Department of Microbiology, University of Otago, PO Box 56, Dunedin, New Zealand1
OmniGene Bioproducts Inc., Cambridge, MA 02138, USA2
Author for correspondence: Clive W. Ronson. Tel: +64 3 4797701. Fax: +64 3 4798540. e-mail: clive.ronson{at}stonebow.otago.ac.nz
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Keywords: biotin biosynthesis, ß-ketoacyl-acyl carrier protein synthase III, horizontal gene transfer
Abbreviations: NMN, nicotinic acid mononucleotide
The GenBank accession number for the sequence reported in this paper is AF311738.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The symbiosis island was discovered in a population of mesorhizobia nodulating a stand of L. corniculatus planted 7 years earlier. The stand was established with a single inoculant Mesorhizobium strain, ICMP 3153, in tussock grassland devoid of naturalized rhizobia capable of nodulating the plant. Genetically diverse mesorhizobia recovered from nodules from the stand all contained identical symbiotic DNA, suggesting that they arose through transfer of symbiotic DNA from the inoculant strain to nonsymbiotic rhizobia present at the site. The symbiotic DNA was located on the chromosome of the strains (Sullivan et al., 1995 ). Subsequently, seven nonsymbiotic strains that belonged to four genomic species of mesorhizobia were isolated from the site where the diverse symbionts were discovered. Phenotypic characterization of the nonsymbionts revealed that they were auxotrophic for biotin and thiamin and all but one were auxotrophic for nicotinate. In contrast, ICMP 3153, R7A, which is a field reisolate of ICMP 3153, and the diverse symbionts were prototrophic for all three vitamins, suggesting that the island contained genes required for vitamin synthesis (Sullivan et al., 1996
). Transfer of the symbiosis island from R7A to nonsymbionts in the laboratory was subsequently demonstrated using selection for prototrophy (Sullivan & Ronson, 1998
).
Little is known about vitamin synthesis and metabolism in rhizobia. Individual strains of some species require one or more of pantothenate, thiamin and biotin for growth, and Azorhizobium caulinodans also requires nicotinate (Dreyfus et al., 1983 ; Graham, 1963
). Loci required for thiamin synthesis are located on plasmids in Sinorhizobium meliloti and Rhizobium etli (Finan et al., 1986
; Miranda-Rios et al., 1997
). An early study of the biotin requirements of several rhizobial species suggested that rhizobia could be divided into prototrophs, auxotrophs and bradytrophs, strains which grow poorly in continuous serial culture in biotin-free media (Wilson & Wilson, 1942
). Recent studies carried out using S. meliloti suggested that both synthesis and uptake of biotin promote rhizosphere colonization (Streit & Phillips, 1996
). The association of vitamin loci with acquired genetic elements, some of which contain known symbiotic genes, together with the results of Streit et al. (1996)
suggest that the loci may play an important role in the plantmicrobe interaction. However, the fact that some wild-type symbiotic rhizobia are vitamin auxotrophs counters this suggestion.
Here we present the analysis of a 13·2 kb region of the Mesorhizobium loti symbiosis island that contains operons required for the syntheses of nicotinic acid mononucleotide (NMN) and biotin. The bio operon included a novel gene, designated bioZ, which was implicated in the synthesis of the biotin precursor pimeloyl-CoA. The region also contained a nifA gene required for nitrogen fixation. The vitamin loci were not required for symbiotic nitrogen fixation or for rhizosphere colonization.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
DNA sequencing.
EcoRI and PstI fragments of p637 were subcloned into pUC19, and the plasmid templates were sequenced using universal M13 forward and reverse sequencing primers and custom primers. Other primers used included a Tn5-specific sequencing primer, 5'-CGTTCAGGACGCTACTT-3'. Sequencing was carried out using thermal cycle sequencing and an Applied Biosystems model 377 autosequencer. The sequence was assembled using the Seqman package (DNASTAR, Madison, WI). Open reading frames (ORFs) likely to encode genes were identified using Genemark.hmm utilizing a heuristically derived model (Besemer & Borodovsky, 1999 ), and through database searches using BLAST version 2.0 (Altschul et al., 1997
).
Tn5 mutagenesis.
An E. coli HB101 clone containing cosmid p637 was mutagenized with Tn5 using phage 467 (de Bruijn & Lupski, 1984
). After phage infection, Tn5 insertions in p637 were selected by preparing cosmid DNA from pooled Kmr Tcr mutants, transforming this DNA by electroporation into HB101 and selecting for Kmr Tcr transformants. The positions of Tn5 insertions were mapped by restriction endonuclease analysis and by cloning BamHI or EcoRIBamHI fragments containing the arms of Tn5 into pUC19 and sequencing these clones using the Tn5-specific primer. Marker exchange of Tn5 insertions into strain R7A was forced by plasmid incompatibility using pPH1JI (Ruvkun & Ausubel, 1981
). To confirm that the exchange had occurred, genomic DNA was prepared from the putative R7A::Tn5 mutants, digested with EcoRI, blotted and probed successively with p637 and p239.
Bacterial crosses.
Crosses were carried out by triparental spot matings on TY plates using the helper plasmid pRK2013 in HB101 (Ditta et al., 1980 ) and using HB101 donor strains. Mesorhizobium transconjugants were selected on G/RDM media supplemented with appropriate antibiotics.
Nodulation tests and symbiotic phenotype.
Nodulation tests were carried out using L. corniculatus as described by Vincent (1970) . Uninoculated plants and plants inoculated with strain R7A were used as controls. After 2 weeks, plants were scored for the presence of nodules. After 6 weeks, shoot length and shoot wet weight were measured. No significant difference in shoot weight or shoot length in comparison to uninoculated controls indicated an ineffective (Fix-) symbiosis.
Rhizosphere colonization studies.
Inocula were prepared from 48 h cultures grown in G/RDM media supplemented with 1 ng biotin ml-1. The cells were washed and diluted, and viable counts were performed as described by Streit et al. (1996) . L. corniculatus seeds were surface-sterilized and germinated as described by Vincent (1970)
. Individual 2-d-old seedlings were transferred into modified Leonard jars containing vermiculite and 0·25x Hoaglands solution, inoculated with 2090 c.f.u. and incubated under controlled conditions as described by Streit et al. (1996)
. Roots were harvested 10 d post-inoculation and ground in 1 ml sterile distilled water with a Dounce homogenizer. Rhizobia were enumerated by plating dilutions on G/RDM media. For dual inoculation experiments where R7A was compared to R7A::Tn5-5, 50 colonies were patched from G/RDM plates used for total enumeration onto G/RDM containing neomycin to determine the ratio of mutant to wild-type.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The nucleotide sequence of the insert in p637 was determined, and a 13·2 kb region containing potential vitamin loci was analysed (Fig. 1). Of 12 ORFs encoding proteins larger than 60 amino acids identified, 11 showed similarity to sequences in the databases (Table 2
). The deduced amino acid sequences of nadA, B and C were similar to those of the nadA, B and C gene products which constitute the pathway for NMN synthesis in E. coli (Penfound & Foster, 1996
). Genes required for biotin synthesis identified on the basis of the similarity of their products to E. coli orthologues were bioB, F, D and A. The p637 genes were coupled by overlapping stop and start codons, suggesting that they are transcribed as a single operon. The start codon of a fifth ORF, bioZ, overlapped the stop codon of bioA. The deduced amino acid sequence of bioZ showed similarity to ß-ketoacyl-acyl carrier protein synthases (KAS IIIs) from several bacterial species, including the fabH gene product of E. coli (Table 2
; Fig. 2
). The deduced amino acid sequences of nifA and panD showed similarity with the NifA protein required for regulation of several nif and fix genes in nitrogen-fixing bacteria, and the PanD protein required for pantothenate biosynthesis, respectively. The translated product of ORF1 showed strong sequence similarity over its entire length with the product of a gene of unknown function from Neisseria meningitidis that is flanked by nadA and nadC in that organism (Table 2
).
|
|
|
Complementation studies with M. loti bio genes
The Bio- phenotype of the R7AbioZ::Tn5 strains implies that bioZ is required for synthesis of the biotin precursor pimeloyl-CoA, as only the bioA, B, D and F genes are required for the synthesis of biotin from pimeloyl-CoA (Demoll, 1996 ). In E. coli, the bioC and bioH genes are involved in pimeloyl-CoA synthesis (Demoll, 1996
; Eisenberg, 1987
). Complementation studies were carried out to determine if bioZ could complement E. coli bioC and bioH mutants, and to confirm the function of the R7A bioBFDA genes. Plasmid pBIO131 containing the Bacillus subtilis bioI gene was used as a positive control for the complementation of bioH and bioC mutants (Bower et al., 1996
). Several plasmids that differed in their symbiosis island gene complement and the orientation of the genes with respect to the vector lac promoter were constructed (Fig. 1
; Table 1
). The results suggested that variations in the level of expression of the R7A bio operon from the various constructs influenced their ability to complement the E. coli bio mutants. E. coli bioA, B, D, F and H mutants were complemented by at least one construct containing the entire symbiosis island bio operon, while bioZ alone complemented the E. coli bioH mutant but not the bioC mutant (Table 3
).
|
In order to further confirm a role for bioZ in biotin synthesis in R7A, the bioZ gene was subcloned from p1013 as a BamHIHindIII fragment into the broad-host-range expression vector pML122 to produce plasmid p1015 (Table 1). This plasmid restored the ability of R7A::Tn5-27 to grow on biotin-free G/RDM containing avidin.
Plasmids p637::Tn5-5, p637::Tn5-27 and p637::Tn5-33 did not convert CJ3 to biotin prototrophy, confirming that bioA and bioZ are required for biotin synthesis in M. loti. To determine whether ORF2 or panD, which flank the bio operon, were also required, regions of p637 were first cloned into pBluescript II KS and subsequently subcloned into pIJ3200 using suitable restriction sites to produce p995, p996 and p997 (Table 1; Fig. 1
). These plasmids converted CJ3 to biotin prototrophy, indicating that neither ORF2 nor panD is required for biotin synthesis in CJ3.
Growth-room rhizosphere colonization studies
Five rhizosphere competition experiments, each with five replicates, were carried out using modified Leonard jars placed in a growth room. Treatments comprised R7A alone, R7A::Tn5-5 (bioA::Tn5) alone, and a 1:1 mix of the two strains. Inocula were prepared by washing and diluting 48 h cultures 105-fold to give 20120 c.f.u. per treatment. Between 105 and 107 bacteria were recovered from each root harvested 10 d post-inoculation, indicating that the bacteria had undergone 1218 doublings. No differences in the abilities of the two strains to colonize the L. corniculatus rhizosphere were detected (R7A, log104·7±1·2; R7AbioA::Tn5, log105·3±1·1; 1:1 mix, log105·4±1·0, with each strain equally represented).
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Biotin is an essential cofactor for carboxylase enzymes and in E. coli and B. subtilis is synthesized from pimeloyl-CoA via four enzymic steps involving the products of the bioA, B, D and F genes. Steps involved in pimeloyl-CoA synthesis are less well-characterized but differ between the two organisms. The functions of the symbiosis island bioA, B, D and F genes were confirmed by their ability to complement respective E. coli bio mutants. However, orthologues of bioZ are not found amongst the known bio genes of other organisms.
Variation in the growth of complemented E. coli bio mutants that was dependent on the particular plasmid used for complementation was observed. For example, p961 complemented the E. coli bioA, bioD and bioF mutants to wild-type growth; however, the E. coli bioB mutant was bradytrophic when complemented with this plasmid. In contrast, the E. coli bioB mutant was fully complemented by p964 containing the entire bio operon fused to the E. coli lacZ promoter, but p964 did not complement the bioA, D or F mutants. Plasmid p964 was also the only plasmid containing the entire symbiosis island bio operon that complemented the bioH mutant, albeit to bradytrophic growth, whereas p1013 restored the mutant to wild-type growth. Problems with complementation and/or viability associated with the expression of bioB have been observed in other organisms, including E. coli and Serratia marcescens (Ifuku et al., 1995 ; Levy-Schil et al., 1993
; Sakurai et al., 1993
). Moreover, S. meliloti strains expressing the E. coli bio operon show reduced viability (Streit & Phillips, 1996
).
Pimeloyl-CoA is a seven-carbon thioester that is the first dedicated intermediate in the biotin biosynthetic pathway. B. sphaericus and B. subtilis both contain a pimeloyl-CoA synthetase, encoded by bioW, which condenses pimelic acid with CoA to form pimeloyl-CoA. bioW complements both E. coli bioC and bioH mutants in the presence of pimelic acid (Bower et al., 1996 ; Gloeckler et al., 1990
). In E. coli, pimeloyl-CoA synthesis does not involve pimelic acid as an intermediate. Instead, a synthetic mechanism similar to a fatty acid synthase reaction involving the condensation of three malonyl-CoA molecules has been suggested based on radiolabelling studies (Ifuku et al., 1994
; Sanyal et al., 1994
). The symbiosis island bioZ gene cannot complement an E. coli bioC mutant, even in the presence of free pimelic acid, and therefore its product does not appear to act as a pimeloyl-CoA synthetase; however, it does complement bioH mutants.
The strong sequence similarity of BioZ to ß-ketoacyl-acyl carrier protein synthase III (FabH) suggests that BioZ is a fatty acid synthase. FabH forms part of the dissociated (type II) fatty acid synthase system in bacteria. It catalyses the first step in fatty acid synthesis, the condensation of an acetyl-CoA starter unit with malonyl-ACP to form acetoacetyl-ACP. The enzyme also has an acetyl-CoA:ACP transacylase activity which involves the interconversion of acetyl-ACP and acetyl-CoA (Tsay et al., 1992 ). Condensing enzymes such as FabH function via an acylenzyme intermediate attached as a thioester to an active cysteine residue. The active-site cysteine of FabH is conserved in BioZ (Fig. 2
). It seems possible that BioZ is involved in a condensation reaction perhaps involving incorporation of a thioester with an odd number of carbon atoms leading to synthesis of pimeloyl-ACP, and also in its transacylation to pimeloyl-CoA. E. coli FabH, which has been shown to carry out a condensation reaction of propionyl-CoA with malonyl-ACP to produce fatty acids with odd numbers of carbon atoms (Heath & Rock, 1996
), may carry out similar steps to BioZ but have a narrower specificity for its transacylation reaction, accounting for the requirement for BioH in E. coli. In this model, BioC would be required for the synthesis of one of the substrates for the initial condensation reaction. It of course remains to be determined whether mesorhizobia contain a bioC orthologue elsewhere in their genome.
The presence of vitamin synthesis genes on the symbiosis island suggests that they may be required for, or augment, the symbiotic interaction. However, mutants with Tn5 insertions in the NMN and biotin synthesis operons were unaffected in nodulation or nitrogen fixation, and the bioA mutant was unaffected in rhizosphere colonization. These results are consistent with the findings that some rhizobia are natural biotin auxotrophs (Graham, 1963 ) and that nonsymbiotic mesorhizobia were present in the rhizosphere of L. corniculatus grown in the field (Sullivan et al., 1996
). However, they are contrary to a previous study which reported that S. meliloti biotin auxotrophs competed very poorly with the parent strain (Streit et al., 1996
). It is possible that the difference in the results reflects the different host plants or rhizobial species used.
Despite the lack of an obvious symbiotic role, the presence of vitamin synthesis operons on the island strongly suggests that they can contribute to a successful plantmicrobe interaction. Conversely, the lack of the operons may benefit the nonsymbiotic mesorhizobia. Acquired genes usually confer a selective advantage only under specific conditions and may be prone to loss under other environmental conditions. The vitamin synthesis genes may have been lost from the nonsymbionts if they were detrimental to the rhizobia for maintaining a saprophytic lifestyle in the absence of the host. For example, vitamin starvation may enable the bacteria to adopt a stress-resistant dormant state, which would only be alleviated in the presence of an environment conducive to growth. In this regard, the nonsymbiotic rhizobia may use the presence of vitamins as an indicator of the presence of a plant root. Certainly the presence of the bio and nad genes on the symbiosis island extends the range of house-keeping genes known to be encoded on accessory genetic elements.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Beringer, J. E. (1974). R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84, 188-198.[Medline]
Besemer, J. & Borodovsky, M. (1999). Heuristic approach to deriving models for gene finding. Nucleic Acids Res 27, 3911-3920.
Bower, S., Perkins, J. B., Yocum, R. R., Howitt, C. L., Rahaim, P. & Pero, J. (1996). Cloning, sequencing, and characterization of the Bacillus subtilis biotin biosynthetic operon. J Bacteriol 178, 4122-4130.[Abstract]
Boyer, H. W. & Roulland-Dussoix, D. (1969). A complementation analysis of the restriction and modification of DNA in Escherichia coli. J Mol Biol 41, 459.[Medline]
de Bruijn, F. J. & Lupski, J. R. (1984). The use of transposon Tn5 mutagenesis in the rapid generation of correlated physical and genetic maps of DNA segments cloned into multicopy plasmids a review. Gene 27, 131-149.[Medline]
Cleary, P. P. & Campbell, A. (1972). Deletion and complementation analysis of the biotin gene cluster of Escherichia coli. J Bacteriol 112, 830-839.[Medline]
Demoll, E. (1996). Biosynthesis of biotin and lipoic acid. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 704709. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
Ditta, G. S., Stanfield, D., Corbin, D. & Helinski, D. R. (1980). Broad host range DNA cloning system for Gram-negative bacteria: construction of a gene bank of Rhizobium meliloti. Proc Natl Acad Sci USA 77, 7347-7351.[Abstract]
Dreyfus, B. L., Elmerich, C. & Dommergues, Y. R. (1983). Free living Rhizobium strain able to grow on N2 as the sole nitrogen source. Appl Environ Microbiol 45, 711-713.
Eisenberg, M. A. (1987). Biosynthesis of biotin and lipoic acid. In Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, pp. 544550. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
Finan, T. M., Kunkel, B., De Vos, G. F. & Signer, E. R. (1986). Second symbiotic megaplasmid in Rhizobium meliloti carrying exopolysaccharide and thiamine synthesis genes. J Bacteriol 167, 66-72.[Medline]
Fischer, H. M. (1994). Genetic regulation of nitrogen fixation in rhizobia. Microbiol Rev 58, 352-386.[Abstract]
Friedman, A. M., Long, S. R., Brown, S. E., Buikema, W. J. & Ausubel, F. M. (1982). Construction of a broad host range cosmid cloning vector and its use in genetic analysis of Rhizobium mutants. Gene 18, 289-296.[Medline]
Gloeckler, R., Ohsawa, I., Speck, D. & 7 other authors (1990). Cloning and characterization of the Bacillus sphaericus genes controlling the bioconversion of pimelate into dethiobiotin. Gene 87, 6370.[Medline]
Graham, P. H. (1963). Vitamin requirements of root nodule bacteria. J Gen Microbiol 30, 245-248.
Hatfield, D., Hofnung, M. & Schwartz, M. (1969). Genetic analysis of the maltose A region in Escherichia coli. J Bacteriol 98, 559-567.[Medline]
Heath, R. J. & Rock, C. O. (1996). Roles of the FabA and FabZ beta-hydroxyacyl-acyl carrier protein dehydratases in Escherichia coli fatty acid biosynthesis. J Biol Chem 271, 27795-27801.
Hirsch, P. R. (1979). Plasmid-determined bacteriocin production by Rhizobium leguminosarum. J Gen Microbiol 113, 219-228.
Ifuku, O., Miyaoka, H., Koga, N., Kishimoto, J., Haze, S., Wachi, Y. & Kajiwara, M. (1994). Origin of carbon atoms of biotin: C-13-NMR studies on biotin biosynthesis in Escherichia coli. Eur J Biochem 220, 585-591.[Abstract]
Ifuku, O., Koga, N., Haze, S., Kishimoto, J., Arai, T. & Wachi, Y. (1995). Molecular analysis of growth inhibition caused by overexpression of the biotin operon in Escherichia coli. Biosci Biotechnol Biochem 59, 184-189.[Medline]
Labes, M., Puhler, A. & Simon, R. (1990). A new family of RSF1010-derived expression and lac-fusion broad-host-range vectors for Gram-negative bacteria. Gene 89, 37-46.[Medline]
Levy-Schil, S., Debussche, L., Rigault, S. & 7 other authors (1993). Biotin biosynthetic pathway in recombinant strains of Escherichia coli overexpressing bio genes: evidence for a limiting step upstream from KAPA. Appl Microbiol Biotechnol 38, 755762.
Liu, Y. N., Tang, J. L., Clarke, B. R., Dow, J. M. & Daniels, M. J. (1990). A multipurpose broad host range cloning vector and its use to characterize an extracellular protease gene of Xanthomonas campestris pathovar campestris. Mol Gen Genet 220, 433-440.[Medline]
Miranda-Rios, J., Morera, C., Taboada, H., Davalos, A., Encarnacion, S., Mora, J. & Soberon, M. (1997). Expression of thiamin biosynthetic genes (thiCOGE) and production of symbiotic terminal oxidase cbb3 in Rhizobium etli. J Bacteriol 179, 6887-6893.[Abstract]
Penfound, T. & Foster, J. W. (1996). Biosynthesis and recycling of NAD. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 721730. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
Ronson, C. W., Nixon, B. T., Albright, L. M. & Ausubel, F. M. (1987). Rhizobium meliloti ntrA (rpoN) gene is required for diverse metabolic functions. J Bacteriol 169, 2424-2431.[Medline]
Ruvkun, G. B. & Ausubel, F. M. (1981). A general method for site-directed mutagenesis in prokaryotes. Nature 289, 85-88.[Medline]
Sakurai, N., Imai, Y., Masuda, M., Komatsubara, S. & Tosa, T. (1993). Molecular breeding of a biotin-hyperproducing Serratia marcescens strain. Appl Environ Microbiol 10, 3225-3232.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Sanyal, I., Lee, S. L. & Flint, D. H. (1994). Biosynthesis of pimeloyl-CoA, a biotin precursor in Escherichia coli, follows a modified fatty acid synthesis pathway: 13C-labeling studies. J Am Chem Soc 116, 2637-2638.
Streit, W. R. & Phillips, D. A. (1996). Recombinant Rhizobium meliloti strains with extra biotin synthesis capability. Appl Environ Microbiol 62, 3333-3338.
Streit, W. R., Joseph, C. M. & Philips, D. A. (1996). Biotin and other water-soluble vitamins are key growth factors for alfalfa root colonisation by Rhizobium meliloti 1021. Mol PlantMicrobe Interact 9, 330-338.[Medline]
Sullivan, J. T. & Ronson, C. W. (1998). Evolution of rhizobia by acquisition of a 500-kb symbiosis island that integrates into a phe-tRNA gene. Proc Natl Acad Sci USA 95, 5145-5149.
Sullivan, J. T., Patrick, H. N., Lowther, W. L., Scott, D. B. & Ronson, C. W. (1995). Nodulating strains of Rhizobium loti arise through chromosomal symbiotic gene transfer in the environment. Proc Natl Acad Sci USA 92, 8985-8989.[Abstract]
Sullivan, J. T., Eardly, B. D., van Berkum, P. & Ronson, C. W. (1996). Four unnamed species of nonsymbiotic rhizobia isolated from the rhizosphere of Lotus corniculatus. Appl Environ Microbiol 62, 2818-2825.[Abstract]
Tsay, J. T., Oh, W., Larson, T. J., Jackowski, S. & Rock, C. O. (1992). Isolation and characterization of the beta-ketoacyl-acyl carrier protein synthase-III gene (fabH) from Escherichia coli K-12. J Biol Chem 267, 6807-6814.
Vincent, J. M. (1970). A Manual for the Practical Study of Root Nodule Bacteria. Oxford: Blackwell Scientific Publications.
Wilson, J. B. & Wilson, P. W. (1942). Biotin as a growth factor for rhizobia. J Bacteriol 43, 329-341.
Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103-109.[Medline]
Received 29 December 2000;
accepted 29 January 2001.