Bulbsociety List
Date: Mon, 28 Jul 2003 12:44:16 -0700 (PDT)

Segregation Distortion and the Law
Karl King

Over the past century — since the rediscovery of Mendelian segregation announced in 1900 — various researchers have found cases that diverged significantly from the "normal" pattern.

Correns, one of the rediscoverers, found an example in maize. Mangelsdorf and Jones studied this case in detail and found that certain lines and specimens gave very significant excesses of defective kernels, while others gave similarly significant deficiencies. The trait is recessive and should be seen in 25% of the kernels of a plant heterozygous for it. They wrote, "An F2 ear with 32.4 percent recessives produced F3 progenies with an average of 23.8 percent recessives. At first glance it appeared that the deviation in F2 was not inherited. Closer examination, however, shows that although the average percentage of defective seeds in F3 is almost normal, the individual progenies show marked deviations, some in one direction, some in another."

And even closer examination revealed that the discrepancies were much greater in the bottom halves of the ears than in the top halves. The "segregation distorter" in this case was a separate gene on the same chromosome that prevented pollen grains possessing it from growing tubes long enough to reach the lower ovules of the ears. Thus, if the defective kernel gene and the short pollen tube gene are on the same chromosome, there will be a deficiency of defective kernels. But if the two genes are on opposite chromosomes in a double heterozygote there will be an excess of defective kernels (deficiency of normals).

In 1905, Cuenot reported on his study of the non-Mendelian pattern of inheritance of yellow coat color in mice. The trait is lethal when homozygous, so yellow x yellow should give about 67% yellow by Mendelian calculation. Cuenot got 72.45%. Some years later Castle and Little acquired a related strain for their own study and came up with 64.77% yellows. That's closer, but they decided to sweeten the pot by combining their results with Cuenot's to get 66.52%.

The problem here is the same as in the case of defective maize kernels. Results from different lines and different generations had to be combined so that the very real discrepancies could be averaged out of existence.

Another segregation distorter was discovered in fruit flies in 1956. Numerous white-eyed females were mated to red-eyed males heterozygous for the white-eye gene. A few of the pairs yielded only red-eyed progeny. Obviously some "gene" was present in those fathers that "promoted" the chromosome carrying it into all the sperm nuclei that managed to fertilize an ovum. Crossing over is suppressed in male fruit flies, so if the red-eye gene happened to be on the chromosome carrying the segregation distorter, it got promoted as well. It was estimated that 3-5% of wild fruit flies possess such a gene.

ASPB Plant Biology Meeting 2002 Fourth Day Briefing, posted 08-13-2002
Ken Feldmann opened his seminar by defining the phenome as the quantitative identification of the form and function derived from genes, but lacking a quantitative, integrative definition. ... He also described work performed to analyze segregation distortion mutants, which are lines that have non-Mendelian segregation ratios for resistance to the marker present in the T-DNA insertion. His group has found that 4% of the lines exhibit segregation distortion, and that for half of these, the altered resistance ratios are passed on to subsequent generations. About 1,000 genes appear to be involved in segregation distortion. Finally, Feldmann described the highly automated approach that CropDesign, Inc. is using for phenotyping of rice.

Thus far, segregation distorters appeared to be exceptions to the general rule of Mendelian segregation. But more recent research casts further doubt on the validity of Mendel's "law". At the annual meeting of the American Society of Plant Biologists, Aug 06, 2002, Ken Feldman of Ceres, Inc. reported on some Arabidopsis thaliana research. "He also described work performed to analyze segregation distortion mutants, which are lines that have non-Mendelian segregation ratios for resistance to the marker present in the T-DNA insertion. His group has found that 4% of the lines exhibit segregation distortion, and that for half of these, the altered resistance ratios are passed on to subsequent generations. About 1,000 genes appear to be involved in segregation distortion."

That's 1000 SD genes in Arabidopsis thaliana alone.

It appears that segregation distorters may be the "law", but that their combined actions, blurred by crossovers and averaged over several generations, create the illusion of tidy Mendelian segregation. It may be significant (or not) that fruit fly research suggested 3-5% of wild specimens carry a segregation distorter, and that Feldman reported on 4% of Arabidopsis thaliana lines having one. Certainly more research is needed, but it could be that the actions of numerous segregation distorters "average out" in the majority of cases so that their individual effects are not easily identified. But in a minority of specimens or lines the unitary influence of one or more SDs is exposed. After all, SDs are as likely to be linked as any other "genes", especially when there are so many known examples. Two SDs of opposite effect may be linked into a Mendelizing unit until separated by a rare crossover into a pair of SD "mutations".

However, Ky et al. (2000) found segregation distortion in 30% of the loci studied in Coffee. The plants in this case were tetraploid, which can introduce other problems for simple Mendelian inheritance.

The important thing to note is that segregation distortion is a genuine phenomenon backed by extensive research. While it may be rarely observed overall, it will be common in certain lines and select specimens.

Furthermore, the phenomenon may become even more pronounced in hybrids. In the January 2002 report of the Plant, Animal & Microbe Genomes X Conference, Faris, Haen and Gill reported on the "Genomic analysis of segregation distortion and recombination on Durum chromosome 5B." They wrote, "Distorted segregation ratios of genetic markers are often observed in progeny of inter- and intraspecific hybrids and may result from competition among gametes or abortion of the gamete or zygote." And that, "detailed analysis of Aegilops tauschii chromosome 5D indicated that it possessed at least three different segregation distortion loci that conferred gametophytic competition among pollen when an F1 plant was used as a male parent."

H. I. Oka of the National Institute of Genetics (Japan) wrote (Rice Genetics Newsletter vol. 6 p. 83. Dec. 1989), "It is well known that in hybrids between glutinous and non-glutinous rice varieties, glutinous homozygotes tend to be deficient in the F2, often being significantly less than 25%. This trend is particularly noticeable in distant crosses."

Fishman, et al. (2001) reported on hybrids of Mimulus guttatus and M. nasutus. "Nearly one-half (49%) of the 255 markers genotyped in our F2 mapping population deviate from the Mendelian expectation of 3:1 or 1:2:1 genotype ratios (for dominant and codominant markers, respectively) at a=0.5. Nearly one-third (31%) show significantly distorted genotypic ratios at a higher threshold (a=0.001). The codominant markers have the highest proportion of loci with distorted ratios (66 and 47% at 0.05 and 0.001, respectively).

Zeller found a possibly related phenomenon in wheat-rye hybrids. "Recently we reported in numerous spontaneous chromosome substitutions and translocations of different origin the preferred tendency of a specific rye chromosome (1R) to replace the complete wheat chromosome 1B or a part of it. We proposed to designate this behaviour of chromosomes of the sub-tribe Triticinae preferential substituting ability (Zeller 1973)."

I have not read the report, but it looks like another example of a hybrid population reverting to one of the parents while retaining a chromosome or part of a chromosome from the other. In wide hybrids we cannot assume that all the chromosomes of the parents will be equally similar or different. That is, both parents may share one roughly homologous chromosome yet differ substantially in all other chromosomes as a result of different chromosomal rearrangements that have occurred since the two species separated from a hypothetical common ancestor. In such cases the tendency will be for hybrid progeny to revert more-or-less quickly towards one or the other parent, except for traits associated with the nearly homologous chromosomes.

And where partial hybrids occur with some frequency (e.g., where cells with complete sets of parental chromosomes are incapable of differentiating as embryos) there will be a similar tendency to revert towards the parental types except for any equivalent (interchangeable) chromosomes or segments.

If it is not already apparent, environment can have a powerful influence over segregation distortion — on competition among pollen as well as on the preferred pattern of chromosome loss in partial hybrids.

For example, Gladiolus tristis is pollinated by night-flying moths. At night. Its pollen tubes grow most rapidly in the range 15-20°C. Other Gladiolus species are pollinated by bees during the day when the ambient temperature is higher than G. tristis pollen tubes can endure. If G. tristis were crossed with one of the bee-pollinated species, we may expect to find segregation in F1 pollen for temperature restrictions on pollen tube growth — and segregation distortion for any "genes" linked to those factors, varying with temperature.

The same general principle must apply to partial hybrids resulting from loss of chromosomes. Even where initial chromosome loss is random, survival of different cell-lines within the embryo would not be random, especially when the parents are adapted to widely different environmental conditions. The common (though far from universal) tendency of partial hybrids to follow the seed parent may be "distorted" where the seed parent carries a specific genetic defect (e.g., Van Tubergen's hybrids of Lilium martagon var. album x L. Hansoni. Several hundred of the offspring resembled the seed parent exclusively, except that none had white flowers) or where the seed parents are enduring inhospitable conditions (e.g., Christy Hensler's hybrids of unhappy Japanese iris pollinated by thriving Siberians. Some of the seedlings resemble the Japanese seed parent very closely except for their tolerance of hot, dry sandy soil inherited from the pollen parent.)

Simply stated, segregation distortion is natural selection operating at the level of gametes. Somewhat similarly, partial hybridization often involves natural selection acting between segregating cell-lines within embryos. In either case, selection will be governed by immediate conditions rather than the long-range needs of the organisms.