Molecular genetic techniques
C. Meghen, D.E. MacHugh and D.G. Bradley
The authors can be contacted at the Department of Genetics, Trinity College, Dublin 2, Ireland.
West Africa differs from the rest of the continent in that significant populations of both subspecies of domestic cattle, Bos taurus and Bos indicus, are found there. Most of Africa is populated with zebu breeds and varieties of stabilized crossbreeds, collectively known as sanga. It is only in regions of West and Central Africa that large numbers of apparently pure taurine breeds remain. These breeds, including N'Dama, Kuri and varieties of West African Shorthorn, are the descendants of the original domestic cattle of the continent and represent a genetic resource unique to this portion of the globe (Lhoste, 1991) (Figure 1).
It now seems likely, in the light of modern archaeology and the molecular results described below, that cattle were originally domesticated in at least two primary locations approximately 10 000 years ago. One of these was centred in the nascent civilizations of the Near East and the other probably in northern India and Pakistan. These utilized two different strains of the wild ox, Bos primigenius, and gave rise to the two modern cattle types.
The first cattle in Africa likely migrated from the Near East through Egypt, possibly interbreeding with local wild variants, and were humpless, or taurine, in character. These animals and their herders moved throughout the western subcontinent, around a much-reduced Sahara, some 7 000 years ago. Through natural selection resulting from a long association with the natural challenges of humid Guinean zones, these cattle have become adapted to living with endemic diseases. In particular, some breeds possess trypanotolerance or varied levels of adaptation to the tsetse fly-transmitted disease, trypanosomiasis.
Humped, or zebu, cattle have a more Eastern origin and, although first appearing in Egypt in ancient times, only started to spread comprehensively through West Africa along with Arabic influences within the last 1 400 years. This subspecies shows marked adaptation to arid environments and predominates in those areas of the world with Sudano-Sahelian pasture ecologies. They possess little or no trypanotolerance and, consequently, may be maintained in much of Africa only through the use of expensive disease-prevention measures such as tsetse control or chemical prophylaxis.
1 Kuri cattle off the Island of Tchongolé in Lake Chad - Bovins Kuri à proximité de l'île de Tchongolé (lac Tchad) - Vacunos Kuri frente a la isla de Tchongolé, en el lago Chad
2 Collecting blood samples from Kapsiki cattle in the Kapsiki Region of northern Cameroon - Collecte d'échantillons de sang de bovine Kapsiki dans la région de Kapsiki dans le nord du Cameroun - Obtención de muestras de sangre de un vacuno Kapsiki en la región de Kapsiki, Camerún septentrional
At the moment, taurine herds predominate in the more southern, humid ecological zones of West Africa, inhabiting an almost continuous swath from the Gambia to Cameroon. However, they are significantly threatened with assimilation or extinction from the incursion of the more numerous zebu into these regions. Because of the attraction of the larger indicine bulls to pastoralists, and because of migrations largely the result of droughts such as those between 1972 and 1983, many taurine populations are decreasing on a scale greater than ever before. Between 1970 and 1981, in the four Sahelian countries of Mali, Mauritania, the Niger and Senegal, the bovine to human ratios fell by 30 to 40 percent, a statistic not parallelled in the more southern countries. The future utilization of the genetic resource represented by the trypanotolerant taurine to harvest the biomass of much of humid sub-Saharan Africa depends on their conservation, promotion and improvement (Bradley et al., 1993a; Philipson 1992). Genetic characterization of these breeds and their traits form one important component in a strategy to expand food production in the tsetse-infested third of the continent's land mass.
It is necessary at this point to decide exactly what is meant by the term "breed". The animal-oriented definition recognizes that breeds differ by the totality of average differences observed in many quantitative and qualitative traits. The differences may overlap, but they have a genetic basis, and these differences taken together provide a unique description (Baker and Manwell, 1991). This definition provides a solid basis for the application of population genetic techniques and is in stark contrast to the arbitrary and often colloquial designation of a particular type to one breed or another based on anthropocentric criteria.
It is instructive to make a distinction between those genetic markers (described below) that are used as a tool to identify and locate useful traits and those that are used to describe population structures. Although the same marker systems can be utilized for the two different approaches, the experimental design and analysis of data differ greatly between studies of production traits and those of populations.
The characterization of production traits in cattle using molecular markers is currently a very active area of research. The identification of such markers requires the examination of either large pedigrees or populations that are carefully phenotyped. Markers themselves may be polymorphisms within genes that are suspected to actually confer a particular trait (candidate genes) or, more commonly, they may be anonymous regions of the genome that are examined with the expectation that they will be adjacent, i.e. linked, to genes of importance. One approach is to look for a correlation between differences in marker allele frequencies and trait variations in selected isolates. Care must be taken that the origins of trait-positive and trait-negative groups are similar, however. For example, a difference in marker frequency between populations of trypanosusceptible zebu and trypanotolerant taurine cattle would be a result of the distant relationship between the two subspecies rather than a measure of any particular biological character. Another approach attempts to correlate the patterns of segregation of marker and trait in well-defined pedigrees. This is a sensitive and often resourceintensive approach, but it is not subject to the errors mentioned above. A large project of this nature using trypanotolerant N'Dama and susceptible Boran progenitors is under way at the International Laboratory for Research on Animal Diseases (ILRAD), Nairobi, Kenya (Kemp and Teale, 1991).
Population-level characterization is distinctly different from that of trait-based characterization. A primary aim of the population geneticist is to detail in an unambiguous and often numerical manner parameters that are useful in describing a population. These parameters are quite varied and rarely does any one investigator need to assess all of them (for a review see Hillis and Moritz, 1990). In the case of West African cattle, the aspects of particular interest include genetic distance between breeds, genetic variation within breeds, gene flow and admixture between breeds. To pursue the population genetic approach, then, it is often most desirable to work with subsamples taken at random from each population or breed.
To characterize the cattle breeds of West Africa, genetic markers extracted from blood samples isolated on-site must be used (Figure 2). These markers should be selectively neutral, so that observed similarities are attributable to common ancestry and not convergence. The markers of choice must also be polymorphic, thereby enabling the quantification of variations between and within breeds and further to provide a method of detecting the genetic admixture between breeds.
Traditionally, external or internal phenotypic characters have been used to ascribe a given animal to a breed. These extremely varied characters have ranged from skin colour to size of testes, from the presence or absence of the bifid process to milk butterfat content. One of the problems associated with this type of analysis is the difficulty in combining different measures in order to provide a useful tool for the description of a given breed. Objective interbreed comparisons, especially across broad geographical ranges, have not been attempted using this kind of data. Because of the influence of both natural and artificial selection on phenotypic traits, it is unlikely that any such comparison would produce meaningful results. There are other techniques such as blood typing, karyotyping and immunological assays that have been used to provide comparative data, but, as with gross external and internal phenotypic characters, the data lack the power to resolve the differences between closely related cattle breeds. It would be ideal to be able to test the genome directly and, fortunately, with the advent of several new molecular genetic techniques, this has become possible.
What follows is an overview of some of the techniques currently available to the molecular geneticist that have particular relevance to population studies. Some techniques will not be described here as they have little practical application in the study of genetic variation within and between cattle breeds. These include multilocus and single locus genetic fingerprinting, as well as VNTRs (variable number of tandem repeats).
A widespread method in population genetics, a protein solution is electrophoresed through a gel and an enzymespecific reaction highlights-one locus whose alleles (allozymes) may migrate different distances because of differences in charge (Queller, Strassmann and Hughes, 1993). Protein polymorphisms, although still widely used in population studies, are of limited value in the assessment of genetic variation at the level of cattle breeds. This is largely because of the relatively low levels of polymorphism found in protein loci, resulting in a lower taxonomic limit to the resolving power of protein electrophoresis. Manwell and Baker (1980) extracted data from nearly 1 000 papers from which they analysed ten polymorphic loci in ten major breed groups. They were able to distinguish each group based on the analysis of the protein data, but within the groups homogeneity was noted. In West Africa, the interest is in the variation present within breed groups, therefore, allozyme data is not particularly useful. There are many drawbacks to the general use of protein electrophoresis, and these have been described elsewhere (Hillis and Moritz, 1990).
A study of restriction fragment length polymorphisms (RFLPs) involves comparing the number and size of deoxyribonucleic acid (DNA) fragments produced by the digestion of DNA with various restriction enzymes. The banding pattern is generated by the presence or absence of restriction cutting sites and these in turn are produced by mutation. The fragments are blotted to membranes and probed with cloned radiolabelled DNA that binds to a single locus. This technique can be applied to nuclear DNA or to mitochondrial DNA (also to chloroplast DNA in the case of plants). It has applications in the study of genetic distance, population variation, gene flow, effective population size, patterns of historical biogeography and analyses of parentage and relatedness. Since mutational events are generally the product of base substitutions, however, the rate of mutation is likely to be extremely low (10-7 to 10-8 per generation), and this results in a similar problem to that of proteins, which is, a lack of resolving power when dealing with very closely related groups. This has been demonstrated by Theilmann et al. (1989), who
carried out a study of nine RFLPs in six breeds of cattle, with the result that only Brahman (Bos indicus) cattle differed significantly from the other breeds, which were all Bos taurus.
Deoxyribonucleic acid sequencing can be used to address virtually any systematic problem and many problems in population genetics, but it is not necessarily the most efficient or cost-effective method. Most typically, DNA sequence data is used to construct molecular phylogenies (Figure 3), and the presupposed distance between taxa affects the selection criteria of the genomic region to be sequenced. If two organisms are distantly related, then conserved sequence regions should be used. Conversely, genetic distance between closely related species is determined by using variable sequences. Once again, however, the genetic difference between cattle is low enough to make direct sequencing an unlikely candidate for population analyses. In order to uncover the level of variation needed to distinguish breeds, the stretches of DNA to be sequenced would be prohibitively long. Additionally, when trying to assess such parameters as intraspecific hybridization, geographical variation or individual relatedness, large sample sizes are required, yet DNA sequencing at this time is not sufficiently automated to allow the routine screening of large samples.
Simple tandem repeats (STRs), or microsatellites, are a relatively new class of genetic marker. In the past few years, they have become the marker of choice for gene mapping and are increasingly being used for population studies (MacHugh et al., in preparation; Wall et al., 1993). Microsatellites consist of tandem repeats of very short nucleotide motifs from one to six base pairs long, the dinucleotide repeat CA being the most common in mammalian genomes. A typical microsatellite locus may consist of a stretch of DNA with the base sequence CA repeated 12 times, i.e. (CA)12. When the unique sequence flanking both ends of the repeated sequence is known, the microsatellite can be preferentially amplified using the polymerase chain reaction (PCR). Different length classes (alleles) vary in the number of repeats and can be separated using polyacrylamide gel electrophoresis (PAGE) (Figure 4). This class of marker is highly polymorphic, displaying many different alleles for a given locus.
Recent studies have shown that the mutation rate is in the order of 10-4 per generation (Weber and Wong, 1993; Dallas, 1992). This means that, although new length classes are generated at a rate fast enough to allow the distinction of breeds, the rate is not so fast that relationships are obscured by homoplasy (identity of alleles as a result of separate mutation events as opposed to common ancestry). The primary disadvantage of this technique is that a prior knowledge of the DNA sequence is required to allow the design of PCR primers. This causes the greatest problems when trying to use microsatellites in species for which there is little or no sequence data. There are public-domain databases of accumulated sequence data, such as GenBank and EMBL, however, where cattle and other economically important species are well represented. These can be searched for microsatellite sequences. Alternatively, genomic libraries can be probed for STRs and sequenced directly.
3 D-loop sequence from typical Bos indicus and B. taurus cattle - La boucle-D de bovine Bos indicus et B. taurus typiques - Secuencia del lazo D de vacunos de los tipos Bos indicus y B. taurus
Schematic representation of the amplification of a microsatellite using PCR and the separation of the different allele length classes on a gel - Représentation schématique de l'amplification d'un microsatellite par la méthode ACP et de la séparation des différentes classes de longueur des allèles sur un gel - Representación esquemática de la ampliación de un microsatélite utilizando la reacción en cadena de la polimerasa y la separación de los distintos grupos de longitudes de los alelos en un gel
Random amplified polymorphic DNA (RAPD) is another recently identified genetic technique (Williams et al., 1990). Although levels of genetic variation are high and a prior knowledge of DNA sequence is not necessary, this technique has not yet been widely accepted as a tool for population studies. There are a number of reasons for this. First, it is not yet fully understood how the genetic variation observed is generated and, consequently, it can be difficult to reconstruct evolutionary histories. Second, consistency of results is not guaranteed as minor differences in experimental conditions can produce erratic results. Even under carefully controlled conditions there can be ambiguity in the scoring of bands separated on a gel. Third, RAPDs are dominant markers and heterozygotes are typically scored as homozygotes, which decreases their information content.
From the preceding discussion it is apparent that there are various techniques available to the population geneticist and that some of these are better suited to the study of West African cattle than others. At this point, however, it is useful to outline a totally different aspect of the description of cattle breeds. Not only can different techniques be used to describe the variation present, but parts of the genome that have distinct modes of inheritance can also be sampled. These are autosomal, mitochondrial and Y-chromosomal DNA. For the remainder of this discussion it shall be shown how a synthesis of different techniques, as applied to different parts of the genome, can provide a great deal of information about the genetic structure of cattle populations with particular reference to those in West Africa.
Animal mitochondrial DNA is easily isolated as it is present in high copy number in most somatic cells. It evolves five to ten times more rapidly than nuclear DNA, and a particular region, the D-loop, evolves even faster. It is maternally inherited, without recombination, and therefore can provide an unambiguous maternal phylogeny. Mitochondrial DNA has found a special place in the population geneticist's repertoire of molecular tools. Although a mitochondrial phylogeny offers a relatively clear picture of the evolutionary history of a single genetic element, it is only one component of the entire organismal genealogy. The strictly maternal inheritance of mitochondrial DNA can cause misinterpretation of data and consequently the misreading of resultant phylogenies.
One of the approaches used in a genetic distance study of seven tropical and six temperate breeds of cattle was that of RFLP and sequence analysis of mtDNA (Loftus, 1992; Loftus et al., 1992). The breeds included African Bos taurus (N'Dama), African Bos indicus (Fulani, Butana and Kenana), Indian B. indicus (Hariana, Tharparkar and Sahiwal) and European B. taurus (Aberdeen Angus, Charolais, Friesian, Hereford, Jersey and Simmental). From the RFLP data, two mitochondrial lineages were observed, one representing European and African mitochondria, including the zebu breeds, and the other representing Asian B. indicus mitochondria (Figure 5). Within the European and African groups, no further substructuring could be detected using this technique. Complete mtDNA sequence analysis for 26 individuals two from each of the 13 breeds - verified the dichotomy between the Asian and Afro-European clades but failed to provide enough sequence variation to significantly separate the European and African clades (Figure 6). The hypervariable D-loop region from 64 more animals was sequenced but similarly did not reveal any further population substructuring.
Although the mitochondrial data may seem limited in its usefulness, it was possible to-calculate divergence times for the two main lineages, which gave an estimate of between 200 000 and 1 million years. This led to the inevitable conclusion that modern domestic cattle are the product of two separate domestication events, and not one as was previously believed (Loftus et al., 1994). Furthermore, the peculiar absence of indicine mitotypes in African B. indicus has provided an interesting insight into the historical molecular biogeography of African cattle. It can be postulated that the expansion of zebu cattle in Africa was brought about primarily through the transmission of zebu male genes into the native taurine population. This hypothesis is strengthened by supporting work carried out on the Y-chromosome.
Y-chromosomes within species tend to be very similar at the sequence level (Bradley, unpublished data). If polymorphisms were found, it would be possible to use the Y-chromosome to trace paternal origin in a way not dissimilar to that of mtDNA typing. An RFLP study using a VNTR-type probe has shown some variation in the Y-chromosomal DNA. This has been used to provide a sensitive assay for genetic admixture in West African trypanotolerant cattle populations (Bradley et al., 1993b). The probe detected a polymorphism that distinguished a taurine from an indicine haplotype. Using this marker, it was possible to detect an asymmetrical dissemination of genetic variation. Some of the morphologically pure taurine N'Dama cattle of West Africa possess the indicine Y-chromosomes (Figure 7).
This paradoxical observation is most easily interpreted as resulting from the predominant male-mediated introgression of zebu genes in both recent and historical times, the strongest influence occurring between breeds that are in close proximity. This approach, although useful as a quick method of diagnosing zebu introgression in West African taurine cattle, does not provide enough information to construct an informative phylogeny or to quantitatively assess levels of admixture.
5 Restriction fragment patterns from Bgl II digested Bos indicus and B. taurus mtDNA - Fragments de restriction d'ADNmt de Bos indicus et B. taurus digérés par l'enzyme Bgl II - Modelos de fragmentos de restricción del ADNmt de Bos indicus y B. taurus digeridos con Bgl II
6 A phylogenetic tree for mtDNA of cattle breeds from three continents. American bison were included so that the tree could be rooted - Arbre phylogénétique de l'ADNmt des races bovines de trois continents. Le bison américain est inclus pour que l'arbre ait ses racines - Arbol filogenético del ADNmt de las razas de vacunos de tres continentes. Se han incluido los bisontes americanos de manera que pueda aparecer la raíz del árbol
7 Stacked column graph illustrating the number of zebu and taurine Y haplotypes observed in each of the populations tested - Histogramme illustrant le nombre d'haplotypes Y des zébus et des taurins observés dans chacune des populations testées - Histograma de los haplotipos Y de cebú y taurinos observados en cada una de las poblaciones estudiadas
8 Introgression of autopomorphic zebu alleles into some of the taurine populations of West Africa - Introgression d'allèles autopomorphiques de zebus dans certaines populations de taurins d'Afrique occidentale - Introgresión de los alelos autopomórficos de cebú hacia algunas de las poblaciones taurinas de Africa occidental
The study of nuclear polymorphisms provides further valuable information concerning the genetic structure of cattle populations, information that could not be obtained from the study of mitochondrial DNA or- the Y-chromosome. There are several techniques that can be used to assay levels of autosomal variation and some of these have been discussed already. Work carried out by MacHugh et al. (in preparation) has shown that microsatellites are perhaps the most informative markers for population studies at the subspecific level. Using 12 microsatellite loci, it was possible to detect substantial variation between six breeds of European cattle (MacHugh et al., in preparation). The extension of this work to 20 polymorphic loci, not only in European breeds but also in West African breeds, should provide a robust measure of genetic distance between the various cattle groups. The information that can be extracted from microsatellite data is not limited solely to measures of genetic distance. The great divergence between taurine and zebu cattle, as demonstrated using mitochondrial data, has led to the occurrence of some microsatellites with autopomorphic alleles (alleles unique to a population or subgroup). These can be powerful tools in quantifying the genetic admixture between the two main cattle types of West Africa. Figure 8 shows quite clearly the kind of information that can be obtained with this type of datum; the N'Dama breed can be seen to vary in the level of zebu admixture according to its geographical proximity to the neighbouring zebu breeds, the Gobra and Maure. The pie charts show an east-west, north-south cline of increasing zebu influence. As further data are collected from the other West African cattle breeds and as more autopomorphic loci are characterized, an even clearer picture of the genetic relationships of these cattle should emerge. Quantifying the extent of crossbreeding is complicated somewhat by the fact that the zebu cattle of the region are also crossbred, but this can be taken into account when estimating levels of zebu-taurine admixture.
Anderson, S., de Bruijn M.H.L., Coulson, A.R., Eperon, I.C., Sanger, F. & Young I.G. 1982. Complete sequence of bovine mitochondrial DNA: conserved features of the mammalian mitochondrial genome. Cambridge. J. Mol. Biol., 156: 683-717.
Baker, C.M.A. & Manwell, C. 1991. Population genetics, molecular markers and gene conservation of bovine breeds. In C.G. Hickman, ed. Cattle genetic resources, p. 221-304. Amsterdam, the Netherlands, Elsevier Science Publishers.
Bradley, D.G., MacHugh, D.E., Meghen, C., Loftus, R.T., Sharp, D.M. & Cunningham E.P. 1993a. Genetic characterisation of indigenous cattle breeds: first results and implications for genetic improvement, p. 37-44. FAO Animal Production and Health Paper No. 110. Rome, FAO.
Bradley, D.G., MacHugh, D.E., Loftus, R.T., Sow, R.S., Hoste, C.H. & Cunningham, E.P. 1993b. Zebu-Taurine variation in Y-chromosomal DNA: a sensitive assay for genetic introgression in West African trypanotolerant cattle populations. Anim.. Genet. (In press)
Dallas, J.F. 1992. Estimation of microsatellite mutation rates in recombinant inbred strains of mouse. Mamm. Genome, 3: 452456.
Hillis, D.M & Moritz, C. 1990. Molecular systematics. Sunderland, MA, USA, Sinauer Associates, Inc. Publishers.
Kemp, S. J. & Teale, A.J. 1991. International laboratory for research for animal diseases: bovine genome mapping and trypanotolerance. Proceedings of a workshop held at ILRAD, Nairobi, Kenya, 9-11 April 1991. p. 29-30.
Lhoste, P. 1991. Cattle genetic resources of West Africa. In C.G. Hickman, ed. Cattle genetic resources, p. 73-89. Amsterdam, the Netherlands, Elsevier Science Publishers.
Loftus, R.T. 1992. Mitochondrial DNA phylogeny of European, African and Indian cattle breeds. Dublin University. (Ph.D. thesis)
Loftus, R.T., MacHugh, D.E., Bradley, D.G., Sharp, P.M. & Cunningham, E.P. 1992. Mitochondrial DNA and inferred relationships between European, African and Asian cattle. Anim. Genet., 23 (supple. 1): 65.
Loftus, R.T., MacHugh, D.E., Bradley, D.G., Sharp, P.M. & Cunningham E.P. 1994. Evidence for two independent domestications of cattle. Proc. Natl. Acad. Sci. (In press)
MacHugh, D.E., Loftus, R.T., Bradley, D.G., Sharp, P.M. & Cunningham E.P. Microsatellite DNA variation within and among European cattle breeds. (In preparation)
Manwell, C. & Baker, C.M.A. 1980. Chemical classification of cattle I. Breed groups. Anim. Blood Groups Biochem. Genet., 11: 127-150.
Philipson, J. 1992. Genetic resources of cattle, p. 129-156. FAO Animal Production and Health Paper No. 104. Rome, FAO.
Queller, C., Strassmann, J.E. & Hughes, C.R. 1993. Microsatellites and kinship. Trends in Ecol. Evol., 8: 285-288.
Theilmann, S.L., Skow, L.C., Baker, J.F. & Womack J.E. 1989. Restriction fragment length polymorphisms for growth hormone, prolactin, osteonectin, alpha crystallin, gamma crystallin, fibronectin and 21-steroid hydroxylase in cattle. Anim.. Genet., 20: 257-266.
Wall, W.J., Williamson, R., Petrou, M., Papaioannou, D. & Parkin, B.H. 1993. Variation of short tandem repeats within and between populations. Hum. Mol. Genet. 2, 7: 1023-1029.
Weber, J.L. & Wong, C. 1993. Mutation of human short tandem repeats. Hum. Mol. Genet. 2, 8: 1123-1128.
Williams, J.G.K., Kubelik A.R., Livak K.J., Rafalski, J.A. & Tingey, S.V. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acid Res., 18, 22: 6531-6535.