Horse Genetics. Ernest Bailey

suggested that at least one of the mutations that causes a white coat could cause embryonic losses. In matings between white horses, both solid color (non-white) and white foals were always produced, demonstrating that the gene for white coat color (W) was dominant to the non-white gene (w). However, the ratio of white to colored foals (28:15) from matings of heterozygous white horses (W/w) did not correspond to the expected ratio of 3:1 that would be expected in matings of known heterozygotes (Table 5.3). In this case, the results more closely approximated a 2:1 ratio. This could be explained if homozygous white (W/W) offspring are lost as embryos, and only the heterozygous white and the non-white offspring make it to term, as shown in Table 5.3. Consistent with this observation, the authors of the study did not observe true-breeding white horses (homozygotes) in their herd (Pulos and Hutt, 1969).

Table 5.3. If all classes of offspring are equally viable, the proportion of white to colored offspring from matings between white horses heterozygous for the dominant White gene W is expected to be 3:1., i.e. 3 white, 1 colored. If one homozygous class is lethal (example, 25%), then the proportion of white to colored would be 2:1. Relatively large numbers of offspring might be needed to distinguish between a 3:1 and a 2:1 ratio.

       Partial dominance among alleles

      Partial dominance, incomplete dominance, and semi-dominance are terms used interchangeably. Essentially, these terms mean the presence of just a single allele affects the phenotype and the presence of two alleles has a greater impact. A good example of this is the effect on hair color by the the dilution allele (CR) at the Cream locus (C). When a chestnut horse has one copy of the CR allele, e.g. heterozygous CR/cr, it appears as a palomino. The rich, red chestnut coat is diluted to a light cream color and the mane and tail are very pale. When a chestnut horse has two copies of the CR allele, the dilution is even greater with the horse appearing almost white. Chestnut horses homozygous for CR are called “cremellos” (Chapter 8). The effect of the variant may be additive, depending on how the phenotype is measured.

       Co-dominance among alleles

      Co-dominance is a lot like partial dominance in that when the gene is present it is expressed. However, it does not have an impact on its co-dominant allelic partner. Blood group markers and SNPs are co-dominant. Their presence can be detected whenever they are present, but they do not have an effect on the expression of their alleles. Most of us are familiar with the ABO human blood groups. The ABO blood group is encoded by a single locus. One of the alleles makes the A blood group substance. Another makes the B blood group substance. When the A factor is present it is detected. When the B factor is present it is detected. The detection of the A phenotype has nothing to do with the detection of B. When neither A nor B are present, they are not detected, and the individual is referred to as type O.

      However, the ABO system is also a good example of the nature of recessive and dominance interactions. Since there is no O protein (O is a null allele) there is no way to distinguish a homozygote for A (genotype A/A) from a heterozygote for A and O (genotype A/O). As a result, the allele for O is recessive to the allele for A. Therefore, the alleles for A and B are co-dominant but the allele for O is recessive.

      In summary, dominance, recessive, partial dominance, and co-dominance are descriptions of the interactions among alleles.

       Expected ratios, statistical tests and alternative models

      Learning to predict and recognize simple 1:1, 1:2:1, 3:1, and 2:1 trait ratios among offspring is extremely useful for building genetic models of trait transmission. These ratios may be complicated, as we have seen, by the interactions of more than one gene, and may not be obvious unless large numbers of offspring are available. Statistical testing (e.g. the chi-square test) may be needed to determine whether the observed ratios match the expected ratios. (Consult a basic genetics or statistics text for information on how to apply such tests.)

      If the data obtained do not match expected ratios, then alternative proposals need to be considered, such as:

      1. The trait is genetic, but the hypothesized transmission mechanism is incorrect (e.g. more than one gene may be involved).

      2. The trait is produced by environmental influences, not genes.

      3. The gene shows “reduced penetrance” (i.e. the phenotype is modified by environment or other gene combinations, so that the effect of the mutant gene is not easily recognized).