We are investigating volatile organic compound (VOC) based detection diagnostics with two major citrus diseases: HLB and CTV. We performed simultaneous biomarker extraction using cutting-edge analytical chemistry approaches with off-gassed chemicals in field studies to evaluate disease detection based on these gaseous chemical species using a portable differential mobility spectrometer (DMS) system. The main idea is to combine data of individual compounds to develop an integrated panel of markers to obtain a more robust means of early detection for these diseases. The CTV sampling was done at the Lincove Station, CA (summer 2010) and Pauma Valley, CA (Feb 2011). DMS measurements of healthy, mild and severely CTV infected Valencia citrus was performed. In addition to the mobile VOC analysis, we collected leaf samples from all three groups for later real time PCR gene expression analysis, (15 leafs per tree, 6 trees per class, 3 class in total). We will analyze for genes that seem to be strong host biomarkers of early infections. We also collected whole-leaf metabolite extracts that will be subjected GC-TOF/MS analysis to investigate the secondary metabolism of host plant in response to CTV infection. Our initial results indicate a set of 18 VOC compounds (with distinct chemical identifications) that appear associated with milder forms of the CTV disease. We found that co-infection with stubborn also produced a distinct VOC metabolite profile, and we can still identify a clear ‘CTV’ signal even in the presence of the confounding co-infection factor. We have acquired additional VOC samples from severely CTV infected group only, in order to investigate if the same volatiles are up-regulated in these phenotypically distinct trees. Future work would include a return visit to Puma Valley to increase the number of DMS measurements, and a return visit to Lindcove (if new infections are discovered). Additional sampling in other regions of CA will help to quantify how geographical variation may (or may not) affect the detection of disease specific biomarkers. We are also investigating VOC diagnostics for HLB detection. Prior to in-field sampling, we completed preliminary studies including sample collection protocol optimization from greenhouse citrus on the UC Davis campus. The in-field testing was carried out with collaborators at Lake Alfred, FL throughout this year (2010-2011) to account for weather and seasonal (trees blossoming, fruit harvest, etc.) variations. Currently, 3 trips were made in November 2010, December 2010 and January 2011. Sample collection was carried out in parallel using polymer-based absorption devices (SPME and Twister), as well as two portable GC/DMS analysis units. Approximately, 80 samples were collected up to date using SPME, and analyzed using GC/MS. Approximately 100 individual GC/DMS runs were recorded for the same pair-matched trees used for the SPME sample collection. In addition, approximately 160 samples were taken using Twister devices (December 2010), and the GC/MS analysis of these samples is on-going. For all of the above experiments, trees of one varietal (Hamlin) were included. Half of all samples were taken for HLB-infected and presumed healthy trees selected by scouts at the CREC facility in Lake Alfred FL. A number of trees with symptoms varying from relatively mild to severe were included in the study. For the next stage, we will include PCR-positive symptomatic trees, PCR-positive asymptomatic trees and PCR-negative presumably asymptomatic trees. This will allow us to assess feasibility of the DMS-based chemical sensing for early-stage asymptomatic disease detection. The collected data are being analyzed and a mathematical model is currently being developed for the differentiation of HLB-sick and healthy trees based on our data.
We have completed the target selection and design of MOLigo probes for detection of Candidatus Liberibacter, Xylella fastidiosa, Xanthomonas axonopodis pv. citri, and citrus tristeza virus. Characterization probes for Liberibacter have also been designed, and testing of all Liberibacter probes will begin this week using synthetic DNA targets. Because there has been a large influx of new Liberibacter sequence data since probes were designed, we have initiated an in silico evaluation of all probes to determine if they will detect and characterize the new isolates. We met with Dr. John Hartung of USDA ARS in Beltsville, MD and Dr. Bill Schneider of USDA ARS Ft. Detrick to discuss samples for the validation panels. Both have agreed to send purified nucleic acids from infected plants, insect vectors, and healthy plants. We will assemble these samples into our validation panels for testing once the assays appear to be working properly on synthetic targets. We recently learned that Luminex plans to discontinue in January 2012 the type of non-magnetic beads that we are using to develop these assays. Because we plan to transfer the technology to service labs, we need to either locate another source of non-magnetic beads, or move our development efforts to the magnetic beads. We have contacted people at Luminex and Radix, a Luminex partner about the availability of tagged, non-magnetic beads past January 2012. We will proceed with development on the type of bead that will be available for the foreseeable future. The advantage of the magnetic beads is that they perform as well or better than the non-magnetic beads, and are the only beads compatible with Luminex’s new, less expensive machine. This machine only has a 50-plex capacity, but costs around $35K, making it feasible to deploy our citrus assays to additional labs at lower startup costs. Our concern is that the magnetic beads will cost more, which will increase the cost of each assay. Still, it is anticipated that the total assay cost will still be a fraction of the cost per marker as compared to RT-PCR assays currently in use. We will continue to pursue options for tagged, non-magnetic beads. Additionally, we have moved the oligo design software to a new server, which is online and accessible to us and the public at http://moligodesigner.lanl.gov . We are in the process of automating the multiplex function of the design tool, which will facilitate probe selection and assay design. This will make the iterative re-design of probes and assays should we encounter some probes that do not perform according to our specifications.
The project main objective is to select antibody fragments type single chain fragment variable (scFv) against Huonglongbing (HLB), Citrus Verigated Chlorsis (CVC) and Citrus Stubborn Disease the latter as a proof-of-concept, and the development of immunoassays that incorporate these scFvs. We have requested Stubborn samples from many collaborators, but they are not available yet as a pure culture. However, during this time, we have constructed one scFv library from chicken spleens that were immunized with bacterial proteins isolated in Fort Detrick from cultures of Candatus Libribacter asiaticus (CaLas). The immunization protocol and deliver of spleens were performed in 100 days, and mRNA was immediately extracted to construct the library, using recombinant PCR, into an M13 vector. To validate the library, we performed a Phage Display selection of specific scFvs against Xyllela OMPs. Three highly reactive fragment antibodies, from 12 reactive clones, were further secreted in culture in a TOP10 vector, and HPLC-purified for further validation with ELISA and western blots. All three antibodies strongly recognized the expected 42 KDa MopB protein. Fluorescence microscopy analysis of scFvs in Xyllela cells is under investigation. Interestingly, the same antibodies could recognize the related protein in a related pathogen Xanthomonas juglandis OMPs giving confidence that these antibodies should detect the mopB in CVC. The detection of mopB in Xylella were comparable with a polyclonal antibody against Xylella MopB developed by Dr. Bruening. Additionally, we have also started working with fluorescence spectroscopy based detection. Besides training and calibration of equipment, we have established the parameters to work with specific range of wavelengths for our reactions with quantum dots (q-dots). In order to work under field conditions, we decide to use q-dots with the excitation at 480nm and emission in the spectrum of 500-900 nm. Protocols for antibodies conjugation to q-dots are under development. Tests of antibody stability are also under way. In conclusion, we have established a very simple conventional ELISA with the antibodies against bacteria OMPs, and the assay parameters will be translated to fluorescence spectroscopy in the near future to improve sensitivity of the test.
The recently published Ca. Liberibacter asiaticus (Las) strain psy62 genome, derived from a psyllid, revealed a prophage-like region of DNA in the genome, but phage have not been associated with Las to date. In the present study, shotgun sequencing and a fosmid DNA library of curated Las strain UF506, originally derived from citrus symptomatic for HLB, revealed two largely homologous, circular phage genomes, SC1 and SC2. SC2 encoded putative adhesin and peroxidase genes that had not previously been identified in Las and which may be involved in lysogenic conversion. SC2 also appeared to lack lytic cycle genes and replicated as a prophage excision plasmid, in addition to being found integrated in tandem with SC1 in the UF506 chromosome. By contrast, SC1 carried suspected lytic cycle genes and was found in nonintegrated, lytic cycle forms only in planta. Phage particles associated with Las were found in the phloem of infected periwinkles by transmission electron microscopy. The average phage head size was ~21.30 nm and the average neck length was ~ 5.03 nm; based on the relatively short neck length, the observed phage particles belong to the Podoviridae. In psyllids, both SC1 and SC2 were found only as prophage. Similar phage DNA sequences (corresponding to both SC1 and SC2 and including putative lysogenic conversion genes) were found in Ca. L. americanus (Lam) strain ‘S’o Paulo’ isolated from infected citrus in Brazil in collaboration with Dr. Nelson Wulff at Fundecitrus. This Lam genome is about one year from completion. It currently has a total of 1,203,790 bps on 355 contigs, with the avg. length of the contigs at 3,390 bps. In planta, including citrus, both SC1 and SC2 can enter a phage-programmed cell death cycle called the lytic replication cycle. Once the lytic cycle is engaged, cell lysis is assured by a variety of SC1-encoded genes. Surprisingly, a potential colicin IA and potential cognate colicin immunity protein were identified on SC1 and SC2, respectively. Colicins are anti-bacterial proteins that kill any competitor cells lacking the colicin immunity protein. The fact that the SC2 replicon carries a putative immunity protein would render Las cells with SC2 resistant to any colicins produced by SC1, and also would prioritize SC2 preservation in a given Las strain. Assuming infectious capability of both phages, SC2 might infect a Liberibacter strain and confer a selective advantage by virtue of its lysogenic conversion genes (confering major pathogenic advantages), while SC1, by itself, is predicted to be lethal to any strain it can infect (unless it was previously or simultaneously infected with SC2). The hypothesized lethality of SC1 would confer a selective advantage to strains that shed infectious SC1 particles by virtue of its colicin gene alone, which could eliminate potential competition by similar bacteria in the phloem cell or in the psyllid alimentary canal. This mechanism might also ensure essentially monoculture infections of citrus by only a single Las strain (with SC1 and SC2) at a given time. Over time in a grove setting, this could lead to survival of only the fittest of multiple Liberibacter strains in a case of competition by two or more populations. In Brazil, Lam used to be the predominant form of predominant form of Liberibacter causing HLB, but has been replaced by Las as the predominant species. It is possible that the Las colicin kills Lam. Further investigation of Liberibacter colicins and activation of the phage lytic cycle seems warranted as potential control strategies for Liberibacters generally.
The overall objectives of this project were: 1) Development of rapid regeneration of HLB-affected periwinkle by optimizing growth conditions for the diseased cuttings; 2) Screening and evaluating anti-microbial molecules for suppression of Liberibacter in periwinkle. In the past three months, the research was mainly focused on optimizing a citrus grafting system for screening chemical compounds in HLB control. 1. Developing an optimized citrus grafting system for screening chemical compounds in HLB control The HLB-affected scions from different phenotypic types (yellow shoot and Blotchy mottle) and different cultivars (lemon and grape fruit) were grafted onto the Las-free citrus rootstocks (sour orange and grapefruit). The results showed that more than 60% of HLB-affected lemon scions survived and 70% of the inoculated rootstocks were infected at 6 months after grafting with high Las bacterial titers of HLB-affected scions. If the Las bacterial titers in the grafted scions were low (less than 1000), all grafted scions survived and grew better, and less than 30% rootstocks were infected with low bacterial numbers. The results also showed that plant growth regulators (PGRs) increased the survival rates of the grafted scions, but had no effects on the infection rates. 2. Evaluation of chemical compounds against Las bacteria using the optimized grafting systems HLB-affected lemon branches were sampled and soaked in the PS (penicillin and streptomycin), MDL (metronidazole), Kasumin and DBNPA (2, 2-dibromo-3-nitrilopropionamide) solutions overnight, cut into 3 buds for scions (S1, S2 and S3) and then grafted onto the Las-free grapefruit (R). More than 80% of the scions treated with PS survived and grew better than the untreated controls, and no rootstocks and scions tested positive by qPCR for Las bacteria. When the HLB-affected scions were treated with MDL, more than 60% of the scions survived, but more than 95% rootstocks and grown scions were infected by Las bacteria. DBNPA-treatment partly inhibited the Las bacteria with 40% of infection rates in the grafted scions and rootstocks. Only 4.4% HLB-affected scions survived when treated with kasumin at 1mg/ml, showing the phytotoxic effects. 3. Continue to monitor the effect of antibiotics on HLB-affected citrus trees in the field In June of 2010, the screened antibiotic combinations (PS and KO) have been applied to the HLB-affected citrus in the field by trunk-injection. The primarily results showed that the antibiotic combinations decreased the Las bacteria titer in the treated citrus as compared to controls. In the past three months, the trunk injection of antibiotic combination was continued in Oct and Dec. The Las bacterial number was quantified by qPCR. Two abstracts were submitted in 2nd International HLB meeting, one manuscript has been accepted in phyopathology.
Fluorescence and visible-near infrared spectrometry is currently under investigation for detecting HLB. The Multiplex’ fluorescence data were collected from 30 healthy, nutrient-deficient, and HLB infected trees (Hamlin and Valencia) under field and laboratory conditions from CREC groves. In addition, fluorescence and visible-near infrared data from about 498, 53, and 549 corresponding citrus leaves were collected from different varieties of citrus trees (Southern Garden groves) under laboratory conditions. The PCR analysis of the leaf samples from CREC groves has been performed. Preliminary analysis on the fluorescence data (21 features) indicated that yellow fluorescence (green excitation) and red fluorescence (UV) could indicate stressed leaf condition; while blue-to-red fluorescence (UV) could indicate HLB infection under laboratory conditions. Similarly, fluorescence excitation ratios (red & UV) and yellow fluorescence (red) could indicate stress- and HLB detection under field condition, respectively. Further analysis is ongoing to evaluate the fluorescence features that can be used as a predictor for stressed-condition and HLB infection in citrus leaves. Aerial images (color,multispectral and hyperspectral cameras) were acquired from a grove near Clewiston in Hendry County and CREC groves on Dec. 3, 2010. Reflectance data from healthy and infected canopies (with RTK-GPS recorded coordinates) were measured. The pre-processing and analysis of these aerial images are ongoing. Analysis procedure and results for previous set of data (2007, 2009) have been reported to a journal article, ‘Citrus greening disease detection using aerial hyperspectral and multispectral imaging techniques’, submitted to the Journal of Applied Remote Sensing. In this paper, based on the calibration set data, a spectral library was created for healthy and infected canopies, and some of the HLB detection algorithms that were tested using ENVI hyperspectral imaging software are spectral angle mapping (SAM), linear spectral unmixing, mixture tuned matched filtering, and spectral feature fitting. Comparing the ground-truth data, SAM with multispectral data yielded a detection accuracy of 87%. Currently, support vector machine and neural networks in addition to the previously-tested algorithms are being evaluated to validate the results with the new data set. Based on the completed preliminary studies on data acquisition and sample collection protocols optimization, field data from Hamlin trees (CREC groves) were collected in Nov. and Dec. 2010 for the estimation of citrus volatiles. The sample collection was carried out using the SPME (for GC/MS analysis), Twister devices and two portable gas chromatography-differential mobility spectrometry (GC/DMS) units. Approximately 80, 100 and 160 samples were collected using SPME, GC/DMS, and Twister devices with about equal number of samples from HLB-infected and healthy trees. Preliminary data analysis on the volatiles data collected using the GC/DMS units using principal component analysis (PCA) indicated that the first two PCs presented a clear separation between HLB and healthy samples. Using the leave-one-out validation strategy, the principal component regression yielded a classification accuracy of 83% on this sample set. These results show a very promising direction of using portable GC/DMS unit for field detection. At CREC, leaf volatiles from an HLB-infected tree leaf (new leaves) were collected at times 1, 4, 8, and 16 h and temperatures range from 47-75’F with a 1 cm tri-phase SPME fiber. Volatiles collected were very localized and tend to vary depending on leaves and branches sampled. The whole day sampling appears to collect more volatiles than overnight sampling, presumably due to higher daytime temperatures. We are continuing to sample positional differences in greening trees and comparing them to control trees to determine branch to branch and tree to tree variation.
At least a couple of most promising silica nanogel based copper (Cu) formulations (which demonstrated superior biocidal efficacy in laboratory based testing) will be systemically evaluated in the field in the upcoming season. The Co-PI (Jim Graham at the CREC, Lake Alfred, FL)and his team will conduct this trial at the Fort Pierce, FL facility. Santra group at the UCF have synthesized the materials for this upcoming trial. Material characterization is currently on-going to check product quality. A number of commercial Cu products including Kocide will be used as positive control whereas pure silica nanogel (without Cu) will be used as negative control. Laboratory data successfully demonstrated improved biocidal efficacy against X. Alfalfae (indicative of improved Cu bioavailability) and strong adherence properties to citrus leaf surface (indicative of improved longevity). Preliminary field trail data collected so far indicate that nanogel based Cu formulations are quite effective against citrus canker. Details will be available in future reports.
The HLB disease causing ‘Candidatus Liberibacter asiaticus’ (CLas) pathogen was successfully detected using PCR probes. The digoxigenin (DIG) based DNA-PCR probes showed promising results compared to riboprobes (RNA probes). Though highly specific, RNA probes did not improve hybridization of the tissue bolts and occasionally non-specific hybridization was observed in healthy tissue blots. Of all the DNA-PCR probes, EF ‘Tu probes were much promising in their specificity to detect the ‘CLas’ pathogen (both in HLB infected citrus plants as well as HLB positive psyllids). These EF ‘Tu probes were also effective in detecting CLas’ pathogen even in the non-symptomatic tissues and in different developmental stages of ‘CLas’ pathogen carrying psyllids. The sensitivity of EF ‘Tu probes against detecting HLB infected citrus plants directly from the field was also tested. The EF ‘Tu DIG labeled tissue blot hybridization was shown to be positive for the field sampled HLB diseased plants and we did not find any hybridization signal for the healthy plants. Tissue imprints, once on the membrane can be stored for later use, can be shipped easily and generally are stable. The results of these observations were presented as a poster presentation in the 2nd International HLB Conference at Orlando. Salient feature of this presentation included results of hybridization of a branch containing symptomatic and asymptomatic leaves which showed strong hybridization signal in the leaf midrib, veins, petiole and the main branch stem. These results confirmed hybridization observed with detached plant parts. We are in the process of doing tissue-blots of very young leaves and actively growing shoot tips which have been shown to contain high HLB concetrations. Real time PCR (qPCR) evaluations of different probes employed in this study were performed and results suggested that the hybridization signals by the probes in the tissue imprints were directly correlated with the copy number in CLas genome.
The research proposal involves expression of putative effectors of ‘Candidatus Liberibacter asiaticus’ (CLas) in citrus using Citrus tristeza virus in an effort to understand if the putative effectors contain determinants for pathogenicity to citrus. CLas, is a fastidious, Gram-negative, phloem-limited, alpha-proteobacterium is the causative agent of Huanglongbing (HLB) also known as citrus greening, and is one of the most destructive diseases of citrus worldwide. It is believed that many pathogenicity islands and prophage related gene clusters observed in the CLas genome contain many hypothetical proteins unique to CLas which may function as putative bacterial effectors required for pathogen virulence, multiplication etc., and insect transmission determinants. Towards this end, as a first step we have amplified at least 20-25 hypothetical protein genes by polymerase chain reaction (PCR) using specific primers designed based on the CLas sequence, and the total nucleic acids isolated from HLB infected citrus as template. Amplification was not observed with the template from healthy citrus which indicated the specificity of the putative effector amplicons. The genes encoding putative effector proteins were cloned into citrus tristeza virus vector, behind heterologous beet yellows virus CP sub-genomic RNA controller element engineered between the CPm and CP genes of the CTV vector in a binary vector. By the use of citrus tristeza virus (CTV) vector, we could express putative effectors of the CLas directly inside the phloem of citrus. Initially, the CTV vector with engineered putative HLB effectors was used to transform Agrobacterium (EHA 105) and agro-infiltrate Nicotiana benthamiana seedlings for transient expression in inoculated leaves and subsequent systemic spread. The expectation is that the expression of the putative effectors of CLas in N. benthamiana would induce a phenotypic response. We are using N. benthamiana as an intermediate host since the direct agroinfiltration of citrus has not been possible. The CTV virions containing HLB effectors from N. benthamiana will be purified on sucrose step gradient and will be used to inoculate citrus plants by bark-flap inoculation. The resulting systemic spread and expression of the putative effectors throughout citrus trees will enable us to understand the role of the putative effectors in disease induction. We are at a stage to transfer the CTV virions from the N. benthamiana seedlings to citrus. Recently, early infection of N. benthamiana seedling (6- 8 weeks as opposed to 10- 12 weeks) with CTV vector was observed (Choaa and Dawson, personal communication) when a foreign was engineered under CTV CP promoter between the p23 ORF and the 3′ non-translated region, and we plan to use this vector in future experiments in addition to the expression of HLB effectors in citrus using as originally proposed.
To date, 400 different media formulations have been tested for their ability to support growth of Candidatus Liberibacter asiaticus. Inoculum was prepared by washing the bacteria from biofilms formed on the surface of alimentary canals of the psyllid, Diaphorina citri, which were reared on citrus plants infected with huanglongbing. Two days ago, a four to six fold increase in the number of the bacterium was detected in one medium formulation (BM397) five days after it was inoculated. Several other medium formulations exhibited lesser degrees of increased numbers. The culture in medium BM397 was transferred to fresh medium and will be monitored for further growth. New cultures in medium BM397 and the other media that had promising results are being established to determine if the these results are repeatable. Medium BM 397 has an extremely complex formulation that will need to be optimized if it continues to support growth of Liberibacter asiaticus. In summary, our approach to culturing Liberibacter asiaticus appears to be working working.
This report will be present according to objectives of project, i.e., diagnosis of Liberibacter asiaticus by both, serological and molecular approaches. Firstly, from the molecular approach the Nested PCR protocol was finalized and definitively the double step is much more sensitive than the one step. We used DNA from a single psyllid or from inoculated citrus plants by psyllids as template for quantitative PCR (QPCR) and the results were compared against Nested PCR one step (NPCR-OS), double step (NPCR-DS), and traditional PCR. We verified that a hundred percent of matches between QPCR and NPCR-DS was obtained for the samples whose threshold cycle (Ct) values ranged from 22 to 32, as determined by QPCR. For the samples whose Ct values ranged from 32.1 to 34, and from 34.1 to 36 there were 98.11% and 96.61% of matches between NPCR-DS and QPCR, respectively. For those samples with Ct values above 36.1 only 54% of results were shared between QPCR and NPCR-DS. The number of positive samples determined by NPCR-OS was lower than determined by NPCR-DS or QPCR, but a higher when compared to traditional PCR for all comparison levels. Finalizing, based on the significant agreement between NPCR-DS and QPCR for the samples with Ct <36 as well as others facts, as the non-frequently Ct around of 38 reported for the negative controls, we used the 36th cycle as a threshold for decide if the samples are positive (Ct <36) or negative (Ct >36). For the serologic-based diagnosis we tested the eight potential antibodies (see last report) against Lib. asiatic symptomatic plant and negative controls. From the eight antibodies, four (T3699-4, T3699-5, T3699-6, T3699-7) shown reactivity against infected samples by ELISA (OD >0.456), but no reaction was observed against health plants or several bacteria species as Xylella fastidiosa, Xanthomonas, Pseudomonas, Agrobacterium, Escherichia coli, and Methilobacterium spp (OD<0.099). Previous tests shown that the reactivity of antibody T3699-7 was statistically more reactive to the others three antibodies, T3699-4, T3699-5, and T3699-6. The ELISA results of that antibody using different citrus species budding infected by Las were quit similar to real time PCR, with a moderate Cohen's Kappa index (0.583; SD 0.316). Both, the moderate value and the high standard deviation (SD) probably are consequence of low number of samples used in the tests. Also, the T3699-7 was used on Dot Blot methodology which shown potentially useful on diagnosis of Lib. asiaticus. Although this project was officially finish on November 2010, we are still working on set-up and optimize the serologic based protocols. The good specificity and reactivity of T3699-7 and the absence of a commercial serologic based test make us excite on continuous this work. The next steps will be to test the T3699-7 on a broad number of samples using both ELISA and DOT-BLOT methodologies.
Previously, a total of 293 ‘Ca. L. asiaticus’ isolates obtained from HLB-affected plants American and Asian countries were used for genetic analysis. Samples were successfully amplified by seven microsatellite primer pairs. The number of alleles and haploid genetic diversity per locus ranged from 2 to 30, and 0.204 to 0.881, respectively. In the clonal corrected data set, genotypic linkage disequilibrium was not detected by pairwise comparison of loci across the overall isolates (P.>.0.01). A total of 147 genotypes (haplotypes) were identified. Overall haploid genetic diversities were the highest in Asian countries. The haploid genetic diversity of the Florida (USA) isolates was lowest among all the geographic groupings. A UPGMA clustering analysis identified three major groups of ‘Ca. L. asiaticus’. Isolates from India were clustered in a distinct group. Most of the isolates from China and other Asian countries, and Brazil were generally grouped in group 1. While some isolates from Florida occurred in group 1, most isolates from Florida were clustered in group 2. The STRUCTURE analysis based on Bayesian modeling further assessed the genetic structure of ‘Ca. L. asiaticus’. This approach utilizes statistical methods to determine the relationships among the isolates without prior population information. In this analysis three different clusters (K) were identified based on the ad hoc statistic .K. The membership of each isolate in a group obtained from STRUCTURE analysis can be estimated as (q), the ancestry coefficient, which varies on a scale between 0’1.0, with 1.0 indicating full membership in a population. Individuals can be assigned to multiple clusters (with values of q summing to 1.0) indicating they are admixed. Individual sample with q.’.0.90 (ancestry coefficient) were considered as having single lineage and individuals with q.<.0.90 were considered as admixed lineage. The result of STRUCTURE analysis is consistent with UPGMA in which isolates from India were grouped in a distinct cluster. Brazilian and most east-southeast Asian isolates were clustered as a single lineage (q.'.0.90). Some isolates taken from central Florida (Polk, Pasco, and Lake Counties) shared the same lineage with east-southeast Asian and Brazilian isolates. Most Florida isolates, however, grouped in a different cluster. Some admixed isolates (q.<.0.90) were found in Florida as well as in Guangxi province in China, and in Cambodia. The seven microsatellites developed in this study are useful for detection, isolate differentiation, and genetic analysis of 'Ca. L. asiaticus'. Our results showed that current 'Ca. L. asiaticus' populations in HLB-affected citrus in Asia and the Americas are comprised of three distinct genetic groups: (1) Indian, (2) predominantly east-southeast Asian and South American (Brazil) and (3) predominantly North American (Florida, USA). While regional differences were observed from the distribution of dominant clusters, the similar genetic makeup of east-southeast Asian and Brazilian isolates lead us to hypothesize that 'Ca. L. asiaticus' populations in Brazilian groves were most likely introduced from east or southeast Asia. The precise sources of the dominant genetic group of 'Ca. L. asiaticus' retrieved from Florida are not clearly resolved from the present analysis. However, less-pervasive groups may have been introduced directly from Asia or via Brazil. While the results here provide some insights into the origins and genetic structure of HLB-associated 'Ca. L. asiaticus', it should be noted that more samples adding to this study will help further increase resolution for differentiation genetic relationships among 'Ca. L. asiaticus' worldwide.
The methods if this project, Laser Capture Microdissection (LCM) followed by 16sDNA and metagenomic 454 XLR pyrosequencing, failed to reveal the presence of L. asiaticus in infected tissue, despite the fact that qPCR controls on parallel infected tissues showed the presence of typical amounts of L. asiaticus. We speculate that this may be attributed to the non-uniform distribution of these bacteria in the infected plant. We attempted to compensate for this by using larger samples for sequence analysis in additional trials, but we still did not find Liberibacter. Other reasons for failing to find Liberibacter cannot be ruled out. The second unexpected finding of our study was that the infected tissues, but not the uninfected control tissues, showed the presence of a wide diversity of bacteria at very low abundance. Should this result prove reproducible, it would suggest that the Liberibacter infection results in increased susceptibility to secondary infections. We hasten to point out that this study was not designed to test that hypothesis, and thus less material was examined from uninfected leaves than from infected, and that DNA yields were disappointingly low throughout this study. Nonetheless, this is an intriguing observation that merits further examination.
The Southern Gardens Diagnostic Laboratory has been in operation since October 2006. To date, the SGDL has received approximately 152,000 grower samples, 17,100 samples from Southern Gardens, and approximately 6,000 individual psyllid samples. In addition, several thousand samples have been submitted by various University and USDA researchers on various projects. For the period covered by this report (July, 2010 to September, 2010), a total of 4,803 samples were processed. Of these, approximately 45.4% were grower samples (2,180), 92% were samples from Southern Gardens (442), and 45.4% were research or psyllid samples from various University, USDA, or private research projects (2,181). Although sample volume varies from year to year and also during a given year, it appears that the total grower sample load will be lower this year compared to the previous year. The total number of grower samples is projected to be in the 30,000 range compared to the previous year’s grower sample load of 44,985 grower samples. However the sample load of research related samples is increasing and combined we expect to be in the range of the samples used to prepare the budget for this project. In addition to the disease testing activities, the lab recently coordinated a blind panel test of infectious psyllids for many of the laboratories involved in psyllid testing in California. In addition to the various labs, a variety of methods were evaluated and the results of the panel testing will be presented at the International HLB meeting in Orlando in January, 2011. Currently we are in the process of organizing a similar infectious psyllid panel test that will be conducted on a national level. We are also in the process of automating a portion of the DNA extraction procedures which should increase the lab efficiency. Validation of the process is expected to be completed during the first quarter of 2011.
The objective of this project is 1) to complete the Las genome sequence and conduct comparative genomics studies on the Liberibacter species; 2) to explore the potential role of the microbial community and genetic diversity of Las bacteria in HLB development; 3) to confirm if Las bacteria are seed-transmissible and their role in HLB development. A complete circular genome of Candidatus Liberibacter asiaticus was obtained using a metagenomics approach and published in MPMI 22:1011-1020, 2009. In collaboration with Dr. Hong Lin at the USDA-ARS in Parlier, California, we have obtained a complete genome sequence of Ca. L. psyllaurous with ca.1.25Mb . We have also obtained a draft genome (approximately 70%) of Ca. L. americanus using multiple displacement amplification and 454 pyrosequencing technologies. A comparative genomics of these close-related bacteria revealed useful information for understanding their pathogenesis and evolutions. We are continuing the screening of our BAC library, and results indicate that the BAC library contains most, if not all sequences of the Las genome with the insets up to 130 kb. Sequencing analysis confirmed our previous published Las psy62 genome sequence. However, in the hyper variable regions, genetic diversity was observed among the clones. We have characterized the ATP translocase from Las and proved its function using a heterologous E. coli system. This data was published in J. Bacteriol. 192:834-840, 2010. We are currently developing an antibody-based “drug” to target this protein, aimed at disrupting ATP import, which may be important for its survival. We have also characterized the individual genes of two putative zinc operons in Las, with an overall aim of interfering with the ability of Las to regulate zinc uptake. Seed transmission of Las was tested in grapefruit, sweet orange, and trifoliate orange. Relatively high titers of Las were detected from both seed coats and inner seed coats collected from HLB-affected citrus plants. A very low titer of Las was detected from the embryos and seedlings using nested PCR and real-time PCR. Most, if not all the seedlings did not show typical HLB symptoms and contained a relatively low Las bacterial titer for HLB, even in the three to four year old seedlings. The results indicated that the seed-transmitted Las could not cause typical HLB disease by themselves, which suggested “Detection of Candidatus Liberibacter asiaticus was NOT necessarily equal to the presence of “HLB disease” in plants.” Psyllid transmission study on the Las-positive seedlings was performed. High percentage of psyllids acquired Las bacterium but did not have the same bacterial levels as those from HLB-affected citrus plants. In vitro culture of Las bacterium appeared to be possible. Ten to thirty-five folds increase of Las bacterium in liquid media were observed. We are continuing to modify growth conditions for the improvement of the Las growth rate in media.
(863) 956-8817
(863) 956-5894
700 Experiment Station Road
Lake Alfred, FL 33850
Email Us
