INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Xylella fastidiosa is a fastidious gram-negative, xylem-limited bacterium (26) that causes a range of economically important plant diseases, including citrus variegated chlorosis (CVC) (3, 23); Pierce's disease (PD) of grapevine; alfalfa dwarf; leaf scorch of almond, coffee, elm, sycamore, oak, plum, mulberry, maple, and oleander; and periwinkle wilt (for reviews, see references 19 and 20).

CVC is a major problem in Brazil, where over 70 million sweet orange trees (34%) are affected. The disease also occurs in Argentina, under the name "pecosita" (7, 9). CVC affects all commercial sweet orange varieties. Affected fruits are small and hardened and thus of no commercial value. Rapid dissemination of CVC comes from the use of infected nursery trees and transmission of X. fastidiosa by several xylem-feeding sharpshooter insect vectors.

The genome sequence of the CVC strain of X. fastidiosa, clone 9a5c, was recently determined, and the nature of genes that were identified by annotation suggests a number of potential pathogenicity mechanisms, such as cell-wall hydrolysis, adhesion, intervessel migration, and toxicity (26). However, gene function must be determined experimentally. One way is through the study of relevant mutants. Production and analysis of such mutants require bacterial transformation with appropriate plasmids.

Broad-host-range and/or suicide plasmids have been used extensively in genetic studies of various gram-negative, phytopathogenic bacteria, including Xanthomonas spp., which are known to be phylogenetically related to X. fastidiosa (27). However, attempts to transform X. fastidiosa 9a5c by conjugal transfer from an Escherichia coli donor strain or by electrotransformation with plasmids pUFR047 (6), pLAFR6 (2), pVSP61 (1), pDSK602 (17), pSUP2021 (25), and pUIRM504 (14) were unsuccessful (P. B. Monteiro and J. Renaudin, unpublished data). Therefore, we have constructed plasmids containing the chromosomal replication origin (oriC) of X. fastidiosa, a strategy which we developed for transformation of Spiroplasma citri, a plant-pathogenic mollicute related to low-guanosine-plus-cytosine gram-positive bacteria (22, 28). Here we report: (i) the construction of plasmids containing all or part of the oriC region of X. fastidiosa, as well as a kanamycin resistance gene driven by an X. fastidiosa rRNA promoter; (ii) the transformation of two X. fastidiosa citrus isolates with these plasmids; and (iii) integration of the plasmids into the bacterial chromosome at the rRNA promoter by homologous recombination. The results suggest that plasmids based on the X. fastidiosa oriC are promising tools for specific gene targeting through homologous recombination in X. fastidiosa.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains. E. coli XL-1 Blue {recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F' proAB lacI^q ZM15 Tn10 (Tet^r)]} was used as the host strain for subcloning experiments and for propagation of plasmids. Strains of X. fastidiosa were isolated and subcultured from symptomatic twigs of sweet orange or coffee as previously described (3, 13). Their geographical origin and host species are indicated in Table 1. X. fastidiosa cells were grown in PW medium (5) at 29°C in the dark with low-speed, rotatory agitation (100 rpm).

Primers and PCR amplification. The sequences of all primers used in this study are described in Table 2. Primers RP1 and RP2 were used to amplify the promoter region of rRNA operon 1 of X. fastidiosa 9a5c (fragment rop in Fig. 1A). Primers OR1 and OR2 were used to produce a 1,893-bp DNA fragment (fragment oriC) comprising the dnaA gene and the 5' and 3' flanking intergenic regions. Amplification with primers OR3 and OR2 yielded a 366-bp product corresponding to the region between the genes dnaA and dnaN (fragment ori in Fig. 1A). Primers KAX1 and KA2 were used to amplify the coding sequence of the Tn903 kanamycin resistance (Kan^r) gene with plasmid pUC4K (Amersham Pharmacia Biotech, Inc., Piscataway, N.J.) as the template. For simplicity we called this fragment "kan^r " in Fig. 1A. Primer KAX1 contains the 10 nucleotides (underlined in Table 2), including the ribosome binding site (boldface characters in Table 2), immediately upstream of the coding sequence of the X. fastidiosa dnaN gene. PCRs were performed in a thermal cycler (Perkin-Elmer Cetus Corp., Norwalk, Conn.) using Pfu turbo DNA polymerase as described by the supplier (Stratagene, La Jolla, Calif.). Amplification was achieved over 40 cycles of 30 s at 92°C, 1 min at 60°C, and 4 min at 72°C, with an additional step of 10 min at 72°C. For the amplification of fragments rop, oriC, and ori, we used purified genomic DNA of X. fastidiosa as the template. Cloned PCR products were sequenced with a T7 sequencing kit (Amersham Pharmacia Biotech, Inc.). Amplification reactions with primer pairs RP1-KA2 and M13 universal primer (M13U)-RP3 were carried out in a 20-µl reaction mixture containing 1 µl of X. fastidiosa culture or 50 ng of target DNA, 50 mM Tris-HCl (pH 8.8), 2 mM MgCl₂, 200 µg of bovine serum albumin per ml, 0.05% W1 detergent, 0.2 mM deoxynucleoside triphosphates, each primer at 1 µM, and 2.5 U of Taq DNA polymerase (GIBCO/BRL Life Technologies, Inc., Gaithersburg, Md.). After a denaturation step of 4 min at 94°C, amplification was achieved over 45 cycles of 1 min at 92°C, 50 s at 60°C, and 3 min at 72°C, with a final step of 10 min at 72°C. The PCR product obtained with primer pair M13U-RP3 was purified from agarose gel and directly sequenced. Primers OR1, OR2, OR3, RP1, RP2, and RP3 (Table 2) were designed from the X. fastidiosa genome sequencing data accessible online at http://www.lbi.dcc.unicamp.br/xylella/.

View larger version (29K): [in this window] [in a new window]   FIG. 1.   (A) Partial restriction maps of plasmids p16K, p16KdAori, and p16Kori. pBS, pBluescript; rop, promoter region of X. fastidiosa ribosomal operon 1; kanr, kanamycin resistance gene from Tn903; ori, X. fastidiosa dnaA-dnaN intergenic region. The arrows indicate the direction of transcription. (B) Gene organization of the X. fastidiosa oriC region and nucleotide sequence of the dnaA-dnaN intergenic region. Genes rpmH, dnaA, dnaN, recF, and gyrB are indicated by open arrows. Positions of primers OR1, OR2, and OR3 are indicated by black arrows. Boxed sequences R1 to R5 are putative dnaA boxes. The two GATC sites are double underlined; the stretch of ACC triplets is single underlined. The AT clusters are underlined with a dotted line. The upper and lower asterisks indicate the translational stop codon of dnaA and the initiation codon of dnaN, respectively.

Construction of plasmids p16K, p16KdAori, and p16Kori. The promoter region of the X. fastidiosa rRNA operon 1 was obtained as an 831-bp DNA fragment by PCR amplification of genomic DNA with primer pair RP1-RP2. The coding sequence of the kanamycin resistance gene was amplified with the primer pair KAX1-KA2 with pUC4K as the template. Both fragments were digested with EcoRI. These fragments were then ligated with T4 DNA ligase, and the resulting ligation products were restricted with KpnI. The KpnI-digested DNA was ligated to the dephosphorylated, KpnI-linearized pBS+ vector (Stratagene, La Jolla, Calif.), and the final ligation mixture was used to electrotransform E. coli XL-1 Blue competent cells. Primer pair RP1-KA2 was used in a PCR to screen the transformed, kanamycin-resistant clones. The relative orientations of the promoter and Kan^r gene fragments were determined by double digestion of the recombinant plasmids with enzymes EcoRI and SmaI. The recombinant plasmid carrying the Kan^r gene downstream of the promoter and in the appropriate orientation for transcription was named p16K (Fig. 1A). To construct plasmid p16KdAori, a 1,893-bp oriC fragment containing the dnaA gene and its flanking regions was obtained by amplification of genomic DNA from X. fastidiosa 9a5c with primer pair OR1-OR2 (Fig. 1B). The PCR product was digested with BamHI and inserted at the BamHI site of the pBS+ vector. The cloned fragment was rescued from the recombinant plasmid as a SmaI-HincII fragment of 1,903 bp and inserted into the XbaI-linearized, Klenow-filled-in, plasmid p16K to produce the final plasmid construct, p16KdAori (Fig. 1A). The relative orientations of the oriC and the Kan^r fragments were checked by PCR with the primer pair KAX1-OR2. In plasmid p16KdAori, the dnaA and Kan^r genes are in the same orientation. To construct the p16Kori plasmid, a 366-bp fragment encompassing the dnaA-dnaN intergenic region was amplified with primer pair OR3-OR2 (Fig. 1B). The PCR product was digested with BamHI, ligated to dephosphorylated, BamHI-linearized pBS+ vector, and cloned into E. coli. The cloned fragment was rescued from the recombinant plasmid as a SmaI-HincII fragment of 376 bp and transferred to XbaI-linearized, Klenow-filled-in plasmid p16K to yield the recombinant plasmid p16Kori (Fig. 1A). Using the primer pair RP1-OR2, it was determined by PCR that the dnaA box region was cloned in an orientation reverse that of the Kan^r gene.

Transformation of X. fastidiosa and selection of transformants. A single colony from a freshly streaked PW agar plate was dispersed in 2 ml of liquid PW medium (minus MgSO₄, plus L-histidine [0.1%, wt/vol]) by vortexing, and the cell suspension was incubated at 29°C for 4 days. A sample (0.3 ml) of this culture was used to inoculate 30 ml of PW. After a 4-day incubation under the same conditions, the culture was transferred to a chilled polypropylene tube and the cells were collected by centrifugation (2,600 × g for 15 min) at 4°C. The pelleted cells were washed twice in 30 ml of chilled ultrapure (milli-Q; Millipore) water and once in cold 10% glycerol. The final cell pellet was resuspended in 0.3 ml of 10% glycerol and kept on ice. For electroporation, an 80-µl aliquot of cell suspension was mixed with 5 to 10 µg of DNA in 5 µl of TE buffer (10 mM Tris-HCl [pH 8], 1 mM EDTA). The mixture of cells and DNA was transferred to a cold 0.2-cm electroporation cuvette and kept on ice for 1 min. The cells were electroporated at 2.5 kV, 200 , and 25 µF to generate a pulse of approximately 6 ms. After electroporation, the cells were resuspended in 1 ml of PW plus L-histidine (0.1%, wt/vol) recovery medium and incubated at 29°C for 6 h without agitation. The transformants were selected by plating (250 µl of cell suspension per plate) on PW agar medium supplemented with kanamycin (5 µg per ml). The plates were wrapped with Parafilm to prevent desiccation and incubated for 20 days at 29°C. The kanamycin-resistant colonies were picked individually and grown in broth medium containing 5 µg of kanamycin per ml. During propagation, the antibiotic concentration was progressively increased to 20 µg/ml.

DNA isolation and Southern blot hybridization. Large-scale and small-scale preparations of plasmid DNA amplified in E. coli were carried out according to standard procedures (24). X. fastidiosa DNA was isolated as described by Chen and coworkers (4), with the following modification. After the cells were lysed with lysozyme, the preparation was heated at 60°C for 1 h in the presence of 0.5% sodium dodecyl sulfate and 0.1 mg of proteinase K/ml of solution. X. fastidiosa extrachromosomal DNA was prepared by the alkaline lysis method (24). For Southern blot hybridizations, restricted DNA was denatured in 0.4 N NaOH and blotted to positively charged nylon membranes by the capillary transfer procedure with 10× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate). Hybridizations with biotinylated probes were carried out according to the procedures outlined in the labeling kit, BioPrime, and the detection kit, Blue Gene (GIBCO/BRL Life Technologies, Inc.). The biotinylated probes used in this study were obtained by PCR amplification with the following primer pairs: KA1-KA2 for the kanamycin resistance gene (probe kan), OR2-OR3 for the replication origin (probe ori), and RP1-RP2 for the promoter region of rRNA operon 1 (probe rop).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

oriC region of X. fastidiosa. The oriC region used in this work was selected from the X. fastidiosa genome sequence (26). This region comprises the dnaA gene and the two flanking intergenic regions, one between genes rpmH and dnaA and the other between dnaA and dnaN (Fig. 1B). Plasmid p16KdAori contains the entire oriC region, while in plasmid p16Kori, only the 366-bp intergenic region downstream of dnaA is present (Fig. 1A). This fragment shares features with the oriC region of other gram-negative bacteria. In particular, it contains five dnaA boxes (consensus sequence TTATCCACA), two GATC Dam methylation sites in between dnaA boxes R3 and R4, and two AT-rich sequences of approximately 35 nucleotides each (Fig. 1B). In addition, the region downstream of the dnaA boxes contains a distinctive stretch of 24 nucleotides consisting of 8 repeats of the ACC triplet that could represent the counterpart of the 13-mer repeats in the E. coli oriC region (18).

Transformation of X. fastidiosa isolates with plasmids p16K and p16KdAori. Five citrus strains and two coffee strains of X. fastidiosa (see Table 1 for origins of the strains) were electrotransformed with plasmids p16K and p16KdAori. Transformation assays with all seven strains, using 5 or 10 µg of plasmid p16K, yielded no kanamycin-resistant transformants, indicating that p16K did not replicate or integrate into the host chromosome and that spontaneous resistance was not detected for the level of kanamycin used in the experiments. In contrast, kanamycin-resistant transformants were obtained for the citrus strains J1a12 and B111 electroporated with plasmid p16KdAori but not for the other five strains of X. fastidiosa tested. With both strains J1a12 and B111, the transformation efficiency was found to be very low, approximately 10 transformants per µg of plasmid DNA. Ten kanamycin-resistant colonies of each strain were grown separately in liquid PW medium containing 5 µg of kanamycin per ml. The presence of p16KdAori sequences in these transformants was subsequently demonstrated by PCR amplification with primer pair RP1-KA2 and by hybridization of total DNA with the kan, ori, and rop probes (data not shown).

Maintenance of plasmid p16KdAori in X. fastidiosa transformants. To determine whether the plasmid was maintained as free extrachromosomal DNA or had integrated into the X. fastidiosa chromosome, the transformants carrying p16KdAori were subcultured twice and grown in liquid medium for a period corresponding to approximately 40 generations. The maintenance of the plasmid was monitored by Southern blot hybridization of total and extrachromosomal DNA with various probes. In the experiment represented in Fig. 2, total and extrachromosomal DNAs of two independent X. fastidiosa J1a12 transformants were digested with SphI or BamHI and hybridized with the kan probe. The bands of 7.1 kbp and 2 kbp correspond, respectively, to the SphI and BamHI fragments of the purified plasmid (Fig. 2A, lanes 1 and 7) and were not detected in the transformed cells, either in extrachromosomal DNA (Fig. 2A, lanes 2 and 6) or in the total DNA (Fig. 2A, lanes 3 and 5). Instead, an 11.1-kbp SphI fragment and an 8.4-kbp BamHI fragment were detected in the transformants (Fig. 2A, lanes 3 and 5) but not in untransformed cells (Fig. 2A, lanes 4). The absence of p16KdAori as a free extrachromosomal element was confirmed by the inability to transform E. coli with DNA extracted from the X. fastidiosa transformants. These results suggested that the plasmid or part of the plasmid had integrated into the bacterial chromosome.

View larger version (46K): [in this window] [in a new window]   FIG. 2.   (A) Southern blot hybridization between SphI- or BamHI-digested DNA extracted from X. fastidiosa transformants and the kan probe. Lanes 1 and 7, p16KdAori; lanes 2 and 6, extrachromosomal DNA extracted from transformant clones 1 (lanes 2) and 2 (lanes 6); lanes 3, 4, and 5, total DNA extracted from transformant clone 1 (lanes 3), from untransformed cells (lanes 4), and from transformant clone 2 (lanes 5). (B) Southern blot hybridization between EcoRI- or KpnI-digested DNA extracted from X. fastidiosa transformants and the ori probe. Lanes 1 and 2, total DNA extracted from untransformed cells of strains 9a5c (lanes 1) and J1a12 (lanes 2); lanes 3, p16KdAori; lanes 4, total DNA extracted from X. fastidiosa transformant clone 1. (C) Southern blot hybridization of EcoRI- or KpnI-digested DNA from untransformed and p16KdAori-transformed X. fastidiosa with the rop probe. Lanes 1, total DNA extracted from X. fastidiosa 9a5c; lanes 2, total DNA extracted from untransformed strain J1a12; lanes 3, p16KdAori; lanes 4, total DNA extracted from J1a12 transformant clone 1. Sizes are indicated in kilobase pairs.

Mapping the p16KdAori insertion site in X. fastidiosa transformants. To determine the position at which p16KdAori had integrated into the chromosome, genomic DNAs from three independent clones were digested with various enzymes and hybridized with the ori and the rop probes (Fig. 2B and C, respectively). Identical results were obtained with each of the three transformants. In Fig. 2B, the hybridization patterns of DNA extracted from the X. fastidiosa transformant are compared to those of the DNA extracted from untransformed cells from the same strain (J1a12). For comparison, DNA from strain 9a5c was also included. In the case of the transformant (Fig. 2B, lanes 4), the ori probe was found to hybridize with two EcoRI fragments, of 5.1 kbp and 4.2 kbp, or two KpnI fragments, of 7 kbp and 5.1 kbp. One fragment from each pair (5.1-kbp EcoRI fragment or 7-kbp KpnI fragment) comigrated with that detected in the untransformed cells (Fig. 2B, lanes 1 and 2), whereas the other comigrated with the fragment obtained from the purified p16KdAori. Having determined (from the experiment that produced Fig. 2A) that there was no free plasmid in the transformants, these results indicate that the plasmid was not integrated at the oriC region of the chromosome. To determine whether plasmid integration occurred by recombination at the rRNA promoter region, the restricted DNA was hybridized with the rop probe (Fig. 2C). As was expected from the occurrence of two rRNA operons in the X. fastidiosa genome (26), the probe hybridized with two DNA fragments in untransformed cells: two EcoRI fragments of 13.1 kbp and 2 kbp and two KpnI fragments of 6.9 kbp and 4.5 kbp (Fig. 2C, lanes 1 and 2). However, with the DNA extracted from transformed cells, three fragments hybridized with the probe regardless of the enzyme (EcoRI or KpnI) used to restrict the DNA (Fig. 2C, lanes 4). Interestingly, none of these three fragments is found in the purified plasmid (Fig. 2C, lanes 3). The 2-kbp EcoRI fragment or the 4.5-kbp KpnI fragment corresponds to one of the two fragments detected in the untransformed cells, whereas the other two fragments, with sizes of 12.4 kbp and 1.6 kbp for EcoRI and 7.3 kbp and 1.2 kbp for KpnI, differ in size from those in the untransformed cells. These data indicate that, in the transformants, the promoter region of one of the two rRNA operons is duplicated. This suggests that the p16KdAori plasmid has integrated into the chromosome by homologous recombination via a single crossover event, leading to the duplication of the insertion site sequences. Indeed, further Southern blot analyses showed that the restriction fragments hybridizing with the rop probe (Fig. 3A) had sizes that matched those predicted from the map in Fig. 3B, showing recombination between the rRNA promoter fragment carried by the plasmid and the promoter region of rRNA operon 1 of the chromosome. As shown in Fig. 3A, the 1.6-kbp EcoRI fragment, 1.2-kbp KpnI fragment, 7.3-kbp SmaI fragment, 6.2-kbp HindIII fragment, and 3.9-kbp XbaI fragment (lanes 3, 12, 18, 21, and 24, respectively) were not detected in either the purified plasmid (lanes 1, 10, 16, 19, and 22, respectively) or the untransformed cells (lanes 2, 11, 17, 20, and 23, respectively). This indicates that plasmid integration did occur in the promoter region of the rRNA operon 1. 

View larger version (32K): [in this window] [in a new window]   FIG. 3.   (A) Southern blot hybridization of DNA from untransformed and p16KdAori-transformed X. fastidiosa J1a12 (clone 1) with the rop probe. Genomic DNA from untransformed cells (lanes 2, 5, 8, 11, 14, 17, 20, and 23) and from cells transformed by p16KdAori (lanes 3, 6, 9, 12, 15, 18, 21, and 24), as well as purified p16KdAori (lanes 1, 4, 7, 10, 13, 16, 19, and 22), were digested with EcoRI (lanes 1 to 3), EcoRI plus BglII (lanes 4 to 6), HincII (lanes 7 to 9), KpnI (lanes 10 to 12), SmaI plus KpnI (lanes 13 to 15), SmaI (lanes 16 to 18), HindIII (lanes 19 to 21), and XbaI (lanes 22 to 24), respectively. Sizes are indicated in kilobase pairs. (B) Schematic representation of p16KdAori integration by recombination at the promoter region of rRNA operon 1 of X. fastidiosa. The 1.6-kbp and 12.4-kbp EcoRI fragments, 0.9-kbp EcoRI-BglII fragment, 1.4-kbp HincII fragment, 6.2-kbp HindIII fragment, 1.2-kbp and 6.7-kbp KpnI fragments, 7.3-kbp SmaI fragment, and 3.9-kbp XbaI fragment are indicated. The positions of primers M13U and RP3 are indicated by short thick arrows. Abbreviations: BI, BamHI; BII, BglII; E, EcoRI; HII, HincII; HIII, HindIII; K, KpnI; Sm, SmaI; Sp, SphI; X, XbaI. For other abbreviations, see the legend to Fig. 1.

Transformation of X. fastidiosa with plasmid p16Kori. The oriC fragment of plasmid p16KdAori comprises the dnaA gene as well as the 366-bp dnaA-dnaN intergenic region. This region contains five dnaA-box-like sequences and two AT-rich clusters and is thought to represent the genuine replication origin of the X. fastidiosa chromosome (26). To determine whether this region alone is able to function as an autonomous replicating sequence, we constructed plasmid p16Kori by inserting this DNA fragment into plasmid p16K (see Materials and Methods). Transformation of X. fastidiosa strain J1a12 with p16Kori did yield kanamycin-resistant colonies, suggesting that the 366-bp oriC fragment was able to promote plasmid replication in X. fastidiosa. Three colonies were picked and grown in the presence of kanamycin. After two passages in PW medium containing 5 µg of kanamycin/ml, total DNA as well as extrachromosomal DNA was prepared and analyzed by Southern blot hybridization with the ori and rop probes (Fig. 4A and B). As shown in Fig. 4, the 5.1-kbp EcoRI fragment hybridizing with the ori probe in the untransformed cells (Fig. 4A, lane 1) is still detected in the transformant (Fig. 4A, lane 4). In contrast, the 13.1-kbp EcoRI fragment hybridizing with the rop probe (Fig. 4B, lane 1) is not detected (Fig. 4B, lane 4). These results indicate that, as previously described for plasmid p16KdAori, p16Kori has also integrated in the rRNA promoter region. In agreement with this conclusion, the hybridization patterns of DNAs from p16Kori and p16KdAori transformants are identical (Fig. 4B, lanes 3 and 4). In addition, attempts to purify extrachromosomal DNA from the p16Kori transformant by the alkaline lysis procedure failed to reveal the presence of free plasmid (data not shown).

View larger version (66K): [in this window] [in a new window]   FIG. 4.   Southern blot hybridization between EcoRI-restricted DNA extracted from X. fastidiosa J1a12 transformed with plasmids p16KdAori or p16Kori and the ori (A) or rop (B) probe. Lanes 1, total DNA from untransformed cells; lanes 2, plasmid p16KdAori; lanes 3 and 4, total DNAs from cells transformed with p16KdAori (lanes 3) or with p16Kori (lanes 4); lanes 5, plasmid p16Kori. Sizes are indicated in kilobase pairs.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Originally, oriC plasmids were developed to study the initiation mechanism of chromosome replication by isolating chromosomal fragments with autonomous replication (ars) activity (12, 15). In our studies, transformation of CVC strains of X. fastidiosa to kanamycin resistance by electroporation was obtained with plasmids containing all (p16KdAori) or part (p16Kori) of the X. fastidiosa oriC region, but not with plasmid p16K, which lacks the oriC sequences. These results suggest that the two plasmids do possess ars activity even though, in the experiments reported here, we failed to detect the plasmids as free extrachromosomal DNA at the time when DNA was harvested from the transformed cells for analysis. To get enough material, the cells were harvested after two passages, which represented about 40 generations. We propose that during this period the oriC plasmids do replicate before integration. In agreement with this assumption, we have detected faint bands in some Southern blot experiments that could represent free plasmid DNA. Also, the oriC fragment of p16Kori, i.e., the 366-bp intergenic region downstream of the dnaA gene, has features in common with the oriC regions of other gram-negative bacteria, in particular five dnaA boxes and two AT-rich sequences. The requirement of these elements for ars activity has not been tested so far. However, our results represent the first experimental indication that sequences in the intergenic region downstream of the dnaA gene do possess ars activity.

In gram-negative bacteria, oriC plasmids are usually maintained extrachromosomally. In E. coli, incompatibility between oriC plasmids and the chromosome is seen only under very special conditions, such as reduced activity of DnaA protein (for a review, see reference 12). In contrast to E. coli, the gram-positive bacterium Bacillus subtilis strictly regulates the number of oriC copies within a cell (16). In this organism, oriC plasmids cannot be maintained and have a tendency to integrate into the chromosome. Surprisingly, in X. fastidiosa transformants, we found that plasmids p16Kori and p16KdAori were not maintained as free extrachromosomal elements but, instead, had integrated into the host chromosome. As indicated above, no free plasmid was detected in the transformed cells after two subcultures in liquid medium. These results suggest that, in X. fastidiosa, DnaA expression could be more tightly regulated than in E. coli and that therefore the limited level of DnaA expression does not allow extrachromosomal replication of oriC plasmids. Interestingly, in all 17 transformants tested, plasmid integration was found to occur by homologous recombination involving a single crossover event between the rRNA promoter region carried on the plasmid and the homologous sequences present in the chromosome. No recombination at oriC was observed, in contrast to the situation found to occur in the plant pathogen S. citri (22). In this case, the oriC plasmid pBOT1 integrates into the host chromosome by homologous recombination at the oriC region. The reason for which, in X. fastidiosa, plasmid integration preferentially occurred at the rRNA promoter region rather than at oriC is not known.

Five CVC strains and two coffee strains of X. fastidiosa were subjected to transformation assays, but only two CVC strains, J1a12 and B111, could be transformed. Thus, transformation of X. fastidiosa seems to be strain dependent for reasons that have to be investigated further. In the early stages of this work, attempts to transform X. fastidiosa 9a5c with several conjugative or suicide plasmids were unsuccessful. In these assays we used only CVC strain 9a5c. The experience gained with the oriC plasmids suggests that strains other than 9a5c could perhaps be transformed with some of the plasmids described above.

Transformation of CVC X. fastidiosa strain LAR20 has recently been reported (21). However, the shuttle plasmid pEcoR#10 used in these studies proved to be unstable. In contrast, in our case, we have determined that p16KdAori and p16Kori, once integrated into the chromosome, are stably maintained regardless of the presence or absence of kanamycin as the selection pressure.

X. fastidiosa is the first plant-pathogenic bacterium whose genome sequence has been determined (26). Although several putative virulence genes are present in X. fastidiosa, very little is known experimentally about the pathogenicity mechanisms compared with those of other gram-negative bacteria, such as Ralstonia solanacearum or Xanthomonas campestris, for which the availability of gene transfer systems was a key factor in the development of genetic studies. In the plant-pathogenic mollicute S. citri, an organism related to low-guanosine-plus-cytosine gram-positive bacteria, oriC plasmids were used as vectors for both expression of cloned genes in spiroplasma cells (11, 22) and gene inactivation through homologous recombination (8, 10). The stable transformation of CVC strains of X. fastidiosa by oriC plasmids represents the first step toward functional genomic studies. It opens the way for the expression of cloned genes in X. fastidiosa. In addition, it can be expected that the integrative property of X. fastidiosa oriC plasmids will contribute to the construction of mutants by specific gene inactivation through homologous recombination. In this respect, the experimental infection of the herbaceous plant Catharanthus roseus by X. fastidiosa 9a5c that we have recently described (14a) should help us to evaluate the pathogenicity of the mutants.