Nucleotide sequence of Rhizobium meliloti insertion sequence ISRm1: homology to IS2 from Escherichia coli and IS426 from Agrobacterium tumefaciens.

Nucleotide sequencing of Rhizobium meliloti insertion sequence ISRm1 showed that it is 1319 nucleotides long and includes 32/31 nucleotide terminal inverted repeats. Analysis of five different insertion sites using sequencing primers complementary to sequences within the left and right ends demonstrated that ISRm1 generates five bp direct repeats at the sites of insertion. Although ISRm1 has shown a target preference for certain short regions (hot spots), there was no apparent similarity in the DNA sequences near the insertion sites. On one strand ISRm1 contains two contiguous open reading frames (ORFs) spanning most of its length. ISRm1 was found to have over 50% sequence homology to insertion sequences IS2 from Escherichia coli and IS426 from Agrobacterium tumefaciens. Their sizes, the sequences of their inverted repeats, and the characteristics of their insertion sites are also comparable, indicating that ISRm1, IS2 and IS426 belong to a class of related insertion sequences. Comparison of the proteins potentially encoded by these insertion sequences showed that the two ORFs found in ISRm1 are also present in IS2 and IS426, suggesting that they may be functional genes.     

Identification and nucleotide sequence of Rhizobium meliloti insertion sequence ISRm3: similarity between the putative transposase encoded by ISRm3 and those encoded by Staphylococcus aureus IS256 and Thiobacillus ferrooxidans IST2.

The insertion sequence ISRm3 was discovered simultaneously in different Rhizobium meliloti strains by probing Southern blots of total cellular DNA with 32P-labeled pTA2. This plasmid is indigenous to strain IZ450 and fortuitously contained four copies of ISRm3. By using an internal EcoRI fragment as a specific probe (pRWRm31), homology to ISRm3 was subsequently detected in over 90% of R. meliloti strains tested from different geographical locations around the world. The frequency of stable nonlethal ISRm3 transpositions was estimated to be 4 x 10(-5) per generation per cell in strain SU47 when grown in liquid culture. The entire nucleotide sequence of ISRm3 in R. meliloti 102F70 is 1,298 bp and has 30-bp terminal inverted repeats which are perfectly matched. Analysis of six copies of ISRm3 in two strains showed that a variable number of base pairs (usually eight or nine) were duplicated and formed direct repeats adjacent to the site of insertion. On one DNA strand, ISRm3 contains an open reading frame spanning 93% of its length. Comparison of the putative protein encoded with sequences derived from the EMBL and GenBank databases showed significant similarity between the putative transposases of ISRm3 from R. meliloti, IS256 from Staphylococcus aureus, and IST2 from Thiobacillus ferroxidans. These insertion sequences appear to be distantly related members of a distinct class.     

The Sinorhizobium meliloti insertion sequence (IS) elements ISRm102F34-1/ISRm7 and ISRm220-13-5 belong to a new family of insertion sequence elements

The Sinorhizobium meliloti insertion sequence (IS) elements ISRm102F34-1 and ISRm220-13-5 are 1481 and 1550 base pairs (bp) in size, respectively. ISRm102F34-1 is bordered by 15 bp imperfect terminal inverted repeat sequences (two mismatches), whereas the terminal inverted repeat of ISRm220-13-5 has a length of 16 bp (two mismatches). Both insertion sequence elements generate a 6-bp target duplication upon transposition. The putative transposase enzymes of ISRm102F34-1 and ISRm220-13-5 consist of 449 or 448 amino acid residues with predicted molecular weights of 50.7 or 51.3 kDa and theoretical isoelectric points of 10.8 or 11.1, respectively. ISRm102F34-1 is identical in 98.9% of its nucleotide sequence to an apparently inactive copy of an insertion sequence element, designated ISRm7, which flanks the left-end of the nodule formation efficiency (nfe) region of plasmid pRmeGR4b of S. meliloti strain GR4. ISRm102F34-1 and ISRm220-13-5 are closely related since they show an overall identity of 57.0% at the nucleotide sequence level and of 47.3% at the deduced amino acid level of their putative transposases. Both insertion sequence elements displayed significant similarity to the Xanthomonas campestris ISXc6 and its homolog IS1478a. Since none of these insertion sequence elements could be allocated to existing families of insertion sequence elements, a new family is proposed. Analysis of the distribution of ISRm102F34-1/ISRm7 in various local S. meliloti populations sampled from Medicago sativa, Medicago sphaerocarpa and Melilotus alba host plants at different locations in Spain revealed its presence in 35% of the isolates with a copy number ranging from 1 to 5. Furthermore, ISRm102F34-1/ISRm7 homologs were identified in other rhizobial species.     

Distribution of insertion sequence ISRm1 in Rhizobium meliloti and other gram-negative bacteria

An internal 0.9 kb segment of Rhizobium meliloti insertion sequence ISRm1 was used as a probe to determine the distribution of ISRm1 in strains of R. meliloti and other Gram-negative bacteria. The insertion sequence was detected in 80% (12/15) of R. meliloti strains from different parts of the world. Its copy number ranged from one to at least eleven. The ISRm1 copies detected showed variation in their internal restriction sites and their degree of homology to the probe. ISRm1 was found in a variety of genomic restriction fragments, and was detected in plasmids, including the nod and exo megaplasmids of R. meliloti. Other rhizobia found to contain ISRm1 were a strain of R. leguminosarum biovar phaseoli and two Rhizobium isolates capable of nodulating both Medicago sativa and Phaseolus vulgaris. It was also found in a diazotrophic soil bacterium isolated from the roots of wetland rice.     

Isolation and analysis of IS6120, a new insertion sequence from Mycobacterium smegmatis

Insertion sequence IS6120 from Mycobacterium smegmatis was identified by its ability to transpose into different sites in the lambda repressor gene, cl857, carried on an Escherichia coli/mycobacteria shuttle plasmid. IS6120 is a novel 1.5 kb insertion sequence, which has 24-bp imperfect terminal inverted repeats and generates 9-bp duplications of the target DNA following insertion. IS6120 is present in at least three copies in M. smegmatis but was not found in other species, including Mycobacterium tuberculosis. Nucleotide sequence analysis revealed that IS6120 contains two open reading frames, one of which encodes a putative transposase with similarities to those found in IS256 from Staphylococcus aureus, IST2 from Thiobacillus ferrooxidans, and ISRm3 from Rhizobium meliloti. The fact that IS6120 does not recognize a consensus target sequence for insertion and has no homologous sequences in the other strains studied makes IS6120 useful for transposon mutagenesis in mycobacteria.     

Characterization of IS666, a newly described insertion element of Mycobacterium avium

The insertion sequence IS666 was isolated from Mycobacterium avium strain 101. IS666 is a 1474 bp insertion sequence belonging to the IS256 family, that includes IS6120 from Mycobacterium smegmatis, IS1166 and IS1295 from Rhodococcus sp. IGTS8, IST2 from Thiobacillus ferrooxidans, IS256 from Staphylococcus aureus, and ISRm3 from Rhizobium meliloti. IS666 has 24 bp imperfect inverted repeats that fit the consensus described for the family, and generates 9 bp duplications upon insertion into the host DNA with no apparent specificity in the target sequence. In contrast with its two closest homologues, IS1166 and IS6120, IS666 contains a single ORF that would codify a transposase of 434 aa. IS666 is restricted to M. avium, where it is present in 21% of the isolates in a number ranging between 1 to 7 copies.     

Identification of Rhizobium-specific intergenic mosaic elements within an essential two-component regulatory system of Rhizobium species.

Analysis of the DNA regions upstream of the phosphoenolpyruvate carboxykinase gene (pckA) in Rhizobium meliloti and Rhizobium sp. strain NGR234 identified an open reading frame which was highly homologous to the Agrobacterium tumefaciens chromosomal virulence gene product ChvI. A second gene product, 500 bp downstream of the chvI-like gene in R. meliloti, was homologous to the A. tumefaciens ChvG protein. The homology between the R. meliloti and A. tumefaciens genes was confirmed, because the R. meliloti chvI and chvG genes complemented A. tumefaciens chvI and chvG mutants for growth on complex media. We were unable to construct chvI or chvG insertion mutants of R. meliloti, whereas mutants carrying insertions outside of these genes were readily obtained. A 108-bp repeat element characterized by two large palindromes was identified in the chvI and chvG intergenic regions of both Rhizobium species. This element was duplicated in Rhizobium sp. strain NGR234. Another structurally similar element with a size of 109 bp was present in R. meliloti but not in Rhizobium sp. strain NGR234. These elements were named rhizobium-specific intergenic mosaic elements (RIMEs), because their distribution seems to be limited to members of the family Rhizobiaceae. A homology search in GenBank detected six more copies of the first element (RIME1), all in Rhizobium species, and three extra copies of the second element (RIME2), only in R. meliloti. Southern blot analysis with a probe specific to RIME1 showed the presence of several copies of the element in the genome of R. meliloti, Rhizobium sp. strain NGR234, Rhizobium leguminosarum, and Agrobacterium rhizogenes, but none was present in A. tumefaciens and Bradyrhizobium japonicum.     

Phylogeny of fast-growing soybean-nodulating rhizobia support synonymy of Sinorhizobium and Rhizobium and assignment to Rhizobium fredii.

We determined the sequences for a 260-base segment amplified by the polymerase chain reaction (corresponding to positions 44 to 337 in the Escherichia coli 16S rRNA sequence) from seven strains of fast-growing soybean-nodulating rhizobia (including the type strains of Rhizobium fredii chemovar fredii, Rhizobium fredii chemovar siensis, Sinorhizobium fredii, and Sinorhizobium xinjiangensis) and broad-host-range Rhizobium sp. strain NGR 234. These sequences were compared with the corresponding previously published sequences of Rhizobium leguminosarum, Rhizobium meliloti, Agrobacterium tumefaciens, Azorhizobium caulinodans, and Bradyrhizobium japonicum. All of the sequences of the fast-growing soybean rhizobia, including strain NGR 234, were identical to the sequence of R. meliloti and similar to the sequence of R. leguminosarum. These results are discussed in relation to previous findings; we concluded that the fast-growing soybean-nodulating rhizobia belong in the genus Rhizobium and should be called Rhizobium fredii.     

Isolation and characterization of functional insertion sequences of rhizobia

Rhizobia are a group of bacteria that form nodules on the roots of legume host plants. The sequenced genomes of the rhizobia are characterized by the presence of many putative insertion sequences (IS) elements. However, it is unknown whether these IS elements are functional and it is therefore relevant to assess their transposition activity. In this work, several functional insertion sequences belonging to the IS1256, IS3, IS5, IS166, and IS21 families were captured from Rhizobium tropici, Rhizobium sp. NGR234 and Sinorhizobium meliloti, using pGBG1 as a trapping system. In silico analysis shows that homologs of rhizobia mobile elements are present in distantly related genomes, suggesting that Rhizobium IS elements are prone to genetic transfer.     

Characterization and primary structures of DNA-binding HU-type proteins from Rhizobiaceae

The DNA-binding HU-type proteins from several species of Rhizobiaceae including Rhizobium meliloti, two strains of Rhizobium leguminosarum with highly different phenotypic characters and Agrobacterium tumefaciens, were characterized and their amino acid sequences were determined. HU-type proteins isolated from R. leguminosarum L18 and A. tumefaciens are identical and show slight differences with the R. meliloti HU-type protein. On the other hand the R. leguminosarum L53 HU-type protein is quite different from the proteins cited above; several amino acid substitutions encountered in this protein result in significant changes in the folding of the polypeptide chain. The biochemical characteristics of these proteins are in good agreement with the respective position of these bacteria in the phylogeny determined by numerical taxonomy.     

A proline tRNA(CGG) gene encompassing the attachment site of temperate phage 16-3 is functional and convertible to suppressor tRNA

Several temperate bacteriophage utilize chromosomal sequences encoding putative tRNA genes for phage attachment. However, whether these sequences belong to genes which are functional as tRNA is generally not known. In this article, we demonstrate that the attachment site of temperate phage 16-3 (attB) nests within an active proline tRNA gene in Rhizobium meliloti 41. A loss-of-function mutation in this tRNA gene leads to significant delay in switching from lag to exponential growth phase. We converted the putative Rhizobium gene to an active amber suppressor gene which suppressed amber mutant alleles of genes of 16-3 phage and of Escherichia coli origin in R. meliloti 41 and in Agrobacterium tumefaciens GV2260. Upon lysogenization of R. meliloti by phage 16-3, the proline tRNA gene retained its structural and functional integrity. Aspects of the co-evolution of a temperate phage and its bacterium host is discussed. The side product of this work, i.e. construction of amber suppressor tRNA genes in Rhizobium and Agrobacterium, for the first time widens the options of genetic study.     

Isolation and characterization of a DNA replication origin from the 1,700-kilobase-pair symbiotic megaplasmid pSym-b of Rhizobium meliloti.

A 4-kb fragment active as an autonomously replicating sequence (ARS) from the Rhizobium meliloti symbiotic megaplasmid pSym-b was isolated by selecting for sequences that allowed a normally nonreplicative pBR322 derivative to replicate in R. meliloti. The resulting Escherichia coli-R. meliloti shuttle plasmid (mini-pSym-b) containing the ARS also replicated in the closely related Agrobacterium tumefaciens, but only in strains carrying pSym-b, suggesting that a megaplasmid-encoded trans-acting factor is required. The copy number of mini-pSym-b was approximately the same as that of the resident megaplasmid, and mini-pSym-b was unstable in the absence of antibiotic selection. An 0.8-kb DNA subfragment was sufficient for replication in both R. meliloti and A. tumefaciens. The minimal ARS exhibited several sequence motifs common to other replication origins, such as an AT-rich region, three potential DnA binding sites, a potential 13-mer sequence, and several groups of short direct repeats. Hybridization experiments indicated that there may be a related ARS on the other megaplasmid, pSym-a. The pSym-b ARS was mapped near exoA, within a region nonessential for pSym-b replication. These results suggest that the R. meliloti megaplasmids share conserved replication origins and that pSym-b contains multiple replication origins. Since the mini-pSym-b shuttle vector can coexist with IncP-1 broad-host-range plasmids, it is also now possible to use two compatible plasmids for cloning and genetic manipulation in R. meliloti.     

Identification of Rhizobium plasmid sequences involved in recognition of Psophocarpus, Vigna, and other legumes

Symbiotic DNA sequences involved in nodulation by Rhizobium must include genes responsible for recognizing homologous hosts. We sought these genes by mobilizing the symbiotic plasmid of a broad host-range Rhizobium MPIK3030 (= NGR234) that can nodulate Glycine max, Psophocarpus tetragonolobus, Vigna unguiculata, etc., into two Nod- Rhizobium mutants as well as into Agrobacterium tumefaciens. Subsequently, cosmid clones of pMPIK3030a were mobilized into Nod+ Rhizobium that cannot nodulate the chosen hosts. Nodule development was monitored by examining the ultrastructure of nodules formed by the transconjugants. pMPIK3030a could complement Nod- and Nif- deletions in R. leguminosarum and R. meliloti as well as enable A. tumefaciens to nodulate. Three non-overlapping sets of cosmids were found that conferred upon a slow-growing Rhizobium species, as well as on R. loti and R. meliloti, the ability to nodulate Psophocarpus and Vigna, thus pointing to the existence of three sets of host-specificity genes. Recipients harboring these hsn regions had truly broadened host-range since they could nodulate both their original hosts as well as MPIK3030 hosts.     

Rhizobium meliloti insertion element ISRm2 and its use for identification of the fixX gene.

Two of the three plasmids of the wild-type Rhizobium meliloti 41 (pRme41a and pRme41c) carry a copy of ISRm2, a 2.7-kilobase-long transposable element. ISRm2 is terminated by 22-base-pair (bp) inverted repeat sequences, exhibiting some homology to the inverted repeats of elements generating 9-bp target sequence duplication. Transposition of ISRm2 results in a duplication of 8 bp in length, rather rare among transposable elements. DNA sequences homologous to an internal fragment of ISRm2 were found in several Rhizobium species. Transposition of ISRm2 into fixation and nodulation genes located on the symbiotic plasmid pRme41b was detected at a high frequency. Exact locations of two copies of ISRm2 which transposed into the nod-nif region on the megaplasmid were determined. In one case, integration into the protein-coding region of the hsnD gene that determines a host specificity function of nodulation occurred. In the other mutant, ISRm2 was localized upstream of nifA, where a short open reading frame coding for a new fix gene (fixX) was identified. The product of fixX is a ferredoxin carrying a characteristic cluster of cysteine residues. On the basis of the observation that the arrangement of the ISRm2 copies is identical in the free-living wild-type cells and in nitrogen-fixing nodules, we concluded that the involvement of ISRm2 transposition in the development of nitrogen-fixing symbiosis is unlikely.     

Osmotic regulation of beta(1-2) glucan synthesis in members of the family Rhizobiaceae.

High osmolarity in the culture medium of growing Agrobacterium tumefaciens strongly inhibited the accumulation of cellular beta(1-2) glucan. However, the enzymatic system required for the synthesis of this polysaccharide from UDP-glucose was not repressed by high osmolarity. Mutants of A. tumefaciens and Rhizobium meliloti affected in beta(1-2) glucan synthesis were unable to grow normally in low-osmolarity media.     

Expression of an agrocin-encoding plasmid of Agrobacterium tumefaciens in Rhizobium meliloti

Plasmid RP4 was used to mobilize the agrocin 84-encoding plasmid, pAg396, from Agrobacterium tumefaciens strain 396 to A. tumefaciens C58 and C58CI as well as Rhizobium meliloti. It was transferred to, but not stably maintained in, R. leguminosarum. It could not be transferred to R. lupini, R. japonicum or R. trifolii. Plasmid pAg396 did not segregate in R. meliloti and produced levels of agrocin comparable to the parental strain A. tumefaciens 396. The potential of agrocin producing R. meliloti in biological control of crown gall is being investigated.     

Expression of the Agrobacterium tumefaciens chvB virulence region in Azospirillum spp.

Inner membranes of Azospirillum brasilense incubated with UDP-glucose were unable to synthesize beta-(1-2) glucan and lacked the 235-kilodalton intermediate protein known to be involved in the synthesis of beta-(1-2) glucan in Agrobacterium tumefaciens and Rhizobium meliloti. Inner membranes of A. brasilense strains carrying a cosmid containing the chromosomal virulence genes chvA and chvB of Agrobacterium tumefaciens formed beta-(1-2) glucan in vitro and synthesized the 235-kilodalton intermediate protein. No DNA homology to the chvB region was found in different wild-type strains of A. brasilense, but the introduction of a cosmid containing the Agrobacterium tumefaciens chvA and chvB regions yielded strains in which DNA hybridization with the chvB region was detected, provided that the strains were grown under an antibiotic selective pressure.     

Genetic analysis of the gentamicin resistance region of pPH1JI and incorporation into a wide host range cloning vehicle

A region of the IncP plasmid pPH1JI encoding resistance to gentamicin, spectinomycin, and streptomycin was characterized by subcloning, deletion, and insertion mutagenesis. Approximate locations of these resistance determinants were established. A 1.6-kb HindIII-SphI segment of this region expresses gentamicin resistance (Gmr) in Escherichia coli when inserted into various plasmid vectors; this DNA segment encodes a polypeptide of 17.5 kDa. Incorporation of this fragment into an IncP cloning vehicle produced a Gmr wide host range vector, pRAR209, which confers levels of Gmr comparable to those expressed by pPH1JI in E. coli, Agrobacterium tumefaciens, and Rhizobium meliloti.     

Reiterated DNA sequences in Rhizobium and Agrobacterium spp.

Repeated DNA sequences are a general characteristic of eucaryotic genomes. Although several examples of DNA reiteration have been found in procaryotic organisms, only in the case of the archaebacteria Halobacterium halobium and Halobacterium volcanii [C. Sapienza and W. F. Doolittle, Nature (London) 295:384-389, 1982], has DNA reiteration been reported as a common genomic feature. The genomes of two Rhizobium phaseoli strains, one Rhizobium meliloti strain, and one Agrobacterium tumefaciens strain were analyzed for the presence of repetitive DNA. Rhizobium and Agrobacterium spp. are closely related soil bacteria that interact with plants and that belong to the taxonomical family Rhizobiaceae. Rhizobium species establish a nitrogen-fixing symbiosis in the roots of legumes, whereas Agrobacterium species is a pathogen in different plants. The four strains revealed a large number of repeated DNA sequences. The family size was usually small, from 2 to 5 elements, but some presented more than 10 elements. Rhizobium and Agrobacterium spp. contain large plasmids in addition to the chromosomes. Analysis of the two Rhizobium strains indicated that DNA reiteration is not confined to the chromosome or to some plasmids but is a property of the whole genome.