![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Applied and Environmental Microbiology, August 2007, p. 4905-4914, Vol. 73, No. 15
0099-2240/07/$08.00+0 doi:10.1128/AEM.02522-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Wageningen University and Research Centre, Plant Research International BV, Droevendaalsesteeg 1, 6709 PB Wageningen, The Netherlands,1 Wageningen University and Research Centre, Biological Farming Systems, Marijkeweg 22, 6709 PG Wageningen, The Netherlands2
Received 30 October 2006/ Accepted 12 May 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
However, only very few studies have investigated the physiological effect or molecular interaction between human bacterial pathogens and a plant host, e.g., the model plants Medicago and Arabidopsis. For Arabidopsis thaliana, it was shown that the opportunistic human pathogen Pseudomonas aeruginosa PA14 attached to the leaf surface, congregated at the stomata or wounds, and then invaded the leaves and colonized the intercellular spaces (35). The bacterium was also able to make circular perforations in mesophyllic cell walls to allow penetration. From this study, it was concluded that Pseudomonas aeruginosa PA14 is a facultative pathogen of A. thaliana that can cause local and systemic infection, eventually leading to plant death. Also, mutants of human-pathogenic Staphylococcus aureus (36) that contained disrupted genes involved in animal pathogenesis were attenuated in the ability to infect A. thaliana. This suggested that the same regulators that mediate synthesis of virulence factors essential for animal pathogenesis are also required for plant pathogenesis (36). Resistance of A. thaliana to S. aureus was mediated by a direct effect of salicylic acid (SA) on the pathogen affecting attachment on the root surface and reducing pathogen virulence.
Different Salmonella spp. were able to endophytically and epiphytically colonize Medicago sativa (11). A recent study revealed that colonization of Medicago truncatula by S. enterica serovar Typhimurium resulted in the induction of SA-dependent and -independent plant defenses (23). From this observation, the induction of both plant defense pathways was correlated with the bacterial gene expression of type III secretion system-Salmonella pathogenicity island effector proteins, whereas the presence of flagella only induced the SA-dependent plant defense induced by expression of the PR1 gene.
Although these studies give direction to a specific host-pathogen interaction, until now no research has been described to study the gene expression of plants during colonization by human-pathogenic bacteria, such as Salmonella spp.
The objectives of this study were to investigate the physiological and molecular responses of Lactuca sativa to S. enterica serovar Dublin during plant colonization. Colonization of lettuce plants by S. enterica serovar Dublin was studied by comparing the prevalence and degree of colonization on surface-disinfected and untreated plants grown in nutrient water-agar and in manure-amended soil. The epiphytic and endophytic presence of S. enterica serovar Dublin was investigated to provide insight into the capability of S. enterica serovar Dublin to invade plant tissue and to proliferate in or on the plant. To reveal generally and differentially expressed genes upon colonization of lettuce with S. enterica serovar Dublin over time, cDNA amplified fragment length polymorphism (cDNA-AFLP) gene expression profiling was performed. Transcript-derived fragments were subjected to sequence analysis and grouped by gene function. Subsequent gene expression profiling of selected genes was performed using reverse transcriptase PCR (RT-PCR).
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A liquid culture of Salmonella enterica serovar Dublin, grown overnight at 30°C in tryptic soy broth, was kindly provided by H. Aarts (RIKILT, The Netherlands). The culture was maintained by both plating on selective Hektoen enteric agar (Biotec Laboratories Ltd., United Kingdom) and overnight incubation in buffered peptone water (BPW) at 37°C. An Escherichia coli JM109 culture (obtained from the collection of Plant Research International BV) was maintained on Luria broth (LB) plates and in liquid LB medium by overnight incubation at 37°C.
Surface disinfection of lettuce plants colonized with Salmonella serovar Dublin.
To determine the efficiency of surface disinfection, 35 6-week-old lettuce plants (grown in soil) were inoculated with 20 µl of 2 x 108 CFU/ml of S. enterica serovar Dublin. In total, 10 µl of the inoculum was spread across the surface of one leaf and 10 µl was spread across the bottom of another leaf. After 5 min of incubation at room temperature, the plants were cut at the transition point, and from 25 plants, the leafy parts were disinfected by rinsing for 10 s in 70% ethanol and twice in water. Subsequently, each plant (leafy part) was ground in 1 ml of BPW, and a dilution series (diluted 100x and 1,000x) was prepared from the suspension. Each dilution was plated (40 µl) on Hektoen enteric agar, in duplicate, and incubated overnight at 37°C prior to colony counting. The means and standard errors of the numbers of Salmonella CFU recovered were calculated, and the surface disinfection efficiency was determined by the ratio between the mean number of CFU recovered from surface-disinfected plants and the mean number of CFU recovered from nondisinfected plants.
Association of Salmonella serovar Dublin with lettuce grown in manure-amended soil.
Fresh manure was collected from a Dutch organic dairy farm. Soil was collected from a field (60 kg of top layer of 20 cm) from the organic experimental farm the Droevendaal (Wageningen, The Netherlands). The soil consisted of 89% sand, 8% silt, 3% clay, total nitrate (N) and carbon (C) loads of 2,135 mg/kg and 22,400 mg/kg, respectively, and 11% moisture and had a pH of 7.14. The manure contained 28.7% acid detergent fiber, 40.3% neutral detergent fiber, total dissolved organic N and C loads of 740 mg/kg and 8,167 mg/kg, respectively, 220 mg/kg ammonium, and 8.14 mg/kg nitrate and had a pH of 6.8. Both substrates tested negative for the presence of Salmonella spp., which was determined by plating cells directly on selective Hektoen enteric agar and by testing the total DNA extracts from 10-ml BPW enrichments of three random samples of 1 g of each substrate by real-time PCR analysis (26).
Manure was inoculated with 108 CFU of S. enterica serovar Dublin/g wet weight and mixed thoroughly before addition to soil at a weight ratio of 1:10. The final number of S. enterica serovar Dublin CFU was 107/g fresh mixture. In total, 74 pots of 50 ml with 50 g of S. enterica serovar Dublin-contaminated soil-manure mixture were prepared. The negative control pots (74 in total) contained non-S. enterica serovar Dublin-inoculated manure-soil mixture. One lettuce seed was added to each pot (148 in total) and allowed to germinate in a greenhouse at 18°C and 80% humidity. After 6 weeks, each plant was harvested by cutting the plant at the stem just above the soil. The plants were each weighed and thoroughly washed in 30 ml of sterile water prior to analysis. Next, for both treatments, all plants were randomly divided into two sets of 38 plants. Each plant of the first set was ground in 1 ml of BPW. From the second set, each plant was surface disinfected as previously described, followed by grinding in 1 ml of BPW. Each suspension of ground plant material was plated (40 µl) on Hektoen enteric agar, in duplicate. In addition, the wash fraction was centrifuged and the pellet resuspended in 100 µl of BPW prior to being plated in duplicate on Hektoen enteric agar (40 µl/plate). After overnight incubation at 37°C, the total number of Salmonella CFU was counted for each plate. To determine a significant difference in plant weight between both treatments, a paired t test was performed for all tested plants per treatment.
Association of S. enterica serovar Dublin with lettuce grown on Hoagland's agar.
Sterilized lettuce seeds (120) were allowed to germinate for 3 weeks on 0.5% Hoagland's agar (pH 6.8) in closable growing units (10 x 15 x 8 cm) placed in a growth chamber at 20°C with 12-h light-dark intervals. To assess the colonization of lettuce by S. enterica serovar Dublin over time, the 120 3-week-old lettuce plants were inoculated at the root site with 10 µl of 107 CFU/ml of S. enterica serovar Dublin without wounding the roots. Every 2 days for a period of 20 days, the shoots of 12 lettuce plants were cut off just above the agar and weighed. To determine the prevalence, degree of colonization, and localization of S. enterica serovar Dublin (endophytic or epiphytic) associated with lettuce plants, the harvested leafy parts of six plants were not surface disinfected, whereas the leafy parts of the other six plants were surface disinfected as described earlier. The leafy parts were ground in 0.5 ml of BPW. A dilution series (nondiluted and diluted 10x and 100x) was prepared from each leaf suspension, and 40 µl of each dilution was plated onto Hektoen enteric agar, in duplicate. For both treatments (with and without surface disinfection), the prevalence and degree of colonization were determined by calculating the mean S. enterica serovar Dublin CFU subtracted by the error of surface disinfection efficiency.
Lettuce response to bacterial colonization.
The response of lettuce plants to colonization by S. enterica serovar Dublin was compared to the response to colonization by E. coli JM109. Water-inoculated plants were used as controls. Thirty seeds were sprouted on sterile Hoagland's agar in separate tubes in a closable growing unit (50 x 30 x 25 cm) for 3 weeks in a growth chamber at 20°C and 80% humidity. Next, every 10 sprouts were carefully inoculated at the roots with 10 µl of 107 CFU/ml of S. enterica serovar Dublin, 10 µl of 107 CFU/ml of E. coli JM109, or water. After 5 weeks, each surviving plant per treatment was cut at the transition point between the stem and roots and then weighed. To determine significant differences between the treatments with respect to plant death, nonparametric analysis (Kruskal-Wallis test with asymptotic significance) was performed based on the number of surviving plants and the weights of these plants.
Preparation of plant tissue cross sections.
From a different set of lettuce plants that were inoculated in a similar manner to that described above, 3 weeks after inoculation with S. enterica serovar Dublin, the transition region was cut for microscopic analysis before being cross-sectioned (in total, 1 cm of transition region was obtained for each plant). First, the cut transition regions were incubated overnight in fixative (96% ethanol-acetic acid [3:1 {vol/vol}]). After fixation, the tissues were transferred to a graded series of sucrose solutions with increasing concentrations, of 5, 10, 20, 30, 40, and 50% (wt/vol), in phosphate-buffered saline (PBS; pH 7.4). Samples were kept at each concentration for 30 min. Next, the samples were embedded in Tissue-Tek OCT compound (Miles, Elkhart, IN) while ensuring a vertical position of the samples. Cross sections of 20 µm and 200 µm in thickness were cut from each sample by using a cryostat (Microm HM 500 O; Microm Laborgeräte GmbH, Waldorf, Germany) at –30°C. The tissue sections were transferred to poly-L-lysine (p-1524; Sigma Chemical Co., St. Louis, MO) (0.1% in milliQ water [wt/vol])-coated slides, dried at 60°C for 15 min, and stored at –20°C until further use. To label the potentially present S. enterica serovar Dublin in each cross section, the slides were first incubated at 70°C for 15 min, washed once with milliQ water (without shaking) at room temperature, dried at 70°C, washed twice with milliQ water at room temperature, dried at 70°C, washed with PBST (PBS including 0.1% Tween) for 2 min at room temperature, washed with PBS with 2% bovine serum albumin for 10 min, and washed twice with PBST at room temperature. Next, the cross sections were incubated in the dark for 60 min at room temperature with 200 µl of labeling mix, which consisted of PBS, pH 7.4, 10 µg of fluorescein isothiocyanate-labeled polyclonal antibody to Salmonella common structural antigens (KPL Europe, Guildford, United Kingdom), and 10 µl of FA rhodamine counterstain (Difco Laboratories, Detroit, MI). After being stained, the cross sections were washed three times with PBS, pH 7.4, before 100 µl of mounting solution (Vectashield mounting solution for fluorescence; Vector Laboratories, Inc., Burlingame, CA) was added and slides were covered and sealed. Each slide with a cross section was analyzed using a fluorescence microscope including a charge-coupled device camera.
mRNA and DNA preparation for gene expression analysis.
For gene expression analysis, 60 seeds were germinated in Hoagland's agar, and after 3 weeks of growth, 30 plants were inoculated close to the roots with 10 µl of water and 30 plants were inoculated with 10 µl of 107 CFU/ml S. enterica serovar Dublin. Every 2 or 3 days for a 3-week period (10 time points, including time zero), three water-inoculated plants and three S. enterica serovar Dublin-inoculated plants were harvested by cutting the plants just above the agar surface. At each time point, the leafy parts of each treatment plant were pooled prior to being weighed and subsequently ground in liquid nitrogen and stored at –80°C. Total RNA was extracted from the ground samples by using a QIAGEN Plant RNeasy kit (Westburg, Germany) according to the supplier's protocol (including DNase treatment). The total RNA eluates were aliquoted in several portions before being stored at –80°C. Plant mRNAs were purified from 45 µl of each total RNA sample by using an Oligotex mRNA purification kit (Invitrogen). For DNA extraction from ground plant material, a QIAGEN Plant DNeasy kit was applied as described by the supplier's protocol. The purified DNA was dehydrated/dried using a speed vacuum concentrator, resuspended in 50 µl of milliQ water, and stored at –20°C until further use.
cDNA-AFLP differential gene expression analysis of S. enterica serovar Dublin-colonized lettuce.
Basic principles of cDNA-AFLP analysis were followed as described by Bachem et al. (3, 4). Each first-strand cDNA synthesis reaction mix was prepared by incubating 10 µl of purified mRNA with 2 µl of poly(dT) primer (5 µM) for 10 min at 70°C, followed by incubation on ice. Next, 4 µl of first-strand buffer (Invitrogen), 2 µl 0.1 M dithiothreitol, and 1 µl of a 10 mM concentration of each deoxynucleoside triphosphate (dNTP) was added to each RNA sample and shortly incubated at 37°C. First-strand synthesis was started by adding 1 µl of Superscript II RT RNase H– to each reaction mix. The sample was incubated for 1 h at 37°C before ending the synthesis by incubation on ice.
Second-strand synthesis was performed for each first-strand reaction in a total volume of 150 µl containing 500 nmol of each dNTP, 10 units of E. coli DNA ligase, 40 units of E. coli DNA polymerase I, and 2 units of E. coli RNase H (Invitrogen), with incubation for 2 h at 16°C. Two microliters of T4 DNA polymerase was then added, followed by a short incubation of 5 min at 16°C. The final double-stranded cDNA was purified by phenol-chloroform-isoalyl alcohol (1:1:24) and sodium acetate precipitation. The precipitated DNA was resuspended in 25 µl of water.
To perform restriction digestion of the prepared cDNA, 10 µl of cDNA sample was added to 40 µl of restriction mix containing a final concentration of 1x RL buffer with 10 units of MseI and 50 units of EcoRI. The mix was incubated overnight at 4°C.
Subsequently, 20 µl of cut cDNA sample was added to 30 µl of ligation mix, with final concentrations of 5 pmol of EcoRI adapter, 50 pmol of MseI adapter, 1x T4 DNA ligase buffer (Invitrogen), and 5 units of T4 DNA ligase. Ligation was performed by incubating the ligation sample for 90 min at 37°C. The ligated cDNA samples were stored at –20°C until further use.
Preamplification was performed using primers directed to the adapters that were ligated to the cDNA (zero reaction). Each PCR mix of 25 µl consisted of PCR buffer (10 mM Tris-HCl, 1.5 mM MgCl2, 50 mM KCl, pH 8.3), 150 nmol of each dNTP, 63 pmol each of primer EcoRI00 (5' GACTGCGTACCAATTC 3') and primer Mse00 (5' GATGAGTCCTGAGTAA 3'), 1 unit of Taq polymerase (Gibco BRL), and 5 µl of ligated cDNA sample. The reaction mix was incubated for 2 min at 94°C, followed by 35 cycles of 30 s at 94°C, 30 s at 56°C, and 90 s at 72°C.
Selective PCR was performed as described above, using the primers EcoRI19 (33P labeled; 3' +2 overhang [GA]) and Mse11 to Mse26 (2-nucleotide 3' overhang, with each possible combination) with 5 µl of a 50x dilution of preamplification product in a total volume of 20 µl. Primer EcoRI19 was labeled with [
-33P]ATP by using polynucleotide kinase prior to PCR. The PCR profiles were as follows: 1 cycle of 94°C for 30 s, 65°C for 30 s, and 72°C for 1 min, followed by 12 cycles of 94°C for 30 s, 65°C (–0.7°C/cycle) for 30 s, and 72°C for 1 min and 23 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 1 min. PCR products were size fractionated in a 5% polyacrylamide gel run for 1.5 h at 80 W.
Isolation of differentially expressed gene fragments and confirmation by RT-PCR.
AFLP gels were vacuum blotted and dried on Whatman 3MM paper for 1 h and subsequently exposed to X-ray films for 2 weeks. After film development, bands of interest were selected and cut from the blotted gel on Whatman paper.
The small paper cuttings were stored in a microtiter plate with 100 µl of RNase-free water and heated for 5 min at 95°C to elute the DNA from the paper. Next, 5 µl of eluted sample was amplified again by PCR (according to the zero-reaction protocol). Samples were analyzed in a gel and sequenced using primer E00. Each transcript-derived fragment (TDF) sequence was compared against all sequences in the nonredundant database, using the tBlastX program with the EMBL database and the TIGR expressed sequence tag (EST) libraries for L. sativa and A. thaliana.
To allow for gene confirmation and determination of expression profiles by RT-PCR, primer sets were designed based on alignments of TDFs with the most probable sequence hits from the EMBL database and the TIGR EST database. Each primer set was designed such that one primer was located inside both sequences of the alignment and one primer was located outside the TDF but inside the sequence obtained from the TIGR EST database. In addition, other primer sets were designed based on specific lettuce genes related to plant-pathogen interaction, namely, pathogenicity-related gene 1 (PR1), gene 4 (PR4), and gene 5 (PR5) and defender against apoptotic death (DAD-1). All primer sequences and corresponding genes are displayed in Table 1.
|
All amplification products were analyzed by electrophoresis using a 1% Pronarose gel containing 0.5 µg/ml of ethidium bromide. The intensities of the resulting bands were normalized for each gene-specific primer set, using the most intensive band as the 100% expression level, to allow proper comparisons of the time series for both noninoculated and inoculated samples.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Colonization of lettuce grown in manure-amended soil by Salmonella serovar Dublin.
Since salmonellae are frequently isolated from bovine manure, it was hypothesized that lettuce plants grown in manure-amended soil can be colonized by these salmonellae. To investigate the prevalence and degree of infection of lettuce plants with S. enterica serovar Dublin, lettuce seeds were applied to soil that was amended with noninoculated manure or S. enterica serovar Dublin-inoculated manure. From the seeds applied to soil amended with noninoculated manure, 61 of 74 seeds germinated. In soil amended with S. enterica serovar Dublin inoculum, 56 of 74 lettuce seeds germinated. This difference was not significant (chi-square value = 1.02). The mean weight of the leafy parts of the plants grown for 6 weeks on S. enterica serovar Dublin-inoculated manure-soil mixture was 0.52 g ± 0.17 g, and the mean weight of the leafy parts of the plants grown for 6 weeks on noninoculated manure-soil mixture was 0.57 g ± 0.15 g. Using these data, no significant difference was observed between both treatments, using analysis of variance (P = 0.153).
The prevalence of S. enterica serovar Dublin found in association with the leafy parts of lettuce plants was 27% (15 of 56 plants). The wash fraction of 15 sampled plants also contained S. enterica serovar Dublin, indicating that 27% of the plants in each case were colonized above the soil with loosely attached S. enterica serovar Dublin. Moreover, three surface-disinfected plants were positive for S. enterica serovar Dublin (5%), suggesting the presence of S. enterica serovar Dublin inside the plant tissue. Among these internally colonized plants, in two cases the wash fraction was also positive for S. enterica serovar Dublin, which indicated the presence of S. enterica serovar Dublin on the plant surface. For this set, the number of S. enterica serovar Dublin CFU recovered from the surface-disinfected plants ranged from 75 CFU to 1,275 CFU per plant. In addition, one nondisinfected plant was also positive for S. enterica serovar Dublin, indicating the presence of internal and/or external colonization of the plant by S. enterica serovar Dublin. These results suggest that lettuce plants can be colonized by salmonellae, in our case S. enterica serovar Dublin, if they are grown in soil amended with contaminated manure.
Colonization of lettuce grown in Hoagland's agar by Salmonella serovar Dublin.
To study the colonization of lettuce by S. enterica serovar Dublin over time, lettuce seeds were germinated on Hoagland's agar, and after 3 weeks, the plants were carefully inoculated at the roots with S. enterica serovar Dublin. The number of plant-associated S. enterica serovar Dublin CFU varied greatly over time among colonized plants, which led to a high standard error. Therefore, no correlation between the number of S. enterica serovar Dublin CFU and time postinoculation was obtained (Table 2). Nevertheless, a large difference in total number of S. enterica serovar Dublin CFU was found between the disinfected (mean of 3,808 ± 1,643 CFU per plant) and nondisinfected (mean of 49,582 ± 30,012 CFU per plant) plants (Table 2). Taking into account the surface disinfection efficiency (99.81%), for this experiment a maximum of 94 CFU (0.19% of 49,582 CFU) was considered a false-positive result for the disinfected plants. This is approximately 40-fold lower than the average number of CFU found inside the disinfected plants.
|
Localization of S. enterica serovar Dublin in association with lettuce plants.
To evaluate whether S. enterica serovar Dublin was able to colonize the plants up to the leaves, different plant parts were tested by grinding the tissues and subsequently plating them on selective Hektoen agar. From this experiment, S. enterica serovar Dublin was found to be associated with lettuce plants, mainly from the root-stem transition point up to the leaves, but not in the leaves (data not shown).
Next, to determine the possible point of entrance of S. enterica serovar Dublin into lettuce plants, cross sections of colonized plants were analyzed by fluorescence microscopy (Fig. 1). Analyses of the cross sections revealed strong growth of S. enterica serovar Dublin on the root surface (Fig. 1A and B) and near emerging lateral roots (Fig. 1C). Moreover, internalization was observed via the intercellular spaces between epidermal cells (Fig. 1D). S. enterica serovar Dublin bacteria were found at the cortex within the parenchyma tissue (Fig. 1E), either still attached to epidermal cells or spreading further through the parenchyma tissue. In a few cases, S. enterica serovar Dublin was also found attached to endodermal cells, inside the pericycle (Fig. 1E), or even inside the vascular system (Fig. 1F), which suggests that the bacterium might be able to pass the endodermis and potentially spread upwards via the vascular system.
|
|
Nonparametric analysis (Kruskall-Wallis test) of the total number of remaining healthy lettuce plants revealed a significant difference between all three treatments (P = 0.008). Comparing two treatments for the number of healthy plants, a significant difference (P = 0.002) was observed between water-inoculated plants and S. enterica serovar Dublin-inoculated plants. Water-inoculated plants compared with E. coli JM109-inoculated plants showed no significant difference (P = 0.057). This indicated that the observed morphological changes of the lettuce plants were induced by the presence of S. enterica serovar Dublin and were not due to the presence of a bacterium in general (in our case, E. coli) or to the depletion of nutrients (healthy controls).
As an additional control, the different treatments were also compared based on the biomass of the healthy, i.e., surviving, plants, which indicated no significant difference (P = 0.141) in the three treatments between all plants. This suggested that the remaining healthy plants receiving the three treatments (2 S. enterica serovar Dublin-treated plants, 6 E. coli-treated plants, 10 water-treated plants) were not colonized or influenced by the bacteria added close to the roots.
Identification of lettuce genes differentially expressed due to Salmonella colonization.
From the physiological response of lettuce plants to S. enterica serovar Dublin colonization and the presence of S. enterica serovar Dublin inside and outside lettuce plant tissues, it was suggested that the plant also responded on a molecular level. To determine the molecular response of lettuce during colonization by S. enterica serovar Dublin over time, gene expression analyses were performed using cDNA-AFLP analysis. The TDFs displayed from cDNA-AFLP analysis were obtained with 16 tested primer sets (EcoRI-T and Mse-NN). On average, for each primer set, 100 to 150 bands were observed, resulting in approximately 2,000 fragments that were analyzed. Although the majority of bands revealed no differential expression profiles between noninoculated and Salmonella-inoculated plants, discriminative bands were also found. In total, 170 bands were selected from cDNA-AFLP analysis, of which 90 bands showed differential expression profiles between both treatments (noninoculated and inoculated) over time. The 170 selected bands were sequenced, among which 68 consisted of more than one sequence, thus leaving 102 sequences valid for further analysis. The 102 sequences were compared, using tBlastX, against EST databases for L. sativa (22,185 EST) and A. thaliana (62,010 EST) (TIGR gene indices). On the basis of sequence homology, these 102 TDFs were grouped into 12 categories of putative function, followed by classification according to the system of Mahalingam et al. (31) (Table 3). Comparing differentially expressed TDFs versus generally expressed TDFs, an increase was observed for the categories related to plant disease/defense, transport, signal transduction, and hypothetical proteins (Table 3). A reduction in gene expression was observed for plant metabolism genes and genes with unknown functions.
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These results rely on the surface disinfection efficiency obtained with artificially inoculated leaves. This does not fully reflect a completely realistic situation. Naturally occurring bacteria are able to form a protective biofilm on the leaf surface that prevents the penetration of disinfectants and subsequent lysis of the bacterial cells (29). This would imply a less efficient surface disinfection in the case of naturally infected plants. On the other hand, biofilms are only (partially) protective against very mild disinfectants, such as chlorine, but are likely much less or even not protective against 70% ethanol. Moreover, from the tissue cross sections tested with fluorescence microscopy, it was evidenced that S. enterica serovar Dublin was present both inside and outside the plant, which suggests that the S. enterica serovar Dublin CFU found after surface disinfection were indeed present endophytically in the lettuce plant tissue.
The invasion process observed in this research was similar to the invasion of barley with S. enterica serovar Typhimurium (28). It was suggested that the invasion process of S. enterica serovar Typhimurium for colonizing plants is similar to that of plant pathogens (28), characterized by a three-phase process of Ralstonia solanacearum infecting hydroponically grown tomato plants (43). First the root surface is colonized, followed by infection of the vascular parenchyma and then invasion of the xylem. This three-phase process was also observed in this study.
S. enterica serovar Dublin was able to colonize lettuce plants endophytically and epiphytically, both under sterile growing conditions and in manure-amended soil. Epiphytic movement of Salmonella cells from the soil or medium to the aerial portions of the plant was allowed via capillary forces to retain a nondisturbed colonization and plant-microbe interaction. S. enterica serovar Dublin first colonized the root surface, reaching a high density of bacterial cells around naturally present openings or wounds. This is in line with demonstrated bacterial growth and rhizosphere colonization stimulated by root exudates (6, 8) and the observed biofilm formation in the lettuce cross sections described in this paper. Subsequently, invasion occurred via wounds that allowed the bacteria to colonize the roots intercellularly (9, 39), but also via intercellular spaces between epidermal cells. Indeed, S. enterica serovar Dublin was found in the parenchyma tissue, inside the pericycle, and attached to and inside the vascular system.
Typically, the stems of sterilely grown plants appeared constricted at the root-stem transition point several days after inoculation of the roots with S. enterica serovar Dublin. Lettuce may have responded in a hypersensitive manner to the intercellular presence of S. enterica serovar Dublin. This may have led to reduced nutrient flow, leaf yellowing, and finally plant death, herewith indicating that S. enterica serovar Dublin might be pathogenic to lettuce under these conditions. A critical point would be that the cell density used for inoculation was rather high for the soil experiments. Indeed, a concentration of 107 cells/g is not often found in the environment. However, this inoculum level was applied as a worst-case scenario to provide insight into the colonization efficiency of S. enterica serovar Dublin in soil-grown plants. To what extent lettuce is still colonized at lower S. enterica serovar Dublin cell densities still needs to be determined.
In view of the symptoms on plants inoculated with S. enterica serovar Dublin, inoculated plants apparently reacted physiologically to colonization by this human pathogen. We also demonstrated for the first time, using cDNA-AFLP analysis, that plant genes were differentially expressed between S. enterica serovar Dublin-inoculated and noninoculated plants. An increase in expression of pathogenicity-related genes was observed, which suggests a similar response of lettuce to colonization by S. enterica serovar Dublin to the response to plant pathogenic bacteria. The expression profiles of at least nine genes were strongly associated with the colonization of lettuce by S. enterica serovar Dublin. Next to four genes, DAD1 (33), PR1 (10), PR4 (38), and PR5 (42), that are known to be related to plant stress, five other genes that had differential profiles between colonized and noncolonized plants, namely, a NAM-like protein gene (13), OEE3 gene (40), PR1 gene (10), Sec1 family transport protein gene (2), Sec6 transport protein gene (46), and bHLH016 transcription factor gene (20), were also obtained from cDNA-AFLP analysis.
The NAM-like protein is involved in shoot development and leaf formation of Petunia plants (13). In line with this, plant growth was stunted when plants were colonized with S. enterica serovar Dublin and gene expression was reduced over time compared to that in healthy plants. Among basic helix-loop-helix (bHLH) transcription factor protein genes, the Arabidopsis genome encodes at least 150 putative bHLH class transcription factors, of which many play key roles in phytochrome signal transduction (20). These transcription factors are suggested to primarily act more as negative regulators than positive regulators of phytochrome signaling (12). Moreover, these bHLH proteins are found to interact specifically with phytochromes. For example, phytochrome-interacting factor 3 (PIF3) mainly acts as a negative regulator in the phytochrome B pathway but acts as a positive regulator of anthocyanin and chlorophyll accumulation (12). The expression profile of the bHLH016 gene identified in this study showed a decrease over time for the colonized plants but an increase over time for noninoculated plants. This difference can explain the development of symptoms such as leaf yellowing, implying a reduction of chlorophyll production, which is in line with the expression profile. In addition, yellow leaves also lead to less phytochrome translocation, which is induced by negative regulation of the bHLH proteins.
OEE3 is one of the three OEEs (OEE1, OEE2, and OEE3), which are nuclearly encoded chloroplast proteins that are bound to photosystem II (40). Reduction of their gene expression has a direct influence on photosystem II, leading to a reduction of photosynthesis (40). This might be a secondary effect of colonization with S. enterica serovar Dublin, since the bacterium is thought to narrow the stem and thus reduce the nutrient flow dramatically.
The expression profiles of the PR1 gene are similar to the previously published expression profile of the PR1 gene of Medicago colonized with Salmonella spp. (11), which implies an SA-directed defense mechanism of the plant upon colonization with Salmonella spp. These results suggest a pathogenicity-related response of lettuce to colonization by S. enterica serovar Dublin.
Interestingly, the expression of genes (sec1 and sec6) involved in the regulation and formation of the actin cytoskeleton (2, 46) was strongly affected during colonization. Inhibition of SNARE (soluble NSF attachment protein receptor) regulatory proteins consequently blocks mRNA transport by depolarization of the actin cytoskeleton, which will eventually lead to cell death. This might explain the fact that leaf yellowing (and eventually plant death) was observed for the colonized plants. The onset of the secretory block is thought to be activated by a specific signal that influences the actin regulatory machinery. The nature of this signal is not yet known. However, taking into account the results described in this paper (a strong reduction of sec1 and sec6 gene expression over time, in contrast to the case for healthy plants), it might be that the signal is activated by certain bacteria during colonization of the host plant. Since Salmonella spp. are known to change/disrupt the actin cytoskeleton prior to invasion of mammalian cells, it might very well be that specific secretory proteins of Salmonella spp. are responsible for the blocking of the SNARE regulatory proteins of lettuce. According to this hypothesis, the expression profiles of these genes might be related to a more specific than general response of lettuce upon colonization by S. enterica serovar Dublin. In that case, these genes would be designated as potential marker genes, which is especially of great interest with respect to food safety. This theory should be investigated further to better understand the molecular interaction between lettuce and Salmonella spp. during colonization and to clearly identify the presence of such marker genes.
In conclusion, previous studies postulated Pseudomonas aeruginosa and Staphylococcus aureus to be plant pathogenic (36). Whether S. enterica serovar Dublin can also be designated as a pathogen for L. sativa cv. Tamburo is not fully validated. Under sterile growing conditions, symptoms (leaf yellowing, stunting) were observed, whereas no symptoms could be observed on lettuce grown in soil. This might indicate that Tamburo lettuce is susceptible to S. enterica serovar Dublin. However, this does not mean that all lettuce cultivars would be equally susceptible to S. enterica serovar Dublin or that all strains of Salmonella enterica would equally efficiently colonize lettuce. A lettuce cultivar-Salmonella strain interaction study would be very interesting and valuable for agriculture and society in order to reduce or even prevent the risk of disease outbreaks related to the consumption of fresh produce.
|
ACKNOWLEDGMENTS |
|---|
We thank H. Aarts from RIKILT, The Netherlands, for providing Salmonella enterica serovar Dublin and M. Raats from Nickerson-Zwaan BV, The Netherlands, for providing the lettuce seeds.
|
FOOTNOTES |
|---|
Published ahead of print on 18 May 2007.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||
