Maize News Letter 77: 30 (2003)
Freising, Germany: Bavarian State Research Center for Agriculture
Chisinau, Moldova: Institute of Genetics

Possible effects of heterofertilization on the induction of maternal haploids in maize
V. Rotarenco, J. Eder

Up to now the mechanism of the induction of haploids in maize by the use of special inducer-lines still remains insufficiently studied. The prevailing hypothesis suggests that the main cause for the development of seeds with a haploid embryo is a single fertilization. According to this hypothesis, a fertilized central nucleus, dividing, stimulates the unfertilized (haploid) egg cell to develop (Chase, 1969; Tyrnov, Zavalishina, 1973; Hohlov, Grishina, 1976; Enaleeva, 1992). The regular double fertilization is distorted after the pollination with pollen of a haploid inducer line.

Bylich and Chalyk (1996) studied the morphological structure of the sperms in the pollen grains of ZMS haploid inducer line. 6.3% of the pollen grains were found to have two sperms morphologically different from each other. According to the authors, morphological defects of one of the sperms influence its functional properties, which often leads to a single fertilization. This, in its turn, results in maternal haploids. It was suggested that the frequency of haploids is, to some degree, related to the frequency of pollen grains of the inducer line having some kind of morphological defects in one of the sperms.

Obtaining haploids every year, we noticed that after natural and artificial pollination their percentage is substantially different. In an isolated plot with free pollination from the inducer line we obtain 2 to 4% of haploids, regardless of the genotype, while after artificial pollination (pollen is applied on ears isolated by bags only once by hand) of the same genotypes, the percentage of haploids is twice and even three times as high.

In our previous research, comparing the output of the haploids after artificial and natural pollination, we concluded that a possible cause of this difference is a delay of pollination (Rotarenco, MNL 76, 2002). But further research in this field showed that the underlying cause is probably different. A possible reason for this difference is assumed to be heterofertilization.

We carried out an experiment comparing the frequency of heterofertilization using MHI haploid-inducer and line X28C, not having haploid-inducing ability. Both lines possess a dominant R1-nj gene (anthocyanin coloration of the top of the endosperm and embryo). The presence of this gene in a haploid-inducer-line makes it possible to select maternal haploids at the level of dry seeds.

In the greenhouse, we planted a homozygous line 092 not possessing any marker genes, but which, as had been discovered earlier, expresses the R1-nj marker gene well when crossed with a line-carrier of this gene. We prepared two pollen mixtures to pollinate the 092 line — the first mixture consisting of 50% X28C pollen and 50% 092 pollen. The other mixture consisted of 50% pollen obtained from the MHI inducer line, and 50% 092 pollen. The proportion was based on the pollen weight. Each of the pollen mixtures was used for the pollination of three plants of line 092. Each ear pollinated produced over 150 seeds.

After harvest we calculated the percentage of heterofertilization (anthocyanin coloration of the top of the endosperm and white embryo, no anthocyanin coloration of the top of the endosperm and colored embryo). When the first pollen mixture was used for pollination (X28C and 092), on average, 1.5% of heterofertilized seeds were observed. In the second case (MHI and 092), we obtained on average 5.3% heterofertilized seeds.

In the second case, when the inducer pollen was used, seeds with haploid embryos were to be expected. To detect the haploids, we germinated all seeds with colored endosperm and white embryo, both after pollination with the first pollen mixture and after pollination with the second one. Using cytological analysis, we calculated the number of chromosomes in their root meristems. The haploids were found only in the second case, when MHI inducer pollen was used, and the percentage of haploids was on average 7.5% (with respect to the seeds with expressed marker gene). Haploids were not counted in the percentage of heterofertilization, mentioned above.

Thus, when the pollen mixture containing the MHI inducer as a pollinator was used, the percentage of heterofertilization was more than three times as high as in the case when the pollen mixture containing line X28C was used. Most likely, the increase in the percentage of heterofertilization is due to the single fertilization, caused by different defects of one of the sperm of the inducer line, as mentioned before.

It is supposed that the main reason for the decrease in the frequency of haploid seeds in the case of natural fertilization could be heterofertilization. Apparently, more pollen tubes reach the embryo sac applying natural pollination, than during the artificial one. So the probability of heterofertilization is higher, which leads to the observed decrease in the percentage of haploid seeds.

This can be explained by the fact that during natural pollination the pollen that falls on a stigma has a good capability of germination, as there may be an opportunity for its renewal during the whole day. In the case of artificial pollination, there is no opportunity for pollen renewal, so, most likely, fewer tubes reach the embryo sac. So, the probability of heterofertilization decreases, and in the case of single fertilization, the probability of the formation of seeds with a haploid embryo increases.

Of course, further, more extensive studies in this field are needed, but, as we already believe, heterofertilization is an interesting factor related to the mechanism of haploid induction and can have an effect on the percentage of maternal haploid plants induced.

CybeRose note: It appears to me that pollen-tube avoidance may be at play. A pollen tube would waste its one chance at life if it traveled down to an ovum that had already been visited by a normal pollen tube. But some time later, after the first pollen tube has died, a second might arrive to complete the double-fertilization that a partially defective pollen tube only started.

If this is the mechanism, it might be exploited for raising hybrids with tripsacum. In that way, some of the kernels will have a hybrid embryo with a normal triploid endosperm.

On the other hand, this hypothetical model seems to conflict with the observation of heterofertilized seeds resulting from a pollen mixture that did not involve a haploid-inducing type.

I still wonder whether the single nucleus is equally probable to fuse with the central cell as with the ovum. If there is a preference, this mght also be found in wide hybrids. And if the foreign pollen should happen be more drawn to the central cell than to the ovum, we might find seeds without embryos — which have actually been reported in some cases. In other cases, a foreign pollen might favor ova over central cells. Considering these speculative possibilities, it would certainly be desirable to pollinate repeatedly, and possibly helpful to use pollen mixtures.

Maize breeding bibliography