This explanation is particularly for sheep breeders who want a little more detail about genomics than it simply being “a study of the genetic material of a living thing at a molecular level”. Further information is available below.
What makes up the genetic material of a sheep?
The genetic material of sheep and other living things is made up of similar structures. Firstly, there are chromosomes, of which sheep have 27 pairs, but other animals and plants can have different numbers of pairs. The collection of all 27 pairs of the sheep’s chromosomes together and the genetic map that they create is called the ‘genome’. The chromosomes are found in the nucleus of most of the sheep’s cells.
Each chromosome is made up from a long length of DNA, which stands for deoxyribonucleic acid. DNA looks like an extremely long ladder twisted to a spiral shape, described as a ‘double helix’ as shown in Figure 1.
The ‘rungs’ of the DNA ladder are the basic building blocks that make chromosomes differ. Each rung is made of a pair of nucleotides (also called base pairs) joined together (see Figure 2). There are only four possible nucleotides: Adenine (A), Cytosine (C), Guanine (G) and Thymine (T). The order or sequence of these nucleotides along the ladder is what makes one sheep different from another and from other living things.
When comparing the DNA of animals, only the nucleotides along one side of the ‘ladder’ are recorded. Once those nucleotides are known so are the ones that join them, because an A is always paired with a T and a C is always paired with a G.
Genes are long sections of the DNA ladder made up from many nucleotide pairs in a particular sequence and at a particular position along the length of the chromosome. The type and combination of genes is what makes a sheep different to a cow or a plant. Genes are ultimately responsible for making everything in the body such as different types of cells and production of body substances such as enzymes and hormones.
Genes only make up about 10% of the length of each chromosome, some of the other parts help turn the genes on or off, but as yet, the roles of many other parts of the chromosome are still unknown.
While different genes make sheep different to cattle, differences within each gene make one sheep different to another. When there are variations within a gene, this creates ‘alleles’ of a gene. The chromosomes come in pairs, with each of the pair having the same genes, but they could have different alleles, or variations, of each gene. So each animal carries two copies of each gene (except for genes on the sex chromosome), but those genes may be slightly different if they are different alleles of the gene.
How do alleles interact?
For some genes, the alleles act in a dominant or recessive manner. For other genes the alleles are co-dominant and for yet others, they act in an additive manner.
An example of a gene where there are two alleles with one dominant and the other recessive is for Spider Lamb Syndrome, mostly found in the Suffolk and Hampshire breeds. The two alleles for this gene are written as S and s: the uppercase S denotes the dominant gene, in this case the normal, non-Spider Lamb Syndrome allele, and the lowercase s is for the recessive Spider Lamb Syndrome allele.
As each sheep carries two copies of the gene and there are only two alleles, the possible combinations they can carry are: S/S, S/s or s/s.
The S/S animal with both dominant genes will be normal. The S/s animal with one dominant and one recessive allele will also be normal because the dominant allele overrides the recessive allele. However, the s/s animal will be affected with the disease as both alleles are for the disease. The S/s animal will also be described as a “carrier” of the disease.
Red, white and roan colouring in shorthorn cattle is a gene where two alleles are responsible for these three colours. The red (R) allele and the white (r) allele are actually co-dominant. In this case an R/R animal would be red, an r/r animal would be white, but the R/r animal is actually a mixture of red and white, being roan.
The Booroola allele (FecBB) is an example of an additive effect, where the alleles increase ovulation rate. Sheep with no Booroola alleles, FecB+/FecB+ have the normal ovulation rate for that breed. Sheep with one Booroola allele, FecB+/FecBB, have a higher ovulation rate and sheep with two Booroola alleles, FecBB/FecBB, have an even higher ovulation rate.
The increase in ovulation rate depends on the breed or strain of sheep, but as an example, in some strains of Merinos it could mean an increase from a normal rate of about 1 or 2 eggs per ovulation, to 2 or 3 eggs per ovulation with one Booroola allele and 4 to 5 eggs per ovulation if they have two.
When a sheep carries two of the same alleles, they are described as heterozygous for that allele. They are homozygous if they carry two different alleles for the one gene.
An exception to pairs of genes occurs with one pair of the sheep’s chromosomes, called the sex chromosomes. These are referred to as the X and Y chromosomes. Ewes have two X chromosomes to make up their pair, whereas rams have an X and a Y for their pair. In simple terms the Y is the same as its X partner, but is missing one “leg” of the chromosome. For genes that are located on that leg, rams only carry one copy, whereas ewes have two.
An example of a gene on the X chromosome, but not on the Y is colour blindness in humans. It is a recessive gene, so women can be colour blind if they are homozygous for the allele, that is, they carry the recessive allele on both of their X chromosomes. If they are heterozygous, carrying one normal and one colour blind allele, they will not be colour blind as it is a recessive condition. Men carry only one gene, if it is the recessive colour blind allele, then they will be colour-blind.
What are SNPs?
Many production traits of sheep, such as weight, are not controlled by just one or a few major genes (like those described above), but are affected by many genes, each with allele variations. As such, a way is needed to look at numerous genes at once and to see how they affect performance traits.
Unfortunately, most sheep genes are not yet identified or their role is not yet defined. However, at more than 50,000 places along the genome, the sequence of nucleotides is known for a short distance. Each of these sequences has a known variation in one of the nucleotides. These variations are called a Single Nucleotide Polymorphism (SNP) which is pronounced ‘snip’.
For example, where most of the sheep population may have an Adenine (A) nucleotide at a particular locus (or place along the genome) the remainder of the population instead has a Guanine (G) nucleotide. Despite the possibility that within the sheep population all four nucleotide variations could exist at that position, this is extremely rare; mostly there are only two variations at a particular SNP.
To explain how these variations came about, assume that many thousands of years ago all sheep were the same. At some point when the cells that created eggs and sperm were reproducing in a sheep, a mistake occurred (called a mutation) so that a SNP variation appeared in the DNA of the offspring. Over time, more of these mistakes occurred creating many SNPs along the chromosomes.
Some differences in the DNA had no effect, some differences that were so bad that the embryo died or offspring did not survive to reproduce (such as with Spider lamb Syndrome described earlier), and yet others created differences in the animal’s looks or performance (such as the Booroola allele), some of which were an improvement for survival.
Generally, those differences that had some effect were part of a gene and created an allele or variation different to the normal gene found in the original sheep (these original alleles of genes are called ‘wild-type’ and often have a + in their abbreviation). For some genes the wild-type allele has naturally or artificially been bred out of sheep because it created an effect that was less desirable than the new allele.
Alleles of genes can be more than just a simple SNP, for instance, the duplication of a small section of the DNA in a gene or the reversal of a section of the DNA are some of the ways that new alleles for a gene can appear.
Knowing the actual genes and their role is the ultimate goal, and much of this work has been completed for the human genome (at great expense), but it will be some time, if at all, before the sheep genome will be mapped and all gene effects known.
Meanwhile, the next best thing—a SNP—is used to link the genes to performance. SNPs can be found anywhere along the genome, either within a gene, close to a gene, or far from a gene. Remember, they are just a part of the DNA, but are where a difference has occurred.
Alleles, described earlier, are variations in an actual gene. A particular SNP could actually be part of an allele variation in a gene, but more likely, it is somewhere else along the chromosome (because genes are only about 10% of the chromosome length).
If the SNP is a key part of the allele variation (such as SNP 1 in Figure 7 below) then it could be 100% accurate at indicating the gene effect—where the SNP and the allele variation are one and the same. If the SNP is not part of the allele variation, it may still predict the allele variation if they are close together, such as SNP 2. The further from the gene the SNP is (such as SNP 3) the less accurately the SNP will predict the allele variation. A SNP that is a great distance away from the gene is not useful for its prediction; however, if it is close to some other gene, it may be a predictor for that other gene.
How does genetic recombination affect SNPs and genes?
The reason that the distance between the SNP and the allele is important is because breakages can occur in between them. Sometimes a part from one of a pair of chromosomes separates and crosses over with the corresponding part of the other chromosome in the pair; this is called recombination. This can break up associations that have occurred between a particular allele and a particular SNP.
As such, SNPs that are very close to or in the gene are more accurate predictors of the allele, as the chance of the recombination occurring between them decreases the closer they are together.
To visualize this, imagine tying 3 knots in a metre-long piece of string: knot 1 is very close to one end, knot 2 is only 5 mm further on, and knot 3 is close to the other end. If 100 cuts were randomly made along the piece of string most of them would be between knots 2 and 3, maybe only one or none of them would be between knots 1 and 2. Thus knot 2, representing a SNP, is more likely to stay linked to a gene that it is very close to (knot 1), rather than a gene farther away (knot 3).
Time also increases the chance of the recombination occurring between the SNP and gene, as there are simply more opportunities for it to occur.
Both the breed and the bloodline of a sheep can affect the accuracy of a SNP’s ability to predict a gene’s effect because of the recombination factors explained above. This is because either a breed or a bloodline is a narrower population that has few, if any, introductions from other breeds or bloodlines. They are created by focusing on types, which tends to breed out some allele (and nearby SNP) variations.
Sheep from the one breed or bloodline have chromosome pairs that are more likely to be similar, that is, they have less SNP and allele variations. As such, when recombination does occur between the SNP and gene on one chromosome they simply end up with the same SNP and allele anyway because the other chromosome is likely to have been the same.
Where a different breed or bloodline has existed for some time, there may have continued to be a number of allele variations over that time, and recombination opportunities have allowed the breaking of the relationship between the particular SNP and particular allele. So in one breed/bloodline a particular SNP may be strongly related to a particular allele, in another breed/bloodline the same SNP may be associated with the same allele plus another. The SNP then becomes a good predictor in the first breed, but less accurate in the second breed or if the two breeds were crossed.
Which SNPs are associated with which genes?
Researchers already know about many thousands of SNPs—it is known where they are located on the genome and what nucleotide variation occurs, but whether those SNPs mean anything is unknown for many of them.
A population of sheep with full pedigree and extensive performance records for traits such as eye muscle depth, fibre diameter and worm egg count is tested to show which SNP variation occurs at each SNP locus for each animal. This is called a genome-wide SNP assay.
For each performance trait, every SNP is then studied to see whether animals with a particular SNP variation tend to be at one extreme or the other for the trait. As an example, it might be found that most animals with a higher fleece weight tend to have an Adenine (A) nucleotide at a particular SNP, whereas, those with lighter fleeces tend to have a Thymine (T) at the same SNP.
Most of the 50,000 SNPs being investigated will show absolutely no relationship with a particular trait, but a small number of SNPs will show some association. Some SNPs will be shown to predict a few percent of the variation in a trait; other SNPs may predict 5 to 10% of its variation. In most cases it will be a different set of SNPs that help to predict the performance of each particular trait.
Ultimately, a test will be developed that just uses the SNPs that have been shown to predict traits. As an example, this test may include prediction of yearling fibre diameter performance. There could be 12 SNPs, each with different levels of prediction. When combined, the 12 SNPs together may predict 49% of the genetic variation in that trait, with each of the SNPs predicting 1-10% of that variation. Of course, not all SNPs are known or near enough to a gene affecting the trait, so the remaining proportion of genetic variation for that trait cannot be tested in this way.
Are SNP tests practical, useful and valuable?
While the SNP test (or DNA assay) itself is done in a laboratory, the sample collection is a fast and practical process for sheep breeders, that could easily be done with little stress on new-born lambs or older sheep. Currently, two methods are being used by the Sheep CRC: a nasal swab, where a special “cotton bud” is wiped around inside the sheep’s nostril, and the second is an ear sample, where a tiny piece of the ear is removed (with normal ear marking pliers or a special ear-punch). The samples are packaged and posted (with no need for ice-bricks) to the lab.
The test is particularly useful to speed up breeding progress that is currently limited by the extended time for some traits to be expressed and tested, especially for wool traits and reproduction. Knowing the genetic merit of animals when only a few months old allows ewes to be bred via JIVET programs (Juvenile In Vitro fertilization and Embryo Transfer) and rams to be prepared and have semen collected before 6 months old. This can dramatically shorten the generation interval and accelerate a breeding program.
Also, many new opportunities will be opened to breeders with SNP tests for traits that are either expensive or difficult to measure (staple strength, feed efficiency), or require the animal to be slaughtered to gain the information (e.g. dressing percentage, lean meat yield, retail colour stability of meat).
The ability to gain a prediction of performance from a newborn lamb has considerable value, because to get the same accuracy of prediction would otherwise require many animals in the drop to be kept to a later age when most would ultimately not be required (especially in the case of rams). Maintaining those animals comes at a considerable cost, especially in pasture and supplements, labour and animal health costs, which could otherwise be used to run more breeding ewes.
The value of SNP testing can be weighed up by comparing the additional profits from running more breeders, being able to shorten the generation interval to make faster genetic gain, testing for traits not able to be otherwise tested and the cost of SNP testing early in life against costs of maintaining the extra animals, their later traditional testing costs, and lowered income (if selling culls as rams rather than wethers). While current SNP tests are expensive at the research stage, they will be lower as commercial tests, when fewer SNPs need to be tested and more testing volume occurs.
What is a SNP chip?
The current research SNP testing is done with the Ovine SNP 50, DNA analysis research chip (shown in Figure 9) from Illumina. This was developed by Illumina in conjunction with the International Sheep Genomics Consortium. The “SNP chip” is a small glass plate that has 12 panels, with each panel able to test about 50,000 SNPs for one animal, showing what nucleotide is present at each particular SNP position.
The results are generated for each animal as a list of the 50,000 SNPs and one of the two options beside them. By themselves, the results have no meaning until they are combined with the knowledge of their associations with traits. This is done by geneticists at the Animal Genetics and Breeding unit (a partnership between University of New England and Industry and Investment NSW) who work with Sheep Genetics to prepare genomic estimated breeding values. These values are combined with an animal’s existing Australian Sheep Breeding Values (ASBVs) to prepare enhanced ASBVs.
Do you want to know more?
There is a vast quantity of information to be found on genomics on the internet; simply search for some of the terms you have read in this article.
Wikipedia “the free encyclopaedia” is a great source of information on any topic you can imagine. Start with Wikipedia’s page on genomics at this link http://en.wikipedia.org/wiki/Genomics or in a search engine (such as Google), search for any term plus the word wiki, e.g. DNA wiki.
During LambEx 2012 the Sheep CRC hosted a genomics breakfast workshop - download the papers from the workshop below.