Root mucilage from pea and its utilization by rhizosphere bacteria as a sole carbon source.

Plant roots secrete a complex polysaccharide mucilage that may provide a significant source of carbon for microbes that colonize the rhizosphere. High molecular weight mucilage was separated by high-pressure liquid chromatography gel filtration from low molecular weight components of pea root exudate. Purified pea root mucilage generally was similar in sugar and glycosidic linkage composition to mucilage from cowpea, wheat, rice, and maize, but appeared to contain an unusually high amount of material that was similar to arabinogalactan protein. Purified pea mucilage was used as the sole carbon source for growth of several pea rhizosphere bacteria, including Rhizobium leguminosarum 8401 and 4292, Burkholderia cepacia AMMD, and Pseudomonas fluorescens PRA25. These species grew on mucilage to cell densities of three- to 25-fold higher than controls with no added carbon source, with cell densities of 1 to 15% of those obtained on an equal weight of glucose. Micromolar concentrations of nod gene-inducing flavonoids specifically stimulated mucilage-dependent growth of R. leguminosarum 8401 to levels almost equaling the glucose controls. R. leguminosarum 8401 was able to hydrolyze p-nitrophenyl glycosides of various sugars and partially utilize a number of purified plant polysaccharides as sole carbon sources, indicating that R. leguminosarum 8401 can make an unexpected variety of carbohydrases, in accordance with its ability to extensively utilize pea root mucilage.

Plants secrete substances that mimic bacterial N-acyl homoserine lactone signal activities and affect population density-dependent behaviors in associated bacteria.

In gram-negative bacteria, many important changes in gene expression and behavior are regulated in a population density-dependent fashion by N-acyl homoserine lactone (AHL) signal molecules. Exudates from pea (Pisum sativum) seedlings were found to contain several separable activities that mimicked AHL signals in well-characterized bacterial reporter strains, stimulating AHL-regulated behaviors in some strains while inhibiting such behaviors in others. The chemical nature of the active mimic compounds is currently unknown, but all extracted differently into organic solvents than common bacterial AHLs. Various species of higher plants in addition to pea were found to secrete AHL mimic activities. The AHL signal-mimic compounds could prove to be important in determining the outcome of interactions between higher plants and a diversity of pathogenic, symbiotic, and saprophytic bacteria.

A rapid bioassay for screening rhizosphere microorganisms for their ability to induce systemic resistance.

ABSTRACT We developed a rapid and miniaturized bioassay for screening large numbers of rhizosphere microorganisms for their ability to induce systemic resistance to bacterial leaf spot of radish caused by Xanthomonas campestris pv. armoraciae. In this bioassay, Pantoea agglomerans strain E278Ar controlled symptoms of disease as effectively as 2,6-dichloroisonicotinic acid when applied to the roots of seedlings produced in growth pouches in a soilless system. E278Ar essentially did not migrate from seedling roots to the foliage. This suggests that induction of systemic resistance could best explain the observed reduction in disease severity. Three mini-Tn5Km-induced mutants of strain E278Ar were isolated that had lost the ability to induce resistance. The bioassay also was used to demonstrate that the fungal biocontrol agent Trichoderma hamatum strain 382 induces systemic resistance in radish. The bioassay required only 14 to 18 days from seeding until rating for disease severity, which is 10 to 14 days less than earlier bioassays.

Tn5-induced and spontaneous switching of Sinorhizobium meliloti to faster-swarming behavior.

Tn5 mutants of Sinorhizobium meliloti RMB7201 which swarmed 1.5 to 2. 5 times faster than the parental strain in semisolid agar, moist sand, and viscous liquid were identified. These faster-swarming (FS) mutants outgrew the wild type 30- to 40-fold within 2 days in mixed swarm colonies. The FS mutants survived and grew as well as or better than the wild type under all of the circumstances tested, except in a soil matrix subjected to air drying. Exopolysaccharide (EPS) synthesis was reduced in each of the FS mutants when they were grown on defined succinate-nitrate medium, but the extent of reduction was different for each. It appears that FS behavior likely results from a modest, general derepression of motility involving an increased proportion of motile and flagellated cells and an increased average number of flagella per cell and increased average flagellar length. Spontaneous FS variants of RMB7201 were obtained at a frequency of about 1 per 10,000 to 20,000 cells by either enrichment from the periphery of swarm colonies or screening of colonies for reduced EPS synthesis on succinate-nitrate plates. The spontaneous FS variants and Tn5 FS mutants were symbiotically effective and competitive in alfalfa nodulation. Reversion of FS variants to wild-type behavior was sporadic, indicating that reversion is affected by unidentified environmental factors. Based on phenotypic and molecular differences between individual FS variants and mutants, it appears that there may be multiple genetic configurations that result in FS behavior in RMB7201. The facile isolation of spontaneous FS variants of Escherichia coli and Pseudomonas aeruginosa indicates that switching to FS behavior may be fairly common among bacterial species. The substantial growth advantage of FS mutants and variants wherever nutrient gradients exist suggests that switching to FS forms may be an important behavioral adaptation in natural environments.

Relationships between C4 dicarboxylic acid transport and chemotaxis in Rhizobium meliloti.

The relationship between chemotaxis and transport of C4 dicarboxylic acids was analyzed with Rhizobium meliloti dct mutants defective in one or all of the genes required for dicarboxylic acid transport. Succinate, malate, and fumarate were moderately potent chemoattractants for wild-type R. meliloti and appeared to share a common chemoreceptor. While dicarboxylate transport is inducible, taxis to succinate was shown to be constitutive. Mutations in the dctA and dctB genes both resulted in the reduction, but not elimination, of chemotactic responses to succinate, indicating that transport via DctA or chemosensing via DctB is not essential for C4 dicarboxylate taxis, although they appear to contribute to it. Mutations in dctD and rpoN genes did not affect taxis to succinate. Aspartate, which is also transported by the dicarboxylate transport system, elicited strong chemotactic responses via a chemoreceptor distinct from the succinate-malate-fumarate receptor. Taxis to aspartate was unaltered in dctA and dctB mutants but was considerably reduced in both dctD and rpoN mutants, indicating that aspartate taxis is strongly dependent on elements responsible for transcriptional activation of dctA. Methylation and methanol release experiments failed to show a significant increase in methyl esterification of R. meliloti proteins in response to any of the attractants tested.

Role of divalent cations in the subunit associations of complex flagella from Rhizobium meliloti.

Rhizobium meliloti, a symbiotic, nitrogen-fixing soil bacterium with complex flagella, as well as other members of the family Rhizobiaceae, rapidly lost motility when suspended in buffers lacking divalent cations but retained good motility in buffers containing calcium, magnesium, barium, or strontium. Loss of motility was associated with loss of flagella from the cells. Analysis of flagella by sedimentation, gel electrophoresis, and electron microscopy revealed that removal of divalent cations from the complex flagella of R. meliloti resulted in extensive dissociation of the flagellar filaments into low-molecular-weight subunits. Accordingly, divalent cations such as calcium and magnesium that are normally present at high concentrations in the soil solution may be crucial to the assembly and rigidity of complex flagella.

Chemotaxis of Rhizobium meliloti towards Nodulation Gene-Inducing Compounds from Alfalfa Roots.

Luteolin, a flavone present in seed exudates of alfalfa, induces nodulation genes (nod) in Rhizobium meliloti and also serves as a biochemically specific chemoattractant for the bacterium. The present work shows that R. meliloti RCR2011 is capable of very similar chemotactic responses towards 4',7-dihydroxyflavone, 4',7-Dihydroxyflavanone, and 4,4'-dihydroxy-2-methoxychalcone, the three principal nod gene inducers secreted by alfalfa roots. Chemotactic responses to the root-secreted nod inducers in capillary assays were usually two- to four-fold above background and, for the flavone and flavonone, occurred at concentrations lower than those required for half-maximal induction of the nodABC genes. Complementation experiments indicated that the lack of chemotactic responsiveness to luteolin seen in nodD1 and nodA mutants of R. meliloti was not due to mutations in the nod genes, as previously thought. Thus, while nod gene induction and flavonoid chemotaxis have the same biochemical specificity, these two functions appear to have independent receptors or transduction pathways. The wild-type strain was found to suffer selective, spontaneous loss of chemotaxis towards flavonoids during laboratory subculture.

Growth and Movement of Spot Inoculated Rhizobium meliloti on the Root Surface of Alfalfa.

Inoculum droplets of approximately 10 nanoliter volume and containing about 10 Rhizobium meliloti cells were placed onto the root surface of alfalfa seedlings in plastic growth pouches at either the root tip, the position of the smallest emergent root hairs, or at a site midway between these points. The droplets were initially confined to an area of about 0.2 square millimeter at the point of application. By 48 and 96 hours after inoculation, the inoculum bacteria and their progeny were distributed over several centimeters of the root between the initial site of deposition and the growing root tip, reaching densities of 10(3) to 10(4) bacteria per centimeter near the site of initial deposition and decreasing exponentially from that point toward the root tip. Graphite particles deposited on the root surface close to the growing tip were similarly distributed along the root length by 48 and 96 hours, suggesting that passive displacement by root cell elongation was primarily responsible for the spread of bacteria. A nonmotile mutant of R. meliloti colonized alfalfa roots to the same extent as the wild type and was usually distributed in the same manner, indicating that bacterial motility contributed little under these conditions to long distance spread of the bacteria. However, when applied in low numbers, R. meliloti mutants defective in motility or chemotaxis were considerably less efficient in initiating nodules near the point of inoculation than the wild type. This implies that motility and/or chemotaxis contribute significantly to local exploration for suitable infection sites. Almost all nodules on the primary root formed within a few millimeters of the spot-inoculation site, indicating that, under our experimental conditions, movement and multiplication of R. meliloti on the root surface were not sufficient to maintain an adequate population in the infectible region of the root during root growth.

Rhizobium meliloti exopolysaccharide Mutants Elicit Feedback Regulation of Nodule Formation in Alfalfa.

Nodule formation by wild-type Rhizobium meliloti is strongly suppressed in younger parts of alfalfa (Medicago sativum L.) root systems as a feedback response to development of the first nodules (G Caetano-Anollés, WD Bauer [1988] Planta 175: 546-557). Mutants of R. meliloti deficient in exopolysaccharide synthesis can induce the formation of organized nodular structures (pseudonodules) on alfalfa roots but are defective in their ability to invade and multiply within host tissues. The formation of empty pseudonodules by exo mutants was found to elicit a feedback suppression of nodule formation similar to that elicited by the wild-type bacteria. Inoculation of an exo mutant onto one side of a split-root system 24 hours before inoculation of the second side with wild-type cells suppressed wild-type nodule formation on the second side in proportion to the extent of pseudonodule formation by the exo mutants. The formation of pseudonodules is thus sufficient to elicit systemic feedback control of nodulation in the host root system: infection thread development and internal proliferation of the bacteria are not required for elicitation of feedback. Pseudonodule formation by the exo mutants was found to be strongly suppressed in split-root systems by prior inoculation on the opposite side with the wild type. Thus, feedback control elicited by the wild-type inhibits Rhizobium-induced redifferentiation of host root cells.

Host-specificity mutants of Rhizobium meliloti have additive effects in situ on initiation of alfalfa nodules.

Pairs of Rhizobium meliloti nod mutants were co-inoculated onto alfalfa (Medicago saliva L.) roots to determine whether one nod mutant could correct, in situ, for defects in nodule initiation of another nod mutant. None of the Tn5 or nod deletion mutants were able to help each other form nodules when co-inoculated together in the absence of the wild-type. However, as previously observed, individual nod mutants significantly increased nodule initiation by low dosages of co-inoculated wild-type cells. Thus, nod mutants do produce certain signal substances or other factors which overcome limits to nodule initiation by the wild-type. When pairs of nod mutants were co-inoculated together with the wild-type, the stimulation of nodulation provided by individual nodABC mutants was not additive. However, clearly additive or synergistic stimulation was observed between pairs of mutants with a defective host-specificity gene (nodE, nodF, or nodH). Each pair of host-specificity mutants stimulated first nodule formation to nearly the maximum levels obtainable with high dosages of the wild-type. Mutant bacteria were recovered from only about 10% of these nodules, whereas the co-inoculated wild-type was present in all these nodules and substantially outnumbered mutant bacteria in nodules occupied by both. Thus, these mutant co-inoculants appeared to help their parent in situ even though they could not help each other. Sterile culture filtrates from wild-type cells stimulated nodule initiation by low dosages of the wild-type, but only when a host-specificity mutant was also present. The results from our studies seem consistent with the possibility that pairs of host-specificity mutants are able to help the wild-type initiate nodule formation by sustained production of complementary signals required for induction of symbiotic host responses.

When does the self-regulatory response elicited in soybean root after inoculation occur?

The inoculation of soybean (Glycine max L.) roots with Bradyrhizobium japonicum produces a regulatory response that inhibits nodulation in the younger regions of the roots. By exposing the soybean roots to live homologous bacteria for only a short period of time, the question of whether or not early interactions of rhizobia with root cells, prior to infection, elicit this regulatory response has been explored. B. japonicum cells mixed with infective bacteriophages were applied to the roots and then 6 or 24 hours later roots were again inoculated with phage-resistant rhizobia. Mixing of the rhizobia and bacteriophages caused bacterial lysis in 6 to 8 hours and allowed the bacteria to act as live symbionts on the root for only a few hours. However, the interaction of live homologous bacteria with the soybean roots for a few hours did not cause inhibition of nodulation in the younger regions of the roots. Results of these experiments indicate that the self-regulatory response in soybean is not rapidly produced by the early, pre-infection, interactions between rhizobia and the root cells.

Chemotaxis of Rhizobium meliloti to the plant flavone luteolin requires functional nodulation genes.

Luteolin is a phenolic compound from plants that acts as a potent and specific inducer of nodABC gene expression in Rhizobium meliloti. We have found that R. meliloti RCR2011 exhibits positive chemotaxis towards luteolin. A maximum chemotactic response was observed at 10(-8) M. Two closely related flavonoids, naringenin and apigenin, were not chemoattractants. The presence of naringenin but not apigenin abolished chemotaxis of R. meliloti towards luteolin. A large deletion in the nif-nod region of the symbiotic megaplasmid eliminated all chemotactic response to luteolin but did not affect general chemotaxis, as indicated by swarm size on semisoft agar plates and chemotaxis towards proline in capillary tubes. Transposon Tn5 mutations in nodD, nodA, or nodC selectively abolished the chemotactic response of R. meliloti to luteolin. Agrobacterium tumefaciens GMI9050, a derivative of the C58 wild type lacking a Ti plasmid, responded chemotactically to 10(-8) M luteolin. The introduction of a 290-kilobase nif-nod-containing sequence of DNA from R. meliloti into A. tumefaciens GMI9050 enabled the recipient to respond to luteolin at concentrations peaking at 10(-6) M as well as at concentrations peaking at 10(-8) M. The response of A. tumefaciens GMI9050 to luteolin was also abolished by the presence of naringenin.

Efficiency of nodule initiation in cowpea and soybean.

When serial dilutions of a suspension of Bradyrhizobium japonicum strain 138 were inoculated onto both soybean and cowpea roots, the formation of nodules in the initially susceptible region of the roots of both hosts was found to be linearly dependent on the log of the inoculum dosage until an optimum dosage was reached. Approximately 30- to 100-fold higher dosages were required to elicit half-maximal nodulation on cowpea than on soybean in the initially susceptible zone of the root. However, at optimal dosages, about six times as many nodules formed in this region on cowpea roots than on soybean roots. There was no appreciable difference in the apparent rate of nodule initiation on these two hosts nor in the number of inoculum bacteria in contact with the root. These results are consistent with the possibility that cowpea roots have a substantially higher threshold of response to symbiotic signals from the bacteria than do soybean roots. Storage of B. japonicum cells in distilled water for several weeks did not affect their viability or efficiency of nodule initiation on soybean. However, the nodulation efficiency of these same cells on cowpea diminished markedly over a 2 week period. These differential effects of water storage indicate that at least some aspects of signal production by the bacteria during nodule initiation are different on the two hosts. Mutants of B. japonicum 138 defective in synthesis of soybean lectin binding polysaccharide were defective in their efficiency of nodule initiation on soybean but not on cowpea. These results also suggest that B. japonicum may produce different substances to initiate nodules on these two hosts.

Role of Motility and Chemotaxis in Efficiency of Nodulation by Rhizobium meliloti.

Spontaneous mutants of Rhizobium meliloti L5-30 defective in motility or chemotaxis were isolated and compared against the parent with respect to symbiotic competence. Each of the mutants was able to generate normal nodules on the host plant alfalfa (Medicago sativa), but had slightly delayed nodule formation, diminished nodulation in the initially susceptible region of the host root, and relatively low representation in nodules following co-inoculation with equal numbers of the parent. When inoculated in growth pouches with increasing dosages of the parental strain, the number of nodules formed in the initially susceptible region of the root increased sigmoidally, with an optimum concentration of about 10(5) to 10(6) bacteria/plant. The dose-response behavior of the nonmotile and nonchemotactic mutants was similar, but they required 10- to 30-fold higher concentrations of bacteria to generate the same number of nodules. The distribution frequencies of nodules at different positions along the primary root were very similar for the mutants and parent, indicating that reduced nodulation by the mutants in dose-response experiments probably reflects reduced efficiency of nodule initiation rather than developmentally delayed nodule initiation. The number of bacteria that firmly adsorbed to the host root surface during several hours of incubation was 5- to 20-fold greater for the parent than the mutants. The mutants were also somewhat less effective than their parent as competitors in root adsorption assays. It appears that motility and chemotaxis are quantitatively important traits that facilitate the initial contact and adsorption of symbiotic rhizobia to the host root surface, increase the efficiency of nodule initiation, and increase the rate of infection development.

Enhanced nodule initiation on alfalfa by wild-typeRhizobium meliloti co-inoculated withnod gene mutants and other bacteria.

Nodule formation on alfalfa (Medicago sativa L.) roots was determined at different inoculum dosages for wild-typeRhizobium meliloti strain RCR2011 and for various mutant derivatives with altered nodulation behavior. The number of nodules formed on the whole length of the primary roots was essentially constant regardless of initial inoculum dosage or subsequent bacterial multiplication, indicative of homeostatic regulation of total nodule number. In contrast, the number of nodules formed in just the initially susceptible region of these roots was sigmoidally dependent on the number of wild-type bacteria added, increasing rapidly at dosages above 5·10(3) bacteria/plant. This behavior indicates the possible existence of a threshold barrier to nodule initiation in the host which the bacteria must overcome. When low dosages of the parent (10(3) cells/plant) were co-inoculated with 10(6) cells/plant of mutants lacking functionalnodA, nodC, nodE, nodF ornodH genes, nodule initiation was increased 10- to 30-fold. Analysis of nodule occupancy indicated that these mutants were able to help the parent (wild-type) strain initiate nodules without themselves occupying the nodules. Co-inoculation withR. trifolii orAgrobacterium tumefaciens cured of its Ti plasmid also markedly stimulated nodule initiation by theR. meliloti parent strain. Introduction of a segment of the symbiotic megaplasmid fromR. meliloti intoA. tumefaciens abolished this stimulation.Bradyrhizobium japonicum and a chromosomal Tn5 nod(-) mutant ofR. meliloti did not significantly stimulate nodule initiation when co-inoculated with wild-typeR. meliloti. These results indicate that certainnod gene mutants and members of theRhizobiaceae may produce extracellular "signals" that supplement the ability of wild-typeR. meliloti cells to induce crucial responses in the host.

Feedback regulation of nodule formation in alfalfa.

When high dosages of wild-type Rhizobium meliloti RCR2011 were inoculated at two different times, 24 h apart, onto either the primary roots of alfalfa (Medicago sativa L.) seedlings or onto lateral roots on opposite sides of a split-root system, the number of nodules generated by the second inoculum was much smaller than the number generated by the first inoculum. These results provide evidence that alfalfa has an active, systemic mechanism for feedback control of nodulation. Non-nodulating mutants and delayed, weakly nodulating mutants did not elicit a discernable suppression of nodulation by subsequently inoculated wild-type cells. An appreciable number of Rhizobium infections thus seem required to elicit the suppressive response. Mutants in nodulation regions IIb and IIa nodulated extensively in the initially susceptible region of the root, but nodule initiation by these mutants was 100-1000 times less efficient, respectively, than the parent. Nodules formed by these mutants emerged 1 d later than normal. The IIb mutants elicited a relatively strong suppression of nodulation in younger parts of the root, but region-IIa mutants elicited only a weak response. These results indicate that elicitation of the regulatory response need not be proportional to nodule formation and imply that genes in region IIa play an important role in elicitation. At high dosages, the region-II mutants induced the development of thick, short roots in a considerably higher percentage of plants than the wild-type bacteria. Nodules generated by wild-type isolates and region-II mutants did not emerge in strict acropetal sequence, probably because some infections developed more slowly than others. Prior exposure of the root to non-nodulating mutants resulted in nodulation by the parent in regions of the root otherwise too mature to be susceptible, indicating that exposure to these mutants may affect the sequence of root development.

Transposon Mutants of Bradyrhizobium japonicum Altered in Attachment to Host Roots.

Transposon mutants of Bradyrhizobium japonicum 110 ARS were produced and screened for changes in attachment ability. Mutant CFK4 produced twice as many piliated cells, attached in 2.5-fold-higher numbers to soybean root segments, and colonized roots in about 2-fold-higher numbers than did the parental strain, 110 ARS. Mutants CFK35 and CFK38 were reduced in their attachment about 2-fold and 3.5-fold, respectively. This corresponded to reductions in piliated cells in their populations, reduced reaction with anti-pilus antiserum, and reduced hydrophobic attachment. Mutants CFK4 and CFK38 nodulated soybeans at about the same level as the parent strain, but CFK35 induced only pseudonodules. Two-dimensional gel analyses of the proteins from the mutants showed relatively few changes in proteins.

Nitrate induced regulation of nodule formation in soybean.

Nodule formation was inhibited by exposing soybean plants to nitrate in plastic growth pouches. Exposure to 15 millimolar nitrate resulted in a 2.5-fold decrease in the number of nodules formed in the region of the primary root above the mark made at the time of inoculation to indicate the position of the root tip. Serial section analysis of Bradyrhizobium infections in this region revealed that infection initiation was inhibited approximately 3-fold by exposure to nitrate. Both initial cortical cell divisions and infection thread formation were inhibited. If exposure to nitrate was delayed for 18 hours after the time of inoculation, inhibition was much reduced. This indicates that most of the nitrate-sensitive events of infection were functionally complete within less than 18 hours. Exposure to nitrate for periods of 4 to 24 hours after inoculation, followed by transfer to no-nitrate conditions for the remainder of the time, did not result in substantial inhibition of nodule number. This indicates that the effects of nitrate on infection initiation can be almost entirely reversible. Split towel pouches were used to physically separate portions of the primary root exposed to nitrate and portions of the root exposed to rhizobia. In experiments where nitrate was applied either below or above the inoculated region of the primary root, the degree of inhibition of nodulation was not correlated with either the external concentration of nitrate in contact with root cells undergoing infection or with the internal concentration of nitrate in the infectible region of the root. These results indicate that nitrate itself may not directly inhibit infection initiation or induce host regulatory responses.

Role of Pili (Fimbriae) in Attachment of Bradyrhizobium japonicum to Soybean Roots.

Pili (fimbriae) were observed on cells of each of the five strains of Bradyrhizobium japonicum and the one strain of Rhizobium trifolii examined. Pili on B. japonicum were about 4 nm in diameter and polarly expressed. Piliated cells were estimated by transmission electron microscopy and hydrophobic attachment to polystyrene to constitute only a small percentage of the total population. The proportion of piliated cells in these populations was dependent on culture age in some strains. Piliated B. japonicum cells were selectively and quantitatively removed from suspension when cultures were incubated with either soybean roots or hydrophobic plastic surfaces, indicating that pili were involved in the attachment of the bacteria to these surfaces. Pili from B. japonicum 110 ARS were purified and found to have a subunit molecular weight of approximately 21,000. Treatment of B. japonicum suspensions with antiserum against the isolated pili reduced attachment to soybean roots by about 90% and nodulation by about 80%. Pili appear to be important mediators of attachment of B. japonicum to soybean roots under the conditions examined.