48. Distribution and evolution of a transposon-like sequence in 30 varieties of rice

Section of Biochemistry, Molecular and Cell Biology, Cornell University, Ithaca, NY 14853 USA

Previously, we have reported the structure and sequence of the rice mitochrondrial cytochrome oxidase subunit II (COII) gene (Kao et al, 198). Sequence comparison between rice and maize COII genes revealed the existence of a 461 bp insertion sequence within the intron of the rice COII gene. The insertion sequence possesses a seven bp inverted repeat which is flanked by an eight bp direct repeat. Based on its structure, it is likely that the insertion sequence within the rice intron was derived from a transposable element by duplication with the generation of direct repeats flanking the insertion sequence at the site of integration.

The transposon-like structure of the insertion sequence led us to look for the existence of other copies throughout the rice genome. Using an insertion- specific probe (positions 1052)-1358), we carried out Southern blot analysis on total genomic DNA from 30 varieties of rice. As shown in Table 1, all varieties were found to have extr copies of the insertion sequence; some (5 Kb, 7 Kb and 11 Kb) are common to most varieties, and other (1.2 Kb, 1.5 Kb, 2.5 Kb, 4 Kb and 6.5 Kb) are unique to certain varieties. When the same blot was hybridized with an intron-specific probe, only one band (6.5 Kb in O. brachyantha and 3.4 Kb in all other varieites examined) corresponding to the COII gene was detected. This result indicates that the extra copies of the insertion sequence are not necessarily associated with the intron or coding sequences of the COII gene, and are distributed more or less randomly throughout the mitochondrial genome. The band intensity of these copies was weaker than that of the copy in COII from which the probe was derived. This weaker hybridization intensity may be interpreted to mean that the nucleotide sequences of the additional copies has diverged from that of the COII insertion sequence. It is not likely that the additional copies are located on the nuclear genome because there are perhaps several hundred copies of the mitochondrial genomes but only one copy of the nuclear genome per cell. According to sequencing data, the insertion sequence in the COII gene has only short reading frames, implying that it is not capable of coding for any transposase. However, we do not know how these extra copies are different, how well they maintain the transposon-like structure, or whether any of the extra copies are active as a transposable element.

From results presented in Table 1, we also noticed the following facts. (A) A 5 Kb band is present in all japonica and indica varieties but not in any of the wild rice species. This result may be interpreted to mean that the transfer of the insertion-specific sequence into the 5 Kb fragment occurred after the divergence of these subspecies from the wild species. (B) A 2.5 Kb band is present only in Labelle, Lemont, M202, Bell Patna and CP 231 among the indica varieties. Thus, the transfer of the insertion-specific sequence into the 2.5 Kb fragment occurred relatively recently. The 2.5 Kb hybridizing band is found in both the rice variety Labelle and its female parent Belle Patna but not in its male parent Dawn. A likely explanation is that the 2.5 Kb fragment is present in the mitochondrial genome of Bell Patna but not in that of Dawn, and it was maternally inherited in the production of Labelle. Another possibility is that Labelle is more similar to Belle Patna than to Dawn. This is consistent with the observations by Bollich et al. (1973) on physiological properties and by other members in our lab (previous note) on the apparent copy number and thermal stability of a repetitive DNA, which indicate that Labelle resembles Belle Patna more than Dawn. (C) O. brachyantha (FF genome) is the only species that does not have a 3.4 Kb band, but it has a 6.5 Kb band containing the COII gene with its insertion sequence. This result is consistent with the fact that among all the rice species and varieties listed in Table 1, O. brachyantha is the only one that is not an Eu-Oryza (Tateoka 1964). (D) All varieties of japonica rice have the 1.5 Kb hybridizing band, which means the presence of this gragment is indicative of the japonica variety. It is likely that the existence or absence of a DNA fragment hybridizing to this or other specific probes may be correlated with the pattern of inheritance in certain varieties of rice.

We thank Xinping Zhao for the rice DNA samples. This work is supported by research grants RF84066, Allocation No. 3, from the Rockefeller Foundation, and GM29179 from the National Institutes of Health, U.S. Public Health Service.

Table 1. Summary of a Southern blot analysis on the total genomic DNA from varieties of rice using an insertion-specific probe from the rice COII gene.

Kao, T.H., E. Moon and R. Wu, 1984. Cytochrome oxidase subunit II gene of rice has an insertion sequence within the intron. Nucleic Acids Res. 12: 7305-7315.

Bollich, C.N., J.G. Atkins, J.E. Scott and B.D. Webb, 1973. Registration of Labelle rice. Crop Science 13: 773-774.

Tateoka, T., 1964. Notes on some grasses. SVI. Embryo structure of the genus Oryza in relation to the systematics. Am. J. Botany 51: 539-543.