Xylella fastidiosa Afimbrial Adhesins Mediate Cell Transmission to ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 2009, p. 521–528 0099-2240/09/$08.00⫹0 doi:10.1128/AEM.01921-08 Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Vol. 75, No. 2

Xylella fastidiosa Afimbrial Adhesins Mediate Cell Transmission to Plants by Leafhopper Vectors䌤 Nabil Killiny and Rodrigo P. P. Almeida* Department of Environmental Science, Policy, and Management, University of California, Berkeley, California 94720 Received 18 August 2008/Accepted 11 November 2008

The interactions between the economically important plant-pathogenic bacterium Xylella fastidiosa and its leafhopper vectors are poorly characterized. We used different approaches to determine how X. fastidiosa cells interact with the cuticular surface of the foreguts of vectors. We demonstrate that X. fastidiosa binds to different polysaccharides with various affinities and that these interactions are mediated by cell surface carbohydrate-binding proteins. In addition, competition assays showed that N-acetylglucosamine inhibits bacterial adhesion to vector foregut extracts and intact wings, demonstrating that attachment to leafhopper surfaces is affected in the presence of specific polysaccharides. In vitro experiments with several X. fastidiosa knockout mutants indicated that hemagglutinin-like proteins are associated with cell adhesion to polysaccharides. These results were confirmed with biological experiments in which hemagglutinin-like protein mutants were transmitted to plants by vectors at lower rates than that of the wild type. Furthermore, although these mutants were defective in adhesion to the cuticle of vectors, their growth rate once attached to leafhoppers was similar to that of the wild type, suggesting that these proteins are important for initial adhesion of X. fastidiosa to leafhoppers. We propose that X. fastidiosa colonization of leafhopper vectors is a complex, stepwise process similar to the formation of biofilms on surfaces. layer and most likely are secreted through wax canals (27). In some insects, this wax layer is covered by a cement layer (mucopolysaccharides) formed from secretion of dermal glands (4). X. fastidiosa cells have been shown to colonize specific areas of the foreguts of insects, where they multiply and form a carpet-like biofilm (39). Cells seem to initially attach laterally to the cuticle of insects (2), but in fully colonized insects, X. fastidiosa is always found polarly attached, presumably because a larger cell surface area is exposed to the very dilute sap nutrients, passing through the foregut at 5 to 50 cm/s, being ingested by the insects. This turbulent environment is expected to cause occasional detachment of cells prior to the formation of mature biofilms within vectors (see reference 3 for a discussion of this topic). The interaction of X. fastidiosa with the foregut cuticle differs from those of other xylem-limited bacteria, such as Leifsonia xyli, which can be acquired from plants but are not transmitted by insects (5). Only two studies with X. fastidiosa knockout mutants have addressed aspects of vector transmission (11, 34). However, both studies focused on X. fastidiosa’s cell-cell signaling system, which regulates cascades of genes and pathways, thus allowing the identification of target genes but not identifying specific interactions between vector and pathogen. The rpfF gene (regulation of pathogenicity factors F) encodes an enzyme that synthesizes the signaling molecule diffusible signaling factor (DSF), whereas rpfC is part of a hybrid two-component DSF sensor (11). An rpfF mutant is not transmissible by insects because it does not colonize the foregut of vectors (34), while an rpfC mutant colonizes the insect’s foregut but is transmitted at lower rates than that of the wild type (11). In vitro adhesion assays indicated that the rpfF mutant did not form biofilms, while the rpfC mutant adhered to surfaces more strongly than the wild type did. Targeted gene expression analyses of X. fastidiosa adhesins indicated that hemagglutinin-like proteins (Hxf afimbrial adhesins) and type

Vector-borne diseases have reemerged as a major threat to society in recent decades (8, 21). However, despite the importance of arthropod vectors in these disease systems, substantially more research efforts are directed toward understanding the molecular mechanisms of host-pathogen interactions rather than those occurring between vectors and pathogens. Thus, the disruption of interactions between pathogen and vector represents a potentially large untapped source of disease control alternatives (6, 17, 25). For most systems, however, we lack basic knowledge on how microbes colonize arthropod vectors. The fastidious bacterium Xylella fastidiosa is a xylem (waterconducting) tissue colonizer that causes diseases in many hosts of economic importance, including grape, citrus, coffee, and almond plants (24). Spread of the pathogen occurs by means of xylem sap-feeding leafhoppers (Insecta, Hemiptera, Cicadellidae) (37, 42, 43). Among xylem sap feeders, there is no evidence of vector specificity, but transmission efficiency may vary (18, 40). Unlike many other insect-borne bacterial plant pathogens, which colonize internal tissues of their vectors, X. fastidiosa colonizes the leafhopper’s foregut cuticular lining (i.e., the surface of the cuticle) (39). Furthermore, the loss of vector infectivity after molting and the lack of a latent period (time between pathogen acquisition and inoculation) strongly suggest that the foregut is the site from which X. fastidiosa is transmitted (1, 38). Although the chemical composition of the outermost layer of the leafhopper cuticle has not been studied in detail, it has been described for other insects. The cuticle is composed of proteins, chitin, other polysaccharides, and lipids (4). Lipids cover the cuticle as a wax

* Corresponding author. Mailing address: Department of Environmental Science, Policy, and Management, University of California, Berkeley, CA 94720. Phone: (510) 642-4197. Fax: (510) 643-5438. E-mail: [email protected]. 䌤 Published ahead of print on 14 November 2008. 521

522

APPL. ENVIRON. MICROBIOL.

KILLINY AND ALMEIDA TABLE 1. Xylella fastidiosa Temecula1 strains used in this study Strain

Notes

Source or reference(s)

Temecula1 gfp mutant (KLN59.3) rpfF mutant (KLN61) rpfC mutant rpfF rpfC mutant fimA mutant (6E11) pilB mutant (1A2) fimA pilO mutant (DM12) hxfA mutant hxfB mutant gumH mutant gumD mutant

Wild type GFP-labeled strain Required for synthesis of DSF (PD0407) Two-component sensor kinase (PD0406) Double mutant (PD0406/PD0407) Required for type I pilus (short pilus) synthesis (PD0062) Required for type IV pilus (long pilus) synthesis (PD1027) Required for type I (PD0062) and type IV (PD1693) pilus synthesis Hemagglutinin-like secreted protein (afimbrial adhesin) (PD1792) Hemagglutinin-like secreted protein (afimbrial adhesin) (PD2118) Required for EPS synthesis (PD1391) Required for EPS synthesis (PD1394)

33 33 34 10, 11 10, 11 30 30 26 22 22 D. Cooksey D. Cooksey

I pili (fimbrial adhesin) were associated with adhesion of these knockout strains to glass surfaces, but type IV pili were not (11). Thus, indirect evidence allowed us to hypothesize that some adhesins are important for X. fastidiosa attachment to and colonization of vectors and subsequent inoculation into susceptible hosts, while other adhesins have little or no role in this process. In this study, we sought to determine the nature of X. fastidiosa-vector interactions by using biochemical, molecular, and biological assays. MATERIALS AND METHODS Bacterial strains and insects. We used the X. fastidiosa isolate Temecula1 for all experiments. This isolate belongs to the grape genetic group of X. fastidiosa and has been fully sequenced (47). Specific gene knockout mutants used here were previously produced and characterized by other groups (Table 1). All cells were grown in vitro at 28°C on modified periwinkle wilt solid medium (23), with or without kanamycin at 5 ␮g/ml. We collected cells from plates 7 days after inoculation for biochemical and biological assays. Laboratory-reared, X. fastidiosa-free leafhopper vectors (Graphocephala atropunctata; Hemiptera: Cicadellidae) were kindly provided by David Morgan (California Department of Food and Agriculture) to conduct biochemical experiments. Cell adhesion assays. To coat one-square-centimeter pieces of nitrocellulose membrane (NCM), we incubated solutions with 1 ml of phosphate-buffered saline (PBS [pH 7.4]; 2 mM K2PO4, 8 mM Na2HPO4, 0.14 M NaCl, 2 mM KCl) containing 1% (wt/vol) of different polysaccharides under slow shaking conditions for 1 h at room temperature. Polysaccharides tested included chitosan, chitin, carboxymethyl cellulose, cellulose, and dextran sulfate (Sigma-Aldrich). Afterwards, we performed three 15-min washes with PBS plus 1% Tween 20 (PBS-Tween). We then incubated the NCM pieces for 2 h in PBS with an EDTA-free protease inhibitor cocktail (one tablet per 10 ml) (Sigma-Aldrich) containing X. fastidiosa cells collected from periwinkle wilt solid medium with a final suspension optical density at 600 nm (OD600) of 0.4. We used buffer containing wheat germ agglutinin (WGA) (0.1 mg/ml) or bovine serum albumin (BSA) (0.1 mg/ml) as a control. After three washings with PBS-Tween, proteins bound to the polysaccharide-coated NCM were detected by either Ponceau S (0.1% [wt/vol] Ponceau S in 1% acetic acid) or amido black (0.1% [wt/vol] amido black in 10% acetic acid and 25% isopropanol). In a follow-up experiment, we tested the role of extracellular polymeric substances (EPS) and surface proteins, which may be bound loosely to X. fastidiosa cells grown in vitro, in cell attachment. We used four treatments for this assay, as follows: (i) X. fastidiosa cells were suspended in PBS; (ii) cells were washed twice in PBS-Tween and centrifuged at low speed at 4°C, and the pellets were suspended in PBS; and cells were treated with either (iii) proteinase K (0.1 mg/ml in PBS) for 1 h at 56°C or (iv) pronase E (1 mg/ml in PBS) for 1 h at 37°C and then washed twice. The proteinase-treated cells were then incubated with the polysaccharide-coated NCM pieces as described above. Specificity of cell attachment to polysaccharides. We used different approaches to determine the specificity of X. fastidiosa attachment to polysaccharides. G. atropunctata leafhopper foregut cuticle was obtained by extracting tissue covered by the clypellus region of the head, which included the cibarium and precibarium areas of the foregut. We did not conduct further dissection, and thus extracts included the external and internal cuticles as well as other tissue present

in that region of the head of leafhoppers. The tissue was washed in PBS buffer after dissection. PBS buffer containing 1% (wt/vol) foregut tissue was subjected to five cycles of 2 min of sonication and 2 min on ice. We placed 2-␮l drops of extracts obtained from the foreguts of leafhopper vectors on NCM strips. After being blocked with 6% nonfat milk (Sigma-Aldrich) in PBS, strips were incubated in PBS with a protease inhibitor (Sigma-Aldrich) containing X. fastidiosa cells with an OD600 of 0.4 in the absence or presence of sugars for 2 h. We performed an affinity competition study by saturating X. fastidiosa cell suspensions with different sugars and measuring cell adhesion to the extracts. We suspended X. fastidiosa cells in PBS with 0.5 M of D(⫹)-mannose, D(⫹)-glucose, D(⫹)-galactose, N-acetylglucosamine, chitobiose (N-acetylglucosamine dimer), or chitotriose (N-acetylglucosamine trimer) prior to blotting. After three 15-min washes with PBS-Tween, strips were incubated for 1 h at room temperature in PBS with purified polyclonal immunoglobulins G against total X. fastidiosa at a final concentration of 5 ␮g/ml (provided by B. C. Kirkpatrick, UC Davis). NCM strips were again washed three times and incubated at room temperature for 1 h with alkaline phosphatase-conjugated goat anti-rabbit immunoglobulins G (Sigma-Aldrich). Strips were washed three times in PBS-Tween, and bound antigenantibody complexes were detected with an alkaline phosphatase color developer (NBT/BCIP) according to the manufacturer’s instructions (Sigma). In other experiments, we used water-soluble copolymers (O-glycosylacrylamides) (9) containing ␣-D-glycosyl, ␣-D-mannosyl, and ␣-D-galactosyl ligands to test for the specificity of X. fastidiosa attachment to monomers in the absence of other molecules that could confound results. We placed 2 ␮l of 10-fold serial dilutions of copolymers (stock solutions were 200 ␮l/ml in 0.1 M sodium bicarbonate, pH 9) onto NCM. After incubation in PBS containing 3% BSA (wt/vol) to block nonspecific binding sites, the membranes were incubated overnight with green fluorescent protein (GFP)-labeled X. fastidiosa (33) at an OD600 of 0.4 in PBS under slow shaking conditions at 4°C. After washing of the membrane in PBSTween three times, the attachment of GFP-labeled cells to the copolymers was detected with a UV light and appropriate GFP filters. Images were acquired with an epifluorescence stereomicroscope in the Biological Imaging Facility at UC-Berkeley. To test for X. fastidiosa adhesion to intact surfaces mimicking the adhesion sites within insects, we used hindwings of the leafhopper vector Homalodisca vitripennis (1). H. vitripennis was selected for these assays due to excessive autofluorescence observed with other species. We suspended GFP-labeled X. fastidiosa cells in PBS (OD600 of 0.4) in the presence of a series of different concentrations of N-acetylglucosamine (0, 50, 100, 200, 500, and 1,000 mM). These suspensions were incubated overnight at 4°C in PBS containing the GFPlabeled bacteria. After washing of the wings with PBS-Tween, the adhered cells were visualized as described above. Lastly, to determine the specificity of bacterial adhesion to intact leafhopper surfaces, GFP-labeled Escherichia coli DH5␣ and the plant pathogens Pseudomonas syringae pv. syringae B728a, Erwinia herbicola 299R (31), Xanthomonas campestris pv. campestris 8004 (35), and X. fastidiosa (33) (all with an OD600 of 0.4) were incubated overnight with vector hindwings and visualized after incubation as described above. Attachment profiles of surface protein and other knockouts. We used 10 knockout X. fastidiosa mutants generated by other research groups (Table 1). NCM strips were spotted with 2-␮l drops of vector foregut extracts, forewings (previously prepared as described above for foregut extract), crab shell chitin (1% [vol/wt] in PBS), 100 mg/ml ovalbumin, and 100 mg/ml BSA. Two-microliter aliquots of 25-␮g solutions of X. fastidiosa proteins (acetone extracted; concentration was determined by the Bradford procedure) were spotted as internal

TRANSMISSION OF X. FASTIDIOSA TO PLANTS

VOL. 75, 2009 controls. Strips were then blocked with 6% nonfat milk in PBS. Strips were incubated in PBS with a protease inhibitor (Sigma-Aldrich) containing different X. fastidiosa mutants. After three 15-min washes with PBS-Tween, attached cells were immunologically detected as described above. We used quantitative real-time reverse transcription-PCR to measure the expression of five putative vector adhesion-associated X. fastidiosa genes. RNAs were obtained from the same X. fastidiosa wild-type and knockout strains used for the biochemical study described above. RNA isolation, cDNA synthesis, and real-time reverse transcription-PCR were carried out as described by Chatterjee et al. (11). We normalized all samples in relation to the wild-type results (values of ⬎1 indicate upregulation, and those of ⬍1 indicate repression). All samples were run three times and the results averaged, and each treatment was replicated three times. Insect transmission bioassays. Adult G. atropunctata leafhopper vectors were collected in 2007 on riparian vegetation at Wohler Park near Guerneville, Sonoma County, CA. We kept insects on sweet basil (Ocimum basilicum), moving the colony to new plants once a week or every other week as necessary. Only second-generation adults were used in biological transmission bioassays; those individuals should be free of the bacterium even if some collected adults were infected, as X. fastidiosa is not transovarially transmitted and is lost when insects molt (3). G. atropunctata is an efficient vector of X. fastidiosa from grape to grape compared to other vectors (1). Grapevines (Vitis vinifera) were used for all greenhouse experiments. We mechanically inoculated plants with the X. fastidiosa wild type or with hxfA and hxfB mutants following previously published protocols (23). Three months after inoculation, we cultivated X. fastidiosa cells from plants (23) and used an immunocapture PCR protocol (44) to confirm that plants were infected. Prior to transmission tests, we caged all G. atropunctata adults on healthy grape plants for 4 days to confirm their noninfected status; none of these pretest plants became infected. Groups of four of these adult G. atropunctata insects were then caged on each infected source plant for a 4-day acquisition access period to acquire X. fastidiosa. The four insects were then divided into groups of two individuals and transferred to healthy seedlings for a 4-day inoculation period (n ⫽ 10 for each strain). In a second experiment, individual insects were used for the inoculation access period (n ⫽ 50 for each strain). Plants were maintained in a greenhouse for 3 months. X. fastidiosa cells were then cultivated from plants to estimate the transmission rate. The proportions of infected to healthy plants were compared among the three X. fastidiosa strains, using a one-by-three contingency table analysis. We performed another experiment comparing the colonization of sharpshooter vectors (G. atropunctata) with that of the X. fastidiosa wild type and both hxf mutants. Grapevines were mechanically inoculated with cell suspensions as described above and used for tests after all plants exhibited disease symptoms. Because hxf mutants were hypervirulent compared to the wild type (22; our observations), we did not expect that differences in host colonization would impact this study; if anything, vectors would be expected to acquire cells more often from mutant-infected plants. We confined adult G. atropunctata insects on symptomatic plants for a short pathogen acquisition period of 12 h aimed at reducing the probability of multiple acquisition events and limiting potential differences in cell numbers detected within insects that could arise from variability among insects that acquired X. fastidiosa early (with time for cell multiplication within vectors) compared to those acquiring cells late in the 12-hour period. After the 12-hour acquisition access period, insects were maintained on basil for cell multiplication only, not for reacquisition from plants. At various times after acquisition, we randomly selected insects from basil, extracted the DNA from their whole heads, and used real-time PCR to determine which individuals were X. fastidiosa positive. For positive individuals, we used real-time PCR to quantify the number of X. fastidiosa cells within those individuals (using a modified version of the protocol described in reference 16). SYBR green technology was used instead of TaqMan (i.e., no probe was used), and amplification was performed with a Fast ABI 7500 real-time PCR system (Applied Biosystems). In total, we tested at least four insects per period for each strain. We compared X. fastidiosa populations among strains by two-way analysis of covariance, with strain as the fixed effect and collecting time as a covariate (14). X. fastidiosa populations were log and then square root transformed, and time was log transformed to meet test assumptions.

RESULTS X. fastidiosa surface proteins bind to polysaccharides in vitro. Because the foregut cuticle of leafhoppers should be rich in polysaccharides, we hypothesized that carbohydrates mediate the attachment of X. fastidiosa to vectors. We incubated X.

523

FIG. 1. Attachment of X. fastidiosa cells to polysaccharide-coated NCM pieces. (A) Comparison of X. fastidiosa attachment profiles with a lectin (WGA) and an albumin (BSA). (B) Effects of protease treatments and cell suspension washes on X. fastidiosa attachment to polysaccharides. Note the nonspecific staining in the chitosan treatment.

fastidiosa cells with polysaccharide-coated NCM pieces, determining cell adhesion with an amino acid stain. The lectin WGA was used as a positive control because of its carbohydratebinding profile. WGA was detected bound to all polysaccharide-coated NCM strips. We found X. fastidiosa cells bound to NCM coated with chitosan, chitin, carboxymethyl cellulose, cellulose, and dextran sulfate, similar to what we observed with WGA. As expected, BSA (not a carbohydrate-binding protein) did not bind to any of these compounds, except for some background binding to chitosan (Fig. 1A). To test whether X. fastidiosa surface proteins were implicated in attachment to polysaccharides, we washed cells to remove the secreted proteins and loosely bound EPS (extracellular polymeric substances or gum). Washed cells attached to polysaccharidecoated NCM similarly to unwashed cells (Fig. 1B), suggesting that molecules responsible for cell adhesion properties are not secreted into the environment. We also determined that cells treated with proteases attached to polysaccharides less than untreated cells did (Fig. 1B). These observations indicate the presence of carbohydrate-binding proteins on the outer surface of X. fastidiosa cells. Components of the bacterial surface have affinity for glucose and mannose but not for galactose. To determine the affinity of X. fastidiosa cells for different sugars, we used a competition assay based on the concept that polysaccharide-binding proteins on the surface of X. fastidiosa can be saturated by exogenous molecules, reducing overall cell attachment to leafhopper foregut extracts. D(⫹)-Galactose did not interfere with the binding of X. fastidiosa to foregut extracts, while D(⫹)-mannose had a small effect (Fig. 2A). However, the monomeric moiety of chitin, N-acetylglucosamine, as well as its dimer (chitobiose), its trimer (chitotriose), and its core molecule (glucose), blocked cell adhesion to leafhopper foregut extracts

524

KILLINY AND ALMEIDA


FIG. 2. Carbohydrate-mediated inhibition of X. fastidiosa cell attachment to surfaces. (A) Carbohydrate inhibition of X. fastidiosa attachment to leafhopper foregut extracts spotted on NCM, indicating that cell surface adhesins can be saturated if they are incubated with certain molecules (GlcNac). (B) Adhesion of GFP-labeled X. fastidiosa to carbohydrate-acrylamide copolymer (O-glycosylacrylamides) dilution series. (C) Dilution series of N-acetylglucosamine inhibiting the attachment of GFP-labeled X. fastidiosa to leafhopper hindwings. (D) Specific adhesion of X. fastidiosa to insect hindwings compared to that of other plant-pathogenic bacteria and Escherichia coli.

(Fig. 2A). The affinity of X. fastidiosa cells for sugars was also tested using synthetic copolymers (9), which eliminate potential sources of error compared to the competition assays, as the latter were performed with leafhopper extracts mimicking in vivo conditions that were certainly complex in composition. We also used GFP-labeled X. fastidiosa (33) to limit sample processing. We determined that cells specifically bound to the glucosyl ligand “poly(O-␣-D-glucopyranosylacrylamide) copolymer.” A negligible interaction was obtained with the galactosyl ligand, while binding of X. fastidiosa to the mannosylated copolymer was detected halfway through the dilution series used (Fig. 2B). In order to compare our in vitro observations to in vivo cell adhesion to leafhoppers, we used the hindwings of insect vectors to mimic the cuticular surface of the foregut canal that X. fastidiosa colonizes. The entire exoskeleton of insects is generally assumed to have a similar chemical composition, although details are lacking for this specific system. We used N-acetylglucosamine as a competitor molecule in assays testing

for GFP-labeled X. fastidiosa cell attachment to hindwings. Attachment diminished as the N-acetylglucosamine concentration increased in the dilution series (Fig. 2C). These results indicate that X. fastidiosa binding to polysaccharides in vitro is similar in its characteristics to its binding to the leafhopper cuticle. Lastly, in order to test the specificity of bacterial adhesion to leafhopper hindwings, we tested if other GFP-labeled bacteria, including the plant pathogens Pseudomonas syringae, Xanthomonas campestris, and Erwinia herbicola and Escherichia coli, attached to that surface (Fig. 2D). Interestingly, only X. fastidiosa cells attached to the wings. Thus, X. fastidiosa cells have surface proteins with affinity for polysaccharides on the surfaces of insect wings and for glucosylated molecules, which can be saturated by N-acetylglucosamine and similar molecules. Role of hemagglutinin-like proteins in cell adhesion. We used several X. fastidiosa knockout mutants to identify proteins associated with cell adhesion to leafhopper foregut extracts and other substrates spotted on NCM strips. Adhesion to leaf-


VOL. 75, 2009

FIG. 3. Attachment of X. fastidiosa knockout mutants to NCM pieces coated with vector foregut extracts and other substrates. See Table 1 for a list of the mutants used in this test.

hopper foregut (G. atropunctata) and wing (H. vitripennis) extracts, in addition to crab shell chitin, was observed for the wild type (Fig. 3). We found that only the rpfF, rpfF rpfC, hxfA, and hxfB mutants showed less attachment than the wild type. The genes disrupted in the cell-cell signaling mutants regulate a cascade of pathways and genes (10, 11, 48), including hxfA and hxfB. To confirm our biochemical observations, we quantified the expression of five genes of putative importance in cell adhesion (afimbrial and fimbrial adhesins and gum) for the same mutants used in these assays compared to that in the wild type (Table 2). Our results showed that hxfA and hxfB were downregulated in rpfF and rpfF rpfC mutants, which showed reduced attachment in the biochemical study. Thus, of the genes tested in our study, the afimbrial adhesin genes hxfA and hxfB were implicated in X. fastidiosa cell adhesion to the substrates tested. Hemagglutinin-like proteins are important for X. fastidiosa early attachment to vectors and transmission to plants. We conducted two experiments to determine the role of hxfA and hxfB in X. fastidiosa transmission by sharpshooters to plants. In the first experiment, we confined noninfected G. atropunctata insects to plants mechanically inoculated with wild-type and hxfA and hxfB mutant cells, after which groups of two individuals were moved to healthy plants for 4 days as an inoculation access period. Transmission occurred in all treatments, with the hxfA and hxfB mutants being transmitted less than the wild type (70, 80, and 100%, respectively) (Fig. 4A), albeit with no statistical difference (␹2 test; P ⫽ 0.1864). In a second experiment, we used individuals instead of pairs to more precisely estimate single-insect transmission efficiencies. With this more discriminating approach, we found that the hxfA and hxfB

525

mutants were transmitted at lower rates than the wild type (36, 46, and 88%, respectively) (Fig. 4A) (␹2 test; df ⫽ 1; P ⬍ 0.001). Because X. fastidiosa transmission rates are correlated with the bacterial population in plants, we quantified the infection levels in plants used in these tests. Plants infected with the hxfA and hxfB mutants used for the transmission tests had populations that were ⬃10-fold larger than those of plants infected with the wild type (data not shown; results are similar to those in reference 22), suggesting that the hxfA and hxfB mutants were transmitted less than the wild type because of their impaired interactions with insects rather than because of smaller populations in source plants. We hypothesized that the reduced transmission rates of hxfA and hxfB mutants were due to limited colonization of vectors early in biofilm formation. In order to test this hypothesis, we conducted another experiment and quantified the numbers of X. fastidiosa cells in the heads of vectors over time after a 12-hour pathogen acquisition access period. Overall, 80% of insects that fed on grapevines infected with the wild type were positive for X. fastidiosa, whereas only 38% and 42% of those fed on plants infected with the hxfA and hxfB mutants were infected, respectively (␹2 test; df ⫽ 2; P ⬍ 0.001). We quantified the numbers of cells of these strains within vectors (positive samples only). There were significant effects of strain (F2,54 ⫽ 23.229; P ⬍ 0.0001) and time (F1,54 ⫽ 803.341; P ⬍ 0.0001) and a strain-time interaction (F2,54 ⫽ 5.362; P ⫽ 0.0075). Populations of the two mutants soon after leafhopper access to infected plants were similar to each other but statistically different from that of the wild type (Fig. 4B). Twelve hours after acquisition, we found that insects fed on the wild type averaged 415 detectable cells, whereas averages of 96 and 120 cells were detected in leafhoppers fed on hxfA mutant- and hxfB mutant-infected plants, respectively. However, after 96 h, the bacterial populations of all three strains were similar to each other (Fig. 4B). Thus, the knockouts were impaired in early attachment to insects, but after initial adhesion, their patterns of foregut colonization (i.e., population growth) were similar to that of the wild type (slope of regressions), suggesting that hxfA and hxfB may have redundant roles in relation to vector transmission. DISCUSSION Molecular interactions between insects and pathogens are often neglected in the study of vector-borne diseases. The characterization of vector-pathogen interactions has been especially poor for bacterial plant pathogens, although a large

TABLE 2. Relative quantification of gene expression in different Xylella fastidiosa mutants by quantitative real-time PCR Fold change in expression in X. fastidiosa mutant relative to that in wild type (mean ⫾ SE)a

Gene

hxfA hxfB fimA pilY1 gumJ

rpfC

rpfF

rpfF rpfC

hxfA

hxfB

fimA

pilB

fimA pilO

gumD

gumH

2.8 ⫾ 0.3 2.11 ⫾ 0.5 2.3 ⫾ 0.05 0.9 ⫾ 0.03 3.1 ⫾ 0.08

0.53 ⫾ 0.06 0.11 ⫾ 0.03 0.3 ⫾ 0.04 1.2 ⫾ 0.05 0.4 ⫾ 0.03

0.8 ⫾ 0.07 0.45 ⫾ 0.07 0.5 ⫾ 0.1 1.2 ⫾ 0.06 0.6 ⫾ 0.05

ND 0.9 ⫾ 0.02 1.2 ⫾ 0.03 1.1 ⫾ 0.03 0.9 ⫾ 0.01

0.82 ⫾ 0.03 ND 1.1 ⫾ 0.04 0.9 ⫾ 0.04 0.8 ⫾ 0.03

1.4 ⫾ 0.1 1.3.⫾03 ND 1.3 ⫾ 0.03 0.7 ⫾ 0.01

1.2 ⫾ 0.16 1.6 ⫾ 0.2 1.5 ⫾ 0.2 ND 1.2 ⫾ 0.02

1.1 ⫾ 0.03 1.2 ⫾ 0.01 ND ND 0.7 ⫾ 0.03

2.55 ⫾ 0.4 2.2 ⫾ 0.3 1.4 ⫾ 0.04 0.9 ⫾ 0.03 ND

1.68 ⫾ 0.1 1.4 ⫾ 0.1 1.3 ⫾ 0.03 1.1 ⫾ 0.01 ND

a Quantity of mRNA relative to that in wild-type cells. Equal amounts have a value of 1.0. Normalization was carried out using 16S rRNA as an endogenous control. ND, not determined.

526

KILLINY AND ALMEIDA

FIG. 4. (A) Transmission of X. fastidiosa by leafhopper vectors. Both experiments show that hxfA and hxfB mutants were transmitted less often than the wild type, but results from a larger experiment using individuals instead of groups were statistically significant. Different letters on bars indicate statistically significantly different treatments (P ⬍ 0.05). (B) Bacterial populations within leafhopper vectors over time after a 12-hour pathogen acquisition access period. Data are shown for the wild type (solid regression line), the hxfA mutant (dotted regression line), and the hxfB mutant (dashed regression line). Note the values immediately after acquisition (12-hour period) and 4 days afterwards. Fewer hxfA and hxfB cells adhered to vectors, but after a few days, the populations were of equal size.

number of plant viruses have been studied in detail in relation to their transmission biology (20, 36). X. fastidiosa is not an exception; despite the fact that several genome sequences of this organism are available and much of its biology and ecology are well understood, no studies have determined the molecular basis of cell attachment to leafhopper vectors. We conducted a series of experiments that allowed us to conclude that specific X. fastidiosa surface proteins bind to chitin-like polysaccharides on the foregut cuticular lining of leafhoppers. In addition, we demonstrated that these proteins are required for initial cell adhesion to vectors but that once adhered, mutants with reduced attachment abilities colonized insects similarly to the wild type (Fig. 4B). In fact, several adhesins may be necessary for the first steps in leafhopper colonization; those may also have different or complementary functions in the first stages. These results suggest that X. fastidiosa colonization of surfaces on vectors is a complex multistep process similar to that for other biofilm-forming bacteria (15). The cuticular lining of the foregut of insects is part of the exoskeleton. However, little is known about insect foregut ultrastructure, particularly for sap-sucking insects. In addition, the physical and chemical properties of the cuticles of insects vary within individuals and may change during development. For example, the foregut cuticle of the beetle Costelytra zealandica (Coleoptera, Scarabaeidae) has no free lipid layer such as those present on its external cuticle, but the outermost layer of the foregut cuticle contains glycoproteins or glucolipids that


act as receptors for Serratia entomophila (7). In another study, the foregut cuticle of the cockroach Periplaneta americana was found to be different from the body cuticle, with Mallory’s stain indicating the presence of chitin in the foregut’s cuticle surface (32). Accordingly, we assumed the presence of carbohydrate moieties in leafhopper foregut surfaces colonized by X. fastidiosa. We demonstrated that X. fastidiosa has an affinity for polysaccharides, acting similarly to lectins, and that proteins on the cell surface are responsible for such activity. Specificity tests indicated that X. fastidiosa cells bind to chitin and similar carbohydrates but not to galactose. These biochemical observations, coupled with our finding that X. fastidiosa attaches to leafhopper wings but other plant-pathogenic bacteria do not, suggest that adhesion to these surfaces is a specific process. Although we did not characterize the surface of the insect foregut, we used vector hindwings as proxies for the foregut to mimic the interactions of cells and the intact insect surface. In our study, we found that leafhopper wings can be stained with Alcian blue 8GX, which is specific for glycosaminoglycans similar to the polysaccharides we found to which X. fastidiosa cells can bind (data not shown). This finding suggests the presence of a layer of polysaccharides on the surface of the leafhopper wing cuticle. To determine which cell surface proteins were associated with adhesion to vectors, we tested a series of sitedirected knockouts of genes of putative importance in vector colonization. X. fastidiosa has a limited number of annotated candidate adhesins, including type I and IV pili (short and long fimbriae, respectively) and hemagglutinin-like proteins (HxfA and HxfB). We found that hxfA and hxfB mutants were deficient in cell adhesion in vitro, as were two cell-cell signaling mutants (rpfF and rpfC rpfF mutants). The rpfF mutant is not transmissible to plants and does not colonize the foregut of vectors (34); the rpfC rpfF mutant is also nontransmissible (unpublished data). To determine if the levels of hxfA and hxfB expression were reduced in these adhesion-deficient mutants, we compared the transcription levels of the afimbrial adhesin genes, two pilus genes, and one gum gene for all mutants tested biochemically. hxfA and hxfB expression levels were consistently lower than those in the wild type only in the knockouts that did not adhere to leafhopper foregut extracts (Table 2). Thus, although we focused our study on a limited number of candidate proteins, our results demonstrate that X. fastidiosa adhesion to carbohydrates is mediated by HxfA and HxfB. Other afimbrial adhesin homologs have been identified in X. fastidiosa genomes (45) and may also be important in cell-carbohydrate interactions. No studies have been conducted on those proteins. Furthermore, X. fastidiosa genomes still have a large number of hypothetical open reading frames, and screens for carbohydrate-binding proteins may identify new adhesins of importance in this system. It seems that the role of carbohydrate-binding proteins in X. fastidiosa is similar to that in other bacteria, especially Vibrio spp. The attachment of Vibrio alginolyticus and Vibrio cholerae to the chitinous exoskeleton of marine arthropods is reduced if cells are treated with N-acetylglucosamine, N,N⬘-diacetylchitobiose, N,N⬘,N⬙-triacetylchitotriose, pronase E, chitin, or WGA, while no reduction in adhesion is observed with D-glucose, fucose, or fructose (46). These observations closely match our results with X. fastidiosa, except that D-glucose interferes with

VOL. 75, 2009

the attachment of X. fastidiosa to foregut extracts, while it does not reduce Vibrio sp. adhesion to the exoskeletons of marine arthropods. For V. cholerae, a role in the attachment to chitin has been demonstrated for surface proteins, including mannose-sensitive hemagglutinin, which promotes the adherence of cells to zooplankton (12, 46, 49). A 134-kDa protein, chitovibrin, is secreted by the marine bacterium Vibrio parahaemolyticus and is inducible by chitin and chitin oligomers. Chitovibrin shows no apparent enzymatic activity but exhibits a strong affinity for chitin and chito-oligomers (19). In V. cholerae, membrane proteins are able to bind chitin particles, and two of these peptides (molecular masses, 36 and 53 kDa) bind specifically to GlcNAc (46). On the other hand, V. cholerae not only attaches to chitin but also degrades it by its extracellular chitinases (13). It seems that chitin consumption regulates gene expression, as microarray expression profiling and mutational studies of V. cholerae growing on a natural chitin surface or with the soluble chitin oligosaccharides (GlcNAc)2-6, GlcNAc, or glucosamine dimer (GlcN)2 identified three sets of differentially regulated genes (29). In addition, V. cholerae can acquire new genetic material by natural transformation during growth on chitin (28). It remains to be determined if X. fastidiosa uses chitin as a carbon source. Except for the observation that X. fastidiosa’s cell-cell signaling system controls genes associated with vector transmission (11, 34), our previous knowledge of interactions in this system was based on indirect data and assumptions. Chatterjee et al. (10) summarized a hypothesis suggesting that X. fastidiosa has separate plant colonization and insect acquisition phases, which are controlled by accumulation of the signaling molecule DSF. Indeed, hxfA and hxfB are upregulated in the presence of DSF, supporting this hypothesis and corroborating work done with the cell-cell signaling-deficient rpfF mutant. However, our results show that both HxfA and HxfB are required for optimal initial adhesion to vectors, but their importance for biofilm maturation seems limited. In the future, a double knockout mutant for these genes will need to be tested to determine if these proteins have redundant roles in mediating cell adhesion to carbohydrates (hxfA and hxfB share high sequence similarity, with the exception of two deletions in hxfB in relation to hxfA), a test not technically possible until recently (41). We propose that X. fastidiosa colonization of vectors is similar to the formation of biofilms on surfaces. Scanning electron microscopy observations we made previously (2) support this hypothesis. We hypothesize that cells initially adhere laterally to the foregut cuticle via carbohydrate-binding proteins, such as HxfA and HxfB (Fig. 5A and B). Since these proteins are assumed to occur throughout cells, adhering laterally increases the cell surface area in contact with the substrate and streamlines the bacteria to the flow of xylem sap ingested by the insect vector. After initial adhesion, cells may produce large quantities of EPS that can result in the concentration of resources and DSF in microcolonies. As the colony size increases, cells at the center of the biofilm become polarly attached to the foregut surface (Fig. 5C), potentially through polar short type I pili, increasing the surface area for nutrient absorption. Cellcell attachment may be mediated by hemagglutinin-like or other cell surface proteins (22). Lastly, a typical mature X. fastidiosa biofilm within vectors is formed, with all cells polarly


527

FIG. 5. Hypothetical model of X. fastidiosa colonization in leafhopper vectors. (A) Cells initially attach laterally to the cuticle of insects, a process mediated by HxfA, HxfB, and possibly other carbohydratebinding proteins. Microcolonies are established (B) and change in morphology (C), with cells in the center becoming polarly attached to increase the exposed surface area; type I pili may be important for polar attachment. (D) Mature biofilm, with newly divided daughter cells not attached to the leafhopper cuticle being subject to detachment from the biofilm. Scanning electron microscope pictures are unpublished images obtained in a previous study (2).

attached (Fig. 5D). At this stage, newly divided cells are not anchored on the cuticle of insects and may occasionally be detached from vectors and inoculated into plants. This hypothesis (summarized in Fig. 5) may be useful to guide future studies on this system by providing testable questions, as up until now no data on these interactions, with the exception of microscopy observations, were available. ACKNOWLEDGMENTS We thank Steve Lindow, Bruce Kirkpatrick, Harvey Hoch, Tom Burr, and Don Cooksey for providing us with X. fastidiosa mutants. We also thank David Morgan for providing us with some of the sharpshooters used here and Sandy Purcell and our lab colleagues for helpful discussions, especially Matt Daugherty for assistance with statistical analyses. This work was supported with funding from the California Department of Food and Agriculture (PDRP). REFERENCES 1. Almeida, R. P. P., and A. H. Purcell. 2003. Transmission of Xylella fastidiosa to grapevines by Homalodisca coagulata (Hemiptera, Cicadellidae). J. Econ. Entomol. 96:264–271. 2. Almeida, R. P. P., and A. H. Purcell. 2006. Patterns of Xylella fastidiosa colonization on the precibarium of leafhopper vectors relative to transmission to plants. Ann. Entomol. Soc. Am. 99:884–890. 3. Almeida, R. P. P., M. J. Blua, J. R. S. Lopes, and A. H. Purcell. 2005. Vector transmission of Xylella fastidiosa: applying fundamental knowledge to generate disease management strategies. Ann. Entomol. Soc. Am. 98:775–786. 4. Andersen, S. O. 1979. Biochemistry of insect cuticle. Annu. Rev. Entomol. 24:29–59. 5. Barbehenn, R. V., and A. H. Purcell. 1993. Factors limiting the transmission of a xylem inhabiting bacterium Clavibacter xyli subsp. cynodontis to grasses by insects. Phytopathology 83:859–863. 6. Beard, C. B., C. Cordon-Rosales, and R. V. Durvasula. 2002. Bacterial symbionts of the Triatominae and their potential use in control of Chagas disease transmission. Annu. Rev. Entomol. 47:123–141. 7. Binnington, K. C. 1993. Ultrastructure of the attachment of the bacterium Serratia entomophila to foregut cuticle of Costelytra zealandica (Coleoptera, Scarabidae) and a review of nomenclature for insect epicuticular layers. Int. J. Insect Morphol. Embryol. 22:145–155. 8. Callaway, E. 2008. The green menace. Nature 452:148–150. 9. Chadli, A., M. Caron, M. Ticha ´, R. Jourbet, D. Bladier, and J. Kocourek. 1992. Development of screening methods for detection of carbohydratebinding proteins by use of soluble glycosylated polyacrylamide-based copolymers. Anal. Biochem. 204:198–203. 10. Chatterjee, S., R. P. P. Almeida, and S. E. Lindow. 2008. Living in two worlds: the plant and insect lifestyles of Xylella fastidosa. Annu. Rev. Phytopathol. 46:243–271. 11. Chatterjee, S., C. Wistrom, and S. E. Lindow. 2008. A cell-cell signaling sensor is required for virulence and insect transmission of Xylella fastidiosa. Proc. Natl. Acad. Sci. USA 105:2670–2675. 12. Chiavelli, D. A., J. W. Marsh, and R. K. Taylor. 2001. The mannose-sensitive hemagglutinin of Vibrio cholerae promotes adherence to zooplankton. Appl. Environ. Microbiol. 67:3220–3225.

528

KILLINY AND ALMEIDA

13. Connell, T. D., D. J. Metzger, J. Lynch, and J. P. Folster. 1998. Endochitinase is transported to the extracellular milieu by the eps-encoded general secretory pathway of Vibrio cholerae. J. Bacteriol. 180:5591–5600. 14. Crawley, M. J. 2007. The R book. John Wiley & Sons, West Sussex, England. 15. Davey, M. E., and G. A. O’Toole. 2000. Microbial biofilms: from ecology to molecular genetics. Microbiol. Mol. Biol. Rev. 64:847–867. 16. Francis, M., H. Lin, J. C. Cabrera-La Rosa, H. Doddapaneni, and E. L. Civerolo. 2006. Genome-based primers for specific and sensitive detection and quantification of Xylella fastidiosa. Eur. J. Plant Pathol. 115:203–213. 17. Franz, A. W. E., I. Sanchez-Vargas, Z. N. Adelman, C. D. Blair, B. J. Beaty, A. A. James, and K. E. Olson. 2006. Engineering RNA interference-based resistance to dengue virus type 2 in genetically modified Aedes aegypti. Proc. Natl. Acad. Sci. USA 103:4198–4203. 18. Freitag, J. H., and N. W. Frazier. 1954. Natural infectivity of leafhopper vectors of Pierce’s disease virus of grape in California. Phytopathology 44: 7–11. 19. Gildemeister, O., B. Zhu, and R. Laine. 1994. Chitovibrin: a chitin-binding lectin from Vibrio parahaemolyticus. Glycoconj. J. 11:518–526. 20. Gray, S. M., and N. Banerjee. 1999. Mechanisms of arthropod transmission of plant and animal viruses. Microbiol. Mol. Biol. Rev. 63:128–148. 21. Gubler, D. J. 2008. The global threat of emergent/reemergent vector-borne diseases, p. 43–64. In Vector borne diseases: understanding the environmental, human health, and ecological connections. The National Academies Press, Washington, DC. 22. Guilhabert, M. R., and B. C. Kirkpatrick. 2005. Identification of Xylella fastidiosa antivirulence genes: hemagglutinin adhesins contribute a biofilm maturation to X. fastidiosa and colonization and attenuate virulence. Mol. Plant-Microbe Interact. 18:856–868. 23. Hill, B. L., and A. H. Purcell. 1995. Multiplication and movement of Xylella fastidiosa within grapevine and four other plants. Phytopathology 85:1368– 1372. 24. Hopkins, D. L., and A. H. Purcell. 2002. Xylella fastidiosa: cause of Pierce’s disease of grapevine and other emergent diseases. Plant Dis. 86:1056–1066. 25. Ito, J., A. Ghosh, L. A. Moreira, E. A. Wimmer, and M. Jacobs-Lorena. 2002. Transgenic anopheline mosquitoes impaired in transmission of a malaria parasite. Nature 417:452–455. 26. Li, Y., G. Hao, C. D. Galvani, Y. Meng, L. De La Fuente, H. C. Hoch, and T. J. Burr. 2007. Type I and type IV pili of Xylella fastidiosa affect twitching motility, biofilm formation and cell-cell aggregation. Microbiology 153:719– 726. 27. Locke, M. 1974. The structure and formation of the integument in insects, p. 123–213. In M. Rockstein (ed.), The physiology of Insecta, vol. 6. Academic Press, New York, NY. 28. Meibom, K. L., M. Blokesch, N. A. Dolganov, C. Y. Wu, and G. K. Schoolnik. 2005. Chitin induces natural competence in Vibrio cholerae. Science 310: 1824–1827. 29. Meibom, K. L., X. B. Li, A. T. Nielsen, C. Y. Wu, S. Roseman, and G. K. Schoolnik. 2004. The Vibrio cholerae chitin utilization program. Proc. Natl. Acad. Sci. USA 101:2524–2529. 30. Meng, Y. Z., Y. X. Li, C. D. Galvani, G. X. Hao, J. N. Turner, T. J. Burr, and H. C. Hoch. 2005. Upstream migration of Xylella fastidiosa via pilus-driven twitching motility. J. Bacteriol. 187:5560–5567.

APPL. ENVIRON. MICROBIOL. 31. Miller, W. G., J. H. J. Leveau, and S. E. Lindow. 2000. Improved gfp and inaZ broad-host-range promoter-probe vectors. Mol. Plant-Microbe Interact. 13:1243–1250. 32. Murthy, R. C. 1975. Structure of the foregut cuticle of Peripplaneta americana. Entomol. Exp. Appl. 32:312–317. 33. Newman, K. L., R. P. P. Almeida, A. H. Purcell, and S. E. Lindow. 2003. Use of a green fluorescent strain for analysis of Xylella fastidiosa colonization of Vitis vinifera. Appl. Environ. Microbiol. 69:7319–7327. 34. Newman, K. L., R. P. P. Almeida, A. H. Purcell, and S. E. Lindow. 2004. Cell-cell signaling controls Xylella fastidiosa interactions with both insects and plants. Proc. Natl. Acad. Sci. USA 101:1737–1742. 35. Newman, K. L., S. Chatterjee, K. A. Ho, and S. E. Lindow. 2008. Virulence of plant pathogenic bacteria attenuated by degradation of fatty acid cell-tocell signaling factors. Mol. Plant-Microbe Interact. 21:326–334. 36. Ng, J. C. K., and K. L. Perry. 2004. Transmission of plant viruses by aphid vectors. Mol. Plant Pathol. 5:505–511. 37. Purcell, A. H. 1982. Insect vector relationships with procaryotic plant pathogens. Annu. Rev. Phytopathol. 20:397–417. 38. Purcell, A. H., and A. H. Finlay. 1979. Evidence for noncirculative transmission of Pierce’s disease bacterium by sharpshooter leafhoppers. Phytopathology 69:393–395. 39. Purcell, A. H., A. H. Finlay, and D. L. Mclean. 1979. Pierce’s disease bacterium: mechanism of transmission by leafhopper vectors. Science 206: 839–841. 40. Redak, R. A., A. H. Purcell, J. R. S. Lopes, M. J. Blua, R. F. Mizell, and P. C. Andersen. 2004. The biology of xylem fluid-feeding insect vectors of Xylella fastidiosa and their relation to disease epidemiology. Annu. Rev. Entomol. 49:243–270. 41. Reddy, J. D., S. L. Reddy, D. L. Hopkins, and D. W. Gabriel. 2007. TolC is required for pathogenicity of Xylella fastidiosa in Vitis vinifera grapevines. Mol. Plant-Microbe Interact. 20:403–410. 42. Severin, H. H. P. 1949. Transmission of the virus of Pierce’s disease of grapevines by leafhoppers. Hilgardia 19:190–206. 43. Severin, H. H. P. 1950. Spittle-insect vectors of Pierce’s disease virus. II. Life history and virus transmission. Hilgardia 19:357–382. 44. Shapland, E. B., K. M. Daane, G. Y. Yokota, C. Wistrom, J. H. Connell, R. A. Duncan, and M. A. Viveros. 2006. Ground vegetation survey for Xylella fastidiosa in California almond orchards. Plant Dis. 90:905–909. 45. Simpson, A. J. G., et al. 2000. The genome sequence of the plant pathogen Xylella fastidiosa. Nature 406:151–157. 46. Tarsi, R., and C. Pruzzo. 1999. Role of surface proteins in Vibrio cholerae attachment to chitin. Appl. Environ. Microbiol. 65:1348–1351. 47. van Sluys, M. A., et al. 2003. Comparative analysis of the complete genome sequence of Pierce’s disease and citrus variegated chlorosis strains of Xylella fastidiosa. J. Bacteriol. 185:1018–1026. 48. Wang, N., W. Feil, and S. E. Lindow. 2007. Identification of traits of Xylella fastidiosa conferring virulence to grape and insect transmission by analysis of global gene expression using DNA microarrays. Phytopathology 97:S120. 49. Zampini, M., L. Canesi, M. Betti, C. Ciacci, R. Tarsi, G. Gallo, and C. Pruzzo. 2003. Role for mannose-sensitive hemagglutinin in promoting interactions between Vibrio cholerae El Tor and mussel hemolymph. Appl. Environ. Microbiol. 69:5711–5715.