Endophytic Bacteria Associated to Sharpshooters (Hemiptera: Cicadellidae), Insect Vectors of Xylella fastidiosa subsp. pauca

1Department of Genetics, Escola Superior de Agricultura “Luiz de Queiroz”, University of São Paulo, Piracicaba, SP, Brazil 2Department of Entomology, Plant Pathology and Zoology, Escola Superior de Agricultura “Luiz de Queiroz”, Universidade de São Paulo, Piracicaba, SP, Brazil 3Núcleo Integrado de Biotecnologia, University of Mogi das Cruzes, Mogi das Cruzes, SP, Brazil 4Department of Entomology, University of California Riverside, Riverside, CA, USA 5Institute of Natural Sciences, Federal University of Alfenas, Alfenas, MG, Brazil


Introduction
Citrus variegated chlorosis (CVC) was first reported in Brazil in 1987 [1,2] and by 2000, the disease affected already 34% of the 200 million sweet orange trees in state of São Paulo. By 2005, the percentage had increased to 43%, and CVC was present in all citrus growing regions of Brazil [3,4]. The disease is caused by the xylem-limited Gram-negative bacterial pathogen, Xylella fastidiosa subsp. pauca [5,6].
Endophytic microorganisms live within plants without causing apparent harm to the host. We have studied the possible use of endophytes as vehicles to control both phytopathogens and insects [7][8][9][10]. Endophytes colonize ecological niches similar to that of phytopathogens [11,12]. The co-localization enables the interaction between both, pathogens and endophytes, creating the possibility to endophytes to act as biocontrol agents [13,14,15]. Biocontrol activity of the endophytic bacterial community may be through the synthesis of structural compounds, such as siderophores and extracellular enzymes [16,17], and the induction and expression of general molecular-based plant immunity [16]. Alternatively, they may act by niche competition, preventing pathogens from becoming established in a host [18].
In citrus, several endophytic bacteria were isolated and Methylobacterium spp. and Curtobacterium flaccumfaciens were further determined as the main endophytic species interacting with X. fastidiosa subsp. pauca [19,20]. These species vary in population density when CVC-affected and asymptomatic plants are compared. Later, Lacava et al. [21] reported that the growth of X. fastidiosa subsp. pauca was inhibited by endophytic C. flaccumfaciens and stimulated by Methylobacterium sp. A similar effect was demonstrated by the reduced severity of the X. fastidiosa subsp. pauca colonization in plants priory colonized by C. flaccumfaciens [22].
Cicadellinae leafhoppers, commonly named sharpshooters, are xylem-feeder insects [23]. An association has been observed between the xylem-feeding habit of sharpshooters and their ability to transmit X. fastidiosa [24][25][26][27]. In Brazilian citrus groves, Dilobopterus costalimai, Oncometopia facialis and Acrogonia citrina are the most common sharpshooters found [25][26][27]. After acquisition of X. fastidiosa by the insects, colonies of bacterial cells were visible in the cibarium and pre-cibarium of transmitting insects attached to the foregut walls [21,28] and also in insect honeydew [29]. The transmission efficiency of bacteria is a measure of the ability to successfully acquire bacteria from an infected plant and transmit to healthy ones. The transmission rates for the main species associated with CVC vary from 1% to 5% [26,30,31].
Many aspects can influence the transmission of a pathogen by an insect vector such as the low concentration of X. fastidiosa subsp. pauca cells in the citrus plants [22] and the low number of colonized vessels in infected plants [32]. The interaction between different bacteria inside the insect foregut may also influence the transmission, as once inside the foregut, bacterial interaction, such as competition for nutrients, space and other complex interactions could occur. In this context, the aims of this work were (i) to access the bacterial population associated with the main sharpshooters responsible for the transmission of X. fastidiosa subsp. pauca in citrus; (ii) to evaluate the diversity of heterotrophic bacteria by amplified ribosomal DNA restriction analysis (ARDRA); (iii) to compare the bacterial community presents in the insect heads from distinct species and collected from citrus at distinct period of the year by denaturing gradient gel electrophoresis (DGGE).

Isolation of heterotrophic bacteria from insects
At each sampling date, ten insects of each species (O. facialis, A. citrina and D. costalimai) were selected for bacterial isolation (total of 90 analyzed insects: 3 species x 3 dates x 10 insects). Vectors were surface sterilized by immersion in 75% ethanol for two min, transferring them to a container with sodium hypochlorite solution (2% available Cl -) for two min, and rinsing them twice in sterile double-distilled water [33]. After sterilization their heads were cut off, and triturated separately in 1 ml of saline solution (NaCl 0.85 %). An aliquot of 100 μl of this suspension and 1:10 dilutions were plated on non-selective TSB 5% medium (Tryptone Soy Broth, DIFCO) amended with agar (1.5%). Plates were incubated at 28°C for 10 days, colonies were counted, classified based on their morphology and the number of colony forming unities per insect head (CFU/insect head) was determined.

DNA extraction from bacterial isolates
DNA was extracted from fresh bacteria cultures of 120 isolates (Insect Associated Bacterium -IAB) with the following protocol according [24]. A 1.5 ml sample of an overnight bacterial culture was centrifuged for 2 min at 12.000 Χ g and resuspended in 1 ml of TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0), recentrifuged, and resuspended in 500 μl of TE buffer plus 0.5 g of 0.1 mm diameter glass beads and 30 μl of 10% sodium dodecyl sulfate. The cells were homogenized for 30 seg in a bead beater (Braun cell homogenizer; B. Braun, Melsungen, Germany). A 500 μl volume of Tris-buffered phenol was then added; the solution was mixed and centrifuged for 10 min at 12.000 Χ g. The aqueous phase was extracted once with phenol-chloroform-isoamyl alcohol (25:24:1) and once with chloroform-isoamyl alcohol (24:1), and the DNA was precipitated with isopropanol (1/6 actual volume) incubated for 5 min at room temperature and collected by centrifugation (10 min at 12.000 Χ g). The pellet was washed with 70% ethanol, air dried, and resuspended in 30 μl of double-distilled water.

PCR for 16S rDNA amplification
The amplification of the 16S rDNA from the isolates was carried out in a 50μl reaction volume containing 1μl bacterial DNA (approx. 20 ng), 0.4 μmol -1 of each primer (R1378 and PO27F, details in Table  1), 200 μmol -1 of each dNTPs, 3.75 mmol -1 of MgCl 2 , 2.5 U Taq DNA polymerase (Invitrogen, São Paulo, Brazil) in 10 mmol -1 Tris-HCl (pH 8.3) and 10 mmol -1 KCl. The amplification protocol consisted of an initial step at 94°C for 4 min, followed by 30 amplification cycles of 94°C for 1 min, 62°C for 1 min and 72°C for 1 min with a final extension at 72°C for 10 min. Amplification products were separated by electrophoresis by spotting 5μl of the PCR reaction mixture onto 1% agarose gel and visualized under ultra-violet light after staining with ethidium bromide (0.5 μg ml -1 ).

ARDRA and ribotype selection
After the amplification of bacterial 16S rDNA from 120 bacterial isolates (40 isolates from each insect species) the amplified fragments (1360bp) were digested with the restriction enzyme AluI (AG/CT) (Fermentas, Vilnius, Lithuania) for 5 hours at 37°C and visualized in a 1.4% agarose gel stained with ethidium bromide (0.5 μg ml -1 ) under UV light. The resulting patterns were analyzed, and distinct cleavage patterns were considered as ribotypes [34].

Sequencing of the 16S rDNA gene from the selected isolates
Later, isolate representatives of main ribotypes had their 16S rDNA genes partially sequenced and blasted against the nr/nt database at the GenBank (http://www.ncbi.nlm.nih.gov/BLAST) and classified at Ribosomal Data Project (http://rdp.cme.msu.edu/). Also, the phylogeny of obtained sequences was reconstructed in a tree composed by Insect Associated Bacteria 16S rDNA sequences, sequences from type strains of the genera Methylobacterium and Curtobacterium (obtained from Ribosomal Data Project database), and sequences from previously isolated citrus endophytic bacteria from these genera [17]. Alignment and clustering were performed by Kimura-2 parameter and Neighbour-joinning methodology. The phylogenetic and molecular analyses were conducted using MEGA version 4.1 [35], where 1000 replication of bootstrap analyses were performed.

DNA extraction from insect heads
After surface sterilization the heads of insects were cut off, frozen with liquid nitrogen and triturated, separately, in 1 ml of extraction buffer (10 mM Tris-HCl, 1mM EDTA, pH 8.0, SDS 10%). The total DNA extraction was carried out according to the bacterial DNA extraction protocol [33]. From each insect species, DNA from 15 heads was extracted at each sampling date.

Detection of Methylobacterium mesophilicum and Curtobacterium flaccumfaciens by specific PCR
DNA extracted from insect heads was tested for the presence of M. mesophilicum and C. flaccumfaciens, the most important endophytes associated with X. fastidiosa subsp. pauca, by using specific primers [36,37]. Due to a limitation in the amount of bacterial DNA from one single insect head, a nested approach was applied. First, universal bacterial 16S rDNA PCR amplification, using primers PO27F and     (Table 1), was preformed as described below. Second, two microliters of the first reaction was re-amplified in a nested PCR performed with primers MMV6 and R1378 (Table 1), which generated a product of 300bp when DNA from M. mesophilicum is present. Similarly, the detection of C. flaccumfaciens was made with primers CF16S and 530R (Table 1), which generated a product of 400bp. The PCR reactions consisted of an initial denaturation step of 94°C for 7 min, followed by 25 cycles of 94°C for 1 min, 56°C and 62°C (for M. mesophilicum and C. flaccumfaciens respectively) for 1 min, and 72°C for 1 min and a final extension of 10 min at 72°C. The percentage of positive results was determined.

Assessment of the bacterial community in sharpshooter heads by DGGE
The total DNA extracted from insect heads was subjected for DGGE analysis. A nested approach was applied as previously described for specific PCR procedures. First amplification was performed using primers PO27F and R1378 ( Table 1). The second amplification was made with DGGE primers R1378 and U968GC (Table 1) as previously described [38]. A PCR mixture was made up of 1 µl of extracted DNA (approx. 5 ng) from the insect heads, 5 µl of 10X Stoffel buffer (10 mM Tris-HCl [pH 8.3], 10 mM KCl), 20 pmol of each primer (Table 1), 200 µmol of each deoxynucleoside triphosphate, 3.75 mM MgCl 2 , 0.5 µl of formamide and 2.5 U of Taq DNA polymerase Invitrogen (Carlsbad, CA, USA) in a 50 µl final volume. For the amplification protocol, denaturation was carried out at 94°C for 5 min, followed by 30 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 2 min, and a final extension of 10 min at 72°C. Five micro liters of the PCR product was analyzed by electrophoresis in a 1.4% agarose gel with 0.5X TBE buffer before running the DGGE gel.
DGGE was performed as previously described [33,39] using the INGENY phor U2 apparatus (INGENY, Goes, The Netherlands). PCR samples were loaded onto 6.0% polyacrylamide gels in 0.5X TAE buffer (20 mM Tris-acetate, 0.5 mM EDTA, pH 7.4). The polyacrylamide gels were made with denaturing gradients ranging from 45 to 65% (where the 100% denaturant contained 7 M urea and 40% formamide). The gels were run for 16 h at 100 V and 60°C, after which the gels were soaked for 30 min in SYBR Green I nucleic acid stain (1:10,000 dilution; Molecular Probes, Leiden, The Netherlands) and immediately photographed under UV light.
In order to compare the DGGE fingerprinting, gels were submitted to analysis by the software GelcomparII (Applied Maths, Kortrigk, Belgium), where images were converted, normalized and band patterns were compared. Cluster analysis of DGGE patterns was performed using the UPGMA method (unweighted pair group method with arithmetic mean) based on the similarity calculated by densitometric Pearson correlation.

Isolation of cultivable bacteria from insects
A total of 17.230 bacterial colonies were counted in all isolations from different species and at distinct sampling time. These colonies were classified into three distinct morphological groups; G1) pinkpigmented colonies which appear on plates after seven days of incubations, typically Methylobacterium spp.; G2) dark pink or orange fast grower colonies, typically Curtobacterium spp.; G3) other colonies. The concentration of bacteria (CFU/insect head) belonging to each morphological group is represented in (Figure 1).
The results show a differential composition of the heterotrophic bacterial communities according to the sharpshooter species and depending on the month of sampling. Abundance of total bacteria was higher from samples taken in March compared to numbers from May and June. Also specific groups varied in frequency along the period of

ARDRA and sequencing
A subsample of the total number of colonies obtained (120 colonies, 40 isolated from each insect species) was subjected to the genotypic characterization by ARDRA. In total, 16 cleavage patterns were observed, determining the ribotypes constituting the heterotrophic bacterial communities from sharpshooter heads. Among these ribotypes, the colonies from the two targeted groups, Curtobacterium spp. and Methylobacterium spp., has revealed a fidelity in the pattern of the 16S rDNA gene cleavage, allowing this approach to measure the proportion of these bacteria within the sampled colonies ( Figure 2). Other bacterial haplotypes were characterized randomly and sequences have shown the affiliation of the isolates to the genera Bacillus, Brachybacterium, Brevibacillus, Brevundimonas, Nocardia, Paenibacillus, Pseudoclavibacter, Rhodococcus, Sphingomonas and Staphylococcus (Table 2).
Correlating the number of ribotypes found in colonies from each insect species, we observed a higher occurrence of the ribotype II (Curtobacterium spp.) in species O. facialis and A. citrina, and a similar trend but in lower number for the occurrence of Methylobacterium spp. (ribotype I) (Figure 2).

Confirmation of the
Curtobacterium spp. and     Table 3).

Phylogeny of sharpshooter isolates and similarity to citrusendophytic bacteria
Another point found in this work was the comparative phylogeny of isolates from both targeted groups (Insect Associated Bacteria -IAB) with endophytic isolates obtained in previous works. UIA-12-R, UIA-13-R, UIA-14-R and ER1.6 are16S rDNA sequences from Curtobacterium flaccumfaciens and AR1.6/11, SR1.6/2 and SR3/27 are from Methylobacterium species previously isolated from citrus [17].One can observe the clustering of IAB isolated in this work with endophytic strains isolated previously from citrus trees, more specifically C. flaccumfaciens strain UIA-12-R ( Figure 3) and Methylobacterium strains SR1.6/2 and SR3/27 ( Figure 4). In general, bacteria isolated in this study clustered together.

Analysis of the bacterial communities of sharpshooter heads by DGGE
The DGGE analysis showed variable fingerprinting according to the period when insects were collected. The first separation was observed for samples collected in June, with a further sub-branching for samples from May and March ( Figure 5). These variations were based on changes in the abundance of bands and also in the intensity of similar bands found in distinct samples. Also, within each major branching it is possible to separate the DGGE pattern according to the vector species, with a more deterministic separation for the species A. citrina ( Figure 5). For other species, D. costalimai and O. facialis, the species-selection was lower, as evidenced by the mixing of samples from these species of sharpshooters.

Discussion
The role of associated microbial community in the pathogenic potential of X. fastidiosa has been discussed in the literature [12,14,17,18,37], where authors state that these microorganisms can partially inhibit or promote the growth of X. fastidiosa inside the citrus plants. However, there was a lack of knowledge about the occurrence and frequency of endophytic bacteria commonly found inside the insect vectors, which transmit pathogens from plant-to-plant. In order to fill this gap this study was conducted, using three insect species described as X. fastidiosa vectors, generating the first insights in this important aspect, allowing further focused studies on the biological control or symbiotic control of this pathogen in citrus groves. We focused our approaches in detecting and quantifying two major citrus endophytes in the selected insects. However, a culture-independent approach was also used to describe changes in bacterial communities associated with these vectors.
Differences in bacterial communities associated with insects suggest complex interactions between insects, host plants, bacteria and environment. In March, when the temperatures were around 25°C and the precipitation 160 mm the bacterial community was denser with a higher number of bacterial groups, when compared to the months of May and June when the temperatures and the precipitation were lower (20°C, 15°C and 80 mm, 60mm respectively). Each insect contained an associated bacterial community that was altered in time and in different environmental conditions. Climate shifts affect not only the bacterial community in plants but also the behavior of insects [31], which choose alternative host plants during different periods of the year, and host plant conditions (citrus or adjacent ones), which change depending on temperature and precipitation on the area.
The DGGE technique has been used to study bacterial communities [39]. Reeson et al. [40] described the bacterial communities associated with wasps (Vespula germanica) by applying DGGE-based methods as well as Lacava et al. [33] who used DGGE to characterize microbe in glassy-winged sharpshooter (GWSS), the main insect vector of X. fastidiosa subsp. piercei [6] causing Pierce's disease (PD) in the USA [22]. Lacava et al. [33] found M. extorquens and C. flaccumfaciens as part of the bacterial community of GWSS in the USA. In addition, Kirkpatrick and Wilhelm [41] have also isolated strains of C. flaccumfaciens as part of endophytic bacterial community of grapevine in California, USA. Also, C. flaccumfaciens was consistently isolated as an endophytic bacterium from citrus plants in Brazil [17,18].
In the present study, Curtobacterium sp. was the most abundant bacteria colonizing insect heads. Curtobacterium flaccumfaciens was implicated in playing an important role in the prevention of CVC symptoms in citrus trees [17,18,33]. The presence of the citrus endophyte, Curtobacterium sp., colonizing the insect heads could explain why the transmission efficiency of X. fastidiosa subsp. pauca by vectors is low (5 to 10%), when compared to the transmission of X. fastidiosa subsp. fastidiosa by GWSS, which cause Pierece's disease in grapevines (45%) [26,31]. Table 3 illustrates that the genus Curtobacterium, can play an important role in the transmission of X. fastidiosa subsp. Pauca. The presence of this Curtobacterium has been cited in several works where the interaction of these bacteria with X.fastidiosa had been studied. High population of Curtobacterium were observed in asymptomatic citrus plant [19,20], A less severe colonization of plants by X.fastidiosa was observed in plants previously colonized by C.flaccumfaciens [22] and in this study, a high population of Curtobacterium was observed associated to insects with lower transmission rate of X.fastidiosa.These observations show a colonization pattern what suggest an interaction between both species where Curtobacterium would affect negatively the growth of X.fastidiosa and so the transmission and development of CVC symptoms. Curtobacterium could be influencing pathogen adhesion to the vector foregut or inhibiting growth of the pathogen.
The bacterial communities associated with vector insects and plants differ in abundance through the yearly season. Endophytic bacteria could influence disease development by reducing the insect transmission efficiency due to competition with pathogens in host plants and also in insect foreguts. In addition the bacterial communities in the foregut of insect vectors of X. fastidiosa subs p. pauca changed with time, environmental conditions and in different insect species. However, members of the genus Curtobacterium were consistently detected in the sharpshooters foregut and are commonly isolated from the xylem of citrus plants [17], and because of this, they may be candidates for biological control.
In a recent review, Lacava et al. [15] suggested C. flaccunfaciens is a potential candidate for biological control of CVC because an antagonism between C. flaccumfaciens and X. fastidiosa subsp. pauca was strongly indicated in vitro and in vivo [17,18,33] including inhibition of growth of X. fastidiosa subsp. pauca and reduced severity of the disease symptoms in the presence of this phytopathogen. The ability demonstrated by C. flaccumfaciens to colonize plant tissues in the presence of X. fastidiosa subsp. pauca and the reduction of disease symptoms caused by X. fastidiosa subsp. pauca [19] are prerequisites for the use of this endophytic bacterium as a biocontrol agent. Since members of the genus Curtobacterium were consistently detected in Acrogonia citrina Marucci  (Signoret), two of the main insect vectors of X. fastidiosa subsp. pauca as demonstrated in the present study, they fulfill another requirement of candidates for biological control of X. fastidiosa subs p. pauca [19], i.e. they can colonize both the insect vectors of X. fastidiosa subsp. pauca and citrus plants.