Department of Biological Sciences, Faculty of Science, National University of Singapore, 10 Kent Ridge Crescent, Singapore 1192601
Author for correspondence: K. Y. Leung. Tel: +65 8747835. Fax: +65 7792486. e-mail: dbslky{at}nus.edu.sg
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Keywords: Edwardsiella tarda, GFP, carp epithelial cells, infection kinetics
Abbreviations: EPC, epithelioma papillosum of carp; GFP, green fluorescent protein; MAPK, mitogen-activated protein kinase
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Little information is available on what virulence factors may contribute to E. tarda infections. The pathogenic mechanism and the invasion pathways of E. tarda are not clear. Several virulence factors have been implicated, including haemolysins (Janda et al., 1991a ; Hirono et al., 1997
), dermatoxins (Ullah & Arai, 1983
), a lethal toxin (Suprapto et al., 1996
), anti-phagocyte killing (Ainsworth & Dexiang, 1990
), congo-red binding and serum resistance (Janda et al., 1991b
). There has been no study to correlate virulence factors of fish isolates with virulence in fish. In murine models, LD50 of human isolates did not correlate with plasmid content, chemotactic motility, serum resistance or exoenzymes including haemolysins (Janda et al., 1991b
). The ability to invade epithelial cells is a common virulence character in various fish pathogens, such as Aeromonas hydrophila (Tan et al., 1998
) and Vibrio species (Wang et al., 1998
). However, the ability of fish isolates of E. tarda to penetrate and replicate in mammalian epithelial cells is in question. Janda et al. (1991a
) reported that fish isolates were among a group that was not consistently positive for invasion of human HEp-2 cells. A human E. tarda isolate containing a mutation in the haemolysin gene was reported to be defective for entry into HEp-2 cells, suggesting that the phenotypes of cytotoxicity and invasion are linked (Strauss et al., 1997
). Information on interactions between E. tarda and fish cells is lacking and this area therefore requires study in order to understand the initial pathogenetic process in fish.
The recent introduction of the green fluorescent protein (GFP) from the jellyfish Aequorea victoria as an endogenous fluorescent tag promises to provide a means of both rendering the bacteria visible and tracing their activity in living host cells (Prasher et al., 1992 ; Chalfie et al., 1994
; Valdivia et al., 1996
). Many mutants of GFP have been generated for improved fluorescence (Crameri et al., 1995
). There are many advantages of tagging bacteria with GFP. It fluoresces independent of co-factors, substrates or any additional gene products. In addition, it is autonomous, sensitive, stable, specific, easy to use, non-toxic and does not interfere with cell growth and function (Chalfie et al., 1994
; Valdivia et al., 1996
). With such a tool, invasion pathways of E. tarda in fish and fish culture cells can be analysed using small doses of bacteria to mimic the process of naturally acquired E. tarda infections.
This study illustrates the usefulness of fluorescent proteins in tracking the fish pathogen, E. tarda, in vitro [epithelioma papillosum of carp (EPC) cells] and in vivo (blue gourami). Our purpose was to illustrate that E. tarda invade and replicate inside carp epithelial cells. At the same time, infections with virulent and avirulent strains were carried out in fish to study their different infection kinetics. By understanding the pathogenesis of edwardsiellosis, we hope to develop new approaches to combat this disease.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Survival assay in blue gourami serum.
Blood was collected from the caudal vein of naive blue gourami, and serum was separated from the clot by centrifugation at 4 °C. Bacteria grown in TSB at 25 °C for 24 h were collected by centrifugation and washed three times in PBS (137 mM NaCl, 4·3 mM Na2HPO4, 1·4 mM KH2PO4, pH 7·2). The cell suspension was mixed with blue gourami serum to give a final serum concentration of 50% and the bacterial count was adjusted to approximately 5x107 c.f.u. ml-1. The microcentrifuge tubes were incubated at 25 °C and samples were removed at 1 h for serial dilutions and plate counts on TSA. Serum survival ability was calculated by dividing the number of viable bacteria after the serum treatment by the number of viable bacteria before treatment, as described previously (Leung et al., 1995 ). A value greater than one was taken as indicating serum resistance and a value less than one was taken as indicating serum sensitivity.
LD50 determinations.
To assess the virulence of the E. tarda strains, LD50 determination was conducted using naive blue gourami, Trichogaster trichopterus (Pallas). The fish, each weighing approximately 13 g, were purchased from commercial fish farms and acclimatized for more than 1 month. Three groups of 10 fish were injected intramuscularly with 0·1 ml PBS-washed bacterial cells adjusted to the required concentrations. Fish were monitored for mortalities for 7 d and LD50 values were calculated by the method of Reed & Muench (1938) .
Cell culture.
All tissue culture reagents were obtained from Gibco. EPC cells from Cyprinus carpio (Wolf & Mann, 1980 ) were grown in minimal essential medium (MEM) with Hanks balanced salts solution (Sigma), 10 mM HEPES pH 7·3, 2 mM glutamine, 0·23% NaHCO3 and 10% (v/v) heat-inactivated foetal bovine serum at 25 °C in a 5% (v/v) CO2 atmosphere. Cells were grown in 75 cm2 flasks and divided at least once a week by trypsin/EDTA treatment and dilution at 1:10 in fresh media.
Cytoskeleton and signal transduction inhibitors.
Inhibitors and their concentrations used are listed in Table 3. The concentrations of inhibitors used were according to the manufacturers recommendations. For genistein and staurosporine, concentrations used were those described by Rosenshine et al. (1992)
. DMSO was used to dissolve all chemicals. Prior to use, the inhibitors were further diluted in MEM supplemented with 10% foetal bovine serum and added to the cultured fish cells.
|
Adherence and internalization assays.
These were performed as described by Elsinghorst (1994) and Wang et al. (1998)
with minor modifications. EPC monolayers were grown for 72 h in 24-well tissue culture plates to 100% confluence. They were then infected with E. tarda, at m.o.i.s of 1:1 and 1:3, for 30 min at 25 °C as described above. A control, Escherichia coli strain TOP10F' (Clontech) was also included. To measure the number of bacteria adhering to the monolayers, the monolayers were washed six times with Hanks salts, lysed with 1% (v/v) Triton X-100 in PBS, and then the bacterial numbers quantified by plate counting. To measure internalization, the monolayers were washed three times with Hanks salts to remove bacteria remaining in the medium and incubated for 2 h in MEM containing gentamicin (100 µg ml-1) (Sigma) to kill any residual extracellular bacteria. Afterwards, the EPC cells were washed three times with Hanks salts to remove any remaining gentamicin. One millilitre of 1% Triton X-100 in PBS was added to each well to lyse the EPC cells and the number of viable intracellular bacteria was quantified by plate counting. The adherence and internalization rates were calculated from the mean of two wells in quadruplicate experiments. In studies of different signal transduction inhibitors, EPC monolayers were incubated with each of the inhibitors for 30 min prior to infection and throughout the internalization assay. The number of viable intracellular bacteria was determined after 2·5 h as described above. The number of viable intracellular bacteria was expressed as a percentage of untreated controls and was determined in triplicate experiments. To ensure that the E. tarda strains were not affected by the inhibitors, bacterial viability counts were determined after 2·5 h of treatment. Viabilities of bacteria were expressed as a percentage of untreated controls and were determined in triplicate experiments.
Intracellular replication assay.
This was performed as described by Leung & Finlay (1991) . The intracellular replication of strains PPD130/91 and PPD125/87 was assessed as described for the internalization assay with the following modifications. Instead of treating the cells with 1% Triton X-100 in PBS after gentamicin treatment, the monolayers were washed with Hanks salts and fresh MEM containing 10 µg gentamicin ml-1 was added to inhibit any growth of the remaining extracellular bacteria. This medium was removed after incubation for the appropriate time (at 3·5, 4·5, 5·5 and 6·5 h post-inoculation), the monolayers were lysed with 1% Triton X-100 in PBS and the number of viable intracellular bacteria was quantified as described for the internalization assay.
Transformation of E. tarda strains with pGFPuv and pBFP2.
Plasmids pGFPuv and pBFP2 (Clontech) were transformed into E. tarda strains using a standard CaCl2 transformation protocol (Sambrook et al., 1989 ). Plasmids pGFPuv and pBFP2 have an ampicillin resistance marker and transformed E. tarda were plated out on TSA containing ampicillin. Colonies that were resistant to ampicillin and fluoresced bright green (pGFPuv) or blue (pBFP2) under UV light were selected.
Plasmid stability.
This assay was performed to check for the ability of transformed E. tarda PPD130/91(pGFPuv) and PPD125/87(pBFP2) to retain their plasmids. A single colony of the bacteria was inoculated into a test tube containing 5 ml TSB. This bacterial culture was continuously subcultured by reinoculating to another test tube containing 5 ml fresh TSB daily for 7 d. The bacterial culture was sampled daily for 7 d to quantify the bacterial number by plate counting. Plasmid stability was checked by examining the ability of the bacterial colonies to fluoresce with the aid of an UV lightbox.
Confocal microscopy and immunofluorescence study.
EPC cells were seeded onto glass coverslips and placed in each well of a 24-well tissue culture plate, as described above. The monolayers were then inoculated with 5 µl E. tarda tagged with pGFPuv, prepared as above. At appropriate time intervals, the coverslips were removed and examined using a Nikon Optiphot microscope attached to a Bio-Rad MRC 500 confocal system (Lasersharp). A blue laser, of excitation wavelength 488 nm, was used to view E. tarda tagged with pGFPuv. Optical sectioning of infected EPC cells was carried out serially at 2 µm intervals along the apical and basal axes of the cell.
Rhodamine phalloidin (Molecular Probes) was used to label actin in EPC cells according to the manufacturers instructions. Labelled actin was visible under the confocal laser scanning microscope using an excitation wavelength of 514 nm. Images obtained at x600 to x1800 were photographed with a Polaroid freeze-frame recorder using Kodak T-max 100 and Kodak colour ISO 100 film. The frequency of the co-localization of actin patches and bacteria was also measured as number of co-localizations per 100 EPC cells in triplicate experiments.
Study of infection kinetics of E. tarda in blue gourami.
The intramuscular route of administration was used in the study of the infection kinetics of E. tarda in vivo. There were five treatment groups four single infections and one mixed. For the single infections, fish were infected separately with E. tarda PPD130/91, PPD130/91(pGFPuv), PPD125/87 or PPD125/87(pBFP2) at an inoculum size of 105 c.f.u. In the mixed infection study, fish were infected with E. tarda PPD130/91(pGFPuv) and PPD125/87(pBFP2) in a 1:1 ratio, with 5x104 c.f.u. of each strain. The fish were challenged with 0·1 ml bacterial suspension via intramuscular injection. A control group of fish each received 0·1 ml sterile PBS. For 7 d, four fish from each group were sampled daily. The gall bladder, heart, intestine, kidney, liver and spleen were aseptically removed. Blood was collected aseptically from the caudal vein. A piece of body muscle dissected from the site of injection, measuring approximately 1x1 cm, was also taken. The samples from each treatment were pooled together by organ type and placed into sterile sample bags (Whirl-Pak). They were then diluted with 1 ml PBS and homogenized with a Stomacher Lab-Blender, model 80 (Seward Medical). The homogenates were serially diluted and plated in triplicate onto appropriate media, either TSA or TSA containing ampicillin, and incubated at 25 °C for 48 h. Bacterial counts were then done with the aid of an UV lightbox.
Histological examination of the body muscle and liver of blue gourami.
Fish were infected with PPD125/87(pBFP2), PPD130/91(pGFPuv) and the mixed inoculum as described above. A control group received only PBS. At 5 d post-inoculation, body muscle and whole liver samples were dissected from fish from each treatment group. These tissues were immediately mounted in Jung Tissue Freezing Medium (Leica Instruments) before being snap-frozen in liquid nitrogen and then cut into 15 µm sections on a cryostat (model CM1850; Leica Instruments). These tissue sections were then examined using an Axiovert 25 CFL phase-contrast inverted microscope (Carl Zeiss) at x200 and x1000 magnifications. To view the fluorescent bacteria, a blue filter of excitation wavelength 488 nm was used. Images were photographed with a camera attached to the microscope using Kodak colour ISO 100 film.
Statistical analysis.
All data from morphological, adherence and internalization assays were expressed as mean±SEM. The data were analysed using one-way ANOVA and a Duncan multiple range test (SAS software, SAS Institute). Values of P<0·05 were considered significant.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Interactions between E. tarda and EPC cells
All E. tarda strains examined adhered to and invaded EPC cells (Table 2). Adherence and internalization rates were similar regardless of the m.o.i. The adherence rates were always higher than internalization rates, with the exception of PPD130/91. The internalization rate of the virulent strains ranged from 6·7 to 39·7%. Among these strains, AL9379 had the highest internalization rate, together with a high adherence rate. Avirulent strains all had internalization rates which were less than 4·4%, except PPD499/84 and PPD125/87. These two strains exhibited greater adherence (>100%) to EPC cells, compared to the other avirulent strains. Adherence rates greater than 100% may be due to high adherence capability and rapid bacterial growth of these strains during the 30 min incubation in MEM. The Escherichia coli control did not adhere to or invade EPC cells, indicating that invasion by E. tarda is a specific process.
|
Intracellular replication of E. tarda strains in EPC cells
Both PPD130/91 and PPD125/87 were able to replicate inside EPC cells. The number of PPD130/91 inside the EPC cells increased significantly every hour throughout the experimental period, from 5·0x104 (2·5 h post-inoculation) to 7·0x105 (6·5 h post-inoculation). The number of PPD125/87 also increased from 1·4x105 c.f.u. (2·5 h post-inoculation) to 2·1x105 c.f.u. (3·5 h post-inoculation). Addition of gentamicin to inhibit any growth of the remaining extracellular bacteria slowed down morphological change in the PPD125/87-infected monolayers for 3·5 h. However, the intracellular bacterial population sampled at 4·5 h post-inoculation decreased (2·0x105 c.f.u.). At 6·5 h post-inoculation, there was further reduction in the intracellular population of PPD125/87 in the EPC cells (1·1x105 c.f.u.). This decrease in bacterial numbers correlated with the ongoing changes in EPC cell morphology and the monolayers were eventually lysed. Thus, the decrease in bacterial numbers could be due to intracellular bacteria being killed by extracellular gentamicin leaking into the non-intact EPC cell monolayers.
Effects of inhibitors on internalization of E. tarda strains in EPC cells
Bacterial uptake usually involves the exploitation of host cell functions to facilitate invasion. Different classes of inhibitors (Table 3) were used to better understand the roles of the cellular cytoskeleton and signal transduction pathways in the internalization of two E. tarda strains in EPC cells. The internalization of PPD130/91 and PPD125/87 was very sensitive to cytochalasin D, a specific blocker of actin polymerization. Inhibition of internalization by cytochalasin D was observed at concentrations as low as 0·1 µg ml-1. Therefore, it is most likely that these two strains enter EPC cells by triggering rearrangements of the actin microfilament system. Genistein, a specific tyrosine kinase inhibitor, significantly inhibited internalization of the two E. tarda strains, suggesting that E. tarda internalization requires the involvement of protein tyrosine kinase. On the other hand, staurosporine, a non-selective protein kinase C inhibitor, significantly accelerated internalization of PPD125/87. This chemical did not inhibit or accelerate the internalization of PPD130/91. PD098059, a mitogen-activated protein kinase (MAPK) kinase inhibitor, significantly inhibited the internalization of PPD130/91, but failed to have any effect on the internalization of PPD125/87 by EPC cells. None of the inhibitors interfered with bacterial viability or the viability of EPC cells (data not shown).
Transformation of E. tarda with pGFPuv and pBFP2
To study the invasion pathways of E. tarda inside EPC cells and in a fish host, pGFPuv and pBFP2 plasmids were transformed into these bacteria. These transformed strains fluoresced an intense green (pGFPuv) or blue (pBFP2) when illuminated with UV light. The stability of the two plasmids was tested in vitro, by subculturing in TSB daily for 7 d, using E. tarda PPD130/91(pGFPuv) and PPD125/87(pBFP2), respectively, and was found to be maintained at 100% (n=3) in the strains. In our experiments done in vivo, bacteria that were recovered from various organs obtained from PPD125/87(pBFP2)-infected fish still retained their fluorescence even on the fourteenth day after injection into the fish. These results suggest that fluorescent protein expression is very stable in the bacterial population. Transformed PPD125/87 and PPD130/91 had similar LD50 values, effects on EPC cell morphology, haemolysin production and serum survival ability as the untransformed strains (Table 1). Adherence rates of transformed PPD125/87 were lower but were still within the same range as other avirulent strains. Internalization rates of PPD125/87 after transformation remained the same. There was no change in the adherence and internalization rates of PPD130/91 after transformation with pGFPuv and pBFP2. Thus, the incorporation of pGFPuv and pBFP2 plasmids did not cause any significant changes in E. tarda virulence. We then chose PPD130/91(pGFPuv) and PPD125/87(pBFP2) to study the bacterialhost interactions in vitro and in vivo.
Evidence for E. tarda internalization
Internalized E. tarda tagged with fluorescent protein could be visualized under the confocal microscope. Early in infection, the virulent strain, PPD130/91, multiplied inside EPC cells without any significant morphological change in the EPC cells (data not shown). Many tightly clustered E. tarda PPD130/91(pGFPuv) were seen inside the host cells 30 min post-inoculation. The monolayer and the cell nuclei appeared to be intact despite the presence of intracellular bacteria. At 8 h post-inoculation (Fig. 1a), the cell morphology had totally changed, cells becoming round, transparent and distorted. The nuclei of these infected cells had also disintegrated completely. Similarly, the avirulent strain PPD125/87 was found inside host cells after infection. However, in contrast to infection with PPD130/91, infected cells were already rounded and distorted, with most of the monolayer detached from the coverslip in the early stages of infection (3 h post-inoculation, data not shown). Fig. 1(a)
is a high-power micrograph of an infected cell at its last stage (8 h post-inoculation) of infection with PPD130/91(pGFPuv). Bacteria were found inside the EPC cells by optical sectioning carried out serially at 2 µm intervals along the z axis. Many rod-shaped E. tarda could be seen clearly in the mid-region (middle plane) of the infected cell, suggesting the cluster of bacteria was intracellular.
|
Infection kinetics of E. tarda in fish
Single infection kinetics of PPD130/91(pGFPuv) (Fig. 2a) and PPD125/87(pBFP2) (Fig. 2b
) and their untransformed counterparts (data not shown) were found to be similar. In single infections with E. tarda PPD130/91(pGFPuv), high bacterial numbers were detected in blue gourami (Fig. 2a
). On day 1 of the experiment, they were highest in the body muscle bacteria had multiplied 400-fold (4·4x107 c.f.u.). Bacteria were also detected in high numbers in the liver, followed by the kidney, blood, heart, spleen and gall bladder. By day 5, the bacterial population in the muscle had proliferated to 2·1x109 c.f.u. In the fish organs, rapid bacterial growth was also observed highest in liver, then in the kidney, spleen, heart, gall bladder, blood and intestine. Mortalities in the infected fish were observed from day 3 onwards and all were dead by day 5 of the experiment. On the other hand, in the single infections with PPD125/87(pBFP2) (Fig. 2b
), there was a decline in the bacterial population over a period of 7 d. Although bacteria were detected in all parts of the fish on day 1, there was no significant growth. Body muscle only contained 3·1x104 c.f.u. of PPD125/87(pBFP2). By day 7 of the experiment, there were insignificant numbers of bacteria in all organs except the spleen. No mortality was observed in these infected fish. In the mixed infections, both avirulent and virulent E. tarda were detected in blue gourami. On day 1, the PPD125/87(pBFP2) population had multiplied 10-fold in the body muscle and growth continued until day 3 (Fig. 2d
), after which the population decreased. Numbers of PPD125/87(pBFP2) in the heart, kidney, spleen, liver and blood were higher than in the single infection. By day 5, the number of PPD125/87(pBFP2) in the fish had fallen to low levels, but were slightly higher in the spleen. PPD130/91(pGFPuv) had multiplied 1000-fold in the body muscle by day 1 and growth persisted over the experimental period (Fig. 2c
). However, the numbers of PPD130/91(pGFPuv) in the rest of the organs were lower than in the single infections and decreased by day 5. Despite the low numbers of both PPD130/91(pGFPuv) and PPD125/87(pBFP2) in the fish, mortalities were observed in the infected fish from day 3 onwards and the rest were dead by day 5 of the experiment.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
This is the first study on the interaction of E. tarda with fish epithelial cells in vitro. The epithelial cell line was used because it is likely to be the first cell type that a fish pathogen would encounter in vivo before establishing infection. Using our tissue culture model, E. tarda strains induced cytotoxic changes in EPC monolayers and two different types of pattern were observed. Rapidly cytotoxic (represented by PPD125/87) and slowly cytotoxic (represented by PPD130/91) strains have been identified among E. tarda (Table 1). In our study, rapid induction of cytotoxicity in EPC cells by E. tarda was not related to virulence in our fish model. Our results agreed with reports of other researchers. Janda et al. (1991a
) suggested that haemolysins and other exoenzymes were responsible for the cytotoxic effect observed in HEp-2 cells. Later, they reported that haemolysins and other exoenzymes in E. tarda did not play a major role in bacterial virulence and subsequent establishment of disease in mice (Janda et al., 1991b
). The slowly cytotoxic strains, such as PPD130/91 and AL9379, did not have drastic effects on host cell morphology for an extended period of time. The rapid continuous growth of E. tarda in MEM, after a long time, would eventually affect the integrity of the EPC monolayers. This failure to induce cytotoxic effects rapidly in vitro is also seen in human pathogens such as Shigella flexneri, Y. enterocolitica, and Salmonella species (Finlay & Falkow, 1988
, 1989
).
Adherence to and invasion of host cells are often the initial steps of the pathway taken by many pathogenic bacteria (Finlay & Cossart, 1997 ; Finlay & Falkow, 1997
). In our study, all the E. tarda strains adhered to and invaded EPC cells (Table 2
). The adherence and internalization rates of E. tarda were much greater than those of other fish pathogens such as Vibrio species (Wang et al., 1998
). The virulent strain, PPD130/91, entered fish cells efficiently, being found intracellularly at 30 min post-inoculation (data not shown).
Usually, antibody-labelled bacteria are visualized after fixing cells at different time points during infection, but this limits the tracing of the pathway taken by the live bacteria. Fluorescein dyes and commercial labels such as Live and Dead (Molecular Probes) fade rapidly due to photobleaching and newly proliferated bacteria do not have sufficient stain to give a strong signal. Thus the true total bacterial population is underestimated. These conventional methods are also time-consuming and inconvenient. With GFP, E. tarda PPD130/91 was visualized easily under the confocal or epifluorescence microscope. These bacteria were able to survive and multiply over time while the EPC monolayer remained intact (data not shown). Our intracellular replication assays further supported the suggestion that PPD130/91 was able to multiply rapidly inside EPC cells. In later stages of infection, degeneration of cell nuclei and other cellular organelles could also be observed (Fig. 1a).
The mechanism employed by E. tarda to enter fish epithelial cells appears to be microfilament dependent (Fig. 1b). Cytochalasin D, a drug which inhibits actin filament polymerization, significantly inhibited the internalization of both PPD130/91 and PPD125/87 by EPC cells. Similar observations have also been reported by Janda et al. (1991a)
on a human isolate of E. tarda where inhibitors of microfilaments (cytochalasins B and D) affected the invasive process. Polymerized actin filaments are part of the host cytoskeletal structure responsible for many active cell processes such as phagocytosis, cell movement and membrane ruffling. Actin polymerization has been associated with bacterial internalization by host cells in A. hydrophila (Tan et al., 1998
), Shigella flexneri (Clerc & Sansonetti, 1987
) and Salmonella spp. (Finlay et al., 1991
). GFP allowed us to visualize the association of E. tarda with actin polymerization.
In addition to recruitment of host cytoskeletal elements, there is increasing evidence for the induction of host cell signalling pathways by invasive bacteria. These bacteria are able to exploit the host cells existing signalling pathways, including activation of protein tyrosine kinases and of phospholipase C, leading to increased intracellular Ca2+ levels and protein kinase C activity, to stimulate their uptake. In the present study, genistein, a specific inhibitor of protein tyrosine kinase with negligible effect on the activity of other serine and threonine kinases (Akiyama et al., 1987 ), was found to inhibit the internalization of both PPD130/91 and PPD125/87. This finding suggests that E. tarda internalization involves protein tyrosine kinase activation. This seems to be a common mechanism among fish pathogens, including Vibrio spp. (Wang et al., 1998
) and A. hydrophila (Tan et al., 1998
), which also require the activation of protein tyrosine kinase for internalization. Staurosporine, a relatively non-specific protein kinase C inhibitor (Tamaoki et al., 1986
), substantially accelerated the internalization of PPD125/87. This acceleration of invasion by staurosporine might be associated with some signalling pathways other than protein kinase C. It has been shown in rat PC12 cells that staurosporine enhanced tyrosine phosphorylation of certain proteins (Rasouly & Lazarovici, 1994
). If this is the case, then activation of tyrosine kinases may have enhanced the invasion of E. tarda and inhibition of tyrosine kinases with genistein blocked the internalization process. The internalization of E. tarda PPD130/91 by EPC cells probably involved MAPK kinase since PD098059, a MAPK kinase inhibitor, blocked internalization.
The pathogenesis of Edwardsiella infection is a complex interplay between the bacteria and different types of host cells. Although tissue culture models may be invaluable in gaining further understanding about the disease under study (Quinn et al., 1997 ), it is essential to examine the invasion pathway taken in a natural host to fully understand the pathogenesis of bacterial infections. Only a few studies have been done on infection kinetics of E. tarda in fish (Kusuda & Ishihara, 1981
; Rashid et al., 1997
) and the pathogenesis of edwardsiellosis is not fully understood. In the previous studies, antibiotics were used as markers to tag E. tarda. This conventional method of bacterial detection is indirect and may overestimate the number of bacteria of interest. This point is important when bacterial sampling is done in the gastrointestinal tract of the experimental animal where a wide variety of bacteria is present. Identification of the bacteria introduced would be tedious and may not be 100% accurate. The use of fluorescent proteins is a direct and visual means of detection, effective in screening in the presence of other bacteria. In our study, the two chosen strains were characterized by various virulence factors and used as representatives of virulent and avirulent groups. An intramuscular infection model was used to simulate infection by physical injury under natural conditions. From our results, the ability of both virulent PPD130/91(pGFPuv) and avirulent PPD125/87(pBFP2) to spread to various organs within one day (Fig. 2a
, b
) indicated that they might be able to survive in the muscle and blood of blue gourami. Although both strains were serum-resistant (Table 1
), they exhibited contrasting infection kinetics. PPD130/91(pGFPuv) proliferated very rapidly in all the fish organs (Fig. 2a
), suggesting the presence of additional virulence factors that protected it against the host defence system. Bacterial numbers were highest in the body muscle, but there were substantial numbers in the other organs. In contrast, avirulent PPD125/87(pBFP2) was unable to grow inside the fish (Fig. 2b
). Thus, virulence in E. tarda is multifactorial and is related to additional factors besides serum resistance, possibly including factors such as haemagglutination, production of haemolysins and siderophores, and ability to bind Congo red, as reported by Janda et al. (1991b)
.
In addition to examination of single infections, studies of mixed infections are also possible with two bacterial strains expressing different fluorescent proteins. Using fluorescent proteins, two bacterial populations can be differentiated confidently and easily. These infections yielded interesting information on the ability of two bacterial strains, each at doses below the LD50, to complement the pathogenicity of each other. In the presence of PPD130/91(pGFPuv), PPD125/87(pBFP2) was able to proliferate over the first day after infection, with bacterial numbers in the body muscle as high as 107 cells ml-1 by day 3 (Fig. 2d). PPD125/87(pBFP2) may have gained some protection against the host defences from PPD130/91(pGFPuv). However, the benefits that PPD125/87(pBFP2) gained were short-lived, as seen by the subsequent decline in its population (Fig. 2d
) and also in its inability to spread throughout the fish. Establishment of avirulent bacteria in the muscle early in infection may be possible because the body of a non-immune host requires time to detect and elicit its defences against a new foreign agent (Iwama & Nakanishi, 1996
). This may explain why PPD125/87(pBFP2) could proliferate in the muscle in the first 3 d after infection. After this time, the ensuing inflammation may elicit more effective defences. PPD125/87(pBFP2) may not be able to resist these host responses, unlike the virulent PPD130/91(pGFPuv) strain. Mortalities similar to those in single infections with PPD130/91(pGFPuv) were noted. High numbers of PPD130/91(pGFPuv) in the muscle and their presence in other organs of the fish may have contributed to the mortalities. Tissue culture studies indicated that both avirulent PPD125/87 and virulent PPD130/91 have the ability to adhere to and invade host cells, but avirulent E. tarda appears to lack the ability to spread in vivo (Evelyn, 1996
).
Fluorescent proteins allow visualization of the association of bacteria with the host tissues in infection experiments. Histopathological studies confirmed the presence of PPD130/91(pGFPuv) in fish infected with this strain and the mixed inoculum (Fig. 3). Histological sections of fish that were infected with avirulent PPD125/87(pBFP2) appeared to be similar to those from uninfected fish. Conventional histology using indirect antibody labelling requires longer processing times for bacterial detection, and is tedious and expensive. Using our methodology, bacteria in cryostat sections were directly visualized without any immunological treatment.
Our present study describes a novel method, using fluorescent proteins, to conveniently study PPD130/91(pGFPuv) and PPD125/87(pBFP2) in vitro and in vivo. We have found that avirulent and virulent E. tarda strains are mostly serum-resistant and are able to adhere to, enter and replicate in fish cells using internalization mechanisms involving host microfilaments and protein tyrosine kinase. Virulent PPD130/91(pGFPuv) has additional virulence factors that result in establishment of disease in fish. Fluorescent proteins have opened up the possibility of following infection by microscopy. Therefore, in future it will be possible to follow the bacterial infection at various time points, especially early in infection, using very low doses of E. tarda. Using this tool, bacterialhost interactions can be further investigated in pursuit of the identification of the virulence factors involved, thus contributing to our understanding of the pathogenesis of Edwardsiella infections.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S., Itoh, N., Shibuya, M. & Fukami, Y. (1987). Genistein, a specific inhibitor of tyrosine-specific protein kinases. J Biol Chem 262, 5592-5595.
Amandi, A., Hiu, S. F., Rohovec, J. S. & Fryer, J. L. (1982). Isolation and characterization of Edwardsiella tarda from fall chinook salmon (Oncorhynchus tshawytscha). Appl Environ Microbiol 43, 1380-1384.[Medline]
Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W. & Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263, 802-805.[Medline]
Clerc, P. & Sansonetti, P. J. (1987). Entry of Shigella flexneri into HeLa cells: evidence for directed phagocytosis involving actin polymerization and myosin accumulation. Infect Immun 55, 2681-2688.[Medline]
Crameri, A., Whitehorn, E. A., Tate, E. & Stemmer, W. P. C. (1995). Improved green fluorescence protein by molecular evolution using DNA shuffling. Nature Biotechnol 14, 315-319.
Elsinghorst, E. A. (1994). Measurement of invasion by gentamicin resistance. Methods Enzymol 236, 405-420.[Medline]
Evelyn, T. P. T. (1996). Infection and disease. In The Fish Immune System Organism, Pathogen, and Environment, pp. 339-359. Edited by G. Iwama & T. Nakanishi. San Diego: Academic Press.
Farmer, J. J. & McWhorter, A. C. (1984). Genus X. Edwardsiella. In Bergeys Manual of Systematic Bacteriology, pp. 486-491. Edited by J. G. Holt & N. R. Krieg. Baltimore: Williams & Wilkins.
Finlay, B. B. & Cossart, P. (1997). Exploitation of mammalian host cell functions by bacterial pathogens. Science 276, 718-725.
Finlay, B. B. & Falkow, S. (1988). Comparison of the invasion strategies used by Salmonella cholerae-suis, Shigella flexneri and Yersinia enterocolitica to enter cultured animal cells: endosome acidification is not required for bacterial invasion or intracellular replication. Biochimie 70, 1089-1099.[Medline]
Finlay, B. B. & Falkow, S. (1989). Salmonella as an intracellular parasite. Mol Microbiol 3, 1833-1841.[Medline]
Finlay, B. B. & Falkow, S. (1997). Common themes in microbial pathogenicity revisited. Microbiol Mol Biol Rev 61, 136-169.[Abstract]
Finlay, B. B., Ruschkowski, S. & Dedhar, S. (1991). Cytoskeletal rearrangements accompanying Salmonella entry into epithelial cells. J Cell Sci 99, 283-296.[Abstract]
Hirono, I., Tange, N. & Aoki, T. (1997). Iron-regulated haemolysin gene from Edwardsiella tarda. Mol Microbiol 24, 851-856.[Medline]
Janda, J. M., Abbott, S. L. & Oshiro, L. S. (1991a). Penetration and replication of Edwardsiella spp. in HEp-2 cells. Infect Immun 59, 154-161.[Medline]
Janda, J. M., Abbott, S. L., Kroske-Bystrom, S., Cheung, W. K. W., Powers, C., Kokka, R. P. & Tamura, K. (1991b). Pathogenic properties of Edwardsiella species. J Clin Microbiol 29, 1997-2001.[Medline]
Kubota, S. S., Kaige, N., Miyazaki, T. & Miyashita, T. (1981). Histopathological studies on edwardsiellosis of tilapia. I. Natural infection. Bull Fac Fish Mie Univ 9, 155-165.
Kusuda, R. & Ishihara, S. (1981). The fate of Edwardsiella tarda bacteria after intramuscular injection of eels, Anguilla japonica. Bull Jpn Soc Sci Fish 47, 475-479.
Leung, K. Y. & Finlay, B. B. (1991). Intracellular replication is essential for the virulence of Salmonella typhimurium. Proc Natl Acad Sci USA 88, 11470-11474.[Abstract]
Leung, K. Y., Yeap, I. V., Lam, T. J. & Sin, Y. M. (1995). Serum resistance as a good indicator for virulence in Aeromonas hydrophila strains isolated from diseased fish in South-East Asia. J Fish Dis 18, 511-518.
Meyer, F. P. & Bullock, G. L. (1973). Edwardsiella tarda, a new pathogen of channel catfish (Ictalurus punctatus). Appl Microbiol 25, 155-156.[Medline]
Nakatsugawa, T. (1983). Edwardsiella tarda isolated from cultured young flounder. Fish Pathol 18, 99-101.
Plumb, J. A. (1993). Bacterial Diseases of Fish. Edited by N. R. Bromage, V. Inglis, & R. J. Roberts. Oxford: Blackwell Scientific Publications.
Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendgast, F. G. & Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111, 229-233.[Medline]
Quinn, F. D., Newman, G. W. & King, C. H. (1997). In search of virulence factors of human bacterial disease. Trends Microbiol 5, 20-26.[Medline]
Rashid, M. M., Nakai, T., Muroga, K. & Miyazaki, T. (1997). Pathogenesis of experimental edwardsiellosis in Japanese flounder Paralichthys olivaceus. Fish Sci 63, 384-387.
Rasouly, D. & Lazarovici, P. (1994). Staurosporine induces tyrosine phosphorylation of a 145 kDa protein but does not activate gp140trk in PC12 cells. Eur J Pharmacol 269, 255-264.[Medline]
Reed, L. J. & Muench, H. (1938). A simple method of estimating fifty per cent end points. Am J Hyg 27, 493-497.
Rosenshine, I., Donnenberg, M. S., Kaper, J. B. & Finlay, B. B. (1992). Signal transduction between enteropathogenic Escherichia coli (EPEC) and epithelial cells: EPEC induces tyrosine phosphorylation of host cell proteins to initiate cytoskeletal rearrangement and bacterial uptake. EMBO J 11, 3551-3560.[Abstract]
Sae-Oui, D., Muroga, K. & Nakai, T. (1984). A case of Edwardsiella tarda in cultured colored carp, Cyprinus carpio. Fish Pathol 19, 197-199.
Sambrook J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Strauss, E. J., Ghori, N. & Falkow, S. (1997). An Edwardsiella tarda strain containing a mutation in a gene with homology to shlB and hpmB is defective for entry into epithelial cells in culture. Infect Immun 65, 3924-3932.[Abstract]
Suprapto, H., Hara, T., Nakai, T. & Muroga, K. (1996). Purification of a lethal toxin of Edwardsiella tarda. Fish Pathol 31, 203-207.
Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto, M. & Tomita, F. (1986). Staurosporine, a potent inhibitor of phospholipid/Ca2+ dependent protein kinase. Biochem Biophys Res Commun 135, 397-402.[Medline]
Tan, E., Low, K. W., Wong, W. S. F. & Leung, K. Y. (1998). Internalization of Aeromonas hydrophila by fish cells can be inhibited with a tyrosine kinase inhibitor. Microbiology 144, 299-307.[Abstract]
Ullah, M. A. & Arai, T. (1983). Pathological activities of the naturally occurring strains of Edwardsiella tarda. Fish Pathol 18, 65-70.
Valdivia, R. H., Hromockyj, A. E., Monack, D., Ramakrishnan, L. & Falkow, S. (1996). Applications for green fluorescent protein (GFP) in the study of hostpathogen interactions. Gene 173, 47-52.[Medline]
Van Damme, L. R. & Vandepitte, J. (1984). Isolation of Edwardsiella tarda and Plesiomonas shigelloides from mammals and birds in Zaire. Rev Elev Med Vet Pays Trop 37, 145-151.[Medline]
Wakabayashi, H. & Egusa, S. (1973). Edwardsiella tarda (Paracolobactrum anguillmortiferum) associated with pond-cultured eel disease. Bull Jpn Soc Sci Fish 39, 931-936.
Wang, X. H., Oon, H. L., Ho, G. W. P., Wong, W. S. F., Lim, T. M. & Leung, K. Y. (1998). Internalization and cytotoxicity are important virulence mechanisms in vibriofish epithelial cell interactions. Microbiology 144, 2987-3002.[Abstract]
White, F. H., Simpson, C. F. & Williams, L. E. (1973). Isolation of Edwardsiella tarda from aquatic animal species and surface waters in Florida. J Wildl Dis 9, 204-207.[Medline]
Wolf, K. & Mann, J. A. (1980). Poikilotherm vertebrate cell lines and viruses, a current listing for fishes. In Vitro 16, 168-179.[Medline]
Received 21 June 1999;
revised 20 September 1999;
accepted 4 October 1999.