Human and medical genetics
Friedrich Vogel and Arno G. Motulsky
Recent developments in human and medical genetics
Human genetics and disease
Public health policy considerations regarding amniocentesis
Other chromosomal anomalies
Mutations and hereditary diseases
Frequencies of hereditary diseases
Complex multifactorial anomalies and diseases
Common chronic diseases
Mental diseases and mental retardation
Novel methods to study complex diseases
Cancers: somatic genetic diseases
Cytogenetics of cancers
Molecular applications to cancer therapy and prevention
Genetics as a basic science of medicine and public health
Genetic variability in the ‘normal’ range
Genetic polymorphisms and disease
Genetic variability in reaction to drugs: pharmacogenetics
Genetic variation in the reaction to food and other environmental factors: ecogenetics
Lactose digestion polymorphism
Applications of genetic knowledge in medical practice
Assessment of genetic risks and genetic counselling
Organization of genetic counselling
Public health or community genetics
Recent developments in human and medical genetics
During recent decades, human genetics has seen enormous progress. Up to the late 1950s, many traits with simple monogenic modes of inheritance were already known. Most of them were hereditary diseases, usually rare. However, twin and family studies indicated a genetic component for more common complex diseases. So-called ’empiric risk’ figures to predict recurrence risks were useful for genetic counselling in such instances, but the nature of the underlying genetic variability was largely unknown. In the 1950s, methods for microscopic study of human chromosomes were developed, and the normal chromosome number was established to be 46 (Tijo and Levan 1956). Soon afterwards, some previously unexplained birth defects were found to be due to numerical or structural chromosomal aberrations; for example, Down syndrome as trisomy 21, the Klinefelter syndrome as 47,XXY, and the Turner syndrome as 45,X. Many different structural defects of chromosomes were discovered as deletions, or translocations (Vogel and Motulsky 1996). In the late 1960s, amniocentesis followed by analysis of fetal cells suspended in the amniotic fluid was shown to be useful for genetic, in most instances cytogenetic, diagnosis of various fetal anomalies. Together with improvements in the biochemical and molecular characterization of monogenic diseases, this development led to an increased demand for genetic diagnosis and counselling, mainly in industrialized countries such as North American and western Europe, but later in other parts of the world as well. To meet this demand, many genetic counselling centres were established.
In the early 1970s, a scientific revolution began. The concept of ‘molecular disease’ had been proposed in 1949 by Pauling and his coworkers when they discovered that sickle cell anaemia was caused by a genetically determined, electrophoretically detectable defect of the haemoglobin molecule (Pauling et al. 1949). In the 1970s, methods were developed for studying the genetic material (DNA) directly (Watson et al. 1987, 1992; Gelehrter and Collins 1990; Vogel and Motulsky 1996). This led to the discovery of many DNA genetic polymorphisms, that is, variants without phenotypic effects that were usually located outside coding DNA sequences. As such variants are inherited as Mendelian traits, they are useful markers for localizing genes by genetic linkage studies. The chromosomal localization or position of a disease-producing gene can often be demonstrated by showing linkage with such a DNA marker. A new methodology of defining genetic disease—positional cloning—now became possible. This approach allowed isolation of the gene followed by determination of the nature of the various mutations that interfered with gene function. This approach has become a highly successful strategy for identifying genes and for elucidating their mechanism of action. A total of about 7000 genes had been mapped by March 2001 as compared with fewer than 3000 in 1994 (World Wide Web URL http://www.ncbi.nlm.nih.gov/omim; see also McKusick (1986, 1987, 1988, 1994)). This success with localization of single genes has raised hopes that similar approaches could identify genes involved in complex diseases of multifactorial origin. In principle, linkage studies should permit identification of the major genes involved in the causation of such diseases, especially when combined with modern methods of genetic epidemiology such as segregation analysis (for a review of computerized analytical systems, see Fischer et al. (1996)). So far, the results of such studies in complex diseases (except for certain monogenic subtypes) have been disappointing. However, the approach remains sound and has been applied successfully to genetic counselling and prenatal diagnosis for monogenic diseases.
Human genetics and disease
Chromosomes consist of a continuous DNA structure and of certain histone and non-histone proteins. The microscopically visible chromosomes transport the genetic material and normally guarantee a regular distribution of DNA to daughter cells in cell division. As a rule, human chromosomes are studied in metaphase (that is, after replication of the DNA double helix), but immediately before the replicated chromosomes are distributed to the two daughter nuclei. Recently developed methods of ‘chromosome painting’ permit the study of chromosomes in interphase, that is, when the nucleus does not divide (Vogel and Motulsky 1996). However, most diagnostic chromosome studies are being performed on metaphase chromosomes of readily available blood lymphocytes. In short-term cultures, lymphocytes are first induced to divide and are then stained and photographed. Individual chromosomes can be identified by their length, shape, and banding pattern. The respective chromosomes of all nucleated cells in an individual (such as fibroblasts) are identical and can be studied in a similar manner to lymphocytes.
The most common numerical anomalies are the trisomies—a chromosomal complement is present in triplicate rather than in duplicate. In most instances, such trisomies are caused by an error during reduction division (meiosis). Normally, the number of chromosomes is reduced by half (in humans, from 46 to 23) to form germ cells; two chromosomes that should be distributed to daughter cells may stick together (‘non-disjunction’) and remain in one cell, making that egg (or sperm) became a trisomic fertilized egg. Most trisomies lead to a severely malformed embryo; as a rule, such an embryo is spontaneously miscarried during the first trimester of pregnancy. According to international statistics, about 15 per cent of recognized pregnancies end in spontaneous abortion and about 50 per cent of these are caused by various chromosomal aberrations. Many of them are trisomies. The high proportion of postnatal survival in trisomy 21 (Down syndrome) is an exception (about 25 per cent of affected fetuses), reflecting the small size of this chromosome, carrying relatively few genes.
The incidence of Down syndrome (trisomy 21) is about 1 in 700 live births; trisomy 21 is caused by non-disjunction of chromosomes within maternal germ cells in about 80 per cent of cases. Other autosomal trisomies include trisomy 13 and trisomy 18, which are much rarer and lead to much more severe malformations causing spontaneous abortions or early death postnatally. X-chromosomal trisomies include the XXY (Klinefelter syndrome) and XXX conditions and are often associated with mild mental retardation.
The risk of having a child with trisomy due to maternal non-disjunction increases with advancing age of the mother; a 45-year-old mother has a 10- to 20-fold risk of giving birth to a trisomic child as compared with a woman of 20. Therefore, the incidence of trisomies in various populations depends critically on the age distribution of the reproducing female population.
Public health policy considerations regarding amniocentesis
In most developed countries, prenatal diagnosis (amniocentesis or chorionic villus biopsy) is offered to pregnant women of ‘advanced’ maternal age to detect Down syndrome, thereby allowing the option of pregnancy termination. The selected ‘cut-off’ age for performing prenatal diagnosis varies in different countries. In the United States, an age of 35 years is generally chosen. In some European countries, a somewhat older age is selected, for example 36, 37, or 38 years. Since fewer women initiate pregnancies as they become older, the total number of antenatal procedures becomes smaller as the ‘cut-off’ age becomes higher. However, as older women have more affected fetuses, the proportion of detected cases rises. For instance, the incidence of trisomy 21 at amniocentesis is about 1 in 250 at 35 years, 1 in 150 at age 37 years, 1 in 100 at age 39 years, and 1 in 60 at 41 years. The selection of the cut-off age becomes a difficult issue of public health policy. Delaying the age for antenatal diagnosis becomes less expensive for publicly supported health services by reducing the number of antenatal procedures. However, as most cases of Down syndrome in a population are born to mothers younger than 35 years of age, fewer affected fetuses among the entire population of all pregnant women will be detected. In a pluralistic health service scheme such as that in the United States, many younger women select amniocentesis at an age younger than 35 years in order to reduce their chances of having a child with Down syndrome. In the European health system that has grown out of the belief that the state has some responsibility for the health of its citizens (Häfner 1999), stricter regulations exist—especially regarding payment for diagnostic efforts. The rate of uptake for amniocentesis usually does not reach 50 per cent of women above the recommended cut-off age. The highest rates have been observed in Denmark where about two-thirds of women above the age of 35 undergo the procedure (Galjaard 1994). These considerations point to the desirability of simple screening tests (such as low levels of a-fetoprotein and detection and (hopefully) analysis of embryonic cells in the maternal blood) to detect trisomies in women of all maternal ages (see below).
Other chromosomal anomalies
Structural chromosomal aberrations are much rarer than trisomies. In deletions, part of a chromosome is lacking. In reciprocal translocations, chromosome parts are exchanged between chromosomes. The resulting phenotypes depend on whether chromosomal material is lacking or is increased, and differ with the chromosomes involved. Different chromosome anomalies have some features in common (Table 1). A detailed description of such chromosomal syndromes has been given by Schinzel (1984).
Table 1 Common findings in autosomal chromosome abnormalities
A group of chromosomal anomalies known as contiguous gene syndromes were discovered when some unusual phenotypes suggesting simultaneous transmission of two or more genetic diseases were analysed at the DNA level. In these instances, no chromosomal defect could be seen by conventional techniques. However, the method of fluorescence in situ hybridization may permit a diagnosis. Such patients often suffer from signs of more than one hereditary disease (for example, X-linked Duchenne muscular dystrophy and chronic granulomatous disease). Molecular studies in such cases reveal a deletion spanning more than one gene but not extensive enough to be discovered by conventional microscopic studies.
This category comprises conditions that are conventionally called ‘hereditary diseases’. They follow Mendel’s laws and are sometimes referred to as Mendelian diseases. If a heterozygote (that is, a person carrying one mutant gene) shows the anomalous phenotype, the mode of inheritance is dominant. Such a disease will be transmitted from one parent to 50 per cent of the offspring on average. This transmission occurs irrespective of gender. In autosomal recessive inheritance, the abnormal phenotype occurs in homozygotes only, that is, when both homologous chromosomes carry a mutation in the same gene. Hence both parents must be (at least) heterozygous for this gene. Within such families, both parents are unaffected; among their children, there is a ratio of one affected homozygote to three unaffected children (two of which will be heterozygous). As both parents have to carry a mutation in the same gene, it is not surprising that children from matings between relatives (such as first cousins who have inherited the abnormal gene from a common ancestor) run an increased risk of being affected, particularly if the gene is very rare.
These modes of inheritance are observed if the mutant gene is located on one of the 22 pairs of autosomes. If the mutant gene is located on the X chromosome, the mode of transmission is influenced by the mechanism of sex determination. Women have two X chromosomes, men carry only one X chromosome. Hence males carrying a single X-linked recessive gene will show the mutant phenotype; they are referred to as hemizygotes. In contrast, heterozygous females for recessive X-linked traits will usually be unaffected clinically but transmit the mutant gene to 50 per cent of their sons who will then be affected. Transmission from father to son cannot occur with X-linked inheritance.
The various Mendelian modes of inheritance suggest different biochemical mechanisms. Recessive diseases are often caused by enzyme defects. Heterozygotes produce about half the amount of enzyme protein compared with the normal homozygotes. In most instances, this reduced amount is sufficient for the maintenance of normal function (but see Vogel (1984)). Mendelian dominance (that is, manifestation of a clinical abnormality in heterozygotes) requires a more complex interaction between the gene products of the two alleles. Such interaction occurs, for example, if a protein is needed for building a structure. To give an example from daily experience, a wall that is built from 50 per cent normal and 50 per cent defective bricks will be defective. Examples are dominant diseases that are caused by the presence of an abnormal collagen, an important component of connective tissue. Patients with osteogenesis imperfecta suffer, among other clinical signs, from frequent bone fractures after trivial trauma. Other mechanisms of Mendelian dominance have been discussed elsewhere (Vogel and Motulsky 1996).
The phenomenon of anticipation has been observed by perceptive physicians for about a century. Certain diseases, such as myotonic dystrophy, tend to have an earlier onset and a more severe course in successive generations. For a long time human geneticists tried to explain this finding as reflecting bias, because it did not fit the theoretical model of Mendelian transmission. More recently, however, it has become possible to explain anticipation by a novel type of mutation. Certain genes contain sequences of many DNA base triplets. For unknown reasons, this system becomes unstable, and the number of triplets tends to increase from one generation to the next. This phenomenon leads to earlier onset and a more severe course of such diseases (Sutherland and Richards 1995). The number of diseases caused by such triplet amplifications is still increasing. In most of them, the function of the nervous and/or muscular system is impaired.
Another non-Mendelian phenomenon is genomic imprinting. The manifestations of certain mutations in offspring may vary with the sex of the parent who carries the mutant gene. In some instances, the mutant gene must be transmitted through the maternal germ line in order to lead to the mutant phenotype; in others, it has to pass the paternal line (Hall 1990; Sapienza and Hall 1995). Such observations show that the maternal and paternal genomes do not always contribute equally to the phenotype of the child. Further analysis of anticipation and genomic imprinting promises new insights into the genetic determination of embryonic development.
Mutations and hereditary diseases
About 1 per cent of all newborns suffer, or will suffer later in life, from an autosomal dominant or X-linked recessive disease (United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) 1986, 1988, 1992). Within this group the dominantly inherited hypercholesterolaemias are among the most common; about 1 in 500 individuals carry a mutation for this condition, which strongly predisposes to an early onset of coronary heart disease. Matings between two such heterozygotes may occur, so that homozygous children are observed occasionally. They usually suffer from coronary heart disease in childhood or adolescence (Goldstein et al. 1995). Reproduction by heterozygotes for this type of hypercholesterolaemia is hardly impaired, as the first clinical signs among heterozygotes only occur at the age of about 40 or 50 years or even later. Thus most people have already had their children. For this reason, almost all persons carrying this gene have inherited it from one parent, and new mutations are extremely rare. This also holds true for Huntington disease for the same reason as the average age at onset is between about 40 and 50 years. However, many other dominant or X-linked diseases have severe clinical manifestations that prevent reproduction. Examples are acrocephalosyndactyly, a severe dominantly inherited malformation syndrome, and the X-linked recessive Duchenne type of muscular dystrophy, a progressive muscular disease with death at about age 20. If such a severe disease is dominantly inherited, the affected individuals are, as a rule, unable to transmit the mutant gene to the next generation. Thus the great majority of patients are ‘sporadic’ cases, and are the only affected persons in their families. The conclusion that a dominant mutation is responsible for the disease is based on three arguments.
Reproduction is observed occasionally, and in these rare instances a 1 : 1 ratio of affected and unaffected offspring is observed.
The probability of many (not all) dominant mutations increases with the age of the father. This increase is not as pronounced as that with the mother’s age in trisomies, and its extent appears to differ somewhat between mutations. However, the risk for a man of 45 of fathering a child with such a mutant autosomal dominant phenotype may be four to five times as high as that for a man of 25. This paternal age effect has a strong influence on the population incidence of such diseases. If identical mutation rates are assumed, the risk of having a child with a disease caused by such a mutation in Pakistan has been estimated to be almost twice that in Bulgaria, simply because more men have children at a relatively advanced age in Pakistan (Modell and Kuliev 1990) (Table 2) (the United States is similar to Bulgaria).
Table 2 Relative mutation rates in European countries in relation to father’s age
The third and most direct argument is direct demonstration of a mutation in the responsible gene. Of course, this requires direct knowledge of this gene, but the evidence is now available for an increasing number of such genes (Vogel and Motulsky 1996). Many such mutations are point mutations in the strictest sense, that is, only one base pair of the approximately 6 to 7 × 109 base pairs in the diploid human genome (46 chromosomes) is mutated. Other mutations are deletions spanning a part of the affected gene (see Cooper and Krawczak (1993) for a detailed report on the molecular basis of human mutations). Newly discovered mutations are being documented in databases such as the Human Gene Mutation Database (Cardiff) (http://www.uwcm.ac.uk/uwcm/mg/hgdm0.html), which has links to many disease-specific registries.
In X-linked recessive diseases that prevent male patients from reproduction (such as Duchenne muscular dystrophy) only about a third of the affected patients are ‘sporadic’ cases; in the majority, the mutant gene is transmitted by the clinically unaffected heterozygous mothers. This is of practical importance because potential female carriers in such families often request genetic counselling and prenatal diagnosis. Here, risk determination must be supported by DNA studies regarding the presence or absence of the mutation or, if the mutation is unknown, by family studies using DNA markers that segregate together with the mutation.
The fraction of new mutants among patients having a disease that is caused by a dominant or X-linked recessive mutation is proportional to the degree to which this disease reduces reproduction of these patients, that is, to the severity of natural selection against this mutant. In diseases where early death prevents reproduction, most instances are cases who owe their disease to a fresh mutation (Table 3). When the mutation rates (that is, the probabilities of mutations in germ cells) can be estimated from epidemiological data in human populations, mutation rates per fertilizing germ cell and per generation have been estimated to range from somewhat less than 104 to about 106 (Vogel and Rathenberg 1975). However, there is good evidence that some mutations leading to hereditary diseases have a lower frequency (Stevenson and Kerr 1967).
Table 3 Approximate percentages of patients affected by new mutations in autosomal dominant disorders
The mutation rates for some dominant mutations are assumed to have decreased in recent generations, as fewer ‘older’ fathers, who produce more mutant sperm, have offspring in modern populations. However, a possible (small) increase may be anticipated because of more exposure to possibly mutagenic agents, such as ionizing radiation or mutagenic chemicals (see below). Selection is also changing for some diseases, mainly because of medical intervention. In the X-linked haemophilias, for example, patients are now being treated successfully with clotting factor preparations. This means higher reproduction rates and, with a constant mutation rate, a higher incidence. A counterbalancing factor could be genetic counselling and prenatal diagnosis followed by pregnancy termination, which could lead to fewer affected individuals in future generations.
Another example is retinoblastoma, the malignant eye cancer of children. About 40 per cent of sporadic cases are due to a new germ cell mutation transmitted to children. Until about 130 years ago, almost all affected children died. Today, however, at least 80 per cent survive due to surgical or radiation treatment and lead a normal life. If the mutation rate remains the same, successful therapy will lead to a substantial increase in the disease (Vogel 1979). Genetic counselling and prenatal diagnosis are likely to counteract this trend. In many dominant and X-linked conditions, a trend towards increase of a disease is unlikely as no treatments are yet available. Examples are acrocephalosyndactyly, achondroplasia, and osteogenesis imperfecta.
Frequencies of hereditary diseases
Whereas autosomal dominant and X-linked diseases have similar rare frequencies in most populations, this is not the case in autosomal recessive conditions. These diseases only occur if the patient is homozygous for a mutant gene; the risk for homozygosity increases with increasing genetic similarity of his or her parents, and as a rule, with the degree of their biological relationship. The increased incidence of these diseases among children from matings between first cousins is the best-known example, but the principle also holds for more remote degrees of relationship. Therefore, frequencies for autosomal recessive diseases depend largely on the breeding structure of the population. In a relatively small (and, for a long time in the past, more or less isolated) population a mutant gene might become relatively common just by chance (‘genetic drift’), but it might also have had a selective advantage at some time in the past. The more common a gene is among heterozygotes, the more the risk increases that two heterozygotes will mate and produce homozygous offspring. This risk is particularly high in populations originating from relatively small ‘founder’ groups that have lived in relative isolation for several generations. Examples from the North American continent are the French Canadians in Quebec (Scriver 1992) and the Amish of Pennsylvania. In such ‘isolate’ populations, a given mutation was introduced by a single founder and increased in frequency with subsequent expansion of the population.
In contrast, the high incidence of sickle cell anaemia among Africans and their descendants elsewhere in the world is an example of selection. Falciparum malaria was hyperendemic in Western and Central Africa, and heterozygotes for the sickle cell gene had a selective advantage by dying less frequently from the endemic infection. The high incidence of Tay–Sachs disease and several other recessive diseases among Ashkenazi Jews has not been explained conclusively. Recent evidence makes a founder effect most likely (Motulsky 1995). The proportion of consanguineous matings, particularly first-cousin matings, has decreased in the last 100 years from a few per cent to a few per thousand. This trend presumably has appreciably reduced the number of patients suffering from autosomal recessive diseases—even below the rate expected from genetic equilibrium between new mutations and selective disadvantage of affected homozygotes. Therefore, industrialized Western countries are currently enjoying a situation in which the incidence of autosomal recessive diseases is as low as it probably ever has been in human history. UNSCEAR has estimated this incidence at about 0.1 per cent of newborns. The actual frequency is likely to decrease further, as prenatal diagnosis with pregnancy termination has become possible in an increasing number of such diseases. However, consanguineous matings are still common in some populations of developing countries—especially in Arab countries and in parts of India. In such consanguineous matings, in addition to autosomal recessive diseases the rates of congenital malformations among children and, according to some reports, the rates of stillbirths, abortions, and neonatal deaths are increased due to homozygosity of mutant genes that are transmitted from common ancestors of the two mates (Vogel and Motulsky 1996). Considerations of public health alone would argue for discouraging consanguineous matings to prevent ill health. Yet, the custom of consanguineous marriages in societies where they are common is an important and integral part of their culture. The institution of public health policy to reduce such marriages, therefore, needs to be carefully considered.
Complex multifactorial anomalies and diseases
In many diseases, there are neither a microscopically visible chromosomal aberration nor a simple monogenic mode of inheritance. Familial aggregation and a higher concordance of monozygotic compared with dizygotic twins point to a contribution of genetic factors. Three broad disease groups can be distinguished: birth defects, common chronic diseases, and mental diseases.
Some malformation syndromes are caused by a numerical or structural chromosomal aberration (see above). In a few others, a monogenic mode of nheritance can be demonstrated. Most often, however, neither explanation holds true. Sometimes a slight familial aggregation is observed. The model of multifactorial inheritance in combination with a threshold effect is sometimes invoked. Exogenous agents such as exposure to a radiation, teratogenic drugs, or viral infections during early pregnancy can rarely be demonstrated. Sometimes a malformation could be the result of entirely random events without specific genetic or environmental causes. Statistics on the incidence of birth defects in various populations exist, but, owing to differences in definitions and ascertainment, the results are difficult to compare. More reliable data are available from Hungary, which has had a system of direct ascertainment, registration, and extensive study of birth defects since about 1970 (Czeizel and Sankaranarayanan 1984) (Table 4). Data from many other populations have been collected in the World Atlas of Birth Defects, the first edition of which was published in 1998 by the International Center for Birth Defects in co-operation with the European Registration of Congenital Anomalies and the World Health Organization. But for many countries, even some of the industrialized countries of the West, sufficiently reliable data are not available.
Table 4 Relative frequencies (per 10 000 births) of congenital anomalies in the United States, Hungary, and British Columbia, Canada
Common chronic diseases
Details of the common chronic diseases are given by King et al. (1992). Many people, unless killed by an injury or an acute infection, die from chronic diseases, including coronary heart disease, cancers, diabetes, high blood pressure, and others. Coronary heart disease and diabetes illustrate the principle of aetiological heterogeneity; such diseases are caused by different aetiologies, but manifest with similar phenotypes. Typically, a genetic predisposition combined with environmental influences causes the disease. We illustrate this with diabetes mellitus.
There are several types of diabetes: type 1 and type 2 diabetes are most common. Type 1 diabetes often manifests during childhood or adolescence. The onset often follows an acute viral infection that appears to precipitate the destruction of insulin-producing islet cells in the pancreas, often by an autoimmune mechanism. Therapy requires insulin substitution. There is a strong association with HLA types (DR3 and DR4). A major gene locus in the HLA region (chromosome 6p) has been identified by linkage studies. Another locus near the insulin gene (11p), as well as other less certain predisposing loci, has also been detected (see also Davies et al. 1994).
Type 2 diabetes usually manifests in middle or advanced age, often associated with obesity and overnutrition. The specific genes involved are not known but are under active investigation. Contrary to type 1 diabetes, concordance of monozygotic twins borders on 100 per cent. Relatives are frequently affected; however, except for rare autosomal dominant subtypes (such as mature onset diabetes of the young), monogenic inheritance cannot be demonstrated. Type 2 diabetes disappeared almost completely under conditions of undernutrition in central Europe near the end and after the Second World War. In contrast, the frequency of type 1 diabetes did not change during this period when severe food shortages were widespread. These observations demonstrate the importance of food intake in the pathogenesis of type 2 diabetes. Many individuals do not appear to be genetically equipped to cope with overnutrition. Insulin therapy is usually not required in type 2 diabetes; weight reduction, exercise, and occasionally drug therapy are sufficient. Type 2 diabetes is an excellent example of how a disease may appear to be largely genetically determined from a geneticist’s point of view, whereas, for a nutritionist, this condition is a typical product of environmentally determined overnutrition in Western affluent societies. Both views are correct!
Some decades ago Neel (1962) proposed the hypothesis that type 2 diabetes might be caused by a ‘thrifty genotype’. He suggested that the gene or genes underlying diabetes might be an adaptation to long-lasting conditions of food shortage and starvation. Genes that increased mobilization of carbohydrate may have enabled their carriers to survive and reproduce. There is circumstantial evidence in favour of this hypothesis. India is a country in which the majority of the population have suffered for a long time from food shortages. Indians who have emigrated and are living under affluent conditions have a higher frequency of type 2 diabetes. Among some Amerindian tribes, diabetes and obesity also have become very common under the conditions of the current Western American diet. Type 2 diabetes, despite its genetic determination, can often be prevented by avoiding overnutrition.
Many rare forms of diabetes also exist. Often, diabetes is only one part of a more complex syndrome. Rare mutations affecting the insulin molecule or insulin receptors have been identified. Such findings are typical for many ‘complex’ diseases. Upon detailed clinical and pathophysiological study, rare types can be distinguished from the more common variety. Monogenic modes of inheritance are often identified in the rarer subtypes and may aid in the elucidation of the more frequent forms. Familial hypercholesterolaemia is an example (Goldstein et al. 1995).
Table 5 (Czeizel et al. 1988) gives some data on lifetime prevalence of some common chronic diseases in Hungary.
Table 5 Some multifactorial diseases in Hungary (1977–81)
Mental diseases and mental retardation
The third group of complex diseases comprises mental diseases and mental retardation (Propping 1989; Tsuang and Faraone 1990, 2000; Gottesman 1991; Vogel and Motulsky 1996). Their significance for public health is high.
The two main groups of mental diseases are schizophrenia and affective disorders, which appear to represent separate diagnostic entities with little overlap. The world-wide incidence of schizophrenia is about 0.5 to 1 per cent. The concordance rate is much higher in monozygotic than in dizygotic twins, but varies between studies and as a rule is higher when rates in twins of institutionalized index patients are compared with twin registry data of entire populations (mainly in Scandinavian countries). As an overall average, about 50 per cent of monozygotic pairs tend to be concordant for clinically verified schizophrenia. Other relatives are also more frequently affected than members of the general population, but such familial aggregation of schizophrenia could also be caused by pathogenic factors in the family environment. Evidence in favour of a genetic interpretation of the twin and family data was provided by various adoption studies. Children of schizophrenic parents who were adopted at a very young age by mentally healthy couples showed the same frequency of schizophrenia as children who had lived in a family with one affected parent. Studies of this type and comparisons of mothers and fathers of schizophrenic adoptees excluded intrauterine maternal influences. However, the operation of genetic factors does not mean that the environment is not important at all. After all, the concordance of monozygotic twins is only 50 per cent, allowing considerable opportunities for environmental or as yet unexplored endogenous and random factors. Life events such as loss of a parent or partner, loss of a job, a major somatic disease, and many other factors may trigger the outbreak of schizophrenia. The ‘vulnerability’ concept, which emphasizes complex interactions between genetic factors and negative life experiences, is popular among psychiatrists.
Within the affective disorders two groups can be distinguished: manic-depressive (or bipolar) disease and simple depression. In manic-depressive disease, episodes of mania are observed in addition to episodes of depression. Among relatives, persons with typical manic-depressive manifestations are observed side by side with others who only suffer from depressions. Concordance of monozygotic twins for affective disorders is of the order of about 70 per cent. The population incidence is slightly lower than that of schizophrenia. In families of probands with endogenous depression, mainly depressions are found; empirical risks are slightly lower than in relatives of manic-depressive patients. The incidence of endogenous depressions is difficult to determine. There is a continuum ranging from transient mood fluctuations to severe depressive episodes without readily recognizable external causes.
Epidemiological studies of entire populations have yielded a surprisingly high prevalence of rates of, mostly transient, psychiatric symptoms that, according to psychiatric criteria in the United States and Western Europe, would have required some form of therapy. However, only a small fraction of these individuals is seen by psychologically or psychiatrically trained caregivers. The severity of the symptoms, the socio-economic and educational status, and general economic conditions determine whether such therapy is obtained.
From a public health point of view, alcoholism is particularly important. In industrialized Western countries, several per cent of the population can be regarded as alcoholics. Much alcoholism has an environmental explanation: if alcoholic drinks were not available at relatively low prices, alcoholism would be less widespread. Another reason is inducement to social drinking by group pressure. Family, twin, and adoption studies suggest the role of genetic susceptibility factors in alcoholism (Omenn and Motulsky 1972; Omenn 1988; Propping 1992), but no specific genes have been identified. The reaction of the brain, as assessed by EEG (Propping 1977; Vogel 2000), appears to be one such factor. Persons exhibiting relatively regular a waves in their resting EEGs do not show a major change of EEG patterns after moderate alcohol intake. In contrast, individuals with a poorly developed a rhythm often develop a regular a pattern under such conditions. They ‘feel much better’ after alcohol intake and therefore may be more susceptible to the development of alcoholism.
A genetic variant of the enzyme aldehyde dehydrogenase, which determines the second step in the metabolic decomposition of alcohol, is particularly common in Oriental populations (about 50 per cent among Japanese). The variant enzyme acts more slowly than the ‘normal’ type, leading to the accumulation of acetaldehyde after alcohol intake. The resultant ‘flushing’ (red face, perspiration, increased heart rate, malaise) appears to deter gene carriers from drinking excessively as concluded from the very low frequency of this variant among Japanese alcoholics compared with controls (Harada et al. 1982; Shibuya 1988). Therefore, this common polymorphism appears to protect against alcoholism and alcoholic liver disease, and can be considered as an antialcoholism gene.
Another group of conditions of great societal significance is mental retardation. It is useful to distinguish between mild (high grade, subcultural) and severe (low grade, mental deficient) individuals. If an IQ of less than 69 is taken as a criterion of mental retardation, the mild variety comprises about 2 to 3 per cent of the population. Only about 0.25 per cent are categorized as severely affected (IQ < 50). The mild group generally shows a high concordance of monozygotic twins and a high incidence of similar cases in the family; they can be interpreted as constituting the lower end of the bell-shaped IQ distribution in the population. The severely mentally deficient group is very heterogeneous. It comprises patients who have suffered intrauterine or postnatal brain damage, those with chromosomal aberrations such as Down syndrome (trisomy 21), and many others. Mental retardation is also one of the relatively constant common clinical signs of autosomal chromosomal aberrations. Some autosomal dominant diseases (for example, tuberous sclerosis) and many autosomal recessive conditions contribute to the pool of the many Mendelian disorders that cause mental retardation.
In recent years, X-linked mental retardation has attracted special attention (Sutherland and Richards 1995). It had been known for some time that severe mental retardation is much more common in males than in females, but this observation was explained by assuming ascertainment biases. We now know that a great number of X-linked types of mental retardation exist. One common type, often associated with a characteristic facial physiognomy, is characterized by a microscopically visible attenuation at the tip of the long arm of the X chromosome (fragile X). This common defect has an incidence of 1 in 4000 (Tariverdiau and Vogel 2000). It is the second most common single genetic condition causing mental retardation, after Down syndrome. The mutation causing this anomaly is an amplification of a base triplet, similar to the basic defect in myotonic dystrophy and in Huntington disease. This amplification only occurs in the female germ line (Vogel and Motulsky 1996). It is not entirely clear why the fragile X syndrome is so common. In addition to a high mutation rate, a selective advantage of female heterozygous carriers by increased reproduction in earlier times has been discussed.
Patients with any type of severe mental retardation pose a societal problem, as many of them need to be taken care of during their entire lifetime. Individuals and families with mild and borderline mental retardation are increasingly requiring attention and social aid. In rural societies of the past, it was easy to find adequate jobs for such individuals—simple farm and garden work. In modern industrialized societies, such work is increasingly done by machines and it is becoming more difficult to find suitable and adequately paid occupations. The medical geneticist can aid families by offering prenatal diagnosis. This is possible for Down syndrome, familial chromosomal aberrations such as various translocations and the fragile X syndrome, and the many autosomally recessive metabolic defects diagnosed with biochemical and/or DNA methods.
Novel methods to study complex diseases
So far, only traditional methods of genetic analysis—twin, family, and adoption studies—have been mentioned. Molecular genetics has provided new and efficient tools such as linkage studies using DNA markers with subsequent isolation of disease genes and their mutations. This strategy has proved successful for analysis of genes and mutations determining many hereditary diseases with simple Mendelian modes of inheritance. Under such modes of inheritance, each individual, based on phenotype, can be attributed to a specific genotype. The LOD score method for linkage study is used to localize the mutant gene to a chromosome; the likelihood of linkage (or cosegregation) of a marker gene and a disease gene compared with no linkage is estimated (Ott 1991). If the mode of inheritance is not definite, the ‘affected sib pair’ method or, more generally, the ‘affected family member’ method, as well as the examination of ‘haplotype sharing’ are preferred, but the sample sizes required are much larger than those needed for the classical LOD score method. The principle of these methods is based on the increased probability that two family members who are affected with the same hereditary or partially hereditary disease will also share part of the haplotypes of other genetic variants—especially DNA markers that are located on the same chromosomes—as the mutant genes responsible for this disease. The probability for such haplotype sharing is higher, the closer such markers are located to the disease gene in question (Van der Meulen and te Meerman; in Edwards et al. 1997). In addition to linkage studies, the principle of haplotype sharing can also be used for studying the origin and age of mutant genes in populations (te Meerman and Van der Meulen 1997). Geneticists are now applying this approach to pedigrees with many complex classic diseases and the major psychoses. Improvements of the basic strategy are being sought by studying extensive pedigrees suggesting autosomal dominance, or investigating families in relatively isolated populations.
In Western societies, about 25 to 30 per cent of the population will die from malignant neoplasias such as various cancers and leukaemias. Many cancers appear to have an environmental cause, and an individual’s lifestyle influences morbidity, as shown by the temporal trend of the age-specific mortality of certain cancers over recent decades. Cancer of the stomach has become much rarer, whereas lung cancer has seen a marked increase, first in males and later in females, definitively related to cigarette smoking. The decrease of stomach cancer is most probably due to improved food hygiene. Viral infection can also predispose to cancer. Many primary liver carcinomas are observed in countries where hepatitis B infection is common. The Epstein–Barr virus is associated with Burkitt’s lymphoma in Western Africa.
What is the role of genetics? Concordance figures in monozygotic twins are not impressive and are not very much higher than in dizygotic twins. Modest familial aggregation is frequently observed but might be entirely environmental in origin. Nevertheless, impressive pedigrees suggesting autosomal dominant inheritance for certain ‘cancer families’ have been published.
The elucidation of the genetics of retinoblastoma—a malignant eye tumour—provided a clue to the genetic pathogenesis of cancer in general. About 40 per cent of all sporadic cases are caused by an autosomal dominant gene mutation that can be transmitted to the next generations. About 60 per cent of such patients suffer from bilateral retinoblastoma. The remaining 40 per cent are unilateral. All non-inherited cases are unilateral; they do not transmit the mutant gene to their offspring. Both the inherited and the non-inherited varieties are relatively rare (about 1 in 15 000 to 20 000) (Knudson 1971; Vogel 1979; Vogel and Motulsky 1996), suggesting an explanation that was confirmed later by direct molecular studies. The carriers of the inherited type are heterozygous for a germinal gene mutation in all their body cells, including the retina. The product of the normal allelic partner of this gene normally appear to prevent tumour formation. However, a somatic mutation in this allelic partner gene in a retinal cell abolishes its normal tumour suppressor function. When this happens in a person who already carries the germinal mutation, cell divisions proceed in an uncontrolled way and retinoblastoma develops. In the non-inherited form, these same tumour suppressor genes have to undergo a somatic mutation on both homologous chromosomes in the same cell in order to produce a tumour. The probability of two such rare events in one cell is much smaller than the probability of a single event. In familial adenomatous polyposis, which is rare, the large bowel is studded with epithelial polyps. Sooner or later, one or several will develop into a cancer. The various molecular events have been analysed carefully: a single mutation of a tumour suppressor gene is usually not sufficient; mutations of several such tumour promoter and suppressor genes are required for tumour development (Kinzler and Vogelstein 1995).
Cancers: somatic genetic diseases
Cancers can be considered as a special type of genetic disease. Sometimes, a single inherited mutation may segregate in a family and somatic mutations of the previously normal partner allele will set the stage for tumour formation. Early age at onset and bilateral tumours in paired organs (eyes, breast) are often observed. All cancers appear to have a mutational origin in a single cell. In contrast, for most cancers, somatic mutations occur in the affected individual only and are not inherited from parent to child.
Several breast cancer genes (Br-CA1 and Br-CA2) and colon cancer genes (mismatch repair genes) have recently been identified as being responsible for 5 to 10 per cent of all cases of breast and colon tumours, respectively. The cancers are not always specific for a given origin; additional malignancies are sometimes seen, such as ovarian cancer with the Br-CA1 gene mutation. Characteristically, not all persons who inherit the germinal cancer gene develop tumours, as the required additional somatic mutations do not always occur.
Cytogenetics of cancers
A specific chromosomal aberration in chromic myeloid leukaemia was found in the 1960s: a translocation between chromosomes 9 and 22 (the Philadelphia chromosome). This was the first example of a unique and specific chromosome aberration in neoplastic tissue. Irregular and non-specific abnormalities of cell division causing various chromosomal abnormalities are common in malignant cells and appear to be secondary effects. However, tumour-specific chromosomal aberrations are increasingly found—mainly in leukaemias, but in some solid tumours as well (Andrews et al. 1994). A translocation may lead to irregular growth and a malignancy if a gene necessary for an important step of oncogenesis comes under the control of an unrelated regulatory gene. For example, lymphomas are observed when genes coding for immunoglobulin components such as k or l chains are positioned close to such control genes.
Molecular applications to cancer therapy and prevention
So far, the results of ongoing molecular studies have had no direct influence on cancer therapy. However, experimental attempts at somatic gene therapy for cancer are being studied. Examples are the introduction of normal tumour suppressor genes or of genes that aim to stimulate immune destruction of the tumour. Leukaemias can now be treated by eliminating all cells within the haematopoietic system by massive irradiation or cytostatic treatment, and transplantation of stem cells from an individual with a compatible HLA phenotype—preferably a sibling. Molecular studies, preferably using the polymerase chain reaction method, are now able to find out whether all mutant cells were eliminated, or if a few of them have remained and might cause a relapse (van Dongen et al. 1998).
Molecular insights are also being applied to cancer prevention. Relatives of patients whose breast or colon cancers are caused by single mutant germinal genes detectable by molecular techniques are at high risk and warrant testing. Subjects found to carry the mutant genes need surveillance with more conventional methods, such as mammography and colonoscopy. Molecular tests need to be standardized before general introduction. Moreover, as several different genes and many different mutations at a given tumour gene may be responsible in different affected families, methods need to be developed to detect these mutations. Testing of family members (who may have risks as high as 50 per cent) has a much higher priority than testing in the general population, even for relatively high-frequency tumour genes (for example, about 1.5 per cent of the Ashkenazi Jewish population are heterozygotes for one of three breast cancer genes). Many problems abound. It will be difficult to convey to prospective testees that a negative test for a given breast cancer gene, for example, does not exclude the development of the more common non-familial type of breast cancer.
The recent developments in cancer genetics show how the concepts of cancer genes and their mutations explain the phenomenon of malignant growth. Genetics of families and populations are now being supplemented by the study of the genetics of cell populations.
Genetics as a basic science of medicine and public health
Medicine and public health are more than scientific fields. They are professions that deal with the causes and management of disease, but need science for optimal practice. In the early nineteenth century, pathology was the leading science of medicine. In the late nineteenth century, pathology was supplemented by bacteriology, which offered for the first time a rational aetiological concept of disease. A specific single cause explained microbial infections. This was great progress, as it opened the way for causal therapy, which finally arrived in the middle of the twentieth century with the development of chemotherapy and antibiotics.
During the last two decades human genetics has assumed the role of major paradigm in medicine. First, many monogenic diseases could be explained by mutations of enzymes or various proteins. This advance had immediate therapeutic consequences, allowing successful treatment by removing a noxious metabolite or by substituting a protein that was lacking. An example of the first strategy is the phenylalanine-restricted diet in phenylketonuria; the second strategy is epitomized by factor VIII substitution in haemophilia A. The development of cytogenetics identified the fundamental cause of many birth defects as chromosomal defects, but the mechanisms by which chromosomal aberrations determine complex phenotypes remain a challenge for research in developmental genetics. In recent years, the genetic material and the genes themselves became accessible to analysis, allowing direct elucidation of defects in genetic diseases, including cancers. These developments in molecular and cellular biology have brought genetic concepts and methods into biomedical research in general. The genetic paradigm has become the major conceptual framework within which biomedical research is currently being performed. The term ‘molecular medicine’ is increasingly applied to such work.
Genetic variability in the ‘normal’ range
Genetic polymorphisms and disease
A polymorphism is a monogenic trait that exists in the population in at least two phenotypes (and presumably at least two genotypes), neither of which is rare, that is, neither of which occurs with a frequency of less than 1 to 2 per cent. Often we find more than two alleles and more than two phenotypes for a single locus. The first human polymorphism was the ABO blood group discovered in 1900. ABO, Rh, and other blood groups are genetic differences of surface antigens of red blood cells. Polymorphisms of serum proteins and various enzymes are also known. A very important group of polymorphisms comprises the surface antigens of cells (such as lymphocytes) involved in the immune response (major histocompatibility complex); these highly polymorphic HLA types are particularly interesting because they are largely responsible for the rejection of organ and skin transplants if the HLA types of donor and recipient are not carefully matched (Tiwari et al. 1987). Organization of matching of immunologically compatible organs such as kidneys, hearts, or livers has become a major international endeavour; special organizations have been founded, for example Eurotransplant in Leiden (The Netherlands).
In addition to polymorphisms that are detected by phenotypic variation, there are many heritable differences in the base sequence of the DNA that do not influence the phenotype and can be detected only by direct studies of DNA. Most such DNA polymorphisms are found outside coding genes. They may consist simply of an exchange of single DNA base or involve variation in numbers of repeated dinucleotides or trinucleotides. Methods for identification include the use of one or other of many different restriction endonucleases that cut DNA sequences specifically recognized by each one of these enzymes. The most popular method at present is the use of the polymerase chain reaction to amplify the dinucleotide or trinucleotide variants.
DNA polymorphisms are utilized for many practical problems, such as identification of individuals for forensic purposes in criminal investigations or for identifying disputed paternity. In medical genetics, their main use is for linkage investigations in families. Association studies of polymorphic DNA markers with diseases are often carried out by comparing affected patients with controls. If a gene responsible for a disease cosegregates with the marker gene among patients, linkage may be present, but the interpretation of such findings is often difficult because of genetic heterogeneity of the disease and other complexities such as ethnic differences between patients and controls. Currently, major interest centres on the discovery of single nucleotide polymorphisms (SNPs) that occur with a frequency of 1 in 1000 nucleotide base pairs. Complete maps of the thousands of SNPs distributed over the genome should be helpful for both association and linkage studies.
Polymorphisms at the phenotypic level may have a direct influence on susceptibility to ‘complex’ diseases; they may contribute to the multifactorial complex of genes involved in causation. The best-known examples are the ABO blood groups and the HLA system. Many diseases are slightly more common in carriers of certain ABO blood types than in others (Vogel and Motulsky 1996). Group A is found more frequently in cancers of the stomach, salivary glands, mouth and pharynx, and ovary, as well as in thrombotic diseases. Group O is more common in peptic ulcers. Moreover, there is evidence that some infectious diseases had ABO blood group associations at a time when treatments were unavailable. The total contribution of such genes to the disease aetiology is small. Associations with certain HLA types are observed mainly for diseases involving autoimmune processes. The strongest association has been found between ankylosing spondylitis and HLA B27. The risk for this disease among carriers of this HLA type is almost 90 times as high as that of other HLA types. Another disorder with a very strong, but unexplained, HLA association is narcolepsy (for a full tabulation of studies, see Tiwari and Terasaki (1985)). An increasing number of genetic polymorphisms—especially those of genes involved directly or indirectly in the various defence mechanisms against infective agents—have been shown to lead to differences in the response to infective agents, for example, course and outcome of infectious diseases (Hill 1996; Vogel and Motulsky 1996). This aspect of medical genetics appears to be relatively neglected—probably because, at present, infectious diseases are playing a minor part in morbidity and mortality in Western countries, where most medical geneticists are working. But in developing countries, infectious diseases are still very important. Therefore, a shifting of emphasis to this field among scientists of these countries would be worthwhile (Vogel 1998). Owing to increasing resistance of germs to antibiotics, infections are becoming increasingly dangerous in the Western world as well.
Genetic variability in reaction to drugs: pharmacogenetics
The pioneers of human genetics during the first decades of the twentieth century, such as Garrod and Haldane, hinted that inherited biochemical variation might explain unusual reactions to drugs and foods. In the 1950s, a few abnormal untoward reactions to drugs were shown to be caused by a genetically determined variation of enzymes (Motulsky 1957). Mutations of the enzyme glucose-6-phosphate dehydrogenase explained haemolytic reactions caused by ingestion of fava beans and by a variety of drugs, including the antimalarial agent primaquine. Variation in the enzyme pseudocholinesterase was found to cause prolonged apnoea on administration of suxamethonium, a drug widely used to relax muscles during surgery. Genetic differences in acetyltransferase activity explained marked individual differences in the blood level of isoniazid, a drug often used in tuberculosis therapy. Genetic variation in a component of the P-450 system of the liver (which is involved in the metabolism of foreign substances) causes defective oxidation of a wide variety of drugs, with certain adverse reactions. A variety of other monogenic pharmacogenetic traits have been described (Evans 1993; Vogel and Motulsky 1996).
In addition to these monogenic traits, a series of twin studies has shown that genetic factors are involved in the metabolism of many other drugs, as measured by plasma concentration, half-life, and other parameters. Among them are such frequently used drugs as antipyrine, dicumarol, aspirin, halothane, and others. Hence genetic variation in drug metabolism is a widespread and regularly observed phenomenon. If untoward drug reactions occur, genetic variation should be considered as one of the possible explanations.
Genetic variation in the reaction to food and other environmental factors: ecogenetics
One of the drugs for which genetic variation in metabolism has been demonstrated by repeated twin studies is ethanol. Ethanol is usually ingested for pleasure as an alcoholic drink. There are wide individual differences not only in alcohol metabolism, but also in its effects on the brain (see Chapter 10.2 and Chapter 10.3).
Lactose digestion polymorphism
Another ecogenetic polymorphism affects lactose digestion (Flatz 1992). The disaccharide lactose occurs widely in nature, but large amounts are found only in mammalian milk. To be absorbed, lactose must be hydrolysed to glucose and galactose. This is achieved by lactase, an enzyme located at the surface of intestinal cells. In all lactose-producing mammals, intestinal lactase activity is high during the suckling period, declines after weaning, and remains low in adolescent and adult animals. The human species was long considered an exception to this rule as high lactase activity appeared to be expressed throughout the lifetime. However, early studies had been performed among individuals of European origin who do maintain such intestinal lactase activity during adult life. With widespread population screening, it was found that most non-European adult humans had low lactase activity similar to other adult mammals.
Family studies have shown persistence of lactase production in adults to be genetically determined as a Mendelian dominant trait. This means that homozygotes or heterozygotes for the lactase persistence allele (LAC*P) digest lactose as adults. The genetic mechanism of persistence of lactose expression in the intestine is unknown. A regulatory gene is probably responsible. In most populations, both alleles (LAC*P and LAC*R for lactose restriction) are present. The frequency of poor lactose digesters (LAC*R/LAC*R) ranges between 1 and 96 per cent. Populations in subtropical Africa, Eastern Asia, Australia, and native Americans have frequencies of poor lactose digestion of between 90 and 100 per cent. A high rate of persistent lactose absorption is found only in populations who depend on milk from their animals—desert people in Arabia and northern Africa—as well as western and central Europeans. Most southern and eastern Europeans exhibit an intermediate distribution with frequencies of ‘malabsorbers’ ranging from about 30 to 90 per cent. The precise reasons for these enormous differences are largely unknown. Most observers agree that natural selection must have played a part. Arabian and African populations largely depend on milk for their protein supply, so that those who inherited a gene allowing absorption of lactose after weaning may have had a higher chance of survival. This explanation is less likely for Europe, as survival in northern Europe never appeared to depend critically on the milk supply. Protection of the gene for persistent lactose absorption (Lac*P) against vitamin D deficient rickets, which was common in central and northern Europe, has been suggested.
Lactose-containing foods lead to increased peristalsis, colonic irritation, and diarrhoea in subjects with lactose malabsorption. However, clinically significant signs are rare, as affected individuals usually reduce their milk consumption. Some well-intended support programmes for children, for example in Africa, have been disappointing because children fed too much milk developed diarrhoea. This polymorphism is a good example of how relative the concepts ‘normal’ and ‘abnormal’ are. Lactose absorption in adults, which was first regarded as normal, turned out to be the exception when the trait was studied globally. Lactose malabsorption appeared abnormal at first, but was later shown to be the rule among most populations of the world.
Further discussion can be found in various UNSCEAR reports, Sankaranarayanan (1988), Neel and Schull (1991), and Vogel (1992).
The fact that energy-rich radiation can influence mutations was first established in 1927 in Drosophila. ‘Classical’ radiation genetics developed from these results. When DNA was identified as the genetic material, it was soon demonstrated that the genetic material of all living beings is susceptible to radiation-induced damage. Extensive studies on the mouse have elucidated the principles of radiation mutagenesis for the mammalian genome. The mouse data, together with results from direct observations on spontaneous and induced mutations in humans, have been used by international committees to estimate the potential genetic effects of radiation in human populations and to predict genetic damage in relation to a radiation dose. Often, the so-called ‘doubling dose’ is estimated, that is, the radiation dose that doubles the spontaneous mutation rate, assuming that radiation has occurred more than about 8 to 12 weeks before fertilization. Table 6 and Table 7 show recent estimates provided by two international committees (UNSCEAR and the Committee on the Biological Effects of Ionizing Radiation (BEIR)). The estimates agree fairly well regarding the induction of chromosomal aberrations and monogenic diseases. The estimates on malformations and complex diseases remain vague at best (for the available data and a detailed discussion, see Czeizel and Sankaranarayanan (1984), Czeizel et al. (1988, 1990), and Vogel (1992)).
Table 6 Radiation risk estimates for genetic diseases: estimated increase per 100 rem (1 Sv) low-dose rate radiation (UNSCEAR estimates)
Table 7 Radiation risk estimates for genetic diseases: estimated increase per 1 rem (0.01 Sv) low-dose rate radiation (BEIR estimates)
Most estimates cited in Table 6 and Table 7 are based on studies in mice. Directly observed information from humans is rare. A large body of data from human beings became available from follow-up studies of the survivors of the atomic bombings in Hiroshima and Nagasaki in August 1945. Joint American and Japanese research teams are continuing to study survivors, as well as their offspring born after the bombing. Direct teratogenic effects of microcephaly and mental retardation were observed in fetuses irradiated during the 18th to 25th weeks of fetal life. Search for mutations in survivors’ germ cells initially utilized offspring parameters such as stillbirths, major malformations, and death during early life. Later, additional end-points were introduced, such as chromosomal studies and inherited protein variation (Neel and Schull 1991). The result can be summarized in one sentence. Despite the fact that about 70 000 children were examined and that all possible statistical biases were considered with painstaking precision, no definite genetic effects among the offspring could be proved.
Assuming that such effects must have occurred, as in all other species, a genetic doubling dose was estimated from the small but statistically insignificant differences in ‘untoward pregnancy outcomes’ and early mortality of children of irradiated survivors compared with controls. These doubling doses were higher (about 200 rem for acute radiation and about 400 rem for chronic radiation) than those estimated from earlier mouse studies. More recently, a reassessment of the mouse data suggests no difference between the species. No such estimates were possible for the most well-defined human data, chromosomal aberrations, and protein variants, as children of the irradiated group had even fewer untoward results than the non-irradiated control group. In any case, it can be reasonably concluded that ionizing radiation in doses that modern populations (including occupational groups exposed to low-level radiation) might receive have few untoward effects on the health of future generations. Similar conclusions were reached in less rigorous studies of offspring of populations in India and China who had lived for generations on ground containing radio-active isotopes (Vogel 1992). More recent studies seem to show that ionizing radiation causes mainly larger deletions outside of transcribed genes; therefore, its effects are mostly not visible in the phenotype of offspring (Sankaranarayanan 1999; Sankaranarayanan and Chakraborty 2000a,b,c). But studies on offspring of Japanese atomic bomb survivors largely failed to show any increase of mutations that change DNA markers (Neel 1995). Exposure to very high radiation doses will either kill the individual or lead to sterility by killing germinal stem cells and, therefore, does not damage future generations. These findings do not mean that radiation protection should be neglected, as there will always be an additional finite risk of mutations for offspring.
Most importantly, definitive risks to the irradiated atom bomb survivors themselves were detected. Clearly, increased risk for leukaemias was demonstrated a few years after atomic bombing and for solid malignant tumours 15 to many years later. The cancer risk was highest with radiation exposure at young ages. These and other data permit some recommendations to be made for radiation protection.
Radiation should be kept to a minimum.
If radiation therapy is necessary and protection of gonads is technically impossible, a time period of at least 8 weeks—even better, 3 months—should elapse between the end of exposure to ionizing irradiation and fertilization. This avoids fertilization of germ cells that have been irradiated in a postmeiotic state of development, when they are particularly susceptible to genetically relevant radiation damage.
In women, the days immediately around fertilization are particularly dangerous, and so any radiation exposure should be avoided at that time.
Radiation in general should be avoided during pregnancy. If radiation exposure with low dosage (such as after diagnostic radiography or isotope diagnostic procedures) has occurred in early pregnancy, the fetal risks are extremely small.
A second possible source of genetic damage to future generations are chemical mutagens. These are strongly reactive substances that are able to react with DNA to cause genetic alterations. Many agents are used in medical therapy as cytostatic agents. As their cytostatic action is based on interaction with DNA, potential mutagenic effects can hardly be avoided without compromising the desired therapeutic effects. However, such drugs are mostly used on patients who either have already reached their postreproductive age or will not have any more children because of poor health (Vogel and Jäger 1969). There is much concern about so-called environmental mutagens, that is, substances present in small doses due to many different chemical sources in environmental pollution. Numerous naturally occurring chemicals have been shown to be mutagenic in bacterial test systems used as surrogate models for chemical mutagenesis (Ames et al. 1990). Mutagenous assay results are important for risk assessment (see Chapter 8.8 and Chapter 8.9). Nature appears to have endowed our species with mechanisms that protect against mutagenic influences. We agree with Ames et al. (1990) that the potential of chemical mutagenesis has been exaggerated, but few direct data exist and unpleasant surprises remain possible.
Applications of genetic knowledge in medical practice
Further discussion can be found in Fuhrmann and Vogel (1983), Harper (1993), and Vogel and Motulsky (1996).
Assessment of genetic risks and genetic counselling
The potentially most important application of genetic knowledge for public health is genetic counselling. This activity consists of an accurate diagnosis, an assessment of genetic risks by appropriate information regarding these risks, and a discussion of the various reproductive options. Diagnosis may be technically complex and may require various procedures, particularly a variety of biochemical and molecular tests, including study of fetal cells following prenatal diagnosis. Making diagnosis may be difficult, but various reference books (McKusick 1994) and websites are now available (http://www3.ncbi.nlm.nih.gov/omim) (Fischer et al. 1996). Counselling has to start with the construction of a family chart or pedigree. Even clinically similar or identical diseases may have different modes of inheritance, with different consequences for the risk. Risk assessment may be simple, but sophisticated statistical techniques are sometimes required to integrate the various data. Genetic counselling usually requires expert knowledge. Apart from the scientific and technical aspects necessary for optimal genetic counselling, the genetic counsellor must be empathic and sensitive to the many psychological aspects raised by the genetic problems facing a family or a patient.
Genetic counselling is indicated for increased risks of occurrence of a disease or birth defect in a family. This may be evident if the counsellee or a close relative suffers from a disease for which a genetic cause is known or can be assumed, if the counsellee is at risk of being a healthy carrier of a disease gene, or if the two partners are close relatives such as first cousins. Often, a couple have already had a child with a birth defect or genetic disease and are concerned regarding the recurrence risk for the next child. An increased risk may also exist if the affected patient is the only diseased person in an otherwise healthy family (for example, in autosomal recessive diseases, dominant new mutation, trisomy 21, and various multifactorial diseases).
Chromosomal study is a common diagnostic test. The main indications are as follows.
Suspected Down syndrome: even if the clinical picture is definite, chromosome studies are indicated as recurrence risk is higher for translocation Down syndrome than for free trisomy 21.
Disturbances of sex development, for example, Klinefelter’s (XXY) and Turner’s (X0) syndromes.
Combinations of various anomalies, such as small size for date, retarded mental and motor development, multiple birth defects, and dysmorphic face and/or other body parts.
Suspected X-linked mental retardation.
Habitual abortions, to rule out translocations (after other causes have been excluded).
In order to make meaningful plans for appropriate cytogenetic study, the chromosome laboratory needs precise data on the clinical and genetic aspect of the case under study.
Further details are given by Harper (1993) and Becker et al. (1995).
Prenatal diagnosis may involve non-invasive as well as invasive methods. The most frequently used non-invasive technique is ultrasonic examination of the fetus. In some countries, particularly in Europe, ultrasonic examination has become a routine procedure that is performed in each pregnancy. Precise assessment of the gestational age and fetal position is possible in early pregnancy. Later studies permit the detection of abnormal fetal growth as well as of a growing number of birth defects. Interpretation of abnormal ultrasound patterns requires considerable experience. Ideally, a suspicious finding should lead to referral to specialist ultrasonographers with experience in fetal pathology in centres with high-quality ultrasound equipment.
A frequently used non-invasive prenatal test is a-fetoprotein screening in the blood of pregnant women. Based on the levels of a-fetoprotein, often supplemented with biochemical tests such as chorionic gonadotropin and unesterified oestriol (triple-marker screening), it is possible to recognize an increased risk for several malformations, including neural tube defects (increased a-fetoprotein level), Down syndrome (decreased a-fetoprotein level), and others. If such abnormal biochemical patterns are found, careful ultrasound study and amniocentesis to verify a diagnosis of neural tube defect by study of a-fetoprotein and other biochemical parameters in the amniotic fluid needs to be performed. Some authors are recommending maternal a-fetoprotein blood screening for all pregnant women as the method is harmless and can help in the recognition of several fetal abnormalities. The large number of false-positive as well as false-negative tests with this approach justifies some scepticism (Andrews et al. 1994), and there is no general agreement regarding the institution of such programmes in all countries. In the United States, over 2 million pregnant women (about half of all pregnancies) are said to undergo maternal a-fetoprotein testing every year.
The most common invasive procedures are amniocentesis and chorionic villus sampling. Amniocentesis is carried out at 14 to 16 weeks of gestation and samples cells of fetal origin in the amniotic fluid. Chorionic villus biopsy is performed between 9 and 12 weeks of gestation. The risk of fetal loss following amniocentesis ranges between 0.5 and 1 per cent and is somewhat higher for chorionic villus biopsy, which has the marked advantage of being carried out early in pregnancy. The most common indications for amniocentesis or chorionic villus biopsy are maternal age above 35 (see above), Down syndrome or any other chromosomal aberration syndrome in a previous child, balanced translocation in one parent, neural tube defect in a previous child and/or in one of the parents, and monogenic hereditary disease where prenatal diagnosis (DNA or biochemical) is possible.
Organization of genetic counselling
The organization of genetic counselling and prenatal diagnosis differs in various countries of the world. In The Netherlands, for example, both activities are concentrated in a few centres in which all facilities are available. Health insurance does not pay for activities outside these centres. In Germany, genetic counselling and prenatal units exist in all university institutes of human genetics; there are about 30 such units for a population of more than 80 million. In addition, qualified doctors are permitted to perform genetic counselling and/or certain routine genetic diagnostics tests, which therefore tend to be shifted from university institutes to private practice.
In the United States, genetic counselling and prenatal centres exist in most medical schools as well as in many larger hospitals. Counselling prior to routine prenatal diagnosis (that is, as carried out for maternal age indications) is usually performed by obstetricians or their nurse assistants prior to the procedure. However, such personnel are only recently beginning to be well informed about medical genetics. Counselling for other than routine indications is usually performed by qualified medical geneticists (MD level) and increasingly by genetic counsellors, who have been trained in human and medical genetics and its practical applications for 2 years following a 4-year college education. These genetic counsellors often work in centres with a team that includes medical geneticists (MD level), biochemical and molecular geneticists (PhD level), and cytogeneticists (PhD level), but a growing proportion is attached to health maintenance organizations and to groups of obstetricians and/or paediatricians. These new professionals have filled an important gap in the provision of genetic services, particularly to population groups which, because of geographical location and other reasons, have not had access to expert genetic advice.
It has been recommended that one genetic and prenatal diagnosis unit should serve a population of a million. Location in or affiliation with a university hospital is a frequent and useful arrangement. A variety of personnel are required. Several medical geneticists, ideally with different special knowledge in various subareas of medical genetics and various clinical specialties, are desirable. Non-doctoral genetic counsellors or specially trained nurses are needed to deal with the many problems of patient communication and follow up. Expert PhD scientists, together with a certain number of technicians, are required to carry out the many different types of laboratory studies.
As in other fields of medicine, quality control is essential. This is relatively easy for laboratory work such as for chromosomal or DNA diagnosis. Here, it has become customary in many countries that licensing agencies or scientific or professional societies distribute anonymous specimens to the laboratories and compare results (Andrews et al. 1994).
Quality control for genetic counselling is more difficult but requires that personnel engaged in this practice must have undergone supervised training and, ideally, certification by examination. This is available in the United States where a specialty board of medical genetics exists for clinical genetics as well as for several aspects of laboratory genetics (cytogenetics, biochemical genetics, and molecular genetics). A different board for non-doctoral genetic counsellors has also been established. Primary care practitioners, obstetricians, paediatricians, internists, and oncologists will require more training in genetics and genetic counselling in medical schools and during postgraduate training in order to advise patients about the meaning and interpretation of the many new genetic tests that are becoming available.
Public health or community genetics
Aspects of human and medical genetics described so far are mainly the concern of the medical profession, including biologists, non-medical biochemists, and other helpers in research and medical practice. In recent years, however, emphasis has shifted toward studies on clinically healthy persons, individual differences in disease susceptibilities, and appropriate diagnostic and preventive measures in the ‘normal’ population (Modell and Kuliev 1989). Conferences are being held that centre around these problems, and a new journal Community Genetics has been founded. In an inaugural editorial of this journal L.P. ten Kate (1998) described these problems as follows:
Community genetics … encompasses all activities to enable the identification of people … with increased genetic risk who want to acquire this knowledge in order to make informed decisions. [It] minimizes the number of people who would like to know that they are at increased risk, but do not know yet. However, we should not bargain on ethical principles of autonomy, doing good and not harm, justice, and providing equal access and solidarity.
Disease susceptibilities and individual risks have also been considered increasingly in other areas of medical genetics; it is the major difference that phenotypically healthy population groups are now being approached actively, and are offered genetic services. One major offer of this kind is genetic screening.
P>Genetic screening involves the study of all individuals of a population or population group for the presence of a certain genetic variant, disease, or carrier state (Vogel and Motulsky 1996). Such screening is generally recommended if effective treatment or preventive measures are possible. Neonatal screening of newborns for phenylketonuria is the best-known example. This condition is one of the most common inborn errors of metabolism (about 1 in 12 000 births in populations of European origin) and is carried out in most European countries and in the United States. The enzyme defect leads to a build-up of phenylalanine in the body, including the brain, which results in profound mental retardation. Diagnosis involves measurement of phenylalanine level in a small drop of blood on filter paper. Restriction of phenylalanine from the diet in affected infants permits normal development. Another treatable condition that is screened neonatally is hypothyroidism, which is frequently not genetic in origin. Such programmes have been successful in that initial screening is followed by specific tests that are sensitive and specific. The required treatments are highly successful and prevent severe mental retardation. Even though these diseases are relatively rare, most European countries and the United States have introduced neonatal screening for these conditions.
Many other diseases have been suggested for testing newborn infants. These include conditions such as sickle cell anaemia, Duchenne muscular dystrophy, and cystic fibrosis. Sickle cell anaemia testing of newborns has been widely performed in the United States as early detection allows antibiotic therapy to prevent infant and childhood mortality. As racial identification of specimens is difficult, all newborns are tested for sickle cell anaemia.
No effective curative or preventive treatment is available for Duchenne muscular dystrophy, but newborn testing has been occasionally recommended to identify carrier mothers for genetic counselling. There is no general agreement regarding this recommendation. Some authoritative groups feel that identification of genetic disease in children for purposes of genetic counselling in the parent should never be carried out. More studies with emphasis on the psychosocial aspects are required.
Cystic fibrosis testing of newborns that might allow earlier treatment has not been definitely shown to affect the natural history of the disease and, therefore, is not generally recommended. Various other inborn errors in metabolism such as maple syrup urine disease, galactosaemia, and homocystinuria are being screened in some areas, but these conditions are very rare and the required metabolic treatment is not successful in all cases. Rigorous pilot studies that are clearly labelled as such should be required before newborn testing for a given condition is generally recommended as a service procedure. Even though such pilot studies need to involve a large number of newborns, a clear distinction must be maintained between such quasi-experimental programmes compared with generally recommended screening studies for all newborns.
Another type of screening deals with heterozygote detection of relatively common recessive diseases to identify matings at risk of producing affected offspring. Among Ashkenazi Jews, 3 to 5 per cent of the population are heterozygotes for the autosomal recessive Tay–Sachs disease. Screening for the carrier state of the gene during pregnancy, or ideally even earlier, offers the possibility of avoidance of the disease, such as by prenatal diagnosis of couples where both partners are carriers. Tay–Sachs disease has become extremely rare among Jews in the United States and Israel by using this approach followed by pregnancy termination. Other recessive diseases common in this population such as Canavan’s disease are now being added. The thalassaemias are very common in many populations of tropical and subtropical areas, for example, in the islands of Sardinia and Cyprus. A screening programme for heterozygous carriers of b-thalassaemia with subsequent prenatal diagnosis in couples at risk has led to a reduction of affected homozygotes by about 70 to 80 per cent. Attempts at screening for the sickle cell gene among the African-American population of the United States have been less successful. Insufficient information and failure to discriminate between the common sickle trait and the rare sickle cell anaemia have led to serious misunderstandings and even to discrimination of carriers on the job market. ‘Population screening for haemoglobin disorders thalassaemias and sickle cell (thalassaemias and sickle cell disorders) has been practised on a large scale for over 20 years, and basic concepts and methods of community genetic have developed within this framework’ (Modell and Kuliev 1998).
Screening for the cystic fibrosis gene, particularly in northern and central European populations, is sometimes recommended as cystic fibrosis is common in these populations (1 in 2000 births). The current life expectancy for a child with cystic fibrosis is about 30 or, with optimal medical care, about 40 years; the disease is not as devastating as phenylketonuria, Tay–Sachs disease, or b-thalassaemia. As a practically feasible genotyping programme does not detect all potential mutations among persons of central and northern European origin, an affected child may be born despite the fact that one (or even both) parents tested negative. Different sets of testing panels for cystic fibrosis mutations need to be used in other populations as the frequency of mutations differs in various ethnic groups. These reasons and the considerable expense of running a programme that needs to be carried out on the entire Caucasian population have led to restraint in instituting population screening for cystic fibrosis.
To arrive at a clearer picture of a complex situation, various approaches, together with their advantages and disadvantages, have been discussed (Decruyenaere et al. 1998; Schmidtke 1998). A 1997 National Institutes of Health consensus group recommended testing of pregnant women under certain, restrictive conditions, but did not recommend a general population screening for cystic fibrosis. In the opinion of Schmidtke (1998), there remain ‘many unresolved, and perhaps unresolvable, psychological and ethical problems.’ Therefore, in his opinion, ‘a CF [cystic fibrosis] carrier screening program is premature at the best’.
Much more extended genetic screening can be envisaged in some future scenarios. It appears likely that we shall have the power to predict a wide variety of specific risks and health hazards based on genetic testing. A committee of the Institute of Medicine of the National Academy of Sciences of the USA strongly recommended against combining many different tests at one time, such as at birth (Andrews et al. 1994). Tests for untreatable diseases should not be combined with tests for preventable or treatable conditions. Children should not be screened for any disorders unless treatment or prevention during childhood is available.
Somatic gene therapy (treatment of certain genetic diseases and types of cancer by manipulation of certain genes in body cells outside the germ line) falls into the wide field of medical therapy; it has little impact on public health policy (except, of course, the large amount of money that would be necessary). Recently, however, some scientists (especially molecular biologists) have discussed gene therapy in germ cells not only for the treatment of diseases, but for improving certain abilities of ‘normal’ individuals. Most medical geneticists are much more conservative. For the prevention of diseases, such manipulations are not necessary; selection of appropriate zygotes after pre-implantation diagnosis would achieve the same goal. However, in some countries, manipulation of early zygotes by such a procedure is prohibited by law. Attempts at improvement of human individuals by manipulating normal germ cells meets with many technical and ethical problems that require much more extensive discussion. The purpose of public health genetics is not genetic improvement of human beings—it is not eugenics. Rather, human suffering from illness and disease should be diminished by intelligent and careful utilization of genetic knowledge.
Ames, B.N., Profet, M., and Gold, L.S. (1990). Dietary pesticides (99.99 per cent of all natural) and nature’s chemicals and synthetic chemicals: comparative toxicology. Proceedings of the National Academy of Sciences of the USA, 87, 7777–86.
Andrews, L.B., Fullarton, J.E., and Holtzman, N.A. (1994). Assessing genetic risks. National Academy of Sciences Press, Washington, DC.
Becker, R., Fuhrmann, W., Holzgreve, W., and Sperling, K. (1995). Pranatale Diagnostik und Therapie. Wiss Verlagsgesellschaft, Stuttgart.
BEIR (Committee on the Biological Effects of Ionizing Radiation). (1990). Health effects of exposure to low levels of ionizing radiation (BEIR V). National Academy of Sciences Press, Washington, DC.
Cooper, D.N. and Krawczak, M. (1993). Human gene mutation. BIOS Scientific, Oxford.
Czeizel, A. and Sankaranarayanan, K. (1984). The load of genetic and partially genetic disorders in man. I. Congenital anomalies: estimates of detriment in terms of years of life lost and years of impaired life. Mutation Research, 128, 73–103.
Czeizel, A., Sankaranarayanan, K., Losence, A., et al. (1988). The load of genetic and partially genetic diseases in man. II. Some selected common multifactorial diseases: estimates of population prevalence and of detriments in terms of lost and impaired life. Mutation Research, 196, 254–92.
Czeizel, A., Sankaranarayanan, K., and Szondi, M. (1990). The load of genetic and partially genetic diseases in man. III. Mental retardation. Mutation Research, 232, 291–303.
Davies, J.L., Kawaguchi, Y., Bennett, S.T., et al. (1994). A genome-wide search for human type 1 diabetes susceptibility genes. Nature, 371, 130–6.
Decruyenaere, M., Evers-Kieboms, G., Denayer, L., and Welkenhuysen, M. (1998). Uptake and impact of carrier testing for cystic fibrosis. Community Genetics, 1, 23–35.
Edwards, J.H., Pawlowitzki, I.H., and Thompson, E. (1997). Genetic mapping of disease genes. Academic Press, London.
Evans, D.A.P. (1993). Genetic factors in drug therapy. Clinical and molecular pharmacogenetics. Cambridge University Press.
Fischer, C., Schweigert, S., Spreckelsen, C., and Vogel, F. (1996). Programs, databases and expert systems for human geneticists—a survey. Human Genetics, 97, 129–37.
Flatz, G. (1992). Lactase deficiency: biological and medical aspects of the adult human lactase polymorphism. In The genetic basis of common disease (ed. R.A. King, J.I. Rotter, and A.G. Motulsky), pp. 305–25. Oxford University Press.
Fuhrmann, W. and Vogel, F. (1983). Genetic counseling. Springer-Verlag, New York.
Galjaard, H. (1994). Genetic technology in health care: a global view. International Journal of Technology Assessment in Health Care, 10, 527–45.
Gelehrter, T.D. and Collins, E.S. (1990). Principles of medical genetics. Williams and Wilkins, Philadelphia, PA.
Goldstein, J.L., Hobbs, H.H., and Brown, M.S. (1995). Familial hypercholesterolemia. In The metabolic and molecular bases of inherited diseases (ed. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle), pp. 1981–2030. McGraw-Hill, New York.
Gottesman, II. (1991). Schizophrenia genes. The origin of madness. W.H. Freeman, New York.
Häfner, H. (1999). Ideengeschichte der Gesundheitspflege. In Gesundheit, unser hochstes Gut? (ed. H. Häfner), pp. 5–28. Springer-Verlag, Berlin.
Hall, J.G. (1990). Genomic imprinting: review and relevance to human diseases. American Journal of Human Genetics, 46, 857–73.
Harada, S., Agarwal, D.P., Goedde, H.W., et al. (1982). Possible protective role against alcoholism for aldehyde dehydrogenase isozyme deficiency in Japan. Lancet, ii, 827.
Harper, P.S. (1993). Practical genetic counselling. Wright, Bristol.
Hill, A.V.S. (1996). Genetics in infectious disease resistance. Current Opinions in Genetics and Development, 6, 348–53.
Institute of Medicine (1992). Advances in understanding genetic charges in cancer: impact on diagnosis and treatment decisions in the 1990s. National Academy of Sciences, Washington, DC.
King, R.A., Rotter, J.I., and Motulsky, A.G. (1992). The genetic basis of common diseases. Oxford University Press, New York.
Kinzler, K.W. and Vogelstein, B. (1995). Colorectal tumors. In The metabolic and molecular bases of inherited diseases (ed. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle), pp. 643–63. McGraw-Hill, New York.
Knudson, A.G., Jr (1971). Mutation and cancer: statistical study of retinoblastoma. Proceedings of the National Academy of Sciences, 68, 820–4.
Levy, H.L. (1973). Genetic screening. Advances in Human Genetics, 4, 1–104.
McKusick, V.A. (1986). The morbid anatomy of the human genome: a review of gene mapping in clinical medicine, Part 1. Medicine, 65, 1–33.
McKusick, V.A. (1987). The morbid anatomy of the human genome: a review of gene mapping in clinical medicine, Parts 2 and 3. Medicine, 66, 1–63, 237–96.
McKusick, V.A. (1988). The morbid anatomy of the human genome: a review of gene mapping in clinical medicine, Part 4. Medicine, 67, 1–19.
McKusick, V.A. (1994). Mendelian inheritance in man. Johns Hopkins University Press, Baltimore, MD.
Modell, B. and Kuliev, A. (1989). Impact of public health on human genetics. Clinical Genetics, 36, 206–98.
Modell, B. and Kuliev, A. (1990). Changing paternal age distribution and the human mutation rate in Europe. Human Genetics, 86, 198–202.
Modell, B. and Kuliev, A. (1998). The history of community genetics: the contribution of the haemoglobin disorders. Community Genetics, 1, 3–11.
Motulsky, A.G. (1957). Drug reactions, enzymes and biochemical genetics. Journal of the American Medical Association, 165, 835–7.
Motulsky, A.G. (1995). Jewish diseases and origins. Nature Genetics, 9, 99–101.
Neel, J.V. (1962). Diabetes mellitus: a ‘thrifty’ genotype rendered detrimental by ‘progress’? American Journal of Human Genetics, 14, 353–62.
Neel, J.V. (1995). New approaches to evaluating the genetic effects of the atomic bombs. American Journal of Human Genetics, 57, 1263–6.
Neel, J.V. and Schull, W.J. (1991). The children of atomic bomb survivors. National Academy Press, Washington, DC.
Omenn, G.S. (1988). Genetic investigations of alcohol metabolism and of alcoholism. American Journal of Human Genetics, 43, 579–81.
Omenn, G.S. and Motulsky, A.G. (1972). A biochemical and genetic approach to alcoholism. Annals of the New York Academy of Sciences, 197, 16–23.
Ott, J. (1991). Analysis of human genetic linkage (revised edn). Johns Hopkins University Press, Baltimore, MD.
Pauling, L., Itano, H.A., Singer, S.J., and Wells, I.C. (1949). Sickle cell anemia: a molecular disease. Science, 110, 543–8.
Propping, P. (1977). Genetic control of ethanol action on the central nervous system. Human Genetics, 35, 309–34.
Propping, P. (1989). Psychiatrische Genetik. Springer-Verlag, Berlin.
Propping, P. (1992). Alcoholism. In The genetic basis of common disease (ed. R.A. King, J.I. Rotter, and A.G. Motulsky), pp. 837–65. Oxford University Press.
Sankaranarayanan, K. (1988). Prevalence of genetic and partially genetic diseases in man and the estimation of genetic risks of exposure to ionizing radiation. American Journal of Human Genetics, 42, 651–62.
Sankaranarayanan, K. (1998). Ionizing radiation and genetic risks. IX. Estimates of the frequencies of Mendelian diseases and spontaneous mutation rates in human populations: a 1998 perspective. Mutation Research, 411, 129–78.
Sankaranarayanan, K. (1999). Ionizing radiation and genetic risks. X. The potential ‘disease phenotypes’ of radiation-induced genetic damage in humans: perspectives from human molecular biology and radiation genetics. Mutation Research, 429, 45–83.
Sankaranarayanan, K. and Chakraborty, R. (2000a). Ionizing radiation and genetic risk. XI. The doubling dose estimates from the mid-1950s to the present and the conceptual change to the use of the human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations. Mutation Research, 453, 107–27.
Sankaranarayanan, K. and Chakraborty, R. (2000b). Ionizing radiation and genetic risk. XII. The concept of ‘potential recoverability correction factor’ (PRCF) and its use for predicting the risk of radiation-inducible genetic disease in human live births. Mutation Research, 453, 129–81.
Sankaranarayanan, K. and Chakraborty, R. (2000c). Ionizing radiation and genetic risk. XIII. Summary and synthesis of papers VI to XII and estimates of genetic risks in the year 2000. Mutation Research, 453, 183–97.
Sapienza, C. and Hall, J.G. (1995). Genetic imprinting in human diseases. In The metabolic and molecular bases of inherited diseases (ed. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle), pp. 437–58. McGraw-Hill, New York.
Schinzel, A. (1984). Catalogue of unbalanced chromosome aberrations in man. DeGruyter, Berlin.
Schmidtke, J. (1998). A commentary on the NIH consensus development statement ‘Genetic testing for cystic fibrosis.’ Appendix: conclusions and recommendations of the NIH consensus development statement on genetic testing for cystic fibrosis. April 14–16, 1997. Community Genetics, 1, 53–6.
Scriver, C.R. (1992). What are genes like that doing in a place like this? Human history and molecular prosopography. In Genetic diversity among Jews: diseases and markers at the DNA level (ed. B. Bonné-Tamir and A. Adam), pp. 3219–29. Oxford University Press.
Shibuya, A. (1988). Genotypes of alcohol metabolizing enzymes in Japanese with alcoholic liver diseases: a strong association of the usual Caucasian type aldehyde dehydrogenase (ALDH) with the disease. American Journal of Human Genetics, 43, 744–8.
Stevenson, A.C. and Kerr, C.B. (1967). On the distribution of frequencies of mutation in genes determining harmful traits in man. Mutation Research, 4, 339–52.
Sutherland, G.R. and Richards, R.I. (1995). Single tandem repeats and human genetic disease. Proceedings of the National Academy of Sciences of the USA, 92, 3636–41.
Tariverdiau, G. and Vogel, F. (2000). Some problems in the genetics of X-linked mental retardation. Cytogenetics and Cell Genetics, 91, 278–84.
te Meerman, G.J. and Van der Meulen, M.A. (1997). Genomic sharing surrounding alleles identical by descent: effects of genetic drift and population growth. Genetic Epidemiology, 14, 1125–30.
ten Kate, L.P. (1998). Editorial. Community Genetics, 1, 1–2.
Tiwari, J. and Terasaki, P.I. (1985). HLA and disease associations. Springer-Verlag, New York.
Tiwari, J., Terasaki, P.I., and Mickey, M.R. (1987). Factors influencing kidney graft survival in the cyclosporine era: a multivariate analysis. Transplantation Proceedings, 19, 1839–41.
Tjio, H.J. and Levan, A. (1956). The chromosome numbers of man. Hereditas, 42, 1–6.
Tsuang, M.T. and Faraone, S.V. (1990). The genetics of mood disorders. Johns Hopkins University Press, Baltimore, MD.
Tsuang, M.T. and Faraone, S.V. (2000). Genetics of schizophrenia. Seminars in Medical Genetics. American Journal of Medical Genetics, 97, 1–106.
UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). (1986). United Nations, New York.
UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). (1988). United Nations, New York.
UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). (1992). United Nations, New York.
UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). (1993). Sources and effects of ionizing radiation. United Nations, New York.
Van der Meulen, M.A. and te Meerman, G.J. (1997). Association and haplotype sharing due to identity by descent, with an application to genetic mapping. In Genetic mapping of disease genes (ed. J.A. Edwards, I.H. Pawlowitzki, and E. Thompson), pp. 115–35. Academic Press, London.
Van Dongen, J.J.M., Seriu, T., Panzer-Grunmayer, E.R., et al. (1998). Prognostic value of minimal residual disease in acute lymphoblastic leukaemia in childhood. Lancet, 352, 1731–8.
Vogel, F. (1979). Genetics of retinoblastoma. Human Genetics, 52, 1–54.
Vogel, F. (1984). Relevant deviations in heterozygotes of autosomal recessive diseases. Clinical Genetics, 25, 381–415.
Vogel, F. (1992). Risk calculations for hereditary effects of ionizing radiation in humans. Human Genetics, 89, 127–46.
Vogel, F. (1998). Gedanken über die Zukunft der Humangenetik. Medizinische Genetik, 10, 33–9.
Vogel, F. (2000). Genetics and the electroencephalogram. Springer-Verlag, Berlin.
Vogel, F. and Jäger, P. (1969). The genetic load of a human populations due to cytostatic agents. Human Genetics, 7, 287–304.
Vogel, F. and Motulsky, A.G. (1996). Human genetics: problems and approaches. Springer-Verlag, Berlin.
Vogel, F. and Rathenberg, R. (1975). Spontaneous mutation in man. Advances in Human Genetics, 5, 223–318.
Watson, J.D., Hopkins, N.H., Roberts, J.W., et al. (1987). Molecular biology of the gene. Benjamin Cummings, Menlo Park, CA.
Watson, J.D., Gilman, M., Witkowski, J., and Zoller, M. (1992). Recombinant DNA. W.H. Freeman, New York.
WHO (World Health Organization) (1998). World atlas of birth defects. WHO, Geneva.