Bulblist 3/1/2004

Heterofertilization and Mentoring

Heterofertilization occurs when the ovum and the central cell are fertilized by sperm cells delivered by different pollen tubes. Thus, the embryo (developing from the fertilized ovum) and the endosperm (from the fertilized central cell) have different "fathers".

Rotarenco and Eder (2003) studied hererofertilization in maize by using pollen parents carrying the dominant R1-nj gene, which gives anthocyanin pigmentation to the top of the endosperm and embryo. The seed parent, 092, lacks the gene, but expresses it well in its seeds when the pollen parent carries it.

092 was pollinated with a mixture of pollen from 092 and X28C, half and half (by weight). The mature seeds were examined for anthocyanin pigment. 1.5% had pigmented embryo-unpigmented endosperm or white embryo-pigmented endosperm. Ignoring the possibility of an extraordinary increase in the rate of mutation, these seeds must have been fertilized by two pollen grains one from X27C, carrying the R1-nj gene, the other from 092.

In a parallel study, the haploid-inducing line MHI was used as the source of the R1-nj gene. Haploid-inducing maize strains produce a small percentage (6.2% in the ZMS strain) of pollen grains with morphologically distinct pairs of sperm cells. One is normal, the other defective. When these are delivered to an egg sac by the pollen tube, either the ovum or the central cell will be fertilized, but not both. Since MHI carries the R1-nj gene, it is possible to examine the seeds and see what happened at fertilization. If both the endosperm and embryo have anthocyanin pigment, then *presumably* one of the pollen grains with two normal sperm cells was responsible. Pigmented embryo/unpigmented endosperm indicates that one of the half-defective pollen grains was involved, and that the embryo is a diploid hybrid (the endosperm having developed without fertilization). However, if the embryo is white and the endosperm pigmented, the embryo is almost certainly a haploid derived entirely from the seed parent.

092 was pollinated by a mixture of pollen (half and half by weight) from 092 and MHI. 7.5% of the seeds were found to be haploids. Of the remaining diploids, 5.3% had resulted from heterofertilization.

So what has all this to do with mentor pollen?

Some interspecific and wider crosses are not successful because the endosperm fails to develop normally. Or, having developed for a while it degenerates. Cases have been reported in tomato hybrids, and in iris hybrids, among others. Or, the endosperm may develop as it should while the embryo does not. Embryo rescue and endosperm culture have been used to get past these problems, but most amateurs (I think) don't care to bother with such techniques.

But if heterofertilization occurs as frequently in plants other than maize, pollen mixtures may be expected to give viable hybrid seeds that do not require special techniques to develop into healthy plants. Pollen from one parent might fertilize the ovum successfully, while the endosperm fails. And pollen from a different parent, even of the same species, might have more success in the endosperm but less in the embryo. Together they could produce a single seed with a good embryo and a good endosperm.

According to Breshnev, etal. (1960), "The first crosses of ordinary tomato with L peruvianum were effected in 1933, 1934 and 1935 in the Kamennaya Step Experimental Breeding station of Voronezh Region; some tiny seeds were obtained, of which only one germinated. The hybrid plant had intermediate structure with a dominance of paternal L. peruvianum characters. There were irregularities of the process of pollen formation (univalents, laggard chromosomes, etc)"

"When methods of pollinating two or three times and pollination with a pollen mixture of L. peruvianum and the maternal variety were used, quite a number of plants of the intermediate type were obtained. In subsequent generations these intermediate forms continued to appear; intermediate forms comprised plants the fruit of which were yellow, red and orange when ripening and had two to six small locules. Variability in size and color of the fruit could be observed even within the same plant."

It is interesting to note that Rotarenco and Eder (2003) also found that a single hand-pollination by haploid-inducing maize varieties resulted in 2 to 3 times as many haploids as open pollination with the same parents. This suggests that heterofertilization can occur sequentially; the second (or third?) fertilization correcting whatever didn't get done the first time. Failure of one or the other fertilization actually increases the probability of heterofertilization, even when both pollen grains derive from the same pollen parent.

If heterofertilization can occur in Iris, it could make it much easier to raise hybrids of TB varieties x Onco-Regelio species. Ordinarily, the hybrid seeds fail to develop a firm endosperm, so the embryos must be cultured artificially. Perhaps a pollen mixture containing some TB pollen would give seeds with hybrid embryos and "pure" TB endosperms that could be raised normally.

The whole concept of "scientific breeding" with its insistence on "pure" pollen from a single parent has hindered research as much as it has helped. Multiple pollination — simultaneously and sequentially — is the rule in Nature, not the exception. Bees, hummingbirds, moths, butterflies, ants, bats and all the other pollinators are rarely fastidious about their handling of pollen. Nor do they avoid flowers that have already been pollinated. In consequence, hybrids occur in Nature that are nearly impossible when we restrict ourselves to "scientific" breeding methods.

The use of pollen mixtures should be part of every breeder's "bag of tricks" for obtaining difficult hybrids. It's a good idea, however, to use pollen parents with dominant traits not present in the seed parent, and differing one from the other. For example, if we wish to cross two species, it would be better to use a white-flowered variety as the seed parent. Pollen from the seed parent can be mixed with pollen from a red-flowered species that is not very compatible. In this way we may be pretty sure that seedlings with pink or red flowers are hybrids, while any white-flowered plants probably are not.

Special preference should be given to any pollen parent with dominant traits expressed early in development. Distinctive pigmentation of cotyledon, hypocotyl or leaf is very desirable, as is a distinctive leaf-shape. Such marker traits can save us years of waiting for our plants to reveal their hybrid origin (or not).