Proteome analysis of differentially displayed proteins as a tool for the investigation of symbiosis.

Two-dimensional gel electrophoresis was used to identify differentially displayed proteins expressed during the symbiotic interaction between the bacterium Sinorhizobium meliloti strain 1021 and the legume Melilotus alba (white sweetclover). Our aim was to characterize novel symbiosis proteins and to determine how the two symbiotic partners alter their respective metabolisms as part of the interaction, by identifying gene products that are differentially present between the symbiotic and non-symbiotic states. Proteome maps from control M. alba roots, wild-type nodules, cultured S. meliloti, and S. meliloti bacteroids were generated and compared. Over 250 proteins were induced or up-regulated in the nodule, compared with the root, and over 350 proteins were down-regulated in the bacteroid form of the rhizobia, compared with cultured cells. N-terminal amino acid sequencing and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry peptide mass fingerprint analysis, in conjunction with data base searching, were used to assign putative identity to nearly 100 nodule, bacterial, and bacteroid proteins. These included the previously identified nodule proteins leghemoglobin and NifH as well as proteins involved in carbon and nitrogen metabolism in S. meliloti. Bacteroid cells showed down-regulation of several proteins involved in nitrogen acquisition, including glutamine synthetase, urease, a urea-amide binding protein, and a PII isoform, indicating that the bacteroids were nitrogen proficient. The down-regulation of several enzymes involved in polyhydroxybutyrate synthesis and a cell division protein was also observed. This work shows that proteome analysis will be a useful strategy to link sequence information and functional genomics.

Rhizobia can induce nodules in white clover by "hijacking" mature cortical cells activated during lateral root development.

We examined a range of responses of root cortical cells to Rhizobium sp. inoculation to investigate why rhizobia preferentially nodulate legume roots in the zone of emerging root hairs, but generally fail to nodulate the mature root. We tested whether the inability to form nodules in the mature root is due to a lack of plant flavonoids to induce the bacterial genes required for nodulation or a failure of mature cortical cells to respond to Rhizobium spp. When rhizobia were inoculated in the zone of emerging root hairs, changes in beta-glucuronidase (GUS) expression from an auxin-responsive promoter (GH3), expression from three chalcone synthase promoters, and the accumulation of specific flavonoid compounds occurred in cortical cells prior to nodule formation. Rhizobia failed to induce these responses when inoculated in the mature root, even when co-inoculated with nod gene-inducing flavonoids. However, mature root hairs remained responsive to rhizobia and could support infection thread formation. This suggests that a deficiency in signal transduction is the reason for nodulation failure in the mature root. However, nodules could be initiated in the mature root at sites of lateral root emergence. A comparison between lateral root and nodule formation showed that similar patterns of GH3:gusA expression, chalcone synthase gene expression, and accumulation of a particular flavonoid compound occurred in the cortical cells involved in both processes. The results suggest that rhizobia can "hijack" cortical cells next to lateral root emergence sites because some of the early responses required for nodule formation have already been activated by the plant in those cells.

Transgenic white clover. Studies with the auxin-responsive promoter, GH3, in root gravitropism and lateral root development.

We report improved method for white clover (Trifolium repens) transformation using Agrobacterium tumefaciens. High efficiencies of transgenic plant production were achieved using cotyledons of imbibed mature seed. Transgenic plants were recovered routinely from over 50% of treated cotyledons. The bar gene and phosphinothricin selection was shown to be a more effective selection system than nptII (kanamycin selection) or aadA (spectinomycin selection). White clover was transformed with the soybean auxin responsive promoter, GH3, fused to the GUS gene (beta-glucuronidase) to study the involvement of auxin in root development. Analysis of 12 independent transgenic plants showed that the location and pattern of GUS expression was consistent but the levels of expression varied. The level of GH3:GUS expression in untreated plants was enhanced specifically by auxin-treatment but the pattern of expression was not altered. Expression of the GH3:GUS fusion was not enhanced by other phytohormones. A consistent GUS expression pattern was evident in untreated plants presumably in response to endogenous auxin or to differences in auxin sensitivity in various clover tissues. In untreated plants, the pattern of GH3:GUS expression was consistent with physiological responses which are regarded as being auxin-mediated. For the first time it is shown that localised spots of GH3:GUS activity occurred in root cortical tissue opposite the sites where lateral roots subsequently were initiated. Newly formed lateral roots grew towards and through these islands of GH3:GUS expression, implying the importance of auxin in controlling lateral root development. Similarly, it is demonstrated for the first time that gravistimulated roots developed a rapid (within 1 h) induction of GH3:GUS activity in tissues on the non-elongating side of the responding root and this induction occurred concurrently with root curvature. These transgenic plants could be useful tools in determining the physiological and biochemical changes that occur during auxin-mediated responses.

Rhizobium inoculation and physical wounding result in the rapid induction of the same chalcone synthase copy in Trifolium subterraneum.

The gene or genes encoding chalcone synthase (CHS) in the legume Trifolium subterraneum (subterranean clover) were induced within 6 hr after inoculation with Rhizobium leguminosarum bv. trifolii strain ANU843. No induction was found in uninoculated controls or plants inoculated with either the nodulation-deficient R. l. bv. trifolii strain ANU845 (pSym-) or R. meliloti strain 1021, which is capable of nodulating alfalfa but not clover. Morphological examination of the interaction between the legume and bacteria in this system showed that root hair distortion (a marker of the early events in the interaction) was apparent within 10 hr after inoculation. This indicated that CHS induction could occur before any detectable sign of rhizobial penetration of root hairs. The addition of a crude preparation of R. l. bv. trifolii lipooligosaccharide signals (Nod metabolites) to the plant growth medium had no effect on the expression of CHS over 36 hr, although root hair distortion was apparent over this time. These treatments were then contrasted with physical wounding. Wounding the plants led to a rapid induction of CHS, occurring within 2 hr. Sequence analysis of cloned CHS cDNA from pools sampled after Rhizobium inoculation or wounding treatments showed the gene designated CHS5 was the major CHS species in both treatments. Conserved sequences were found in promoters of CHS5 and soybean Gmchs7, a gene which has overlapping expression patterns. These findings support the view that the induction of the phenylpropanoid pathway is involved in the very early events of the Rhizobium infection of legumes.

Extension of host range of Rhizobium leguminosarum bv. trifolii caused by point mutations in nodD that result in alterations in regulatory function and recognition of inducer molecules.

The positive activation of several nodulation genes in strain ANU843 of Rhizobium leguminosarum biovar trifolii is mediated by the product of the nodD gene and by the interaction of NodD with plant-secreted inducer and anti-inducer compounds. We have mutagenized the nodD gene of strain ANU843 with nitrosoguanidine and have found that the ability of the mutated nodD products to interact with inducer and anti-inducer compounds is affected by the amino acid sequence in at least two key regions, including a novel area between amino acids 77 and 123. Several novel classes of mutants were recognized by phenotypic and molecular analysis of the mutant nodD genes. Classes 1 and 4 mutants were able to induce nodA expression independently of the addition of inducer and anti-inducer compounds and were unable to mediate autoregulation of the nodD gene. Classes 2 and 3 mutants retained several properties of the wild-type nodD, including the ability to interact with inducer and anti-inducer compounds and the capacity to autoregulate nodD expression. In addition, class 2 mutants showed an inducer-independent ability to mediate nodA expression to 10-fold higher levels over control strains. The class 3 mutant showed reactivity to compounds that had little or no inducing ability with the wild-type nodD. An alteration in NodD function was demonstrated with classes 2 and 3 mutants, which showed greatly enhanced ability to complement a Tn5-induced mutation in the nodD1 gene of strain NGR234 and to restore nodulation ability on the tropical legume siratro. Mutants of nodD possessing inducer-independent ability to activate nod gene expression (classes 1, 2, and 4) were capable of extending the host range of R. l. bv. trifolii to the nonlegume Parasponia. DNA sequence analysis showed that single base changes were responsible for the altered phenotypic properties of five of six mutants examined. Four of the six mutations affected amino acid residues in a putative receiver domain in the N-terminal end of the nodD protein.

The relationship between nodulin gene expression and the Rhizobium nod genes in Vicia sativa root nodule development.

The role of the Rhizobium nod genes in the induction of nodulin gene expression was examined by analyzing nodules formed on vetch roots by bacterial strains containing only the nod region. Introduction of an 11-kb cloned nod region of the R. leguminosarum sym plasmid pRL1JI into sym plasmid-cured rhizobia conferred on the recipient strains the ability to induce nodules in which all nodulin genes were expressed. This proves that from the sym plasmid only the nod region is involved in the induction of nodulin gene expression. A transconjugant of Agrobacterium carrying the same nod region induces nodules in which only early nodulin gene expression is detected. Thus, the nod region is essential for the induction of early nodulin gene expression. In this case, nodule cytology may indicate that a defense response of the plant interferes with the induction of late nodulin gene expression. Indirect evidence is presented that indeed the Rhizobium nod genes are also in some way involved in the induction of the expression of late noduling genes. The combination between histological data and pattern of nodulin gene expression furthermore reveals a correlation between nodule structure and nodulin gene expression. This correlation may aid in speculations about the functions of nodulins.

Identification of a nodD-dependent locus in the Rhizobium strain NGR234 activated by phenolic factors secreted by soybeans and other legumes.

Transfer of the strain NGR234nodD 1 gene into the narrow host range R. trifolii strain ANU843 on either a 6.7-kb HindIII or 17-kb XhoI fragment broadens the host range of this bacterium to include the tropical legumes Vigna unguiculata, Glycine ussuriensis, Leucaena leucocephala, and siratro (Macroptilium atropurpureum). Contrary to previous data (Bassam et al. 1986), mutagenesis of the 17-kb XhoI fragment with a mini-Mu lac transposon (Mu dII1734) showed that a functional nodD 1 gene was essential for extended host range. Gene expression studies using both Mu dII1734 fusions and a promoter-cloning vector indicated that several loci, including the nodD 1 gene, are constitutively expressed. No evidence was found for regulation of the strain NGR234 nodD 1 gene by its product. Another locus nod-81, was induced only in the presence of exudates from various plant species, including soybean (Glycine max). Whereas the expression of nod-81 was dependent on the presence of a functional nodD 1 gene product, a regulatory nod-box DNA sequence was not detected 5' to this gene by using available oligonucleotide hybridization probes. The nod-81 locus was induced by genistein, daidzein, naringenin, and coumestrol from both cotyledon and root tissue of freshly germinated soybean seedlings. A broad spectrum of commercially available phenolic compounds stimulated induction of the nod-81 locus, including some that antagonize nod gene induction in other Rhizobium species. The nodD 1 gene product from strain NGR234 was shown to determine the spectrum of compounds that induce nod-81 expression.

Plasmids and stability of symbiotic properties of Rhizobium trifolii.

A conjugal plasmid which encodes both peak nodulation genes and nitrogenase genes, and which is labeled with the transposon Tn5, was transferred to a wild-type Rhizobium trifolii strain to examine the stability and expression of the host range and fixation (Fix+) phenotypes. Transconjugates were isolated which were shown to initially form nitrogen-fixing nodules (Nod+ Fix+) on both clovers and peas. These hybrid strains were then repeatedly passaged through either pea or clover nodules or onto a solid agar medium to determine whether these broadened-host-range characteristics were stably maintained. An instability was noted in the capacity of some of these hybrids to form nitrogen-fixing nodules on all of the host plants used. The broadened nodulation ability was, however, more readily maintained. In some cases, the changes in the Nod+ Fix+ phenotype could be attributed to demonstrable changes in the plasmid profile of the hybrid strains, whereas in other cases no demonstrable plasmid alterations could be detected.