16. Analysis of total chloroplast DNA RFLPs in cultivated and wild species of rice

Andreas Dally and Gerard Second1

ORSTOM/CEPE, CNRS, BP 5051, 34033 Montpellier Cedex, France. 1) Present Address: Department of Plant Breeding, 252 Emerson Hall, Cornell University, Ithaca, NY 14853-1902 USA

We have improved the non-aqueous method for the isolation of chloroplasts from freeze-dried leaves. The changes include the mechanization of grinding, the use of less toxic solvents, and differential centrifugation in the presence of Triton to remove contaminant DNA. Total chloroplast DNA was digested by restriction enzymes and restriciton patterns from single plans of various species of the Eu-Oryza section of Genus Oryza were obtained.

Chloroplast DNA from 320 plants were digested by EcoR 1 and Ava I restriction enzymes. Thirty-two distinguishable patterns ("plastotypes") were observed and further characterized by digestion with Bam H1, Hin dIII, Sma I, Pst I, Bst EII and Sa1 I restriction enzymes.

A careful determination of apparent molecular weights of restriction bands allowed for the distinction of 112 mutations. Forty of these were recognized as restriction site mutations for 220 restriction sites surveyed representing 1300 base pairs of DNA sequences.

A cladistic analysis of the 112 mutations allowed us to produce a cladogram of the 32 plastotypes. Its "root" was deduced from a preliminary study of the Oryzeae tribe with digestion by enzyme Bsp x1 (Fib. 1).

This cladogram agreed completely with the cytological distinction of 5 basic genomes, and the recognition of the sativa and latifolia groups of species in the section Eu-Oryza. It also largely agreed with the finer genetic structure as revealed by isozyme electrophoresis patterns. Additional information on the maternal parents of allotetraploids was obtained. Several strong discrepancies between the relationships at nuclear and chloroplastic marker levels suggested some past nucleocytoplasmic subsitution events.

Observations of cultivated rices

Ten plastotypes were distinguished in cultivated rice. One corresponded to O. glaberrima, and 9 others were found in O. sativa. They clustered in the cladistic analysis in two groups which, as shown in Table 1, corresponded largely to the indica-japonica differentiation with one dominant type each. Intermediate indica-japonica varieties shared the japonica plastotype more often thatn the indica plastotype. The rare plastotypes corresponded 1) to certain cytoplasms which induced male sterility (see below) and 2) to unique varieties with particular features.

Some observations of cytoplastmic male-sterile plants

Thirteen cytoplasmic male sterile (CMS) O. sativa plants were analyzed. They represented 10 different lines related to at least two different sources of CMS: Wild Abortive (WA) and Gambiaca.

Two different plastotypes were found: a1, the most common plastotype found in indica varieties, and a2, not found in fertile cultivated varieties but observed in various lines of O. rufipogon. They differed by a restriction site substitution detectable in Sma I and Ava I digestion patterns (Table 2).

WA source of CMS, with 11 plants studied appeared to be heterogeneous with different plants presenting one or the other plastotype, sometimes in the same line. One plant even showed a half and half mixture of the two restriction patterns a1 and a2.

This result suggested possible biparental inheritance of chloroplsts in rice. Because a few controlled hybrids so far studeied always showed maternal inheritance of chloroplasts, biparental inheritance must not be frequent. Its detection could be favored under the breeding scheme used to maintain CMS lines, i.e. pollination with a fertile variety. Further studies are necessary to elucidate the possible relation of this finding with the origin and instability of CMS.

Fig. 1. A cladogram showing interrelationships among 32 plastotypes distinguished. Capital letters stand for the nuclear genomes and small letters for the plastotypes. The length of a branch is approximately proportional to the number of mutations specific to that branch. The arrow indicates the "root".

Table 1. Different plastotypes observed in O. sativa

===============================================================================
Plastotype    No. of       Glaszmann's             Remarks
              varieites    isozyme grouy
===============================================================================
indica
a1              19           I,II          Most indica varieites and some GMS lines
a1               9            I            Most GMS lines analyzed
b                1            I            An African cultivar
c                1            I            A Thai upland cultivar
japonica
e1              19      VI,I,II,III,IV,V   All japonica varieties, most 
                                           intermediate indica-japonica and
                                           some indica varieties
e2               2           II            Two varieties from Iran (Gherdeh
                                           and Gharib)
e4               1           IV            A Thai cultivar (Howm Om)
e5               1                         A land race of Western India with 
                                           affinity to O. glaberrima*
===============================================================================
* Reported by Lolo and Second (this volume)
===============================================================================

Table 2. Plastotypes observed in CMS plants

References

Glaszmann, J.C., 1987. Isozymes and classification of Asian rice varieties. Theor. Appl. Genet. 74: 21-30.