Genetics: Basic genomic concepts
When cells divide, as they do throughout an organism's life, each new cell must acquire a complete complement of DNA. A cellular mechanism called mitosis ensures this by duplicating the DNA and apportioning a full set of chromosomes to each of the two new cells.
When sex cells or gametes (sperm or egg) are produced, by a special cell division mechanism called meiosis (figure 9), the genetic complement must be halved. Two sequential divisions produce four gametes, each containing only one representative ofeach homologous pair plus one sex chromosome. Fertilisation then restores the full chromosome complement.
During the first division of meiosis, the members of each homologous pair separate (or 'segregate') to the two resulting cells independently of the members of any other pair, so that the 23 chromosomes in the different gametes produced by an individual represent different assortments of the original 46. This is an important source of the variation that results from sexual reproduction.
Variation also results from a process called crossing over and recombination (figure 10). During the first division of meiosis, homologous chromosomes can swap portions of their DNA, which means that the chromosomes in the gametes may contain different sets of alleles from the chromosomes of the individual who produced them. There are typically 1-2 crossovers per chromosome during meiosis.
It is possible to follow the inheritance of alleles down through the generations. Because of recombination and the independent segregation of chromosomes in meiosis, most genes will show inheritance patterns that are independent of the patterns shown by other genes. For example, the genes determining eye colour will be inherited independently of the genes determining shape of the ear lobes. However, genes that are close together on the same chromosome will tend to be inherited together more often than expected, as they are less frequently separated by recombination.
The frequency with which two genes are inherited together, or linked, depends on the length of the DNA between the two genes: the longer the gap, the greater the chance that recombination will separate them (in figure 10, genes a and b are closely linked, but recombination separates them from C). Linkage between genes is the basis of genetic mapping: using patterns of inheritance to determine the relative positions of genes on the chromosomes. Linkage analysis in affected families has been an important technique for locating the genes implicated in many Mendelian diseases.
Inheritance patterns in Mendelian disease
Cystic fibrosis, Huntington disease and haemophilia are all examples of Mendelian disorders.
Cystic fibrosis is an autosomal recessive disorder (figure 11): that is, an individual will only develop the disease if he or she has two copies of the mutant allele. An individual with one normal and one CF allele will be an unaffected carrier. Although a carrier does not develop disease, he or she has a 50% chance of passing on the mutant allele to each child; if two carriers conceive a child, there will be a 25% of the child inheriting a mutant allele from both parents and of therefore of being born with the disease.
In the case of Huntington disease, the mutant allele is autosomal and dominant (figure 12). This means that individuals with the disease may have either one or two copies of the mutant allele. A person with one parent who has Huntington disease has a 50% chance of having inherited the mutant gene.
Haemophilia is an example of a sex-linked genetic disease (figure 13). This means that the gene associated with the disease is on one of the sex chromosomes, almost invariably (as in the haemophilia example) the X chromosome. Females have two X chromosomes, so a female with one normal and one haemophilia mutant allele (which is recessive) will be a normal carrier. If her partner is unaffected, all their daughters will also be unaffected (either normal or carriers) but each son has a50% chance of carrying the mutant allele on his single X chromosome, and therefore of being affected.
Some single-gene diseases are caused by mutations not in nuclear DNA, but in the DNA of the mitochondria. An example is Leber's hereditary optic neuropathy (LHON), in which blindness results from damage to the optic nerve. Because the mitochondriaare passed on via the cytoplasm of the egg, mitochondrial diseases show maternal inheritance. The severity of mitochondrial diseases can vary depending on the proportion of the mitochondria that carry the mutation.
THE RELATIONSHIP BETWEEN GENOTYPE AND PHENOTYPE
Although the genes implicated in many Mendelian diseases and some other human characteristics have now been identified, it is often not straightforward to predict an individual's phenotype from their genotype. This is because the phenotype may be influenced both by other genes and by non-genetic (environmental) factors.
The likelihood that a person carrying a disease-associated genotype will develop the disease is known as the penetrance of the condition. Penetrance has a time component. Both Huntington disease and cystic fibrosis are virtually 100% penetrant: cystic fibrosis at birth and Huntington disease by the age of about 70. Mutations in the BRCA1 gene that are associated with familial breast cancer have a lifetime penetrance of 60-85%. Variations in penetrance are caused by the modifying effects of other genes and/or by environmental factors (figure 14).
Most Mendelian diseases were originally classified from their phenotype, in other words their clinical manifestations. As molecular causes have been identified for these diseases, it has turned out that in many cases the same or a similar phenotypecan result from a different underlying genotype. This phenomenon is known as genetic heterogeneity. Genetic heterogeneity resulting from disease-causing mutations at different genetic loci is called non-allelic or locus heterogeneity. An example is autosomal dominant polycystic kidney disease, which in most affected families is caused by mutation of a gene on chromosome 16 but can also arise from mutation of a different gene, on chromosome 4. When different mutations at the same locus cause the same disease, this is known as allelic heterogeneity. Allelic heterogeneity is very common; for example, hundreds of different pathogenic mutations have been discovered in the BRCA1 and BRCA2 genes, and in the CFTR gene associated with cystic fibrosis.
The spectrum and relative frequencies of disease-associated alleles may also vary in different populations as a result of their different genetic ancestry; for example cystic fibrosis is most common in populations originating from the north and west of Europe, whereas sickle-cell anaemia is more common in African or Afro-Caribbean populations and Tay Sachs disease is relatively more prevalent in Ashkenazi Jewish communities (table 2).
Even if a disease phenotype is fully penetrant, the severity and symptoms of the disease can vary in different affected individuals. Sometimes it is possible to predict phenotype from genotype with reasonable accuracy. For example, myotonic dystrophyis caused by an expanded tract of repeated DNA sequence in a gene on chromosome 19. The larger the expansion, the more severe the disease, and the earlier the age of onset. In other cases there is no clear relationship between the genotype at a known causal locus and the disease phenotype, presumably because the phenotype is also influenced by other genetic and environmental factors. An example of a disease showing such variable expressivity is Marfan syndrome, a disease affecting the skeletal system, the eyes and the heart, which can vary widely, even among affected individuals from the same family. Variable expressivity can be considered the rule rather than the exception for virtually all genetic disease, suggesting that the concept of the 'single-gene disease' is in fact very much an over-simplification.
COMMON MULTIFACTORIAL DISEASES, GENE-ENVIRONMENT INTERACTIONS AND THE ROLE OF GENES IN HEALTH AND DISEASE
Although in the past it has been common to think of diseases as having predominantly either a genetic or an environmental cause, most if not all diseases arise from a complex interplay of multiple genetic and environmental factors. Even conditions that we regard as totally environmental, such as injury in a motor accident, may be influenced by genetic factors - in this example, perhaps, factors influencing risk-taking behaviour.
The disease phenylketonuria (PKU) illustrates the difficulty of trying to label a disease as either genetic or environmental, or some fixed and quantifiable mixture of the two. PKU is caused by a mutation that leads to loss of function of the enzymephenylalanine hydroxylase. Its absence allows the build-up of phenylalanine hydroxylase, which is toxic for the developing brain. However, in the absence of phenylalanine, no disease develops: although we tend to think of PKU as a genetic disease, both a specific genotype and an environmental factor are needed for disease to develop.
Whether we think of a disease as primarily genetic or primarily environmental in aetiology depends on the relative prevalence of the causative genetic and environmental factors. For example, in the past when the bacteria that cause tuberculosis were ubiquitous in the UK, doctors understood that it was inherent 'consititutional' (i.e. genetic) factors that determined which individuals succumbed to the disease. Now that the tubercle is rare in most developed countries, TB is thought of entirely as an infectious (environmental) disease caused by exposure to the bacteria, and genetic susceptibility is generally not considered.
The discipline of public health genetics is centred on the critical involvement of genes as determinants of health, and their interactions with other genetic, environmental or social factors. Although some diseases tend to fall at one end or the other of the spectrum between genetic and environmental causes (Figure 15), most fall somewhere in the middle. This is an important principle because in terms of prevention, a central goal of public health, it means that risk of disease may be modified by altering the environmental part of the interaction. The high incidence of diabetes in Pima Indians has been explained as the result of exposure of a particularly 'thrifty' genotype to a Western diet. Genetic factors are clearly at work but in practical terms it is by changing the diet that it will be possible to reduce the incidence of diabetes in this population.
Identifying genes and gene-environment interactions implicated in common disease
Where genetic effects are relatively small, as they are expected to be for the polymorphisms associated with susceptibility or resistance to common diseases, it is difficult to demonstrate a reliable association between a genetic variant and a disease. Genetic epidemiologists use various types of study design to investigate gene-disease relationships, including cohort studies, family-based linkage or association studies or, more commonly, population-based association studies. The latter are essentially case-control studies in which DNA is compared from groups of healthy (control) and affected individuals, to identify whether a specific genetic variant(s) shows a statistically significant association with the disease phenotype (box2).
Other genetic epidemiology studies seek to identify and quantify the combined contributions of genetic and environmental factors to disease risk. For example, tobacco smoke is a known environmental toxin and differential genetic susceptibility to the carcinogenic effects of tobacco smoke has been investigated. The N-acetyltransferase genes, NAT1 and NAT2, are involved in the metabolism and detoxification of aromatic monoamines, carcinogenic compounds found in tobacco smoke. Synthesis of evidence from several studies suggests that smokers carrying certain allelic combinations of the NAT2 gene that confer a 'slow acetylator' phenotype are more susceptible to bladder cancer than those with the 'fast acetylator' phenotype. Genetic effects on reactions to cigarette smoke are likely to be complex, however, involving interactions among variants of several genes.
Genetic epidemiology studies are difficult and expensive to carry out. Large numbers of cases and controls are generally needed in order to obtain statistically robust results. There are many potential problems including genotyping errors, difficulties in defining the phenotype of interest (potentially leading to aetiological heterogeneity in the disease group), various sources of possible bias, and confounding due to population stratification. Type 1 statistical errors (false positive findings) are likely in studies that test a large number of different variants at different loci; stringent criteria for significance are needed in these studies.
As a consequence of these difficulties, very few reported associations have been independently verified. For this reason, and because any individual polymorphism is expected to have only a small effect on disease risk (increasing or decreasing it by less than 50%), there are currently no clinical applications of genetic testing for susceptibility to common disease. Tests for multiple genetic variants may have greater predictive value but it is likely to be many years before the evidence base for such tests can be assembled and evaluated. As part of this evaluation there is a need for social and behavioural research, to improve understanding of how people perceive risk, and whether knowledge of genetic susceptibility would be likely to act asa motivator for preventive action or would lead to counter-productive feelings of fatalism. Genetic susceptibility tests may turn out to be most useful as part of a risk assessment 'package' that includes biochemical and physiological measurements as well as lifestyle factors.
© Public Health Genetics Unit 2006