Endophytic bacteria are ubiquitous in most plant species, residing latently or actively colonizing plant tissues locally as well as systemically. Several definitions have been proposed for endophytic bacteria; in this review endophytes will be defined as those bacteria that can be isolated from surface-disinfested plant tissue or extracted from within the plant, and that do not visibly harm the plant. While this definition does not include nonextractable endophytic bacteria, it is a practical definition based on experimental limitations and is inclusive of bacterial symbionts, as well as internal plant-colonizing nonpathogenic bacteria with no known beneficial or detrimental effects on colonized plants. Historically, endophytic bacteria have been thought to be weakly virulent plant pathogens but have recently been discovered to have several beneficial effects on host plants, such as plant growth promotion and increased resistance against plant pathogens and parasites. In general, endophytic bacteria originate from the epiphytic bacterial communities of the rhizosphere and phylloplane, as well as from endophyte-infested seeds or planting materials. Besides gaining entrance to plants through natural openings or wounds, endophytic bacteria appear to actively penetrate plant tissues using hydrolytic enzymes like cellulase and pectinase. Since these enzymes are also produced by pathogens, more knowledge on their regulation and expression is needed to distinguish endophytic bacteria from plant pathogens. In general, endophytic bacteria occur at lower population densities than pathogens, and at least some of them do not induce a hypersensitive response in the plant, indicating that they are not recognized by the plant as pathogens. Evolutionarily, endophytes appear to be intermediate between saprophytic bacteria and plant pathogens, but it can only be speculated as to whether they are saprophytes evolving toward pathogens, or are more highly evolved than plant pathogens and conserve protective shelter and nutrient supplies by not killing their host. Overall, the endophytic microfloral community is of dynamic structure and is influenced by biotic and abiotic factors, with the plant itself constituting one of the major influencing factors. Since endophytic bacteria rely on the nutritional supply offered by the plant, any parameter affecting the nutritional status of the plant could consequently affect the endophytic community. This review summarizes part of the work being done on endophytic bacteria, including their methodology, colonization, and establishment in the host plant, as well as their role in plant–microbe interactions. In addition, speculative conclusions are raised on some points to stimulate thought and research on endophytic bacteria.Key words: endophytic bacteria, methods, localization, diversity, biological control.

The external and internal colonization of potato and Arabidopsis roots by the biocontrol strain Rhizobium etli G12 containing a plasmidborne trp promoter green fluorescent protein transcriptional fusion, pGT-trp, was studied in the presence and absence of the root-knot nematode Meloidogyne incognita. Plant colonization behavior and biocontrol potential of the marked strain G12(pGT-trp) was not altered compared with the parental strain. Plasmid pGT-trp was stable for more than 80 generations without selection and conferred sufficient fluorescence to detect single bacterial cells in planta. Although bacteria were found over the entire rhizoplane, they preferentially colonized root tips, the emerging lateral roots, and galled tissue caused by Meloidogyne infestation. Internal colonization of potato roots was mainly observed in epidermal cells, especially root hairs. G12(pGT-trp) colonization was also observed in inner Arabidopsis root tissues in areas of vascularization. In the presence of M. incognita, G12(pGT-trp) colonized the interior of nematode galls in high numbers. In some cases, bacterial colonization even extended from the galled tissue into adjacent root tissue. The internally colonized sites in roots were often discontinuous. Fluorescence microscopy of gfp-tagged rhizobacteria was a sensitive and a rapid technique to study external and internal colonization of plant roots by bacteria interacting with nematodes.

Investigations were conducted to determine how a systemic plant-colonizing bacterium Enterobacter asburiae JM22 enters cotton plant tissues. Passive uptake was excluded for JM22 by experimentation with glutaraldehyde-fixed (killed) bacterial cells applied to seeds and leaves; no bacteria were found internally or externally on roots or leaves. In contrast, application of live JM22 cells led to colonization of external and internal root and leaf tissues. Active penetration of JM22 in the absence of external wounding was demonstrated for cotton seedlings germinated on water agar and inoculated with the bacterial suspension. The mean internal bacterial population density for seedlings was 3.8 × 103 CFU/g surface-disinfected radicle tissue. Studies of in planta enzymatic activity demonstrated hydrolysis of wall-bound cellulose in the vicinity of JM22 bacterial cells. The same phenomenon was observed for a cortical root colonizing bacterium, Pseudomonas fluorescens 89B-61, a plant growth-promoting strain with biocontrol potential against various pathogens.Key words: endophytic bacteria, cotton, cell wall hydrolysis.

Investigations were conducted to determine if biological control agent Pseudomonas fluorescens 89B-61 could colonize cotton tissues systemically and if internal colonization by a known endophytic bacterium, Enterobacter asburiae JM22, was influenced by the presence of other plant-associated bacteria. Following seed treatment, Pseudomonas fluorescens 89B-61 colonized cotton roots both externally and internally at mean population densities of 8.7 × 105 CFU/g and 1.1 × 103 CFU/g, respectively. However, bacteria were not detected in cotyledons, leaves, or stems. After inoculation onto leaves, Pseudomonas fluorescens 89B-61 established a mean internal population density of 1.6 × 104 CFU/g leaf tissue. Following stem injection, Pseudomonas fluorescens 89B-61 did not colonize roots or leaves. Pseudomonas fluorescens 89B-61 was localized on the root surface concentrated in grooves between epidermal cells, below collapsed epidermal cells, and in intercellular spaces close to the root epidermis, as identified by immunogold labeling of the bacterial membrane. Combined application of E. asburiae JM22 with another endophyte, Paenibacillus macerans Tri2-10, resulted in significantly lower internal populations of E. asburiae JM22 compared with treatment with E. asburiae JM22 alone. However, when coinoculated with a rhizosphere colonist, Micrococcus agilis strain 2RD-11, the colonization density of E. asburiae JM22 was not negatively affected. The results suggest that the internal colonization of cotton by bacteria with biological control activity may be an important aspect in their capacity to protect host plants against plant pathogens. The extent of internal colonization was shown to be influenced by other bacterial colonists.Key words: endophytic bacteria, location, interaction, cotton.

No abstract