IRRI, Los Baños, The Philippines

Background and objective
‘Blast’ disease of rice and over 50 graminaceous species is caused by the heterothallic ascomycete Magnaporthe grisea (Hebert) Barr. Several dispersed repetitive DNA elements have been used to characterize isolates and populations of the pathogen: MGR586, grasshopper, and MAGGY [1]. Based on the distribution of these elements, the species has been considered to comprise asexual and genetically isolated host-limited forms [2]. In the Indian Himalayas. rice and several millet hosts of M. ;grisea are cultivated in the same fields and are frequently subject to simultaneous blast epidemics. The pathogen population in this region presents a unique opportunity to test the hypothesis that host-delimited forms of the pathogen are genetically isolated.

Materials and methods
Isolates were collected from fields in one village, Bandrakoti, where rice (>80 isolates) and two millets, Setaria italica (>60 isolates) and Eleusine coracana (>60 isolates) are grown in near-by plots, or sometimes inter-sown. EcoR1-digests of genomic DNA were probed with MGR586, MAGGY and grasshopper. They were further probed for single (or low-copy) loci developed as part of an M. ;grisea mapping project (gift of S. A. Leong). Fertile isolates from rice and S. italica were crossed and single ascospore progeny probed with MGR586.

Results and conclusions
All rice-derived isolates had high copy number (>50 bands) MGR586 and MAGGY. Isolates from E. ;coracana had high copy number (>30) grasshopper, no MAGGY, and few (<10) to no MGR586. Isolates from S. ;italica had low copy MGR586 (<15 bands); but, about half had no MAGGY and high copy grasshopper. The other half had high copy number MAGGY, but no grasshopper. Thus, rice- andE. ;coracana-derived isolates had clearly distinct multiple copy probe profiles consistent with genetic isolation. The more complex pattern from S. ;italica suggested a bridge between rice and E. ;coracana pathogens. Population studies have revealed only very rare intermediate multilocus profiles that would be expected if sexual recombination were occurring between the different groups. However, crosses between isolates with different multilocus profiles yielded single ascospore progeny with multilocus profiles highly skewed towards one or the other parents (e.g. >50% of the progeny from crosses between isolates from rice and S. ;italica had >40 MGR586 bands). Thus, multilocus data are not conclusive that sexual recombination does not occur between isolates from different hosts. Single locus probes revealed a very similar pattern as the multilocus probes, including the dimorphic profiles within the S. ;italica pathogens. However, within each host-of-origin group occasional atypical alleles were encountered. These were typical of alleles from one of the other groups. More single-copy probes are being used to determine whether there is limited gene flow between groups. The existing data suggest strong host limitation in M. ;grisea, even where they are in close contact in this particular ecology. Gene flow appears to be limited, and this may reflect an early phase of speciation.

1. Zeigler RS, Leong SA, Teng, PS. 1994. Rice Blast Disease. Wallingford: CABI. 626 pp.
2. Shull V, Hamer, J. 1997. Fungal Genetics and Biology 20, 59-69.