1.8.29
PHYSICAL MAP AND ORGANIZATION OF THE RICE BLAST FUNGUS, MAGNAPORTHE GRISEA, CHROMOSOME 7

H ZHU, B BLACKMON, M SASINOWSKI and RA DEAN

Department of Plant Pathology and Physiology, and Clemson University Genome Center, Clemson University, Clemson, SC 29634, USA

Background and objectives
The rice blast fungus, Magnaporthe grisea, is one of the most destructive plant pathogens worldwide and has been adopted as a key model organism for studying various aspects of the host-plant interaction. The fungus infects a wide range of grasses, in addition to rice, and is ideally suited for Mendelian and molecular genetic studies. It is a haploid, heterothallic Ascomycete and has a relatively small genome size ranging from 40 to 50 Mbp permitting separation of intact chromosomes by PFGE. Numerous genes controlling the infection process, signal transduction pathways, and recognition of the host plants have been identified. Over 400 genetic markers have been mapped to the seven chromosomes. A crucial aspect of further molecular dissection of M. grisea is the generation of a comprehensive physical map. Previously, a large-insert (130 kbp) bacterial artificial chromosome (BAC) library was constructed using a rice-infecting strain 70-15 [1]. This deep coverage library (>25 genome equivalents) is particularly suitable for the construction of a physical map and sequence ready contigs.

Materials and methods
For constructing physical maps using large insert libraries several methods are commonly used, including hybridization and fingerprinting. Each method has a number of deficiencies, therefore we adopted a strategy of applying both simultaneously to create a robust physical map. Our initial goal was to reconstruct a physical map of chromosome 7 using BAC clones. In a single-copy chromosome region, two non-overlapping BAC clones can be bridged with shared clones identified by hybridization using their labelled ends as probes. When two single-copy BAC clones are separated by a repetitive region, they are less likely to share a bridging BAC clone; however, the BAC clones between the two probes may still overlap each other. To confirm the physical relationship in both scenarios, BAC DNA fingerprinting is required. Thus our strategy involved hybridization of chromosome 7-specific BAC clones containing single-copy DNA and RFLP markers on chromosome 7 to the BAC library, fingerprinting all chromosome 7-specific BAC clones identified by hybridization, and contig assembly based on both hybridization and fingerprinting analyses.

Results and conclusions
Chromosome 7 was isolated from a CHEF gel and used to probe filters of a sub BAC library containing single-copy clones. 147 clones were identified. These were used in a hybridization without substitution approach along with 19 RFLP markers from chromosome 7 to identify a total of 624 BAC clones. BAC clones were fingerprinted by HindIII digestion, separated by analytical agarose gel electrophoresis, and stained with SYBRgold. Gel images were captured by a Fluorimager and restriction fragment bands called using IMAGE software. Hybridization contigs were constructed using a random cost algorithm, while fingerprinting contigs were constructed using the software package FPC.

Results from both methods were in good agreement. The combined data produced two anchored contigs covering over 95% of the 4.2 Mbp chromosome 7. The genetic and physical maps closely agreed, with the exception of a single chromosome inversion. This may reflect strain variation. Based on the contig maps, a minimum BAC tile was created to investigate gene expression and organization of the chromosome. Whole-genome reconstruction is in progress and current status will be presented.

References
1. Zhu H, Choi S, Johnston AK et al., 1997. Fungal Genetics and Biology 21, 337-347.