Дисертації з теми "Leguminosarum"

Corrigendum attached to back cover. Includes bibliographical references (leaves 160-190). Investigates the role of rhizopines in rhizobial competition for nodulation, and to isolate the rhizopine synthesis genes in Rhizobium leguminosarum bv. viciae.

Rhizobia are a diverse group of symbiotic alpha-proteobacterial diazotrophs which enter a relationship with specific leguminous plants, in which the plant supplies the bacteria with required compounds whilst the bacteria reduce atmospheric nitrogen into ammonia that the plant uses as a nitrogen source. Modification of rhizobial strains has produced mutants more effective at fixing nitrogen, which in turn results in an increase in biomass of host plants under laboratory conditions but these strains are frequently out competed by wild-type strains in field studies or lost in the intervening years of a crop rotation. This study aimed to establish a library of mutants and a system for screening these strains en masse to identify some of the genes involved in competitive rhizosphere colonization in Rhizobium leguminosam.

Rhizobium leguminosarum is a Gram negative, motile, nitrogen-fixing soil bacterium. Due to the scarcity of iron in the soil bacteria have developed a wide range of iron scavenging systems. The two types of iron scavenging systems used are indirect and direct. In-silico analysis of the genome identified a unique direct iron scavenging system the Hmu operon. This system has been identified in other closely related rhizobium species and is believed to be involved in utilizing heme compounds as a sole source of iron. We have attempted to characterize the role of the Hmu operon in iron utilization by monitoring the growth of R. leguminosarum ATCC 14479 in hemin supplemented media. Growth curves show that it is capable of using hemin as a sole source of iron. The outer membrane profiles were analyzed for the presence of hemin binding proteins.

So much depends upon a red wheel barrow glazed with rain water beside the white chickens - William Carlos Williams 1923. Carbon catabolism of Rhizobium lequminosarum MNF3841 was studied in free-living cells grown in chemostat and in bacteroids. Enzymes of the Entner-Doudoroff (ED) pathway, the pentose phosphate (PP) pathway and the TCA cycle were present, though the absence of phosphofruetokinase prec1uded the operation of the complete Embden- Meyerhof-Parnas (EMP) pathway. The low activity of fructose-bisphosphate aldolase in sugar-grown cells indicated that recycling of glyceraldehyde 3-phosphate (produced by the ED pathway) to fructose 6-phosphate is unlikely. Further catabolism of glyceraldehyde 3-phosphate is probably achieved via the enzymes of the latter part of the EMP pathway which were shown to be present in this organism. In phosphate-limited chemostat culture the activities of invertase, glucose-6-phosphate de hydro- genase, the ED enzymes and 6-phosphogluconate dehydrogenase were two- to three-fold lower in cells grown on fumarate compared to the activities in cells grown on sucrose. Glucose-6-phosphate dehydrogenase also showed modulation of activity due to the nature of the growth limitation with oxygen-limited cells possessing only 50% of the activity of phosphate-limited cells when fumarate was the carbon source. None of the other sugar catabolic enzymes, nor any of those of the TCA cycle showed any modulation in response to the growth substrate or the nature of the growth limitation. Since modulation of some sugar catabolic enzymes was demonstrated in free-living cells in response to growth substrate, the preferences of free-living cells for C4-dicarboxylates, or sugars, were further investigated. In chemostat culture under phosphate-limitation MNF3841 co-utilised fumarate in combination with glucose, or sucrose, or glucose plus fructose. A slight preference for Ca-dicarboxylates was indicated, since the inhibition of sugar utilisation by fumarate was greater than the inhibition of fumarate utilisation by equivalent concentrations of glucose, or sucrose, or glucose plus fructose. Though the obvious importance of Ca-dicarboxylates as carbon sources for both free-living rhizobia and bacteroids is recognised, the ancillary enzymes required for their catabolism have not yet been identified. R. lequminosarum MNF3841 catabolised Ca-dicarboxylates and L-arabinose vi a the TCA cycle with the requirement for acetyl CoA being met by the action of malic enzyme and pyruvate dehydrogenase. Malic enzyme was present in sugar-grown free-living cells though higher levels were observed when fumarate or L-arabinose was the growth substrate. Manganese-dependent malic enzyme activity was evident with either NADP* or NAD* as the cofactor and the activity was stimulated by the presence of KC1. The activity of pyruvate dehydrogenase, which is also required for the catabolism of sugars via the TCA cycle, was higher in sucrose-grown cells than those grown on fumarate. In addition to the TCA cycle and the ancillary enzymes (malic enzyme and pyruvate dehydrogenase) the growth of rhizobia on C4-dicarboxylates (and other substrates which feed into the TCA cycle such as L-glutamate, L-aspartate, L-histidine and L-arabinose) also requires a system of gluconeogenesis. This is accomplished in MHF3841 vi a phosphoenolpyruvate carboxykinase (PEPCK), fruetose-bisphosphate aldolase and fructose-bisphosphatase in conjunction with enzymes of the EMP pathway. In addition R. lequminosarum MNF3085, a PEPCK-deficient mutant, failed to grow on succinate, pyruvate, L-arabinose or L-glutamate, yet grew as well as MNF3841 on glucose, sucrose and glycerol showing that PEPCK is essential for gluconeogenesis. PEPCK and fruetose-bisphosphate aldolase were rapidly derepressed following transfer of cells from a medium with sucrose as the carbon source to one with fumarate as the carbon source. In chemostat culture, the addition of 0.1 mM sucrose caused an 80% inhibition of PEPCK and fructose-bisphosphate aldolase synthesis and 0.4 mM sucrose caused complete inhibition of PEPCK synthesis. Although Ca-dicarboxylate transport was rapidly inducible in free-living cells, bacteroids of MNF3841 isolated from pea nodules could immediately transport 1 4 C-succinate. Furthermore, blocking pyruvate dehydrogenase with arsenite resulted in bacteroids immediately accumulating pyruvate and malate from fumarate indicating that bacteroids in the nodules are in receipt of C*- dicarboxylates. Bacteroids isolated on a Percoll gradient had activities of TCA cycle enzymes, pyruvate dehydrogenase and malic enzyme up to six-fold higher than those in free-living cells, whereas the activities of sugar catabolic enzymes in bacteroids were 2- to 14- fold lower than those in free-living cells grown on sucrose. These activities are a further indication that Ca-dicarboxylates (and not sugars) are the principal form of carbon catabolised by bacteroids. Additionally bacteroids of MNF3841 contained low levels of PEPCK and fruetose-bisphosphate aldolase. The bacteroid-associated PEPCK activity was clearly of bacterial and not plant origin because of its nucleotide requirement and the fact that bacteroids of MNF3085 (PEPCK deficient in the free-living form) contained no PEPCK activity. MNF3085 nodulated and fixed nitrogen as effectively as t he parent which demonstrates that the capacity to synthesise sugars via gluconeogenesis is not required for an effective symbiosis. Thus these data suggest that although bacteroids of MNF 3085 receive sufficient sugar to compensate for their gluconeogenic defect, there is insufficient sugar available to bacteroids of the wild type to completely repress the synthesis of PEPCK. A low amount of sugar available to the bacteroid suggested by these data would be in keeping with the very low activities of sugar catabolic enzymes in the bacteroid. These data in conjunction with the transport of C4-dicarboxylates by bacteroids immediately after their isolation and the elevated activities of the enzymes of C4-dicarboxylate catabolism in bacteroids indicate that C4-dicarboxylates are indeed the major carbon substrate used by them for N2 fixation.

legume-rhizobia symbioses, free-living rhizobia colonise root nodules and develop into N2 fixing specialists known as bacteroids. During bacteroid development, rhizobia must adapt to the nodule environment, consisting of reactive oxygen species, low oxygen, antimicrobial secondary metabolites, low pH and in some nodules, antimicrobial peptides. This study offers a holistic insight into the processes required by R. leguminosarum during bacteroid development in nodules formed on four legumes: Pisum sativum, Vicia faba, Vicia hirsuta and Phaseolus vulgaris. Initially, a high-throughput mutagenesis strategy was used to target genes upregulated during bacteroid development. Screening forty-two mutants on P. sativum identified some moderate phenotypes but more importantly, highlighted functional redundancy between certain gene products. A clear example of functional redundancy was seen between the Mn2+ transporters SitABCD and MntH. Single mutations in sitA or mntH did not cause a symbiotic phenotype whereas the double mutant could not form bacteroids on P. sativum, V. faba or V. hirsuta. Intriguingly, no symbiotic phenotype for the double mutant was observed on P. vulgaris. In addition to Mn2+ transporters, a Mg2+ channel, MgtE, that is essential for growth in Mg2+-limited medium at low pH was identified. As with the Mn2+ transporters, the requirement of MgtE during symbiosis depended upon the species of the hostlegume. Reasons for host-dependent requirement of SitABCD, MntH and MgtE are discussed. The requirement of three O2-responsive regulators that govern regulatory pathways essential to N2 fixation was also investigated. FnrN appears to be the major O2- responsive regulator required for symbiosis but in addition to fnrN, two genes, fixL and fixLc, need to be mutated to prohibit N2 fixation. Other findings include a putative toxin-antitoxin system that hinders N2 fixation when disturbed.

L'etude d'une population naturelle de r. Leguminosarum a porte deux annees de suite sur quatre ensembles de bacteries nodulant des especes vegetales differentes cultivees dans un meme champ dans des conditions experimentales choisies pour permettre l'etude du biovar phaseoli (haricot), biovar trifolii (trefle) et de deux echantillons du biovar viceae (pois et lentille) en estimant l'influence de facteurs de milieu sur la diversite de ces populations; genotype vegetal, espace, temps et historique cultural. L'analyse des profils plasmidiques d'une centaine d'isolats de chacun de ces ensembles a montre que les rhizobium peuvent etre regroupes en types de similarite totale ou en classes definies par la presence de quelques plasmides communs, le plus souvent cryptiques. Des similitudes entre profils de restriction d'adn total ou isoenzymatiques des isolats ainsi regroupes revelent des caracteristiques chromosomiques communes. Ces resultats permettent, en outre, de rapprocher des isolats de biovars differents. La diversite estimee par le nombre de profils differents et les dominances de classes plasmidiques varient selon les ensembles etudies. Les isolats de ces ensembles possedent egalement des capacites variables a fixer l'azote en symbiose: les rhizobium nodulant le trefle et le pois sont fixateurs alors que le cinquieme, pour la lentille, et la moitie, pour le haricot, sont inefficients. Certaines classes plasmidiques permettent de regrouper des isolats de capacites fixatrices homogenes. Ce niveau de caracterisation est ainsi valide. En comparant les differents ensembles etudies, le genotype vegetal apparait comme un facteur determinant l'expression de la diversite microbienne lors de la mise en place d'une population nodulante et en tant que culture specifique precedant un isolement. Le facteur temps ne semble pas modifier la diversite decrite. Concernant le facteur espace, cette diversite est du meme

Rhizobium leguminosarum is a gram-negative soil bacterium that requires iron for survival. However, iron becomes insoluble in the presence of oxygen at physiological pH. In response, Rhizobia species have used siderophore mediated iron transport systems to meet their iron requirements. R. leguminosarum ATCC 14479 produces the trihydroxymate siderophore vicibactin and we hypothesize that the import of the ferric iron-vicibactin complex is energized by the TonB-ExbB-ExbD system. Here, we have identified a putative tonB gene. A tonB mutant was created and compared with wild type in its ability to transport 55Fe-vicibactin. Also, the putative TonB of R. leguminosarum ATCC 14479 is interesting due to its estimated size compared to the TonB of E. coli.. Many groups have attempted structural analysis of the C-terminus of TonB in E. col with inconsistent results. We were successful in expressing 2 different sized TonB C-terminals (120 and 200 amino acids) using pET17b in E. coli.

This thesis reports investigations into the iron nutrition and metabolism of Rhizobium leguminosarum biovar viciae MNF710; this organism excretes a trihydroxamate siderophore which I have named hydroxamate K This is the first report of a trihydroxamate siderophore produced by root nodule bacteria. Using transposon Tn5 mutagenesis, three mutants of MNF 710were isolated, two of which proved useful. Mutant MNF7101 excretes a variety of hydroxamate siderophores but not hydroxamate K The major product is a trihydroxamate chemical different from hydroxamate K but still structurally similar. Hydroxamate K is a trihydroxamic acid, which appears to consist of three molecules each of N5-hydroxyornithine and 3-hydroxybutyrate, joined together by alternating peptide and ester linkages. The N5-hydroxyornithine is also acetylated to produce the hydroxamic acid groups involved in iron chelation. The major trihydroxamate siderophore produced by mutant MNF7101 carries a positive charge and lacks the acetyl groups found in hydroxamate K MNF7101 therefore appears to lack a specific acetylase activity required for the acetylation of N5 - hydroxyornithine in MNF710. Mutant MNF7102 does not produce hydroxamate K or any other siderophore, but can still transport 55Fe complexed with it. This mutant appears to be defective in either the induction of hydroxamate K biosynthesis, or in a vital biosynthtic gene. Iron deficiency in MNF710 induces the production of three iron-repressible outer membrane proteins (IROMPs) which are probably receptors for iron- siderophore complexes. Cells of MNF710 can transport 55Fe complexed to hydroxamate K, ferrioxamine B, ferrichrome, ferrichrome A, rhodotorulate, citrate and nitrilotriacetic acid (NTA). MNF7102 will also transport these iron-siderophore complexes. Neither MNF710 nor MNF7102 will transport 55Fe complexes to enterobactin or to the pyoverdines from Pseudomonas aeruginosa ATCC15692 or ATCC17400. In R. leguminosarum MNF710 all of these transport systems are ironregulated, but in MNF7102 iron transport mediated by rhodotorulate, citrate and NTA is constitutive. Other root nodule bacteria, like Rhizobium leguminosarum biovar trifolii WU95 and Rhizobium meliloti U45, also transport 55Fe complexed to siderophores they do not produce. In some instances these transport systems are iron-repressible (hydroxamate K, ferrioxamine B and ferrichrome transport in WU95) and in others they are constitutive (rhodotorulate, citrate and NTA transport in both WU95 and U45, and ferrichrome and ferrichrome A transport in U45). When grown with a high iron concentration, R. leguminosarum MNF710 appears to accumulate iron in the form of a bacterioferritin and other iron storage proteins identifiable by Mossbauer spectroscopy. While this is the first indication of iron storage proteins in root nodule bacteria, the roles these may play in the regulation of iron metabolism within the cells are unknown. Biosynthesis of hydroxamate K is iron-regulated, with excretion occurring only when cells of R. leguminosarum MNF710 are iron-stressed. However, induction of the biosynthetic genes for hydroxamate K appears to be influenced by some form of stored iron within the cells. A model has been developed in which hydroxamate K production is controlled by an "immediately available" form of iron which is in dynamic equilibrium with the storage iron. Regardless of the external iron concentration it appears that the iron concentration within the cell must be maintained above a critical level; if not, siderophore synthesis is derepressed and continues until the stored iron concentration is restored. The significant finding from this work is that the free-living form of R. leguminosamm MNF710 has the potential to use three different strategies to ensure survival and growth under iron-deficient conditions - specific siderophore production and uptake, utilization of siderophores produced by other microorganisms, and short-term iron storage for later growth.

The entry of aromatic substrates into Rhizobium leguminosarum biovar viciae MNF300 and Rhizobium leguminosarum biovar trifolii WU95 has been investigated. Two independently-obtained Tn5- insertion mutants of R. leguminosarum bv viciae, MNF3030 and MNF3036, and one Tn5-233-insertion mutant of R. leguminosarum bv trifolii, MNF9013, each unable to grow on 4-hydroxybenzoate as sole carbon source, were used to study the transport of 4- hydroxybenzoate and protocatechuate. All of these mutants were unable to take up labelled 4-hydroxybenzoate but were able to transport labelled protocatechuate. Enzyme assays on the mutants revealed that they lacked 4-hydroxybenzoate hydroxylase (PobA) activity, suggesting that uptake could be via metabolic drag. However, inhibition of 4-hydroxybenzoate uptake in the wild-type using metabolic poisons and competitive inhibitors suggested that uptake was via an active system. To determine how 4-hydroxybenzoate entered these cells, the mutated gene was isolated by complementing the mutants with a gene library of R. leguminosarum by viciae B155. Complementation simultaneously restored both the ability of the mutants to take up and grow on 4-hydroxybenzoate as sole carbon source and the activity of 4-hydroxybenzoate hydroxylase. In addition, the region of DNA flanking Tn5 in one of the mutants was cloned. Hybridization studies revealed that a labelled 4.0 kb DNA fragment isolated from the B155 gene library and able to complement the mutants hybridized to two DNA fragments (2.2 kb and 1.8 kb) of an EcoR1 genomic digest of MNF300 DNA and to two DNA fragments (2.2 kb and 7.6 kb) of EcoR1 genomic digests of the mutants MNF3030 and MNF3036. This pattern indicated that an EcoR1 site present in MNF300 was absent from B155 and that Tn5 (5.8 kb) had inserted into the 1.8 kb fragment in MNF3030 to produce the 7.6 kb band. Using labelled fragments of the 4.0 kb B155 complementing DNA, the 4.0 kb fragment was mapped. A 2.0 kb EcoR1/Pst1 fragment, which encompassed the 1.8 kb fragment in MNF300 into which Tn5 had inserted in the mutants, also had the capacity to complement the mutants MNF3030, MNF3036 and MNF9013, restoring their ability to grown on 4-hydroxybenzoate. The 2.0 kb fragment of B155 and 1.5 kb of DNA sequences flanking Tn5 in MNF3030 were sequenced. The 2.0 kb B155 complementing DNA contained only one complete gene; it showed a high degree of homology to sequences for pobA (4-hydroxybenzoate hydroxylase) found in the database GenBank for Pseudomonas aeruginosa and Acinetobacter calcoaceticus. DNA sequences from the mutant MNF3030 showed that approximately 80% of the pobA gene had been cloned and that Tn5 had inserted directly into the gene. In addition, the amino acid sequences derived from DNA sequences upstream of pobA on the complementary strand revealed homology to pobR, the regulatory gene for pobA in Acinetobacter calcoaceticus. The evidence supporting an active uptake system was found to be explicable in terms of direct effects on 4-hydroxybenzoate hydroxylase, since both hydroxyaromatic analogues of 4-hydroxybenzoate and metabolic poisons also inhibited the enzyme. It is therefore proposed that 4-hydroxybenzoate enters cells of R. leguminosarum by diffusion, either passively or facilitated by a binding protein. Induction of the pathway for 4-hydroxybenzoate catabolism occurs in response to 4-hydroxybenzoate, possibly by the binding of this substrate to the product of pobR, which induces a conformational change enabling the transcription of pobA. The product of pobA, 4-hydroxybenzoate hydroxylase, converts 4-hydroxybenzoate to protocatechuate and the further metabolism of this substrate allows the entry of more 4-hydroxybenzoate. Thus 4-hydroxybenzoate is "pulled" into the cell by metabolic "drag".

Mestrado em Toxicologia
A contaminação do meio ambiente por metais pesados é muito comum, e pode resultar de uma variedade de fontes antropogénicas, entre as quais indústrias metalúrgicas, descargas de efluentes, fundições, minas e aplicação pesticidas e fertilizantes contaminados. A quantidade de metais acumulada no solo, resultante da poluição ambiental, depende da escala de emissão, do transporte do metal a partir da fonte e também da retenção do metal no solo. Os microrganismos do solo são sensíveis a exposições prolongadas a metais pesados. Assim, neste trabalho pretendeu-se isolar Rhizobium de diferentes locais que sofreram contaminação por metais pesados durante longos anos, por forma a determinar a sua tolerância ao cádmio em meio artificial suplementado com diferentes concentrações de CdCl2. Além disso, determinaram-se alguns parâmetros físico-químicos que influenciam a biodisponibilidade dos metais e as concentrações destes elementos nos solos. A variabilidade de tolerâncias verificada entre os isolados de Rhizobium sugere que por detrás da maquinaria metabólica básica, existem outros mecanismos que permitem os isolados mostrarem diferentes respostas ao stresse provocado pelo cádmio. Este facto levou-nos a aprofundar algumas características macromoleculares dos isolados com o objectivo de identificar mecanismos de tolerância e/ou características intrínsecas dos próprios isolados que lhes possam conferir essa tolerância ao metal. De acordo com os resultados obtidos podemos dizer que VE e AS foram os locais que apresentaram concentrações mais elevadas de metais, sendo por isso os mais contaminados. A concentração máxima permitida por lei para o Zn foi largamente ultrapassada em VE, e em AS a concentração verificada esteve muito próxima do limite máximo permitido. Os resultados obtidos evidenciaram que a tolerância dos isolados de Rhizobium foi diferente entre os isolados estudados. De acordo com a CMT os isolados podem ser divididos em 4 categorias: os sensíveis (0-125 μmol.l-1 CdCl2), os medianamente tolerantes (125 e 210 μmol.l-1 CdCl2), os tolerantes (250 e 500 μmol.l-1 CdCl2) e os extremamente tolerantes que toleraram concentrações ≥ 750 μmol.l-1 de cádmio. Os isolados de BTA foram os mais tolerantes uma vez que 46,7% destes toleraram concentrações entre 250 e 500 μmol.l-1 de cádmio, seguindo-se os isolados de AS e SB. No entanto, para concentrações ≥ 750 μmol.l-1, esta tendência apenas foi verificada para BTA. Os isolados de BTB são na sua maioria medianamente tolerantes (70%), enquanto que os isolados de MA são medianamente tolerantes e tolerantes, ao contrário dos isolados de MB que são todos sensíveis. Os perfis polipeptídicos dos isolados de Rhizobium foram bastante diferentes entre si, evidenciando uma grande variabilidade entre os isolados estudados. A presença de cádmio induziu o aumento/diminuição de alguns polipeptídeos, sugerindo uma adaptação das células às condições de stresse. Na sua maioria, as alterações provocadas pelo cádmio corresponderam a diminuições da expressão proteica, sugerindo que o metal afecta o metabolismo básico das células, inibindo passos metabólicos importantes para a sobrevivência e crescimento do rizóbio. vi Os isolados de Rhizobium estudados também apresentaram diferentes perfis plasmídicos. A maioria dos isolados tolerantes apresentaram 2 plasmídeos de pesos moleculares iguais a 485 e 415 kb, sugerindo que a tolerância destes isolados está relacionada com a presença de plasmídeos. Todos os isolados estudados apresentaram concentrações de cádmio intracelular significativamente inferiores ao cádmio adsorvido à parede. Nos isolados sensíveis, medianamente tolerantes e tolerantes a quantidade de cádmio adsorvida foi em muitos casos 100 vezes superior à verificada no interior da célula. Estes resultados sugerem a existência de mecanismos de exclusão ou sequestração extracelular do metal nos isolados medianamente tolerantes e tolerantes. Nos isolados extremamente tolerantes, a quantidade de cádmio adsorvida à parede foi em média 15 vezes superior à concentração de cádmio intracelular, isto significa que muito provavelmente existem mecanismos de sequestração intracelular. Em suma, podemos concluir que a resposta destas bactérias ao stresse provocado pelo cádmio é um fenómeno complexo, estando a tolerância de Rhizobium dependente de vários mecanismos que se desencadeiam paralelamente. Com efeito, parece evidente a existência de um mecanismo de tolerância comum a todos os isolados, que se caracteriza pela presença e/ou indução do aumento de LPS que contribuem para a sequestração extracelular dos metais. Contudo, apesar dos LPS conferirem um certo grau de tolerância, estes não são suficientes para suportar níveis mais elevados de stresse. Assim, a presença de outros mecanismos é essencial. De facto, nos isolados tolerantes a existência de 2 plasmídeos (485 e 415 kb) indica-nos um possível envolvimento destes na tolerância ao cádmio. Por outro lado, nos isolados extremamente tolerantes, podemos inferir que a tolerância se deve à presença de agentes intracelulares que impedem o metal de interferir com os processos metabólicos mais importantes. Os isolados tiveram diferentes origens e estiveram sujeitos a diferentes tipos de contaminação, no entanto estes factores não influenciaram os mecanismos de tolerância ao cádmio.
Environment heavy metals contamination is now widespread. Soils may become contaminated from a variety of anthropogenic sources such as smelters, mining, industry and application of metal-containing pesticides and fertilizers. The amounts of cadmium accumulated in soils from environmental pollution will depend on the scale of emission, the transport of the metal from the source to the site and the retention of the metal once it has reached the soil. The soil microorganisms are very sensitive to long exposures to moderate heavy metal concentrations. Therefore, the present work aimed to isolate Rhizobium from different locations, which have suffered heavy metal contamination since years ago, in order to determine cadmium tolerance in artificial media supplemented with different concentrations of CdCl2. Furthermore, we have determined some physic-chemical parameters that influence the availability of metals, and also their concentrations in soil. The heavy metal tolerance variability among Rhizobium isolates suggests that beyond basic metabolic machinery there also exist variations that allow isolates to display distinct tolerance responses to cadmium stress. This fact, leads us to investigate some macromolecular characteristics from isolates in order to identify tolerance mechanisms and/or intrinsically characteristics that permit isolates to enhance tolerance. In agreement with results we can say that VE and AS presented the highest concentrations of metals, and are therefore the most polluted soils. The maximum permissible concentration to zinc was largely overcome in VE, and in AS the registered concentration was near the limit. Our results showed different tolerances among Rhizobium isolates, according CMT, the isolates can be divided in 4 groups: sensitive (0-125 μmol.l-1 CdCl2); moderately tolerant (125 –210 μmol.l-1 CdCl2); tolerant (250-500 μmol.l-1 CdCl2); and extremely tolerant (≥ 750 μmol.l-1 CdCl2). BTA isolates were the most tolerant ones since 46,7 % of them tolerated concentrations between 250 e 500 μmol.l-1 of CdCl2, followed by AS and SB isolates. However, in concentrations up to 750 μmol.l-1 this tendency was only checked to BTA. The isolates from BTB are mainly moderately tolerant (70%), while MA isolates are moderately tolerant and tolerant, unlike MB isolates that are all sensitive. The electrophoretic patterns of whole-cell soluble proteins of Rhizobium isolates were very distinct, showing high variability between them. Cadmium induced increases/decreases of some polypeptides, which indicates an attempt of cells to adjust to the new adverse conditions. Most of the alterations corresponded to decreases in protein expression, suggesting a deleterious effect on basic cell metabolism, which imposed an inhibition of important metabolic pathways hence inhibiting survival and development of Rhizobium. The studied Rhizobium isolates also presented different plasmid profiles. Most tolerant isolates presented 2 plasmids with molecular weights of 485 and 415 kb, suggesting the possible involvement of these plasmids in cadmium tolerance. All isolates presented smaller quantities of cadmium inside the cell when compared with concentrations adsorbed to cell wall. viii In sensitive, moderately and tolerant isolates the quantity of metal adsorbed was, in many cases 100 times higher than inside cell. These results suggest the existence of exclusion mechanisms or extracelular sequestration of metal in moderately tolerant and tolerant isolates. In extremely tolerant isolates, the amount of cadmium adsorbed to cell wall was, on average, 15 times higher than cadmium absorbed, which probably means that exists intracellular sequestration mechanisms. In summary, we can conclude that the response of these bacteria to cadmium stress is a complex phenomenon, and the tolerance of Rhizobium is dependent of a variety of mechanisms, which unchained parallel. It seems that there is a common tolerance mechanism to all isolates, which is characterized by the presence and /or induction of LPS. These molecules contribute to extracelular sequestration of metals. However, even though LPS conferred a degree of tolerance, they are not sufficient to support higher stress. Thus, the presence of others mechanisms are essential. In fact, in tolerant isolates, smaller quantity of cadmium intracellular and the existence of 2 plasmids (485 and 415 kb), indicated the possible involvement of these plasmids in cadmium tolerance. On the other hand, in extremely tolerant isolates, we can assume tolerance is related with the presence of intracellular agents that prevent metal to interfere with the most metabolic processes. Even though isolates were originated from different locations and had been exposed to different kinds of heavy metal contamination the tolerance mechanisms were not influenced.

Forming symbiosis with legumes and fixing nitrogen, rhizobia can be ranked among the most ecologically significant bacteria in the world. Two features of rhizobia viz. metabolism and host specificity are important determinants of their growth and their ability to form symbiosis with plants. Investigations of these two features are common in studies conducted using laboratory based experimental and computer based bioinformatics methods. Rhizobium is the largest genus of root-nodulating bacteria. Hence, abundant data can be collected by investigating members of this group. 72 isolates of root-nodule bacteria were collected from a field adjacent to Wentworth College, University of York from plants of genus Vicia and Trifolium; 36 from each. Their metabolic assays showed diversity. Analysis of genome data revealed the presence of five contiguous genes in isolates from Vicia, as well as in the reference strain Rlv. 3841; but absent in the isolates from Trifolium. Limited information was available on the five genes. Work in this thesis promotes our understanding of R. leguminosarum genotype and phenotype. Cross nodulation assays showed that all biovar viciae strains retained their ability to form symbiosis plants of genera Vicia, Pisum and Lathyrus. Metabolic study and celC phylogeny showed that clusters of different species - type strains of R. pisi and R. fabae, and type strains of R. phaseoli and R. etli were closely related to each other. Studies on distribution of the five biovar-specific genes showed that they were widely distributed in biovar viciae strains with one exception – the type strain of R. fabae did not have any of five bvs genes, which suggested these bvs genes were not essential for nodulating Vicia plants. Single gene mutation of the five genes in Rlv. 3841 and subsequent competition nodulation tests suggested that inactivation of the bvs4 and bvs5 genes could lead to decreased competitiveness in mutants as compared to wild type on Vicia sativa, Vicia faba and Pisum sativum plants. Complementation of bvs5 gene restored competitiveness. Enzyme assays for the two genes confirmed the activities described in their annotation viz. aliphatic nitrilase for bvs4 and sulfite oxidase for bvs5. Mutation of bvs4 results in a loss of bvs5¬-encoded enzyme activity, indicating that these two genes are probably in an operon. In conclusion, the study shows that relations among rhizobial strains and species can be observed by studying bacterial metabolism. The work hints at a novel competitive mechanism that requires further investigation.

PW7007 cells were grown in media containing 20 µM iron, then transferred to media containing no added iron to monitor the induction of siderophore biosynthetic genes. A rise in the expression of the vbsS gene above basal level was measurable 4 h post-transfer in control cultures. Addition of the potential precursor amino acids L-glutamate and Lomithine or the amino acid L-arginine did not alter the time or extent of induction of expression. Addition of the siderophores vicibactin, desferrioxamine, enterobactin or vicibactin/enterobactin together caused expression of the siderophore synthesis gene to be measurable earlier than in control cultures, at 2 h post-transfer. This early induction is probably due to the siderophores binding all the residual iron in the low-iron cultures and rendering it ‘invisible’ to the cells until suitable outer membrane receptors and uptake proteins can be synthesised. Addition of citrate to the low-iron cultures caused the induction of synthesis genes to occur at the same time as control cultures (4 h post-transfer), but with a reduced level of expression. The citrate-iron complex is constitutively taken up by WSM710 and so can immediately supply some of the cell’s iron needs. Expression of vbsS was repressed by the addition of iron to cells growing in low-iron medium, regardless of whether cells had iron reserves (cells that had been grown in high iron medium prior to transfer into low iron medium) or no iron reserves (cells which had been grown in low iron medium prior to subculture in low iron medium). It is suggested that this is because synthetase activity is non-rate-limiting for the synthesis of vicibactin; cells can still manufacture vicibactin even with a low residual level of synthetase. The synthesis of vicibactin was then studied in planta through monitoring the expression of the labelled siderophore biosynthetic gene (vbsS) in PW7007. Vetch plants were inoculated with WSM710 or PW7007 (which differ only in their siderophoreproducing abilities) and grown in low and high iron hydroponic media. Analysis of the nitrogen and iron concentration in plant tops, the weight of plant tops, the number of nodules per plant and the histological characteristics of those nodules in terms of bacteroid content, revealed no inoculant-based difference between plant nodules with bacteroids from a strain able to synthesise siderophores and those nodulated with a siderophore-negative strain. The only inoculant-based difference detected was a higher level of iron accumulated in nodules containing PW7007 bacteroids compared with nodules occupied by WSM710. Even nodules containing PW7007 from plants grown in low-iron medium had more iron than nodules containing WSM710 from plants grown in high iron medium. The reason for this accumulation of iron in nodules occupied by non-siderophore producing cells is currently unexplained. Nodules harvested from plants grown in high and low iron-media and containing PW7007 were stained to assay activity of the gusA reporter gene (an indicator of vbsS expression). Nodules from both iron-deficient and iron-sufficient plants stained deep blue. However, after the bacteroids were extracted and assayed for specific activity, the level of actual gusA activity was found to be very low, at levels similar to those of uninduced laboratory-grown bacterial cultures. The intense colour development seen in whole nodule stains was probably due to low-level basal expression of (3-glucuronidase leading to product building up within the nodule. This emphasises the need for specific activities to be assayed on extracted bacteroids, so that a true picture of the level of expression of marked genes, such as those involved in siderophore production, can be determined. The lack of expression of the vicibactin biosynthetic gene, vbsS, shows that siderophore synthesis in nodules does not occur even in severely iron-deficient plants.

Chemotaxis is the process by which motile bacteria navigate chemical gradients in order to position themselves in optimum environments for growth and metabolism. Sensory input from both the external environment and the internal cellular environment are sensed by chemotaxis transducers and transduced to a two-component system whose output interacts with the flagellum thereby regulating motility. Chemotaxis has been implicated in establishing the endosymbiotic relationship between the motile alpha-proteobacterium Rhizobium leguminosarum biovar viciae and its host Pisum sativa, the pea plant. An approach combing bioinformatical sequence analysis, molecular genetics, and behavioral analysis was used to characterize the chemotaxis system of R. leguminosarum and determine its contribution to this bacterium s lifestyle. A genome search revealed the presence of two chemotaxis gene clusters, che1 and che2. Homologs of each che cluster are major chemotaxis operons controlling flagellar motility in other bacterial species. For this reason we sought to determine the contribution of each che cluster to chemotaxis in R. leguminosarum. We found that while both che clusters contribute to the regulation of motility, che1 is the major che cluster controlling chemotaxis. Using competitive nodulation assays we determined that che1, but not che2, is essential for competitive nodulation. The major che cluster, che1, encodes a chemotaxis transducer, IcpA-Rl, with a globin coupled sensor domain. Chemotaxis transducers with a globin coupled sensor domain comprise a large class of proteins found in bacteria and archaea. These proteins have been shown to bind heme and sense oxygen and are therefore termed HemATs for heme-binding aerotaxis transducers. However, sequence analysis of IcpA-Rl reveals that it lacks the requisite amino acid residues for heme-binding and is therefore unlikely to sense oxygen. We present evidence that IcpA-Rl is likely an energy transducer and represents a novel function of the globin coupled sensor domain in sensing energy related parameters.

This study describes the use of monoclonal antibodies to investigate molecular components of the Rhizobium cell surface that might be important for symbiotic interactions with the host legume. Components that have been identified include lipopolysaccharide and both membrane-associated and secreted proteins. Differences were observed in the structure and antigenicity of lipopolysaccharide (LPS) from free-living rhizobia compared with that of endosymbiotic bacteroids. Culture pH, oxygen concentration and carbon source were all found to be important factors that could affect the expression of LPS antigens within the nodule. Mutants with altered LPS demonstrated that complete LPS structures are necessary for effective symbiosis. Monoclonal antibody nM 25 identified a protease-sensitive epitope that appeared to be attached to a particular species of LPS macromolecule (identified by MAC 114 antibody). Immunocytochemical localization studies of pea nodule sections revealed that JIM 25 antigen, present on the cell surface, was expressed in infection threads but its expression was low in the symbiotic zone containing mature bacteroids. A 38kDa secreted protein was identified by JIM 24 antibody. Fractionation of nodule extracts by differential centrifugation suggested that the protein was present in the peribacteroid space. A 55kDa membrane protein recognized by MAC 115 antibody proved to be a species-specific marker for R. leguminosarum. The structural gene(s) for this protein are encoded on a 2.8kb EcoRI fragment. Localised mutagenesis of this DNA region with the transposon TnPhoA (which carries a promoter-less gene for alkaline phosphatase) provided evidence that the 55kDa protein was' membraneassociated. However, attempts to "marker-exchange" the transposon-induced mutations from cosmid DNA into the Rhizobium genome were unsuccessful, suggesting that the 55kDa protein is essential for growth of free-living rhizobia. In a DNA region adjoining this 2.8kb EcoRI fragment that encodes the 55kDa protein, a new gene has been described, termed muc. When present on a cosmid the muc gene from R. leguminosarum conferred non-mucoid colony morphology on R. meliloti strain B287. "Marker-exchange" of muc::Tn5 mutations from the plasmid to R. ieguminosarum 8002 (bv. phaseo/i) resulted in derivatives that had lost the ability to nodulate Phaseolus beans. However, marker-exchange into R. leguminosarum B556 (bv. viciae) resulted in mutants that showed no abnormal symbiotic phenotypes on peas. The cosmid carrying the muc and 55kDa protein determinants (pIJ1639) was subjected to saturation transposon mutagenesis with TnPhoA This study revealed several new genes that probably encode membrane-associated or secreted proteins. In some cases gene transcription was dependent on the presence of hesperitin [which is known to be an activator of Rhizobium nodulation (nod) genes]. A 4.6kb EcoRI fragment adjacent to the 2.8kb fragment described above was also found to encode essential functions that prevented the construction of chromosomal mutants by marker-exchange.

Mestrado em Métodos Biomoleculares Avançados
O crescente aumento de metais pesados no solo, devido a poluiçâo ambiental, afecta directamente a comunidade microbiana bem como as relações simbióticas que alguns destes microrganismos estabelecem com leguminosas. Uma dos metais pesados mais prejudiciais e tóxicos é o cádmio. Devido à sua elevada toxicidade, este metal constitui uma grave ameaça, não só para as comunidades bacterianas presentes nos solos, mas também para os outros organismos, uma vez que a transferência de iões metálicos ao longo da cadeia alimentar e uma realidade e um problema sério de saúde pública. Sendo a glutationa um importante agente antioxidante e quelante de metais, este trabalho foi delineado de modo a perceber qual o papel deste tripéptido na tolerância ao cádmio em Rhizobium leguminosarum. Inicialmente alguns isolados de R. leguminosarum foram expostos a diferentes concentrações de cádmio e foi testada a influência da glutationa no seu crescimento. Dos isolados que apresentaram um aumento significativo de tolerância na presença desse tio1 (4 em 5), foram seleccionados dois com graus de tolerância distintos, E20-8 (tolerante) e NI-2 (sensível), para a realizaçâo do restante trabalho experimental. Posteriormente foi avaliado o nivel de stresse oxidativo dos dois isolados, através da determinação dos níveis de GSH e de GSSG intracelulares e das actividades de algumas das principais enzimas envolvidas no stresse oxidativo (SOD, CAT, GR e GPX). Os resultados demonstraram que a exposição ao cádmio induziu stresse oxidativo nas células, o que foi comprovado pelo aumento na concentração de GSSG e pela indução de actividade das enzimas estudadas. A adição de glutationa ao meio de crescimento teve um efeito protector, principalmente no isolado sensível. Os dados obtidos sugerem que o aumento de tolerância ao cádmio em R. leguminosarum não está relacionado com uma maior eficiência do mecanismo antioxidante per si, mas com níveis mais elevados de glutationa intracelular. Mostrou-se ainda que o fornecimento de glutationa ao isolado sensível aumentou a sua tolerância. Tendo já sido demonstrado que a capacidade de aumentar a síntese de glutationa estava relacionada com uma maior tolerância ao cádmio. Assim é plausível concluir que a glutationa desempenhe um papel fundamental na tolerância ao cádmio nesta espécie e que esse papel não se limite á sua actividade como antioxidante. Assim, foi estudado o papel da glutationa como agente quelante nesta espécie e, através da adaptação de um protocolo inicialmente desenvolvido para complexos de fitoquelatinas. As diferentes fracções de cádmio extraídas foram separadas: o cádmio fracamente ligado a parede foi removido por extracção aquosa em banho de ultrasons. O cádmio intracelular foi obtido por extracções sequenciais com tampão HEPES, de modo a manter a estabilidade de possíveis complexos. O cádmio ligado á membrana e a parede celular foi extraído com ácido. O cádmio intracelular foi separado por cromatografia de exclusão molecular e as fracções correspondentes aos picos de cadmio foram colhidas e analisadas por HPLC. Verificou-se que a percentagem de cadmio retido na parede celular era semelhante nos dois isolados, demonstrando a importância deste mecanismo, mas confirmando contudo que esta não pode justificar a diferença de tolerância entre os isolados. O isolado tolerante apresentou uma maior concentração de cadmio intracelular. A análise por HPLC revelou que a glutationa era o principal agente quelante de cádmio em Rhizobium, sendo responsável pela complexação de 75% do cádmio intracelular no isolado tolerante e por 28% no isolado sensivel. A adição de glutationa ao meio de crescimento aumentou a eficácia da complexação de cádmio, passando para 90% no isolado tolerante e para 53% no sensível. Os dados apresentados nesta tese contribuem certamente para a compreensão da tolerância aos metais pesados em bactérias. No entanto, os resultados deste trabalho não se cifram só em respostas, mas também em questões. A razão pela qual a adição de glutationa diminui o cádmio intracelular na estirpe tolerante e qual o destino dos complexos nas células são alguns dos pontos que terão de ser esclarecidos em trabalhos futuros.
The increase in the anthropogenic activities has introduced new types of stresses to soil bacteria communities as well as to symbiotic associations between microbes and legumes. Heavy metal contamination, particularly by cadmium, constitutes a problem. Because of its high toxicity, even at low concentrations, cadmium imposes a serious threat to organisms and food-chain transfer of metal ions has become a major public concern. Since glutathione is an important antioxidant and rnetal chelator. this study was undertaken to determine the role of this tripeptide in cadmium tolerance of Rhizobiurn legurninosarurn. Two strains expressing different degrees of tolerance to cadmium stress were used and the influente of the addition of extracelular GSH to the growth media was determined. Oxidative stress and ROS scavenging enzymes were evaluated, as well as changes in GSH and GSSG levels. Results confirmed that cadmium imposes oxidative stress in Rhizobiurn legurninosarurn, which was characterized by an increase in GSSG formation and an induction of SOD, GPX. GR and CAT activities. Addition of reduced glutathione to the growth media had a protective effect, particularly in the sensitive strain. These findings show that increased tolerance in Rhizobiurn legurninosarum is not related to a higher efficiency of the oxygen scavengers per se, but with higher levels of intracellular glutathione. Furthermore, we demonstrated that glutathione suply to the sensitive strain enhanced tolerance, concluding that glutathione plays a crucial role in cadmium tolerance in this species. In a previous report, it was demonstrated that glutathione played an important role in cadmium detoxification in Rhizobiurn cells. but the molecular role of this tripeptide remained to be elucidated. In this work, we report an efficient extraction of GSH-Cd complexes in two R. legurninosarum strains with different levels of tolerance to cadmium. through a modified protocol, originally developed for metal-phytochelatins complexes. The different sub-cellular Cadmium fractions were extracted: loosely bound cadmium was extracted with water, in an ultrasonic bath; intracellular cadmium was sequentially extracted with HEPES buffer; and wall-bound cadmium was extracted with acid. Intracellular cadmium was separated through size exclusion chromatography, and the amount of cadmium in collected fractions was analysed. Peptide peaks containing the higher cadmium concentrations were analysed by HPLC. In both strains, the same percentage of cadmium was retained in cell walls, hence demonstrating an effective avoidance mechanism, but that was not responsible for the obse~edto lerance differences, displayed by the two strains. Intracellular cadmium accumulation was higher in the tolerant strain and rnetal ions were mainly chelated by small-weight peptides. HPLC analysis revealed that glutathione was the main cadmium chelator in Rhizobirim. being responsible for sequestering 75% of intracellular cadmium in the tolerant strain. The sensitive strain presented a less effective cadmium complexation; only 28% of intracellular cadmium was sequestered by glutathione The presence of glutathione increased the efficiency of this mechanism: the tolerant strain was able to sequester 90% of the intracellular cadmium and the sensitive strain complexed 53%. Our findings add a novel and important aspect to the proposed role of glutathione in heavy metal coping for bacteria. These results can be useful in developing biotechnological strategies for cadmium bioremediation procedures and open novel perspectives for the irnprovement of metal tolerance in soil bacteria. However these findings have raised severa1 questions. The reason why glutathione supply decreases the intracellular cadmium in the tolerant strain and the fate of the complexes within the cell are some of the issues that remain to be elucidated.

Doutoramento em Biologia
A associação simbiótica de plantas leguminosas com bactérias do género Rhizobium é o maior e mais eficiente contribuinte de azoto fixado biologicamente (Somasegaran e Hoben, 1994; Zahran, 1999). No entanto, o constante aumento da poluição em solos agrícolas, nomeadamente a contaminação por metais devido à aplicação de fertilizantes e de lamas, está a tornar-se um problema ambiental cada vez mais preocupante (Alloway, 1995a; Giller et al., 1998; Permina et al., 2006; Thorsen et al., 2009; Wani et al., 2008), influenciando de forma negativa a persistência destas bactérias nos solos agrícolas, assim como a sua eficácia de nodulação (Broos et al., 2005; Wani et al., 2008;. Zhengwei et al., 2005). Desta forma, o estudo dos mecanismos de tolerância de Rhizobium a metais tornou-se uma área de investigação de elevada importância. Com o trabalho apresentado nesta tese pretendeu-se perceber melhor a tolerância Rhizobium leguminosarum ao cádmio (Cd), dando particular atenção a um mecanismo de tolerância previamente descrito em R. leguminosarum (Lima et al., 2006): a complexação intracelular de Cd pelo tripéptido glutationa (GSH). Assim, o principal objectivo deste trabalho foi perceber melhor qual a importância deste mecanismo nos níveis de tolerância de rizóbio ao Cd. Como já tinha sido descrito em trabalhos anteriores (Figueira et al., 2005; Lima et al., 2006), foi possível verificar que a estirpe mais tolerante ao metal apresenta níveis mais elevados de Cd e GSH intracelulares. Demonstrou-se ainda que a tolerância ao Cd está dependente da maior eficiência no mecanismo de complexação observada na estirpe tolerante, logo durante as primeiras 12 h de crescimento. Gomes et al. (2002) verificou que a acumulação de complexos GSH-Cd no citoplasma inibe a entrada de metal na célula. Como neste trabalho se observou um aumento nos níveis de Cd intracelular na estirpe tolerante ao longo do tempo, surgiu a hipótese dos complexos serem excretados para o espaço periplasmático. Os elevados níveis de GSH e de Cd determinados no espaço periplasmático corroboraram esta hipótese. Neste trabalho demonstrou-se ainda que a eficácia do mecanismo de complexação, depende da actividade enzimática de uma isoforma específica de GST, que apresentou um elevado acréscimo de actividade na presença do metal. Desta forma, os resultados desta tese indicam que, a maior tolerância de R. leguminosarum ao Cd, depende da capacidade das estirpes para induzir a síntese de GSH na presença de Cd e, simultaneamente aumentar a actividade enzimática da GST específica, optimizando assim o mecanismo de complexação de Cd intracelular.
The symbiotic association of leguminous plants and rhizobia also has a crucial impact on the nitrogen cycle: estimates are that rhizobial symbioses, with a number greater than 100 important agronomical legumes, contribute nearly half to the annual quantity of nitrogen fixed biologically entering soil ecosystems (Somasegaran and Hoben, 1994; Zahran, 1999). Nevertheless, the permanent increased of metal pollution in agriculture soils, due to the current use of fertilizers and industrial and domestic sludge, is becoming one of the most troublesome environmental problems (Alloway, 1995a; Giller et al., 1998; Permina et al., 2006; Thorsen et al., 2009; Wani et al., 2008), has a negative impact in rhizobia persistence in agricultural soils and its ability to form nitrogen-fixing nodules (Broos et al., 2005; Wani et al., 2008; Zhengwei et al., 2005). So, the study of the tolerance mechanisms of soil bacteria to metals arises as a research area with great importance. In this thesis it was intended to draw attention to the tolerance to Cd of Rhizobium leguminosarum, a specific soil bacteria that establish symbiotic associations with legumes. The present study focused on a particular strategy that Rhizobium leguminosarum possess to tolerate Cd: the intracellular chelation of Cd by the tripetide GSH. Thus the elucidation of the role of GSH as a metal chelating agent in bacteria tolerance to Cd was considered to be the main aim of this work. Higher intracellular Cd and GSH levels were observed in the tolerant strain as it had already been demonstrated by others authors (Figueira et al., 2005; Lima et al., 2006). It was also demonstrated that Rhizobium leguminosarum tolerance to Cd was dependent on the higher efficiency of the chelation mechanism exhibited by the tolerant strain. This mechanism was activated at the beginning of the lag phase (12 h of growth). As the accumulation of GSH-Cd complexes may inhibit intracellular Cd uptake (Gomes et al., 2002), which was not observed in the tolerant strain, the GSHCd complexes were suggested to be transported into the periplasmic space. The results point out this hypothesis as the most viable as high levels of GSH and Cd were found in the periplasmic space but only a small percentage of GSH-Cd complexes were quantified. It was also established that the chelation mechanism occurred in the cytoplasm, and its efficiency appeared to be dependent on the enzymatic activity of a specific GST isoform. So, the ability of the tolerant strain to induce GSH synthesis under Cd exposure and, simultaneously, to increment the activity of a specific GST was point out as the main reasons behind the differences the tolerance to Cd observed between the two strains.

To investigate the role of extracytoplasmic function sigma factors in the Rhizobium-legume symbiosis, the genome sequence of Rhizobium leguminosarum bivoar viciae 3841 was analysed using bioinformatics approaches. In total, twenty sigma factors were identified, including sixteen extracytoplasmic function sigma factors. Twelve of the sixteen ECF sigma factors appeared to be co-transcribed with a regulatory gene. The role of the ECF sigma factors in responding to legume produced compounds was investigated by growing strain 3841 containing various ecf reporter plasmids with pea root exudate. Three ECF sigma factors showed different expression patterns when grown with root exudate compared to minimal medium. One of the three sigma factor aenes. ecfR, which was induced by pea root exudates, was investigated further and was shown to be expressed in pea infection threads. Over-expression of ecfR was lethal to R. leguminosarum 3841, but co-expression with the adjacent gene, asfR, prevented the lethal effect.