Variation in plastid inheritance between pelargonium cultivars and their hybrids

Heredity 63 (1989) 145—153 The Genetical Society of Great Britain Received 24 October 1988 Variation in plastid inheritance between pelargonium cultivars and their hybrids R. A. E. Tilney-Bassett and A. B. Almouslem School of Biological Sciences, University College of Swansea, Singleton Park, Swansea SA2 8PP, Wales, U.K. The genetic control of plastid inheritance is analysed in cultivars and hybrids of zonal pelargoniums (Pelargonium X Hortorum Bailey). After crosses between maternal parents with normal green (G) plastids and paternal parents with mutant white (W) plastids in their germ cells, the progeny consist of a mixture of green, variegated and white embryos corresponding to a maternal, biparental or paternal inheritance of plastids in each individual. Families of individual embryos from selfed or crossed parents are of two patterns that are highly dependent upon the genotype of the female parent. Type I females produce families in which green embryos are most, variegated intermediate and white least frequent. Type II females produce families in which green and white embryos are of approximately the same frequency and variegated embryos the least frequent. Five type I and six type II cultivars and several hybrids between them were selfed, and 88 crosses of various kinds were made to create a heterogeneous population of 2601 plants that were tested for their plastid inheritance pattern by crossing with a standard male parent as source of mutant plastid; 1298 families were classified as type I and 1303 as type II. The maternal (G + V), biparental (V) and paternal (V +W) plastid percentage transmission was estimated for each family, and the frequencies of percentage classes within the population of type I or type II families expressed as histograms. The type I families fell into a skewed distribution with such a maternal bias that about one-half inherited only maternal plastids, and over two-thirds of them had at least some maternal plastids in every embryo. The average family had a maternal percentage of 987 per cent and paternal of 47 per cent, and no family was below 80 per cent maternal or above 25 per cent paternal. The overall maternal : paternal ratio was about 21: 1. The type II families were more normally distributed with a large variance. The average family had a maternal percentage of 632 per cent and paternal of 381 per cent with an overall maternal : paternal ratio of about 5:3. On average 466 per cent of type I families and 647 per cent of type II families had biparental progeny. The frequencies of families decreased sharply with increasing proportions of biparental progeny. Analysis of variance of a subset of the type II population showed that there was a significant added variance component for maternal percentage, and a significant linear regression of offspring on parents; hence, in addition to the major gene difference determining type I versus type II patterns, there was a significant polygenic component causing variations within each pattern. Alternative hypotheses for the mechanisms determining differences in the pattern of plastid inheritance in Pelargonium, Oenothera, and Medicago are discussed. INTRODUCTION The use of the fluorochrome 4'6-diamidino-2- phenyl-indole (DAPI) to detect plastid DNA in generative and/or sperm cells of pollen from 235 flowering plant species, has led Corriveau and Coleman (1988) to conclude that biparental inheritance of plastids may occur in about 14 per cent of genera, with examples scattered among 19 per cent of the 80 families examined. Their study confirmed the majority of biparental species pre- viously identified by electronmicroscopic and/or genetic evidence (Kirk and Tilney-Bassett, 1978; Sears 1980). As well as the traditional white (albino) plastid mutants, inheritance studies are now using mutants with resistance to herbicide or antibiotic, or plastids differing in a biochemical marker (Börner and Sears, 1986). Moreover, the detection of the inheritance pattern has been improved with the help of tissue culture (Medgyesy et a!., 1986) and the analysis of chloroplast DNAs, as recently reviewed by Smith (1988a). A par- 146 ticularly interesting approach for the analysis of competition between plastids is the creation of hybrid and cybrid cells by protoplast fusion (Kumar and Cooper-Bland, 1986; Kumar and Cocking, 1987). In the present paper, however, we shall report results from making crosses between plants containing green (G) or white (W) plastids in their germ cells, as these have contributed to our interest in the genotypic control of biparental plastid inheritance. During interspecific G x W crosses in Oenothera, Schötz (1954, 1968, 1974, 1975) discovered that the biparental plastid inheritance was R. A. E. TILNEY-BASSETT AND A. B. ALMOUSLEM An insight into the nature of the population variability was begun with the analysis of variance of 36 G x W and 36 reciprocal W x G crosses (Tilney-Bassett, 1976). This was followed by examin- ing the variation between 214 families, which showed signs of a skewed or normal distribution among type I or type II plants respectively (Tilney- Bassett 1984). The investigation has now been expanded by observing the behaviour of a much larger population. The importance of the nuclear genotype has recently been discovered in other plants. Normally plastid inheritance in Petunia is maternal (Cor- strongly biased towards the maternal parent. riveau and Coleman, 1988), and this was confirmed Nevertheless, there was considerable variation, which he attributed to differences in the ability of by Cornu and Dulieu (1988) in crosses between inbred lines and a viable, virescent plastid mutant of P. hybrida "Blue Bedder". But one line, Thl-3, consistently produced up to 2 per cent variegated seedlings indicating a low level of pollen transmission. This was confirmed by creating, through recurrent backcrosses, a strain consisting of the Thl-3 nucleus of P. hybrida and the normal green wild-type plastids to multiply in competition with a constant source of mutant plastids. When Chiu et a!. (1988) reciprocally crossed four wild-type plastids with seven plastid mutants in a constant nuclear background, they confirmed the highly significant role of the plastome in the process of plastid transmission. They concluded that differences in the multiplication rates were important but insufficient to explain all the results. In Pelargonium the study of biparental plastid inheritance has been largely within cultivars, and differences that exist between plastid mutants have not been studied extensively (Abdel-Wahab and Tilney-Bassett, 1981). Instead, most attention has been given to the important role played by the nuclear genome. After G x W crosses the progeny plastids of P. parodii. The advantage of this combination was that the P. parodii plastids produced a 7.4 kbp fragment, when treated with the BamHl restriction enzyme, that was absent from the P. hybr (...truncated)