D.J.S. Hetzel and R.D. Drinkwater1
1 Introduction
Over the past decade there has been widespread concern about the loss of valuable animal genetic resources through indiscriminate crossbreeding, breed replacement and neglect. The concern has intensified over time despite the potential of new techniques to artificially generate new genetic variation. So why all the concern? In short, the elegance, sophistication and subtlety of nature is still poorly understood. Much remains to be learnt about the creation, maintenance and evolutionary role of natural genetic variation before man can sensibly devise ways for its more effective utilization in a sustainable manner. For this reason alone, protection of the world's genetic resource base is a high priority.
The purpose of this paper is to discuss how DNA based technologies can contribute to the conservation and utilization of animal genetic resources. Three main areas have been addressed. These are the storage of DNA from threatened breeds/strains, the analysis of population structure to prioritize breeds for conservation and the mapping of genes unique to adapted breeds. Ways of measuring genetic variability at the DNA level will also be considered since they are central to more effective use of genetic resources. Key technologies for the preservation of gametes or embryos including cryogenic storage and related techniques have recently been reviewed (47) and will therefore not be discussed in this paper.
2 Collection and storage of DNA
The genetic information essential for the development, growth and efficient functioning of animals is housed within the DNA molecule. In higher mammals, around 3000 million individual bits of information are stored in each of the billions of cells that make up the entire organism. When conservation of this genetic information is considered, it is essential to preserve a sample that is representative of the organism, to purify and preserve it in a way that maintains the basic structure of the DNA molecule, and to store the DNA in alternative forms that allow easy access to the genetic information. Biochemical methods are available to meet these requirements.
2.1 Structure and Properties of DNA
Fortunately DNA is a very resilient molecule. The four basic nucleotides are linked by a sugar-phosphate backbone forming a long molecule (eg. a chromosome) that may be up to 100 million nucleotides in length. Two complementary strands are combined in a double helix formation held together by hydrogen bonds. In cattle, the entire genetic information content is found within sixty chromosomes in the nucleus of each cell. To attain the high degree of condensation that is required to allow such large DNA molecules to be physically housed in the cell nucleus, the DNA is wound around protein molecules and stacked in highly space efficient tertiary structures.
It is the extraordinary length of the DNA molecules together with their highly ordered packing that imparts mechanical and chemical stability. DNA can be treated with organic solvents, exposed to natural enzymes, heated to 75°C, desiccated and stored for thousands of years and its basic structure can still remain intact. It is however not totally impervious to damage. To be conserved efficiently it must be collected and stored with care.
2.2 Harvesting DNA
DNA is present in virtually every mammalian cell type. Any tissue, therefore, is a source of DNA. The ease of extracting the DNA is usually dependent on the tissue type. In dense tissues such as skin and tendon, DNA is difficult to obtain as the cells cannot be dissociated easily. Blood, however, provides an excellent source of loose cells and accessible DNA. A variety of DNA extraction procedures exist (1, 2) each of which incorporate the same basic principles. The cells must be ruptured, endogenous DNA degrading enzymes denatured and removed, the DNA packing protein removed, all lingering traces of any organic solvents removed, and the DNA re-solubilized in a stable buffer. These procedures apply to specialized organelle DNA such as mitochondrial DNA, as well as nuclear genomic DNA. Care must be taken to minimize the physical breakage of DNA, as large molecular size is essential for some analytical procedures.
2.3 Storage of DNA
DNA can be stably stored in a variety of ways. The method of choice often relates to the quantities of DNA required and to its intended use.
2.3.1 Tissue
The most basic storage form is frozen whole tissue. In this form, large amounts of DNA and RNA can be maintained with little preparation. The feasibility of this procedure is evident in the ability to recover DNA from whole mammals that have been entombed in glaciers for several thousand years. The key to the success of this method is rapid freezing of the tissue, and constant low temperature during the storage. RNA is not as well suited as DNA to long term storage in this form. An obvious drawback is the potential bulk of the tissue samples, with the exception of frozen semen. Semen samples have a high concentration of DNA and represent a convenient tissue for collection as well as long term storage. In addition, use can be made of existing artificial breeding infrastructures.
2.3.2 Cells/nuclei
Tissue bulk can be minimized by reducing the tissue to a whole cell or cell nuclei preparation before freezing. By removing the cellular cytoplasm in the nuclei preparation, the potential for autolysis by endogenous nucleases is also reduced. An alternative to dead cell storage is the generation of cell cultures from tissues such as skin or lymphocytes. This provides a pure, dissociated cell type that can be multiplied before harvesting the cells for storage as intact cells or as cell nuclei. The cell lines can also be stored live by freezing in liquid nitrogen, stored, and then revived and re-multiplied as necessary. However there is a risk of somatic mutations occurring as the cell line multiplies, thus creating material that is not identical to the original animal source.
2.3.3 Pure DNA
Purified DNA is certainly the preferred choice for the long term stable storage of DNA. Once dissociated from all protein, and free from bacterial or fungal contaminants, DNA can be stored indefinitely at - 20°C. A 10ml/10mg quantity of DNA extracted from any tissue and distributed in a number of “banks” could provide the basic resource from the conservation of the entire genetic information from a representative individual.
2.3.4 Libraries
Within the context of genetic conservation, there not only needs to be a means of storing a genetic resource; there also needs to be a method by which the information stored can be retrieved, studied and if necessary reused. The cloning of DNA into “libraries” can achieve this. A gene library is a representative portion of the genomic DNA that has been inserted into a “vector”. Amplification of a library allows any single DNA fragment to be multiplied many times in number thereby facilitating analysis. Libraries can also be derived from an RNA rather than DNA source. Such libraries contain the genes that were transcriptionally active in the original tissue. Both RNA and DNA libraries can be used as a storage source of genetic information. However the biological limitations of cloning mean that libraries are not always fully representative. These libraries do, however, offer the only means to identify and isolate entire genes, and as such offer an opportunity to investigate important production characteristics found in exotic and threatened domestic animals.
2.4 DNA and RNA Storage in Conservation Programmes
At the present time, live animals cannot be regenerated from isolated DNA. Therefore DNA storage is complementary rather than an alternative to cryopreservation of semen and embryos and the in situ preservation of animals. DNA does contain the essential genetic information that allows each animal to express its particular phenotype. It is therefore a crucial resource in investigating the function of each gene. As the function of newly discovered genes are elucidated, their role in animal growth development and survival can be determined. DNA cloning together with gene transfer technologies allow single genes, modified genes, or gene systems to be expressed in novel genetic backgrounds. These technologies make very effective use of stored DNA. A situation can be envisaged where a desirable production gene is identified using DNA marker analysis on stored DNA samples of a particular breed, the gene is isolated from a DNA or RNA library made from a representative of the breed, and the gene reintroduced and expressed in the genome of the donor breed. Hence the unique gene is conserved and propagated.
The storage of RNA as cDNA libraries derived from RNA must be an integral part of a conservation programme. RNA allows individual genes to be rapidly isolated, and the gene product expressed. The collection of a representative sample of RNA from an organism is very difficult due to tissue and time specific expression of genes. Hence a broad based collection of RNA could not be considered in the same manner as DNA. Conservation of RNA stocks in the form of cDNA libraries would have to be undertaken on a more selective basis.
The type and amount of stored genetic material should be related to the risk of losing a particular breed and the cost of storing different forms. In general DNA storage is cheap relative to live animal or embryo storage. For breeds at low risk, storage of purified DNA is adequate insurance. However for high risk breeds, DNA, tissue and RNA (in library form) should be preserved for future use. DNA banks should be viewed as genetic information resources for future use.
3 DNA sequence variation
In the process of domestication, limited population size and an increasing reliance on “superior” animals, genetic variation is reduced (3). Techniques for the analysis of variability are therefore an essential ingredient for conservation programmes. These techniques can be used to analyze the phylogeny of breed divergence, to follow gene segregation within populations, and ultimately, to associate nucleotide variation with changes in gene function and expression of animal phenotype. The most appropriate techniques for assessing DNA variation in a variety of applications include sequencing, Restriction Fragment Length Polymorphisms and conformational polymorphisms.
3.1 Sequencing
DNA sequencing provides the highest level of resolution that can be obtained in genome analysis (4). Sequence analysis reveals the fundamental structure of genes and allows an understanding of how nucleotide variation can influence gene expression. The concept of total genome sequencing is rapidly gaining acceptance with new automated sequencing technologies beginning to demonstrate the feasibility of the task. The sequencing of full mammalian genomes is understandably a massive undertaking, but not an unrealistic one. It is already practical to use DNA sequencing to analyze nucleotide variation in large animal populations. Using Polymerase Chain Reaction (PCR) and cycle sequencing technologies, highly variable sections of the genome can be rapidly analyzed in a moderate number of individuals.
3.2 Restriction Fragment Length Polymorphisms
3.2.1 Simple Length Polymorphism
Nucleotide variation can be highlighted by bacterial restriction enzymes which cut genomic DNA at specific sequences. Variation at the enzyme target site or insertions or deletions of DNA sequences between two enzyme sites produce fragments of variable length. The cut genomic DNA is size sorted by electrophoresis in agarose gels and then transferred to nylon support membranes. The specific DNA fragments of interest can be highlighted on the membranes by hybridization with labelled DNA probes. Unfortunately, simple RFLPs only exploit a fraction of the total variation in DNA sequence because only mutations which change the relative position of restrictive enzyme sites are detected. This is an important consideration in domestic animals since variability is lower than other outbred mammals (3).
RFLP systems can be made more informative by altering the analysis technique to produce multiple alleles. This can be achieved in two ways. Firstly by using a long DNA probe (e.g. 50 kbp), a haplotype of many restriction fragments may be revealed with the same enzyme (5). Secondly, the polymerase chain reaction (PCR) (6) can be used to generate a fragment in the region of interest, which is then cut with numerous high cutting frequency restriction enzymes. The resulting multi-fragment pattern can be visualized on a high resolution acrylamide gel. Both procedures do however have practical drawbacks. In haplotyping, the large DNA probe often contains highly repetitive DNA which mask the RFLP signal, and the multicut PCR method requires the nucleotide sequence of the marker region to be known. In many cases, the difficulties of RFLP systems can be overcome by simply using DNA markers that are more polymorphic.
3.2.2 Minisatellites
Another class of RFLPs arises not due to point mutations or chromosomal rearrangements but to variability in the number of repetitive DNA elements. The polymorphism is revealed by restriction enzymes which cut the DNA outside the repetitive sequence.
Minisatellites are fragments of DNA characterized by the tandem repetition of a sequence usually around 25 bp in length (7). The DNA is visualized by gel electrophoresis, blotting and DNA probing. Variation in the number of repeats can be high, with some loci in humans being heterozygous in 99% of individuals (8). In cattle the mean heterozygosity of 50 minisatellite loci was found to be 51% (5). Minisatellite loci exist in high numbers in any genome, and there is a strong sequence similarity between many of these loci. Thus when the genomic DNA is visualized by a minisatellite probe at low DNA hybridization stringencies, many loci become visible simultaneously. This results in a DNA fingerprint, which permits the absolute identification of animals, and under certain circumstances, provides a means of rapidly screening genomes to find DNA markers that are linked to production genes. When used at high stringency, minisatellite probes reveal only two alleles, but with a high probability that the locus is heterozygous and with different length alleles in each parent in a pedigree. This greatly expands the amount of genotype information available for analysis of genetic linkage between a DNA marker and a production trait. One possible limitation to minisatellite systems is a non-random distribution of genomic loci, with a bias towards chromosome telomeres (9, 10). There may be some regions of a genome which do not have minisatellite sequences.
3.2.3 Microsatellites
DNA microsatellites are highly variable polyallelic systems composed of DNA repeated in tandem at each locus (11, 12). The tandem repeats in microsatellites are usually simple dinucleotides (such as (CA)n) with each dinucleotide repeated on average twenty times. The length of each allele is determined by PCR analysis using unique oligonucleotide primers flanking the repeat sequence. The DNA products are visualized on sequencing gels. There are thought to be around 100,000 microsatellite loci in any animal genome (12), which means that any position on the genome lies within 25–50 kbp of a microsatellite. With such a widespread genome coverage and the ability to run up to five microsatellite loci in a single ‘multiplexed’ PCR, microsatellite systems have an analytical power approaching the level of minisatellite systems and are more convenient to use.
3.3 Conformational polymorphism
3.3.1 Single strand conformational polymorphism (SSCP)
SSCP techniques are a rapid and relatively simple means of detecting most DNA sequence changes. They rely on the concept that the conformation of a DNA strand is altered by a nucleotide change, and this new conformation can be detected as a mobility shift in gel electrophoresis (13,14). The method does require either sequence information for PCR primer sites at the site to be analyzed, or a DNA probe that will bind at the site.
3.4 DNA variability, conservation and animal improvement
Quantifying genetic variability at the DNA level is an important requirement for rational conservation and improvement programmes. The following sections will discuss in detail how DNA variation can be used to define population structures and how it can be associated with phenotypic variability.
4 Population structure analysis
The selection of breeds or strains of livestock for conservation or improvement programmes can be hampered by an inadequate description of population structure both within and between populations. Geographic isolation over time has built up a plethora of genetic types but the magnitude of genetic differentiation has rarely been quantified. The situation is further clouded when recent crossbreeding has occurred. A key element of the conservation strategy must be the characterization of breeds and strains to provide an overall picture of genetic diversity. The choice of appropriate populations for conservation or improvement should be based on a combination of phenotypic and genetic data.
4.1 General approach
A preliminary classification of livestock diversity, based largely on phenotypic observations has been made (15). However further refinements are necessary. DNA based measures of variation are potentially very useful in this work. The procedure basically involves four steps:
Obtain representative samples from described populations - sample numbers will be determined from breeding structure and planned analysis methods.
Measure between and within population genetic variation. A number of alternative systems have been discussed in section 3 and their applicability to population structure analysis will be considered below.
Calculate indices of genetic similarity/dissimilarity between populations. A large number of indices have been proposed, through the most widely used index, probably because of its conceptual appeal, is that of genetic distance (16). Other measures (17) may be more useful for short term evolution such as the divergence between livestock breeds.
Construct phylogenetic trees to describe both the current population structure and the evolutionary history of sub populations.
Clearly the success of each study is very dependent on the effectiveness of the sampling procedure and the extent to which variation in the total genome can be reassessed. A major limitation in the past has been the limited capacity to measure genetic variability at the DNA level. However, this limitation is now removed.
4.2 Measurement of genetic variability
Genetic variability can be measured in nuclear DNA or in the maternally inherited mitochondrial DNA. The latter type has the advantage of being unaffected by meiotic recombination. On the other hand, mitochondrial DNA is more subject to genetic bottlenecks. Thus it is advisable to use both sources of genetic information. Where the genetic distance between populations is likely to be relatively small, sensitive measures of variability are required. For this reason, protein polymorphisms or RFLPs are generally not sufficiently variable for efficient use. On the other hand, repetitive sequences such as mini-and micro-satellites, although characterized by hypervariability, will underestimate the rate of genetic divergence because repeat units are both added and subtracted during evolution.
Conventional RFLPs, which reveal less than 20% of total sequence variation, are now being supplemented with conformational polymorphisms such as SSCP, which are of course equally applicable to nuclear and mitochondrial DNA analysis.
However, with the advent of direct sequencing methods, which avoid the need for subcloning steps, the sequence analysis of discrete regions of DNA is now being used for genetic divergence studies. (eg 18) Mitochondrial DNA has been widely used for evolutionary studies because of its rapid evolution. In particular the displacement (D) loop contains several hypervariable (non coding) regions which can be readily sequenced. (18) Such studies have recently been initiated in cattle in an attempt to quantify differences between African, Indian and European cattle breeds. (D. Bradley, personal communication) With automation of sequencing now achieved, it is likely that future studies in other livestock species will also use sequence analysis as the primary method. However phenotype information and, where appropriate protein polymorphism data, should also be considered.
5 Gene mapping
The future use of animal genetic resources will be based on identification of genes associated with specific phenotypic characteristics. Techniques for rapidly mapping and isolating genes and characterizing them in terms of DNA structure and function are being refined. It is likely that in the near future, breeds and strains of livestock with unique or rare characteristics will be able to be characterized at the molecular level. The advantage of this approach is that utilization of unique or rare alleles will be greatly facilitated through programmes to introduce variants of genes into other populations, or to increase the frequency of such alleles within existing populations.
In the following sections, the current status of gene maps for livestock will be briefly reviewed and strategies for mapping individual genes will be discussed.
5.1 Gene maps
Gene maps have three main functions. First, they serve as repositories for gene mapping data ie. the location, order and spacing of genes or genetic markers. Second, maps can be used to facilitate mapping of new genes by providing reference points for searches. Third, by cross-referencing gene mapping information from one mammalian species to another, homologous chromosomal regions can be identified thereby permitting the extrapolation of information on gene location. Thus gene maps are an extremely valuable resource for mapping studies.
5.1.1 Physical maps
Physical gene maps indicate the location of genes on specific chromosomes, chromosomal regions or syntenic groups. The latter comprise groups of genes known to reside on a chromosome of unknown identity. The coordinates and therefore units of a localization will depend on the mapping method used. For example, at present, somatic cell hybrid analysis in cattle can map genes to either the chromosomal or syntenic group level (21). However, in situ hybridization results in assignment to chromosomal region with reference to a banded karyotype (22). Recent developments in the use of nonradioactive methods of hybridization to interphase chromosomes have greatly increased mapping resolution (24). Pulsed field gel electrophoresis is used for fine scale mapping where distances are measured in base pairs of DNA (23).
A recent summary on the status of mammalian gene maps (Table 1), demonstrates that relatively few of the estimated 50 – 100,000 genes have been localized on mammalian genomes. However, research effect in this area is increasing worldwide, particularly with recent initiatives in human genome mapping.
As far as livestock are concerned, the cow is the best documented species. However, it must be realized that only around 25% of the mapped genes have been regionally assigned as distinct from localization to only a chromosome or syntenic group. Thus the task ahead for livestock gene mapping is considerable.
On the other hand, there will be considerable spinoffs to livestock from genome research in other species. Much of the basic work on genome organization and function will be carried out in man, mouse and other experimental organisms such as yeast, Drosophila, C. elegans, and E. coli. As will be discussed later, opportunities for making use of this information are considerable.
Table 1. Current status of physical gene maps in various mammalian species
| Haploid Chromosome No. | No. Syntenic Groups | Number of Genes Mapped | |
| Human | 23 | 23 | 2500 |
| Mouse | 20 | 20 | 2200 |
| Cattle | 30 | 29 | 250 |
| Sheep | 27 | 19 | 50 |
| Pig | 19 | 17 | 80 |
| Source: Ref (29) | |||
5.1.2 Genetic maps
Genetic or linkage maps document the order and spacing of genes or genetic markers where distances are measured in recombination units.
The frequency of recombination between homologous chromosomes, a normal event in meiosis, is generally proportional to the distance between two linked loci and can therefore be used as a unit of chromosomal length. The unit, is known as the Morgan (M), and one centimorgan (cM) corresponds to the distance separating two loci exhibiting a 1% recombination rate. For short distances, eg. less than 30cM, the relationship between physical distances and recombination rate is essentially linear. However over longer distances, interference and the occurrence of double crossovers can reduce the concordance.
Measuring recombination rates or linkage relies on being able to distinguish parental and recombinant genotypes. It therefore requires polymorphism to be defined at each locus, and the analysis of segregation from doubly heterozygous individuals (25). Recently, it has become possible to genotype individual gametes ie. sperm and ova (26). However, genetic contribution is generally inferred for the genotypes of offspring and linkage studies are carried out within families.
As discussed earlier, there are a wide range of polymorphic markers available. However the most numerous and in many ways the most practical are based on DNA sequence variation. Accordingly, such markers are now being widely studied. Each marker permits the segregation of a discrete chromosomal region to be followed. Pairs of flanking markers provide even greater precision.
It is difficult to assess the current status of genetic maps of livestock for two reasons. First, there is currently a large amount of activity around the world directed at linkage maps in all of the economically important livestock species. Thus even a recently published list is quickly out of date. Second, some research is being carried out in the private sector and for reasons of competitive advantage, is not being published. To take the linkage map of the cow as an example, the most recent published report cites a total of 8 linkage groups comprising 64 markers (27). However, our own group alone has mapped a further 80 markers in the past two years and we are aware of another group which has mapped more than 100 other markers.
The challenge in the future will be to collate all the linkage information generated within a species into a common framework. In other words, there is a need to avoid the construction of a large number of genetic maps for a single species with little cross-referencing between them. This will arise if different markers are mapped in different sets of families. Although there will probably be some differences in recombination rate between strains and breeds of the same species, their magnitude will be relatively small in most cases. Given the crude state of livestock genetic maps, it is more important at this stage to collate information across breeds within a species.
To this end, the use of common sets of reference families have been proposed (28). By sharing a limited number of families for use worldwide, the collation of linkage data is greatly simplified. Reference families for cattle are being distributed (28) and such families are being prepared in other livestock species.
A complete linkage map is a powerful tool for gene mapping since it permits systematic and thorough searches of the genome for genes and genetic regions associated with specific traits. An initial goal is a 20cM index marker map ie. where the maximum spacing between polymorphic markers on the chromosome is 20cM. Given that it is not presently possible to choose from where on the genome the markers are isolated, the initial strategy must be to randomly map markers. At some later stage, chromosomal regions which are underpresented with markers can be targeted.
Assuming a total map length of 30M as for humans, and a marker spacing of 20cM, an index maker map of the genome will require around 150 polymorphic markers. In order to achieve this target when markers are isolated at random, in excess of 500 markers will be required for each livestock species. However, 80% genome coverage could be achieved with around half that number (30). Index maker maps, based on a subset of available markers will be then used in the systematic analysis of the genomes.
5.1.3 Comparative maps
Comparative gene mapping involves the mapping of homologous gene loci in multiple species. The principal aim is to identify the boundaries of genomic conservation between evolutionary divergent species. Although the ultimate goal of comparative mapping is to understand the pathways by which chromosomal evolution has accompanied speciation, it has proved to be a very powerful means of extrapolating mapping data from one species to another based on the establishment of conserved chromosomal segments.
Extensive gene and chromosomal homologies have now been identified between man and mouse (31) and between man, mouse and the cow (32). Although the boundaries of conserved segments need further definition, it is now possible to both interpolate and extrapolate the position of genes between these three species. Thus as genes are mapped as one species, their likely position in the other species can be predicted. This will be an important source of mapping information in the future since genome research in human and mouse has accelerated rapidly.
Chromosomal homologies between closely related livestock species have been observed at the cytogenetic level. For instance, the banded karyotypes of sheep and cattle have been matched up almost perfectly and homology has been confirmed by in situ localization of a number of genes (33). Although some minor rearrangements affecting the spacing, and in some cases the order of genes can be expected, it will be possible to exchange information between the cow and sheep gene maps.
5.2 Strategies for mapping genes
There are basically two approaches to mapping genes for which the gene product is unknown. The first of these, referred to as the candidate gene approach, relies on some background biochemical knowledge and is therefore limited in its application. The second approach which defines a linked genetic marker as an intermediate step towards cloning the gene is more generally applicable because it requires no knowledge of gene function. The two approaches are not mutually exclusive. Once the search for a gene is narrowed down to a discrete region, other genes may suggest themselves as candidates where previously they were not obviously connected. So in practice, a combination of both approaches is appropriate, especially as mammalian gene maps become denser, providing a larger number of potential candidates.
5.2.1 Candidate genes
The candidate gene approach is an attractive option because it may give a quick answer and involves an element of hypothesis testing. However, the selection of candidates should be based on solid biochemical data or physiological observations. It should be appreciated that an altered level of a particular enzyme or other gene product does not imply an alteration in the structural gene for that product. Gene expression is frequently very complex, involving a number of regulatory genes, in some cases located on distant parts of the genome to the structural gene. Nevertheless, the hypothesis of whether variability at or around the site of a gene exists, can be tested if appropriate reagents and/or sequence information and pedigrees are available. The requirements include multigeneration pedigrees with appropriate genotype and or phenotype data. The optimum pedigree structures will differ depending on whether simple or polygenic inheritance is expected. Secondly, there is a requirement to measure genetic variability at the candidate gene locus. DNA markers are most frequently used for this purpose. By analyzing for co-segregation at the candidate gene locus and locus of interest, the hypothesis can be tested. Since the candidate gene marker effectively tags a segment of DNA, the involvement of other genes on that segment is tested simultaneously.
There are reports in the literature where candidate gene approaches have been successful. Presumably there are many other unpublished studies which have produced negative results. In cattle, a number of single gene disorders have been mapped and diagnostic marker tests developed via candidate genes. These include Pompe's Disease (34), and Bovine Leucocyte Antigen Deficiency (M. Georges, personal communication). The gene for malignant hypothermia in man and pigs was mapped in a similar way (36). In all cases, detailed biochemical data suggested a strong candidate.
5.2.2 Linked markers
Another approach to mapping genes involves an initial localization to a chromosomal segment via anonymous genetic markers. Detailed analysis of the segment can refine the localization. This approach procedure, part of the so called ‘Reverse Genetics’ approach to gene analysis, is inherently more laborious, but, if used methodically has a higher probability of success. The approach is equally applicable to simply inherited and polygenic effects although the optimum pedigree structures will vary and there is a limit to the size of gene effects detectable in the polygenic situation.
The basic procedure involves genotyping multigeneration families, which have been phenotyped for the trait of interest (Fig 1). The probability of detecting a linked marker or markers is directly related to the distribution of markers and the number of informative matings for linkage analysis. Ideally, the markers will be chosen from an index marker map to ensure complete screening of the genome with a minimum number of markers. However, such maps are not yet available for livestock. Assuming there are sufficient informative meioses to be analyzed, linkage between one or more of the markers and the gene or associated phenotypic effect should be found. By localizing the markers, if they have not already been mapped, the chromosomal segment carrying the gene of interest can be defined. This is a very significant step for two reasons. First, it considerably narrows down the area to be analyzed for closer markers. Second, an approximate localization of a gene allows a re-evaluation of possible candidate genes which have previously been mapped to this region. Homologous regions in other genomes can also be surveyed, thereby expanding the number of potential candidates. Recent reports where the approach of firstly localizing a gene marker followed by successfully selecting from candidate genes in the region include studies on Waardenburg's syndrome (37) and hypertension (38).
Isolation of a gene by using positional cloning techniques is still not a trivial task, although recent successes include Type 1 - Neurofibromatosis (39) and the testis determining factor (40). Nevertheless, once a gene has been mapped by flanking markers, it can be efficiently tagged within families, and used in marker based selection programmes.
5.3 Analysis of quantitative traits
By and large, the traits related to growth, reproduction, fitness, milk production, parasite and disease resistance and tolerance to climatic variables are characterized by continuous variation. Differences between individuals are of a quantitative rather than qualitative nature. This has led to the assumption that such traits are controlled by a large number of genes, each of small effect (see 41). The true genetic basis of quantitative traits has until recently, remained largely unknown, because studies were general based on statistical methods of low power. More recently and with the use of molecular genetic techniques, a more intensive analysis of quantitative traits has begun.
Evidence is accumulating that variability in such traits may be associated with a small a combination of a small number of genes of intermediate to large effect eg. 0.2 – 1.5 phenotypic standard deviations, and an unknown number of genes of smaller effect. Recent studies on yield traits in maize (42), water use efficiency in tomatoes (43), fruit mass, and fruit pH in tomatoes (44) and predisposition to hypertension in rats (38) have strongly suggested a range of gene effects. The contribution of intermediate - large effect genes to the total genetic variation, will be small if the ir frequency is either low or high. As molecular techniques for genome analysis in livestock are refined, it is becoming feasible to analyze the genetic basis of quantitative traits. It is clear that the utilization of such genes, which may be unique to specific populations, will be facilitated by genetic markers. However the resources required to map such genes are substantial and it will therefore be important to optimize experimental designs in order to maximize the probability of success.
5.3.1 Experimental designs for quantitative trait analysis
Quantitative trait analysis measures the association between specific chromosomal segments tagged by markers and phenotypic effects. The phenotypic effects are assumed to be due to the presence of one gene or several genes on the chromosomal segment. Matings are only informative if one or both of the parents are heterozygous for both the marker and gene. The design of QTL mapping experiments fall into two categories viz analysis of crosses between breeds/lines and analyses of families within populations.
Design considerations have recently been reviewed (46). In an extension of the crossed line design, M. MacKinnon (unpublished) has evaluated within family analysis of F1 sires resulting from crosses of divergent breeds.
General conclusions from these studies are;
The analysis of either backcross or F2 families is highly efficient where alternative alleles have been fixed or allele frequencies are very different (eg. 0.8 vs 0.2) in the two breeds/lines. However if allele frequency differences are less than 0.5, there is considerable loss of power over some within family designs.
F1 cross sire families are frequently as efficient for quantitative trait analysis as F2 families, since the availability of polyallelic markers considerably reduce the need to genotype dams. However progeny group sizes must be larger than normal, eg. 100 – 150, for reasonable efficiency. F1 cross sire families are inefficient at detecting rare alleles in one of the parental breeds and are less sensitive to dominance effects than the F2 design.
The analysis of families within straightbred populations generally has low power because the frequency of sires heterozygous at both marker and quantitative trait locus is normally low. Progeny testing can however significantly improve power. (45)
Overall, the power of mapping experiments can be further improved by selective genotyping (35), interval mapping (35) and the use of DNA pools (19). Nevertheless, the required animal and laboratory resources remain substantial. Yet such experiments are likely to be the best approach to analyzing the genetic basis of quantitative traits, and are directly applicable to exotic and rare livestock breeds.
Crosses between resistant and susceptible cattle breeds are being generated in an attempt to understand the inheritance of trypanosomiasis in Africa (20). Other breeds with unique characteristics or adaptations could be utilized in a similar way.
6 Concluding remarks
It has been suggested that DNA technologies can assist livestock conservation and improvement in three major ways. Firstly, various forms of DNA can be stored to preserve genetic variability for future use. Secondly DNA sequence polymorphism and associated sequencing techniques provide the means to decide which livestock populations should be targeted in conservation programmes. And thirdly, gene mapping technologies are capable of identifying the discrete regions of DNA which account for unusual performance characteristic adaptations. Such information will have direct use in future breed improvement programmes. It is therefore essential that not only is livestock genetic diversity preserved, but also that investigations continue on its biological basis. These investigations will provide the foundations for future long term use of livestock with special characteristics.
Figure 1 Steps Involved in the Reverse Genetics Approach to Gene Localization and Characterization
| Segregating families | ||
| ↓ | Genotype with markers | |
| (± candidate gene markers) | ||
| Linked marker | ||
| ↓ | ||
| Map linked marker (if previously unmapped) | ||
| ↓ | ||
| Evaluate other markers in the region | ||
| Closely linked marker | ||
| ↓ | ||
| Positional cloning | ||
| Isolate gene | ||
| ↓ | ||
| Study structure/function | ||
7 References
1. Sambrook J., Frisch E.F., and Maniatis T., (1989). “Molecular Cloning - A Laboratory Manual”, 2nd Edition. Cold Spring Harbor Laboratory Press, USA.
2. Davis L.G., Dibner M.D., and Battey J.F., (1986). “Basic Methods In Molecular Biology”, Elsevier Science Publishing Co, Inc., New York.
3. Hilbert, P., et al., (1989). Analysis of genetic variation in the Belgian Blue cattle breed using DNA sequence polymorphism at the growth hormone, low density lipoprotein receptor, alpha-subunit of glycoprotein hormones and thyroglobulin loci, Anim. Genet. 20:383–394.
4. Sanger F., Nicklen S., Coulson A.R., (1977). DNA sequencing with chain terminating inhibitors, Proc. Natl. Acad. Sci. U.S.A. 74:5463–5467.
5. Georges M., Mishra A., Sargent L., Steele N., Zhao X., (1990). “Progress toward a primary DNA marker map in cattle”, Proc. 13th World Congr. on Genetic Applications to Livestock Production, Edinburgh 107–112.
6. Saiki R.K., et al., (1988). Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science 239:487–491.
7. Jeffreys A.J., Wilson V., Thein S.L., (1985). Hypervariable ‘minisatellite’ regions in human DNA, Nature 314:65–73.
8. Smith J.C., et al., (1990). Highly polymorphic minisatellite sequences: Allele frequencies and mutation rates for five locus-specific probes in a caucasian population, J. Forensic Sci. Soc. 30:19–32.
9. Royale N.J., Clarkson R.E., Wong Z., Jeffreys A.J., (1988). Clustering of hypervariable minisatellites in the proterminal regions of human autosomes, Genomics 3:352–360.
10. Drinkwater R.D., Hediger R., Hetzel D.J.S., (1990). “Polymorphism and location of minisatellite sequences in sheep”, Proc. 13th World Congr. on Genetics Applied to Livestock Production, Edinburgh 129–132.
11. Weber J.L., May P.E., (1989). Abundant class of DNA polymorphisms which can be typed using polymerase chain reaction, Am. J. Hum. Genet. 44:388–396.
12. Luty, J.A., et al., (1990). Five polymorphic microsatellite VNTR's on the human X chromosome, Am. J. Hum, Genet. 46:776–783.
13. Orita, M., Iwahana, H., Kanazawa, H., Hayashi, K., and Sekiya, T. (1989) Detection of Polymorphisms of Human DNA by Gel Electrophoresis as Single-strand Conformational Polymorphisms. Prog. Nat. Acad. Sci., 86: 2766–2770
14. Orita, M., Suzuki, Y., Sekiya, T., and Hayashi, K. (1989) “Rapid and Sensitive Detection of Point Mutations and DNA Polymorphisms using the Polymerase Chain Reaction.” Genomics, 5, 874–879
15. Mason I.L., (1988). Evolution of Domesticated Animals. Longman. Harlow. U.K.
16. Neim, M. (1972). American Naturalist 106:283–292.
17. Reynolds J., Weir B.S., Cockerham C.C., (1983). Genetics 105:767–779.
18. Vigilant L., Stoneking M., Harpending H., Hawkes, K., Wilson A. (1991) “African Populations and the Evolution of Human Mitochondrial DNA” Science 253:1503– 1507
19. Plotsky, Y., Cahener, A., Haberfeld, A., Lavi, U., Hillel, J. (1990) “Analysis of genetic association between DNA fingerprint bands and quantitative traits using DNA mixes.” Proceedings of World Congress on Genetics Applied to Livestock Production. 13:133–136
20. Soller, M. (1990) “Strategies and opportunities for mapping QTL in agriculturally important animals.” Womack J.E. ed., New York, Cold Spring Harbor Laboratory Press, p53–60
21. Adkison L.R., Skow L.C., Thomas T.L., Petrash M., Womack J.E. (1988). Somatic cell mapping and restriction fragment analysis of bovine genes for fibronectin and gamma crystallin. Cytogenet. Cell. Genet. 47:155–159.
22. Hediger, R., Johnson, S.E., Barendse, W., Drinkwater, R., Moore, S. and Hetzel, D.J.S., (1990). Assignment of the growth hormone gene locus to 19q26-qter in cattle and to 11q25-qter in sheep by in situ hybridization, Genomics 8:171–174.
23. Threadgill D.W., Womack J.E., (1991). Synteny mapping of human chromosome 8 loci in cattle. Animal Genetics. 22:117–122.
24. Trask B.J., (1991). Fluorescence in situ hybridization: applications in cytogenetics and gene mapping. Trends in Genetics 7:p149–154.
25. Ott J., (1985). Analysis of Human Genetic Linkage. Johns Hopkins University. Baltimore.
26. Li H., Gyllensten U.B., Cui X., Saiki R.K., Erlich H.A., Arnheim N., (1988). Amplification and analysis of DNA sequences in single human sperm and diploid cells. Nature 335:414–417.
27. Womack J.E., (1990). Status of the Bovine Gene Map. In “Mapping the genomes of agriculturally important animals”. Womack J.E., ed. New York. Cold Spring Harbour Laboratory Press, p115–121
28. Hetzel D.J.S., (1991). The Use of Reference Families For Genome Mapping in Domestic Livestock. In a Gene-Mapping Techniques and Applications, ed Schook L.B., Lewin H.A., McLaren D.G., Marcel Dekker Inc. New York, p51–64.
29. O'Brien S.J., and Marshall-Greaves J.A. (1992). Report of the Committee on Comparative Gene Mapping, Human Gene Mapping Workshop 10.5, Cytogenet. Cell. Genet. (in press).
30. Kashi Y., Soller M., Hallerman E., Beckmann J.S., (1986). Restriction Fragment Length Polymorphisms in Dairy Cattle Genetic Improvement. In Proc. of the Third World Congress on Genetics Applied to Livestock Production, Nebraska, 12:57–63.
31. Nadeau J.H., (1989). Maps of linkage and synteny homologies between mouse and man. Trends in Genet 5:82–86.
32. Womack J.E., (1987). Comparative gene mapping: a valuable new tool for mammalian development studies. Dev. Genet. 8:281–293.
33. Hediger, R., Ansari H.A., Stranzinger G.F., (1991). Chromosome banding and gene localizations support extensive conservation of chromosome structure between cattle and sheep. Cytogenetics and Cell Genetics. 57:127–134.
34. Hetzel D.J.S., Drinkwater R.D., Aspden W.J., (1989). Towards an RFLP marker for Pompe's disease in cattle, Anim. Genet. 20:25.
35. Lander, E.S., Botstein, D., (1989) Mapping mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121:185–199
36. MacLennan D.H., Duff C., Zorzato F., Fujii J., Phillips M., Korneluk R.G., Frodis W., Britt B.A., and Worton R.G. (1990). Nature 343:559–562.
37. Tassabehji M., Read A.P., Newton V.E., Harris R., Ballings R., Gruss P., Strachan T., (1992). Waardenburg's syndrome patients have mutations in the human homologue of the PAX-3 paired box gene. Nature. 355:635–636.
38. Jacob H.J., Lindpaintner K., et al (1991). Genetic Mapping of a Gene Causing Hypertension in the Stroke-Prone Spontaneously Hypertensive Rat. Cell. 213–224.
39. Wallace M.R. et al (1990). Science 249:181–186.
40. Sinclair A.H., Berta P., Palmer M.S. et al (1990). A gene from the human sex-determining region encodes a protein with homology to a conserved DNA-binding motif. Nature 346:240–244.
41. Falconer D.S., (1981). Introduction to quantitative genetics, Second edition Longman, London and New York.
42. Edwards M.D., Stuber C.W., Wendel J.F., (1987). Molecular - marker - facilitated investigations of quantitative - trait loci in maize. I. Numbers, genomic distribution and types of gene action. Genetics 116:113–125.
43. Martin B., Nienhuis J., King G., Schaeffer A. (1989). Restriction fragment length polymorphisms associated with water use efficiency in tomato. Science 243:1725– 1727.
44. Paterson A.H., Lander E.S., Hewitt J.D., Peterson S., Lincoln S.E., Tanksley S.D. (1989). Resolution of quantitative traits into Mendelian factors by using a complete linkage map of restriction fragment length polymorphisms. Nature 335:721–726.
45. Weller, J.I., Kashi, Y., Soller, M., (1990) Power of “daughter” and “granddaughter” designs for genetic mapping of quantitative traits in dairy cattle using genetic markers. J. Dairy Sci 73:2525–2532
46. Soller M., (1991). Mapping Quantitative Trait Loci Affecting Traits of Economic Importance in Animal Populations using Molecular Markers. In Gene Mapping Techniques and Applications' ed Schook L.B., Lewin H.A., McLaren D.G., Marcel Dekker Inc. New York.
47. Brem, G., Brenig, B., Muller, M., Springmann, K., (1989) Ex situ cryoconservation of genomes and genes of endangered cattle breeds by means of modern biotechnological methods. FAO Animal Production and Health Paper No.76
