Applications 2631 Spatial Distribution

The microscopic investigation of the localization of human or plant pathogenic bacteria on plant surfaces has provided new insights into their ecology in that environment. The observations by confocal microscopy that GFP-labeled Salmonella enterica formed microcolonies on the leaves of cilantro plants provided evidence that this enteric pathogen has the ability to colonize plants in a preharvest environment and may cause outbreaks due to contamination that occurred in the field [19]. Furthermore, S. enterica developed high-density populations and large heterogeneous aggregates in the vein area of the leaf, thus following spatial colonization patterns similar to those of the natural plant microflora on uninoculated plants [20]. In this same study, Brandl and Mandrell used confocal scans in the xz plane to obtain optical cross-sections of healthy and diseased cilantro leaves, and demonstrate that S. enterica cells had gained access to internal tissue while growing in the plant lesion whereas they remained on the cuticle layer while colonizing a healthy leaf [19] (Figure 26.2). Optical cross-sectioning by CLSM is an effective way to demonstrate unequivocally the internalization of bacteria in plant openings or tissue without the potential artifacts created by mechanical sectioning, which may contaminate internal tissue with external bacteria.

Confocal microscopy of GFP-labeled cells was used in several studies to demonstrate that enteric pathogens attach to and multiply at high densities in the damaged tissue of a variety of fruits and vegetables [6,19,21]. Burnett et al. combined CLSM and digital image analysis to count GFP-tagged E. coli

figure 26.2 (Color insert follows page 594) CLSM micrograph of GFP-labeled s. enterica serovar Thompson cells after their inoculation and incubation on the leaves of cilantro plants. S. Thompson cells were observed on healthy (A) and diseased (B) leaf tissue. The lower panels show cross-sections of the tissue in the top panel that were acquired by optical sectioning in the xz plane. They reveal that the bacterial cells are located on top of the cuticle on the healthy leaf (A), but have invaded the damaged tissue of the diseased leaf (B). Bars, 10 prn. (From Brandl, M.T. and Mandrell, R.E., Appl. Environ. Microbiol., 68, 3614, 2002.)

figure 26.2 (Color insert follows page 594) CLSM micrograph of GFP-labeled s. enterica serovar Thompson cells after their inoculation and incubation on the leaves of cilantro plants. S. Thompson cells were observed on healthy (A) and diseased (B) leaf tissue. The lower panels show cross-sections of the tissue in the top panel that were acquired by optical sectioning in the xz plane. They reveal that the bacterial cells are located on top of the cuticle on the healthy leaf (A), but have invaded the damaged tissue of the diseased leaf (B). Bars, 10 prn. (From Brandl, M.T. and Mandrell, R.E., Appl. Environ. Microbiol., 68, 3614, 2002.)

O157:H7 cells that attached at various depths in healthy or punctured apple tissue during inoculation [21]. From this approach, they concluded quantitatively that the pathogen infiltrated intact tissue and natural openings to a greater extent under negative than positive temperature differential.

The ability of plant pathogens and human enteric pathogens to become internalized in plant tissue, where they are shielded from the adverse effects of environmental conditions or chemical control agents, has been the focus of much interest in plant microbiology. Confocal microscopy has been a valuable means of probing the internal tissue of plants inoculated with GFP-tagged bacteria. Hallmann et al. were among the first to use the CLSM to observe the endophytic localization of a bacterial species, Rhizobium etli, within Arabidopsis thaliana roots [22]. Using a similar approach, endophytic colonization of A. thaliana and alfalfa roots by E. coli O157:H7 and S. enterica, respectively, has been demonstrated in plant systems in vitro [23,24]. A. thaliana and various other plant species have transparent roots, which lend themselves well to optical sectioning. In contrast, opaque plant tissues, e.g., potato tubers, make visualization of internalized GFP-labeled bacteria more challenging. Although opaque plant tissue can be cleared with various techniques, these generally inhibit or destroy GFP fluorescence [22]. In an experimental soil system that attempted to simulate field conditions, albeit with high inoculum levels, Solomon et al. found evidence for the transmission of GFP-tagged E.coli O157:H7 from contaminated manure to the internal tissue of lettuce leaves [25]. However, the low resolution of their published confocal micrograph reveals the difficulty of imaging fluorescent bacterial cells located under several layers of opaque leaf tissue. Also challenging is the localization of the internalized bacterial cells within the plant tissue in the epifluorescence mode during browsing before image acquisition in the confocal mode; browsing large numbers of fields of view or samples for the visualization of rare events is impractical by confocal microscopy. In our experience, this problem occurs even with brightly fluorescent bacterial cells.

Fluorescence microscopy has been instrumental to the discovery that bacteria can form biofilms on plants. Morris et al. used epifluorescence microscopy and AO staining of microbial cells to demonstrate the presence of natural biofilms on leaves of a variety of vegetables [26]. Since this first report, biofilm formation on plants has been shown in several studies using various microscopy techniques. Monier and Lindow developed a method to study the frequency, size, and localization of bacterial aggregates in situ on leaf surfaces [27]. Quantitative data were obtained by digital image analysis of epifluorescence micrographs of AO-stained bacteria present on leaf samples. The analysis was performed based on the outlining of the profiles of single bacteria or bacterial aggregates, by thresholding on their bright fluorescence intensity against the less fluorescent plant background. This method was applied to the quantification of the size of aggregates formed by S. enterica in the cilantro phyllosphere using GFP as an intrinsic fluorescent bacterial label instead of AO [28], and is illustrated in Figure 26.3.

26.3.2 Cell-Cell Interactions

The characterization of individual microbial aggregates isolated from plants revealed that they harbor a wide range of microorganisms including numerous species of Gram-negative and Gram-positive bacteria, as well as yeasts and filamentous fungi [26,29]. The heterogeneous composition of aggregates suggests that a complex pattern of microbial interactions is possible even at such small scales in the plant environment. While the spatial organization of epiphytic bacterial populations had remained obscure until recently, the use of marker genes conferring the production of fluorescent proteins combined with fluorescent stains has proven to be a valuable tool and has provided new insight into our understanding of bacterial interactions on plant surfaces.

Despite the heterogeneity of the plant surface habitat, which makes it a difficult task to study bacterial interactions in situ, a few studies attempting to decipher the factors shaping the structure of epiphytic communities have

figure 26.3 Schematic diagram illustrating the basis for digital analysis of fluorescence images. (A) Epifluorescence micrograph of cellular aggregates of GFP-labeled S. enterica on a cilantro leaf. Each GFP-labeled S. enterica single cell or aggregate in the image can be identified by thresholding on the bright pixels originating from the GFP fluorescence, which is of higher intensity than the background pixels originating from the leaf surface (inset). This thresholding yields objects (B) for which a variety of parameters, such as total number of pixels or mean pixel intensity, can be automatically measured with the image analysis software. Because of the highly heterogeneous spatial distribution of bacteria on plants, as well as variations between plants, this type of analysis requires the acquisition of a large number of images from random fields of view of multiple plant samples in order to yield unbiased data.

figure 26.3 Schematic diagram illustrating the basis for digital analysis of fluorescence images. (A) Epifluorescence micrograph of cellular aggregates of GFP-labeled S. enterica on a cilantro leaf. Each GFP-labeled S. enterica single cell or aggregate in the image can be identified by thresholding on the bright pixels originating from the GFP fluorescence, which is of higher intensity than the background pixels originating from the leaf surface (inset). This thresholding yields objects (B) for which a variety of parameters, such as total number of pixels or mean pixel intensity, can be automatically measured with the image analysis software. Because of the highly heterogeneous spatial distribution of bacteria on plants, as well as variations between plants, this type of analysis requires the acquisition of a large number of images from random fields of view of multiple plant samples in order to yield unbiased data.

been reported. Normander et al. have reported the significance of bacterial distribution on genetic exchange in the phyllosphere using GFP as an indicator of plasmid transfer [30]. Conjugation was observed under CLSM to occur primarily in the interstitial spaces of epidermal cells and vein cells, and in stomata; bacterial aggregation had a great stimulatory effect on plasmid transfer. Such data are pertinent to assessing the risk associated with the dissemination of antibiotic resistance genes among bacterial cells on plants.

Monier and Lindow tested the spatial partitioning of cells within aggregates on leaf surfaces by establishing different pairwise mixtures of three different epiphytic bacterial species that were tagged either with GFP or CFP [31]. The spatial structure of the resulting aggregates was studied in situ on leaves by epifluorescence microscopy. Digital image analysis was employed to quantify the degree of segregation of the GFP- and the CFP-marked strains and revealed that the fraction of cells in direct contact ranged from 0.2 to 8.0%. The highest segregation occurred between two bacterial species exhibiting negative interactions (Figure 26.4).

Fluorescence microscopy has proven useful also for the assessment of various bacterial genes hypothesized to have a role in cell-cell interactions on plants. For example, by comparing the behavior of GFP-labeled parental and mutant strains in situ on plants under the microscope, the function of the adhesin encoding hecA gene in the attachment and aggregation of Erwinia chrysanthemi on plants was confirmed [32]. In contrast, comparison of GFP-labeled S. enterica parental and LuxS mutant strains by digital image

figure 26.4 (Color insert follows page 594) Epifluorescence micrograph of P. agglomerans and P. fluoresceins cells labeled with CFP (blue arrow) and GFP (green arrow), respectively, and inoculated onto the leaves of bean plants incubated under humid conditions. The dual labeling with fluorescent proteins provided a good means to quantify the degree of spatial segregation between aggregates of these two bacterial species by image analysis. Bar, 20 p,m.

figure 26.4 (Color insert follows page 594) Epifluorescence micrograph of P. agglomerans and P. fluoresceins cells labeled with CFP (blue arrow) and GFP (green arrow), respectively, and inoculated onto the leaves of bean plants incubated under humid conditions. The dual labeling with fluorescent proteins provided a good means to quantify the degree of spatial segregation between aggregates of these two bacterial species by image analysis. Bar, 20 p,m.

analysis of their aggregate sizes on leaves revealed that production of the autoinducer-2 molecule for cell-cell signaling had no detectable role in aggregate formation by this human pathogen in the cilantro phyllosphere [28].

Besides the fluorescent proteins, or in combination with them, fluorescent dyes can provide a means to view the microbial composition and possible interactions on plant surfaces. The LIVE BacLight™ bacterial Gram stains (Molecular Probes) impart green and red fluorescence to Gram-negative and Gram-positive bacterial cells, respectively. The DNA dye SYTO® 9, which is included in the assay, also stains fungi, and therefore the assay allows for easy visualization of the fungal and bacterial composition of the plant microflora (Figure 26.5) [33]. In a study of the interaction of S. enterica with the common bacterial epiphyte P. agglomerans in the cilantro phyllosphere, the SYTO® 62 dye, which emits in the red region of the spectrum, enabled the observation under CLSM that these two species were part of larger bacterial aggregates (Chapter 2, Figure 2.3B), and that they attached to fungal hyphae (Figure 26.6).

26.3.3 Measurement of Biological Parameters

The assessment of bacterial cell viability, although complex in its interpretation, is central to our understanding of how bacteria survive in a given habitat. Also, there is an increasing need to understand the physiology of bacteria in the plant environment in order to design efficient strategies to control plant colonization by human or plant pathogens, or to sanitize fruits and vegetables. Fluorescent reporters may provide information about the

Intestine Xray With White Spots

figure 26.5 (Color insert follows page 594) CLSM micrograph of the microbial community on a leaf section of field-grown bean plants. The microbes were stained with LIVE BacLight Gram stain. Single cells and large mixed aggregates composed of fungi (bright green filaments, white arrow), and putative Gram-negative (green cells) and Gram-positive (red/orange cells) bacterial cells are present on the red autofluorescent leaf and in the vicinity of a glandular trichome (yellow arrow). Bar, 20 p.m.

figure 26.5 (Color insert follows page 594) CLSM micrograph of the microbial community on a leaf section of field-grown bean plants. The microbes were stained with LIVE BacLight Gram stain. Single cells and large mixed aggregates composed of fungi (bright green filaments, white arrow), and putative Gram-negative (green cells) and Gram-positive (red/orange cells) bacterial cells are present on the red autofluorescent leaf and in the vicinity of a glandular trichome (yellow arrow). Bar, 20 p.m.

physiological status of bacterial cells in complex ecosystems. They can be used to determine cell viability via the assessment of basic cell functions such as reproductive ability, membrane integrity, and respiration, or to measure cellular parameters, such as pH and levels of various ions.

26.3.3.1 Kogure Assay for Cell Viability

The ability of bacterial cells to grow and multiply has been the gold standard to demonstrate cell viability. In an approach based on the Kogure assay [34], Wilson and Lindow used a direct viable count method to examine the viability of epiphytic populations of Pseudomonas syringae on bean plants under desiccation stress [35]. The method consisted of incubating cells recovered from bean leaves in low-percentage yeast extract and in nalidixic acid to provide substrates for growth and to prevent cell division, respectively. The cells were then stained with DAPI, and cells that were fluorescent and elongated (growing cells in which division is inhibited by nalidixic acid) were counted as viable cells under the epifluorescence microscope. The increasing frequency of viable but nonculturable cells of the pathogen R. solanacearum during infection of tomato

figure 26.6 (Color insert follows page 594) CLSM micrograph of GFP-labeled S. enterica cells (yellow arrow) and DsRed-labeled P. agglomerans cells (white arrow) attached to a SYTO® 62-stained fungal hypha (blue arrow) in the phyllosphere of cilantro plants. The bright blue round objects are the chloroplasts of the leaf vein epidermal cells. The SYTO® 62 stain which emits at 680maxnm was assigned the pseudocolor blue. The image was acquired by excitation with argon, krypton, and He/Ne lasers (Leica Microsystems, Wetzlar, Germany). Bar, 20 p.m.

figure 26.6 (Color insert follows page 594) CLSM micrograph of GFP-labeled S. enterica cells (yellow arrow) and DsRed-labeled P. agglomerans cells (white arrow) attached to a SYTO® 62-stained fungal hypha (blue arrow) in the phyllosphere of cilantro plants. The bright blue round objects are the chloroplasts of the leaf vein epidermal cells. The SYTO® 62 stain which emits at 680maxnm was assigned the pseudocolor blue. The image was acquired by excitation with argon, krypton, and He/Ne lasers (Leica Microsystems, Wetzlar, Germany). Bar, 20 p.m.

plants was reported in a similar study [36]. This method has the advantage of examining viability directly via the ability of the cells to grow, but provides little information about the spatial distribution of the viable and nonviable cells at the microscale in situ on plants.

26.3.3.2 Indicators of Membrane Integrity

Biological stains that report on bacterial membrane activity can be useful to probe bacterial cell viability directly on plants. These stains penetrate only cells that have a compromised cytoplasmic membrane, and therefore are presumably nonviable. They include the DNA dyes ethidium bromide, TO-PROTM-3 and SYTOX® Green (Molecular Probes), and propidium iodide (PI), which is probably the most commonly used. The LIVE/DEAD BacLight bacterial viability assay (Molecular Probes) is a popular and simple method for determination of bacterial cell viability in which live cells fluoresce green due to staining with SYTO® 9, and dead cells fluoresce red due to staining with PI. The spectra of both stains are sufficiently close to detect both the green and red cells with a fluorescein filter. Using this assay, Warriner et al. reported the differential survival of E. coli and S. enterica cells inside and on the surface of bean sprouts after their treatment with sodium hypochlorite [37]. Also, PI staining was combined with immunostaining to detect specifically viable and nonviable cells of E. coli O157:H7 introduced onto cut lettuce and exposed to sodium hypochlorite [11].

Monier and Lindow developed a protocol based on PI staining, epifluo-rescence microscopy, and digital image analysis to determine the viability of individual bacterial cells directly on plants [38]. In their studies, PI was used to map the distribution of viable and nonviable GFP- or CFP-labeled cells on leaves (1) before and after they were exposed to desiccation stress, (2) in homogenous versus heterogeneous aggregates, and (3) after they landed in an aggregate of the same species versus in that of another species. These studies demonstrated quantitatively the importance of aggregation in the survival of epiphytic bacteria [38], the existence of antagonistic interactions at the bacterial scale within mixed aggregates [31], and the differential fate of immigrant bacteria to leaf surfaces depending on resident bacteria and on leaf anatomical features at the landing site [39].

Because the correlation between membrane integrity and physiological status of the cell is still controversial, cell viability data obtained with stains such as PI should be interpreted carefully. It is important to emphasize that the choice of a fluorescent viability probe is critical, and whether a certain fluorescent probe is suitable for viability assessment under the conditions tested has to be assessed.

26.3.3.3 GFP Fluorescence and Cell Viability

With the general excitement about GFP as an intrinsic label for bacteria, a thorny issue that has received little attention is whether GFP or its variants can serve as an indicator of cell viability. That is, can all fluorescent GFP-tagged cells observed under the microscope be considered as viable cells? Lowder et al. reported a strong correlation between Pseudomonas fluoresceins cell death and leakage of GFP from cells; most dead cells (as assessed by viability staining) were not GFP fluorescent, but a small percentage were dead and retained green fluorescence [40]. We have made similar observations with cultured GFP-labeled S. enterica cells that were exposed to the lethal stress of high temperature, desiccation, or calcium hypochlorite treatment [41]. Thus, it appears that GFP, which is considered as a stable fluorochrome in living cells, is lost rapidly upon cell death in bacteria. However, because of the small percentage of fluorescent GFP cells for which the cell status is unclear, it is preferable to confirm cell viability with an additional method when inferring specifically on the viability of GFP-tagged cells.

26.3.3.4 Other Fluorescent Indicators of Bacterial Physiology

Although many studies have investigated the detection of metabolically active bacteria in aquatic environments with fluorogenic substrates as indicators of bacterial respiratory or enzymatic activity [42,43], few of these dyes have been used so far to investigate bacterial activity on plants. Similarly, the assessment of intracellular pH has been performed at the single bacterial cell level by fluorescence ratio imaging with fluorescent pH indicators (e.g., 5- (and 6-) carboxyfluorescein, BCECF, SNAFL, SNARF) [42,43], and with GFP [44], but their utility for in situ probing of bacterial pH on plants has not been explored. With the increasing interest in bacterial survival to acid stress in fresh-cut fruits and vegetables, such an approach may prove to be almost essential. Additionally, the use of fluorescent probes in combination with flow cytometry in antimicrobial research has been widely reported. Fluorescent probe technology and microscopy may be applied successfully to quantify the effect of decontamination agents on human or plant pathogens on agricultural plants.

26.3.4 Bacterial Gene Expression In Situ ON PLANTS

26.3.4.1 GFP as a Reporter of Gene Expression

The combination of fluorescent markers with reporter gene technology has proven to be a powerful tool to study the behavior of bacteria on plants. The fusion of fluorescent reporter genes to bacterial genes of interest allows for the measurement of the transcriptional activity of that gene at the single bacterial cell level under the microscope, rather than at the population level. Thus, the distribution of transcriptional activity of a gene, and, potentially, the role of its phenotype, can be assessed in particular environments.

Unless the experiments are performed in a gnotobiotic system, an additional marker is required to distinguish the bacterial cells under study from those belonging to the indigenous microflora. This ensures that the bacterial cells in which transcriptional activity of the gene of interest is low or off, and therefore in which the reporter signal is low, will be detected. In the first study of this type on plants, Brandl et al. used a transcriptional gfp fusion to an auxin (IAA) biosynthetic gene of E. herbicola in combination with FISH to assess the distribution of IAA synthesis in this bacterial species on bean leaves [12] (Figure 26.7). Plants were inoculated with a strain of E. herbicola transformed with the gfp fusion and incubated to allow for colonization to occur. Then, the bacterial cells were washed off the leaves and subjected to FISH on microscope slides with tetramethylrhodamine-labeled 16S rRNA probe specific to E. herbicola. The green fluorescence intensity of bacterial cells that were stained red by FISH was measured by analysis of digital images acquired under the epifluorescence microscope. The frequency distribution of GFP fluorescence intensity per cell revealed that a small proportion of the E. herbicola cells on the leaves expressed the IAA gene at very high levels, suggesting that there were microsites on the leaf that were conducive to high production of IAA. With the same approach, subsequent studies demonstrated the heterogeneous distribution of the availability of sucrose [15] and fructose [14] to E. herbicola, and of iron to Pseudomonas syringae [13], on plant surfaces. In all of the above studies, FISH enabled the specific labeling of bacterial cells at the strain level

Spatial distribution of ipdC-gfp expression in situ on the leaf

Distribution frequency of ipdC-gfp expression figure 26.7 Schematic diagram of fluorescence microscopy strategies to investigate the distribution of gene expression at the bacterial cell level on plants. The protocols make use of dual labeling with GFP as a reporter of transcriptional activity, and with rhodamine as a marker for FISH to identify specific bacterial cells among the natural plant microflora. Spatial distribution of ipdC-gfp expression is assessed by visualization under CLSM of the GFP fluorescence in bacterial cells that were identified by FISH, performed directly on plant samples (A). Frequency distribution data of the activity of transcriptional fusions to gfp are acquired by measuring the fluorescence of individual bacterial cells that were washed off the plant surface and identified by FISH (B). (From Brandl, M.T., Quinones, B., and Lindow, S.E., Proc. Natl. Acad. Sci. USA, 98, 3454, 2001.)

Spatial distribution of ipdC-gfp expression in situ on the leaf

Distribution frequency of ipdC-gfp expression figure 26.7 Schematic diagram of fluorescence microscopy strategies to investigate the distribution of gene expression at the bacterial cell level on plants. The protocols make use of dual labeling with GFP as a reporter of transcriptional activity, and with rhodamine as a marker for FISH to identify specific bacterial cells among the natural plant microflora. Spatial distribution of ipdC-gfp expression is assessed by visualization under CLSM of the GFP fluorescence in bacterial cells that were identified by FISH, performed directly on plant samples (A). Frequency distribution data of the activity of transcriptional fusions to gfp are acquired by measuring the fluorescence of individual bacterial cells that were washed off the plant surface and identified by FISH (B). (From Brandl, M.T., Quinones, B., and Lindow, S.E., Proc. Natl. Acad. Sci. USA, 98, 3454, 2001.)

within a given species. Additionally, the rhodamine label of the 16S rRNA probe provided a fluorescent signal with an emission spectrum sufficiently distinct from that of GFP to prevent misinterpretation of the fluorescent signals in each microscope filter set. Optical crosstalk is a major issue in multilabeling experiments.

Frequency distribution analysis of bacterial gene expression has been performed mostly by epifluorescence microscopy to allow for fluorescence measurements of a large number of bacterial cells that were recovered from plant tissue. Brandl et al. developed a method for the assessment of bacterial GFP fluorescence in situ on leaves under the CLSM (Figure 26.7) [12]. The method consisted in performing FISH on leaf disks that were fixed in paraformalde-hyde and then covered with a thin film of low-percentage agar to prevent disruption of the spatial distribution of the bacterial cells during hybridization procedures. The transcriptional activity of GFP reporter fusions was assessed subsequently in 16S rRNA-labeled E. herbicola cells, through the agar, by confocal microscopy (Figure 26.8). In this manner, spatial patterns of gene expression in a specific bacterial population could be established.

figure 26.8 (Color insert follows page 594) CLSM micrograph of the spatial distribution of E. herbicola cells harboring an ipdC-gfp fusion in the vicinity of a glandular trichome on the surface of bean leaves. The same field of view was imaged sequentially with a rhodamine (A) and a GFP (B) emission filter. E. herbicola cells were detected by FISH with a rhodamine-labeled 16S rRNA probe and performed on leaf disks mounted in agar (A). Large variations in the expression of ipdC-gfp were detected among the population of rhodamine-labeled cells (B). White and yellow arrows show E. herbicola cells with high and low levels of ipdC-gfp expression, respectively. Width of white square, 5 p,m. (From Brandl, M.T., Quinones, B., and Lindow, S.E., Proc. Natl. Acad. Sci. USA, 98, 3454, 2001. Copyright 2001 National Academy of Sciences, USA.)

figure 26.8 (Color insert follows page 594) CLSM micrograph of the spatial distribution of E. herbicola cells harboring an ipdC-gfp fusion in the vicinity of a glandular trichome on the surface of bean leaves. The same field of view was imaged sequentially with a rhodamine (A) and a GFP (B) emission filter. E. herbicola cells were detected by FISH with a rhodamine-labeled 16S rRNA probe and performed on leaf disks mounted in agar (A). Large variations in the expression of ipdC-gfp were detected among the population of rhodamine-labeled cells (B). White and yellow arrows show E. herbicola cells with high and low levels of ipdC-gfp expression, respectively. Width of white square, 5 p,m. (From Brandl, M.T., Quinones, B., and Lindow, S.E., Proc. Natl. Acad. Sci. USA, 98, 3454, 2001. Copyright 2001 National Academy of Sciences, USA.)

The discovery of the fluorescent protein DsRed caused much excitement because of its potential to be used as an intrinsic bacterial label along with a transcriptional fusion to GFP in gene expression studies. Despite reports that GFP and DsRed can be detected in a single cell upon simultaneous excitation [45], it appears that DsRed is also a good acceptor molecule, with GFP or CFP as a donor, in fluorescence resonance energy transfer (FRET) imaging [46]. The interaction between these fluorescent proteins may confound the quantitative interpretation of the fluorescent signals, and therefore the usefulness of DsRed for dual-labeling with GFP or its variants still needs to be ascertained.

26.3.4.2 Practical Note on the Use of GFP for Gene Expression Measurements

Because of the great stability of GFP, unstable GFP variants with a short half-life have been constructed to measure transient bacterial gene expression in time-course experiments on plants [47-49]. These destabilized variants prevent the accumulation of GFP under basal or noninduced gene expression conditions and are more accurate reporters of transcriptional activity at a given time point. Also, GFP fluoresces poorly under low oxygen conditions, and some variants like EGFP have decreased fluorescence at a pH between 7.0 and

4.5, with only 50% of its fluorescence at pH 6.0; other factors affecting GFP chromophore formation include temperature, chemical denaturants, and certain solvents [50]. In addition, an effect of bacterial growth rate on GFP fluorescence intensity of individual cells has been reported [49]. Therefore, the effect of experimental conditions on the accumulation and the function of GFP as a chromophore per se should be tested with a constitutively expressed gfp to avoid misinterpretation of GFP fluorescence data.

26.3.4.3 FISH for the Detection of Bacterial mRNA

Although not reported in plant studies at the present, bacterial mRNA detection and quantification by FISH has been successfully performed in single bacterial cells in environmental samples. With as many as five fluorescently labeled oligodeoxynucleotide probes (depending on abundance of the transcript), the mRNA of two enterobacterial genes that are induced at different stages of growth was targeted to profile the physiological activity of Enterobacteriaceae in a waste water microbial community [51]. The mRNA profile was obtained in conjunction with rRNA FISH for the identification of the bacterial cells at the taxon level within the community. In other cases where abundance of target mRNA is low, signal amplification is necessary to detect fluorescence in single cells. Pernthaler and Amann have developed a FISH protocol for the sensitive detection of low-abundance mRNAs at the single bacterial cell level by enzymatic amplification of the fluorescence signal emitted from long oligonucleotide probes that were labeled at high density [52]. This enabled the detection of the expression of a single gene in methanotrophic bacteria present in sediment samples. The application of such powerful reporter systems to the detection of mRNA in single bacterial cells may offer a means of probing specific bacterial functions in natural microbial consortia in the plant environment. It remains to be determined, however, whether detection of the hybridization signal can be achieved against the often autofluorescent background on plants.

26.3.4.4 Immunolabeling of Gene Products

Immunofluorescence labeling represents an alternative method to FISH to investigate the localization of specific bacteria on plants, but it is suitable also for quantitative analysis of proteins or their enzymatic products. For example, patterns of regulation of a Ralstonia solanacearum virulence gene (eps) were determined by quantifying the amount of p-galactosidase protein present in single cells of a transformant of this plant pathogen that carried an eps-lacZ reporter fusion [53]. Quantitative measurements were performed by digital analysis of the fluorescence of single R. solanacearum cells recovered during infection of tomato plants, and then immunolabeled against p-galactosidase. Immunofluorescence microscopy with antibodies against the R. solanacearum exopolysaccharide EPS I54 and a specific Xanthomonas axonopodis lipopoly-saccharide [55] proved useful also to determine the spatiotemporal production of these virulence factors during progression of disease in their respective plant host. Such an approach may have great potential in the investigation of the surface components of plant-associated or human bacterial cells while they grow or survive in the plant environment. The increasing availability of very bright fluorescent antibody conjugates that allow for the detection of even small amounts of molecules at the single cell level makes this method worthy of consideration.

Recent breakthroughs in the development of fluorescent bioconjugates that are bright and span the visible range have contributed to the arsenal of strategies that microbiologists can employ to investigate the molecular biology of bacteria on plants. However, both FISH and immunofluorescence labeling in situ on plant surfaces require great attention to avoid the perturbation of the bacterial cells during the many washes involved in these procedures.

Organic Gardeners Composting

Organic Gardeners Composting

Have you always wanted to grow your own vegetables but didn't know what to do? Here are the best tips on how to become a true and envied organic gardner.

Get My Free Ebook


Post a comment