The aim of this study was to investigate the role of invasin in a bacterial fish pathogen Edwardsiella tarda.
The aim of this study was to investigate the role of invasin in a bacterial fish pathogen Edwardsiella tarda.
In this study, an in-frame deletion mutant of invasin (Δinv) in Edw. tarda H1 was constructed through double crossover allelic exchange to explore the function of invasin in virulence to fish. Meanwhile, an invasin overexpression strain (inv+) was obtained by electrotransformation of a low-copy plasmid pACYC184 carrying the intact invasin into the Δinv mutant. Several virulence-associated characters of the mutants and wild-type strain were tested. Compared with the wild-type H1, haemolytic activity and biofilm formation were decreased in Δinv, while increased significantly in inv+. In addition, the invasin overexpressing strain inv+ exhibited increased internalization into Epithelioma Papulosum Cyprini (EPC) cells. Moreover, in zebrafish model, Δinv showed decreased virulence compared with H1, while inv+ restored the virulence of wild type completely.
The results demonstrated that invasin of Edw. tarda plays essential roles in haemolytic activity, biofilm formation, adherence, internalization and pathogenicity of this bacterium.
This study revealed the role of invasin in Edw. tarda infection and provided useful information for further unveiling the pathogenesis of Edw. tarda.
Edwardsiella tarda is a Gram-negative bacterial pathogen with a broad host range including fish and humans (Janda and Abbott 1993a). Edw. tarda infection is responsible for tremendous economical loss of a variety of cultured fish in Asia, especially Japan and India, and also in the United States (Herman and Bullock 1986; Castro et al. 2006; Lima et al. 2008). It causes systemic infections in fish by entering skin, gill and intestine, where it invades several kinds of cells such as epithelial cells and macrophages and replicates intracellularly (Ling et al. 2000; Wang et al. 2010). Several virulent factors of Edw. tarda have been identified, including abilities to invade epithelial cells (Janda et al. 1991), haemolysins (Chen et al. 1996; Hirono et al. 1997), siderophores, serum resistance (Mathew et al. 2001) and biofilm formation (Zhang et al. 2008).
In many infectious diseases, invasion of epithelial cells by bacterial pathogens is associated with the initiation of infection (Meyer et al. 1997). Invasins, which have previously been identified to be either bacterial surface proteins or host cell receptors, are required to initiate the internalization of a bacterium into the host organism (Cossart and Sansonetti 2004). Invasin of Yersinia is an outer membrane protein required for efficient uptake of the bacteria into M cells (microfold cells) (Isberg et al. 1987). The N-terminal approx. 500 amino acids of Yersinia pseudotuberculosis invasin reside in outer membrane, while the C-terminal 497 residues are considered to bind integrins (Hamburger et al. 1999). However, little is known about the factors that mediate adherence and invasion of Edw. tarda to fish cells. A putative invasin protein (750 amino acids) was identified in the genome sequence of Edw. tarda, but its significance and detailed function are still not clear.
In this study, the function of the putative invasin of Edw. tarda was explored. An in-frame deletion mutant of invasin (Δinv) and the corresponding complemented strain (inv+) of Edw. tarda H1 were constructed. Virulence-associated properties, including the haemolytic activity, biofilm formation, invasion to Epithelioma Papulosum Cyprini (EPC) cells (Srinivasa Rao et al. 2001), and the lethality in fish model, were investigated. This study constitutes an effort to further unravel the invasion mechanisms of Edw. tarda, to develop an attenuated vaccine to combat Edw. tarda infection.
The zebrafish (Dario rerio) used for virulence tests in this study are cultured animals, and all the experiments are carried out in strict accordance with the regulations of local government.
Bacterial strains and plasmids used in this study were described in Table 1. Edw. tarda H1 was isolated from the ulcered muscle of diseased turbot (Scophthalmus maximus) with ‘red body’ in an aquaculture farm of Shandong province in August 2007 (Dong et al. 2009). The authenticity of Edw. tarda H1 was verified by 16S rRNA gene sequencing and gyrB-based PCR test (Lan et al. 2008), and the pathogenicity was verified using the zebrafish model. Plasmids were introduced into Escherichia coli strains by transformation and into Edw. tarda strains by mating with E. coli SY17-1 (λpir). Edw. tarda strains were grown at 28°C in Tryptic Soy broth (TSB) or on Tryptic Soy agar (TSA), while E. coli strains were grown at 37°C in Luria broth (LB). When required, antibiotics (Sigma, St Louis, MO, USA) were added to the medium at the following final concentrations: ampicillin (Amp, 50 μg ml−1), chloramphenicol (Cm, 25 μg ml−1) and polymyxin B (PxB, 12·5 μg ml−1).
|Strains or plasmids||Characteristics||References or source|
|H1||Pathogen isolated from a mariculture farm in Wenden, China. PxBr, Tcr||Dong et al. (2009)|
|Δinv||H1, in-frame deletion of inv||This study|
|inv +||H1, overexpression of inv in Δinv complemented with intact inv||This study|
|SY327 (λpir)|| |
Δ (lac pro) argE(Am) rif malA recA56
rpoB λ pir, host for π-requiring plasmids
|Miller and Mekalanos (1984)|
|SY17-1 (λpir)|| |
Tpr Smr recA thi pro rK- mK-
RP4:2-Tc:MuKm Tn7 λ pir (thi pro hsdR hsdM+ recA RP4-2-Tc:Mu-Km-Tn7)
|Simon et al. (1983)|
|JM109||endA1 hsdR17 gyrA96 Δ(lac proA) recA1 relA supE44 thi F′ (lacIq lacZΔM15 proAB+ traD36)||Yanisch-Perron et al. (1985)|
|pUCm-T||Cloning vector, Ampr||Sangon, Shanghai|
|pRE112||pGP704 suicide plasmid, pir dependent, oriT, oriV, sacB, Cmr||Edwards et al. (1998)|
|pRE112Δinv||pRE112 derivative containing inv bp1-48 fused in-frame to bp2053-2253, Cmr||This study|
|pACYC184||Cmr, Tcr||Fermentas Life Sciences|
Plasmid DNA and genome DNA preparation, recombinant DNA techniques, bacterial transformation and agarose gel electrophoresis were carried out following standard techniques described by Frederick (1995). All enzymes (TaKaRa, Dalian, China) were used according to the manufacturer's instructions.
All primers used in this study were listed in Table 2. The invasin gene of Edw. tarda H1 has an open reading frame of 2253 base pairs, which is identical to ETAE_3034 in Edw. tarda EIB202 (Wang et al. 2009). The invasin mutant (Δinv) of Edw. tarda H1 was constructed using an in-frame deletion strategy. Initially, PCR amplifications were performed to generate the upstream fragment (292 bp) and the downstream part (684 bp) of inv with primer pairs inv-V1/inv-V2 and inv-V3/inv-V4, respectively. The fusion of the two fragments was performed by overlap PCR (Mo et al. 2007) using primers inv-V1 and inv-V4. The fused segment was then cloned into suicide vector pRE112 to produce the insertional construct. The resulting construct was transformed into E. coli SY327 (λpir) and then introduced into E. coli SY17-1 (λpir) (Liang et al. 2003) for mating into Edw. tarda H1 by conjugation. The transconjugants with single-crossover insertion in the H1 chromosome were obtained by screening on TSA plates containing Cm and PxB. Allelic exchange between the chromosomal gene and the in-frame deleted plasmid copy was achieved in a second crossover event, which was counter-selected on TSA containing 10% sucrose to determine the excision of pRE112 from the chromosome. The target mutant was screened by Cm susceptibility and PxB resistance and determined by PCR using primers inv-V1/inv-V4.
To construct the complementary strain inv+, a 2·46-kb fragment containing the putative promoter and open reading frame of invasin was amplified using primers inv-U and inv-D (Table 2). The PCR product was introduced into the BamHI and SphI sites of low-copy plasmid pACYC184. The constructed plasmid was electrotransfered into the Δinv mutant to obtain the complementary strain (inv+). Cm- and PxB-resistant transformants were screened, and the presence of the plasmid was confirmed by PCR analysis and sequencing.
Expression of invasin operon in inv+ was monitored by qRT-PCR analysis. Different Edw. tarda strains were grown in TSB to OD540 of 0·5, and total RNA was extracted from each strain as described by Simms et al. (1993). The RNA was subjected to DNase I treatment to exclude the genomic DNA contamination. The first-strand cDNA was synthesized using equal amount (1 μg) of DNase I–treated RNA as template and primed with random primers (6 mer). The qRT-PCR was carried out in an ABI 7500 detector (Applied Biosystems, Foster City,CA, USA) using the SYBR ExScript RT-PCR kit (Takara). Each assay was performed in triplicate using 16S rRNA gene as the reference gene according to the method described by Tian et al. (2008). The primers (16s-qF/16s-qR, inv-qF/inv-qR) for qRT-PCR (Table 2) were designed using Primer Express software (Applied Biosystems) with predicted products in the 100–200 bp size range.
Overnight cultures of Edw. tarda strains H1, Δinv and inv+ were adjusted to the same density based on OD540, and 300 μl bacterial cultures were diluted 1 : 100 into 300 ml of fresh LB in triplicates and then cultivated at 28°C with constant agitation. The optical absorbance at 540 nm was measured every 2 h until the bacteria reached the stationary phase.
Edwardsiella tarda strains H1, Δinv and inv+ were point inoculated on TSA plates containing 0·3% (w/v) agar, and the swimming motility was evaluated by measuring the swimming distance of the bacteria. Swarming motility was assessed on TSA plates containing 0·5% (w/v) agar and 5% (w/v) glucose as described by Rashid and Kornberg (2000). The plates were analysed after incubation for 24 h. Each experiment was performed in triplicate.
Haemolysis assay was performed as described by Xu et al. (2010). Edwardsiella tarda strains grown in TSB for 12 h were harvested by centrifugation at 5000 g, washed with PBS and adjusted to 1 × 108 cells ml−1. Sheep erythrocytes were washed three times in PBS and resuspended in M9 medium (Collins and Thune 1996) at the final concentration of 10% (v/v). One hundred microlitres of bacterial culture was inoculated into 5 ml of M9 medium containing 10% erythrocytes and then cultured at 28°C with constant agitation. The OD540 of each sample was tested at 3 h intervals after 200 μl of the supernatant was collected by centrifugation (5000 g, 5 min).
Biofilm assay was performed as described previously (Stepanović et al. 2007) with minor modification. Briefly, overnight cultures of Edw. tarda were adjusted to OD540 of 0·5 and diluted 1 : 20 with fresh TSB. Two hundred microlitres of the cell suspension was inoculated into 96-well microtiter-plates. Wells containing uninoculated growth medium were used as negative control. The plates were incubated statically at 28°C for 2 days. The suspensions were removed, and the wells were washed three times with sterile distilled water before methanol fixation for 20 min. The methanol was then removed, and the plates were inverted for at least 30 min. The adherent bacteria were strained with 150 μl 2% crystal violet for 15 min at room temperature and washed gently with water. The plates were then air-dried, followed by redissolving the samples with 150 μl of 95% ethanol for 30 min. Biofilm formation was measured at 570 nm using an ELISA reader (SUNRISE™ Tecan Group Ltd, Männedorf, Switzerland).
Adherence and internalization assays were performed as described by Mathew et al. (2001) with minor modification. The EPC cell line (Fijan et al. 1983) was derived from fathead minnow Pimephales promelas and maintained in minimal essential medium (MEM) (Gibco, Carlsbad, CA, USA) supplemented with 10% sterile foetal bovine serum (FBS) (Gibco) at 25°C. Briefly, EPC cells were grown in 24-well tissue culture plates to 100% confluence. Overnight cultures of Edw. tarda were washed three times with PBS and added to EPC monolayers at a multiplicity of infection (MOI) of 10 : 1 and incubated for 3 h at 28°C. To measure adherence, the monolayers were washed three times with PBS, lysed with 1% (v/v) Triton X-100 in PBS, and the bacterial numbers were quantified by plate counting. To measure internalization, the monolayers were washed three times with PBS and incubated for an additional 2 h in MEM containing 200 μg ml−1 gentamycin to kill extracellular bacteria. The monolayers were then washed three times with PBS, lysed with 1% Triton X-100 in PBS, and the bacterial numbers were quantified by plate counting. The adherence and internalization rates were calculated from the mean of three wells in triplicate experiments.
The LD50 values were determined with zebrafish (Dario rerio) according to He et al. (2012). Healthy zebrafish (average weight = 0·3 g) from quarantined stocks were acclimatized in laboratory conditions at 22°C for more than 1 week. Three paralleled groups containing 10 fish each were injected intraperitoneally with 20 μl of bacterial suspension at 103–107 CFU ml−1. The control groups were injected with 20 μl of PBS. Mortalities were recorded over a period of 2 weeks after infection. The LD50 values were calculated as described by Wardlaw (1995).
Statistical analysis was performed using SPSS 19.0 software (IBM Corporation, Chicago, IL, USA), with the Tukey–Kramer test for multiple comparisons. Results were expressed as mean ± SEM with P < 0·05 and P < 0·01 taken to show statistical significance and distinct significance, respectively.
Using the double-selection strategy of allelic exchange mutagenesis with suicide vector pRE112, we deleted a 2004-bp core portion of the invasin and obtained the Δinv mutant with loss of an internal region of invasin from 16 to 684th amino acid residues. The mutant was identified by the resistance to PxB and inability to grow with Cm. The correct deletion was confirmed by DNA sequencing of the resulting PCR product.
Complementary strain was constructed by electrotransformation of the plasmid pACYC184 harbouring the intact invasin gene into Δinv. It was selected by PxB- and Cm- resistant, and the existence of the plasmid was confirmed by PCR analysis and sequencing. The Δinv and inv+ mutants were grown on TSA for thirty generations, indicating that they were steadily maintained.
The in-frame deletion mutant Δinv was unable to transcribe invasin to mRNA as determined by qRT-PCR analysis, while the expression level of invasin in inv+ increased by 163-fold compared with the wild-type H1. Therefore, we redefined inv+ as an invasin overexpression Edw. tarda strain.
Edwardsiella tarda H1, Δinv and inv+ grew at the same speed and represented similar growth patterns (data not show), and all the strains reached the maximum cell density at 16 h, indicating that invasin has no effect on growth of Edw. tarda.
There were no differences among strains of Edw. tarda H1, Δinv and inv+ in swimming (P < 0·05) and swarming motility (Table 3). They all had the ability to swim and swarm. This result demonstrated that invasin is not essential for the motility in Edw. tarda.
|Swimming motility (mm)||9·6 ± 1·8||8·9 ± 0·9||9·4 ± 1·5|
|Adherence to EPC (%)||19·8 ± 1·3||17·8 ± 0·8||21·5 ± 1·7|
|Internalization into EPC (%)||2·5 ± 0·3||2·2 ± 0·2||7·3 ± 0·5a|
|LD50 to zebrafish (CFU g−1)||(2·20 ± 0·04) × 104||(2·00 ± 0·03) × 105||(2·14 ± 0·04) × 104|
No distinguishable growth difference in M9 medium was observed among Edw. tarda H1, Δinv and inv+ (data not show). Compared with the wild-type H1, haemolytic activity of inv+ increased for about 48·7% (P < 0·01), while the Δinv decreased for about 10·4% (P < 0·05) (Fig. 1).
As shown in Fig. 2, the OD570 of Edw. tarda H1, Δinv and inv+ was 0·435 ± 0·04, 0·304 ± 0·03 and 0·762 ± 0·09, respectively. Biofilm formation of inv+ increased for 75·2% compared with the wild-type H1 (P < 0·01). On the other hand, Δinv was attenuated in biofilm production, producing 30·1% less than that of H1 (P < 0·05), indicating that the biofilm formation is positively associated with invasin in Edw. tarda.
Adherence and internalization to epithelial cells are important processes in bacterial infection (Meyer et al. 1997). There is no significant difference between the wild-type strain and the mutants in adherence to EPC cells. Compared with the wild-type strain, Δinv showed similar internalization rates, while inv+ exhibited a three-fold increase (P < 0·01), underlining the significance of invasin overexpression in internalization to epithelial cells (Table 3).
The LD50 of H1, Δinv and inv+ strains in the zebrafish model was 2·20 × 104, 2·00 × 105 and 2·14 × 104 CFU g−1, respectively. The Δinv mutant exhibited a 10-fold decrease in virulence compared with the wild-type strain, while the virulence was completely restored in inv+ (Table 3).
Invasin is a protein that allows enteric bacteria to penetrate cultured epithelial cells. Investigation the role of invasin might help to elucidate the mechanism of bacterial invasion to host cells. An invasin in-frame deletion mutant (Δinv) and a complementary strain (inv+) of Edw. tarda were constructed in the current study. The qRT-PCR analysis showed that the invasin was excessively expressed in inv+ without any extrinsic inducer. This indicates the expression level of invasin might be controlled tightly by complex and multifactorial regulatory mechanisms from the genome (Liu et al. 2012).
Haemolysin was reported to be an important virulence factor in the pathogenesis of Edw. tarda. Two types of haemolysins, an extracellular hole-forming haemolysin (Chen et al. 1996) and cell associated (Watson and White 1979; Janda and Abbott 1993b), were identified. When Edw. tarda H1 was inoculated onto the plates containing 5% sheep erythrocytes, no obvious haemolytic activity was observed, indicating that Edw. tarda H1 may only possess a weak haemolytic activity (data not shown). Nevertheless, inv+ exhibited a significant increase in haemolytic activity compared with the wild type. Therefore, haemolysins of Edw. tarda might be positively regulated by invasin.
Biofilms are aggregates of micro-organisms that mediate attachment to a surface. It was considered important in the pathogenesis of pathogen, as it could be used as an advantage to resist environmental challenges and avoid immune, cellular and chemical systems of host defence (Wakimoto et al. 2004). In the biofilm formation assay, Δinv produced thinner biofilm than the wild type, while the biofilm of inv+ was thicker, suggesting that invasin is a positive contributor in biofilm formation of Edw. tarda. It might function directly by enhancing biofilm production or indirectly by coordinating with other regulators of the bacteria such as the quorum-sensing system (Matthew and Greenberg 2005), which needs to be further investigated.
Adherence to and invasion of host cells are often the initial steps of uptake in many bacterial pathogens (Finlay and Falkow 1988; Finlay and Cossart 1997). The invasin overexpression strain inv+ exhibited increased adherence and especially internalization to EPC cells, indicating that the invasin of Edw. tarda may directly facilitate bacterial uptake by host cells.
In our results, inv+ displayed significantly increase in biofilm formation and internalization to EPC cells; however, inv+ did not display increased virulence to zebrafish. Different from cultured cells, live fish face a complex environment. Overexpression of one gene may not be sufficient to generate significant difference in virulence towards fish. After Edw. tarda was internalized, invasin may not provide further advantages for the intracellular replication and dissemination of bacterial pathogen.
The Edw. tarda invasin contains an Ig-like domain, which is often found in bacterial surface proteins such as intimins and invasins that are involved in pathogenicity (Pell et al. 2010). Furthermore, its predicted gene product shares 54% of similarity with invasin from Y. pseudotuberculosis IP 32953 (Genbank ID: YP_ 070195.1). It was reported that the Y. pseudotuberculosis invasin promotes bacterial entry by binding to members of the β1 integrin family of the host cell and activating reorganization of the host cytoskeleton to form pseudopods that envelop the bacteria (Hamburger et al. 1999). Meanwhile, Edw. tarda strains were able to adhere and enter fish cells using internalization mechanisms involving host microfilaments and protein tyrosine kinase (Ling et al. 2000). The invasin investigated in the current study shares similarity with the membrane region of the Y. pseudotuberculosis invasin, but lacks the integrin-binding region, which might explain the comparable levels of internalization rates in Δinv and the wild-type strain. Despite of the lack of integrin-binding region, overexpression of invasin in Edw. tarda facilitated the internalization into fish cells. It is suggested that Edw. tarda uses an invasion mechanism that is distinct from Y. pseudotubercuiosis, and the invasin binding to the integrins of host cells may not be essential for the uptake of bacteria. Nevertheless, multiple factors are required for bacterial pathogens to bind and enter eukaryotic cells to establish infection (Isberg and Leong 1990). Bacteria–host interactions should thus be further investigated to identify the virulence factors involved, which could contribute to our understanding of the pathogenesis of Edw. tarda infections.
In conclusion, the current study demonstrated that invasin is important for haemolytic activity, biofilm formation and virulence of Edw. tarda. Although it may not be the only contributor in invasion of host cells, the invasin of Edw. tarda plays a significant role in promoting virulence during infection.
This work was supported by grants from the International Science and Technology Cooperation Programme of China (No. 2012DFG31990) and the National High Technology Research Development Program of China (863 Programs, No. 2008AA092501).