online retrievalsystem for biological sequence banks. Biochimie 78, 364–369(1996).

SAMBROOK, J. E., FRITSCH, E. F. and MANIATIS, T.A.: Molecularcloning: a laboratory manual, 2nd ed. Cold Spring HarborLaboratory, New York, Cold Spring Harbor Laboratory1989.

SAITOU, N. and NEI, M.: The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol. Biol.Evol. 4, 406–425 (1987).

SHEARING, D.: Karoo. South African wild flower guide 6. Botan-ical Society of South Africa, Kirstenbosch, South Africa 1994.

SNEATH, P. H. A., SOKAL, R. R.: NUMERICAL TAXONOMY.The principles and practices of numerical classification. SanFrancosco, W. H. FREEMAN 1973.

STRIJDOM, B. W.: South African studies on biological nitrogen-fixing systems and the exploitation of the nodule bacterium-legume symbiosis. S. Afr. J. Sci. 94, 11–23. (1998).

SY, A., GIRAUD, E., JOURAND, P. GARCIA, N., WILLEMS, A., DE LA-JUDIE, P., PRIN, Y., NEYRA, M., GILLIS, M., BOIVIN-MASSON, C.DREYFUS, B.: Methylotrophic Methylobacterium bacterianodulate and fix nitrogen in symbiosis with legumes. J. Bacte-riol. 183, 214–220 (2001).

THOMPSON, J. D., GIBSON, T. J., PLEWNIAK, F., JEANMOUGIN, F.,HIGGINS, D. G.: The ClustalX windows interface: flexiblestrategies for multiple sequence alignment aided by qualityanalysis tools. Nucl. Acid. Res. 25, 4876–4882 (1997).

UEDA, T., SUGA, Y., YAHIRO, N., MATSUGUCHI, T. RemarkableN2-fixing bacterial diversity detected in rice roots by molecu-lar evolutionary analysis of the nifH gene sequences. J. Bacte-riol. 177, 1414–1417 (1995).

VAN BERKUM P., EARDLY, B. D.: Molecular evolutionary system-atics of the Rhizobiaceae, pp. 1–24. In: The Rhizobiaceae,Molecular Biology of Plant-Associated Bacteria, (H. P.

SPAINK, A. KONDOROSI, P. J. HOOYKAAS, eds). Dordrecht:Kluwer Academic Publishers 1998.

VAN WYK, B-E.: A synopsis of the genus Lotononis (Fabaceae:Crotolarieae). Contributions from the Bolus Harbarium.Cape Town: The Rustica Press 1991.

VINCENT, J. M.: A manual for the practical study of root-nodulebacteria. Oxford, Blackwell Scientific Publications 1970.

WEISBURG, W. G., BARNS, S. M., PELLETIER, D. A. and LANE, D.J.: 16S Ribosomal DNA amplification for phylogenetic study.J. Bacteriol. 173, 697–703 (1991).

WILSON, K.: Preparation of genomic DNA from bacteria, section2.4.1–2.4.5. In: Current protocols in molecular biology. (F.M. AUSUBEL, R. BRENT, R. E. KINGSTON, D. D. MOORE, J. A.SMITH, J. G. SEIDMAN, K. S. STRUHL, eds.) New York, JohnWILEY & SONS 1994.

YOUNG, J. P. W.: Phylogeny and taxonomy of rhizobia. Plant Soil186, 45–54 (1996).

YOUNG, J. P. W., HAUKKA, K. Diversity and phylogeny of rhizo-bia. New Phytol 133, 87–94 (1996).

YOUNG, J. M., KUYKENDALL, L. D., MARTINEZ-ROMERO, E.,KERR, A., SAWADA, H.: A revision of Rhizobium Frank 1889,with an emended description of the genus, and the inclusionof all species of Agrobacterium Conn 1942 and Allorhizobi-um undicola de Lajudie et al.1998 as new combinations: Rhi-zobium radiobacter, R. rhizogenes, R. rubi, R. undicola andR. vitis. Int. J. Syst. Evol. Microbiol. 51, 89–103 (2001).

Corresponding author:J. B. JAFTHA, Department Microbiology and Plant Pathology,Agricultural Building, Room 11–4, University of Pretoria, Preto-ria, 0001, South AfricaTel.: ++27-12-420 4562; Fax: ++27-12-420 3266;e-mail: [email protected]

characterization of pigmented methylotrophic bacteria...

Documents