What makes Pseudomonas bacteria rhizosphere competent?

The rhizosphere, a term introduced by 20), is defined as the area influenced by the root system. In comparison with root-free soil, the rhizosphere forms a nutrient-rich niche for bacteria as a result of exudation of compounds such as organic acids, sugars and amino acids. Mites, nematodes and microorganisms are found at elevated levels in the rhizosphere (4). The enormous impact that soil bacteria can have on plants is illustrated by an experiment in which gnotobiotic wheat seedlings were each inoculated with one of 150 fluorescent Pseudomonas strains isolated from field-grown wheat. Of these isolates, 40% stimulated root growth and 40% inhibited root growth, whereas only 20% showed no effect (7). The importance of rhizosphere competence or root colonization in beneficial plant–microbe interactions is underlined by two studies. 33) showed that inadequate colonization leads to decreased biocontrol activity. 6) reported an inverse correlation between the population size of the biocontrol strain Pseudomonas fluorescens 2-79 and the number of take-all lesions observed on wheat. Effective root colonization can contribute to the following processes: (i) causing disease when pathogens are involved; (ii) disease control by microbes that produce antifungal metabolites (AFMs); (iii) phytostimulation, i.e. plant growth promotion, usually by the production of phytohormones; (iv) biofertilization, i.e. the process by which microbes increase the availability of nutrients, such as nitrogen, phosphate and micronutrients, to the plant; (v) as more and more bacteria are isolated that are able to degrade potentially hazardous organic chemicals in nature, colonization can also play a prominent role in bioremediation promoted by plants, a process called phytoremediation. Despite its important role in these processes, the bacterial traits involved in the process of rhizosphere competence are still poorly understood. In many applications of beneficial root-colonizing microbes, the cells are applied to the seed (24). We have coated germinated seedlings of various crops with cells of the aggressively colonizing Pseudomonas fluorescens biocontrol strain WCS365 (12). After growth of the plants for 7–10 days, most bacteria are found on and just below the seed, whereas the bacterial concentration rapidly decreases along the root becoming approximately 100-fold lower at the root tip (29; 40; 34; 5; 9). Microscopic inspection of the roots shows that many bacteria are present in microcolonies (9). These are assumed to be ideal sites for processes that require a high density of bacteria (18), such as control of plant diseases by the production of AFMs (31; 9) and bacterial conjugation (17). In order to identify traits and genes involved in root colonization, two approaches can be distinguished. (i) Mutants are isolated in traits expected to be involved in colonization. Subsequently, the colonizing abilities of wild type and mutants are compared. (ii) Randomly obtained mutants are screened for their ability to colonize. Subsequently, the responsible gene and trait are identified. Using the latter approach, we reasoned that the easiest way to detect colonization mutants was to use the parental strain as an internal control. Surprisingly, it appeared that quite a number of the isolated mutants were only impaired when tested in competition. However, it should be realized that competition mimics the situation in raw soil, which contains approximately 108 bacteria g−1. As it is preferable for many applications that bacteria colonize as much of the root as possible, successful colonization is often evaluated as the ability to reach the root tip after seed inoculation. Although we realize that, for most applications, only the rhizosphere competence in raw soil counts, we found it convenient to use a gnotobiotic system, which has the advantage over raw soil of higher numbers of bacteria and a higher reproducibility. So far, interesting results observed in the gnotobiotic system appear to occur also in raw soil. The most severely impaired Pseudomonas colonization mutants appeared to be non-motile mutants. This role of flagella was observed in various wild-type strains (38; 34), in sand (34; 13) as well as in soil (38) and on both potato (38) and tomato (34; 13). It should be noted that, in other laboratories, no role for flagella in the colonization of wheat (21) and soybean (32) has been detected. Mutants impaired in the synthesis of the O-antigen of lipopolysaccharide (LPS) are impaired in rhizosphere competence, when tested both alone in soil on potato (39) and tomato (13) as well as in competition on tomato (34; 13). Mutants that lack the O-antigen of LPS have a slightly decreased growth rate compared with their respective parental strains (13). Therefore, the inability of these mutants to colonize may result from either the lack of O-antigen or their lower growth rate. The observation that the growth rate of one colonization mutant with a shortened O-antigen (13) is normal indicates that the O-antigen certainly plays a role in rhizosphere competence independent of the growth rate. The role of growth rate in colonization was confirmed by the identification of many growth rate-impaired mutants among colonization mutants obtained by random screening (13). The ability to synthesize amino acids and vitamin B1 was shown by 35) to be essential for root colonization by P. fluorescens WCS365. This indicates that the amount of amino acids exuded by the root is too low to complement the mutations physiologically. Mutants of P. fluorescens WCS365 impaired in their ability to grow on root exudate sugars behave like the wild type in competitive colonization. In contrast, a mutant impaired in the utilization of organic acids shows a strongly reduced competitive colonization ability. These results, as well as the observation that organic acids are the major carbon source found in tomato exudate, indicate that utilization of organic acids is the nutritional basis for tomato root colonization by P. fluorescens WCS365 (B. J. J. Lugtenberg et al., in preparation). An early step in the establishment of a plant–bacterium interaction is attachment of cells to the root. Fimbriae were reported to play a role in the attachment of P. fluorescens 2-79 to corn roots (37). In our root colonization studies with several Pseudomonas wild-type strains, we have never been able to establish a role of attachment in colonization (unpublished results; 9). Cell-surface proteins that have been shown to be involved in the attachment of Pseudomonas spp. to plant roots include the outer membrane protein OprF of P. fluorescens OE 28.3 (27) and an agglutinin isolated from P. putida strain Corvallis that mediates agglutination of bacterial cells to a glycoprotein on the plant root (2). This bacterial protein encoded by the aggA locus was reported to be involved in the adherence to roots and in competitive colonization with the parental strain on the roots of bean (2), potato, tomato and grass (19). Although agglutinin plays a major role in the adherence and colonization abilities of P. putida strain Corvallis to bean and cucumber, the role of agglutinins is not general for all biocontrol strains. No agglutination-dependent adherence and root colonization could be demonstrated for 30 different Pseudomonas isolates on tomato, potato and grasses (19). The attachment and virulence of Agrobacterium tumefaciens was also impaired in mutants in a periplasmic binding protein-dependent (ABC) transport system (26). After a laborious screening procedure in which approximately 2200 Tn5 or Tn5lacZ mutants of P. fluorescens WCS365 were screened, 16 competitive colonization mutants remained that were not impaired in any of the previously mentioned colonization traits. The notion that growth rate contributes to colonization (34) was underlined by the observation that 5 out of the 16 isolated colonization mutants had a slightly decreased growth rate. The remaining 11 rhizosphere competence mutants shared the following properties. (i) They were not impaired in colonization when tested alone, indicating that they are not supersensitive to compounds exuded by the root (13). (ii) They were severely impaired in competitive root colonization on all tested crops including tomato, wheat and radish, indicating broad-host-range mutations. One of the transposon insertions was found in the nuo-4 gene encoding a subunit of NADH ubiquinone oxidoreductase. This protein is part of the respiratory chain and is involved in the generation of the proton motive force, which can be used as an energy source for the synthesis of ATP, for active transport of various nutrients and for rotation of flagella (3). A second NADH dehydrogenase (encoded by ndh ) is downregulated by low oxygen. We observed that the nuo-4 mutant grows more slowly in low oxygen than its wild-type P. fluorescens WCS365. Therefore, it is tempting to speculate that the rhizosphere environment is microaerophilic. A two-component system is also involved in the colonization ability of P. fluorescens WCS365. A transposon insertion has been shown to be present in the sensor kinase member (ColS) of a ColS/ColR two-component system (14). No evidence was found that the ColR/ColS two-component system was involved in the suspected colonization traits chemotaxis and transport of exudate compounds (14). We now have preliminary indications that some exudate compounds are able to induce the colR/colS promoter and that this system may be related to the cell's internal pH (L. Dekkers, unpublished results). The action of a site-specific recombinase homologue (Sss) also appears to be essential for colonization (15). These enzymes promote conservative reciprocal recombination (which does not require DNA synthesis) between two small (approximately 15 bp) homologous DNA sequences. Depending on the orientation of these two sequences, this will lead either to inversion or to an excision of the DNA fragment situated between these sites (30). Colony phase variation brought about by site-specific recombinases is able to influence the synthesis of fimbriae (1; 23), lipopolysaccharides (11), surface lipoprotein antigens (25) and flagella (41). DNA rearrangements generate subpopulations that enable bacteria to adjust to sudden environmental changes (16). Using the strategy of generating subpopulations, a bacterial population is able to respond at all times quickly and adequately to environmental changes, even if only a few cells of a subpopulation are surviving. We interpret the impaired colonization ability of P. fluorescens sss (15) mutant PCL1233 as being locked in a subpopulation that is not rhizosphere competent (15). Understanding traits and analysing genes involved in rhizosphere competence is crucial for the success of applications for beneficial purposes, such as the delivery of green chemicals for the control of plant diseases, for phytostimulation and bioremediation of the environment through the use of plants. Using the mutant approach, it has become clear that many bacterial genes are involved in rhizosphere competence. Among the identified rhizosphere competence traits are traits from which a role could be expected, such as growth rate, a surface polysaccharide, the biosyntheses of crucial components such as amino acids and utilization of the major exudate carbon source. Interestingly, a beginning has been made with the identification of traits from which it was less likely to predict a role in rhizosphere competence, such as NADH dehydrogenase, enzymes involved in DNA rearrangements and in the maintenance of the internal pH. The picture emerges that the study of colonization genes identifies many genes that are a prerequisite for rhizosphere competence but do not play a role in the usual laboratory media. As colonization is often considered to be the limiting step in biocontrol, we introduced the colonization operon containing the sss homologous operon into the other wild-type P. fluorescens strains F113 and WCS307. In both cases, introduction leads to an enhanced colonization ability. As colonization is often the limiting factor in biocontrol, the new constructs were also tested for their ability to control tomato foot rot caused by the fungal pathogen Fusarium oxysporum f. sp. radicis lycopersici. It appeared that strain WCS307 had acquired biocontrol abilities. This observation confirms the crucial role of colonization in biocontrol (12). We can conclude that improvement of colonization by genetic engineering is a realistic goal. Rhizosphere competence may have many analogies with competence of virulent bacteria to occupy niches. It is easy to predict a role for rhizosphere competence in causing plant root diseases. In addition, the pattern begins to emerge that several traits involved in rhizosphere competence may also be involved in the detrimental action of human and animal pathogens. (i) LPS is involved in the colonization of the mouse large intestine by Salmonella typhimurium (28). (ii) We reported that a Pseudomonas htrB mutant is impaired in root colonization, possibly through a slight growth defect (13). HtrB is a lauroyl transferase involved in lipid A biosynthesis (10). 22) reported that a Salmonella typhimurium htrB mutant is severely limited in its ability to colonize organs of the lymphatic system of mice. (iii) The human pathogen P. aeruginosa strain PAO1 was reported to be an excellent wheat rhizosphere colonizer that is even able to protect wheat and cucumber from the fungal pathogens Gaeumannomyces graminis and Pythium ultimum respectively (36). (iv) Among Vibrio cholerae mutants impaired in colonization of the mouse intestine, mutants were found in the biosynthesis of biotin and lipolysaccharide. In addition, seven auxotrophs were found to be attenuated in colonization (8).