An official website of the United States government.
<OL> <LI> Develop an informatics system to automate design of plant-pathogen diagnostic microarrays. We will focus on ribosomal sequences for fungi and bacteria and genomic sequences for viruses. The probes will be filtered by our own criteria of taxon specificity and through the Affymetrix chip design system. <LI> Implement a Phase 1 chip using the new rapid, low-cost, design capability of NimbleExpress through Affymetrix (target: ~250,000 probes). Produce 10 phase 1 chips. <LI> Test Phase 1 chips in both the Lawrence and Goodwin labs on artificially inoculated and naturally infected plant tissue samples. We will determine which probes and labeling methods are optimal to positively indicate a small number of test pathogens (Alternaria, Septoria, Phytophthora) as well as which probes for all pathogens which have non-specific binding against three crops (wheat, corn, soybean). <LI> Review the successes and failures of the Phase 1 chip and design the Phase 2 chip, incorporating new genomic data and input on priorities from stakeholders. The chips will provide approximately 35,000 probes and will be fabricated by the standard Affymetrix technology which provides a lower per-unit cost. Produce one lot of chips (90 +/-5). <LI> Test Phase 2 chips on a broader regimen than Phase 1 and additionally solicit samples from or provide chips directly to APHIS centers interested in testing the system in real-world application. <LI> Pursue strategies to transition the approach to routine APHIS monitoring of plant pathogens. Provide a plan for utilization of the newest technologies to enable cost-effective implementation.
NON-TECHNICAL SUMMARY: Plant biosecurity requires robust methods to identify both known and novel plant pathogens. Quick and accurate diagnosis of plant legions would increase our ability to respond appropriately in cases of newly introduced dangerous pathological agents. The goal of the project is to assess the potential of DNA microarrays for identifying pathogens in infected plant tissue.
<P>
APPROACH: The methodology will measure sequence-specific RNA molecules in affected plant tissue to determine which species are present in a sample. Ribosomal RNA (rRNA) is a good choice for this because is the most abundant class of nucleic acid in living cells in terms of copies of specific sequences and has highly conserved sites suitable for priming molecular manipulations. RNA, though not rRNA, can also be used to detect viral pathogens and we expect messenger RNA (mRNA) transcripts to be abundant in heavy infection. Total RNA will be amplified and labeled using standard Affymetrix procedures modified by using random hexanucleotide and/or conserved rDNA primers. To design the array of oligonucleotide probes (25-mers are used on the Affymetrix platform), we will need to assemble and analyze all available rRNA sequences for fungal and bacterial plant pathogens and all genomic DNA for viruses. We will analyze this to generate a set of probes that include species-specific primers and more conserved probes shared among all species in a genus or higher group. We will also screen the probes for cross hybridization to host-plant rRNA and mRNAs (using EST unigene sets). We plan to apply 20-40 probes per pathogen species, plus various control probes such as to host plant rRNA. This will provide a robust single assay to distinguish among nearly a thousand pathogens. The design will be tested in two phases. Phase 1 will use the NimbleExpress fabrication procedure (low design fee, higher cost per chip) which provides approx. 250,000 probes. Ten such chips will be made and used to screen a large number of candidate probes for cross-hybridization and other performance characteristics. Phase 2 will utilize the standard Affymetrix photolithography fabrication to provide 35,000 probes. A single order, ~90, Phase 2 chips will be made and used in a variety of validation experiments. Validation will occur in three contexts: laboratory-controlled infections; field-collected samples from a single lab (collaborator Stephen Goodwin); and field-collected samples from multiple independent labs. Statistical interpretation will utilize empirical methods (e.g., discriminant functions or support-vector machine) to optimize interpretations of species previously looked at, and extrapolations to generate scoring functions for species as yet untested.