Conjugative Plasmid Transfer in Xylella fastidiosa Is Dependent on tra and trb Operon Functions

ABSTRACT

The insect-transmitted plant pathogen Xylella fastidiosa is capable of efficient horizontal gene transfer (HGT) and recombination. Natural transformation occurs at high rates in X. fastidiosa, but there also is evidence that certain strains of X. fastidiosa carry native plasmids equipped with transfer and mobilization genes, suggesting conjugation as an additional mechanism of HGT in some instances. Two operons, tra and trb, putatively encoding a conjugative type IV secretion system, are found in some but not all X. fastidiosa isolates, often on native plasmids. X. fastidiosa strains that carry the conjugative transfer genes can belong to different subspecies and frequently differ in host ranges. Using X. fastidiosa strain M23 (X. fastidiosa subsp. fastidiosa) or Dixon (X. fastidiosa subsp. multiplex) as the donor strain and Temecula (X. fastidiosa subsp. fastidiosa) as the recipient strain, plasmid transfer was characterized using the mobilizable broad-host-range vector pBBR5pemIK. Transfer of plasmid pBBR5pemIK was observed under in vitro conditions with both donor strains and was dependent on both tra and trb operon functions. A conjugative mechanism likely contributes to gene transfer between diverse strains of X. fastidiosa, possibly facilitating adaptation to new environments or different hosts.

IMPORTANCE Xylella fastidiosa is an important plant pathogen worldwide, infecting a wide range of different plant species. The emergence of new diseases caused by X. fastidiosa, or host switching of existing strains, is thought to be primarily due to the high frequency of HGT and recombination in this pathogen. Transfer of plasmids by a conjugative mechanism enables movement of larger amounts of genetic material at one time, compared with other routes of gene transfer such as natural transformation. Establishing the prevalence and functionality of this mechanism in X. fastidiosa contributes to a better understanding of HGT, adaptation, and disease emergence in this diverse pathogen.

INTRODUCTION

Xylella fastidiosa is a diverse plant pathogen affecting a wide range of host plants worldwide (1). As a fastidious organism, X. fastidiosa can effectively colonize only xylem vessels of host plants and the foregut of insect vectors (2). X. fastidiosa is broadly categorized into multiple subspecies, which often have distinct host ranges. The emergence of new diseases caused by X. fastidiosa is thought to be, in part, a result of frequent horizontal gene transfer (HGT) and recombination events between related strains and in some cases between subspecies (3–5). A significant amount of this HGT is likely due to natural transformation and has been extensively characterized under a variety of in vitro conditions (6–9). X. fastidiosa can acquire DNA fragments of up to 3.5 kb by natural transformation, with higher transformation and recombination frequencies on solid medium or in microfluidic chambers designed to mimic natural habitats (6, 7, 9). However, genomic evidence suggests that, in certain instances, recombination between X. fastidiosa strains has occurred on a much larger scale, leading to the emergence of highly recombinant strains with the ability to infect new hosts (3).

The presence of large plasmids carrying putative type IV secretion systems (T4SSs) in different X. fastidiosa strains suggests that conjugation may be another mechanism of HGT in this pathogen (10). Two X. fastidiosa strains, i.e., M23 (X. fastidiosa subsp. fastidiosa), which causes Pierce's disease of grapes, and Riv5 (X. fastidiosa subsp. multiplex), which was isolated from ornamental plum, carry nearly identical 38-kb plasmids (pXFAS01 and pXF-Riv5) containing two operons, tra and trb, that encode all of the necessary functions for a conjugative T4SS (10). The trb homologs in X. fastidiosa are related to mating pair formation genes present in Yersinia pseudotuberculosis strain IP31758 (10). The tra module in X. fastidiosa appears to be derived from a separate lineage and encodes DNA transfer functions. In addition to the large plasmids present in strains of X. fastidiosa subsp. fastidiosa and X. fastidiosa subsp. multiplex, these conjugative transfer genes are present in some strains of X. fastidiosa subsp. pauca, a pathogen of citrus (11). Although it is logical that these plasmids can be transferred between the different strains and subspecies of X. fastidiosa, it has not been shown that the tra and trb operon functions are intact and that conjugative gene transfer does in fact occur in this pathogen. Using a mobilizable vector, pBBR5pemIK, we show that plasmid transfer takes place between different strains and subspecies of X. fastidiosa under in vitro growth conditions, providing an alternate mechanism of HGT that may contribute to the existence of recombinant strains.

RESULTS

tra and trb operons are present in at least four subspecies of X. fastidiosa.

The prevalence of tra and trb operons in X. fastidiosa was determined by BLASTN searches using the traG and trbD nucleotide sequences from M23 as queries (Table 1). Both traG and trbD were found in several strains of X. fastidiosa subsp. fastidiosa, X. fastidiosa subsp. multiplex, and X. fastidiosa subsp. pauca. Additionally, trbD but not traG was found in X. fastidiosa subsp. sandyii strain Ann-1. Although these gene sequences are located on plasmids in some strains, there are several instances in which one or both homologs are encoded by chromosomal sequences. These conjugative transfer genes also do not appear to be universally present in all strains of X. fastidiosa, suggesting that conjugative transfer is not an essential function in this species, although it may contribute to genetic variation and recombination.

TABLE 1.

Distribution of tra and trb operon genes in Xylella fastidiosa

X. fastidiosa strain	X. fastidiosa subspecies	Host(s)	Conjugative transfer operon(s)	Gene locationa	Reference
Temecula	fastidiosa	Grape	None	NA	33
Stag's Leap	fastidiosa	Grape	trb		39
M23	fastidiosa	Grape, almond	tra, trb	Plasmid	32
GB514	fastidiosa	Grape	None	NA	40
CoDiRO	fastidiosa	Olive	tra, trb	Plasmid	41
Dixon	multiplex	Almond	tra, trb	Chromosome	19
M12	multiplex	Almond	None	NA	32
Riv5	multiplex	Plum	tra, trb	Plasmid	GenBank accession no. NC_020121.1
Sy-VA	multiplex	Sycamore	None	NA	42
MUL0034	fastidiosa	Mulberry	trb	Chromosome	3
Griffin-1	multiplex	Oak	None	NA	43
Ann-1	sandyi	Oleander	trb	Chromosome	44
PLS229	taiwanensis	Pear	None	NA	45
BB01	multiplex	Blueberry	tra, trb	Chromosome	46
9a5c	pauca	Citrus	tra, trb	Chromosome	33
6c	pauca	Coffee	None	NA	47
COF0407	pauca	Coffee	tra, trb	Plasmid	GenBank accession no. CM003764.1

^aNA, not applicable.

The functions of the tra and trb genes are essential for transfer of a mobilizable plasmid within X. fastidiosa subsp. fastidiosa.

Conjugation reactions were set up using different strains of X. fastidiosa subsp. fastidiosa as donor and recipient. Donor strains M23 (containing tra and trb) and Temecula (lacking tra and trb) were transformed with the broad-host-range mobilizable vector pBBR5pemIK (12). The recipient strain was a Temecula strain modified to carry an antibiotic resistance marker (chloramphenicol resistance [Cm^r]) inserted in the chromosome for selection. After coculture of the recipient and donor strains on solid medium for 5 days, cells were subjected to double antibiotic selection utilizing the resistance marker on the recipient strain chromosome (Cm^r) and an antibiotic marker on plasmid pBBR5pemIK (gentamicin resistance [Gm^r]). Successful transformants were obtained from the M23 donor strain but not from the Temecula donor strain (Table 2). Putative transconjugants were then confirmed by PCRs targeting plasmid and chromosomal sequences (Fig. 1).

TABLE 2.

Donor and recipient strains for conjugative transformation

Donor straina	Recipient straina	Plasmid	Transformantsb
Temecula	Temecula Cmr	pBBR5pemIK	−
M23	Temecula Cmr	pBBR5pemIK	+
M23	Temecula Cmr	pBBR5pemIKmob−	−
M23 ΔtraG	Temecula Cmr	pBBR5pemIK	−
M23 ΔtraM	Temecula Cmr	pBBR5pemIK	+
M23 ΔtrbD	Temecula Cmr	pBBR5pemIK	−
M23 ΔtraG/traG+	Temecula Cmr	pBBR5pemIK	+
M23 ΔtraM/traM+	Temecula Cmr	pBBR5pemIK	+
M23 ΔtrbD/trbD+	Temecula Cmr	pBBR5pemIK	+
Dixon	Temecula Cmr	pBBR5pemIK	+
E. coli S17-1λ	Temecula Cmr	pBBR5pemIKpBAD-PD1607	−
M23	E. coli Top10	pBBR5pemIK	+

^aAll strains are X. fastidiosa unless otherwise noted.

^bPresence or absence of confirmed transformants. Plus signs indicate that transformants were detected at a rate of 20 to 100 transformants per 10⁴ CFU of recipient cells. Minus signs indicate that no transformants were detected. Twelve replicates were performed for each donor-recipient pair.

FIG 1.

Confirmation of plasmid transfer from M23 to Temecula of X. fastidiosa subsp. fastidiosa. Following coculture of donor strain X. fastidiosa M23 carrying plasmid pBBR5pemIK and recipient strain X. fastidiosa Temecula Cm^r on PD3 plates for 5 days, transformants were selected on PD3 plates supplemented with chloramphenicol (recipient marker) and gentamicin (pBBR5pemIK). Transformants were subcultured and grown individually for DNA extraction and PCR confirmation of plasmid transfer. PCR primers targeted the recipient marker gene locus (3 kb) (A) and pemI-pemK genes located on plasmid pBBR5pemIK (∼750 bp) (B). Lanes (both panels): 1, size markers; 2 to 6, selected transformants; 7, donor strain M23 pBBR5pemIK; 8, recipient strain Temecula Cm^r; 9, PCR negative control.

A modified version of plasmid pBBR5pemIK was constructed by removing the mobilization genes from the plasmid backbone (pBBR5pemIKmob−) (Fig. 2). Plasmid pBBR5pemIKmob− is still able to replicate and to transform X. fastidiosa by other mechanisms, but it lacks the functions necessary to be transferred via a conjugative mechanism. This plasmid was transformed into M23 to test further whether the mechanism of transfer is conjugation rather than natural transformation. Plasmid pBBR5pemIKmob− was not transferred between X. fastidiosa strains during coculture (Table 2), confirming that the route of gene transfer in this case is likely conjugation.

FIG 2.

Modification of pBBR5pemIK to remove mobilization functions. (A) Broad-host-range plasmid pBBR5pemIK (12). (B) Modified version of pBBR5pemIK with mob genes removed by deletion mutagenesis. pBBR5pemIKmob− retained broad-host-range replication (rep), gentamicin resistance (Gm^r), and plasmid stability (pemI-pemK) loci. Maps were made to scale using Vector NTI Advance software (Invitrogen).

M23 knockout mutants in traG, traM, and trbD were created to test whether the observed plasmid transfer depends on the T4SS encoded on M23 plasmid pXFAS01 (10) (Fig. 3). Loss of traG and trbD, which are essential for the function of the T4SS, resulted in no transformants, whereas loss of traM, which encodes a transcriptional activator, did not abolish conjugative transformation. Complemented ΔtraG/traG⁺, ΔtraM/traM⁺, and ΔtrbD/trbD⁺ strains were all capable of producing transformants (Table 2).

FIG 3.

Construction of conjugative transfer deletion mutants in X. fastidiosa strain M23. In X. fastidiosa subsp. fastidiosa M23, the conjugative type IV secretion system is encoded on native plasmid pXFAS01 (10). Light green arrows represent replication and stability factors, dark green arrows represent tra operon genes, blue arrows represent trb operon genes, and pink arrows indicate genes chosen for site-directed mutagenesis. The traM (transcriptional activator), traG (conjugative transfer coupling protein), and trbD (conjugative transfer ATPase) genes were individually deleted by homologous recombination, and the respective open reading frames were replaced with a selective marker for kanamycin resistance. The map was created to scale using Vector NTI Advance software (Invitrogen).

Conjugative plasmid transfer can occur between different subspecies of X. fastidiosa.

X. fastidiosa strain Dixon (X. fastidiosa subsp. multiplex) carrying pBBR5pemIK was used as a donor strain in place of M23 and produced successful transformants of a Temecula recipient strain. These transformants were screened using PCR primers designed to differentiate X. fastidiosa subsp. fastidiosa and X. fastidiosa subsp. multiplex (Fig. 4). Intersubspecies conjugation produced transformants at a rate of 20 to 50 transformants per 10⁴ CFU of recipient cells, on average (n = 12). Plasmids extracted from Temecula transconjugants were consistent with intact pBBR5pemIK, based on restriction digestion (Fig. 5A), indicating plasmid transfer rather than recombination. Because of genetic differences between the Dixon and M23 strains, it was possible to confirm the identity of the transformants using multilocus sequence typing (MLST). All transformants screened were identical to the Temecula recipient strain at the four loci sequenced (see Fig. S1 in the supplemental material).

FIG 4.

Confirmation of transformants from X. fastidiosa subsp. multiplex Dixon and X. fastidiosa subsp. fastidiosa Temecula conjugative plasmid transfer. Following coculture of donor strain X. fastidiosa subsp. multiplex carrying plasmid pBBR5pemIK and recipient strain X. fastidiosa Temecula Cm^r on PD3 plates for 5 days, transformants were selected on PD3 plates supplemented with chloramphenicol (recipient marker) and gentamicin (pBBR5pemIK). Transformants were subcultured and grown individually for DNA extraction and PCR confirmation of transformant identity. PCR primers were specific for the X. fastidiosa subsp. fastidiosa recipient strain (412-bp product) (A) and the X. fastidiosa subsp. multiplex donor strain (638-bp product) (B) (49). Lanes (both panels): leftmost, size markers; 1 to 7, selected transformants; 8, donor strain Dixon pBBR5pemIK; 9, recipient strain Temecula Cm^r; 10, PCR negative control.

FIG 5.

Plasmid confirmation from transconjugants. Plasmids extracted from X. fastidiosa transconjugants (A) or E. coli transconjugants (B) were digested with restriction endonuclease EcoRV, resulting in bands at 4.3 kb and 1.2 kb. Lanes (both panels): 1 to 5, plasmids extracted from transconjugants; 6, pBBR5pemIK.

X. fastidiosa can serve as a donor in interspecies plasmid transfer.

The ability of X. fastidiosa to transfer plasmids to other bacterial species was tested by using M23 carrying pBBR5pemIK as a donor strain and Escherichia coli Top10 as a recipient strain. E. coli transformants were selected on Luria-Bertani (LB) medium because it does not support growth of X. fastidiosa, and a number of successful transconjugants were produced. Plasmids isolated from E. coli transformants were consistent with intact pBBR5pemIK, based on restriction enzyme digestion (Fig. 5B). This suggests that X. fastidiosa strains that possess the conjugative T4SS can readily transfer plasmids to unrelated bacterial species.

The capacity of X. fastidiosa to take up plasmids via a conjugative mechanism from E. coli was tested using E. coli S17-1λ as a donor strain and the marked Temecula strain (Cm^r) as a recipient strain. In this case, plasmid pBBR5pemIKpBAD (arabinose-inducible promoter) (Fig. S2) encoding an X. fastidiosa-specific methyltransferase (PD1607) (13) was used for transfer. In spite of this plasmid modification, which was designed to circumvent an X. fastidiosa restriction-modification (R-M) system (13), no positive transformants were identified after double antibiotic selection for the recipient strain marker and the plasmid. Plasmid pBBR5pemIK could be transferred from E. coli S17-1λ to E. coli Top10, however (data not shown), further suggesting that the barrier to transformation lies with the X. fastidiosa recipient strain.

DISCUSSION

Host switching of X. fastidiosa strains and the emergence of new diseases due to this pathogen are thought to be in part a result of frequent HGT. Here, we demonstrate the ability of X. fastidiosa to transfer plasmids via a conjugation mechanism dependent on functions of the tra- and trb-encoded T4SS. Strains of X. fastidiosa that carry the conjugative T4SS were able to transfer the broad-host-range plasmid pBBR5pemIK to other strains of X. fastidiosa belonging to different subspecies, as well as to unrelated bacteria (E. coli). Because natural transformation occurs at higher rates under conditions designed to mimic the in vivo growth environment of X. fastidiosa (microfluidic chambers and grapevine xylem sap) (2), it likely has a significant influence on HGT and recombination during the life cycle of this pathogen. Conjugative plasmid transfer, which can mediate transfer of DNA segments as large as the plasmids themselves, is not reliant on immediate integration for stability of the transferred DNA and therefore has the potential to move more genetic material in one transfer event than natural transformation. Recombination events of up to ∼3.5 kb in length that are associated with natural transformation in X. fastidiosa have been identified (5). Classic examples of conjugative plasmid transfer, such as the E. coli F-like plasmids, involve transfer of up to 100 kb (14), and the conjugative plasmids identified in X. fastidiosa are up to 38 kb (10). It remains to be seen whether transfer of large native plasmids, such as pXFAS01 found in X. fastidiosa M23, can occur at an efficient rate or whether complete transfer occurs only with smaller, easily mobilizable plasmids, such as pBBR5pemIK. Further investigation of the environmental conditions that affect conjugation efficiency will determine whether conjugation is likely to be a frequent occurrence in vivo, similar to findings observed for natural transformation.

A wide variety of bacterial species undergo conjugative plasmid transfer in natural environments, including soil, water, and mammalian or invertebrate tissues (15–18). The biologically relevant scenarios in which plasmid transfer might occur between different strains of X. fastidiosa would be in the plant hosts or the insect vectors. In some cases, X. fastidiosa strains belonging to different subspecies have been isolated from the same host plants growing in the same location, although it is unclear whether this is a common occurrence (19). Although little is known regarding conjugative plasmid transfer during bacterial colonization of xylem tissue specifically, HGT has been shown to occur in a number of different plant-associated microbial communities (20–22). Ralstonia solanacearum can acquire DNA during infection of tomato plants (22), and a number of Pseudomonas species are able to transfer plasmids by conjugation while colonizing the rhizosphere and phyllosphere environments (20, 21). The distribution of X. fastidiosa within the xylem vessels is often sporadic, with many vessels containing low-cell-density bacterial populations (23). Because of this, it may not be likely that different strains would be present at high enough cell densities for conjugative transfer to occur in the plant. However, heavily infected, symptomatic vines do contain large aggregates of X. fastidiosa and it is possible that, if multiple strains were present simultaneously, then plasmid transfer could occur.

Vectors of X. fastidiosa feed on multiple host plants and could conceivably acquire several different X. fastidiosa strains concurrently (19, 24). Studies of gene transfer in other invertebrate-associated microbial communities have shown insect colonization to be an environment conducive to the process. Conjugation between E. coli and indigenous bacterial species was detected in the gut of the microarthropod Folsomia candida (18). A high frequency of plasmid transfer in Yersinia pestis in the flea midgut is thought to mediate the spread of antibiotic resistance (25). Because X. fastidiosa colonizes insect vector mouthparts through the formation of biofilms, it is very possible that X. fastidiosa strains possessing conjugative transfer functions would have the opportunity to transfer native plasmids to other strains of X. fastidiosa or to other bacterial species in this environment.

Interestingly, there are several X. fastidiosa strains that have one or both of the tra and trb operons encoded on the chromosome, likely the result of an insertion event following the acquisition of partial or intact plasmid sequences. In the case of X. fastidiosa strain Dixon (X. fastidiosa subsp. multiplex), it is apparent that the conjugative T4SS encoded on the chromosome is fully functional. Other strains, such as Ann-1 (X. fastidiosa subsp. sandyi), have homologs to the trb operon but not the tra genes, suggesting partial recombination of the T4SS with the chromosome and a potential loss of conjugative transfer functions. Because the tra and trb functions do not appear to be essential in X. fastidiosa, it is unclear what, if any, selective advantage is provided by the maintenance of a conjugative T4SS. The native plasmids that carry the tra and trb operons do not appear to convey any specific virulence genes, although the ability to undergo HGT via conjugation may increase the chance of acquiring new genetic material that is beneficial to adaptation. In nonheterogeneous E. coli populations, transfer of natural conjugative plasmids has been shown to stimulate the production of biofilms (26, 27), a phenotype essential to the X. fastidiosa life cycle. Multiple studies of plasmid evolution have demonstrated that, in general, conjugative plasmids convey a fitness advantage to bacteria, even in the absence of any direct selective pressure such as antibiotics (28, 29).

At least four different R-M systems have been identified in X. fastidiosa subsp. fastidiosa. Transformation efficiency in X. fastidiosa can be greatly improved by inhibiting type I restriction systems (30) or altering DNA methylation to prevent degradation by a species-specific type II restriction system (13). The prevalence of R-M systems in X. fastidiosa is likely prohibitive to incorporation of foreign DNA that may be acquired by conjugation (13, 30). We were unable to demonstrate acquisition of pBBR5pemIK by X. fastidiosa from an E. coli donor strain despite modification of the plasmid by expression of a X. fastidiosa-derived methyltransferase (13). Due to the vastly different growth rates of the two bacteria and the difficulty of excluding the E. coli donor strain from transformant selection, it is possible that plasmid transfer does occur in this case but not at detectable rates in vitro. Alternatively, although X. fastidiosa appears to readily transfer plasmids to other X. fastidiosa strains and E. coli, conjugation may facilitate plasmid uptake and recombination only from closely related organisms. This alone, however, has implications for intersubspecies recombination, which is thought to play a role in the adaptation of X. fastidiosa to new host plants.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

All bacterial strains used in this study are listed in Table 3. Escherichia coli strain Top10 (Thermo Fisher Scientific) was used as a cloning host. E. coli strain S17-1λ (31) was used as a donor strain for interspecies conjugation. E. coli was grown in LB medium at 37°C unless otherwise indicated, with the addition of antibiotics when necessary. For E. coli, antibiotics were used at the following concentrations: kanamycin, 30 μg/ml; spectinomycin, 100 μg/ml; chloramphenicol, 35 μg/ml; gentamicin, 10 μg/ml. X. fastidiosa M23, Temecula, and Dixon (19, 32, 33) and mutant derivatives were grown in PD3 medium (34) at 28°C. For X. fastidiosa strains, PD3 medium was supplemented with 5 μg/ml gentamicin, kanamycin, or chloramphenicol when necessary.

TABLE 3.

Bacterial strains and plasmids used in this study

Strain or plasmid	Characteristicsa	Source or reference
Strains
Xylella fastidiosa subsp. fastidiosa Temecula	Wild-type X. fastidiosa; grape pathogen	33
Xylella fastidiosa subsp. fastidiosa M23	Wild-type X. fastidiosa; grape and almond pathogen	19
Xylella fastidiosa subsp. multiplex Dixon	Wild-type X. fastidiosa; almond pathogen	11
Temecula Cmr	Temecula with chromosomal insertion of antibiotic resistance marker Cmr	This study
M23 ΔtraM	M23 knockout mutant in traM; Kanr	This study
M23 ΔtraM/traM+	ΔtraM complemented using plasmid pBBR5pemIK; Kanr Gmr	This study
M23 ΔtraM pBBR5pemIK	ΔtraM carrying plasmid pBBR5pemIK; Kanr Gmr	This study
M23 ΔtraG	M23 knockout mutant in traG; Kanr	This study
M23 ΔtraG/traG+	ΔtraG complemented using plasmid pBBR5pemIK; Kanr Gmr	This study
M23 ΔtrbD	M23 knockout mutant in trbD; Kanr	This study
M23 ΔtrbD/trbD+	ΔtraG complemented using plasmid pBBR5pemIK; Kanr Gmr	This study
M23 ΔtrbD pBBR5pemIK	ΔtraM carrying plasmid pBBR5pemIK; Kanr Gmr	This study
Escherichia coli Top10	Cloning host; F− mcrA Δ(mrr-hsdRMS-mcrBC) ϕ80lacZΔM15 ΔlacX74 recA1 araD139 Δ(ara-leu)7697 galU galK rpsL (Strr) endA1 nupG	Thermo Fisher Scientific
E. coli Eam1	DH5α derivative containing IPTG-inducible X. fastidiosa methylase gene (PD1607); Strr Spr	13
E. coli S17-1λ	RP4 Mob+ Smr	31
Plasmids
pCR8/GW/TOPO	TA cloning vector with att sites for Gateway cloning; Spr	Thermo Fisher Scientific
pBBR5pemIK	Broad-host-range vector containing pemI/pemK stability factor; Gmr	48
pBBR5pemIKmob−	pBBR5pemIK modified to remove mobilization functions	This study
pCR8-traM-kan	Mutagenesis construct for traM; Spr Kanr	This study
pCR8-traG-kan	Mutagenesis construct for traG; Spr Kanr	This study
pCR8-trbD-kan	Mutagenesis construct for trbD; Spr Kanr	This study
pBBR5pemIK-traM	Complementation construct for ΔtraM	This study
pBBR5pemIK-traG	Complementation construct for ΔtraG	This study
pBBR5pemIK-trbD	Complementation construct for ΔtrbD	This study
pAX1-Cm	Insertion vector targeting a neutral site on the chromosome; Cmr	37
pBBR5pemIKpBAD	Broad-host-range vector with arabinose-inducible promoter; Gmr	This study

^aCm^r, chloramphenicol resistance; Kan^r, kanamycin resistance; Gm^r, gentamicin resistance; Sp^r, spectinomycin resistance; Str^r, streptomycin resistance; IPTG, isopropyl-β-d-thiogalactopyranoside.

Identification of tra and trb homologs.

Whole-genome sequences of the X. fastidiosa strains in Table 1 were downloaded from the NCBI database. BLAST databases were made for each genome using the makeblastdb command. Additionally, 33 X. fastidiosa plasmid sequences were downloaded from the NCBI database and compiled into a single BLAST database using the makeblastdb command. Using standalone BLAST (35), the traG, traM, and trbD nucleotide sequences were used as queries for BLASTN searches against the X. fastidiosa plasmid database, as well as against each individual X. fastidiosa strain genome database. The BLAST parameters used were as follows: word_size, 11; expect threshold, 10; gap open existence, 0; gap open extension, 2.5; match and mismatch, 1 and −2.

Plasmid and strain construction.

All PCR primers used in this study are listed in Table 4. Plasmid pBBR5pemIKmob− was created by deleting the mob genes from pBBR5pemIK (12), using the Q5 site-directed mutagenesis kit (New England BioLabs). Briefly, the primers mobdelfwd and mobdelrev were used to PCR amplify pBBR5pemIK, excluding the mob open reading frame. This PCR product (1 μl) was added to a ligation reaction with kinase-ligase-DpnI (KLD) mixture for 5 min at room temperature and was transformed into chemically competent E. coli DH5α cells. Plasmid deletion mutants were screened by PCR using the primers mobfwd and mobrev, to confirm removal of the mob open reading frame. Plasmids pBBR5pemIk and pBBR5pemIKmob− were then transformed into X. fastidiosa M23 and Dixon wild-type strains and M23 mutant derivatives ΔtraM, ΔtraG, and ΔtrbD, by electroporation (36).

TABLE 4.

PCR primers used in this study

Primer name	Sequence	Source or reference
mobdelfwd	CGCCTGCCCACGAGCTTGAC	This study
mobdelrev	ACCTGCGGCGTTGTGACAAT	This study
mobfwd	TCAATTTTTTTAATTTTCTCTGGGG	This study
mobrev	ACGAATTGTTAGGTGGCG	This study
NS1-fwd	GTCAGCAGTTGCGTCAGATG	37
NS1-rev	AAAGCTGCCGACGCCAAATC	37
KanFwd	ATCGATGAATTGTGTCTCAAAATCT	This study
KanRev	TGCAGGTCGACTCTAGAGGAT	This study
traGupfwd	ACCGATACGGAATCTTGTGCC	This study
traGuprevkan	CCATGTAAGCCCACTGCAAGACCGCATGATTAATCGTGCC	This study
traGdnfwdkan	TCTTGACGAGTTCTTCTGAATCTGTTTCTTCCTTTCGTCTT	This study
traGdnrev	GCGTAGAAGCTGCATTGCG	This study
traMupfwd	CGACTGGCCTCTCAAGAACA	This study
traMuprevkan	CCATGTAAGCCCACTGCAAGCAGAGCACCGTTGCGTTGTC	This study
traMdnfwdkan	TCTTGACGAGTTCTTCTGATTCGGTGAGTCTTATCTTGGCT	This study
traMdnrev	GTTCAGAACGAGCAAACGCC	This study
trbDupfwd	CCCTTGAACAATACGTTGAGGC	This study
trbDuprevkan	CCATGTAAGCCCACTGCAAGTACGCGAGCCTGGCTTGTA	This study
trbDdnfwdkan	TCTTGACGAGTTCTTCTGATCACCATTGCAATCGCATTG	This study
trbDdnrev	TGGTCGAAGAAACCACGCAT	This study
traM-ORF-fwd	ATTGGGGACCTGTCGAACAC	This study
traM-ORF-rev	GCCTGTACACTCGCCTTGAG	This study
traG-ORF-fwd	GAAACCGTGTCGCCTTTAGC	This study
traG-ORF-rev	TCTGAACTCATGGCGGTTGA	This study
trbD-ORF-fwd	TCAAACACGAGTGGCGGTG	This study
trbD-ORF-rev	GCATTGTCAAATGATGCAAGCG	This study
XF1968fwd	GGAGGTTTACCGAAGACAGAT	49
XF1968rev	ATCCACAGTAAAACCACATGC	49
XF2542fwd	TTGATCGAGCTGATGATCG	49
XF2542rev	CAGTACAGCCTGCTGGAGTTA	49
petCfwd	CTGCCATTCGTTGAAGTACCT	38
petCrev	CGTCCTCCCAATAAGCCT	38
nuoN	GGGTTAAACATTGCCGATCT	38
nuoN	CGGGTTCCAAAGGATTCCTAA	38
holC fwd	GATTTCCAAACCGCGCTTTC	38
holC rev	TCATGTGCAGGCCGCGTCTCT	38
pilU	CAATGAAGATTCACGGCAATA	3
pilU	ATAGTTAATGGCTCCGCTATG	3
Cmfwd	AACGGATCCTATCGTCAATTATTACCTCCA	This study
Cmrev	ACGGATCCTTTTCGACCGAATAAATACC	This study
pemIKfwd	ATTCACTAGTGATTGGTTGCAGTGG	This study
pemIKrev	CGATTTTCGCTATAAGCGTATAAATG	This study
RST31	GCGTTAATTTTCGAAGTGATTCGATTGC	50
RST33	CACCATTCGTATCCCGGTG	50

The ΔtraM, ΔtraG, and ΔtrbD strains in the M23 strain background were created by homologous recombination using marker replacement. DNA fragments (1 kb) upstream and downstream of the open reading frames (traM, traG, and trbD) were PCR amplified using the primer sets traMupfwd/traMuprevkan and traMdnfwdkan/traMdnrev, traGupfwd/traGuprevkan and traGdnfwdkan/traGdnrev, and trbDupfwd/trbDuprevkan and trbDdnfwdkan/trbDdnrev, respectively. These primer sets include a 15-bp sequence that overlaps with the selectable marker used for gene replacement, a kanamycin resistance marker from the EZ-Tn5 transposon mutagenesis kit (Epicentre). The kanamycin marker was amplified using the primers kanfwd and kanrev. Three PCR products for each mutagenesis construct (upstream and downstream flanking sequences and the kanamycin marker) were annealed using overlap extension PCR and cloned into plasmid pCR8/GW/TOPO, following the manufacturer's instructions, to create plasmids pCR8-traM-kan, pCR8-traG-kan, and pCR8-trbD-kan. Constructs were confirmed by sequencing and were transformed into X. fastidiosa M23 by electroporation for homologous recombination. Successful X. fastidiosa transformants were selected on PD3 plates containing kanamycin and were confirmed by PCR amplification and sequencing of the region of recombination.

Strain Temecula Cm^r was created by inserting a selectable marker (Cm^r) into a neutral location on the chromosome (37). X. fastidiosa strain Temecula was transformed with plasmid pAX1-Cm by electroporation. Transformants were selected on PD3 plates containing chloramphenicol and were screened for correct marker insertion by PCR using the primers NS1fwd and NS1rev. All plasmids were propagated in E. coli strain Eam1 (13) prior to electroporation into X. fastidiosa, to improve transformation efficiency.

Plasmid pBBR5pemIKpBAD was created by inserting the arabinose-inducible promoter element araBAD in frame with the cloning site of pBBR5pemIK-GW (12). X. fastidiosa methyltransferase PD1607 (13) was then cloned into the Gateway recombination site of pBBR5pemIKpBAD to create pBBR5pemIKpBAD-PD1607.

X. fastidiosa conjugation.

Donor and recipient strains were grown on PD3 plates for 5 days at 28°C and then harvested with PD3 liquid medium. Cell concentrations were adjusted to an optical density at 600 nm (OD₆₀₀) of 0.3 for both strains. Donor and recipient cells were then mixed together in a 1:1 ratio, and 200 μl of combined cell suspension was spotted on PD3 plates without antibiotics and incubated at 28°C for 5 days. Donor and recipient cells were also plated separately, to serve as negative controls. Cells were then harvested in 1× phosphate-buffered saline (PBS), and serial dilutions were plated on PD3 agar containing chloramphenicol and gentamicin for selection. Antibiotic-resistant colonies were counted after 14 days of incubation at 28°C, and select colonies were subcultured on fresh plates for additional screening.

Confirmation of X. fastidiosa transconjugants.

Double-antibiotic-resistant colonies were grown on PD3 plates with antibiotics for selection for 7 days. Cells were harvested in 1× PBS, and total DNA was extracted using the DNeasy blood and tissue kit (Qiagen). For transformants from M23-Temecula conjugation reactions, 1 μl of total DNA preparation was used as the template for PCR to confirm the identity as X. fastidiosa, the presence of the chromosomal marker (Cm^r) in the recipient strain, and the presence of pBBR5pemIK. Primers used for PCR were RST31/RST33 (X. fastidiosa-specific primers), Cmfwd/Cmrev (recipient chromosomal marker), and pemIKfwd/pemIKrev (plasmid sequence) (38). Some positive clones were further confirmed as the correct strain by sequencing the pilU and cysG genes, which are polymorphic in M23 and Temecula. For transformants from Dixon-Temecula conjugation reactions, antibiotic-resistant colonies were selected and screened as described above except that subspecies-specific primers XF2542fwd/XF2542rev and XF1968fwd/XF1968rev were used to differentiate donor and recipient. Some positive transformants were further confirmed by sequencing the pilU, holC, nuoN, and petC genes, which are polymorphic in X. fastidiosa subsp. fastidiosa and X. fastidiosa subsp. multiplex. Intact plasmids were also extracted from transconjugants and digested with EcoRV to confirm the identity as pBBR5pemIK.

Interspecies plasmid transfer with a X. fastidiosa donor strain.

X. fastidiosa donor strain M23 pBBR5pemIK was grown on PD3 plates for 5 days at 28°C, harvested in liquid PD3 medium, and adjusted to an OD₆₀₀ of 0.25. Two hundred microliters of X. fastidiosa cell suspension was spotted on PD3 plates and incubated for an additional 4 days at 28°C. The E. coli recipient strain (Top10) was grown overnight on LB agar plates, subcultured in LB liquid medium, and grown to an OD₆₀₀ of 0.5. E. coli cells (1 ml) were harvested by centrifugation and washed twice in liquid PD3 medium to remove salts from LB medium. E. coli cells (200 μl) were spotted on top of X. fastidiosa growth on plates, and the plates were incubated for an additional 24 h at 28°C. Cells were then harvested from the plates in 1× PBS, and serial dilutions were plated on LB plates supplemented with gentamicin to select for recipient cells that had acquired the plasmid pBBR5pemIK. Plasmids were extracted from E. coli transformants and digested with EcoRV to confirm the identity as pBBR5pemIK.

Interspecies plasmid transfer with an E. coli donor strain.

X. fastidiosa recipient strain Temecula Cm^r was grown for 5 days on PD3 plates, harvested with PD3 liquid medium, and adjusted to an OD₆₀₀ of 0.25. X. fastidiosa cell suspensions were spotted on PD3 plates and grown for an additional 4 days. E. coli donor strain S17-1λ carrying plasmid pBBR5pemIKpBAD-PD1607 was grown overnight on LB agar plates, subcultured in LB liquid medium supplemented with arabinose at a concentration of 0.02%, and grown on a shaker at 37°C to an OD₆₀₀ of 0.5. E. coli cells (1 ml) were harvested by centrifugation and washed twice in liquid PD3 medium to remove salts from LB medium. E. coli cells (200 μl) were spotted on top of X. fastidiosa growth on plates, and the plates were incubated for an additional 24 h at 28°C. Cells were then harvested in 1× PBS, and serial dilutions were plated on PD3 medium supplemented with chloramphenicol and gentamicin.