Diverse Bacteria Associated with Root Nodules of Spontaneous Legumes in Tunisia and First Report for...

MicrobialEcology

Diverse Bacteria Associated with Root Nodules of SpontaneousLegumes in Tunisia and First Report for nifH -like Genewithin the Genera Microbacterium and Starkeya

Frederic Zakhia1, Habib Jeder2, Anne Willems3, Monique Gillis3, Bernard Dreyfus1

and Philippe de Lajudie1,4

(1) Laboratoire des Symbioses Tropicales et Mediterraneennes, Montpellier, France(2) Institut des Regions Arides, Nahal-Gabes, Tunisia(3) Laboratorium voor Microbiologie, Universiteit Gent, Ghent, Belgium(4) LSTM UMR113 IRD/CIRAD/AGRO-M/UM II/USC-INRA, Campus International de Baillarguet, TA10 / J, 34398 Montpellier Cedex 5, France

Received: 8 April 2005 / Revised: 8 April 2005 / Accepted: 12 April 2005 / Online publication: 6 April 2006

Abstract

We characterized 34 endophytic bacterial isolates associ-ated to root nodules collected from spontaneous legumesin the arid zone of Tunisia by 16S rDNA polymerase chainreaction (PCR)–restriction fragment length polymor-phism, whole cell protein sodium dodecyl sulfate–poly-acrylamide gel electrophoresis (SDS-PAGE), 16S rDNAand 16S–23S rDNA internal transcribed spacer sequenc-ing. Phylogenetically, these isolates belong to the branchescontaining the genera Inquilinus, Bosea, Rhodopseudomo-nas, Paracraurococcus, Phyllobacterium, Ochrobactrum,Starkeya, Sphingomonas, Pseudomonas, Agromyces, Micro-bacterium, Ornithinicoccus, Bacillus, and Paenibacillus.These strains did not induce any nodule formation wheninoculated on the wide host spectrum legume species M.atropurpureum (Siratro) and no nodA gene could be am-plified by PCR. However, nifH sequences, most similarto those of Sinorhizobium meliloti, were detected withinstrains related to the genera Microbacterium, Agromyces,Starkeya and Phyllobacterium.

Introduction

Legume-nodulating bacteria (LNB) are soil bacteria thatcan establish a symbiosis with legume plants. Duringspecific interactions with legumes, these bacteria enterroot tissues via root hairs or directly via wounded tis-sues (Bcrack entry^) and induce nodule formation on

roots and/or shoots [35]. Inside the nodule, they fix di-nitrogen to the benefit of the plant. Until recently, allknown LNB belonged to genera within the a class ofProteobacteria, namely, Rhizobium, Mesorhizobium,Bradyrhizobium, Allorhizobium, Sinorhizobium, and Azo-rhizobium [119]. During the past 3 years, the investiga-tion of new nodule isolates has led to the discovery ofLNB in unexpected genera in the a-Proteobacteria, i.e.,Methylobacterium [94], Devosia [83], Blastobacter [101],and recently, Ochrobactrum [70]. Moreover, strains be-longing to genera in the b-Proteobacteria, i.e., Burkhold-eria [69] and Ralstonia [22] were shown to be LNB. Morerecently, Benhizia et al. [9] reported the associationbetween legume root nodules and strains belonging tothe g class of Proteobacteria.

Bacteria are abundant in soils (up to 109

cells pergram [79]) and diverse (a minimum of 4000–7000 dif-ferent bacterial genomes per gram [99]). Therefore,plants are constantly involved in interactions with a widerange of bacteria, and a number of plant-associated bac-teria colonize the rhizosphere (rhizobacteria), the phyl-losphere (epiphytes), and the inside of plant tissues(endophytes). Several authors have reported techniquesfor plant endophyte isolation [58, 66]. Our group showedthat some photosynthetic Bradyrhizobium sp. strains canboth induce nitrogen-fixing stem nodules on Aeschyno-mene spp. and be natural endophytes of rice roots inAfrica [21]. Other LNB have been reported to enter ben-eficial endophytic associations with cereals, i.e., Rhizobi-um leguminosarum with rice and wheat in Egypt [14, 29,116, 117] or Rhizobium etli with maize [46]. Some ni-trogen-fixing bacteria, like Serratia marcescens, are endo-Correspondence to: Philippe de Lajudie; E-mail: [email protected]

DOI: 10.1007/s00248-006-9025-0 & Volume 51, 375–393 (2006) & * Springer Science+Business Media, Inc. 2006 375

phytic of rice [47], and others have been shown to beplant-growth-promoting (PGP) endophytes like Aceto-bacter, Herbaspirillum [82], Burkholderia [25, 43], andAzoarcus [81]. More and more plant-associated bacteriaare characterized; for example bacteria of the genera Mi-crobacterium, Cellulomonas, Clavibacter, and Curtobacte-rium are able to colonize tissues of agronomic and prairieplants [122]. Endophytic Bacillus strains were isolatedfrom soybean root nodules [5]. Therefore, internal col-onization of plants by various endophytic bacteria is notnecessarily of pathogenic nature and may be beneficial tothe plant. Although the contribution of endophytic di-azotrophic bacteria in fixed nitrogen is very low in com-parison to that provided by LNB to the legumes [60],some plants (Brazilian sugarcane varieties) have beenshown to provide for a substantial percentage of theirnitrogen requirements from N2 fixation [15].

On the other hand, Agrobacterium strains have beenisolated from nodules of many legume species [31], butno definitive explanation of the presence of these bacteriainside nodules could be demonstrated.

We previously investigated nodulation of 42 sponta-neous wild legumes in the arid zone of South Tunisia,where rainfalls do not exceed 180mm [52]. We charac-terized 60 LNB associated to 19 of these legumes by apolyphasic approach [120] and identified them as Rhizo-bium spp., Sinorhizobium spp., Mesorhizobium spp., andBradyrhizobium spp. In the time course of isolation ofthese LNB, we obtained 34 more bacterial isolates be-longing to genera where no species of LNB were de-scribed so far. Here we report on the characterization ofthese isolates. We first performed 16S amplified ribosom-al DNA restriction analysis (ARDRA), a technique de-scribed as a convenient grouping method for several kindsof bacteria like LNB [62, 71, 97], bacteria belonging to thegenera Brevibacterium [18], Bacillus, and Paenibacillus[49]. We also performed sodium dodecyl sulfate–poly-acrylamide gel electrophoresis (SDS-PAGE) of cellularproteins, also widely used for taxonomic analysis ofmany taxons, i.e., in Clostridium [20], Xanthomonas,Campylobacter, lactic acid bacteria [103], and LNB [30,32, 33, 71]. We also submitted strains to 16S rDNA se-quencing, which is most commonly applied for bacterialphylogenetic studies [114]. For strains found phyloge-netically close to Bradyrhizobium, where 16S rDNA isvery conserved among species, we performed partial se-quencing of the 16S-23S rDNA internal transcribedspacer (ITS) known to be more taxonomically informa-tive [111]. New isolates were screened for nodulationcapacity. Since seeds of their original plants of isolationwere not available (except for Retama raetam), we per-formed nodulation tests on Macroptilium atropurpureum,a legume species known to be nodulated by a wide rangeof LNB. Presence of nodA and nifH genes was looked forby polymerase chain reaction (PCR) amplification.

Methods

Bacterial Strains and Culture Media. The new isolatesand their plants of isolation are listed in Table 1. Ref-erence strains included in our study are listed in Table 2.All strains were maintained on yeast mannitol agar (YMA)[106] at 28-C, containing (in grams per liter): mannitol,10; sodium glutamate, 0.5; K2HPO4, 0.5; MgSO4 I 7H2O,0.2; NaCl, 0.05; CaCl2, 0.04; FeCl3, 0.004; yeast extract(Difco, Sparks, MD, USA), 1 (pH 6.8); agar, 20. All strainswere stored at j80-C on the same medium plus 20% v/vglycerol. For protein preparations we used tryptone–yeast(TY) extract medium containing (in grams per liter, pH6.8 to 7) tryptone (Oxoid, Basingstoke, UK), 5; yeast extract(Oxoid), 0.75; KH2PO4, 0.454; Na2HPO4 I 12H2O, 2.388;CaCl2, 1; LabM agar, 20.

Bacterial Isolation Procedure. Roots were carefullyrinsed to eliminate any possibly trapped pocket of soil.Naturally occurring root nodules were collected in naturaand either directly used for bacterial isolation or storeddried in tubes containing CaCl2 [28]. Isolation was per-formed on each nodule separately (two to three nodulesper plant). Upon use, nodules were rehydrated in sterilewater and surface sterilized by immersion in calcium hy-pochlorite (3%) for 5 min. The nodules were then asep-tically rinsed eight times with sterile water. A 100-mLaliquot of the last washing solution was checked for ste-rility by inoculation on YMA agar plates and incubation.Only nodules resulting in a sterile final washing liquidwere further considered for bacterial isolation. Noduleswere then individually crushed in a drop of sterile waterand the suspension was streaked on YMA in Petri dishes.Colonies appeared after several days of incubation at 28-Cin aerobic conditions. They were checked for purity byrepeated streaking on YMA and by microscopic exami-nation of living cells. Additional control of purity wasperformed on Microbacterium strains using Gram stain(Biomerieux, Marcy L’Etoile, France), according to themanufacturer’s instructions.

Plant Cultivation and Nodulation Tests. Seeds weresurface sterilized with 96% sulfuric acid. The time fortreatment with H2SO4 was as follows: M. atropurpureum,20 min; R. raetam, 15 minutes. Seeds were then washed10 times with sterile water to remove all traces of acid.Then they were placed into sterile water for 24 hours at4-C. The seeds were then incubated in sterile Petri dishescontaining 10% water agar for 24 to 48 h at obscurity toallow germination. Plants were then transferred to tubescontaining Jensen seedling slant agar [106] (a nitrogen-free medium; for 1 L: KH2PO4 (20 g/L) + MgSO4 (20 g/L[10 mL], NaCl (20 g/L) [10 mL], CaHPO4 (50 g/L) [20 mL],FeCl3 I 6H2O (4 g/L) [10 mL] and microelements (H3BO3

2.86 g; MnSO4 I 4H2O 2.03 g; ZnSO4 I7H2O O.22 g; CuSO4

376 F. ZAKHIA ET AL.: BACTERIA ASSOCIATED TO SPONTANEOUS LEGUMES IN TUNISIA

Table 1. New isolates and main results

StrainaOther straindesignation Origin (root nodules)

16S-ARDRAcluster

SDS-PAGEgroup 16S rDNA sequencingb

BoseaORS 1414A RAT 304 R. raetam I Sep B. thiooxidanst (97%, 1443 bp),ORS 1496A RAT 1301 As. gombiformis I Nt B. mexicanap (95%, 600 bp)STM 358A RAT 1901 O. vaginalis I Nt B. thiooxidanst (99%, 1365 bp)PhyllobacteriumORS 1419H RAT 800 As. algerianus II 2 Ph. myrsinacearumt (96%, 1438 bp)ORS 1402A, H RAT 003 Ar. uniflorum II 3 Ph. myrsinacearumt (96%, 1449 bp)ORS 1403H RAT 004 Ar. uniflorum II 3 Ph. myrsinacearumt (96%, 1419 bp)ORS 1420H RAT 801 As. algerianus III Sep Ph. myrsinacearumt (98%, 1428 bp)STM 391 RAT 2204 C. villosa III Sep Ph. myrsinacearumt (98%, 1430 bp)STM 370H RAT 3000 La. numidicus III Nt Ph. myrsinacearumt (98%, 1408 bp)StarkeyaORS 1476 RAT 312 R. raetam IV Sep NtORS 1474A, H RAT 310 R. raetam IV Sep St. novellat (98%, 1438 bp)SphingomonasORS 1497A RAT 1302 As. gombiformis V Sep Sp. asaccharolyticat (97%, 1443 bp)STM 397H RAT 2301 Lo. argenteus V Sep Sp. asaccharolyticap (99%, 595 bp)STM 384H RAT 2101c C. villosa V Nt Sp. asaccharolyticap (99%, 466 bp)InquilinusORS 1421A RAT 802 As. algerianus VI 2 I. limosusp (99%, 611 bp)ORS 1422A RAT 803 As. algerianus VI 2 I. limosusp (98%, 674 bp)STM 378A RAT 863b Astragalus sp. VI Sep I. limosusp (98%, 343 bp)STM 364A RAT 2101b C. villosa VI Nt I. limosusp (99%, 472 bp)STM 385A RAT 2101d C. villosa VI Nt I. limosusp (98%, 548 bp)STM 377A RAT 863a Astragalus sp. VI Nt I. limosusp (99%, 126 bp)PseudomonasSTM 368 RAT 2401 M. truncatula VII 1 Pseudomonas syringaet,n pv. phaseolicola

(99%, 1474 bp)ORS 1432H RAT 1001 H. carnosum VII Sep Pseudomonas brassicacearump

(98%, 582 bp)ActinomycetalesORS 1472H RAT 308 R. raetam VIII Sep Mi. flavescenst (97%, 1489 bp)ORS 1417H RAT 603 O. natrix subsp. falcata VIII Sep Mi. flavescenst (98%, 1456 bp)ORS 1418H RAT 604 O. natrix subsp. falcata VIII Sep Mi. flavescenst (98%, 1489 bp)ORS 1480H RAT 317 R. raetam VIII Nt Mi. barkerip (98%, 417 bp)ORS 1437H RAT 15 Ar. uniflorum Sep Nt Agromyces cerinump (98%, 550 bp)ORS 1481H RAT 405 As. armatus Sep Nt My. frederiksbergensep

(100%, 552 bp)STM 379H RAT 2400 M. truncatula Sep 1 Ornithinicoccus hortensisp (97%, 573 bp)OthersSTM 392H RAT 2200 C. villosa Sep Sep Bacillus circulansp (97%, 333 bp)ORS 1478A RAT 314 R. raetam Sep Sep Ochrobactrum grignonensep (95%, 604 bp)STM 388A RAT 2303 Lo. argenteus Sep Sep Paenibacillus lautusp (96%, 450 bp)ORS 1473A RAT 309 R. raetam Sep Sep Pa. rubert (94%, 1443 bp)ORS 1416riH O. natrix subsp. falcata Nt Nt Rhodopseudomonas rhenobacensist

(97%, 1446 bp)aAll strains except ORS 1416ri have been tested for nodulation on Ma. atropurpureum. Strains ORS 1414, ORS 1472, ORS 1473, ORS 1474, ORS 1476, ORS1478, and ORS 1480 have been tested for nodulation on both Ma. atropurpureum and R. raetam.

bThe closest match in NCBI is shown. Between brackets are given, respectively, the percentage of homology with the most similar published 16S rDNAsequence and the number of base pairs considered.

AStrain screened for nodA gene.tTotal 16S rDNA sequence.pPartial rDNA sequence.HStrain screened for nifH gene.nOne sense sequence was determined.Nt: not tested; Sep: separate.Ar.: Argyrolobium; As.: Astragalus; B.: Bosea; C.: Calycotome; H.: Hedysarum; I.: Inquilinus; La.: Lathyrus; Lo.: Lotus; M.: Medicago; Ma.: Macroptilium;Mi.: Microbacterium; My.: Mycobacterium; O.: Ononis; Pa.: Paracraurococcus; Ph.: Phyllobacterium; R.: Retama; Sp.: Sphingomonas; St.: Starkeya.

F. ZAKHIA ET AL.: BACTERIA ASSOCIATED TO SPONTANEOUS LEGUMES IN TUNISIA 377

Table 2. Reference strains used in this study

StrainOther

designation Host plant or originGeographical

originReferenceor source

16S-ARDRAcluster

SDS-PAGEgroup

Agrobacterium rhizogenesORS 1352T LMG 150T – – Sep SepAgrobacterium rubiORS 1353 LMG 159 Rubus sp. USA, 1942 [54] Sep NtAgrobacterium tumefaciensORS 1351T LMG 140T – Sep SepORS 2644T LMG 187T Lycopersicon lycopersicon USA [54] Sep NtAgrobacterium vitisORS 2643 LMG 257 Vitis vinifera Greece LMG Sep SepAllorhizobium undicolaORS 992T LMG 11875T N. natans Senegal [30] Sep NtORS 995 LMG 11876 N. natans Senegal [30] Sep SepAzorhizobium caulinodansORS 571T LMG 6465T Sesbania rostrata Senegal [38] Sep SepBradyrhizobium japonicumNZP 5549T LMG 6138T Glycine max Japan – Nt SepBurkholderia tuberumSTM 678T LMG 21444T Aspalathus carnosa South Africa [69] Sep NtMesorhizobium amorphaeSTM 238 LMG 18932

or 18960Amorpha fruticosa China Sep Nt

Mesorhizobium chacoensepr5 LMG 19008T Prosopis alba Argentina [105] Nt SepMesorhizobium ciceriORS 2738T LMG 17150T C. arietinum L. Spain [74] Sep Nt

LMG 14989T C. arietinum L. Spain [74] Nt SepMesorhizobium huakuiiORS 1752T LMG 14107T Astragalus sinicus China [23] Sep SepMesorhizobium lotiORS 664T LMG 6125T Lotus tenuis New Zealand [51] Sep SepMesorhizobium mediterraneumCa-36T, ORS 2739T LMG 17148T C. arietinum Spain [73] Nt SepORS 2754 Sep NtMesorhizobium plurifariumORS 654 LMG 10056 Leucaena diversifolia Brazil [32] Sep NtORS 1030 LMG 11890 Acacia senegal Senegal [32] Nt SepORS 1032T LMG 11892T A. senegal Senegal [32] Nt SepMesorhizobium tianshanenseORS 2640T LMG 15767T Glycyrrhiza pallidiflora China [24] Sep SepMe. nodulansORS 2060T Crotalaria podocarpa Senegal [94] Sep SepR. etliORS 645T LMG 11937T Phaseolus vulgaris L. Mexico [87] Sep SepRhizobium galegaeHAMBI 540T LMG 6214T,

ORS 668TGalega orientalis Finland [65] Sep Sep

Rhizobium gallicum bv. phaseoliR4384 PHI21 P. vulgaris France [1] Nt SepR. giardinii bv. giardiniiR4385 H152T P. vulgaris France [1] Nt SepR. hainanenseCCBAU 57003 LMG 18075R. huautlenseSO2, STM 279T LMG 18254T Sesbania herbacea Sierra de Huautla,

Mexico[108] Sep Sep

S25 LMG 18256 S. herbacea Sierra de Huautla,Mexico

[108] Nt Sep

R. leguminosarum bv. phaseoliORS 662 LMG 8819T2 Sep Nt


0.08 g; NaMoO4 IH2O: 0.09 g) [1 mL]). For root nodu-lation trials, three plants were routinely tested with eachstrain. Control plants of noninoculated M. atropurpur-eum and R. raetam were included. Plants were grown inchamber set for a 22-C/18-C light/dark thermoperiodand 16:8 h light/dark photoperiod with fluorescent lampsBcold white^ at a photon flux density of 130 mmol mj2 sj1.They were inoculated by using 1 mL of YM bacterialsuspension in exponential stage of growth. Although YMcontains forms of nitrogen (yeast extract and glutamate),the common practice shows that this little residual ni-trogen does not inhibit nodule formation, but on thecontrary, it may act as a plant growth starter for nodu-lation. Roots were observed for nodule formation during6–8 weeks after inoculation.

Analysis of Protein Electrophoretic Patterns. Allstrains were grown at 28-C for 48 h in Roux flasks onbuffered TY medium. Whole-cell protein extracts wereprepared from 80 mg cells and SDS-PAGE was performedusing small modifications of the procedure of Laemmli[61], as described previously [33]. The normalized den-sitometric traces of the protein electrophoretic patternswere grouped by numerical analysis, using theGelCompar 4.2 software package [104]. The similaritybetween all pairs of traces was expressed by the Pearsonproduct–moment correlation coefficient (r) convertedfor convenience to a percent value [77, 78].

DNA Preparation. Cells from 2- to 3-day-old YMAplate cultures were suspended in sterile water and washed

R. leguminosarum bv. trifoliiORS 663 LMG 8820 Sep NtR. leguminosarum bv. viciaeORS 639 LMG VF39SM Sep Nt

LMG 8817T Nt SepRhizobium mongolenseUSDA 1844T, STM 246T LMG 19141T Medicago ruthenica Mongolia, China [102] Sep SepRhizobium tropici IIaORS 651 LMG 9517 P. vulgaris L. Brazil [67] Sep NtR. tropici IIbORS 625T, CIAT 899T LMG 9503T P. vulgaris Colombia [67] Sep SepSinorhizobium adhaerensR-14065T ATCC 33212T Soil USA [19, 112] Nt SepSinorhizobium arborisORS 1755T LMG 14919T Prosopis chilensis Sudan [71] Sep Nt

LMG 15624 A. senegal Sudan [71] Nt SepLMG 15626 A. senegal Sudan [71] Nt Sep

Sinorhizobium frediiORS 669T LMG 6217T G. max China [51] Sep SepSinorhizobium kostienseORS 513 LMG 15613 Pr. chilensis Soudan [71] Sep NtSinorhizobium medicaeM 102, ORS 504 LMG 16580,

HAMBI 1809M. truncatula Syria [39] Sep Nt

M75 LMG 16579,HAMBI 1808

Medicago radiata Syria [39] Nt Sep

Sinorhizobium melilotiNZP 4027T, ORS 665T LMG 6133T Medicago sativa Virginia, USA – Sep SepSinorhizobium morelenseR-13987T, Lc04T LMG 21331T Leucaena leucocephala Mexico [107] Nt SepSinorhizobium saheliORS 609T LMG 7837T Sesbania cannabina Senegal [33] Sep SepORS 611 LMG 7842,

LMG 8310Sesbania grandiflora Senegal [33] Sep Sep

Sinorhizobium terangaeORS 51 LMG 6464 S. rostrata Senegal [33] Sep NtORS 1009T LMG 7834T Acacia laeta Senegal [33] Sep SepORS 1007 LMG 7847 A. laeta Senegal [33] Nt Sep

Nt, not tested; Sep, separate; Ttype strain.A.: Acacia; C.: Cicer; G.: Glycine; M.: Medicago; Me.: Methylobacterium; Ne.: Neptunia; P.: Phaseolus; Pr.: Prosopis; R.: Rhizobium; S.: Sesbania.

Table 2. Continued

StrainOther

designation Host plant or originGeographical

originReferenceor source

16S-ARDRAcluster

SDS-PAGEgroup


twice. The optical density at 620 nm (OD620 nm) wasthen measured, and a volume corresponding to 200 mLmatching an OD620 nm = 1 was centrifuged at 12,000 rpmfor 2 min. The cell pellet was suspended in 100 mL ofsterile water, and 100 mL of Tris-Cl (10 mM, pH 8.3) and20 mL of proteinase K (1 mg L

_1 solution) (Promega,Madison, WI, USA) were successively added. After 2 hincubation at 55-C, proteinase K was denaturated by in-cubation for 10 min in boiling water. The solution wasthen centrifuged at 14,000 rpm for 10 min. The super-natant containing DNA was stored at j20-C until use.

16S ARDRA. Nearly full-length 16S ribosomalDNA was amplified using the universal eubacterial 16SrDNA primers FGPS6 and FGPS1509 (Table 3). PCRamplification was carried out in a 50-mL reaction mix-ture containing template DNA (2 mL), 5 mL of reactionbuffer (Invitrogen, San Diego, CA, USA), 1.5 mL of MgCl250 mM, 4 mL of deoxynucleotide triphosphate (dNTP)2.5 mM each (Sigma-Aldrich, Deisenhofen, Germany),2 mL of each primer (20 mM) (Biotech AG, Ebersberg,Germany), 1 U of Taq polymerase (Invitrogen) and33.2 mL of sterile Milli-Q water. A negative controlwithout template was included in every PCR run. Ampli-fication was performed in a Gene Amp PCR system 2400(PerkinElmer, Nieuwerkerk a/d Ijssel, the Netherlands)using the following program: initial denaturation (5 minat 94-C), 35 cycles of denaturation (30 s at 94-C), an-nealing (30 s at 55-C) and extension (7 min at 72-C),and final extension (7 min at 72-C).

PCR products were checked by electrophoresis in 1%agarose gel in TAE buffer (40 mM Tris-acetate, 1 mMEDTA, pH 8,3) containing two drops of 0.5 mg/mL ethid-ium bromide.

For ARDRA analysis, 10-mL aliquots of PCR productswere digested with MspI, CfoI, HinfI, RsaI, AluI, NdeII,and TaqI [62] for 2 h in a 20-mL final volume as specifiedby the manufacturer (Promega) but with an excess of en-

zyme (5 U per reaction). Restriction fragments were ana-lyzed by horizontal gel electrophoresis using 3% metaphoragarose (FMC Bioproducts, Rockland, ME, USA) in TBEbuffer (89 mM Tris borate, 2 mM EDTA, pH 8,3) con-taining two drops of 0.5 mg mLj1 ethidium bromide. Theelectrophoresis was performed using a 15 � 10-cm tray,12 wells (Easy Cast, Electrophoresis system model B1,OW1, Scientific Inc.) at 50 V for 15 min and at 80 V for 3h. The gel images were scanned, digitized, and stored in acomputer using Perfect Image software (V-5.3, Clara Vi-son, Germany). Profiles were normalized, combined, andclustered using the GelCompar II 2.5 software package(Applied Maths, Austin, TX, USA) [104]. Dice coeffi-cient and unweighted pair group method using averagelinkage (UPGMA) clustering were used.

16S rDNA Sequencing. PCR products of the 16SrRNA gene were run on a 1% agarose gel, the band wasexcised and DNA purified using a QIAquick gel extractionkit (Qiagen, Courtaboeuf, France) and sequenced by usingthe primers FGPS6, FGPS1509, 16S-370f, 16S-1080r, 16S-870f and 16S-1924r (Table 3). Sequence reactions wereperformed using the ABI Prism BigDye Terminator Cyclesequence kit (Applied Biosystems, Foster City, CA, USA)and analyzed on an Applied Biosystems model 310 DNAsequencer. Sequences were assembled using the Autoas-sembler program (PerkinElmer) and aligned with ClustalX and Genedoc software packages. A phylogenetic tree wasconstructed using the neighbor-joining algorithm [86]with PAUP 4 software, and the stability of the groupingswas assessed by performing a bootstrap analysis. Forpartial 16S rDNA sequencing, we used primer FGPS6,FGPS1509, or 16S-1080r.

16S-23S rDNA (ITS) Sequencing. The internalspacer of the 16S-23S rDNA region was amplified usingthe primers 16S-870f and FGPL20540 (Table 3). A touch-down PCR [36] was performed with a PerkinElmer model

Table 3. Primers used for DNA amplification and sequencing

Primer name Primer sequence Target gene Reference

FGPS 6 50-GGA GAG TTA GAT CTT GGC TCA G-30 16S rDNA [72]FGPS 1509 50-AAG GAG GGG ATC CAG CCG CA-30 16S rDNA [72]16S-870f 50-CCT GGG GAG TAC GGT CGC AAG-30 16S rDNA [94]16S-1080r 50-GGG ACT TAA CCC AAC ATC T-30 16S rDNA [94]16S-370f 50-GGC AGC AGT GGG GAA TAT TG-30 16S rDNA [94]16S-1924 rev 50-GGC ACG AAG TTA GCC GGG GC-30 16S rDNA [94]FGPL20540 50-CCG GGT TTC CCC ATT CGG-30 23S rDNA [72]*Br5 50-CTT GTA GCT CAG TTG GTT AG-30 RNAile [111]nodAfbrad 50-GTY CAG TGG AGS STK CGC TGG G-30 nodA [94]nodArbrad 50-TCA CAR CTC KGG CCC GTT CCG-30 nodA [94]nifH (forA) 50-GCI WTI TAY GGN AAR GGN GG-30 nifH [109]nifH (forB) 50-GGI TGT GAY CCN AAV GCN GA-30-30 nifH [109]nifH reverse 50-GCR TAI ABN GCC ATC ATY TC-30 nifH [109]

DNA sequence degeneracies are indicated by using the International Union of Pure and Applied Chemistry conventions, as follows [64]: R, A/G; Y, C/T; W,A/T; V, A/C/G; B, C/G/T and N, A/C/G/T. Inosine (I) was used to reduce the degeneracy of the primers by replacing fourfold-degenerate positions (N) inthe 50 portions [8, 17, 100].


2400 thermocycler. In a total volume of 25 mL, the reactionmixture contained 2 mL of the template DNA, 2.5 mL of 10�buffer (Eurogentec, Cologne, Germany), 1.5 mL 25 mMMgCl2, 2 mL of dNTPs (2.5 mM each), 1 mL of eachprimer (20 mM), 1 U Taq DNA polymerase (Gibco-BRL,Carlsbad, CA, USA). The amplification was performedusing the following program: BTouchdown^ PCR con-sisting of an initial denaturation step (96-C, 5 min) fol-lowed by 20 cycles of denaturation (96-C, 30 s), annealing(from 65-C to 55-C, 30 s) and extension (72-C, 90 s, fol-lowed by 20 cycles of denaturation (96-C, 30 s), annealing(55-C, 30 s), extension (72-C, 90 s), and final extension(72-C, 7 min). The PCR products were purified using aQuiaquick PCR purification kit (Qiagen) according to themanufacturer’s instructions and sequenced using theprimers Br5*, as described by Willems et al. [111], andFGPL20540 [72].

Amplification and Sequencing of nifH Genes. Dueto the distant phylogenetic relationships among nitrogenfixers, the sequences of nifH genes have diverged con-siderably [121]. Therefore, the design of universal nifHprimers requires a high degree of DNA sequence degen-eracy and may result in reduced specificity during PCRamplification. The Bnested^ PCR [13, 40] has been re-ported to be a useful tool to check the presence of nifHsequences within microorganisms [109]. It consists in atwo-step PCR: The first step may produce both theexpected gene and nonspecific amplificates; the secondstep is then performed using one of the previous primersand an internal one, resulting in more specific amplifica-tions. Presence of nifH sequences was checked here byperforming a nested PCR [13, 40] and using threeprimers (Table 3) as described by Widmer et al. [109].The first PCR was carried out with the forward primernifH(forA) and the reverse primer nifH(rev). The second(nested) PCR was performed with the forward primernifH(forB) and the same reverse primer nifH(rev). Theexpected band size was 370 bp.

Both PCRs were carried out in a 25-mL reactionvolume (the reaction mixture is the same as described forIGS amplification except for the addition of 1.25 mL of1% W-1 solution). The template for the first PCR was2 mL of DNA, the template for the second one was 2 mLfrom the first PCR product.

After an initial denaturation consisting of 5 min at95-C, the cycling conditions were as follows (for bothPCRs): denaturation for 11 s at 94-C and for 15 s at 92-C,annealing for 8 s at 48-C and for 30 s at 50-C, and ex-tension for 10 s at 74-C and for 10 s at 72-C. A final10-min extension step at 72-C was performed after thecycling steps. We performed 40 cycles for both PCRs. Thenested PCR products were visualized using a 2% agarosegel and the bands corresponding to 370 bp were excised,purified by Qiaquick kit, and sequenced (as described

above) using the same primers as for the nested PCR.Sequences were assembled and aligned, and a phylogenetictree was constructed as described above (Fig. 4).

Amplification of nodA Gene. Primers nodAfbradand nodArbrad (Table 3) were used for nodA amplifica-tion. A touchdown PCR [36] was performed using thefollowing conditions: an initial denaturation step (94-C,5 min) followed by 20 cycles of denaturation (94-C,30 s), annealing (from 60-C to 50-C, 30 s) and extension(72-C, 42 s), followed by 22 cycles of denaturation (94-C,30 s), annealing (50-C, 30 s), extension (72-C, 42 s andfinal extension (72-C, 7 min). The bands corresponding tothe expected size (550 to 600 bp) were cut, purified, andsequenced as described above. A positive control was add-ed (strain ORS 2060

T

of Methylobacterium nodulans) forwhich these primers were originally designed [94].

Results

Collection of New Bacterial Isolates. Nodules werecollected from wild legumes sampled in geographicallydistant sites in the infra-arid zone of Tunisia. The classicalprocedure was performed for nodule surface sterilizationand bacterial isolation. Generally, a single nodule yielded asingle colony type, but sometimes, two or more colonytypes were recovered from the same nodule and were fur-ther studied. Some strains produced abundant polysac-charides, resulting in confluent growth, others not. Wedid not take any sparse or sporadic colonies into consid-eration. In addition to the legume nodulating bacteria(LNB) characterized previously by our group as Rhizobi-um spp., Sinorhizobium spp., Mesorhizobium spp., andBradyrhizobium spp. [120], we obtained 34 other noduleisolates that were able to grow on YMA medium. These34 isolates originated from nodule samples of the fol-lowing legume species: Argyrolobium uniflorum, Astragalusalgerianus, Astragalus armatus, Astragalus gombiformis,Calycotome villosa, Hedysarum carnosum, Lathyrus nu-midicus, Lotus argenteus, Medicago truncatula, Ononis na-trix subsp. falcata, Ononis vaginalis, R. raetam (Table 1).

16S ARDRA Analysis. For the sake of clarity inthe following, the word Bcluster^ will be used to refer toARDRA grouping and the word Bgroup^ will be used torefer to SDS-PAGE grouping.

Thirty-three new isolates were characterized byARDRA. (The 34th one, ORS 1416ri, for which the com-plete 16S rDNA sequence is reported in Table 1, was notincluded in the ARDRA study.) We included referencestrains representative of known LNB species in our study(Table 2). The 16S ribosomal DNA was PCR-amplifiedand all strains produced a single band of the expectedsize of approximately 1500 bp. Profiles obtained with


Agrobacterium rhizogenes

Cluster I

Cluster II

Rhizobium tropici IIb

Rhizobium leguminosarum



Rhizobium etli

Rhizobium galegae

Rhizobium huautlense

Rhizobium mongolense

Sinorhizobium saheli


Sinorhizobium terangae


Sinorhizobium arboris

Sinorhizobium medicae

Sinorhizobium meliloti

Sinorhizobium fredii

Sinorhizobium kostiense

Agrobacterium tumefaciens


Agrobacterium rubi

Agrobacterium vitis

Allorhizobium undicola


Rhizobium tropici IIa

Mesorhizobium loti

Mesorhizobium tianshanense

Mesorhizobium ciceri

Mesorhizobium mediterraneum

Mesorhizobium huakuii

Mesorhizobium amorphaeMesorhizobium plurifarium

Cluster IV

Cluster V

100

90

80

70

60504030

OR S 1414

OR S 1496

STM 358

OR S 1352T

OR S 625T

OR S 662

OR S 663

OR S 639

OR S 645T

OR S 668T

STM 279T

STM 246T

OR S 1402

OR S 1403

OR S 1419

OR S 609T

OR S 611

OR S 51

OR S 1009T

OR S 1755T

OR S 504

OR S 665T

OR S 669T

OR S 513

OR S 1351T

OR S 2644T

OR S 1353

OR S 2643

OR S 992T

OR S 995

OR S 651

OR S 664T

OR S 2640T

OR S 2738T

OR S 2754

OR S 1752T

STM 238

OR S 654

STM 370

OR S 1420

STM 391

OR S 1478

OR S 571T

OR S 1476

OR S 1474

OR S 1497

STM 397

STM 384

LM G 150T

LM G 8819T2

LM G 8820

LM G VF39SM

LM G 11937T

LM G 6214T

LM G 18254

LM G 19141T

LM G 7837T

LM G 8310

LM G 6464

LM G 7834T

LM G 14919T

LM G 16580

LM G 6133T

LM G 6217T

LM G 15613

LM G 140T

LM G 187T

LM G 159

LM G 257

LM G 11875T

LM G 11876

LM G 9517

LM G 6125T

LM G 15767T

LM G 17150T

LM G 14107T

LM G 18932

LM G 10056

LM G 6465T

Cluster III

Agrobacterium rhizogenes

Cluster I

Cluster II

Rhizobium tropici IIb




Rhizobium etli

Rhizobium galegae

Rhizobium huautlense

Rhizobium mongolense





Sinorhizobium arboris

Sinorhizobium medicae

Sinorhizobium meliloti

Sinorhizobium fredii

Sinorhizobium kostiense



Agrobacterium rubi

Agrobacterium vitis



Rhizobium tropici IIa

Mesorhizobium loti

Mesorhizobium tianshanense

Mesorhizobium ciceri

Mesorhizobium mediterraneum

Mesorhizobium huakuii

Mesorhizobium amorphaeMesorhizobium plurifarium

Cluster IV

Cluster V

100

90

80

70

60504030

OR S 1414

OR S 1496

STM 358

OR S 1352T

OR S 625T

OR S 662

OR S 663

OR S 639

OR S 645T

OR S 668T

STM 279T

STM 246T

OR S 1402

OR S 1403

OR S 1419

OR S 609T

OR S 611

OR S 51

OR S 1009T

OR S 1755T

OR S 504

OR S 665T

OR S 669T

OR S 513

OR S 1351T

OR S 2644T

OR S 1353

OR S 2643

OR S 992T

OR S 995

OR S 651

OR S 664T

OR S 2640T

OR S 2738T

OR S 2754

OR S 1752T

STM 238

OR S 654

STM 370

OR S 1420

STM 391

OR S 1478

OR S 571T

OR S 1476

OR S 1474

OR S 1497

STM 397

STM 384

LM G 150T

LM G 8819T2

LM G 8820

LM G VF39SM

LM G 11937T

LM G 6214T

LM G 18254

LM G 19141T

LM G 7837T

LM G 8310

LM G 6464

LM G 7834T

LM G 14919T

LM G 16580

LM G 6133T

LM G 6217T

LM G 15613

LM G 140T

LM G 187T

LM G 159

LM G 257

LM G 11875T

LM G 11876

LM G 9517

LM G 6125T

LM G 15767T

LM G 17150T

LM G 14107T

LM G 18932

LM G 10056

LM G 6465T

Cluster III

Figure 1. Dendrogram based on the unweighted pair group method using average linkage (UPGMA) clustering of Dice correlationvalues (SD) of normalized and combined 16S ARDRA patterns of Tunisian isolates and reference strains using 7 enzymes (MspI, CfoI, Hinf I,RsaI, NdeII, AluI, TaqI).


seven restriction enzymes were combined and analyzedusing the GelCompar II version 4.2 software. The result-ing patterns comprised 3 to 5 bands (NdeII, RsaI, TaqI,AluI), 5 to 6 bands (HinfI, MspI) or 3 to 6 bands (CfoI).Altogether, the combined profiles resulted in around 38bands for cluster analysis. Reproducibility of the methodwas assessed on several reference species, and differentstrains of a given species shared the same profile. Theclusters were first delineated according to the similarityvalue between strains (data not shown). However, suchclusters grouped strains with a very low level of similarityand appeared very heterogeneous (50% internal similar-ity). Thus, in the light of the 16S rDNA sequences ob-tained later (see Fig. 3), we created ARDRA clusters in away that they contain strains belonging to the same ge-nus. On the other hand, the strains that we left in sepa-rate positions are single strains belonging to differentgenera. Results are shown as a dendrogram in Fig. 1 andrecorded in Table 1. New isolates grouped separatelyfrom reference strains and formed 8 main clusters (I toVIII). Cluster I (52.29% similarity level), heterogeneous,contains three new isolates. Cluster II (100% similaritylevel) contains three new isolates. Cluster III (89,95%similarity level) consists of three isolates. Cluster IV(75.20% similarity level) consists of two new isolates.Cluster V (65.15% similarity level) consists of three newisolates. Cluster VI (75.10% similarity level) is the largestcluster with six strains. Cluster VII (76.11% similaritylevel) consists of two new isolates. Cluster VIII (50.6%

similarity level), heterogeneous, includes four new iso-lates. Strains STM 379, STM 388, STM 392, ORS 1437,ORS 1473, ORS 1478, and ORS 1481 have separatepositions.

SDS-PAGE Analysis. The results of SDS-PAGEanalysis are presented as a dendrogram (Fig. 2 and re-corded in Table 1). As expected, none of the new isolatesgrouped with any of the reference strains. Only threegroups could be delineated at a similarity level of 90%:Group 1 (STM 379 and STM 368) consists of one strainfrom ARDRA Cluster VII and one strain with separateposition in ARDRA. Group 2 consists of ORS 1421,ORS 1422 (ARDRA Cluster VI), and ORS 1419 (ARDRACluster II). Group 3 contains ORS 1402 and ORS 1403(ARDRA Cluster II).

16S rDNA Sequencing. We performed DNAsequencing on almost all isolates (Table 1). The nearlyfull-length 16S rDNA sequences of 15 strains was de-termined. The results of analysis are shown as a phy-logenetic tree (Fig. 3). The studied strains are clearlydistinct from known LNB and are distributed close to thegenera Pseudomonas, Rhodopseudomonas, Bosea, Phyl-lobacterium, Starkeya, Sphingomonas, Paracraurococcus,and Microbacterium. In addition, partial 16S rDNA se-quencing was performed for 17 strains (Table 1). Closestmatches included Inquilinus, Ornithinicoccus, Agromyces,Bacillus, Ochrobactrum, and Paenibacillus.

STM 397

STM 384

ORS 2060T

ORS 1473

STM 377

STM 378

ORS 1422

STM 364

STM 385

ORS 1421

STM 678T

STM 368

ORS 1432

STM 388

STM 392

ORS 1437

ORS 1481

ORS 1417

ORS 1418

ORS 1472

ORS 1480

STM 379

LMG 21444T

Cluster VII

Cluster VIII

Cluster V

Cluster VI

Methylobacterium nodulans

Burkholderia tuberum

Figure 1. Continued


ITS DNA Sequencing. According to 16S rDNAanalyses, strains ORS 1414, ORS 1496 and STM 358belong to the Bradyrhizobiaceae family. We thus per-formed ITS DNA sequencing on two of these strains (ORS1414, STM 358) and compared the obtained sequenceswith all the Bradyrhizobium ITS sequences of our data-base and with those of Afipia, Rhodopseudomonas, Ni-trobacter, and Blastobacter. Sequences of ORS 1414 andSTM 358 grouped outside all of these and are the mostperipheral of the dendrogram (data not shown). On thebasis of these partial sequences it appears that the twostrains are not that close to any of these genera. More-over, the two strains, having an ITS sequence similarity

of only 59%, seem to belong to different species and may-be even to different genera.

Amplification and Sequencing of nifH Genes.

Eighteen strains (see Table 1) were tested for the pres-ence of nifH-like gene sequences by PCR amplification.Amplificates of the expected size (370 bp) were producedby only 10 strains. The PCR products of these strainswere sequenced and only seven sequences showed ho-mology to nifH sequences (as revealed by NCBI BLAST,Table 4). These sequences corresponded to isolates iden-tified in our study as Microbacterium sp., Agromyces sp.,Phyllobacterium sp., and Starkeya sp. The phylogenetic

Group 1

Group 2

Group 3

Sinorhizobium frediiRhizobium mongolenseRhizobium leguminosarum bv viciaeRhizobium etli

Rhizobium hainanenseBradyrhizobium japonicumRhizobium giardinii bv giardiniiMesorhizobium tianshanense

Rhizobium tropici IIbMesorhizobium plurifariumMesorhizobium plurifariumSinorhizobium terangaeSinorhizobium terangaeAzorhizobium caulinodansSinorhizobium saheliSinorhizobium saheliMesorhizobium lotiRhizobium galegaeSinorhizobium melilotiSinorhizobium medicaeRhizobium gallicum bv. phaseoliSinorhizobium adhaerensSinorhizobium morelenseSinorhizobium arborisSinorhizobium arborisAgrobacterium vitisRhizobium huautlenseRhizobium huautlense

Agrobacterium bv 1Allorhizobium undicola

Mesorhizobium chacoenseMesorhizobium mediterraneum

Mesorhizobium huakuiiMesorhizobium ciceriAgrobacterium rhizogenesMethylobacterium nodulans

LM G 6 2 1 7 TLM G 1 9 1 4 1LM G 8 8 1 7 TCF N 4 2 TS TM 3 9 2S TM 3 7 9S TM 3 6 8S TM 1 4 3 2ORS 1 4 1 7ORS 1 4 7 2ORS 1 4 1 8ORS 1 4 2 1ORS 1 4 2 2ORS 1 4 1 9S TM 3 7 8ORS 1 4 9 7LM G 1 8 0 7 5LM G 6 1 3 8 TR4 3 8 5LM G 1 5 7 6 7ORS 1 4 2 0LM G 9 5 0 3 TLM G 1 1 8 9 0LM G 1 1 8 9 2 TLM G 7 8 4 7LM G 7 8 3 4 TLM G 6 4 6 5 TLM G 7 8 4 2LM G 7 8 3 7 TLM G 6 1 2 5LM G 6 2 1 4 TLM G 6 1 3 3 TLM G 1 6 5 7 9R4 3 8 4R-1 4 0 6 5 TR-1 3 9 8 7 TLM G 1 5 6 2 4LM G 1 5 6 2 6LM G 2 5 7LM G 1 8 2 5 4LM G 1 8 2 5 6S TM 3 5 9LM G 1 4 0 TLM G 1 1 8 7 6ORS 1 4 0 3ORS 1 4 0 2ORS 1 4 7 8S TM 3 9 1ORS 1 4 7 6LM G1 7 1 4 8 TLM G 1 9 0 0 8LM G 1 4 1 0 7LM G 1 4 9 8 9LM G 1 5 0 TORS 2 0 6 0ORS 1 4 1 4ORS 1 4 7 4S TM 3 9 7ORS 1 4 8 0ORS 1 4 7 3

1 0 09 08 07 06 05 04 0

ORS 1432

Figure 2. Dendrogram showing the relation-ships between the electrophoretic proteinpatterns of Tunisian legume bacterial isolatesusing GelCompar 4.2 software package. Thedendrogram is based on mean correlationcoefficient (r) values, which were grouped bythe unweighted average pairs groupingmethod (UPGMA).


Figure 3. 16S rRNA genesequence-based dendrogramobtained by neighbor-joiningmethod, showing the phylogeneticpositions of Tunisian isolates.Significant bootstraps (980%) areindicated as percentages (1000 rep-lications). The tree is rooted onMicrobacterium kitamiense.


analysis (Fig. 4) showed that these nifH sequences are allclosely related to that of Sinorhizobium meliloti except forORS 1403 (Table 4). No amplification could be obtainedusing forward primer nifH(forB) with strains ORS 1403and ORS 1474. ORS 1403 reverse sequence [obtainedwith primer nifH(rev)] was 95% homologous with theSinorhizobium xinjiangense nifH gene, whereas the ORS

1474 reverse sequence showed 90% homology with the S.meliloti nifH gene (Table 4).

Nodulation Tests and nodA Gene Amplification.

Although R. raetam grows naturally on sandy soils, itgrew very well in Jensen tubes. The control plants (non-inoculated ones) showed chlorotic symptoms 6–8 weeks

Table 4. Results for nifH gene sequencing

Strain Genus Closest match in NCBIa

ORS 1417b Microbacterium S. meliloti nifH gene (94%, 329 bp)ORS 1418b Microbacterium S. meliloti nifH gene (93%, 338 bp)ORS 1472b Microbacterium S. meliloti nifH gene (96%, 320 bp)ORS 1437b Agromyces S. meliloti nifH gene (93%, 314 bp)ORS 1402b Phyllobacterium-like S. meliloti nifH gene (96%, 342 bp)ORS 1403c Phyllobacterium-like S. xinjiangense nifH gene (95%, 182 bp)ORS 1474c Starkeya S. meliloti nifH gene (90%, 346 bp)aBetween brackets are given, respectively, the percentage of homology with the most similar published 16S rDNA sequence and the number of base pairsconsidered.bResult of sequencing by two primers: forward and reverse.cResult of sequencing for one primer only.

Figure 4. Phylogenetic treebased on nearly full length (370bp) nifH sequence of some newisolates from Tunisia. Newsequences are indicated in bold.The significant bootstrap valuesare indicated as percentages(980%) derived from 1000replications. The tree is rootedon Frankia sp. strain FaCl.


after cultivation. Nodulation tests on M. atropurpureumand R. raetam were negative with all the strains tested.However, some strains (ORS 1414, STM 358, ORS 1419,ORS 1474, STM 397, STM 385, ORS 1432, and ORS 1481)had some effect on plant roots that became inflated. Theplants with inflated roots showed chlorotic symptomssimilar to control (N-free, noninoculated) plants. StrainORS 1472 induced a black color covering M. atropur-pureum and R. raetam roots. On the other hand, 15strains (see Table 1) were screened for nodA gene by PCR.No amplification could be obtained using primers de-signed from a Bradyrhizobium nodA gene alignment(Table 3).

Discussion

During isolation of bacteria from root nodules of spon-taneous legumes growing in the arid zone of Tunisia, weobtained 60 classical rhizobial strains [120] and 34 strainsbelonging to several genera not known to include anydescribed LNB species. Here we further analyzed the lat-ter 34 strains. The fact that they survived standard nod-ule surface sterilization indicates that these bacteria areprotected from chemicals, and suggests that they live in-side the nodule tissue. This is confirmed by the sterility ofthe water used for the last rinsing of the nodules, whichwas checked by inoculation on YMA agar plates and in-cubation. Microscopic observation of a stained semithinsection of a nodule is required to definitively clarify theirlocalization inside the nodule tissue. However, it was ofparticular interest to characterize these strains as severalnew LNB were recently found in genera not expectedto include LNB [22, 69, 70, 83, 94, 101]. This charac-terization could also provide more information on thebacterial communities associated with root nodules. Tocharacterize these new isolates we performed 16S ARDRA,SDS-PAGE of total cell proteins, 16S and ITS rDNA se-quencing. Both in ARDRA and SDS-PAGE analyses, thenew isolates appeared rather diverse and externally dis-tributed from the reference strains of known LNB. Somedegree of inconsistency was observed between resultsfrom SDS-PAGE and ARDRA; that is, SDS-PAGE group 1encompasses only one of the two strains of ARDRAClusters VII and one strain that grouped separately inARDRA; Group 2 includes one strain from ARDRACluster II and two from ARDRA Cluster VI.

The 16S rDNA sequencing results revealed that thesestrains belong to three different phyla: the Proteobacte-ria (Gram-negative), the Actinobacteria, and the Fir-micutes (Gram-positive). Within the Proteobacteria, 23isolates belong either to the a class (Bosea, Starkeya,Phyllobacterium, Sphingomonas, Inquilinus, Ochrobac-trum, Rhodopseudomonas) and two isolates to the g-class(Pseudomonas). Seven isolates grouped within the Actino-bacteria, in Microbacterium, Agromyces, Mycobacterium,

or Ornithinicoccus, all belonging to the order Actinomy-cetales. Two isolates grouped in the phylum Firmicutes,inside Bacillus and Paenibacillus.

In ARDRA, the Proteobacteria new strains weredistributed in eight clusters (I, II, III, IV, V, VI, VII,VIII) or had separate positions.

Together with the two other strains of ARDRACluster I, strain STM 358 belongs to the genus Boseabecause its 16S rDNA sequence is 99% similar to that ofBosea thiooxidans. The genus Bosea and B. thiooxidanswere described for the first time by Das et al. [27] for abacterial isolate from agricultural soil. Since then, thegenus was emended and now encompasses four species[59, 75]. Phylogenetically, this genus belongs to theBradyrhizobiaceae family. According to the standardsestablished in Bradyrhizobium [111], results from ITSsequence analysis suggest that Bosea strains from Tunisiarepresent different species because they share only 59%ITS sequence similarity. Unfortunately, no ITS sequencefrom Bosea is available in GenBank for comparison. Wecould not evidence nodulation capacity among thesestrains nor could we amplify any nodA gene.

Strains ORS 1420, STM 370, and STM 391 (ARDRACluster III) are phylogenetically close to Phyllobacteriummyrsinacearum on the basis of their 16S rDNA sequences(Table 1). The genus Phyllobacterium [57] was proposedto accommodate bacteria isolated from leaf nodules ofRubiaceae and Myrsinaceae tropical plants. However, thenodulation-inducing capacity of Phyllobacterium wasnever clearly demonstrated. Recently the two species ofthe genus, P. myrsinacearum and P. rubiacearum, werefused into P. myrsinacearum [68]. Some Phyllobacteriumstrains exhibit a PGP capacity [12]. Nitrogen-fixing ac-tivity and presence of nifHDK-like genes were so far notdemonstrated in Phyllobacterium [63], and, here, at-tempts to amplify nifH DNA within ORS 1420 andSTM 370 also failed. Plants inoculated with ORS 1420and STM 370 did not produce any root nodule but hadgreener foliage than the control (N-free, noninoculated)plants, suggesting a possible PGP effect. Phyllobacteriumstrains having been reported to develop inside leafnodules, we cannot exclude the hypothesis that theymay also develop in root nodules of compatible hostlegumes.

Strain ORS 1478, close to Ochrobactrum according toits partial 16S rDNA sequence, is the nearest neighbor ofPhyllobacterium sp. strains in our ARDRA results. Thegenera Phyllobacterium and Ochrobacterium are known tobe phylogenetically closely related to each other and tothe Rhizobiaceae [110]. Ochrobactrum strains were firstisolated from human clinical samples [50, 96] and soil[50]. Recently, Ngom et al. [70] reported a symbioticnitrogen-fixing strain belonging to the genus Ochrobac-trum for the first time. Ochrobactrum strains had neverbeen reported as nitrogen fixers before. Complete se-


quencing of 16S rDNA is required to elucidate the tax-onomic position of ORS 1478. We did not screen thisstrain for nifH gene.

Three strains, ORS 1402, ORS 1403, and ORS 1419,formed a distinct ARDRA cluster (Cluster II) betweenSinorhizobium and Rhizobium, suggesting a separategenus. Based on their 16S rDNA sequence, these strainshave P. myrsinacearum as their closest phylogeneticneighbor (96% sequence homology). DNA:DNA hybrid-izations are needed to confirm their taxonomic status.No nodulation capacity could be evidenced within thesestrains in our hands. The presence of an nifH-like gene(homologous to that of S. meliloti or S. xinjiangense) intheir genome (Table 4) suggests that strains ORS 1402and ORS 1403 could be nitrogen fixers.

Strains ORS 1474 and ORS 1476 (ARDRA ClusterIV) originate from R. raetam root nodules. The ORS1474 16S rDNA sequence is 98% homologous to that ofStarkeya novella, formerly Thiobacillus novellus [53],originally isolated from soil [92]. Starkeya was recentlycreated for T. novellus since it was the only species of thegenus Thiobacillus grouping in the a class of the Pro-teobacteria, the other species being members of the bProteobacteria [53]. Within the genus Starkeya, no nitro-gen fixing bacterium is known so far. Within Thioba-cillus, only the species T. ferrooxidans is known to be anitrogen-fixing bacterium [118]. In the strain ORS 1474,affiliated to Starkeya, we detected a nifH-like sequenceclose to that of S. meliloti (Table 4). To our knowledge,this is the first report of an nifH-like gene in the genusStarkeya. Attempts to amplify the forward sequence ofthis gene should be done again using suitable primers.

ARDRA Cluster V included Sphingomonas-like strains(STM 384, STM 397, ORS 1497). The genus Sphingomo-nas was created by Yabuuchi et al. [115] for the formerspecies Pseudomonas paucimobilis, which was referredto as a nitrogen fixer by Barraquio et al. [7]. Strains ofSphingomonas adhaesiva have been reported to be endo-phytic of rice (Oryza sativa) [41]. Here, we could notevidence nodulation capacity nor obtain nifH gene am-plification for both STM 384 and STM 397 strains.

Members of ARDRA Cluster VI (six isolates) belongto the a-Proteobacteria, two of them (ORS 1421 andORS 1422) forming a distinct group by SDS-PAGE anal-ysis. Partial 16S rDNA sequencing on all isolates of Clus-ter VI indicated their close relationship to the recentlydescribed genus Inquilinus, which encompasses clinicalstrains [26]. No reports concerning plant interactions ornitrogen fixation are known for this genus. Total 16SrDNA sequencing is required to elucidate the precisephylogenetic position of these isolates.

In our collection, ORS 1416ri is identified at thegenus level by nearly total 16S rDNA sequencing as aRhodopseudomonas sp. strain (Table 1). Rhodopseudomo-nas palustris has been reported to be endophytic of rice

(O. ridleyi) [41] and some Rhodopseudomonas species arenitrogen fixers [118].

The isolate ORS 1473 occupied a separate position inboth ARDRA and SDS-PAGE dendrograms. The total16S rDNA sequencing showed that the closest 16S rDNAsequence (98% homology) was that of Paracraurococcusruber (a-Proteobacteria), the unique species of the genusParacraurococcus that was created by Saitoh et al. [85] forsoil isolates containing bacteriochlorphyll a.

The ARDRA Cluster VII consisted of two Pseudo-monas-like strains, STM 368 and ORS 1432. The genusPseudomonas formerly consisted of phylogenetically un-related groups of proteobacteria including more than 100validly described species [76, 88]. This genus has un-dergone many changes in its classification, and the for-mer Pseudomonas species are now classified in about 15genera belonging to the a, b, and g subclasses of Pro-teobacteria [3, 55]. The species belonging to the genusPseudomonas sensu stricto are now members of the gsubclass of Proteobacteria [55]. Pseudomonas has beenreported to be endophytic of pea shoot tissues [42] orcolonizing bean root surface and possibly conferringprotection against disease development [2]. Other Pseu-domonas strains are endophytic of numerous nonlegumeplants [66], or clinical [76]. Recently, Benhizia et al. [9]reported Pseudomonas strains to be associated with le-gume nodules. On the other hand, nitrogen fixation hasbeen reported for the species Pseudomonas stutzeri [118].We could not amplify a nifH gene from the strain ORS1432.

The other new isolates were Gram-positive; sevenstrains belonged to Actinomycetales: ARDRA Cluster VIIIand three strains in separate positions (Fig. 1; Table 1). Thegenus Microbacterium has been emended recently tocombine the genera Microbacterium and Aureobacterium[95]. Some species of this genus have already been re-ported to be endophytes of agronomic and prairy plants[122]. The genus Agromyces was established by Gledhilland Casida [44] and includes filamentous bacteria,abundant in soils. This genus includes a species isolatedfrom Androsace plants of the family Primulaceae: Agro-myces albus [37]. The microscopic observation of thestrain ORS 1437 revealed no filamentous bacteria butrod-shaped ones. Total 16S rDNA sequencing is neededto identify this strain. No report of endophytic Agromycesstrains has been published so far. In our collection, thestrain ORS 1481 (separate position in ARDRA) is close tothe genus Mycobacterium according to its 16S rDNApartial sequence (Table 1). The genus Mycobacterium isknown to be the casual agents of two important diseases,tuberculosis and leprosy, but some Mycobacteriumspecies are free-living saprophytes found in salt water,soil, and dust, but could be endophytic of winter wheatroot [84]. The genus Ornithinicoccus has been created forbacteria isolated from a garden soil [45] and no report


on its endophytic ability has been published so far. Total16S rDNA sequencing should be performed to confirmthe affiliation of the strain STM 379 to this genus. Noneof these Actinomycetales strains was able to inducenodule formation on the plants tested. However, wenoticed that ORS 1472 induced a black pigment, whichcovered the whole root system of both M. atropurpureumand R. raetam plants. We previously reported [52] thatsome naturally occurring nodules of R. raetam wereblack. This suggests that contact with this bacteriumresults in this plant reaction. The endophytic associationbetween root nodules and Gram-positive bacteria hasbeen recently reported by Bai et al. [5].

Nitrogen fixation within the Actinomycetales hasbeen demonstrated for Frankia [113], Arthrobacter [16],Streptomyces [34, 56], and Proprionibacterium [6]. Al-though Rao [80] reported C2H2-reducing activity forsome Mycobacterium species, no further characterizationwas reported, and Young [118] does not mention anynitrogen-fixing activity within this genus. Ruppel [84]reported that Mycobacterium and Microbacterium hadsome low nitrogen-fixing activity, as measured by acet-ylene-reducing capacity. Here we found nifH-like genesequences within the actinobacteria Microbacterium andAgromyces-like (Table 4), and to our knowledge, this isthe first report for nifH within these genera. We couldnot amplify any nifH gene within the genera Ornithini-coccus and Mycobacterium. This may be due to the lack ofspecificity of the primers we used. For the genus Orni-thinicoccus, no report about nitrogen fixation has beenpublished so far.

The other Gram-positive strains belong to theFirmicutes phylum. Isolates STM 388 and STM 392 are,respectively, Paenibacillus and Bacillus strains (Table 1).For a long time, the genus Bacillus remained heteroge-neous. In recent years, its taxonomy has changed con-siderably, mostly as a result of phylogenetic informationderived from 16S rDNA sequences. There are now atleast seven other genera that contain former Bacillus spe-cies [11, 91]. One of these, Paenibacillus [4], as well asBacillus itself, include nitrogen fixers [4, 89].

The genus Bacillus comprises rice endophytes [93],PGP bacteria isolated from soybean root nodules [5] andlegume nodulation enhancers [48, 90]. Some Paeniba-cillus strains are root surface colonizers [10] and PGPbacteria [98].

In our collection of endophytic root nodule bacteria,we found strains belonging to the genera Phyllobacte-rium, Sphingomonas, Rhodopseudomonas, Pseudomonas,Microbacterium, Mycobacterium, Bacillus, and Paenibacil-lus. Members of these genera were reported in the lit-erature to interact at different levels with plants. Somecolonize the root surface (Paenibacillus), others arefound inside nodules (Phyllobacterium, Ochrobactrum,Pseudomonas, Bacillus), and others are root endophytic

bacteria (Sphingomonas, Rhodopseudomonas, Pseudomo-nas, Microbacterium, Mycobacterium). The strains studieddid not nodulate M. atropurpureum and we could notevidence any nodA-like gene sequence by PCR amplifi-cation with the primers we used. Probably more primersshould be tested for further nodulation gene screening.

The 16S-ARDRA technique allowed good separationof the studied strains at the genus level but 16S rDNAsequencing was essential for phylogenetic affiliation. Inthis study, sequences homologous to nifH gene of S.meliloti or S. xinjiangense, which are LNB, were found forthe first time within strains belonging to the generaMicrobacterium and Starkeya and within strains belong-ing to genera (very) close to Agromyces and Phyllobac-terium. This point is important because if the studiedisolates had recovered the nifH gene by lateral gene trans-fer during evolution, they might have recovered nod-ulation genes, as these genes are usually transferredtogether. This may also be the first report for nifH-likesequences within the genus Phyllobacterium, if the strainsORS 1402 and ORS 1403 are proven to belong to thePhyllobacterium genus (this study is in preparation).

Acknowledgments

The authors are grateful to Sophie Mantelin, LionelMoulin, and Gilles Bena for helpful discussions and toOdile Domergue, Lucette Maure, Emmanuel Protiere,and Cathy Vadala for technical assistance. FZ is indebtedto IRD France for a PhD fellowship, AW and MG to theFund for Scientific Research (Flanders) for a postdoctoralresearch fellow position and research and personnelgrants, respectively.

References

1. Amarger, N, Macheret, V, Laguerre, G (1997) Rhizobium gallicumsp. nov. and Rhizobium giardinii sp. nov., from Phaseolus vulgarisnodules. Int J Syst Bacteriol 47(4): 996–1006

2. Anderson, AJ, Guerra, D (1985) Responses of bean to rootcolonization with Pseudomonas putida in a hydroponic system.Phytopathology 75(9): 992–995

3. Anzai, Y, Kim, H, Park, JY, Wakabayashi, H, Oyaizu, H (2000)Phylogenetic affiliation of the pseudomonads based on 16S rRNAsequence. Int J Syst Evol Microbiol 50: 1563–1589

4. Ash, C, Priest, FG, Collins, MD (1993) Molecular identificationof rRNA group 3 bacilli (Ash, Farrow, Wallbanks, and Collins)using a PCR probe test. Proposal for the creation of a new genusPaenibacillus. Antonie van Leeuwenhoek 64: 253–260

5. Bai, YM, Daoust, F, Smith, DL, Driscoll, BT (2002) Isolation ofplant-growth-promoting Bacillus strains from soybean rootnodules. Can J Microbiol 48(3): 230–238

6. Baranova, NA, Gogotov, IN (1974) Fixation of molecular nitrogenby propionic bacteria. Mikrobiologiya 43: 791–794 (in Russian)

7. Barraquio, WL, Dumont, A, Knowles, R (1988) A method forcounting free-living aerobic N2-fixing H2-oxidizing bacteria. In:


Bothe, H, Debrujn, FJ, Newton, WE (Eds.) Nitrogen Fixation:Hundred Years After. Gustav Fischer, Stuttgart, pp 790

8. Bartl, S, Weissman, IL (1994) PCR primers containing an inosinetriplet to complement a variable codon within a conservedprotein-coding region. BioTechniques 16: 246–250

9. Benhizia, Y, Benhizia, H, Benguedouar, A, Muresu, R, Giacomini,A, Squartini, A (2004) Gamma proteobacteria can nodulatelegumes of the genus Hedysarum. Syst Appl Microbiol 27(4):462–468

10. Bent, E, Breuil, C, Enebak, S, Chanway, CP (2002) Surfacecolonization of lodgepole pine (Pinus contorta var. latifolia[Dougl. Engelm.]) roots by Pseudomonas fluorescens and Paeni-bacillus polymyxa under gnotobiotic conditions. Plant Soil241(2): 187–196

11. Berkeley, RCW (2002) Whither Bacillus? In: Berkeley, R,Heyndrickx, M, Logan, N, De Vos, P (Eds.) Applications andSystematics of Bacillus and Relatives. Blackwell, Oxford, pp 1–7

12. Bertrand, H, Nalin, R, Bally, R, Cleyet-Marel, JC (2000) Isolationand identification of the most efficient plant growth-promotingbacteria associated with canola (Brassica napus). Biol Fertil Soils33: 152–156

13. Beyer, W, Glockner, P, Otto, J, Boehm, R (1995) A nested PCRmethod for the detection of Bacillus anthracis in environmentalsamples collected from former tannery sites. Microbiol Res 150:179–186

14. Biswas, JC, Ladha, JK, Dazzo, FB, Yanni, YG, Rolfe, BG (2000)Rhizobial inoculation influences seedling vigor and yield of rice.Agron J 92(5): 880–886

15. Boddey, RM, De Oliviera, OC, Urquiaga, S, Reis, VM, DeOlivares, FL, Baldani, VLD, Doebereiner, J (1995) Biologicalnitrogen fixation associated with sugarcane and rice: contribu-tions and prospects for improvement. Plant Soil 174: 195–209

16. Cacciari, I, Giovannozi-Sermanni, G, Grapelli, A, Lippi, D (1971)nitrogen fixation by Arthrobacter species. Ann Microbiol Enzimol21: 97–105

17. Candrian, U, Furrer, B, Hofelein, C, Luthy, J (1991) Use of ino-sine-containing oligonucleotide primers for enzymatic amplifica-tion of different alleles of the gene coding for heat-stable toxintype I of enterotoxigenic Escherichia coli. Appl Environ Microbiol57: 955–961

18. Carlotti, A, Funke, G (1994) Rapid distinction of Brevibacteriumspecies by restriction analysis of rDNA generated by polymerasechain reaction. Syst Appl Microbiol 17: 380–386

19. Casida, LE Jr (1982) BEnsifer adhaerens^ gen. nov., sp. nov.: a bac-terial predator of bacteria in soil. Int J Syst Bacteriol 32: 339–345

20. Cato, EP, Hash, DE, Holdeman, LV, Moore, WEC (1982)Electrophoretic study of Clostridium species. J Clin Microbiol15: 688–702

21. Chaintreuil, C, Giraud, E, Prin, Y, Lorquin, J, Ba, A, Gillis, M, deLajudie, P, Dreyfus, B (2000) Photosynthetic bradyrhizobia arenatural endophytes of the African wild rice Oryza breviligulata.Appl Environ Microbiol 66(12): 5437–5447

22. Chen, WM, Laevens, S, Lee, TM, Coenye, T, De Vos, P, Mergeay,M, Vandamme, P (2001) Ralstonia taiwanensis sp. nov., isolatedfrom root nodules of Mimosa species and sputum of a cysticfibrosis patient. Int J Syst Evol Microbiol 51: 1729–1735

23. Chen, WX, Li, GS, Qi, YL, Wang, ET, Yuan, HL, Li, JL (1991)Rhizobium huakuii sp. nov. isolated from the root nodules ofAstragalus sinicus. Int J Syst Bacteriol 41(2): 275–280

24. Chen, W, Wang, E, Wang, S, Li, Y, Chen, X, Li, Y (1995)Characteristics of Rhizobium tianshanense sp. nov., a moderatelyand slowly growing root nodule bacterium isolated from an aridsaline environment in Xinjiang, People’s Republic of China. Int JSyst Bacteriol 45(1): 153 –159

25. Coenye, T, Vandamme, P (2003) Diversity and significance of

Burkholderia species occupying diverse ecological niches. EnvironMicrobiol 5(9): 719 –729

26. Coenye, T, Goris, J, Spilker, T, Vandamme, P, Lipuma, JJ (2002)Characterization of unusual bacteria isolated from respiratorysecretions of cystic fibrosis patients and description of Inquilinuslimosus gen. nov., sp. nov. J Clin Microbiol 40(6): 2062–2069

27. Das, SK, Mishra, AK, Tindall, BJ, Rainey, FA, Stackebrandt, E(1996) Oxidation of thiosulfate by a new bacterium, Boseathiooxidans (strain BI-42) gen. nov., sp. nov.: analysis ofphylogeny based on chemotaxonomy and 16S ribosomal DNAsequencing. Int J Syst Bacteriol 46(4): 981–987

28. Date, RA (1982) Collection, isolation, characterization andconservation of Rhizobium. In: Vincent, JM (Eds.) NitrogenFixation in Legumes. Academic, New York, pp 95 –109

29. Dazzo, FB, Joseph, AR, Gomaa, AB, Yanni, YG, Robertson, GP(2003) Quantitative indices for the autecological biogeography ofa Rhizobium endophyte of rice at macro and micro spatial scales.Symbiosis 35(1–3): 147–158

30. de Lajudie, P, Laurent-Fulele, E, Willems, A, Torck, U, Coopman,R, Collins, MD, Kersters, K, Dreyfus, B, Gillis, M (1998)Allorhizobium undicola gen. nov., sp. nov., nitrogen-fixingbacteria that efficiently nodulate Neptunia natans in Senegal. IntJ Syst Bacteriol 48: 1277–1290

31. de Lajudie, P, Willems, A, Nick, G, Mohamed, SH, Torck, U,Coopman, R, Filali-Matlouf, A, Kersters, K, Dreyfus, B, Lindstrom, K,Gillis, M (1999) Agrobacterium bv. 1 strains isolated from nodules oftropical legumes. Syst Appl Microbiol 22: 119–132

32. de Lajudie, P, Willems, A, Nick, G, Moreira, F, Molouba, F, Hoste,B, Torck, U, Neyra, M, Collins, MT, Lindstrom, K, Dreyfus, B,Gillis, M (1998) Characterization of tropical tree rhizobia anddescription of Mesorhizobium plurifarium sp. nov. Int J SystBacteriol 48: 369–382

33. de Lajudie, P, Willems, A, Pot, B, Dewettinck, D, Maestrojuan, G,Neyra, M, Collins, MT, Dreyfus, B, Kersters, K, Gillis, M (1994)Polyphasic taxonomy of Rhizobia: emendation of the genusSinorhizobium and description of Sinorhizobium meliloti comb.nov. Sinorhizobium saheli sp. nov., and Sinorhizobium teranga sp.nov. Int J Syst Bacteriol 44(4): 715 –733

34. Ding, J, Sun, H, Su, F, Xu, Q, Huang, Y, Ling, P (1981) Studieson nitrogen fixation by actinomycetes. Acta Microbiol Sin 21:424–427

35. Dommergues, Y, Duhoux, E, Diem, HG (1999) Les arbres fix-ateurs d’azote: caracteristiques fondamentales et role dansl’amenagement des ecosystemes mediterraneens et tropicaux.Editions espaces 34 (CIRAD, FAO, IRD). 499 pp.

36. Don, RH, Cox, PT, Wainwright, BJ, Baker, K, Mattick, JS (1991)Touchdown PCR to circumvent spurious priming during geneamplification. Nucleic Acids Res 19(14): 4008

37. Dorofeeva, LV, Krausova, VI, Evtushenko, LI, Tieje, JM (2003)Agromyces albus sp. nov., isolated from a plant (Androsace sp.).Int J Syst Evol Microbiol 53: 1435 –1438

38. Dreyfus, B, Garcia, JL, Gillis, M (1988) Characterization ofAzorhizobium caulinodans gen. nov., sp. nov., a stem nodulatingnitrogen fixing bacterium isolated from Sesbania rostrata. Int JSyst Bacteriol 38(1): 89 – 98

39. Eardly, BD, Materon, LA, Smith, NH, Johnson, DA, Rumbaugh,MD, Selander, RK (1990) Genetic structure of natural popula-tions of the nitrogen fixing bacterium Rhizobium meliloti. ApplEnviron Microbiol 56(1): 187–194

40. El Fantroussi, S, Mahillon, J, Naveau, H, Agathos, SN (1997)Introduction of anaerobic dechlorinating bacteria into soil slurrymicrocosms and nested PCR monitoring. Appl Environ Microbiol63: 806 – 811

41. Elbeltagy, A, Nishioka, K, Suzuki, H, Sato, T, Sato, YI, Morisaki,H, Mitsui, H, Minamisawa, K (2000) Isolation and characteriza-


tion of endophytic bacteria from wild and traditionally cultivatedrice varieties. Soil Sci Plant Nutr 46(3): 617– 629

42. Elvira-Recuenco, M, Van Vuurde, JWL (2000) Natural incidenceof endophytic bacteria in pea cultivars under field conditions.Can J Microbiol 46(11): 1036 –1041

43. Estrada, P, Mavingui, P, Cournoyer, B, Fontaine, F, Balandreau,J, Caballero-Mellado, J (2002) A N-2-fixing endophytic Burkhol-deria sp associated with maize plants cultivated in Mexico. Can JMicrobiol 48(4): 285 –294

44. Gledhill, WE, Casida, LE, Jr (1969) Predominant catalase-negative soil bacteria. III. Agromyces, gen. nov., microorganismsintermediary to Actinomyces and Nocardia. Appl Microbiol 18:340 –349

45. Groth, I, Schumann, P, Martin, K, Schuetze, B, Augsten, K,Kramer, I, Stackebrandt, E (1999) Ornithinicoccus hortensis gen.nov., sp. nov., a soil actinomycete which contains L-ornithine. IntJ Syst Bacteriol 49: 1717–1724

46. Gutierrez-Zamora, ML, Martinez-Romero, E (2001) Naturalendophytic association between Rhizobium etli and maize (Zeamays L.). J Biotechnol 91(2–3): 117–126

47. Gyaneshwar, P, James, EK, Mathan, N, Reddy, PM, Reinhold-Hurek, B, Ladha, JK (2001) Endophytic colonization of rice by adiazotrophic strain of Serratia marcescens. J Bacteriol 183(8):2634–2645

48. Halverson, LJ, Handelsman, J (1991) Enhancement of soybeannodulation by Bacillus cereus UW 85 in the field and in a growthchamber. Appl Environ Microbiol 57(9): 2767–2770

49. Heyndrickx, M, Vauterin, L, Vandamme, P, Kersters, K, Devos, P(1996) Applicability of combined amplified ribisomal DNArestriction analysis (ARDRA) patterns in bacterial phylogenyand taxonomy. J Microbiol Methods 26: 247–259

50. Holmes, B, Popoff, M, Kiredjian, M, Kersters, K (1988)Ochrobactrum anthropi gen. nov. sp. nov. from human clinicalspecimens and previously known as group Vd. Int J Syst Bacteriol38(4): 406 – 416

51. Jarvis, BDW, Gillis, M, De Ley, J (1986) Intra and intergenericsimilarities between the ribosomal ribonucleic acid cistrons ofRhizobium and Bradyrhizobium species and some related bacteria.Int J Syst Bacteriol 36(2): 129 –138

52. Jeder, H, de Lajudie, P, Dreyfus, B, Le Floch’s, E, Behaeghe, T,Zaafouri, MS, Akrimi, N (1996) Nodulation des legumineusespastorales. Etude de la nodulation des legumineuses pastoralesautochtones des regions arides de Tunisie. 9(1196): 3 –10

53. Kelly, DP, Mc Donald, IR, Wood, AP (2000) Proposal for thereclassification of Thiobacillus novellus as Starkeya novella gen.nov., comb. nov., in the alpha-subclass of the Proteobacteria. IntJ Syst Evol Microbiol 50: 1797–1802

54. Kersters, K, De Ley, J, Sneath, PHA, Sackin, M (1973) Numericaltaxonomic analysis of Agrobacterium. J Gen Microbiol 78: 227–239

55. Kersters, K, Ludwig, W, Vancanneyt, M, Devos, P, Gillis, M,Schleifer, KH (1996) Recent changes in the classification of thepseudomonads: an overview. Syst Appl Microbiol 19: 465 – 477

56. Knapp, R, Jurtshuk, P (1988) Characterization of free-livingnitrogen-fixing Streptomyces species and factors which affect theirrates of acetylene reduction. Abstr Annu Meet Am Soc Microbiol88(219)

57. Knosel, D (1984) The genus Phyllobacterium. In: Krieg, NR,Holt, JG (Eds.) Bergey’s Manual of Systematic Bacteriology, Vol1. Williams & Wilkins, Baltimore, pp 254 –256

58. Kobayashi, DY, Palumbo, JD (2000) Bacterial endophytes andtheir effects on plants and uses in agriculture. In: Bacon, CW,White, JF (Eds.) Microbial Endophytes. Marcel Dekker, NewYork, pp 199–223

59. La Scola, B, Mallet, MN, Grimont, PAD, Raoult, D (2003) Boseaeneae sp. nov., Bosea massiliensis sp. nov. and Bosea vestrisii sp. nov.,

isolated from hospital water supplies, and emendation of the genusBosea (Das et al. 1996). Int J Syst Evol Microbiol 53(1): 15–20

60. Ladha, JK, Reddy, PM (2000) Steps toward nitrogen fixation in rice.In: Ladha, JK, Reddy, PM (Eds.) The Quest for Nitrogen Fixationin Rice. International Rice Research Institute, Manila, pp 33–46

61. Laemmli, UK (1970) Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227: 680 – 685

62. Laguerre, G, Allard, MR, Revoy, F, Amarger, N (1994) Rapididentification of Rhizobia by restriction fragment length poly-morphism analysis of PCR amplified 16S rRNA genes. ApplEnviron Microbiol 60(1): 56 – 63

63. Lambert, B, Joos, H, Dierickx, S, Vantomme, R, Swings, J,Kersters, K, Vanmontagu, M (1990) Identification and plant in-teraction of a Phyllobacterium sp., a predominant rhizobacteriumof young sugar beet plants. Appl Environ Microbiol 56(4): 1093 –1102

64. Liebecq, C (Ed.) (1992) International Union of Pure and AppliedChemistry and International Union of Biochemistry and Molec-ular Biology Biochemical Nomenclature and Related Documents,2nd ed., Portland Press, London

65. Lindstrom, K (1989) Rhizobium galegae, a new species of legumeroot nodule bacteria. Int J Syst Bacteriol 39(3): 365 –367

66. Lodewyckx, C, Vangronsveld, J, Porteous, F, Moore, ERB, Taghavi,S , Mezgeay, M, Vanderlelie, D (2002) Endophytic bacteria andtheir potential applications. Crit Rev Plant Sci 21(6): 583–606

67. Martınez-Romero, E, Segovia, L, Martins-Mercante, F, Franco,AA, Graham, P, Pardo, MA (1991) Rhizobium tropici, a novelspecies nodulating Phaseolus vulgaris L. beans and Leucaena sp.trees. Int J Syst Bacteriol 41: 417– 426

68. Mergaert, J, Cnockaert, MC, Swings, J (2002) Phyllobacteriummyrsinacearum (subjective synonym Phyllobacterium rubiacea-rum) emend. Int J Syst Evol Microbiol 52: 1821–1823

69. Moulin, L, Munive, A, Dreyfus, B, Boivin-Masson, C (2001)Nodulation of legumes by members of the b subclass ofProteobacteria. Nature 411: 948 –950

70. Ngom, A, Nakagawa, Y, Sawada, H, Tsukahara, J, Wakabayashi, S,Uchiumi, T, Nuntagij, A, Kotepong, S, Suzuki, A, Higashi, S,Abe, M (2004) A novel symbiotic nitrogen-fixing member of theOchrobactrum clade isolated from root nodules of Acacia man-gium. J Gen Appl Microbiol 50(1): 17–27

71. Nick, G, De Lajudie, P, Eardly, BD, Suomalainen, S, Paulin, L,Zhang, X, Gillis, M, Lindstrom, K (1999) Sinorhizobium arborissp. nov. and Sinorhizobium kostiense sp. nov. isolated fromleguminous trees in Sudan and Kenya. Int J Syst Bacteriol 49:1359 –1368

72. Normand, P, Cournoyer, B, Simonet, P, Nazaret, S (1992)Analysis of a ribosomal RNA operon in the actinomycete Frankia.Gene 111: 119–124

73. Nour, SM, Cleyet-Marel, JC, Normand, P, Fernandez, MP (1995)Genomic heterogeneity of strains nodulating chickpeas (Cicerarietinum L.) and description of Rhizobium mediterraneum sp.nov. Int J Syst Bacteriol 45(4): 640 – 648

74. Nour, SM, Fernandez, MP, Normand, P, Cleyet-Marel, JC (1994)Rhizobium ciceri sp. nov., consisting of strains that nodulate chick-peas (Cicer arietinum L.). Int J Syst Bacteriol 44(3): 511–522

75. Ouattara, AS, Essokazi, AA, Thierry, S, Cayol, JL, Labat, M,Monroy, O, Macarie, H (2003) Bosea minatitlanensis sp. nov., astrictly aerobic bacterium isolated from anaerobic digester sludge.Int J Syst Evol Microbiol 53: 1247–1251

76. Palleroni, NJ (1984) Genus I. Pseudomonas Migula 1894. In:Krieg, NR, Holt, JG (Eds.) Bergey’s Manual of SystematicBacteriology, Vol. 1. Williams & Wilkins, Baltimore, pp 141–199

77. Pot, B, Gillis, M, Hoste, B, Van De Velde, A, Bekaert, F, Kersters,K, De Ley, J (1989) Intra- and intergeneric relationships of thegenus Oceanospirillum. Int J Syst Bacteriol 39: 23 –34


78. Pot, B, Vandamme, P, Kersters, K (1994) Analysis of electropho-retic whole organism protein fingerprints. In: Goodfellow, M,O’Donnell, AG (Eds.) Modern Microbial Methods. ChemicalMethods in Prokaryotic Systematics. Wiley, Chichester, pp 493 –521

79. Ranjard, L, Brothier, E, Nazaret, S (2000) Sequencing bands ofribosomal intergenic spacer analysis fingerprints for characteriza-tion and microscale distribution of soil bacterium populationsresponding to mercury spiking. Appl Environ Microbiol 66(12):5334 –5339

80. Rao, VR (1973) Non-symbiotic nitrogen fixation in paddy fields.Doctoral thesis, Academy of Sciences USSR, Moscow. In: SoilMicroorganisms and Plant Growth. Subba Rao, 1977. MohanPrimlani for Oxford & IBH Publishing, New Delhi, pp. 82– 97

81. Reinhold-Hurek, B, Hurek, T, Gillis, M, Hoste, B, Vancanneyt, M,Kersters, K, De Ley, J (1993) Azoarcus, gen. nov., nitrogen-fixingProteobacteria associated with roots of Kallar grass (Leptochloafusca (L.) Kunth), and description of two species, Azoarcus indigenssp. nov. and Azoarcus communis sp. nov. Int J Syst Bacteriol 43:574–584

82. Riggs, PJ, Chelius, MK, Iniguez, AL, Kaeppler, SM, Triplett, EW(2001) Enhanced maize productivity by inoculation with diazo-trophic bacteria. Aust J Plant Physiol 28(9): 829 – 836

83. Rivas, R, Velazquez, E, Willems, A, Vizcaino, N, Subbarao, NS,Mateos, PF, Gillis, M, Dazzo, FB, Martınez-Molina, E (2002) Anew species of Devosia that forms a unique nitrogen-fixing root-nodule symbiosis with the aquatic legume Neptunia natans (L.F.)Druce. Appl Environ Microbiol 68(11): 5217– 5222

84. Ruppel, S (1989) Isolation and characterization of dinitrogen-fixing bacteria from the rhizosphere of Triticum aestivum andAmmophila arenaria. In: Vancura, V, Kunc, F (Eds.) Interrela-tionships Between Microorganisms and Plants in Soil. Proceed-ings of an International Symposium, Liblice, Prague, pp. 253 –262

85. Saitoh, S, Suzuki, T, Nishimura, Y (1998) Proposal of Craur-ococcus roseus gen. nov., sp. nov. and Paracraurococcus ruber gen.nov., sp. nov., novel aerobic bacteriochlorophyll a-containingbacteria from soil. Int J Syst Bacteriol 48: 1043–1047

86. Saitou, N, Nei, M (1987) The neighbor joining method: a newmethod for reconstructing phylogenetic trees. Mol Biol Evol 4(4):406 – 425

87. Segovia, L, Young, PW, Martınez-Romero, E (1993) Reclassifica-tion of American Rhizobium leguminosarum biovar phaseoli typeI strains as Rhizobium etli sp. nov. Int J Syst Bacteriol 43(2): 374 –377

88. Skerman, VBD, Mc Gowan, V, Sneath, PHA (1989) ApprovedLists of Bacterial names. American Society for Microbiology,Washington, DC

89. Sneath, PHA (Ed.) Bergey’s Manual of Systematic Bacteriology,Vol. 2, Williams & Wilkins, Baltimore

90. Srinivasan, M, Petersen, DJ, Holl, FB (1997) Nodulation ofPhaseolus vulgaris by Rhizobium etli is enhanced by the presenceof Bacillus. Can J Microbiol 43(1): 1– 8

91. Stackebrandt, E, Swiderski, J (2002) From phylogeny to system-atics: the dissection of the genus Bacillus. In: Berkeley, R,Heyndrickx, M, Logan, N, De Vos, P (Eds.) Applications andSystematics of Bacillus and Relatives. Blackwell, Oxford, pp 8 –22

92. Starkey, RL (1935) Isolation of some bacteria which oxidizethiosulfate. Soil Science 39: 197–219

93. Stoltzfus, JR, So, R, Malarvithi, PP, Ladha, JK, de Bruijn, FJ(1997) Isolation of endophytic bacteria from rice and assessmentof their potential for supplying rice with biologically fixednitrogen. Plant Soil 194(1–2): 25 –36

94. Sy, A, Giraud, E, Jourand, P, Garcia, N, Willems, A, de Lajudie, P,Prin, Y, Neyra, M, Gillis, M, Boivin-Masson, C, Dreyfus, B (2001)Methylotrophic Methylobacterium bacteria nodulate and fix

nitrogen in symbiosis with legumes. J Bacteriol 183: 214 –22095. Takeuchi, M, Hatano, K (1998) Union of the genera Mi-

crobacterium Orla-Jensen and Aureobacterium Collins et al., ina redefined genus Microbacterium. Int J Syst Bacteriol 48: 739 –747

96. Tatum, HW, Ewing, WH, Weaver, RE (1974) MiscellaneousGram-negative bacteria. In: Lennette, EH, Spaulding, EH, Truant,JP (Eds.) Manual of Clinical Microbiology, 2nd ed. AmericanSociety for Microbiology, Washington, DC, pp 270 –294

97. Terefework, Z, Nick, G, Suomalainen, S, Paulin, L, Lindstrom, K(1998) Phylogeny of Rhizobium galegae with respect to otherrhizobia and agrobacteria. Int J Syst Bacteriol 48: 349 –356

98. Timmusk, S, Wagner, EGH (1999) The plant growth promotingrhizobacterium Paenibacillus polymyxa induces changes in Arabi-dopsis thaliana gene expression: a possible connection betweenbiotic and abiotic stress responses. MPMI 12(11): 951– 959

99. Torsvik, V, Goksoyr, J, Daae, FL (1990) High diversity in DNA ofsoil bacteria. Appl Environ Microbiol 56(3): 782 –787

100. Ueda, T, Suga, Y, Yahiro, N, Matsuguchi, T (1995) RemarkableN2-fixing bacterial diversity detected in rice roots by molecularevolutionary analysis of nifH gene sequences. J Bacteriol 177:1414 –1417

101. Van Berkum, P, Eardly, BD (2002) The aquatic budding bac-terium Blastobacter denitrificans is a nitrogen-fixing symbiont ofAeschynomene indica. Appl Environ Microbiol 68(3): 1132 –1136

102. Van Berkum, P, Beyene, D, Bao, GP, Campbell, TA, Eardly, BD(1998) Rhizobium mongolense sp. nov. is one of three rhizobialgenotypes identified which nodulate and form nitrogen-fixingsymbioses with Medicago ruthenica [(L.) Ledebour]. Int J SystBacteriol 48(1): 13 –22

103. Vandamme, P, Pot, B, Gillis, M, Devos, P, Kersters, K, Swings, J(1996) Polyphasic taxonomy, a consensus approach to bacterialsystematics. Microbiol Rev 60(2): 407– 438

104. Vauterin, L, Vauterin, P (1992) Computer-aided objectivecomparison of electrophoresis patterns for grouping and identi-fication of microorganisms. Eur Microbiol 33: 633 – 641

105. Velazquez, E, Igual, JM, Willems, A, Fernandez, MP, Munoz, E,Mateos, PF, Abril, A, Toro, N, Normand, P, Cervantes, E, Gillis, M,Martınez-Molina, E (2001) Mesorhizobium chacoense sp nov., anovel species that nodulates Prosopis alba in the Chaco Aridoregion (Argentina). Int J Syst Evol Microbiol 51: 1011–1021

106. Vincent, JM (1970) A manual for the practical study of root-nodule bacteria. In: International Biological Programme Handbookno. 15. Oxford, Blackwell Scientific Publications Ltd., pp 73–97

107. Wang, ET, Tan, ZY, Willems, A, Fernandez-Lopez, M, Reinhold-Hurek, B, Martınez-Romero, E (2002) Sinorhizobium morelensesp nov., a Leucaena leucocephala-associated bacterium that ishighly resistant to multiple antibiotics. Int J Syst Evol Microbiol52: 1687–1693

108. Wang, ET, Van Berkum, P, Beyene, D, Sui, XH, Dorado, O, Chen,WX, Martınez-Romero, E (1998) Rhizobium huautlense sp. nov., asymbiont of Sesbania herbacea that has a close phylogenetic rela-tionship with Rhizobium galegae. Int J Syst Bacteriol 48: 687–699

109. Widmer, D, Shaffer, B, Porteous, L, Seidler, R (1999) Analysis ofnifH gene pool complexity in soil and litter at a Douglas fir forestsite in the Oregon cascade mountain range. Appl EnvironMicrobiol 65: 374 –380

110. Willems, A, Collins, MD (1993) Phylogenetic analysis of Rhizobiaand agrobacteria based on 16S rRNA gene sequences. Int J SystBacteriol 43(2): 305 –313

111. Willems, A, Coopman, R, Gillis, M (2001) Comparison ofsequence analysis of 16S –23S rDNA spacer regions, AFLPanalysis and DNA–DNA hybridizations in Bradyrhizobium. Int JSyst Evol Microbiol 51: 623 – 632

112. Willems, A, Fernandez-Lopez, M, Munoz-Adelantado, E, Goris, J,


De Vidaver, AK, Martınez-Romero, E, Toro, N, Gillis, M (2003)Description of new Ensifer strains from nodules and proposal totransfer Ensifer adhaerens Casida 1982 to Sinorhizobium asSinorhizobium adhaerens comb. nov. Request for an opinion. IntJ Syst Evol Microbiol 53: 1207–1217

113. Williams, ST (Ed.) (1989) Bergey’s Manual of SystematicBacteriology, vol. 4, Williams & Wilkins, Baltimoret

114. Woese, CR (1987) Bacterial evolution. Microbiol Rev 51(2): 221–271

115. Yabuuchi, E, Yano, I, Oyaizu, H, Hashimoto, Y, Esaki, T,Yamamoto, H (1990) Proposals of Sphingomonas paucimobilisgen. nov. and comb. nov., Sphingomonas parapaucimobilis sp.nov., Sphingomonas capsulata comb. nov., and two genospecies ofthe genus Sphingomonas. Microbiol Immunol 34: 99 –119

116. Yanni, YG, Rizk, RY, Abdelfattah, FK, Squartini, A, Corich, V,Giacomini, A, De Bruijn, F, Rademaker, J, Mayaflores, J, Ostrom,P, Vegahernandez, M, Hollingsworth, RI, Martınez-Molina, E,Mateos, P, Velazquez, E, Wopereis, J, Triplett, E, Umali-Garcia,M, Anarna, JA, Rolfe, BG, Ladha, JK, Hill, J, Mujoo, R, Ng, PK,Dazzo, FB (2001) The beneficial plant growth-promoting associ-ation of Rhizobium leguminosarum bv. trifolii with rice roots. AustJ Plant Physiol 28(9): 845 – 870

117. Yanni, YG, Rizk, RY, Corich, V, Squartini, A, Ninke, K, Philip-

Hollingsworth, S, Orgambide, G, De Bruijn, F, Stoltzfus, J,Buckley, D, Schmidt, TM, Mateos, PF, Ladha, JK, Dazzo, FB(1997) Natural endophytic association between Rhizobium legu-minosarum bv. trifolii and rice roots and assessment of itspotential to promote rice growth. Plant Soil 194(1–2): 99 –114

118. Young, JPW (1992) Phylogenetic classification of nitrogen fixingorganisms. In: Stacey, G, Burris, RH, Evans, HJ (Eds.) BiologicalNitrogen Fixation. Chapman & Hall, pp 43 – 86

119. Zakhia, F, de Lajudie, P (2001) Taxonomy of Rhizobia [mini-review]. Agronomie 21: 569 –576

120. Zakhia, F, Jeder, H, Domergue, O, Willems, A, Cleyet-Marel, JC,Gillis, M, Dreyfus, B, de Lajudie, P (2004) Characterisation ofwild legume nodulating bacteria (LNB) in the infra-arid zone ofTunisia. Syst Appl Microbiol 27(3): 380 –395

121. Zehr, JP, McReynolds, LA (1989) Use of degenerate oligonucleo-tides for amplification of the nifH gene from the marinecyanobacterium Trichodesmium thiebautii. Appl Environ Microbiol55: 2522–2526

122. Zinniel, DK, Lambrecht, P, Harris, NB, Feng, Z, Kuczmarski, D,Higley, P, Ishimaru, CA, Arunakumari, A, Barletta, RG, Vidaver,AK (2002) Isolation and characterization of endophytic coloniz-ing bacteria from agronomic crops and prairie plants. ApplEnviron Microbiol 68(5): 2198 –2208


Diverse Bacteria Associated with Root Nodules of Spontaneous Legumes in Tunisia and First Report for...

Documents

Transcript of Diverse Bacteria Associated with Root Nodules of Spontaneous Legumes in Tunisia and First Report for...