Location: Crop Improvement and Genetics Research2012 Annual Report
1a. Objectives (from AD-416):
The citrus variety Carrizo is the single most important rootstock to the U.S. citrus industry. Generated by a ‘Washington’ navel orange x Poncirus trifoliata cross, Carrizo shows tolerance to diseases such as citrus tristeza virus and foot rot. In addition, this variety has recently been demonstrated to be tolerant to huanglongbing (HLB, citrus greening) (2), viewed as the most serious current threat to citrus production. The purpose of this proposal is to acquire additional sequence for the USDA Public Citrus Genome Database database of Carrizo DNA, which is currently serving as a general database for research programs requiring citrus genomic DNA sequences. The additional sequence will allow for assembly of extended genomic contigs, greatly improving utility of the database as well as allowing alignment to an existing citrus physical map and future genome sequence assemblies from sweet orange. These additional sequences will provide a database of sufficient depth (approximately 8x) to allow assembly of Mb-scale contigs suitable for annotation and posting to the database website to improve the efficiency and utility of search functions. Following release of the sweet orange genome, final assembly of the Carrizo genome will be quickly achieved from the data generated through this proposal, and a haploid P. trifoliata genome will be extracted. Likely alleles conferring resistance to HLB can be identified and quickly mobilized for testing in segregating populations (such as the mapping population being planted by the USDA in Ft. Pierce) and for use in transgenic strategies. These data will also provide a foundation for studying genetics/genomics of rootstock characteristics which will greatly facilitate efficient screening and selection of improved rootstock varieties using marker assisted selection.
1b. Approach (from AD-416):
Funding requested to increase the database depth to approximately 8x via acquisition of approximately 2,400 Mb (6 million sequences) of genomic Carrizo DNA. This data will be acquired in two steps. First, two additional Roche 454 random read runs to generate an approximately 5x genome. Second, two libraries of paired-end fragments will be prepared from Carrizo genomic DNA. These libraries will facilitate assembly of larger genomic contigs by revealing relative orientations and positions of shorter assemblies produced via the random 454 reads. The two libraries will be generated from genomic fragments of 3 kb or 10 kb and subjected to 454 sequence analysis resulting in approximately 1.2 Gb of additional sequence. This sequencing effort will result in sufficient genome coverage to allow computational assembly of contigs suitable for annotation and entry into and existing ARS citrus genome website.
3. Progress Report:
The citrus variety Carrizo, the single most important rootstock in the U.S., was generated by a ‘Washington’ navel orange x Poncirus trifoliata cross, Carrizo shows tolerance to diseases such as Citrus tristeza virus, foot rot and huanglongbing (HLB, citrus greening). We had previously generated a Carrizo genomic sequence database containing 3.5x coverage of the genome (USDA Public Citrus Genome Database, http://citrus.pw.usda.gov/). The primary objective of this research was to acquire additional sequence for this database, which is universally available for unrestricted use by research programs requiring citrus genomic DNA sequences. Funding was sought to increase the database depth from the previous 3.5x coverage to approximately 10X (454 GS FLX). Because of an increase in average read length, the sequence acquired significantly exceeded that indicated in the proposal (15X final coverage). This increased read length significantly improved segregation of the two genomes during the assembly process. Assembly of the sequence data was performed by the Roche 454 gsAssembler version 2.6 software program. Multiple program runs were done based on default or heterologous genome settings which influenced the final scaffold forming numbers; 39,966 versus 31,239, respectively. There were 262,573 total contigs formed and represented in the assembly. The software estimates in both cases predicted 9X coverage and a genome size of 673.9Mb indicating partial segregation of the Poncirus trifoliata and Citrus sinensis genomes. This genome and associated assemblies are now of sufficient quality to provide a valuable tool to scientists applying molecular genetics to citrus improvement. This research relates to Objective 2 of the parent project, the development of molecular tools from fruit trees. This includes identification and isolation of transcriptional control elements, relating this progress to Objective 2 of the parent project.