Genetic variation of proteins (protein polymorphism) is widespread among many animal species. The biological significance of protein polymorphism has been the subject of many studies. This variation has a supporting function for population genetic studies as a source of genetic markers. In farm animals it is used in population genetics and in parentage control, mostly together with blood group polymorphism.
In this study polymorphism of blood proteins in a population of Shetland ponies was investigated to search for (natural) selection contributing to the maintenance of protein polymorphism. Genotype distributions were studied in relation to genetic equilibrium and segregation. The pony population served as a model for a population with the structure of a livestock population but with less artificial selection for quantitative traits. In horses and especially in Shetland ponies there is a high level of blood protein polymorphism.
A random sample of 280 foals and their parents was used. The structure of the population was included to yield an unbiassed estimate of the genotype distributions expected under genetic equilibrium conditions. Seven protein systems were investigated (prealbumin, transferrin, albumin, esterase, haemoglobin, carbonic anhydrase and catalase). They are controlled by alleles at autosomal loci, almost exclusively being codominant. The protein variants were distinguished with starch gel electrophoresis, a traditional procedure for this purpose (Chapter 4). Some of the methods used in the typing of protein variants were modified to increase their resolving power and reliability.
Chapter 2 describes literature on the biological significance of genetic polymorphism (proteins and blood groups) in various animal species. A crucial question is whether polymorphism is subject to selection. There are theoretical, inventorial and experimental approaches.
Only a few links exist between polymorphism research in farm animals and in other species. The study of polymorphism in farm animals was stimulated by the search for relations between protein variants and production traits. Although such relations were observed, especially connected with reproduction, no definite conclusions could be drawn. The results with different breeds were contradictory, and there was a lack of data on the structures of populations sampled. In horses deviations from expected genotype distributions in blood protein systems were scarce and consisted mainly of excesses of albumin heterozygotes in segregation.
In theoretical approaches neutrality of protein polymorphisms was emphasized. Recent observations revealed a large-scale and quantitatively comparable rate of polymorphism in various species (Drosophila,
mouse, man) with comparable degrees of average heterozygosity for proteins, including enzymes. Experiments on enzyme polymorphism contributed to the understanding of selection in the maintenance of polymorphism. For a long time heterozygous advantage had been considered as most important in this respect. More recently other selectional mechanisms (e.g. frequency-dependent selection) have been proposed as effective in the maintenance of polymorphism.
In my study the population structure was included for a valid evaluation of the results. Because no quantitative description of the structure of the Shetland pony population was available, this structure was investigated.
Chapter 3 describes the breeding structure of the Shetland pony population in the Netherlands. This population includes almost 9,000 mares and about 220 stallions. It was only recently closed except for immigration of British ponies. Two stallions had been especially important: Bartje, born in 1945 and the British stallion Spotlight of Marshwood, imported in 1956. They had an average genetic relation of 0.07 and 0.10 to the foals born in 1973. The breeding area was subdivided into two central regions and four peripheral regions to investigate the geographical structure of the population. At least 69% of all stallions migrated during their lifetime to a region other than where they were born, mainly from the centre to the periphery. They were almost exclusively used within the regions in which they were situated. The concept of breeding regions was checked with the gene frequencies of protein systems and used in the evaluation of genotype distributions.
The average inbreeding coefficient of foals, born in 1973 was 0.0125, this estimate being based on pedigrees in 5 generations. Inbreeding via Bartje contributed 10% to this figure and the contribution via Spotlight was 29%. The relatively low inbreeding coefficient could be explained by the open character of the breeding population, including the import of British stallions after 1954.
Chapter 5 gives the results on protein polymorphism. Before the investigation of genetic equilibrium and segregation, gene frequencies were analysed. This analysis was done because a stratification in the material caused by more or less isolated breeding regions may influence the genotype distributions expected over the total breeding area. Also gene frequency differences between parents may influence these distributions. Gene frequencies for all protein systems were calculated separately in all breeding regions for foals, sires and dams. There was no homogeneity in their distribution over the total area for foals at 5 loci, for sires at 3 loci and for dams at 2 loci. The differences were not associated with the stallions' migration pattern. Based on the differences the expected genotype distributions were calculated separately for each region and pooled afterwards in the investigation of genetic equilibrium. This procedure led to an increase in the expected numbers of homozygotes of about 2%, which proved to exert a significant effect upon the final result. There were also gene frequency differences between the foals' parents, possibly leading to an excess of heterozygotes in the foals. This excess was estimated but proved to be negligible (i.e. on average < 1 %).
The investigation of genotype distributions included genetic equilibrium and segregation. All loci were considered separately, and moreover the homozygous or heterozygous genotypes of all loci were pooled to estimate the observed and expected average homozygosity with equilibrium and in segregation.
At most loci the observed genotype distributions agreed with the genetic equilibrium distributions in foals, dams and sires. There was a significant excess of heterozygotes of haemoglobin in dams. In segregation there were significant deviations for prealbumin, involving 3 out of 6 prealbumin alleles, and for albumin. The deviations in albumin consisted of an excess of homozygous foals from homozygous dams.
After pooling homozygous or heterozygous genotypes of all seven loci, a significant excess of average heterozygosity was observed in dams compared with genetic equilibrium, and with segregation a significant excess of homozygosity was found in offspring from dams with a high homozygosity. A hypothesis was forwarded to connect both types of deviations obtained after pooling of loci, but there was no evidence of a link between them. The deviations from genetic equilibrium and in segregation were such that they could be considered as compensatory.
Chapter 6 tries to explain the observed deviations. In literature on horse protein polymorphism only separate loci are dealt with. The excess of like-dam homozygotes in albumin was contradictory to most reported observations, where excesses of heterozygotes in segregation of albumin were mentioned. An excess of heterozygous offspring for haemoglobin was observed earlier, but not in Shetland ponies. There was no apparent relation between the deviations and the functions of proteins as far as these were known.
The deviations in my material might result from selection for fitness, which selection is likely to occur in the pony population. Especially the dams in the sample could be subject to selection for fertility. The excess of heterozygosity in dams may be related with selection for fitness, and might contribute to the maintenance of polymorphism. However, there were opposite deviations to compensate for this possible selection.
It was concluded that in most protein systems there were deviations from genetic equilibrium distributions or in segregational ratios. The causes of these deviations could not be revealed. Their net effect on the maintenance of polymorphism was most likely neutral.
In farm animals artificial selection largely influences the genetic composition of populations. The effect of deviations of the nature and magnitude as observed in the pony population on the maintenance of polymorphism will be negligible in livestock. The study of protein polymorphism in livestock can be used to describe changes in genetic composition of populations, for which purpose the simultaneous use of various polymorphisms and the concept of average homozygosity may be promising.