The hypothesis that dispersin B (DspB), an enzyme from the periodontal pathogen Aggregatibacter actinomycetemcomitans that degrades the extracellular matrix polysaccharide PGA, will inhibit biofilm formation of the soft rot pathogen Pectobacterium carotovorum subsp. carotovorum in infected plants was tested by constitutive expression of DspB in tobacco plants. All the transgenic plants expressed properly folded and active DspB enzyme, although at different expression levels. In virulence assays, even the transgenic plant line D10, which produced a low level of DspB compared to other lines, showed significant resistance against P. carotovorum subsp. carotovorum, suggesting that DspB could be a valuable agent for biological control of P. carotovorum subsp. carotovorum infection in crop plants.
Infectious plant diseases continue to cause human suffering and enormous economic losses. One-third of the total crop losses in the world can be attributed directly to plant diseases (Agrios, 1988). Although pesticides have successfully controlled diseases, their continued and increasing use pose a risk to human health and the environment. Alternatively, application of the plant’s own defence mechanisms, combined with understanding of the complex ecology of real-world disease processes, has the potential for more effective protection against plant pathogens (Baker et al., 1997). In this regard, control of soft rot Pectobacterium carotovorum subspecies such as P. carotovorum subsp. carotovorum (Pcc) are economically important, because they cause serious damage worldwide on a wide variety of plants (Pérombelon & Kelman, 1980; Pérombelon, 2002). For instance, reduction in the yield of potato from this pathogen can be as high as 24% of worldwide production (Sawyer, 1984). Soft rot erwinias cause disease mainly by secretion of cell-wall-degrading enzymes such as pectate lyase (Pel), polygalacturonase (Peh), cellulase (Cel) and protease (Prt) (Barras et al., 1994). In addition, secondary factors such as exopolysaccharides (EPS) (Condemine et al., 1999), lipopolysaccharides (LPS) (Schoonejans et al., 1987), iron assimilation (Enard et al., 1988), the Hrp system (Bauer et al., 1994) and motility (Mulholland et al., 1993) also contribute to virulence in soft-rot Pectobacterium spp.
Polysaccharide is a major component of the EPS matrix in most bacterial biofilms (Sutherland, 2001). For plant pathogens, EPS capsules and slime are important for protection against recognition by plant defence mechanisms. They are used to bind water, keeping the bacteria moist, and for retention of nutrients and ions released from the damaged plant cells (Leigh & Coplin, 1992). Several investigators have reported the role of bacterial polysaccharides in pathogenesis (Coplin & Cook, 1990; Katzen et al., 1998; Dharmapuri & Sonti, 1999). Loss of EPS production has been correlated with loss of virulence in plant pathogens such as Xanthomonas oryzae (Kumar et al., 2003), Ralstonia solanacearum (Saile et al., 1997), Erwinia pyrifoliae (Kim et al., 2002) and Xylella fastidiosa (Rodrigues et al., 2008). One of the best-characterized matrix polysaccharides is a hexosamine-containing polymer produced by several staphylococcal species, including Staphylococcus epidermidis and Staphylococcus aureus, and by Escherichia coli. This polymer, called PIA (also PNAG, PS/A or SAA) in Staphylococcus spp. (Baldassarri et al., 1996; Mack et al., 1996; Maira-Litran et al., 2004) and PGA in E. coli (Vuong et al., 2004), consists of a linear chain of N-acetyl-d-glucosamine (GlcNAc) residues in β(1,6) linkage.
A glycosyl hydrolase that degrades S. epidermidis PIA was recently identified (Kaplan et al., 2004b). This enzyme, named dispersin B (DspB), is produced by the Gram-negative oral bacterium Aggregatibacter actinomycetemcomitans, the causative agent of a severe form of periodontal disease that affects adolescents (Zambon, 1985). It was previously shown that purified DspB not only rapidly and efficiently removed S. epidermidis biofilms, but also prevented biofilm formation on precoated surfaces (Kaplan et al., 2004a; Manuel et al., 2007). Since the analysis of the genome of the Gram-negative plant pathogen P. carotovorum revealed the presence of an operon orthologous to the operon that encodes for PIA in S. epidermidis, it is likely that P. carotovorum also produces a PIA-like substance (Fig. 1). It was therefore hypothesized that the production of DspB within a plant will protect it from Pcc infection. Also, since DspB exhibits optimum activity at pH 5·6, production of DspB by plants will be protective since there is a reduction in pH at the damaged site. To test this hypothesis, tobacco plants were genetically manipulated to express DspB using Agrobacterium-mediated transformation and the ability of DspB to provide resistance against Pcc was analysed.
Materials and methods
Bacterial strains and reagents
Staphylococcus epidermidis [NJ 9709 (Kaplan et al., 2004a)] strains were grown in Trypticase soy broth (Becton Dickinson) supplemented with 6 g yeast extract and 8 g glucose L−1 (Kaplan et al., 2003). Pectobacterium carotovorum subsp. carotovorum (ATCC 15713) was cultured overnight at 28°C in LB medium. Inoculum was prepared by centrifugation (5000 g for 10 min) and diluted to 108 CFU mL−1 with fresh medium.
Binary vector constructs for functional expression of the DspB gene in tobacco
The DspB coding region was amplified from pRC3 (Ramasubbu et al., 2005) by PCR using Pfu Turbo DNA polymerase (Stratagene). The forward primer (5′-CACCATGAATTGTTGCGTAAAAGGCAATTC-3′) was used to create a CACC site (bold) adjacent to the DspB initiation codon in the open reading frame, and the reverse primer (5′-TTAGTGGTGGTGGTGGTGGTGCTCATCCCCATTCGTCTTATG-3′) was designed to include a his-tag and stop codon. The PCR product was directionally cloned into the pENTR TOPO vector (Invitrogen). The DspB construct was sequenced to confirm that the coding region was in frame. LR clonase reactions to transfer DNA fragments from entry clones to Gateway destination vectors were carried out according to the instructions provided by the manufacturer (Invitrogen). The pMDC32 Gateway destination vector was obtained from the ABRC Stock Center (The Ohio State University) and used for the expression of DspB protein.
The DspB/pMDC32 construct was transformed into Agrobacterium tumefaciens LBA4404 cells (Invitrogen) by electroporation, and selected on YM plates containing 50 μg kanamycin mL−1. The tobacco seeds (cv. Havana 38) were obtained from Lehle Seeds. Healthy, fully expanded leaves from 4- to 5-week-old tissue-culture-grown tobacco plants were used for transformation. Leaf discs were co-cultured with A. tumefaciens LBA4404 carrying DspB/pMDC32 over a period of 48 h in MS medium (Murashige & Skoog, 1962). Explants were subcultured in regeneration medium (MS salts and vitamins, 30 g sucrose, 1 mg each of thiamine-HCl and BAP, 0·1 mg NAA, 2·5 g phytagel L−1, pH 5·8, plus 50 mg hygromycin and 250 mg cefotaxime L−1) and the callus was transferred to fresh medium until distinct shoots appeared. Shoots were grown in micropropagation medium (MS salts, 30 g sucrose and 2·5 g phytagel L−1, pH 5·8) supplemented with hygromycin and cefotaxime. Plants from different transgenic lines were maintained under standard conditions (16 h light/8 h dark cycle at 20–22°C).
Leaf extracts were prepared from fresh tobacco leaf tissue (500 mg) of different transgenic plants. Briefly, leaf tissues were ground in liquid nitrogen and immediately transferred to 1·5 mL protein extraction buffer (20 mm HEPES pH 7·4, 1 mm MgSO4, 1 mm CaCl2 and 1% Triton X-100). After vortexing for 30 s, samples were centrifuged at 14 000 g for 10 min at 4°C. The supernatant was collected and glycerol was added to a final concentration of 10% and kept at −80°C. Staphylococcus epidermidis biofilm was prepared as described previously (Kaplan et al., 2004a). Briefly, two loops of colonies scraped from the surface of an agar plate were transferred to a microfuge tube containing 200 μL fresh medium. Cells were homogenized with a disposable pellet Kontes pestle (Fischer) and vortex agitated at high speed for 30 s. Twenty-two microlitres were transferred to a 50-mL polystyrene tube containing 22 mL fresh medium. The resulting inoculum contained 107–108 CFU mL−1. Biofilms were grown in 96-well tissue-culture-treated polystyrene microtitre plates (Corning model 3595, Sigma-Aldrich). Wells filled with inoculum (200 μL) were incubated for 16 h at 37°C. For detachment studies, wells were first rinsed by submerging the entire plate in a tub of cold, running tap water, after which 50 μL leaf extract was added to each well along with 150 μL PBS buffer. Plates were incubated at 37°C for different time periods. After decanting the leaf extract, the biofilms were washed in cold, running tap water and were stained with crystal violet as previously described (Kaplan & Fine, 2002). The absorbance values of the well solutions were determined by using a Bio-Rad Benchmark microplate reader set at 590 nm. All assays were performed in triplicate wells on at least three separate occasions, which exhibited similar results with minimal variation among them.
Twenty-two microlitres of S. epidermidis biofilms were transferred to 16·5 mL fresh medium in a 50-mL polystyrene tube. Wells were filled with 150 μL S. epidermidis inoculum with 50 μL leaf extracts from control and transgenic plants to obtain 107–108 CFU mL−1. Biofilms were grown in 96-well tissue-culture-treated polystyrene microtitre plates. The plates were incubated for 20 h at 37°C after which the plates were washed in cold running water and stained with crystal violet as above. For experiments using purified DspB as a positive control, wells filled with inoculum (107–108 CFU mL−1) (200 μL containing varying concentrations of DspB) were incubated for 20 h at 37°C.
Detachment and inhibition assay for Pcc
Cells were cultured overnight at 28°C in LB. Cells were centrifuged (5000 g for 10 min) and diluted to density of 1 × 108 CFU mL−1 with fresh medium. Biofilms were grown in 96-well tissue-culture-treated polystyrene microtitre plates. Wells were filled with 200 μL inoculum with different concentrations of native and boiled DspB and incubated for 24 h at 28°C. Biofilms were washed with cold running water and residual biofilms were stained with crystal violet as described above. The absorbance values of the well solutions were determined at 590 nm. For detachment studies, wells filled with inoculum (200 μL) were incubated for 24 h at 28°C. Wells were first rinsed with PBS (3×) and treated with different concentrations of native and boiled DspB for 1 h at 37°C. Biofilms were stained with crystal violet and absorbance measured at 590 nm.
Total protein was extracted from leaves using a plant total protein extraction kit (Sigma) according to the manufacturer’s instructions. Protein concentration was estimated using Bio-Rad protein assay reagent. Total protein (10 μg) was resolved on SDS-PAGE and transferred to a nitrocellulose membrane. Immunoblot was performed as per the standard protocol. Briefly, membranes blocked with 5% milk in Tris-buffered saline (TBS) were incubated with purified polyclonal antibodies raised against DspB. After washing, the membranes were incubated with HRP-linked secondary antibodies. Membranes were developed with chemiluminescent substrate (Pierce, Thermo Scientific).
Extraction and detection of polysaccharide in Pcc
Polysaccharide was purified from the bacterium as described by Frank & Patel (2007). Briefly, bacteria were grown overnight in Petri dishes containing LB medium at 28°C. Cells were harvested and washed twice in PBS, resuspended in 0·5 m EDTA, sonicated for 5 min at 40 kHz in a bath sonicator, boiled for 5 min and centrifuged. Supernatant was treated with 200 μg proteinase K for 30 min at 65°C and then the enzyme was heat-inactivated for 15 min at 80°C. Extracts were spotted on PVDF (Bio-Rad) membranes using dot-blot apparatus (Bio-Dot, Bio-Rad), and the blots were blocked with 3% bovine serum albumin (BSA) in TBS overnight. The blots were first probed with goat anti-deacetylated PNAG antibody (1:5000) in TBS-0·1% Tween 20 containing 3% BSA for 90 min at 4°C. After washing, the blots were further probed with HRP-linked rabbit anti-goat antibody (1:5000) in 0·1% TBS Tween 20 containing 3% BSA for 90 min and developed with chemiluminescent substrate (as above). Several controls were tested in this assay, including DspB-treated polysaccharide, N-acetylglucosamine, a tetramer of β(1,6)linked N-acetylglucosamine and a tetramer of β(1,4)linked N-acetylglucosamine.
Bacterial growth and inoculation of plants
Pcc (ATCC 15713) was cultured overnight at 28°C in LB. Inoculum was prepared by centrifugation (5000 g for 10 min) and diluted to different population densities in fresh medium. Leaves were excised from both control and transgenic lines of the same age (6–8-week-old plants) and placed in Petri dishes containing two layers of Whatman filter paper (3 mm) moistened with 4 mL sterile water. The leaves were punctured using a 20-μL pipette tip and inoculated at the site of puncture by adding 3 μL Pcc (5 × 103, 1 × 104 or 1 × 105 per inoculation site), as these levels are likely to be associated with early and late infection (Pérombelon, 2002). The infected leaves were incubated at 28°C for 24 or 48 h and evaluated for the presence of maceration. Bacteria from the infected sites from the transgenic lines were reisolated by excision of the site and transferred to 50 μL sterile water and vortexed. Five microlitres of this were used for reinoculation of control leaves to test their viability and ability to infect new leaves.
Sequence analysis of the products of the PGA-encoding genes in P. carotovorum subsp. aseptica (Pca) was performed (Bell et al., 2004), and homology was determined by conducting a blastp search at the NCBI website. Conserved protein domains and predicted membrane localization domains were obtained from the Entrez Protein Database at the NCBI website.
Sequence analysis of EPS-producing genes in P. carotovorum– the pcaABCD locus
A cluster of four genes in P. carotovorum subsp. aseptica (Pca), ECA4451, ECA4452, ECA4453 and ECA4454, displayed a high degree of sequence similarity to the corresponding E. coli genes pgaABCD and Yersinia pestis genes hmsHFRS, and, to a lesser extent, to gene products of the icaADBC loci of staphylococcal species and the pgaABCD loci of A. actinomycetemcomitans (Fig. 1a). The proteins encoded by these loci are involved in the synthesis of poly-β(1,6)-GlcNAc, the major component of the biofilm matrix.
The gene ECA4454 was predicted to encode an 820-amino-acid outer membrane protein and had 42% amino acid identity to Y. pestis HmsH, whereas ECA4453 was predicted to encode a 671-amino-acid protein with 57% identity to Y. pestis HmsF. Interestingly, S. epidermidis IcaB was shown to be a surface-bound protein that deacetylates PIA (Vuong et al., 2004). Among the four encoded proteins, ECA4452, with 442 amino acids, had the highest identity to Y. pestis HmsR at 73% (Fig. 1b). This protein is a member of the glycosyltransferase family 2 enzymes. The smallest of the four proteins, ECA4451, with 140 residues, had only 25% identity to HmsS of Y. pestis. Exopolysaccharide analysis of P. carotovorum subsp. carotovorum (ATCC 15713) suggested that this bacterium also synthesizes a PNAG/PGA/PIA-like material (Fig. 2a). Therefore, it is likely that the individual proteins encoded by the loci might have similar functions that have been proposed for the corresponding proteins of the PGA-producing bacteria. More importantly, the presence of the pga locus renders the bacterium susceptible to biofilm detachment by the enzyme DspB.
Inhibition of biofilm formation of Pcc by DspB
The question of whether DspB inhibits the formation of biofilm by Pcc was tested using an in vitro microtitre plate assay (Kaplan et al., 2004a; Manuel et al., 2007). As evident in Figure 2b, DspB clearly inhibited the Pcc biofilm formation in a dose-dependent manner, suggesting that expression of DspB in vivo would be a useful method to prevent infection by Pcc. Interestingly, DspB did not detach Pcc biofilms in microtitre plates (Fig. 2c), reminiscent of the effect of DspB on S. aureus as well as A. actinomycetemcomitans biofilms. The biofilms of these bacteria are more complex and contain DNA as well as proteins. A previous study showed that DspB by itself does not detach preformed biofilms from these bacteria (Izano et al., 2008).
Expression of DspB in transgenic tobacco
To achieve production of DspB in plants, the dspB gene of A. actinomycetemcomitans was first cloned under the control of Cauliflower mosaic virus promoter 2×35S, which is known to produce high-level expression of foreign genes in plants (Curtis & Grossniklaus, 2003). The dspB gene was introduced in tobacco (Nicotiana tabacum cv. Havana 38) by standard Agrobacterium-mediated transformation and selection through hygromycin resistance. Among 33 transformants generated, 22 plants were positive for the expression of recombinant DspB protein as determined from the results of western blots (data not shown). From the 22 positive lines, five lines showing a range of DspB expression levels (lowest to highest levels) were selected for further studies. Among these five lines, the expression levels were in the order D10 < D17 < D12 < D21 < D23 (Fig. 3). All transformants were judged to be phenotypically normal as there was no morphological variation in growth and development compared to the wildtype.
Functional analysis of DspB
The detachment assay using S. epidermidis showed that DspB-expressing transgenic tobacco plants produced substantial amounts of the biologically active enzyme although at varying levels. Leaf extract from transgenic plant D10 detached about 60% of biofilm after 45 min of incubation, whereas extract from vector-transformed control plants resulted in no detachment (Fig. 4a). Among the transgenic lines, extract from plant D12 showed the maximum detachment of biofilm (82%) (Fig. 4b) after 45 min. Interestingly, the extracts from the transgenic plants not only detached the preformed S. epidermidis biofilm but also inhibited its biofilm formation, suggesting that the recombinant enzyme in the leaves is biologically active (Fig. 5a). While most of the transgenic plants showed at least 70% inhibition of biofilm formation, line D10 exhibited only 33% inhibition (Fig. 5b). The biofilm detachment and inhibition collectively show that the expressed DspB in these plants is biologically active. The variation of DspB expression between different transgenic lines could be the result of differences in transgene copy number and/or position and length of the transgene integration event.
Bioassay for antibacterial effect of transgenic tobacco
Detached leaves from 6- to 8-week-old plants of the wildtype and the five selected transgenic tobacco lines were assayed for resistance against infection by Pcc. Upon infection, control leaves of the wildtype exhibited maceration, which was visible around the inoculated area after 24 h, and the whole leaf tissues were macerated within 48 h (Fig. 6a). On the other hand, transgenic lines showed no visible maceration and the leaves remained green and healthy even after 48 h with high inoculum density (Fig. 6a). Interestingly, the transgenic line D10 that produced a lower level of biologically active enzyme than the other four lines was also resistant to infection at high inoculum density. Pcc, isolated from the inoculated area of transgenic lines after 48 h and used for reinfection of wildtype plants was found to cause tissue maceration within 24 h, suggesting that the reisolated bacteria were viable (Fig. 6b).
The efficient production of food supplies needed to meet the nutritional requirements of a rapidly growing world population is the major and most important goal of modern-day agriculture (Osusky et al., 2005) since worldwide crop losses attributable to microbial diseases are currently around $720B annually. These losses are particularly untenable considering the predicted increase in human populations, which will eventually be accompanied by a decline in the available cultivable land and irrigation (Borlaug, 2000). Pectobacterium carotovorum has been reported to cause disease in over half of angiosperm families (Ma et al., 2007). Losses in potato and other crops as a result of P. carotovorum infection are substantial and control of infection by this bacterium will be a welcome strategy in increasing worldwide crop production. This bacterium is very versatile, as there are even reports of it promoting the growth of human pathogens on fruits and vegetables (Brandl, 2008).
Pectobacterium carotovorum causes tissue maceration by producing plant cell-wall-degrading enzymes which are under the control of quorum sensing (QS) (Brader et al., 2005). Mutant P. carotovorum that are defective in the release of the autoinducer N-acylhomoserine lactone, a signal for quorum sensing, were shown to be avirulent (Dong et al., 2000). Inoculation of P. carotovorum expressing AiiA, an N-acylhomoserine lactone degradase, failed to cause soft rot disease in detached tissues of potato, aubergine, Chinese cabbage, carrot, celery, cauliflower and tobacco (Dong et al., 2000). Furthermore, transgenic potato and tobacco plants expressing AiiA exhibited retarded or completely absent infection by P. carotovorum (Dong et al., 2001). Interestingly, expression of N-acylhomoserine lactones by transgenic plants also resulted in enhanced resistance to infection by wildtype P. carotovorum (Mae et al., 2001). Methods directed against QS to control cell density or to kill bacteria are prone to the development of bacterial resistance as a result of QS (Defoirdt et al., 2010). In addition, it has been reported that potato plants genetically modified to produce N-acylhomoserine lactones, messengers for controlling cell density, increase the susceptibility of plants to P. carotovorum infection (Toth et al., 2004). Another strategy to control infection by P. carotovorum (During et al., 1993) and to control fire blight (Ko, 1999) is the expression of bacteriophage T4 lysozyme in transgenic plants. Studies to control the production of biofilm EPS matrix have been initiated as an alternative method to combat infections by P. carotovorum. This method is attractive since it avoids resistance development resulting from killing the bacteria.
This study has demonstrated that expression of DspB in transgenic tobacco plants protects tobacco against local infections by the pathogenic bacterium Pcc. The presence of a locus similar to the PGA-encoding genes present in E. coli, Y. pestis and S. epidermidis suggests that the potential bacterial biofilm formation is inhibited in the transgenic plants. The role of biofilm formation in disease pathogenesis is well known (Leigh & Coplin, 1992). Thus, introducing DspB in to plants is a novel method of attenuating biofilm formation and rendering these transgenic plants resistant to Pcc. Interestingly, DspB-producing transgenic lines showed 100% resistance to Pcc infection, including the transgenic line D10, which only produced low levels of DspB expression. This suggests that even lower levels of DspB may be sufficient to destroy the production of EPS and confer resistance to high inoculum densities of pathogenic bacteria (Fig. 6a). Furthermore, although bacteria in the inoculated zone did not show any disease symptoms such as tissue maceration, bacteria isolated from the site of infection were still viable nonetheless (Fig. 6b). Expression of DspB did not lead to abnormal phenotypes among the transformants. The inhibition effect of DspB on the biofilm formation of Pcc (Fig. 2b) suggests that this bacterium is being controlled (its growth prevented) at the site of infection. Somewhat similar results were reported when a viral EPS-depolymerase gene was expressed in transgenic pears, targeting Erwinia amylovora (Malnoy et al., 2005); however, there was limited expression of the depolymerase and only two of the plants expressed the enzyme constitutively.
It is also worth mentioning that although DspB was expressed as a cytosolic protein, the plants displayed significant protection against Pcc infection. While it might be desirable to express DspB as a secreted protein or even as a surface-anchored protein, the results indicate that such expression is unnecessary. It was also observed that topical application of DspB did not lead to resistance against Pcc (data not shown). Previously, antiviral single-chain antibody fragment FV, displayed on the membrane, was shown to confer resistance against Tobacco mosaicvirus (Schillberg et al., 1999). In conclusion, it has been shown that the EPS-degrading enzyme DspB from the oral bacterium A. actinomycetemcomitans can confer resistance to pathogens such as Pcc in plants by targeting a virulence factor, PGA.
This work was supported by USPHS Grants DE16591 (NR) and AI82392 (JBK). We acknowledge Dr Gerald Pier, Harvard Medical School, Boston, MA, USA, for providing the anti-deacetylated PNAG antiserum.
CR and MS equally contributed to the paper; CR, MS, JBK and NR designed the work; CR, MS and MB performed research; CR and NR wrote the paper.