We Innovate Healthcare
Genes and Health
Genes and Health
Content I 1
2
II 3
4
5
6
7
8
9
10
11
12
13
14...
29 downloads
667 Views
2MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
We Innovate Healthcare
Genes and Health
Genes and Health
Content I 1
2
II 3
4
5
6
7
8
9
10
11
12
13
14
?
III 15
16
17
18
19
20
21
22
23
24
25
26
27
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
7
To d a y To d a y A A
B B
D D C C
To m o r r o w To m o r r o w
Diagnostics Diagnostics
A A
D D
B B C C
G G G G G G
C C C C C C
A A A A A A
A A A A A A
T T T T T T
C C A C C A
T T T T T T
A A A A A A
Pharmacogenomics: genes and drug response
19
Proteomics: seeing through the undergrowth
31
Targets for medicine
47
PCR: an outstanding method
65
SNPs: the great importance of small differences
83
DNA chips: choosy fish hooks
95
Basic conditions: ethics, law and society
115
Prospects: more knowledge for medical science
127
A brief glossary of terms
137
Genes and Health
If it were not for the great variability among individuals, Medicine might be a Science, not an Art. This statement by Sir William Osler (The Principles and Practice of Medicine, 1892) is as topical now as it was over a hundred years ago. For it remains true that drugs sometimes work as intended but sometimes do not, that one patient will tolerate a drug but another will not and that drugs sometimes have serious side effects. These differences are due at least in part to our genes, the genetic material that makes each of us unique and that consequently makes each of us react differently to drugs. Genetics, genomics, proteomics and other branches of modern biology can help us to understand the medical consequences of these differences, and in fact have already led to the identification of many genetic factors that influence the action of drugs – whether this be by affecting the way in which the body deals with a drug or by influencing the course of the disease concerned. And a new scientific discipline – pharmacogenomics – that deals specifically with the relationships between our genome and the effects of drugs has now appeared. At the same time, increasing attention is now being paid to the principal targets of drugs, namely proteins. Here again, a new branch of science has appeared, namely proteomics, the study of the totality of, and the complex interrelationships between, the proteins of an organism. Thus, as well as learning more about the genetic information that provides the blueprint for the production of proteins, we are building up an ever more detailed picture of bodily function and malfunction at the molecular level. Acquisition of an understanding of the interplay between hereditary and nonhereditary factors in patients is an essential step on the way to better targeted, more personalised therapy. An important precondition for this has now been satisfied in that for the first time in the history of medicine, diagnosis and therapy are meeting on common ground. Thanks to the new field of molecular diagnostics, both diagnosis and therapy are now focused on the network of genes, proteins and other substances that exist in the human body. This is leading to the development of completely new ways of understanding, detecting, preventing and specifically combating diseases. Applications of molecular biology are in fact now leading to the development of a new approach to diagnosis and therapy known as molecular medicine.
4
Grouped around this term are a multiplicity of modern research techniques and disciplines. These include, in equal measure, pharmacogenomics, the search for new drug targets, proteome research, the search for small but important genetic differences known as SNPs, new techniques such as the PCR and DNA chips, and bioethics. At many events held over the past few years, Roche has attempted to cast light on current developments in medicine and to explain the scientific background and potential implications of these developments. This publication is intended to supplement that information and to introduce the reader to the most important terms used in the new field of molecular medicine. It can help to improve understanding of current developments and can form a basis for the public debate that is being conducted at present about the uses of genetics and genomics in medicine.
5
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
I 1
2
II 3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
?
III
In its search for the causes of disease, medicine has now advanced to the molecular level. Genetics, genomics and proteomics are opening the way to a new and deeper understanding of bodily processes and are providing the tools for more precisely targeted interventions when bodily function is disturbed. For the first time in history, diagnosis and therapy are meeting on common ground.
27
It is a gradual revolution that has been going on for hundreds of years and still has a long way to go. It is not possible to say exactly when it started. Perhaps it was in the 18th century, at a time when cupping (mechanical leeching) was still considered to be an effective remedy for headache, cancer and cholera. At a time when the value of panaceas was taken as much for granted as the belief in witchcraft, physicians and scientists began to turn medicine upside down. They gave a new meaning to the concept of diagnosis and in so doing Terms took their craft closer to the causes of disease. Even in Molecular medicine the application of molecular biology those days every medical (in particular, genetics, genomics and proteomics) to medicine. Genetics the study of inheritance; deals with the laws of intreatment was preceded by heritance and the properties of genes. an examination; for over Genomics the study of the form, function and interactions of 2000 years an ‘imbalance of the genes of an organism. Proteomics the study of the form, function and interactions bodily humors’ had been the of the proteins of an organism. most common finding, and bloodletting and cupping were accordingly the most common forms of treatment. The diagnosis was generally made on the basis of a handful of symptoms and signs – the art of diagnosis had no more than that to offer. In the middle of the 18th century scientists such as the Italian anatomist Giovanni Battista Morgagni for the first time set themselves the task of identifying the seat of a disease within the body of their patients. They recognised that disturbances of function can be correctly treated only if they are correctly understood – and that an unmeasurable and undefinable imbalance of ‘humors’ (liquids) was not adequate for that purpose. Instead, they looked for tangible and testable causes of illness. This relentless search for causes is the driving force of the gradual revolution in medicine that has been taking place since Morgagni’s time: a shift away from symptom-based therapy towards causally based therapy. To this end doctors and researchers are delving ever deeper into the human body. Whereas Morgagni continued to look into the ‘solid components’, and more specifically the organs, of the body, his colleague Marie-FrançoisXavier Bichat (1771– 1802) began to distinguish between the different tissues present in organs. One of the most important steps up to that time was taken in 1858 by the Berlin pathologist Rudolf Virchow, whose work‘Cellularpathologie’drew attention to the cells of which all organs of the body are composed. Later, in the 20th century, increasing attention was paid to life processes within cells. All these efforts and discoveries are ultimately
8
directed towards a single goal, that of more precise medicine. The objectives are to identify the causes of a disease, to consider the possibilities for intervention and then to provide effective treatment. These objectives are more relevant now than ever before, and on the way towards them we are at present taking the next major step, that of replacing cellular pathology with molecular medicine. Genetics, genomics and proteomics are opening up totally new perspectives in diagnosis and therapy.
Accompanying the revolution: Morgagni and Virchow Giovanni Battista Morgagni (1682–1771) studied medicine in Bologna and in 1715 was appointed to the chair of anatomy at the University of Padua. In 1761, while still at Padua, he published his principal work De sedibus et causis morborum per anatomen indagatis (‘On the seats and causes of disease investigated by anatomy’). In this work he departed from the standard practice of the time by concentrating not on the symptoms and signs of disease, but on the location of disease within the organs of the body. In his view, the pathological changes in organs that he demonstrated in many autopsies were the true causes of illness. This view amounted to a rejection of the theory of humors (humoralism) that since the time of Hippocrates had attributed disease to an imbalance between the four bodily liquids blood, phlegm, yellow bile and black bile.
Rudolf Virchow (1821–1902) studied medicine at the Berlin Military Academy. After holding a professorship in Würzburg he took up a chair of pathological anatomy that had been created especially for him at the University of Berlin. He was a political activist, campaigning vigorously for democracy and public provision of healthcare. The journal “Virchows Archiv”, which he established in 1847 together with Benno E.H. Reinhardt, was the organ of scientific pathology. In 1858, with his work Die Cellularpathologie in ihrer Begründung auf physiologische und pathologische Gewebelehre (‘Cellular pathology as based on physiological and pathological histology’), he founded a new theory of pathology in which the cells of the body were regarded as the sites of origin of pathological changes.
Every disease is influenced by a larger or smaller number of factors. These include on the one hand environmental factors such as toxins, radiation, infections, nutrition, age, stress and much more besides, and on the other hand the genetic predisposition that causes our body to react to the environment in a certain way. Small changes in our genes can trigger, prevent, promote or alleviate diseases. The same applies to external influences. Whether, when, and how severely a person falls ill is determined by the combination of all these factors – and proteins play a central role in mediating these effects. They read and make working copies of the genes; they carry out the instructions, while at the same time regulating the
Target: the molecular network of the cell
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
9
action, of the genes; they receive signals from the environment, pass these on and incorporate them into the molecular network of the cell. It is precisely in this interplay of environment, genes and proteins (as well as a variety of other equally important substances that differ from case to case) that drugs exert their effects. They act directly on the molecules that make up our body
The changing role of biology in medicine
chemical synthesis improvement by computer
target molecule
potential drugs
binding assay
biological target search
rational drug design
high-throughputscreening
The tasks of biology in medicine have changed greatly in the past few decades. Over this period the initiative for innovations has passed increasingly from chemistry to biology. Molecular medicine is an expression of this change. Evaluation As recently as the 1970s the principal task of biologists in medicine was to evaluate, i.e. to test the effectiveness of, new substances produced by chemists. Targets As knowledge of the molecular basis of diseases increased, biology was able to provide new targets for drug development. These targets formed, and still form, the basis for the search for new medicines by chemical synthesis. Rational drug design Rational drug development arose as a result of increasing knowledge of the structure, i.e. the form, of proteins. The idea was that drugs would be designed by
10
best candidate
efficacy studies
biological evaluation
computer so as to interact in a highly specific way with a certain part of a target molecule; only afterwards would they be produced by chemists. This approach has not yet brought the success that was hoped for. High-throughput screening New, automated methods then made it possible to test large numbers of substances in hundreds or even thousands of miniaturised assays for certain biological, chemical and physical properties. Experiments such as binding to a target, which previously had to be performed individually in molecular biology laboratories, could now be performed in this way. The desired properties of suitable molecules can then be improved in a further step. Highthroughput screening has already resulted in the development of a number of particularly effective drugs.
– and in this sense are themselves an important environmental influence. The more we know about the actions of molecules in our body, the more effectively we are able to intervene when these actions become disordered. z Every newly discovered molecule that plays a role in the development of a disease constitutes a potential target for drugs. For example, in the past few decades biologists have discovered more and more oncogenes, i.e. cancer-promoting gene variants. Many anticancer agents act by restoring the correct function of the products of these genes (mostly proteins). z Knowledge of the structure, i.e. the three-dimensional form, of a target molecule makes it possible to decide in advance whether a given substance has any potential for use as a drug. Though it has yet to notch up many successes, computerbased rational drug design can greatly reduce the number of substances selected for further development. z If the genetic preconditions for a disease are known, a patient’s individual risk can be determined and appropriate preventive action taken. Sickle-cell anemia is an example of this. In this condition, an inherited modification of a certain component of the gene for hemoglobin, the red blood pigment, results in production of an altered protein that changes shape when the oxygen supply is inadequate. Under these conditions the red blood cells assume the form of a sickle, clump together and block the blood vessels. Carriers of this trait therefore need to avoid great heights and changes in air pressure (e.g. in aeroplanes), among other things. z Many diseases are amenable to intervention at the gene level. For example, genes can be turned on or off by drugs, and one day it may even be possible to replace genes completely by means of gene therapy. It is precisely in this latter field, however, that further intensive research is required. In many cases – e.g. severe hereditary diseases due to mutation of a single gene or a small number of genes – gene therapy, along with the stem cell therapy, offers the only hope of genuine cure. z Drugs do not always have the same effects. The effect of a given drug can be too strong, too weak or absent altogether in people with the same symptoms. Moreover, adverse effects are always likely to occur. Our genes are at least partly responsible for these too; the discipline of pharmacogenetics investigates these relationships and attempts to foresee, and ultimately forestall, such problems.
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
11
Genetics, genomics and proteomics thus provide medicine with a variety of new ways of intervening in the development and progression of diseases. Nevertheless, intervention has not become easier, for the deeper medicine looks into life processes, the more complex are the things it sees. The humoral pathology of Hippocrates distinguished between four humors; Morgagni extended the search for the seat of diseases to a couple of dozen organs; Bichat concerned himself with a few hundred bodily tissues; Virchow directed attention to the body’s cells, of which there are about 100 million million; and each of these cells contains an enormous number of nucleic acids, proteins, sugars, fats and other organic and inorganic substances. And in addition to all this is the far less measurable influence of external factors. Nevertheless, the effort is worthwhile. For in the past, methods of combating complex diseases were based largely on trial and error, precisely because such disorders are not caused by a simple infection or gene mutation, but rather arise as a result of a combination of external and internal, predisposing and protective, and variable or unchangeable influences. This is true of most of the major diseases that afflict people in industrialised countries, e.g. cancer, Alzheimer’s disease, diabetes and cardiovascular disease. Every ray of light that genetics, genomics and proteomics cast on the factors that contribute to these diseases helps in the fight against them.
A multiplicity of possible causes
This is because disease-inducing environmental influences can generally be modified – where as our genetic makeup generally cannot. Among the risk factors that contribute to the development of disease, our genetic predisposition is a constant. And this makes it all the more important for us to learn more about, and where possible to limit, its influence. In the 1980s scientists succeeded in identifying the genetic basis of a number of severe hereditary diseases brought about by a single defective gene. These include Huntington’s disease (Huntington’s chorea), cystic fibrosis (mucoviscidosis) and hemophilia. More refined methods now allow scientists to investigate the genetic causes of more complex diseases in which various genes can exert positive or negative influences. z Monogenic diseases such as Huntington’s disease, cystic fibrosis and hemophilia follow the classical (mendelian) laws of
The central importance of genes
12
Monogenic hereditary diseases: cystic fibrosis
cell membrane
nucleotidebinding region regulatory region
chloride
Defect on chromosome 7.
Impact of genetic defect on salt transport.
The first disease-causing genes to be discovered in the 1980s were associated with hereditary diseases due to specific mutations in individual genes. Such monogenic diseases can be classified on the basis of their pattern of inheritance: Autosomal recessive inheritance – The altered gene must be inherited from both the father and the mother for the disease to occur. Carriers of only one altered gene are healthy, but can transmit the disease to their children. An example of this is cystic fibrosis (also known as mucoviscidosis), which is due to a defect on chromosome 7. In this disease salt transport is disturbed in certain mucosal cells, causing the mucus of the respiratory tract, digestive tract and other organs to be extremely viscid. This results in frequent infections and inflammation. Whereas only half a century ago most cystic fibrosis sufferers died during childhood, those born today can expect to live 40 to 50 years thanks to specific (symptomatic) treatment and diet. Autosomal dominant inheritance – In this case a single altered gene is sufficient to cause the disease to appear, and the disease is transmitted to fifty percent of the children of sufferers. Huntington’s disease (Huntington’s chorea) is an example of such a disease. In this condition a defect in the Huntington gene on chromosome 4 causes production of an incorrectly formed protein known as amyloid (a similar change is seen in Alzheimer’s disease and Creutzfeldt-Jakob
disease). This leads firstly to motor disorders and later to mental deterioration. The disease generally appears between the age of 30 and 40 years and progresses in all cases to death after 5 to 20 years. Physiotherapy and diet can slow progression. Since the responsible gene was found in the mid-1990s, scientists have been working to develop suitable drugs, however as yet none has been approved for use. Sex-linked inheritance – In this case the responsible gene lies on a sex chromosome (gonosome), generally the X chromosome. The presence of one unaffected copy is sufficient to suppress the disease, therefore men are far more commonly affected than women, since they lack a second X chromosome. Hemophilia is such a disease. This is due to an inherited deficiency of a blood coagulation factor, as a result of which the affected person’s blood coagulates very slowly or not at all. Untreated hemophiliacs therefore die young of internal and external bleeding. Previously, hemophiliacs were treated with blood transfusions, and even today the missing coagulation factor is obtained mostly from donor blood. Unfortunately, diseases can be transmitted from the donor to the recipient in this way, though less so than with transfusion of whole blood. Now that the responsible gene has been identified, however, the missing factor can be produced using recombinant DNA technology, thus eliminating the risk of transmitting other diseases.
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
13
inheritance. The pattern of occurrence and non-occurrence of such diseases within affected families is determined by whether only one or both copies of the gene in question need to be altered for the disease to occur. In such cases the responsible genes are relatively easy to identify via studies comparing the genetic material of affected and unaffected members of a family. Genetic testing can be used to advise prospective parents of the risk of having a baby that will be affected by a heritable disease, such as cystic fibrosis. z By contrast, the pattern of inheritance of polygenic diseases, which include type 2 diabetes and most types of cancer, is not so simple, since many genes are involved. Most such diseases tend to cluster (occur with increased frequency) in certain families, but not in such a way that the precise distribution of affected and unaffected individuals can be predicted. This requires larger studies to identify the various genes that influence the disease to a greater or lesser extent. This task is rendered even more difficult by the fact that in this case genes that predispose to the disease can overlap with genes that protect against it. Hopes have therefore been placed in the study of single nucleotide polymorphisms, or SNPs (pronounced ‘snips’), i.e. changes in single subunits of the genome. The presence of such single nucleotide substitutions in important sections of a gene can have profound effects on the function of the corresponding gene product. The finding of an increased frequency of certain SNPs in association with a disease indicates that the genes concerned play an important role in the disease in question. z It is not always easy to separate environmental influences from genetic influences, especially since the environment can influence the behaviour of our genes. Twin studies and adoption studies are helpful here. Monozygotic (‘identical’) twins brought up in different families have an identical genome but are subject to different environmental influences, while dizygotic (‘fraternal’) twins brought up in the same family are subject to essentially identical environmental influences and have a similar, but not identical, genome. Finally, adopted children share essentially the same environment with, but are genetically quite different from, their stepbrothers and stepsisters. z Our genetic makeup also exerts a decisive influence on our predisposition to disease. Where genes that play a role in the
14
The complex interplay of genetic factors in disease Now that the findings of molecular biology are being applied to medicine it has become clear that very few diseases have simple genetic causes. In the vast majority of cases many different genes exert a greater or a lesser influence both on the disease itself and on each other’s action. 1. Complex diseases are multifactorial in origin. Both endogenous, in particular genetic, and exogenous factors play important roles, and the relative influence of these two types of factor can vary. Moreover, factors such as diet, environment and behaviour can influence the body’s reactions independently of the actual causes of disease. 2. Genes can protect against or predispose to the development of complex diseases. The combined influences of protective and predisposing factors result in an overall risk. It is therefore entirely possible for a person to have a large number of disease-promoting gene variants — and to transmit these to their offspring — without themselves ever becoming ill, provided only that they also have a sufficient number of protective gene variants. In the age of molecular medicine, terms such as ‘healthy’, ‘ill’, ‘normal’ and ‘abnormal’ are therefore no longer easy to define. 3. All risk factors — including genetic factors — for a complex disease can be either categorical or quantitative. A gene variant is said to act categorically if a certain disease can occur only in its presence. By contrast, quantitatively acting gene variants act additively (or else their effects are multiplied) up to a critical point at which disease occurs. 4. Complex diseases are polygenic, i.e. they result from the action of a number of different genes. The contributions of the individual genes are difficult to determine and can vary enormously, especially as genes influence each other’s action. 5. Another characteristic of complex diseases is genetic heterogeneity: since a number of genes can be jointly responsible for the occurrence of a disease, different combinations of genes can result in the same clinical picture. This complex interplay of influences means that the genetic
environment
genes
nutrition
disease
lifestyle
causes of complex diseases can generally be determined only via large, statistically complex studies and on the basis of a deep understanding of the molecular processes that take place in cells. Only now that the findings of genetics, genomics, proteomics and bioinformatics are being applied to medicine has this become possible. In the case of complex diseases, however, even tests based on techniques of molecular medicine can at best indicate only an approximate risk that an individual will develop a particular disease. Absolute assertions cannot be made. In the future, genetic testing may increasingly be used to guide patients and healthcare providers in designing optimal treatment strategies based on patient’s genetic variations. For example, pharmacogenetic research is already underway to provide physicians with a better understanding of the influence of genetic variation on an individual’s response to medication.
development of a disease (or, for example, in intolerance of a drug or failure of a drug to exert its expected effects) are known, an individual’s risk of developing that disease can to some extent be determined by appropriate genetic tests. Knowledge of his or her predisposition to a certain disease allows the individual concerned to take appropriate precautions and to modify his or her lifestyle accordingly – and if
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
15
necessary to take preventive medicines. Early prevention is therefore one of the potential applications of molecular medicine. Given, however, that most diseases result from the combined action of a large number of genetic and environmental factors and that predisposing and protective genes can overlap, such tests can only ever indicate a greater or lesser probability that an individual will develop a disease.
The importance of looking for the causes of disease has not changed at all since Morgagni propounded his organ-based pathology: only if we truly understand a disease can we treat it correctly. Nowadays,
Diagnosis and therapy look each other in the eye
Triple influence of genes: hepatitis C
1a, 1b, 2a, 2b, 3a
1b, 2a, 2b, 2c, 3a 4 4
5a
Over the past few years it has become increasingly clear that even infectious diseases are subject to a complex set of genetic influences. This is true not just in regard to the causative pathogens – whose genetic background is often well known – but also in regard to the host. Genetic differences between individuals make some people more, and others less, susceptible to particular infections. In addition, our genes influence the way our body deals with all types of drug, including anti-infective agents. In particular, drugs for use against viruses, which evolve rapidly, often show unsatisfactory efficacy and troublesome side effects. Therefore, if better drugs are to be developed, both the genome of the pathogen and that of the patient must be taken into account. An example of an infectious disease that shows this kind of
16
complexity is hepatitis C, which is expected to become far more common over the next few decades. Untreated, this disease leads to cirrhosis of the liver in about 20% of those infected and to liver 1b cancer in a smaller proportion 2a of patients. The responsible 1b, pathogen, hepatitis C virus 1b, 3a 6 (HCV), occurs in at least six different types, plus subtypes, 3b that occur with different frequency in different parts of the 1b, 3a world (see figure). The existence of these variants has a major influence on the effectiveness of standard therapies. For example, interferon, the most important agent for use in this disease, is relatively ineffective against HCV type 1, the predominant type in Europe and North America. Moreover, this type of HCV generally causes more severe disease than other types. In addition, standard interferon preparations usually cause severe side effects, all the more so because they generally have to be taken three times weekly. Administration at such short intervals is necessary because interferon is broken down very rapidly in the body and therefore acts for only a few hours. Patients are thus obliged to put up with a constant ebb and flow of side effects. Improved drugs such as pegylated interferon have been available for some time now. Used in combination with ribavirin, pegylated interferon significantly increases the efficacy of treatment.
however, scientists are searching at other sites, namely in the genes and proteins of our cells – in other words, at the sites where drugs act. For medicine, this is a major advance in that for the first time in history diagnosis and therapy are, so to speak, looking each other in the eye. For the first time it is possible to determine the causes of a disease on the basis of a patient’s genetic predisposition, to predict the effect of drugs on a disease on the basis of the molecular characteristics of the drugs concerned, and finally to choose therapy that is optimal for the individual patient. Knowledge of the molecular level of the disease process thus opens up completely new approaches to treatment: new targets, new strategies, early prevention and better understanding of the effects and side effects of drugs.
Molecular structure of Pegasys with (right) and without (left) PEG coating
One new way of improving therapy is therefore to increase the period of time during which interferon remains in the body. The product Pegasys works in this way. In this medicine the interferon is enclosed within a coating made up of a branched molecule known as polyethylene glycol (PEG). This delays breakdown of the interferon, with the result that the drug only has to be taken once weekly. This results in greater efficacy and fewer side effects. Genetic factors are important in hepatitis C not only in relation to virus type and drug metabolism, but also in that they partly determine the clinical course of the disease. As mentioned above, infections with HCV type 1 are generally more severe than those with other types of the virus. Nevertheless, infection with any HCV type can show an acute or
chronic course, be mild or severe, and undergo spontaneous cure or lead to liver cancer, the direction taken by the disease in an individual being largely determined by that individual’s genome. At present, however, the genes responsible for these differences are largely unknown. These genes are therefore another important object of HCV research, since depending on the genetic predisposition of the patient (and the virus), treatment may be necessary or unnecessary, a particular drug may be suitable or unsuitable, and cure may be possible or unlikely. Therefore, the more is known about the relationships between the genome of the pathogen and that of the host, the more specific will be the drugs that can be developed for use in this disease.
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
17
At the same time, these new methods and discoveries are a continuation of the revolution that began several hundred years ago – less emphasis on signs and symptoms, more investigation of causes. And despite all the progress that has been made, the possibilities offered by medicine remain limited: the interactions between our genes are so complex, our body is so adaptable, and the influence of our environment and lifestyle is so great that we cannot expect to find definitive answers and one-hundred-percent effective therapies in the very near future. Rather, genetics, genomics and proteomics will first help doctors to avoid ineffective or even dangerous therapies – after all, there is no such thing as a panacea. But even this is a major step forward.
References Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations. Pharmacogenomics 1: 23-26, 2001 Bundesministerium für Bildung und Forschung (ed.): Das nationale Genomforschungsnetz. Bonn, 2003 Geschäftsstelle des Wissenschaftlichen Koordinierungskomitees des Deutschen Humangenomprojekts (ed.): Das Humangenomprojekt – 1st and 2nd edition Healthnet-Services GmbH: (M)Eine Geschichte der Pathologie, Teil 1-7: http://hns.pvs-bw.de/mod.php?mod=userpage&page_id=30
18
Pharmacogenomics: genes and drug response
To d a y
A
B
D
C
To m o r r o w
Diagnostics
A
A drug may work well in one person, but poorly or not at all in another. One person may tolerate a drug well, whereas another develops side effects. This fact is as well known as it is unfortunate. These individual differences are largely due to our genome, the genetic blueprint that makes each of us unique. Thanks to new knowledge and techniques, medicine is now able to take greater account of these differences – thus leading to the development of more effective, safer and better tolerated drugs.
D
B C
People are all different – that’s obvious at first glance. There are extra-long beds, creams for sensitive skin, height-adjustable seat belts, tailored shirts and three dozen standard shoe sizes. Each of us has different strengths and weaknesses, abilities and needs. Environmental factors, chance and above all the small differences in our genomes make each of us unique. But if one person finds a standard bed to be too short and another finds a standard shoe size too big, why should we assume that everyone responds to drugs in the same way? Terms In fact, it has been known for some time that the efficacy Pharmacogenetics describes the influence of genes on the and tolerability of drugs efficacy and side effects of drugs. Pharmacogenomics studies interactions between drugs and vary from one person to the the genome. next. Thus, some patients Pharmacokinetics investigates the uptake, conversion and need a lot more or a lot less breakdown of drugs in the body over time. Environmental factors, diet and genetic predisposition all play a role. of a given drug than most Pharmacodynamics deals with the influence of genes on the people; side effects keep ocinteractions between drugs and their molecular targets. curring unexpectedly; and sometimes a drug that is usually highly effective does not work at all. Our uniqueness is reflected in our body’s response to drugs. Because of this, personalised medicine has emerged as a hot new topic of discussion. Future drugs, it is hoped, will be better adapted to our genetic diversity and dissimilar life circumstances and will be more efficient, more specific and safer. And they will be supported by a battery of fast, simple genetic tests that will enable doctors to select the right drug for their patients’ specific needs.
First example Herceptin
20
A first example of a genetically specific drug has already been introduced in the form of Herceptin, which is used in the treatment of breast cancer. Herceptin is only effective in women with a genetic defect which results in the overproduction of a molecule known as the HER2 receptor. When present in excessive numbers on the surface of certain breast cells, these receptors promote cellular growth, leading to tumours. Herceptin is directed against the receptor; it therefore only helps women who have an increased number of copies of the relevant gene. In all other women this highly specific drug is much less effective. Herceptin can therefore be used only in conjunction with a suitable genetic test. Three such tests are currently available: First, the receptors can be visualised on the surface of tumour cells with the help of
specific antibodies linked to a dye. Second, there is a gene test known as FISH which directly detects the genetic change in question. Recently a third test based on the polymerase chain reaction (see chapter about PCR) has become available – at least for research purposes. Using this technique investigators can copy the relevant DNA section in the laboratory, thus revealing if the dangerous genetic change is present. This development has given rise to much hope and anxiety: Expectations range from the attainment of perfectly personalised drugs to progress for the rich only, and the boundaries between the feasible and the conceivable, the possible and the necessary, science and fiction, quickly become blurred. Whether drugs will ever become as individualised as tailored shirts is more than dubious, but even a few off-the-peg sizes and somewhat more variety would be a huge step forward. The engine driving this progress is genome research.
As early as 1958 the German pediatrician Friedrich Vogel suspected that our genes play an important role in determining our response to drugs. He even proposed a name for the branch of science that investigates this phenomenon: pharmacogenetics, the study of the influence of genes on drug effects. The new approach quickly led to the first application of genetics in medicine. Over 100 relevant genes are now known, and many more will follow as scientists rapidly refine the field of pharmacogenetics with the help
Pharmacogenetics as a research discipline
Pharmacogenomics: genes and drug response
21
of methods and knowledge gained from unravelling the human genome. Meanwhile the subject of enquiry has shifted from individual genes and their effects to the interplay between drugs and the genome – pharmacogenomics. The significance of this new branch of science is far greater than the small change in the name implies. We now know that our genome influences the effects of drugs in at least three ways – and until a few years ago only one of them had really been considered. z ‘Pharmacokinetics’ describes the metabolism of drugs, i.e. their uptake, conversion and breakdown in the body. In some people, for example, a drug fails even before it reaches its site of action. Their body takes up the molecule very slowly or sometimes not at all. In other individuals, conversion of the drug (for example to remove a protective molecular cap) proceeds sluggishly. And a third group of patients breaks the drug down too quickly or too slowly. If a drug is broken down too rapidly, taken up too slowly or converted too slowly, its effects will not be felt. Conversely, if a drug is broken down or excreted too slowly, its effects may be magnified. It then remains too long in the body, and the risk of side effects increases sharply. These differences are due not only to environmental factors and diet but also to people’s genetic makeup. This is because specific proteins in our body are responsible for metabolising drugs. And the blueprint for those proteins resides in our genes. Hence, small differences in the genomes of patients can result in pharmacokinetic differences. The discipline described by Vogel dealing with the relationship between genes and drug metabolism is therefore now regarded as forming part of pharmacokinetics. z ‘Pharmacodynamics’, by contrast, describes the interaction between drugs and their molecular targets. In the classic case this relates to the etiology of a disease, i.e. its underlying molecular causes. Usually the activity of an endogenous protein is impaired. The shape of such proteins is genetically determined. Small differences in our genome can therefore significantly alter the structure of these proteins. And since drugs are usually highly sensitive to such differences – after all, a drug is supposed to act on a very specific target molecule so as to have as few side effects as possible – they may become ineffective if the target molecule is altered. z Palliative drugs, i.e. drugs that bring relief, constitute the third and most complex pharmacodynamic way in which genes can interfere with the activity of drugs. Palliative drugs
22
When drugs do not work: pharmacogenetics
effect: • target molecule(s) • etiological or palliative • potential side effects
conversion: • not always necessary cleavage of a protective group
uptake: • e.g. via specific receptors or channels
breakdown:
cleavage of protective groups
• often by modification and/or cleavage • new compounds may be formed
heart
gut
excretion: • often possible only after conversion or breakdown
: (inactive) drug : (inactive) drug in the bloodstream : protective groups : active drug
ureter
: cleaved, water-soluble degradation product
Our genome influences the effects of drugs at at least three levels: z The metabolism of a drug – i.e. its uptake, conversion or breakdown – may be prevented, slowed or accelerated. Pharmacokinetics investigates the causes of these phenomena. z Modification of the target molecule can directly weaken
z
or prevent the action of a drug. Pharmacodynamics describes these underlying ‘etiological’ differences. If a drug does not act on the cause of a disease but rather on its manifestations, many genes may be involved in and interfere with its effects. These ‘palliative’ differences are also the subject of pharmacodynamic research.
do not act directly on the cause of a disease, but rather on its symptoms. Analgesics, for example, usually do not influence the cause of pain but merely the perception of pain in the brain. Nevertheless, such drugs often successfully counter the cause if (as in painful cramps) the cause is directly related to the symptoms (i.e. the pain itself causes the cramps). The ways in which such drugs counteract a disease or relieve its symptoms are usually highly complex, and the genetic reasons for why a drug might not work as desired can be just as varied.
Pharmacogenomics: genes and drug response
23
Genes have been found for each of these three areas of pharmacogenetics. The vast majority of them act at the pharmacokinetic level. This research area is already being applied to the routine development of new drugs. Particularly in the early phases of clinical testing, subjects are already screened for specific metabolic characteristics. This is generally done in order to obtain a picture of pharmacokinetic differences in the population that is as representative as possible. In this way entire drug families can be classified according to their pharmacokinetic properties, thus shedding light on many problems. For example, an enzyme group known as cytochrome P450 plays an important role in the metabolism of many drugs. One important function of P450 proteins is to make water-insoluble substances soluble in order to facilitate their excretion. Over a third of all drugs and exogenous substances in the body fall into this category. We now know that because of genetic differences the P450 machinery works more sluggishly in some people than in others. In such people, known as poor metabolisers, the molecular disposal of drugs proceeds more slowly. Another characteristic of the P450 family of enzymes is medically important: They convert fat-soluble molecules to make them soluble in water. Bonds in the target substances are broken, new bonds are forged and additional molecular groups are attached. In other words, a new molecule with entirely new properties is created. In the worst-case scenario, a harmless substance may be transformed into a carcinogenic toxin. Many side effects of drugs are due to the work of members of the P450 enzyme family – and because the enzymes’ activity can vary from one person to the next, such side effects tend to occur sporadically.
Prominent example: cytochrome P450
Hurdles for drugs: genetic polymorphisms Genetic differences between individuals strongly influence the function of the corresponding gene products (usually proteins) and in this way affect the activity of drugs in the body. Some of these genetic variants, known as polymorphisms, have been identified. They influence people’s response to drug therapies at the pharmacokinetic or pharmacodynamic level. 1. Pharmacokinetics: Genes coding for enzymes involved in the metabolism of drugs form the largest group of known pharmacogenetic factors. Fluctuations in their activity can slow or accelerate the uptake, conversion or
24
excretion of drugs. As a result, the drugs do not remain in the body long enough to be effective or remain in the body too long so that the risk of dangerous side effects increases. 2. Pharmacodynamics: Some genes have also been found that are directly responsible for the structure of the target molecule, influence the associated signalling pathway or interfere with some other metabolic pathway involved in the manifestations of a disease. The interactions between such genes and drugs can be highly complex.
Personalised medicine: the objectives of pharmacogenomics Dosage Are the patient’s individual pharmacokinetic factors known, and can the dose of a drug be adjusted upward or downward to suit his/her metabolism? This would help ensure that the drug works while at the same time reducing its side effects. Efficacy If it is already known in the initial stage of treatment which drug is likely to work for a patient, a lot of trial and error would be saved. Valuable time would be gained for the treatment, unnecessary expense would be reduced and the patience of the doctor and patient alike would be spared.
Safety Can adverse reactions to a specific drug be predicted? Can the patient be switched to another drug that he/she is more likely to tolerate? Are there alternatives? Can provisions be made in advance for medical supportive measures? Prevention If the cause of the disease is related to genetic factors, the disease can be diagnosed early by means of tests and possibly avoided by initiating specific measures such as diet or exercise.
As with all proteins in the body, this variable activity ultimately depends on our genome, which, in turn, is affected by external factors, including drugs. Cytochrome P450 is a good example of this phenomenon as well. The 50-plus genes that code for this family of enzymes can be activated or suppressed by drugs. In this way drugs may affect each other, intensifying or cancelling each other’s effects, even if they have completely unrelated targets. Thus, the genome and drugs form a complex network of dependencies and uncertainties, of effects and side effects, which ultimately makes our body’s individual response to drugs unique, especially when pharmacodynamic factors are also taken into consideration. Pharmacogenomics can elucidate this interplay and help doctors find out whether, at what dose and with what risk a drug can be administered to a given patient before problems arise in the first place.
Genome and environment act together
The old slogan ‘one size fits all’ still applies to most drugs. If a drug fails to work in too many patients, if its activity fluctuates too strongly or if its side effects are too common or too severe, it is not granted regulatory approval. This means a huge loss for the manufacturer and one fewer promising treatment for patients. The failure of a drug in this respect can be due to at least three reasons: z Are there problems with the pharmacokinetics, i.e. the uptake, conversion or breakdown of the drug, and if so, what can be done about it?
Old hurdles for new drugs
Pharmacogenomics: genes and drug response
25
z Are there too many differences in the target molecule be-
tween individuals, and is it possible to target drugs at more constant regions of the same target? z Is a patient’s disease due to various, possibly unknown, causes, and if so does the drug act on just one of them? Distinguishing between these three possibilities has proved very difficult and often impossible. The consequence is that the development of new drugs is ultimately limited – to the detriment of patients, who have only a limited range of therapeutic options available: z New classes of drugs that exhibit especially marked genetic dependencies are not further pursued. z Important target molecules are not considered because of their variable structure, although perhaps this very changeability may cause disease. z If it is not possible to distinguish between the causes of a disease, often only palliative drugs will produce a uniform effect, since they act on the effects rather than on the cause of the disease. However, it is usually much safer and more effective to eliminate the underlying molecular cause.
To understand the molecular causes of a disease or a drug’s failure to have an effect, we therefore need to look at the genes. We first have to identify those sections of the genome that are involved in causing the disease or in a drug’s activity and metabolism. However, because the metabolic processes on which drugs act are extremely complicated, this is no small task. A promising solution is to look for single nucleotide polymorphisms, or SNPs for short. These variations, which are spread more or less randomly throughout the genome, are thought to determine our individual genetic differences to a large degree. Depending on the position of a SNP within a gene, the corresponding gene product is more or less strongly affected. An enzyme, for example, may be impaired, destroyed or improved – with corresponding implications for drugs that interact with that enzyme. If specific SNPs are repeatedly associated with a disease or with specific side effects or drug failures, it can be assumed that the genes concerned have something to do with the observed disturbance.
SNPs yield the first evidence
26
Once such SNPs have been found, the associated genes have to be identified. Until a few years ago this meant tedious searching and sequencing. Today, however, this procedure can be bypassed. Thanks to the Human Genome Project, which sequenced the entire human genome, the relevant data are already available. The search for the identity and function of a gene is also straightforward, since researchers can conduct a computer search for available data or comparable genes. Even the associated gene products (usually proteins) and their functions can be pinpointed quickly and easily in globally linked databases. The work becomes more difficult when the gene product in question is unknown or has only been investigated in another context. Many proteins occur in numerous variants, exist in a number of complexes or assume different functions in different cells. In such cases protein biochemists are consulted; they investigate the molecule in detail for its function and role in the body. After all, in order for drug development and treatment to take a pharmacogenetically active gene variant into account, the causes and effects have to be identified beyond a doubt. Otherwise there is no gain over the old trial-and-error approach.
Databases to the rescue
The final link in the pharmacogenetic research chain is to search for the presence of the gene variant in a patient. Conducting this search within a reasonable period of time was thought to be impossible just a few years ago. Our genome contains at least 30,000 – 40,000 genes and over three billion building blocks – and a SNP is a substitution of just one of those building blocks at a very specific site within an individual gene. Identifying this site requires tens of thousands of individual experiments. Today all these experiments fit on a single tiny silicon chip. Measuring just one centimeter square, a single DNA chip does the work of a large laboratory. Using modern techniques, scientists can array one hundred to several hundred thousand short DNA fragments or even entire genes on the chip. The DNA fragments, known as oligonucleotides, are like anglers fishing for highly specific genome segments in a solution. The key to how this works is that the oligonucleotides are arrayed on the chip in the form of single-strand DNA segments. If a genome section having the same sequence is present in the test solution, it binds to the fragment on the chip to form double-stranded
10,000 experiments at once: DNA chips
Pharmacogenomics: genes and drug response
27
DNA. The binding process generates a fluorescent pulse which can be detected with the help of a laser. Because the arrangement of the DNA fragments on the chip is known, the method can be used to test for thousands of gene segments simultaneously. And that is precisely what’s needed for fast, simple and affordable genetic testing.
Thus the requirements for more individualised drug therapy have been met – at least as far as the technical side is concerned. Though many approaches are still in the trial phase and problems are likely to be encountered, applications are already emerging, for example a DNA chip that recognises the various pharmacogenetically important gene variants of the P450 enzyme family. This chip is the world’s first commercial pharmacogenomic product. It can be used, for example, to screen test subjects during the process of drug development (even though at present such screening isn’t usually based on a gene test, but on physiological investigations). More specific diagnoses will follow, and in a later step it is expected that it will be possible to predict intolerance reactions and treatment failures for new and existing drugs. This will enable doctors and patients to resort to other drugs or at least to prepare for expected problems. Tailor-made drugs may be a long time in coming. But even these rather modest goals are beset by obstacles. As with all new medical developments, there are ethical and legal reservations as well as financial misgivings and still unresolved scientific and technical issues. In addition, applications of modern genetic and genomic research have consistently raised public misgivings, and particularly in this area pharmacogenomics is in need of a rethink: The essential requirement for more personalised medicine is unobstructed access to genetic data. To be of use to patients, findings must be made available to doctors and pharmacists. In any case, the prescription of a drug with significant pharmacogenetic properties reveals the patient’s diagnosis anyway. And we should also realise that this is already the case with many drugs today. Ethical, legal and societal implications of genetics and genomics in medicine as well as the technical and economic requisites are dealt with in more detail in the chapter on basic conditions.
Potential and obstacles
28
References Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations. Pharmacogenomics 1: 23-26, 2001 Lindpaintner K: Herausforderungen und Verheißungen einer individuell zugeschnittenen Behandlung komplexer Krankheiten. Roche, 2000 Froböse R, Albrecht H: Die ganz persönliche Pille. DIE ZEIT, 15/2002 Ma MK et al.: Genetic basis of drug metabolism. Am J Health Syst Pharm 59(21): 2061-2069, 2002 Kroll W, Hartwig W: Pharmakogenomik. Nachrichten aus der Chemie 50, March 2002 Lifescience.de: Pharmacogenomik – Die Suche nach den idealen Pillen. http://www.lifescience.de/ratgeber/mitte/index2.html Human Genome Project Website: http://www.ornl.gov/ Human Genome Project – Pharmacogenomics: http://www.ornl.gov/hgmis/medicine/pharma.html
Pharmacogenomics: genes and drug response
29
Proteomics: seeing through the undergrowth
Every cell in our body contains at least 100,000 different proteins, and every cell type contains a different set of proteins. Proteins form a vast and highly complex network: they construct and break down molecules, they transport, store and mobilise substances, they allow cells to communicate with each other, they give and receive orders, they keep cells alive and can program cells to die. It is precisely in this network that drugs act, and only now are we beginning to understand how, where, when and why they act. Proteomics can help us to see through this molecular undergrowth.
The two Australian scientists who coined the term ‘proteome’ at a conference in Siena, Italy in 1994 were probably hoping that use of this term would raise the profile of their field of study, protein research. For they coined this term in deliberate analogy to the term ‘genome’, the principal focus of biological research in the 1990s. At that time the project aimed at decoding the human genome was proceeding at breakneck speed with the aid of unprecedented financial, technical and organisational backing. One by one, yeasts, threadworms and even huTerms man beings revealed their Genome the (largely unchangeable) totality of the genes of genetic makeup. By contrast, an organism. proteins, the major products Proteome the (in most cases constantly changing) totality of the proteins of an organism. of all these genomes, were Genomics the study of the form, function and interactions largely ignored at that time, of the genes of an organism. Proteomics the study of the form, function and interactions being regarded as a subject of the proteins of an organism. for basic research by scientists interested only in the acquisition of knowledge for its own sake. Since then, this situation has changed completely. For a number of years now, scientists conducting basic research have not been the only ones wanting to find out exactly what cells produce on the basis of their genome. For proteins bring about the vital processes that take place in an organism and are therefore the most important target for strategies – e.g. those based on the use of drugs – aimed at interfering with these processes. However, whereas a cell can only ever have one genome, a cell’s proteome, that is to say the totality of its proteins, is highly variable. In theory, each cell’s proteome is different at every point in time and at every different site within the cell, since unlike its genetic material, a cell’s proteins are being constantly produced, broken down, altered, moved around, bound and separated. The task of bringing light into this tangled undergrowth is an even greater scientific challenge than that of decoding the human genome, but at the same time it will accelerate progress in gene and genome research.
The importance of proteins lies in the multiplicity of the tasks that they perform. They play a central role in almost all the processes involved in the life of an organism or – viewed on a smaller scale – a cell:
Diverse structures and functions
32
z Structural proteins are responsible for the form and shape of
cells. They form the structural framework of the cell and a large part of the outer envelope of the cell. Bodily structures such as tendons and hairs are made of protein. Structural proteins account for most of the protein in our body. z Metabolic proteins, or enzymes, are responsible for the constant synthesis, rearrangement and breakdown of all the substances that are required by, or formed within, the body; they also provide the energy required for these processes. Even minor disturbances of the complex interplay between these proteins can result in serious diseases. z Signalling proteins are responsible for communication within (and to some extent also outside of) the body. These include hormones and intracellular messenger substances. Many medicines act by interfering with signalling pathways within the body. z Regulatory proteins control the processes that take place within an organism, including correct transcription of DNA, the genetic material. In addition, proteins perform a variety of other tasks, e.g. as antibodies in the immune system, oxygen transporters in the blood and motors in muscle. The complex interplay between all the proteins of the body is as fascinating as it is impenetrable. It is estimated that each type of cell in our body contains about 100,000 different proteins, whereas our genome contains only 30,000 to 40,000 genes and moreover is the same in all cells. Because of the
Diverse and changeable: the structure of proteins
} }
primary structure
secondary structure
tertiary structure
A chain of up to twenty different amino acids (primary structure – the variable regions are indicated by the squares of different colours) arranges itself into three-dimensional structures. Among these, helical and planar regions are particularly common. The position of these secondary structures in relation to one another determines the shape of the protein, i.e. its tertiary structure. Often, a number of proteins form functional complexes with quaternary structures; only when arranged in this way can they perform their intended functions. When purifying proteins, it is extremely difficult to retain such protein complexes in their original form.
quaternary structure
Proteomics: seeing through the undergrowth
33
amount and complexity of the relevant data and knowledge already available or yet to be obtained, attempts to intervene in the world of proteins – e.g. by means of drugs – have until now been based largely on the principle of trial and error. The new discipline of proteomics aims to change this.
As a first step, many laboratories around the world are working to develop as complete a list as possible of all the proteins that occur in the human body. This list is analogous to the sequence of the human genome insofar as it provides no answers by itself, but forms an immensely important basis for further research. Just as gene researchers sifted through the DNA sequence to identify genes and regulatory elements, protein researchers can refer to this protein catalogue when conducting their experiments. At the same time, comparison of the genome with the proteome facilitates the search for new genes and thereby provides genome researchers with an additional tool for interpreting their results. In the next step, the proteome is considered in relation to its time and place of occurrence and above all in relation to the external influences that act upon it. The appearance or disappearance of proteins in the course of an illness or in response to administration of drugs has already been a subject of investigation for decades, however the search for such changes can now be conducted in far more systematic fashion with the aid of proteomics. For example, a ‘differential protein expression analysis’, i.e. a comparison between the proteome of a healthy subject and that of a patient with a given disease under the same conditions, can in theory identify and precisely characterise all the differences between the proteome of these two individuals and make them visible at a glance. In short, proteome research can help to identify the causes and effects of diseases, and of the treatment of diseases, more rapidly, more simply and more precisely.
Protein catalogues to provide order and perspective
Proteome research thus constitutes an important link between various fields and disciplines of medicine: z Diagnosis. The effects of genetic changes first become manifest at the level of proteins. In many cases the question of whether a known mutation actually brings about an illness
Proteome research as a link between disciplines
34
or is counterbalanced by other factors can be answered only by means of protein testing. Proteins are therefore more suitable than genes for use as diagnostic markers of complex diseases. z Therapy. In the vast majority of cases, medicines do not alter the genome. Even when they influence the expression of the genome, they do so by inducing changes in the proteome. Such changes therefore indicate whether, and if so to what extent, an administered drug exerts an effect. A ‘snapshot’ of a patient’s proteome can thus help the doctor to adapt treatment to that patient’s individual requirements. z Research. One of the objectives of proteomics is to identify metabolic and signalling pathways that play a role in the development of diseases. Each newly discovered protein in such a pathway is a potential target for new drugs. Proteomics is expected to provide a major boost for drug discovery. z Development. The more is known about a target molecule, the more simply and rapidly can a drug that acts specifically on that molecule be developed. Knowledge of the proteome can also provide information on potential problems and side effects of a drug before they occur in clinical trials. For example, certain marker proteins can indicate that a cell has been exposed to a substance that is toxic to them. Toxic effects of new drugs can thus sometimes be predicted at an early stage of drug development.
Given its huge importance, protein research has always occupied a central role in the search for new diagnostic methods, treatments and techniques. Most of the methods used today were known long before the term proteomics was coined in Siena. Proteomics is now extending this branch of biology by making it possible to acquire and analyse huge amounts of data in a minimal amount of time – just as in genome research. And this possibility opens up new applications for proteomics. In classical protein experiments, a single protein is isolated, its identity, form and function are determined, the gene that codes for it may be identified and, finally, the location of the protein within the cell is determined. In proteomics the procedure is exactly the same in principle – except that thousands of proteins are analysed, identified and quantified (i.e. the amount present within a cell determined) simultaneously. To achieve this, the
Strategy: standardisation and automation
Proteomics: seeing through the undergrowth
35
clinical sample tissue
protein separation (2D gel electrophoresis) isoelectric point (pl)
molecular weight (kDa)
blood
tissue protein extraction
detail
spot picking (each spot contains a protein) bioinformatics protein identification by database search
mass spectrometry (of each spot)
Proteome analysis using 2D gel electrophoresis and MALDI-TOF mass spectrometry. Proteins are extracted from clinical material such as blood or tissue and
separated by 2D gel electrophoresis. The resulting protein spots are then cut out of the gel and the proteins are identified by mass spectrometry.
various steps in the process are performed largely automatically and in parallel, and the results obtained are analysed using powerful computers and specially developed software programs.
The first step is always that of separating the mixture of proteins in a sample. The most important method used to achieve this – both in classical protein analysis and in proteomics – is two-dimensional (2-D) gel electrophoresis. This technique has been used routinely since the 1980s, and was in fact the subject of the conference in Siena at which the term proteomics was first used. In 2-D gel electrophoresis, the proteins in a sample are applied to a rectangular piece of synthetic gel. Within this gel the proteins are separated firstly according to their charge and then – at
Separation by charge and size
36
Two-dimensional gel electrophoresis: an important separation method control
Alzheimer’s disease
molecular weight SDS-Page
IEF
Production of glial fibrillary acidic protein (GFAP) is increased in Alzheimer’s disease. In two-dimensional gel electrophoresis, proteins are separated according to their charge and size. The first step is separation by charge. This is achieved by means of isoelectric focusing (IEF) using a polyacrylamide gel to which a pH gradient is applied for this purpose. When an electric field is applied to the gel, proteins migrate along the gradient for as long as they possess a net charge. Once a pro-
tein reaches its isoelectric point, i.e. once its net charge is zero, it stops. In the next step, the separated proteins are further sorted according to size. This occurs at a right angle to the direction of the first separation, i.e. in a second dimension. The detergent sodium dodecyl sulphate (SDS) is added for this purpose. The molecules of this substance bind to the proteins to an extent that depends upon the size of the protein molecules. Once again, an electric field is applied, however in this step the rate of migration is determined by the charge of the SDS, which in turn is determined by the size of the protein molecules. This type of separation is known as SDS-PAGE, or sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Proteins that are very similar (e.g. modified forms of the same molecule) are often extremely difficult to separate in conventional gels. In such cases use is made of narrowrange gels with an extremely gentle gradient within the pH range being examined. In this way even minimal differences in charge can be detected.
a right angle to the direction of separation in the first step – according to their size. The proteins are then rendered visible, resulting in a complex pattern of spots in which each spot represents a specific protein. The larger the spot, the more of the protein concerned was present in the sample – as in a map in which larger towns are indicated by larger spots. The pattern of spots, which is referred to as a proteome map, indicates both the identity and the amount of the various proteins present. Under the same conditions a given protein will always be found at the same place; the map thus provides direct information on whether, and if so in what quantity, a particular protein is present in a sample. If the protein of interest is found, the spot containing it is cut out of the gel. In the classical technique this is done by hand, whereas in proteomics this task is performed by robots. In order to permit comprehensive proteome comparisons and highly precise analyses, however, it is desirable that all the proteins in a sample be analysed. Many of these molecules are present in such minute quantities that they cannot be rendered visible in the gel using conventional staining methods. Often, however, it is precisely such proteins that are of interest as potential targets for
Proteomics: seeing through the undergrowth
37
drugs. Moreover, many proteins occur in a number of isoforms, i.e. slightly different variants that lie extremely close together in the two-dimensional gel. In order to permit identification of these less common variants as well, two-dimensional gel electrophoresis is generally complemented by other separation methods.
For example, components of the sample to be analysed can be separated in advance by filtration or centrifugation. Also available are various techniques by means of which proteins can be presorted according to various characteristics – e.g. charge, size, shape or binding behaviour. And the 2-D gel electrophoresis itself can be further refined by limiting both of the separation factors that it employs, namely charge and size, to a certain range within which resolution is particularly high. And for the task of excising the spots, Roche researchers have now developed a grid-shaped tool that cuts the gel into 6000 tiny pieces, each of which can be analysed automatically. In the next step, the proteins to be identified are cut into pieces. Just as genome researchers split DNA into more easily sequenced fragments using nucleases as ‘molecular scissors’, protein researchers use proteases, another kind of enzyme, to cleave amino acid chains at precisely defined points. This results in a mixture of variably sized protein fragments known as peptides. Since proteases act in highly specific fashion – the enzyme trypsin, for example, always cleaves the chain after the amino acids arginine or lysine – the peptide mixture obtained is specific for each protein/protease pair.
Presorting increases resolution
In order to identify the cleaved protein, the peptide mixture is fed into a mass spectrometer. As its name suggests, this device is able to measure the mass (and thus the weight) of molecules. A peptide mixture to be analysed is embedded in a carrier material – a matrix – and subjected to a laser impulse. The matrix transfers the energy of the laser to the peptides, which are thereby ionised, i.e. charged, and vaporised. The charged peptides are now accelerated by a powerful electric field and fly through a flight tube. Small peptides fly faster than large peptides. The time that they take to reach the end of the tube therefore indicates the size of the pro-
A fingerprint for every protein
38
The tiny difference: peptide fingerprinting
UV laser
+
powerful electric field
-
laser beam + ++
+ + +
+
amplifier data analysis in computer matrix peptides
ionisation
paffle plates flight tube (vacuum)
acceleration
In MALDI-TOF MS (matrix-assisted laser desorption ionisation time-of-flight mass spectrometry), the protein sample to be investigated is digested with a specific protease and the resulting peptide mixture is embedded in a matrix in a mass spectrometer. The energy of a laser impulse applied to the matrix is transferred to the peptides, which are thereby ionised and vaporised. The charged peptides
time of flight
detector
measurement
analysis
enter a flight tube along the length of which a powerful electric field is applied. The peptides fly in the direction of the electric field until they reach a detector at the end of the tube – the lighter molecules arriving first, the heavier molecules last. The time taken by the peptides to fly through the tube is measured.
tein fragments. This technique bears the rather cumbersome name of ‘MALDI-TOF MS’ (matrix-assisted laser desorption ionisation time-of-flight mass spectrometry). The result of this measurement is like a fingerprint of the protein under investigation: a distinctive spectrum on the basis of which the molecule can be identified beyond doubt. In fact, protein researchers actually refer to this process as ‘fingerprinting’ and compare their work with forensic science in which fingerprints are used to identify criminals. And just as police compare fingerprints obtained at the scene of a crime with those in their files, protein researchers search their databases for the protein that fits the spectrum they have obtained. To do this, they do not even need to have tested the proteins concerned, since nowadays all the proteins in a database can be exposed to a given protease
Proteomics: seeing through the undergrowth
39
MALDI-TOF spectra with corresponding proteins
1500 2000 2500
3000
3500
4000
4500 5000
1500 2000 2500
3000
3500
4000
4500 5000
Signal broadening caused by carbon isotope
1500
1500 2000 2500
3000
3500
4000
4500 5000
1500 2000 2500
3000
3500
4000
4500 5000
MALDI-TOF MS yields a spectrum – a fingerprint – that is specific for the digestant and the protein present in the sample. Each peak represents a signal of a certain strength obtained at a certain time. The fingerprint can be reproduced at will and can be calculated in virtual fashion for the proteins in a database. The spectrum thus obtained can be used for direct identification of the source protein.
2000
2500
3000
3500
4000
Nevertheless, simple database comparisons of this kind account for only a small part of protein research. Only if the protein under investigation is not found in one of the many databases, and is therefore considered to be possibly unknown, does the really exciting part of the task begin: the work to identify the form and function of the
40
5000
A peculiarity of mass spectra is the broadening of the signal that occurs as a result of the natural occurrence – to an extent of about 1% - of the heavy carbon isotope 13C. With larger peptides there is a high probability that at least one carbon atom in the molecule will have this greater mass. The presence of this isotope broadens the signal, and the resulting strong, broad signal can obscure a weaker signal. Nevertheless, the resolution of the technique can be considerably increased by use of mathematical methods and in particular by use of extremely sensitive spectrometers. The accuracy of modern MALDI-TOF mass spectrometers is rated at about 10 ppm (parts per million).
and subjected to mass spectrometry in virtual form in a computer. To remain with the analogy of forensic science, this is equivalent to a suspect’s fingerprint being worked out from his photograph – a possibility that would presumably make every police officer in the world green with envy!
Form and function go together
4500
13C
new molecule in the hope of deriving some medical benefit from it. For every newly discovered protein is potentially a new target within the molecular network of our body. Determination of the form, i.e. the amino acid sequence and the external shape, of the protein is easier to automate than determination of the function of the protein.Here again,the sequence of amino acid units can be determined in a mass spectrometer; in this case the peptides are removed from the end of the protein and the individual units are determined one by one. Once the amino acid sequence of the protein is known, conclusions can be drawn as to the shape of the whole molecule – though only to a very limited extent. For even though an amino acid chain basically arranges itself into a shape, the shaping of proteins within cells often occurs via highly complex pathways. Many proteins are assembled with the aid of enzymes that hold on to a part of the polypeptide chain while the rest is being formed. In the absence of these ‘molecular chaperones’ the units orientate themselves incorrectly. This process of protein folding is an important area of research, including medical research. For example, incorrectly folded proteins are responsible for the feared brain degeneration of Creutzfeldt-Jakob disease. In this case there appears to be a chain reaction in which incorrectly folded proteins are able to induce pathological changes in the structure of correctly folded proteins.
Until a few years ago determination of the spatial configuration of a correctly folded protein was a task worthy of a doctorate. Now, however, proteome research has largely automated this aspect of science and thereby made it possible to undertake a comprehensive structural analysis of the human proteome. This applies in particular to x-ray crystallography, the most important method for investigating the shape of small to medium-sized proteins. In this technique proteins are cultured en masse in host organisms, mostly bacteria or yeasts, and then crystallised. Previously, culture of crystals was regarded almost as an art form, mastered only by a few specialists with ‘golden hands’. Nowadays it is performed automatically, the only task left to scientists being that of selecting suitable specimens from the mass of cultured crystals. The tiny crystals – at least a twentieth of a millimeter of edge length is required – are then bombarded with x-rays. On their way through the crystal the x-rays are deflected (diffracted) by the protein
Automated structural analysis
Proteomics: seeing through the undergrowth
41
Function follows form: the enigma of protein structure The structure of a protein determines its function. Thus, muscle proteins are fibrous, membrane channels are tubular and enzymes are mostly rounded with one or more depressions into which their substrate fits. Shown left is the drug Herceptin, an antibody which is able to bind via the yellow-highlighted regions to a protein that causes breast cancer. Knowledge of the structure of the target protein is important, since it permits the development of custommade drugs and thus obviates much trial and error. Although the shape of a protein is ultimately determined by its amino acid sequence, it has so far proved almost impossible to predict the structure of a given protein.
molecules, and from the pattern of this diffraction scientists can work out the three-dimensional form of the molecule.
By far the most difficult part of the analysis of a new protein, however, is identification of its function, i.e. its place within the molecular network of the body. Here again, though, technological advances are providing more and more possibilities. These include protein chips: millimeter-sized silicon wafers on which thousands of different proteins are placed. If the new protein binds to one of the molecules on the chip, it can be assumed that it also binds to this molecule in the living cell. Proteins, however, are notoriously unfaithful: the unions they enter into are generally neither exclusive nor enduring. At another time, at another place, under different circumstances they will gladly try out something else – with other proteins, with other molecules or even with DNA. Within each and every cell of our body, therefore, is a complex network of interactions and conditions which the human brain cannot take in at a single glance. Fortunately, computers can perform this task for us. Bioinformatics is the science that attempts to bring order into this chaos. It also creates an important link between genetics, genomics and proteomics, since it combines the data from all
Fishing for the right partner
42
three of these disciplines. Which gene goes with which protein? Under what circumstances is this protein formed, and why? What genetic signals order the production of this protein? Which proteins help regulate such an order? And where are the genes for these proteins located? Genes and their products, proteins, are inseparable. This has long been known to science, and proteomics is therefore a necessary extension of genome research.
Cooperation between these disciplines is correspondingly close. Proteomics generally blends seamlessly with genetic and genomic experiments, whose conclusions it checks and extends. Transcriptome research, which deals with messenger RNA, the working copies of our genetic material, is likewise often combined with experiments conducted at the level of proteins. In order to promote this cooperation and especially in order to consolidate the efforts being undertaken in proteome research throughout the world, proteome researchers have now formed a worldwide body – the Human Proteome Organisation, abbreviated as HUPO – analogous to that formed to promote genome research. This organisation aims to undertake a variety of tasks: z Awareness. As compared with genome research, proteome research is still scarcely known to the general public. Its tasks and objectives, and especially its importance, therefore need to be publicised. The ultimate objective of HUPO in this regard is to achieve acceptance of the proposition that the human proteome merits at least as much support and as many resources as the Human Genome Project. z Coordination. In its latter stages, the decoding of the human genome developed into a much-publicised race between the publicly financed Human Genome Project and the private company Celera Genomics. This unexpected competition certainly resulted in the objective of the project being achieved earlier than had been expected. On the other hand, it also resulted in a lot of work being duplicated – a massive waste of research capacity that HUPO hopes will not be repeated in the case of proteomic research. z Protein catalogue. By analogy with the sequencing of the human genome, HUPO aims to develop a comprehensive protein catalogue with the aid of which, for example, potential genes identified in the Human Genome Project can be assessed.
HUPO - Human Proteome Organisation
Proteomics: seeing through the undergrowth
43
First successes: new signalling pathways in cancer
Dozens of signalling pathways are involved in the development of the various forms of cancer. Each newly discovered pathway provides further potential targets for medical intervention. Proteomics can help in the elucidation of such signalling pathways.
control ICE3
ras-transformed mouse lymphocytes
prevents the cell from undergoing transformation into a cancer cell. In order to be able to exert this control function, ICE3 is split in healthy cells by an enzyme known as granzyme B to form apopain. The 2-D gel of the cancer cells shows a large amount of unsplit, i.e. inactive, ICE3. From a separate experiment with gene chips it is known that the genetic instructions for the production of granzyme B are absent from the cancer cells. From this combination of findings the signalling pathway that operates here can be deduced: Signalling pathway diagram ICE3
HMG2
ras
BTF3a The 2D gel on the left shows a subset of the proteome of a normal cell; that on the right the same subset of the proteome of a cancer cell. The differences are readily apparent. - A protein referred to as ICE3 is present in greater quantity in the cancer cell. - The amount of the high mobility group protein HMG2 is reduced in the cancer cell. - The transcription factor BTF3a is formed in greater amounts in the cancer cell. ICE3 plays an important role in programmed cell death, or apoptosis. This ‘suicide’ of body cells occurs when the genetic material of a cell is severely damaged; in this way it
44
granzyme B
X
apopain
apoptose
uncontrolled cell growth
The two other proteins whose amounts are altered in the cancer cells likewise play a role in cancer. HMG2 binds to deformed DNA. This protein appears to have largely disappeared from the cytoplasm of the cancer cells; this is evidence of genetic damage to the cells. BTF3a is likewise a DNA-binding protein, however its function is to ensure correct transcription of genes. BTF3a had previously been shown to be present in increased amounts only in intestinal cancer cells. This protein therefore has potential for use as a tumour marker.
Proteomics is nevertheless confronted by problems. One of the most important of these is a direct consequence of the central characteristic of the proteome, namely its complexity. The total number of proteins in the human body is now known to be many times greater than the number of genes. There are estimated to be about 100,000 different proteins per cell type in our body; some of these are present in almost all cells, whereas others occur only in a small number of cells. The Roche database, for example, currently contains over 150,000 mass spectra, corresponding to around 4500 proteins. From this it is clear that a great deal of work remains to be done. It is also likely that many proteins, and in particular many modifications, i.e. subsequent alterations to proteins, will prove to be extremely rare. Unlike the situation with the human genome, it will be difficult ever to claim to have catalogued the human protein completely. In fact, this will in all likelihood be impossible, since failure to find any new proteins in a particular period most certainly does not mean that no proteins remain to be discovered (for example, our body is constantly producing new antibodies in response to antigens). This applies even more to protein interactions with other proteins and with other components of the body – such as genes. Only potent interactions of this kind are easy to find. Another problem that confronts proteome research is already well known to its ‘big brother’, genome research: the problem of patenting. As with genome research, it is to be expected that courts of law will eventually determine what can be protected, and how. For as long as uncertainty prevails in this area, however, every publication constitutes a risk for a researching company. The fact that this problem is well known from genome research will at least prove helpful in this regard.
The enormous challenge of proteomics
References Langen H: Proteomics as a new field in biology: applications and potentials. Presentation at Roche Roundtable on Genetics and Genomics, May 2000 Screening – Trends in Drug Discovery: Future Trends in Proteomics. Interview with Hanno Langen in: Screening, 2/2001 Brauckmann B: Protein analysis helps in the evaluation of new drugs. Roche Facets No. 14, 2000 Human Proteome Organisation – Website: http://www.hupo.org/
Proteomics: seeing through the undergrowth
45
Targets for medicine
Without targets there can be no drugs. Small wonder, then, that targets are the most sought-after and fought-over objects in medical research. New methods make the search for targets faster, safer and more effective and thereby lead to advances in medicine, since every new target is another ray of hope in the fight against diseases.
They are the ‘stars’ of medicine. Everywhere they are sought after, pursued and adored. Whenever they are found, they become an object of feverish research. Scarcely any other area of biological research is being pursued with such financial backing and intellectual effort as the search for ‘targets’ for new drugs. This is only to be expected, since it is in drug targets that pharmaceutical research has placed its greatest hopes for new, safer, more efficient and more effective therapies. In its broadest sense, a pharmaceutical ‘target’ is any molecular site within the body that is potentially susceptible to attack by drugs. Most such targets are proteins, though other biomolecules such as DNA, RNA, sugars and fats also have potential as targets for drug action. Common to all such molecules is their key role in metabolic processes and thus their importance for bodily function – and malfunction. Expressed in another way, targets often play a role in the development of disease. And that is what makes them such interesting objects of research.
Because of the multiplicity of their functions and properties, proteins are by far the most important targets for drugs in the body. They play a role in the development and progression of almost all diseases. Since their correct function is directly related to their form, one of the fundamental requirements of a drug is that it be able to distinguish between the correct and incorrect forms of a target molecule. Disease can also be caused by an excess or deficiency of a protein, or by its occurrence at the wrong time or in the wrong place. Since proteins, even when in the body, participate in a wide variety of chemical reactions and interactions, it is relatively easy to influence their actions by means of drugs. Far more difficult a task is to specifically influence only a certain action of a certain protein. Almost all currently used drugs influence the molecular network of the body at the level of proteins.
Targets in the narrow sense: proteins
DNA, the substance that bears our genetic information, controls the vast majority of bodily processes and lays down the framework for the body’s reactions to the environment. Many diseases are due to genes, though in most cases external factors also play a role. Malfunction of a gene can be due to an alteration in the sequence of
Important cause of disease: DNA
48
its building blocks. Alternatively, the instructions issued by a gene may be too strong or too weak, or may be carried out at the wrong time or in the wrong place. Medically important genes are often referred to as target genes. Since substances that directly influence DNA are difficult to find, most drugs act either on gene products or on molecules that regulate, serve or process genes – and most such molecules are likewise proteins.
RNA has a number of functions in the body, though its full role has yet to be elucidated. One of its principal functions is to act as a blueprint for the translation of genes into their products. Messenger RNA (mRNA), the form of RNA that has this function, is singlestranded and can be blocked by RNA with the complementary nucleotide sequence (‘antisense’ RNA). This possibility is already being exploited in research as a rapid – and rapidly reversible – means of switching off particular genes. RNA molecules can also act as enzymes, the best-known example of this function being ribosomes. These complex structures consisting of RNA and proteins are responsible for the synthesis of proteins. RNA molecules also play a key role in the processing of mRNA. And finally, the recent discovery of small nuclear RNA (snRNA), the role of which is still largely unclear, has opened up an exciting new field of research. RNA forms a far greater variety of structures than does DNA and in this sense resembles proteins. This makes it more accessible to drugs than are genes. So far, however, little is known of the role of RNA in disease.
Much more research required: RNA
Other substances that occur in the body are also potential targets for drug action: z Sugars are found, among other places, on cell surfaces, where they serve as markers and permit mutual recognition. They can assume many different forms and are being intensively investigated at present. z Fats not only form a large part of cell membranes, but also serve as hormones, antioxidants and much more besides. They are relatively small molecules that can assume very different forms. z All metabolites, that is to say the starting substances, intermediate products and end-products of our metabolism, are theoretically susceptible to influence by drugs.
Mostly minor roles: sugars, fats, etc.
Targets for medicine
49
insulin
extracellular intracellular p91
*
p60PIK
* *
PI 3'kinase p85α p 110
?? ISSK
* *
IRS-1
ISK RAS
dynamin
Shc SHPTP2 (SYP)
*
*
SOS
B-RAF
GRB 2
*14-3-3 RAF-1
Jun Fos
p70s6k
??
MAPK
* p90Rsk p70s6k GSK3
* PP-1 phosphorylase kinase inactivation
glucose transport
glycogen synthase activation
*
CRE's
PLO
* *
protein synthesis
PKCβ
*
PKC ζ 40S
cGI-PDE (type III) translocation
PKa
MEK
NFr B translocation
transcription factor activation
GLUT4 translocation
??
glycogen deposition
Molecules that play key roles in metabolism are potential drug targets.
With the exception of sugars, such molecules offer medicine only very nonspecific targets for drug action. Moreover, most of them are formed and broken down very rapidly in the body and play only a minor role in metabolism as compared with proteins, in particular. Under some circumstances, however, it can be useful to bind a specific intermediate product of an undesirable metabolic pathway and thereby block production of the end-product of that pathway. Even the rapid turnover of such targets can be advantageous if, for example, the action of a drug needs to be very rapid in onset and brief.
If diseases are to be combated more effectively, targets must play a central role – after all, every form of medical therapy is directed against some kind of target. Two basic possibilities suggest themselves in this respect: either existing therapeutic methods can be improved, or completely new methods of treatment can be developed.
Improve or invent
50
βy NSF SNAP RABs ARFs Ga, βy SNARE`s
7TM
In innovative improvement, the target remains essentially the same, but attempts are now made to influence it by different means, i.e. by developing new drugs. To this end, the new drug must be better adapted to the molecular structure of its target. The development of improved therapies therefore requires a sound knowledge of the target concerned. By contrast, speculative target research aims to develop completely new methods of treatment, i.e. to find new molecular targets for drugs. As compared with innovative improvement, this approach calls for greater financial investment and more extensive scientific input while at the same time having a greater risk of failure: since no pharmaceutical experience is available with them, presently unknown targets may, after prolonged and expensive research, prove unsuitable or too difficult to influence. On the other hand, since agents developed in this way have the potential to bring major advances in therapy, research of this kind can have huge consequences both for manufacturers and for patients. Most of the hopes placed in new targets are located in the field of speculative research. However, it is becoming increasingly difficult to find suitable targets, since most of the molecules that play a role in the development of the major diseases that afflict mankind are probably already known. The economic and scien-
Great risk, great benefit: speculative target research The search for completely new targets has a high failure rate. In many cases a great deal of intensive research has to be conducted before it becomes clear whether a newly discovered target molecule has anything at all to do with a disease, and even then the importance and general role of the molecule in the body's metabolism has to be elucidated. If a potential target is found to play only a minor role in the genesis of disease, active agents developed to act on it are unlikely to be of therapeutic value; if, on the other hand, a potential target is found to play a major role in other important metabolic processes, the undesirable function is likely to be difficult to block precisely and specifically. At the same time, however, speculative target research has the potential to deliver considerably greater therapeutic advances than can be achieved by simple improvement of familiar routes of attack. For example, previously unknown targets can:
z provide a better, or simply a different, way of interfering
z z
z
in a disease process and in this way lead to the development of new, and possibly more effective or better tolerated, drugs; be closer to the molecular cause of a disease than presently known targets and thereby provide a more specific and surer method of attack; play a role in production of the symptoms, rather than the cause, of a disease and therefore also have potential for use as a therapeutic target in other diseases with similar symptoms; be used as targets in diseases that have hitherto been regarded as essentially untreatable due to a lack of any suitable target; these include various types of cancer, since cancer is often due to a variety of causes that need to be treated individually.
Targets for medicine
51
tific outlay required for the development of completely novel therapies is therefore becoming ever greater.
The profusion of techniques now used to find and evaluate new drug targets are derived from a wide variety of scientific disciplines including proteomics and genomics, genetics, molecular biology, chemistry, physics and even informatics and information technology. In addition, scientists make use of results obtained in other areas of research. The various scientific approaches pursue different goals and to some extent build on each other’s findings. Research in this field includes not just the search for new targets, but also analysis of the function of these targets, evaluation of the potential usefulness of targets and the search for suitable sites of attack within targets. For example, some techniques are directed exclusively toward finding new biomolecules without regard to the biological function of these; others are used to elucidate the molecular basis of diseases; others to investigate the properties and functions of molecules that play a role in disease; and yet others to investigate ways of influencing newly discovered biomolecules and ultimately to find optimal remedies, i.e. drugs, for diseases. Most work, however, is conducted at a number of different levels of target research.
Successful search at many levels
The area of drug research that has been most important historically deals with the most important group of targets, namely proteins. Originally, most such work was aimed at identifying the sites and mechanisms of action of new drugs manufactured by chemists. To this end protein biochemists observed the properties of proteins that reacted with active substances and the interactions that resulted from these reactions. As this field of research revealed more and more about the molecular processes that take place in our body, speculation began about the molecular basis of diseases – and this provided drug research with new targets. Today, protein biochemistry is as much concerned with the search for new targets as it is with the evaluation of targets and the choice of suitable agents. This situation has arisen because it is now possible – thanks to rapid progress in unravelling the structure of proteins (e.g. by means of x-ray crystallography
Proteomics and classical protein biochemistry
52
• proteomics • databases • SNPs • gene chips • PCR • transg. mice •
unknown molecule
• e.g. protein
• proteomics • databases • • • • •
structural investigation
• protein structure • possibly prediction of structure
• • databases • SNPs • gene chips • PCR • transg. mice •
genetic background
• associated gene • transcript • regulation
• proteomics • databases • • gene chips • • • HTS
properties
• physical • chemical • biological
• proteomics • databases • SNPs • gene chips • • transg. mice •
• proteomics • databases • SNPs • gene chips • • transg. mice • HTS
function
• e.g. enzyme, hormone, structural protein
Modern target finding and evaluation is supported by a wide variety of methods. The techniques employed vary greatly in technical complexity and cost and often
molecular environment
• associated metabolic pathway • complexes with other proteins • when, where, how, how much
• proteomics • databases • SNPs • gene chips • • transg. mice •
• proteomics • databases • • • • transg. mice • HTS
selection
drug finding
• attack which target where?
• chemical synthesis • test physical, chemical and biological properties
• • databases • • • • transg. mice •
drug evaluation
• preclinical and clinical studies • marketing authorisation procedure
yield useful findings only when their results are considered together. Most techniques are used – in more or less modified forms – at various levels.
and mass spectroscopy; see chapter on proteomics) – to make predictions as to the required properties of new drugs. As far as the search for targets is concerned, great hopes have been raised by the discipline of proteomics, which aims to catalogue and study the entire complement of proteins of an organism. Already, a large number of previously unknown proteins – and thus potential drug targets – have been discovered, their composition and form elucidated and their function described.
A similarly systematic method of searching for new targets can now be conducted in the complete absence of any biological system: many drug targets are now being searched for and evaluated in the worldwide network of research computers. Interconnected databases are a veritable treasure-trove of information that can obviate the need for many lengthy and costly experiments. If, for example, the composition, that is to say the amino acid sequence, of an unknown protein is known, comparison with known mole-
Databases and the Human Genome Project
Targets for medicine
53
cules of similar structure provides important information on the likely properties, and thus also on the function, of the protein concerned. One of the largest sources of data in this field is the sequence of the human genome, which is now known thanks to the work of the Human Genome Project. A frantic search for genes relevant to disease, and thus for targets for new drugs, is now being conducted among the three billion items of genetic information that make up this treasure-trove. Computerisation of biological research has now developed into an independent scientific discipline and career path known as bioinformatics.
Single nucleotide polymorphisms, or SNPs (pronounced ‘snips’) have become increasingly important in recent years. These randomly occurring variations in single DNA subunits are transmitted from generation to generation and are considered to be responsible for many medically important phenomena (see chapter on SNPs) including intolerance, side effects and variations in the effectiveness of drugs. They also play a role in the development of many diseases. Thus, the consistent occurrence of specific SNPs in patients with certain signs or symptoms suggests that these SNPs interfere with the function of disease-related genes. Those genes can then serve as potential targets for drugs. As with proteomics and genomics, the systematic search for SNPs is now giving rise to extensive databases.
SNPs – single nucleotide polymorphisms
The basis for another important technique used in target research was established by the computer industry: several hundred thousand different DNA fragments can now be accommodated on a glass or synthetic chip measuring just 1.5 cm square. This makes it possible, for example, to find genes containing segments with an absolutely specific sequence (see chapter on DNA chips). DNA chips can also be used to investigate the enormous variety of messenger RNAs present in a cell, i.e. the ‘transcriptome’ of a cell. In a simple experiment of this type it is possible to determine whether a gene is transcribed at all in a given tissue under a given set of conditions. In this way the transcriptome too can provide information on potential targets for drugs; moreover, mRNA itself contains possible targets.
DNA chips, the transcriptome and protein chips
54
Virtual laboratory: bioinformatics For a long time now, the work that researchers do in their computers has been at least as important as the work they do in their laboratories. Modern experiments, especially if automated, yield an unimaginable amount of data that could not possibly be analysed without the aid of powerful computers and specially developed computer programs. Bioinformatics, the application of modern information technology to biological research, has thus developed into an independent scientific discipline with a broad range of applications, including the following: z Sequence analysis is the original and core activity of bioinformatics. For example, special programs have been used to sift through the three billion building blocks that make up our genome in order to identify genes and regulatory elements. Such programs can also perform sequence analysis of proteins and translation in both directions (from gene sequence to protein and vice-versa).
z DNA chip experiments provide amounts of data that
z
z
previously would have been acquired only after years of work in major research institutions. In order to permit evaluation of such data within a reasonable time, bioinformatics specialists are constantly developing new programs and databases. Prediction of the structure of proteins is a relatively new and controversial subfield of bioinformatics. Modern supercomputers attempt to predict the three-dimensional form of proteins on the basis of their amino acid sequence. Use of databases and networking of data and laboratories are basic prerequisites for smooth performance and evaluation of experiments. Known information must be accessible everywhere at all times and must be available in a form that is compatible with radically different computers, operating systems and programs.
In theory, the DNA chip technique is also suitable for the investigation of proteins, since in principle any molecule can be investigated on the silicon surface of chips. For this purpose, however, proteins – like DNA – have to be attached to, or ideally synthesised on, the substrate, and this is vastly more difficult to achieve with sensitive, highly complex molecules such as proteins than with a relatively stable and structurally simple molecule such as DNA.
Vanishingly small amounts of DNA can be rendered visible by means of the polymerase chain reaction (PCR). With the aid of the DNA-extending enzyme polymerase, a single DNA molecule can be copied in a chain reaction (‘amplified’) as many times as desired and thus made available for investigation. This technique has made it possible, among other things, to detect the presence of minute amounts of the genetic material of HIV, the AIDS virus, and in this way identify variants of this deadly infectious disease. If a drug is to be used against one of the few, but highly variable, products of viral RNA, PCR can be used to identify the specific variant of this target molecule present in a patient and then select the most appropriate drug. In future PCR will therefore
PCR – polymerase chain reaction
Targets for medicine
55
High-throughput screening system.
become of great value for investigating and classifying targets and for determining the exact concentrations of these in patients (see chapter on PCR).
‘High-throughput screening’ is an important aid in the search for suitable drugs for a given target. In this technique thousands of chemical substances (for example, modifications of a computer-designed molecule intended to fit into the binding site of a protein variant that causes disease) are automatically tested for certain properties (such as binding to the intended target). The data obtained are analysed by computer and candidate substances undergo a further round of improvement. This continues until a molecule with optimal characteristics has been found. Highthroughput screening now performs reliably, rapidly and cheaply a task that only a few years ago required an enormous amount of work on the part of hundreds of biologists, chemists and physicists.
High-throughput screening (HTS)
56
Animal models
What use is a target about which nothing is known? Among the bewildering amount of data provided by genomics and proteomics are many hitherto unknown biomolecules, however these are not (yet) targets. In order to qualify as targets, such molecules must not only play an important role in the body, but also make a significant contribution to the genesis of disease. Unfortunately, however, the function of an unknown gene or protein is not easy to discover, especially as it can vary with environmental conditions and with time and place. In complex diseases such as cancer, Alzheimer’s disease and diabetes it is even more difficult to determine the role and importance of a potential target in the molecular processes of disease – especially as such diseases can appear early or late, occur in more or less severe form, have many causes and show different signs and symptoms. Since it is impossible to identify and characterise all the factors that can influence a complex disease, researchers make use of animal models to investigate such diseases. The effects of specific changes – especially gene variants – on the health of animals are observed, since the environmental factors that operate in such models are known and can be controlled. Different animals have proved useful for investigating particular diseases and addressing specific questions. For example, a large part of our knowledge of human embryonic development and the molecular basis of cancer was obtained via studies on fruit flies and threadworms, since it was in these animals that the genes that play a crucial role in these processes were first found and investigated. In the case of applied research into how disease develops, the most important model is the mouse.
Depiction of a target molecule with bound drug.
Targets for medicine
57
Humans and mice share 99 percent of their genes. This means that Homo sapiens and Mus musculus each possess only about 300 genes that are not present in the other. Scientific findings obtained in mice can therefore be extrapolated with relative ease to humans. The mouse is thus an almost ideal experimental animal, especially as it is easy to rear and breed in a laboratory. For decades, therefore, mice have played a central role in research into diseases. Moreover, in the year 2002 the Mouse Genome Project was brought to a successful conclusion, permitting direct comparison between the human and the mouse genome. Of considerable importance for the role of mice in modern medicine were experiments conducted in the early 1980s in which foreign material was for the first time successfully inserted into the germ line of mice. This can now be done with any human gene. Animals into which a small amount of foreign genetic material has been inserted are known as ‘transgenic’. They are particularly valuable for medical research purposes because the effects of genes on the development of disease can be directly studied in them. In transgenic mice – in contrast to cell cultures or cell extracts, for example – the entire molecular environment of a disease, i.e. the totality of factors that can promote or inhibit the development of a disease, is present. The effects of medical interventions are therefore very similar to those that actually occur in the human body. It is also possible to selectively remove certain genes from the genome of a mouse. Observation of these ‘knockout’ mice (as opposed to ‘knock-in’ mice with additional genes) then reveals the consequences of absence of the gene concerned, and this in turn permits conclusions as to the function of the gene. These two techniques (knock-in and knock-out) can be combined so as to replace a gene with a different one. If the second gene is a variant of the first, the effects of the variation can be directly observed. Transgenic mice are used above all to investigate the molecular basis of diseases, and this work is resulting in the discovery of more and more extremely important targets. Genetically modified laboratory animals are also being used to study targets discovered by other means and, not least, to evaluate potential drugs. In fact, experiments on transgenic mice can often replace those on other animals such as monkeys, and even studies in humans, in the early phases of drug development.
Transgenic mice as a model
58
Important objects of research: transgenic mice Transgenic mice are among the most important tools of molecular medicine. These animals have been genetically modified either by insertion of additional – mostly human – genes into their genome or by selective removal of certain of their genes. Genes can also be altered or else removed and then replaced by others. All such types of transgenic mice are excellent models for investigating the influence of genes and environmental factors on the development and progression of diseases. The breeding and use of animals for pharmaceutical research are strictly regulated in all European countries and in most cases are under the control of independent committees that include representatives of animal protection groups. Two techniques are used to insert foreign genes into animals: 1) The technique of microinjection into the pronucleus of fertilised ova was introduced in the early 1980s and has now been performed successfully in a number of animal species. A solution containing many copies of the desired gene is injected into the maternal or paternal pronucleus (undeveloped cell nucleus) of a fertilised ovum. Copies of the gene are randomly incorporated
into the genome of the resulting zygote. The ovum is then implanted into the uterus of a female animal. About a quarter of the offspring produced in this way contain the desired gene in their genome and subsequently transmit it to their own offspring. This method can be used only to add genes to an animal’s genome. 2) By contrast, the technique of microinjection into the blastocyst can be used to specifically modify or eliminate an animal's own genes. To date the only mammal in which this technique has been performed successfully is the mouse. This technique employs embryonic mouse stem cells – i.e. cells that still have the potential to develop into any kind of cell – whose genome has been altered in the desired way by means of recombinant technology. These cells are injected into a mouse blastocyst (multicellular stage of a developing embryo). The embryo is then reimplanted into the mother, where it develops into a chimeric mouse, only a proportion of whose cells possess the desired genetic modification. Those offspring of such chimeric mice whose ovaries have developed from the inserted stem cells will bear the altered genes in all their cells.
Controlled mutagenesis: blastocyst injection
Microinjection of DNA
DNA embryonic stem cells unfertilised ovum
cell injection into the blastocyst
DNAmicroinjection
donor blastocyst
fertilised ovum
reimplantation in mouse prepared for pregnancy
reimplantation mouse prepared for pregnancy
chimeric mouse breeding
transgenic offspring
offspring with the desired mutation
Targets for medicine
59
The existence of so many technical possibilities does not make the search for targets in any way routine. All hopefulness notwithstanding, it remains true that a newly discovered biomolecule is by no means necessarily also a suitable target for drug research. As mentioned above, decades of research have in any case already resulted in the discovery of a great many targets. This applies in particular to molecules that are relatively common, long-lived and stable and that appear to play a major role in important diseases – molecules, in other words, for which the risk of failure in drug research is relatively small. The days of the gold rush in speculative target finding are therefore over. To continue with this metaphor, the large nuggets have all been found, the claims have been pegged out, whatever gold remains is buried quite deep below ground, and how much remains is more or less known already. The scientists of the Human Genome Project found that our genome contains ‘only’ about 30 000 genes – less than a third the number that had been hoped for: after all, even a lowly threadworm has 20 000 genes. On the other hand, it is now known that these genes can be transcribed and translated into proteins in many different ways – and every type of cell in our body has at least 100 000 different proteins, which is good news for target finding. Nevertheless, this number includes isoforms – the in many cases only slightly differing variants of a given protein – and protein complexes, which in some cases can form and disintegrate rapidly. Estimates of the number of actual targets among proteins therefore range from a few hundred to a few thousand – i.e. not all that many, whatever the exact figure.
The search becomes more difficult
Nevertheless, the search is worthwhile, because we still lack causal therapeutic options for use against far too many of the major diseases that afflict mankind, including cancer, cardiovascular disease, Alzheimer’s disease and diabetes. Even in the case of infectious diseases, medicine is at a very early stage in terms of the possibilities it has to offer – precisely because molecular methods have now made it possible to attack many pathogens selectively. Also unsatisfactorily treated to date are many diseases which only a few decades ago had not been investigated at all or else were incorrectly classified, such as allergies and autoimmune diseases. And finally, many rare diseases have become objects of drug re-
Weighing risks against benefits
60
search now that the existence of global markets and public funding has made the development of drugs for use against them economically viable for private companies. The decision as to whether, when and how a potential target should be further developed depends on a number of factors. First among these is medical benefit: what benefit will a new drug bring to victims of the disease it is intended to treat? In the case of treatments for previously untreatable diseases (or forms of disease) the answer is obvious, however even benefits such as substantially improved tolerability, greater potency or avoidance of side effects can justify the development of a new drug. Another important factor influencing the decision for or against a drug target is a scientific analysis of anticipated expenditure versus risks. In other words, what is the probability that an agent for use against the target can actually be developed, and what expenditure will be required in order to develop such an agent? Before these questions can be answered, certain details of the structure and function of the target must already be known – after all, it is pointless to direct efforts and resources to targets that are difficult to influence if the therapeutic objective might be more easily achievable by influencing other targets.
A third important factor influencing the evaluation of drug targets is economics. Every privately owned pharmaceutical company has to finance its research spending with income obtained from the sale of its products – without money there can be no research. Though this is fundamentally true of any company that develops new products, some particular considerations apply to the field of medicine. z Probably the most important of these – see above – is a duty of care towards patients. The high degree of responsibility borne by the healthcare industry in this regard inevitably influences economic decision-making. This gives rise to conflicts whenever a medically (or socially or politically) correct decision is clearly economically incorrect – in other words, when it is clear from the outset that the cost of developing a medically worthwhile drug can never be recovered via sales of the drug. Some such conflicts can be resolved via ‘orphan disease’ programmes in which the state provides financial and legal support for research into particularly rare diseases and the development of drugs for use against them.
Drug research must also pay for itself
Targets for medicine
61
z Another peculiarity of privately funded pharmaceutical re-
search is the composition of the ‘electoral college’ that participates in the decision-making process. This includes not just the company itself in the form of its management, research, finance and marketing directors, employees, investors and financial analysts, but also patients in the form of patient associations, legislators in the form of regulatory authorities, and the public in the form of the press and a variety of associations. z An additional peculiarity of the pharmaceutical industry is the sheer unpredictability of biology. All scientific and technical advances notwithstanding, interventions in biological systems are always subject to a high degree of uncertainty. Evaluation of targets is therefore often difficult. Even the fact that the total number of potential targets is unknown is important in that it has a substantial impact on the economic significance of each and every target. A newly discovered biomolecule can turn out to be useless after prolonged research, while years later it may suddenly acquire considerable value. The same is true of new drugs: whatever precautions are taken, a new drug may prove to be unsuitable or even dangerous; conversely, a completely new set of indications for a drug may suddenly be found. Since the biological system that forms the subject of medicine is the human body, such risks and uncertainties are of the greatest importance and the cost of researching and developing new drugs is correspondingly high. z Undoubtedly one of the most-discussed aspects of the economics of pharmaceutical research is the question of patents. While patent protection is essential for the survival of any economic sector that undertakes research and development, biological patents are subject to particular problems. Prominent among these is the fact that the distinction between a (nonpatentable) discovery and a (patentable) invention is often difficult to make in biology. It is also argued that overgenerous patent protection can inhibit further research, and thus medical progress. This view is propounded not just – as always – by competitors, but also by a variety of community groups, politicians and basic researchers representing to some extent conflicting interests. After some heated initial discussions and a certain amount of legal toing and froing, the legal and practical arrangements that have now been made, though still somewhat woolly, provide companies with a sensible degree of security for forward planning.
62
research director medical profession
financial analysts
target selection customer advocates
market group
drug hunter
regulatory agency
Many different constituencies with partly overlapping and conflicting interests influence target selection in pharmaceutical research.
The past few decades have seen a significant shift in the pattern of decisions on whether or not to proceed with research on a given target. This is because the development of new drugs has become riskier due to ever-increasing technical costs and a shortage of potentially good targets. Nevertheless, what ultimately determines a company’s success or failure is the ratio of risk to financial return. Thus, a venture with a high risk coupled with a low likelihood of (financial) success is to be avoided. Unfortunately, however, neither of these factors is easy to determine – there are many examples of important drugs developed at great expense that failed as a result of unforeseen problems, while conversely, the success of many ‘blockbusters’ (drugs with sales of over a billion US dollars per year) was not anticipated at the time of their development. Large amounts of money are therefore spent nowadays generating financial projections that attempt to take account of all conceivable risks and forms of success. None of these considerations affects the fundamental value of targets in medicine: without new targets, genuine progress is difficult to achieve. It is therefore with good reason that the word ‘target’, more than any other, embodies the hope for better, more rapidly acting, better tolerated and individualised therapies, since every new target is at the same time a diagnostic tool
Between risk and return
Targets for medicine
63
Good fortune: unexpected blockbusters active ingredient (product)
position in drug class
indication
original sales forecast
Tamoxifen (Novaldex)
1st
breast cancer
£ 100,000
Captopril (Capoten)
1st
hypertension, heart failure
USD 20 million
Cimetidine (Tagamet)
1st
Fluoxetine (Prozac) Atorvostatin (Lipitor)
peptic ulcer
£ 700,000
nd
depression
?
th
hypercholesterolemia
?
2 5
that can improve our understanding of the disease process in patients. We must therefore hope that the ‘pop stars’ of medicine continue to make headlines.
References Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations. Pharmacogenomics 1: 23–26, 2001 Knowles J, Gromo G: Target selection in drug discovery. Nature Rev, Vol 2, January 2003 Brauckmann B: From basic research with genetically modified mice to new forms of medical therapy. Roche Facets No. 14, May 2000 Ebeling M: Mit eigener Bio-IT am Forschungspuls. Internal Roche publication Human Genome Project – Website: http://www.ornl.gov/hgmis/ Geschäftsstelle des Wissenschaftlichen Koordinierungskomitees des Deutschen Humangenomprojekts (ed.): Das Humangenomprojekt – 1st and 2nd edition
64
PCR: an outstanding method
Scarcely any invention has altered biological science so radically in such a short period as the polymerase chain reaction, or PCR. With this technique, minute amounts of DNA can be replicated very rapidly and thereby amplified to such an extent that the DNA becomes easy to detect, study and use for any given purpose. The potential of this technique in medicine has long been known, and ever more applications are being developed. Wherever genes provide clues to the cause or natural history of a disease, PCR is the method of choice.
Long car journeys can sometimes be a godsend. Driving along a monotonous stretch of dark road one April weekend in 1983, American chemist Kary Mullis was struck by an idea that was later to earn him the Nobel Prize: the principle of the polymerase chain reaction. Among the instruments and glassware of his laboratory Mullis might never have had the most momentous and far-reaching idea of his life. Within a few years PCR – short for ‘polymerase chain reaction’ – took the world’s biological laboratories by storm. By the mid1980s the technique was used for the first time to diTerms agnose a disease, when researchers identified the gene DNA deoxyribonucleic acid; the chemical substance of our genes for sickle cell anemia. At RNA ribonucleic acid; the chemical substance that makes about the same time the up the working copies of genes (mRNA), among other things Nucleic acids a chemical term that covers both DNA and method was introduced inRNA; nucleic acids are molecules consisting of long chains of to forensic medicine. The nucleotides linked together polymerase chain reaction Nucleotides the building blocks of DNA; they comprise the four bases adenine, thymine, cytosine and guanine (A, T, C, reaped the highest scientific G; in RNA thymine is replaced by uracil [U]), a sugar and at honour for its inventor in least one phosphate group; without the phosphate group record time: In 1993, just these building blocks are referred to as nucleosides Sequence the order of the nucleotides in DNA (DNA seten years after his historical quence) or RNA (RNA sequence) car journey, Kary Mullis rePrimer a short DNA fragment with a defined sequence that ceived the Nobel Prize for serves as an extension point for polymerases Polymerases enzymes that link individual nucleotides toChemistry. The reason for gether to form long DNA or RNA chains this extraordinary success is Hybridisation (annealing) the joining of two complementary DNA (or RNA) strands to form a double strand that the technique provided Complementary DNA The building blocks of DNA and a solution to one of the most RNA form specific pairings. Two strands whose building blocks pressing problems facing form a sequence of perfect pairings are able to form a stable double strand and are referred to as complementary strands biology at the time – the replication of DNA.
In the PCR procedure trace amounts of DNA can be quickly and repeatedly copied to produce a quantity sufficient to investigate using conventional laboratory methods. In this way, for example, it is possible to sequence the DNA, i.e. determine the order of its building blocks. Theoretically, a single DNA molecule is sufficient. PCR is therefore one of the most sensitive biological techniques ever devised. Given these capabilities, Mullis’s method ultimately ushered in the age of genomics. From the Human Genome Project to the search for targets to the development of gene tests, there are few areas of
Rapid DNA cloning
66
genetic research today that do not depend on PCR. Only with the advent of increasingly sensitive DNA chips in recent years has PCR faced any notable competition (see chapter on DNA chips). But even then it is often necessary to first copy, or amplify, the DNA of interest. For this reason PCR and DNA chips often go hand in hand.
Simple and effective: the PCR principle
denaturing
polymerase Zielsequenz
extension
Polymerase-Kettenreaktion: PCR
DNS
getrennte DNS-Stränge (denaturiert)
hybridisation
1 2
repeat cycles …
3
The polymerase chain reaction serves to copy DNA. It uses repeated cycles, each of which consists of three steps: 1. The reaction solution containing DNA molecules (to be copied), polymerases (which copy the DNA), primers (which serve as starting DNA) and nucleotides (which are attached to the primers) is heated to 95°C. This causes the two complementary strands to separate, a process known as denaturing or melting. 2. Lowering the temperature to 55°C causes the primers to bind to the DNA, a process known as hybridisation or annealing. The resulting bonds are stable only if the primer and DNA segment are complementary, i.e. if the base pairs of the primer and DNA segment match. The polymerases then begin to attach additional complementary nucleotides at these sites, thus strengthening the bonding between the primers and the DNA. 3. Extension: The temperature is again increased, this time to 72°C. This is the ideal working temperature for the polymerases used, which add further nucleotides to the developing DNA strand. At the same time, any loose bonds that have formed between the primers and DNA segments that are not fully complementary are broken. Each time these three steps are repeated the number of copied DNA molecules doubles. After 20 cycles about a million molecules are cloned from a single segment of doublestranded DNA. The temperatures and duration of the individual steps described above refer to the most commonly used protocol. A number of modifications have been introduced that give better results to meet specific requirements.
4 5
PCR: an outstanding method
67
The basic PCR principle is simple. As the name implies, it is a chain reaction: One DNA molecule is used to produce two copies, then four, then eight and so forth. This continuous doubling is accomplished by specific proteins known as polymerases, enzymes that are able to string together individual DNA building blocks to form long molecular strands. To do their job polymerases require a supply of DNA building blocks, i.e. the nucleotides consisting of the four bases adenine (A), thymine (T), cytosine (C) and guanine (G). They also need a small fragment of DNA, known as the primer, to which they attach the building blocks as well as a longer DNA molecule to serve as a template for constructing the new strand. If these three ingredients are supplied, the enzymes will construct exact copies of the templates (see box on page 67). This process is important, for example, when DNA polymerases double the genetic material during cell division. Besides DNA polymerases there are also RNA polymerases that string together RNA building blocks to form molecular strands. They are mainly involved in making mRNA, the working copies of genes.
Copies of copies of copies
These enzymes can be used in the PCR to copy any nucleic acid segment of interest. Usually this is DNA; if RNA needs to be copied, it is usually first transcribed into DNA with the help of the enzyme reverse transcriptase – a method known as reverse transcription PCR (RT-PCR). For the copying procedure only a small fragment of the DNA section of interest needs to be identified. This then serves as a template for producing the primers that initiate the reaction. It is then possible to clone DNA whose sequence is unknown. This is one of the method’s major advantages. Genes are commonly flanked by similar stretches of nucleic acid. Once identified, these patterns can be used to clone unknown genes – a method that has supplanted the technique of molecular cloning in which DNA fragments are tediously copied in bacteria or other host organisms. With the PCR method this goal can be achieved faster, more easily and above all in vitro, i.e. in the test-tube. Moreover, known sections of long DNA molecules, e.g. of chromosomes, can be used in PCR to scout further into unknown areas.
Copying unknown DNA
68
Soon after its discovery the PCR method was refined in several ways. One of the first modifications of the original protocol concerned the polymerases used. Like all enzymes, polymerases function best at the body temperature of the organism in which they originate – 37°C in the case of polymerases isolated from humans. Below this temperature the enzyme’s activity declines steeply, above this temperature it is quickly destroyed. In PCR, however, the two strands of the DNA molecule must be separated in order to permit the primers to anneal to them. This is done by raising the temperature to around 95°. At such temperatures the polymerases of the vast majority of organisms are permanently destroyed. As a result, new enzyme had to be added in the first reaction step of each cycle – a time-consuming and expensive proposition. A solution was found in hot springs. Certain microorganisms thrive in such hot pools under the most inhospitable conditions, at temperatures that can reach 100°C and in some cases in the presence of extreme salt or acid concentrations. The polymerases of these organisms are adapted to high temperatures and are therefore ideal for use in PCR. Today Hot spring: Thermus aquaticus. Nearly all PCR techthe polymerases used in niques in the world now use the Taq polymerases isolatnearly all PCR methods the ed from the bacterium Thermus aquaticus, a microorworld over are derived from ganism that dwells in hot springs at about 70°C. The first such microorganisms. This T. aquaticus strain from which polymerases for PCR were prominent bacterium goes obtained was found in Yellowstone National Park in the by the name of Thermus USA. aquaticus, and its heat-stable polymerase, called Taq polymerase, supports an entire industry. The organism was originally discovered in a 70°C spring near Great Fountain Geyser in Yellowstone National Park in the USA. Employees of Cetus, who Kary Mullis was working for at the time of his discovery, isolated the first samples from the hot spring and then cultivated in the laboratory one of the most useful bacterial strains known today. Meanwhile Thermus aquaticus has been found in similar hot springs all over the world.
Help from hot springs
PCR: an outstanding method
69
The introduction of Taq polymerase has certainly not been the only modification to the PCR method. This was helped by the fact that Mullis published his discovery relatively early – though not without some difficulty. Both Science and Nature, the two most renowned scientific journals, failed to recognise the significance of PCR and rejected the paper describing the method. Moreover, despite global patent protection, the use of the PCR technique is still free and unrestricted for basic researchers thanks to Roche, which owns the rights to the method. In 1991 Roche obtained an exclusive license from Mullis’s former employer Cetus for 300 million dollars. Scientists from all over the world have modified the PCR method in many ways and adapted it for routine diagnostic testing and molecular research. At the same time, more and more new applications are emerging.
Further developments around the world
In the 1990s biology was faced with one overriding preoccupation: the unravelling of the genome. Thanks to huge technical and organisational efforts, first viruses and bacteria, then yeasts, plants and animals relinquished the secrets of their genetic material. This accomplishment would have been unthinkable without PCR, which made it possible to prepare large amounts of DNA within a short time. The simple cloning of DNA has therefore remained one of the main uses of the method. Thus PCR is used whenever the exact sequence of DNA building blocks needs to be determined: e.g. in other genome sequencing projects, in gene research, in the investigation of genomic changes, in the search for targets, etc. An important topic in the field of genomics today is SNPs (pronounced ‘snips’), single nucleotide changes in the genome which appear to account for a large proportion of the genetic differences between individuals (see chapter on SNPs). Among other things, SNPs are responsible for disease susceptibility and for differences in the way patients respond to drugs. In order to detect such hereditary and often widespread variations, scientists have to sequence the genome of many different people in parallel. Genes with SNPs are also potential targets for new drugs. PCR therefore plays a key role in this important area of drug research.
Forerunner of genomics: DNA sequencing
70
adenine thymine guanine cytosine
DNA, our genetic material.
When PCR is used only for detecting a specific DNA segment, the method is referred to as qualitative PCR. Usually the standard protocol is used. Qualitative PCR is an extremely sensitive method which is theoretically able to detect a single DNA molecule in a sample solution. In many cases specific genes are copied in this way, e.g. in order to identify pathological changes. As mentioned earlier, the first gene identified by PCR was the gene responsible for sickle cell anemia. Countless other gene tests have meanwhile been devised. Qualitative PCR is also used around the world in forensic medicine to identify individuals. Usually individual regions of the genome are amplified and examined. However, although these regions differ between people, they reveal nothing about the traits or character of the person in question. PCR can of course be used to detect not only human genes but also genes of bacteria and viruses. One of the most important medical applications of the classical PCR method is therefore the detection of pathogens. Here PCR is replacing immunological methods, in which antibodies against a pathogen are used to identify the pathogen in a patient’s blood. Antibodies are not detectable until several weeks after the onset of an infection, whereas PCR is able to detect the DNA or RNA of the pathogens much more quickly. Moreover, antibodies can remain in the bloodstream long after an infectious disease has resolved. Hence, only qualitative PCR can determine whether an infection
Sensitive determination: qualitative PCR
PCR: an outstanding method
71
has been eradicated, whether it is chronic (and might therefore progress unnoticed) and whether the individual has been reinfected with a different but related pathogen. Many viruses contain RNA rather than DNA. In such cases the viral genome has to be transcribed before PCR is performed, and RTHuman immunodeficiency virus. PCR is therefore used. Sometimes it is also necessary to detect pathogens outside the body. Fortunately, the PCR method can detect the DNA of microorganisms in any sample, whether of body fluids, foodstuffs or drinking water. PCR is therefore used in all these areas. One of the most urgent problems PCR is helping to solve is to determine if donated blood is contaminated. Blood banks are one of the major transmission sources of hepatitis C, for example, and sometimes of HIV. Fast, simple and above all inexpensive testing is essential – and PCR ideally meets all these criteria.
Quantitative PCR provides additional information beyond mere detection of DNA. It indicates not just whether a specific DNA segment is present in a sample, but also how much of it is there. This information is required in a number of applications ranging from medical diagnostic testing through target searches to basic research. Consequently, although quantitative PCR was not described until the 1990s, the method already exists in a number of variants and protocols to meet a broad range of requirements. Theoretically it is possible to calculate the amount of DNA originally present in a sample directly from the amount found at the end of a PCR run. If, for example, there were not one but two double strands at the start of the reaction, exactly twice as much will be present after each cycle. However, this simple approach founders on the fact that conditions for the polymerases are not optimal at the start or end of PCR. At the start the performance of the enzymes is limited by the small amount of template present, while in the final cycles the enzymes’ activity declines as a
More than just yes or no: quantitative PCR
72
result of continuous temperature changes. Moreover, in these later cycles the amount of available nucleotides falls and the newly formed templates increasingly bind to each other rather than to the primers. The effects of all these factors vary greatly depending on small differences in the reaction conditions (temperature, duration of the steps, concentrations of the reagents, etc.). In practical terms it is therefore impossible to draw direct conclusions about the number of molecules in the original sample from the amount of DNA present at the end of PCR. Instead, researchers have developed various methods that determine the number of new DNA molecules formed in the reaction, i.e. after each cycle (see box on p. 74). Because this approach affords continuous observation of the reaction, it is referred to as real-time PCR. Ultimately such experiments involve conjugating the new DNA copies (but not the primers or free DNA building blocks) to a dye, thus making it possible to determine the quantity of template.
Target research
Quantitative PCR is used, for example, to help search for and evaluate targets, i.e. the sites in the body at which new drugs can act. This primarily relates to the discovery of new genes, a task in which PCR is basically used as a DNA copying tool. The same applies to already known genes that come in a number of variants and that are fairly widespread in the population (polymorphisms; see chapter on SNPs). However, for a gene truly to be a target for new drugs, its products must be involved in the development or progress of a disease. The common occurrence of specific gene variants in affected individuals can only serve as an initial signpost. The question ultimately is not whether a specific form of a gene is present or not, but whether observed variations – i.e. changes in a gene sequence, multiple occurrences of a gene or its absence – really have different effects in healthy and ill people. To investigate this – a procedure known as target validation – we need to consider the gene’s products rather than the gene itself. Gene products are usually proteins. Protein research is therefore devising increasingly sophisticated methods to detect, identify and assay its subjects of enquiry (see chapter on proteomics). But the latter task, quantity determination, cannot be satisfactorily performed with available proteomic methods. Furthermore, proteins often differ markedly in their life cycle and activity. As a result, the quantity of a protein provides only lim-
PCR: an outstanding method
73
Shining examples: quantitative real-time PCR Many applications require the amount of DNA originally present in a PCR sample to be determined. Many techniques are available for calculating the number of DNA copies formed during the individual steps of the PCR procedure and thus for deriving the quantity originally present in the sample. They are usually based on the middle, or exponential, phase of the PCR in which the amount of DNA template is approximately doubled in each cycle. z Example: competitive PCR – This method is now largely of historical significance only. It was one of the first quantitative PCR methods developed. In addition to the template of interest, another DNA template having a very similar sequence was added to the same reaction vessel. Both DNA strands were then cloned simultaneously under identical conditions. The amount of template and the amount of ‘competitive’ DNA formed provided at least a rough estimate of the amount of DNA present in the original sample. z Example: real-time PCR – Most of the quantitative PCR methods in use today are based on a 1922 discovery by the American Russ Higuchi, who used the dye ethidium bromide (EtBr). Embedded in double-stranded DNA, EtBr fluoresces when stimulated by light. The observed fluorescence therefore indicates the amount of DNA formed and does so at any given time during the PCR reaction, hence the name real-time PCR. In this method parallel runs are performed with the same known quantity of DNA and a comparative curve is plotted under identical conditions. This presupposes that the sequence of the DNA to be copied is known. Moreover, it is not possible to distinguish directly between the correctly formed product and primers that have annealed to form
forward primer
forward primer
R
probe
R
probe
z
a double strand. Nevertheless, the principle is still used, though usually with other dyes that specifically interact only with the desired DNA product. The experiments have been simplified by the introduction of special equipment such as the Roche LightCycler, which automates the entire procedure, heating and cooling the solutions, stimulating the dye to fluoresce and continuously monitoring the fluorescence. Special computer programs help to analyse the data. Example: TaqMan probes: One way to measure only the desired DNA product during PCR is to use TaqMan probes, short DNA fragments that anneal to a middle region of the template DNA (see below). The probes bear a reporter dye (R) at one end and a quencher (Q) at the other. Quenchers are molecules that quench the fluorescence of dyes in their proximity. The polymerases in the PCR solution are able to break down the TaqMan probes during the doubling of the DNA template. In so doing they free the reporter dye, which then migrates away from the influence of the quencher. Hence the fluorescence of the dye is measurable only if the polymerase has in fact copied the desired DNA strand. Each freed molecule of reporter dye represents a DNA strand that has been formed. TaqMan probes can therefore be used to measure the amount of DNA formed at any given time.
Q
Q
R probe forward primer
74
Q
ited information about the expression of the corresponding gene. A far more sensitive method is available in the form of quantitative PCR, which measures not the proteins themselves but rather the working copies of the corresponding genes, the mRNA. In accordance with the principle of RT-PCR, the mRNA is reverse-transcribed into DNA. Its original quantity is then determined by quantitative real-time PCR. This provides an informative picture of how vigorously a gene is being transcribed and in what form, since in many cases one and the same gene can give rise to different products at different sites in the body. The initial working copies of such genes are cut and stitched back together by the cells in various ways. Consequently, the products can have markedly different properties.
Example: STEP
One method of determining the quantity and nature of working copies of genes in various samples is ‘single target expression profiling’, or STEP. In this technique attention is focused on a specific gene (a target), for which an expression profile is prepared, i.e. its expression is measured in various tissues of healthy and/or ill persons (see box on p. 74). When the results are entered in a diagram it can be readily seen in which areas of the body the gene in question is particularly active or inactive. A diagram of this kind is therefore referred to as a body map. A specific tissue from different people can also be examined. Several hundred individual values can be meas-
Quantitative analysis of human DNA using the LightCycler.
PCR: an outstanding method
75
PCR and target evaluation: STEP Single Target Expression Profiling, or STEP for short, provides an overview of the expression of a specific gene in various samples. In this method the number of working copies of a specific gene is measured by quantitative PCR. The samples can come from different tissues of an individual or from tissues of many different donors. By comparing the profiles in healthy and ill people it is possible to determine, for example, whether a gene in question really is associated with the development or progression of a disease.
colon normal
level
colon cancer
The values obtained are plotted and compared. The example below is a comparative diagram showing the expression profile of various tissues, including samples from the colon of a healthy subject and of a patient with cancer of the colon (circles). The gene in question was suspected of being more active in the presence of colon cancer. However, the STEP procedure showed that the colon-specific gene is equally active in healthy and ill individuals.
ured and plotted. In many cases this reveals whether a target really is associated with a disease. In addition to target evaluation, the STEP method – like other quantitative PCR techniques – is also used in other areas of biological science, e.g. in the development of model systems (for example new cell lines) and in basic research. The establishment of whether, when and how a gene is expressed provides important information on the role of the corresponding gene product in the body’s molecular network. This approach has shed light on many biological processes, e.g. the body’s response to external factors such as administration of drugs.
Another important application of quantitative PCR is in molecular diagnosis, i.e. the diagnosis of diseases based on molecular findings rather than on physiological symptoms. In this connection the diagnosis of viral diseases is an area that is gaining increasing importance. For simple diagnostic testing, i.e. to determine if a pathogen is present in the patient’s body, qualitative PCR is sufficient. However, to follow the progress of a disease and to help
Applications in the field of infectious diseases
76
choose the right treatment doctors often need to know the actual concentration of pathogens present. PCR is one of the few techniques available today that is able to measure the pathogen load. This is an important parameter, for example, in viral infections, which often follow a chronic course and produce no clinical effects for some time, despite infection and in some cases ongoing physical damage. In this case the viral load in the bloodstream can provide an indication of how the disease is progressing. In addition, quantitative determinations serve to monitor the success of treatment. If a drug works in a patient, his/her pathogen load will decline sharply. However, some viruses change so rapidly that they cannot be completely eradicated by the drugs used: they become resistant to them. This often occurs, for example, in viral hepatitis C infection. The hepatitis C virus (HCV) often causes chronic inflammation of the liver, leading to liver cirrhosis or even cancer in a substantial proportion of those affected. Damage to the liver often accumulates over decades without being directly detectable. As mentioned earlier, qualitative PCR has been used for some years now for diagnosing HCV infection. But the quantitative form of the technique opens up whole new perspectives for the treatment of the disease. With the help of this method it is possible to monitor the success of treatment as well as determine how rapidly the disease is progressing, if at all. Unlike with conventional methods, it is also possible to ascertain whether the disease has been eradicated or has become chronic.
Hepatitis C viruses.
PCR: an outstanding method
77
Example: HIV
Another important application in which quantitative PCR is used in the field of infectious diseases is AIDS. Those infected with the causative agent of the disease, the human immunodeficiency virus (HIV), have to take a cocktail of usually three drugs indefinitely, this being the only way to keep the virus in check, if only for a while. HIV mutates extremely rapidly, quickly becoming resistant to drugs. Viruses that are resistant to all three drugs at the same time sometimes even occur, requiring a new cocktail to be used. In order that this moment, known as a viral breakthrough, should not pass unnoticed – rapid proliferation of the viruses could cause the disease to flare up again – the quantity of viral particles in the blood must be measured periodically. A rise in the viral load indicates that the drugs being used are losing their efficacy. Quantitative PCR permits such monitoring and helps doctors adjust the treatment optimally.
Genetic factors are always involved in the development of cancer. Their contribution varies greatly depending on the type of cancer. Genes not only help to determine progression of the disease but can also have a substantial influence on the effectiveness of the available treatments. Identifying the genes that play a role in the development of cancer is therefore an important step towards improving treatment. Both qualitative and quantitative PCR play a crucial role in the fight against cancer. PCR can identify genes that have been implicated in the development of cancer. Often the genes exist in a number of variants with significantly different effects. One example is the gene known as p53, whose product is a central monitor of cellular division. If the function of this monitor is disrupted in a cell, the cell can become cancerous relatively easily. Variants of p53 and similar genes can be detected by qualitative PCR, giving doctors and patients an indication of their personal risk of developing cancer or – if the patient already has cancer – how aggressively it can be expected to progress. Because multiple changes have usually accumulated before cancer actually develops, a reliable test must examine a large number of gene variants. For this reason DNA chips are being increasingly used to screen people for genetic changes (see chapter on DNA chips).
Applications in cancer therapy
78
Meanwhile, quantitative PCR also is gaining importance in the fight against cancer. This is the case where a cancer has a genetic basis but is not due to an altered gene. In some cancers the genes that control cellular division are intact but have been switched off. This can occur through a process known as promoter methylation. In the DNA region containing the start information for reading the downstream gene (the promoter) the cell attaches small molecules (methyl groups) to specific building blocks of the DNA (the cytosine bases). As a result, polymerases that normally read the genes and produce working copies of them are no longer able to dock to the start region. The gene therefore remains silent and no gene product is formed. Only in recent years has it emerged that this mechanism of promoter methylation shuts down vital genes in many cancer cells. ‘Methylation status’ is therefore of crucial importance because it provides information on the chance of a tumour becoming malignant and giving rise to metastases. A simple and reliable method used for detecting these crucial DNA changes is methylation-specific PCR (MSP). In this relatively new technique cellular DNA is first treated with sodium bisulphite, which converts normal cytosine to the RNA building block uracil but leaves methylated cytosine intact. This results in different products depending on the methylation status of the DNA (see box). Specific primers for those products are used in the subsequent PCR procedure. In this way it can be determined if the original DNA was methylated or not.
Example: promoter methylation
A host of other applications
New applications for PCR are still emerging, particularly in the field of medicine. The search for genetic predispositions to diseases is an especially important area of research. In many cases the onset of a disease can be prevented or at least delayed by lifestyle modification or the taking of medications. One example of this is osteoporosis, a loss of bone density that is especially common in postmenopausal women. Because the disease tends to run in families, it is clear that genetic factors are involved in its development and progression. An intensive search for the genes involved is currently under way and has already produced some results. Several promising candidate genes have been identified that appear to be involved in osteoporosis. In this context PCR is not only an aid in the search for the culprit genes, but also of-
PCR: an outstanding method
79
fers the possibility of identifying the responsible gene variants in patients. Similarly, PCR is helping in the investigation and diagnosis of a growing number of diseases. It has also long been a standard method in all laboratories that carry out research on or with nucleic acids. Even competing techniques such as DNA chips often require amplification of DNA by means of PCR as an essential preliminary step.
Identifying inactivated genes: methylation-specfic PCR (MSP)
unmethylated DNA
methylated DNA CH3 CH3
ATC
CGCGTCG
CH3 CH3
ATC
ATC
primer 2 CH3 CH3
CH3
CGCGTCG GCGCAGC
ATC
primer 1
ATC
primer 2
GCGCAGC primer 2
no product
3.
PCR product only with primer 1
In many tumours important genes that control cell growth are switched off by methylation of the promoter region. These changes can be detected by means of methylationspecific PCR (MSP). In MSP all the normal cytosines (C) of the original DNA are converted to the RNA building block uracil. The methylated
80
CH3
CGCGTCG primer 1
2.
UGUGTUG ACACAAC
ATC
NaHSO3
1.
UGUGTUG
CH3
CGCGTCG
ACACAAC primer 1
no product
PCR
PCR product only with primer 2
cytosines, by contrast, remain unchanged (1). The subsequent PCR procedure then uses specific primers for the various products formed (2). Hence, either the original methylated or the unmethylated DNA is copied. The original DNA was therefore methylated or not (3) depending on the primer used to obtain a product.
References Mullis KB, Faloona FA: Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol 155: 335–350, 1987 Higuchi R, Dollinger G, Walsh PS, Griffith R: Simultaneous amplification and detection of specific DNA sequences. Bio/Technology 10: 413–417, 1992 Wilfingseder D, Stoiber H: Quantifizierung von PCR-Produktmengen durch real-time PCRVerfahren. Antibiotika Monitor, Heft 1/2/2002 Reidhaar-Olson JF, Hammer J: The impact of genomics tools on target discovery. Curr Drug Discovery, April 2001 Brock TD: Life at high temperatures. Bacteriology 303: Procaryotic Microbiology, 1994 http://www.bact.wisc.edu/Bact303/b27 Mullis KB: The polymerase chain reaction. Nobel Lecture, December 8, 1993 http://www.nobel.se/chemistry/laureates/1993/mullis-lecture.html Kary Mullis Website: http://www.karymullis.com Deutsches Hepatitis C Forum e.V. Homepage – Qualitativer HCV-RNA Nachweis: http://hepatitis-c.de/pcr1.htm
PCR: an outstanding method
81
SNPs: the great importance of small differences
G G G G G G
Single nucleotide polymorphisms, or SNPs, the tiny differences between individual genomes, make each of us unique. At the same time, however, they are partly responsible for individual differences in the effectiveness and tolerability of drugs. SNPs have therefore become one of the most important objects of medical research, especially as they could also provide clues to new targets.
C C C C C C
A A A A A A
A A A A A A
T T T T T T
C C A C C A
T T T T T T
A A A A A A
Small genetic differences come in many varieties, and the vast majority probably have no consequences at all. Some influence characteristics that are medically irrelevant (for instance, they may contribute to us having our mother’s mouth but our father’s chin). Others are potentially more relevant to health (for instance, they may contribute to us being as restless as our grandmother, or to putting on weight Terms at as early an age as did our grandfather). Sometimes SNPs single nucleotide polymorphisms – differences in these small differences can individual building blocks (base pairs) of DNA that are distributed randomly over the genome and passed from generation have more significant conseto generation. quences (for instance, they Genotype the alternative forms (alleles) of a gene present may increase our risk of dein an individual; generally there is a maximum of two – one from veloping a disease or lower the father and one from the mother. Phenotype the constitution of a living creature that results the likelihood that we’ll refrom the interaction of genotype and environmental influences. spond to a medicine). Very Phenotype refers both to medically irrelevant characteristics rarely – mostly in the case of and to diseases. Pharmacogenetics the branch of science concerned typical inheritable diseases – with the influence of genetic variation on the effectiveness and their mere presence or abside effects of drugs. sence determines whether a Targets the molecules – proteins or small organic molecules – upon which drugs act in our body. family member will develop a disease or remain healthy. Most of these genetic differences between individuals consist of single nucleotide polymorphisms, or SNPs (pronounced ‘snips’). SNPs are randomly distributed variations of the building blocks of our genome that make each of us genetically unique. They contribute to family resemblance with regard both to external features and to the risk of developing certain disorders. In medicine, SNPs have become important parameters in the search for new, safer, more effective and better tolerated drugs aimed at providing more targeted therapy. 99.9 percent of the human genome is identical in all individuals. On average, however, one in every 500 to 1000 base pairs of our genome differs from the one found in the majority of people. These randomly occurring changes are passed from generation to generation and account for a high proportion of the DNA differences between us. It is estimated that between three and six million such variations lie hidden in our genome. When such a variation is present in at least one percent of the population – or to be more precise, of a particular population, i.e. an ethnic group – it is referred to as a single nucleotide polymorphism, or SNP.
SNPs: common, hereditary, stable
84
The small difference: SNPs G G G G G G
DNA sequence
C C C C C C
A A A A A A
A A A A A A
T T T T T T
C C A C C A
T T T T T T
A A A A A A
DNA sequences differing by a single nucleotide
Single nucleotide polymorphisms, or SNPs, are differences in individual building blocks (nucleotides) of DNA that are distributed randomly over the genome. They can occur in any position within or outside of genes and accordingly can have very different effects. SNPs present within the proteinencoding regions of a gene may result in incorporation of an alternative amino acid in the protein for which the gene serves as the blueprint, or template. Depending on where this occurs within the protein and to what extent the alternative amino acid differs from the normally incorporated one, such an amino acid exchange can have a profound influence on the function of the protein.
M M
SNP C
SNP A
individuals with different SNPs
In the following theoretical example, replacement of a cytidine (C) with a guanine (G) in the gene results in formation of an amino acid with completely different properties: instead of the large, basic amino acid glutamine (Gln), the small, neutral amino acid glycine (Gly) is formed.
unchanged SNP Gene AAG-CGA-ATT-AGG › AAG-GGA-ATT-AGG Protein Lys -Gln -Ile -Arg › Lys -Gly -Ile -Arg
Depending on where it occurs, such a variation can have very different effects. SNPs are therefore subdivided into four groups on the basis of their site of occurrence: z rSNPs (random SNPs): Only about ten percent of our genome is made up of genes. The great majority of SNPs are therefore located in what we currently view as ‘silent’ regions of our genome. These SNPs are extremely unlikely to have any perceivable effect on our phenotype, or constitution. In science, many of them are used as markers in the mapping of genes within the genome. z gSNPs (gene-associated SNPs): Many SNPs are situated alongside genes or in introns, the regions of a gene that do not code for a gene product, i.e. do not form part of the template for a protein. The fact that they are inherited with these genes makes gSNPs useful for the study of associations between the gene (and its variants) and certain phenotypes. Mapping gSNPs may be functionally relevant if they influence important control elements of the gene and thereby decrease or increase transcription of a gene.
SNPs: the great importance of small differences
85
The less common, the more important: SNP groups
chromosome
gene
·
·
Genes are arranged in different ways on chromosomes. They generally consist of a promotor region (pr), in which the control elements of the gene are located; exons (ex), the segments of the gene that are translated into the corresponding protein (the gene product); and introns (in), segments located between exons that are not part of the protein template. SNPs occur in all regions of the genome and are subdivided into four groups on the basis of their site of occurrence:
·
ex in
·
pr
rSNPs (random SNPs) are situated at a distance from genes and are used mostly as aids to mapping. They are relatively common (estimated at 3–6 million in humans).
gSNPs (gene-associated SNPs) are situated alongside genes or in introns. They too are used mostly for gene mapping; they can also influence the control of gene activation. gSNPs are less common than rSNPs (estimated at less than 1 million). cSNPs (coding SNPs) are situated in exons and often influence the function of the corresponding gene product (estimated at about 100,000). pSNPs (phenotype-relevant SNPs) are gSNPs or cSNPs that influence the constitution of the organism. These are the most important type of SNP from the point of view of medicine, but are the least common type (probably no more than 10,000).
z cSNPs (coding SNPs): Exons are the coding regions of a gene,
i.e. the sequences of a gene that are translated into the gene product – the protein. SNPs that are present in exons can have a major influence on the function of the protein concerned if they result in incorporation of an alternative amino acid. z pSNPs (phenotype-relevant SNPs): Both gSNPs and cSNPs can influence a person’s phenotype: the former primarily via the amount, and the latter usually via the form, of the protein for which the gene codes. pSNPs are the most important type of SNP from the point of view of medicine. They form one of the foundations of pharmacogenetics, the branch of science concerned with the influence of gene variation on the effectiveness and tolerability of drugs. SNP analysis is now performed on a number of mature technology platforms with very high accuracy and reproducibility.
86
The following examples show how important SNPs are in relation to medical practice. The effects of these sequence variations can be divided into two categories that are of relevance to medicine: on the one hand, interference with the action of drugs, or, to be more precise, with the way in which the body deals with drugs; and on the other hand, involvement in the development and progression of diseases. Variations of the first category may help explain why drugs work more or less well in some people and why they may cause undesirable side effects only in certain people. Knowledge of the position and effects of SNPs should therefore be considered in the development of better and more targeted forms of treatment. Variations of the second category are important reasons why individuals differ in terms of their susceptibility to certain diseases despite living in similar environmental conditions.
Medical effect at two levels
Gene variants can affect the action of drugs in various ways, including: z Absorption. The uptake of a drug into the body can be disturbed, with a corresponding reduction in the effect of the drug. z Activation. Many drugs work only after being converted into a different form in the body, e.g. by removal of a protective molecule. The function of the enzymes that catalyse such reactions may be impaired by SNPs in the gene concerned. z Distribution. To work properly, a drug must reach its site of action at the correct concentration. Since many proteins play a role in this transport, SNPs can have a major influence on drug distribution. If delivered at inappropriate (either too high or too low) concentration to the target, or if delivered to the wrong site, drugs can have undesirable effects. z Breakdown and elimination. As in the example of the P450 family of enzymes (see below), foreign substances (e.g. medicines) commonly need to be chemically altered (broken down) in the body so that they can be eliminated. If this happens too rapidly, the effectiveness of the drug may be reduced, whereas if breakdown is too slow, the drug remains in the body for too long and reaches excessive levels that increase the likelihood of undesirable effects. Moreover, a harmless molecule may be converted into one with dangerous properties, e.g. one that promotes cancer.
SNPs relevant to drug response (pharmacogenetics)
SNPs: the great importance of small differences
87
Breakdown four times faster: cyp2c19 and omeprazole
100
80 cure rate in %
The cyp2c19 gene of the cytochrome P450 gene family plays an important role in the metabolism of fat-soluble drugs. Forexample, the proton pump inhibitor omeprazole works up to four times less well in individuals with a certain variant of this gene (‘mut’) than in those with the most common genotype (‘wt’). This is because the drug is broken down faster in individuals with the mut genotype. In this group of patients the cure rate can accordingly be improved by giving higher doses of the drug.
gastric ulcer peptic ulcer
60
40
20
0 mut/mut
wt/mut cyp2c19 genotype
z Molecular target. SNPs in the gene that codes for the target
molecule of a drug can directly interfere with the action of the drug on its target. The beta2-adrenergic receptor and the anti-asthma drug albuterol provide an example of this (see below).
SNPs assume particular importance whenever they are associated with the effectiveness or tolerability of medicines. The cytochrome P450 proteins are of great importance in the elimination of drugs from the body (metabolism). Many of the P450 proteins occur in a number of variants (based on SNPs). Some of these variants have clearly altered function. Depending on which variant is present, the way the body ‘treats’ certain drugs, and thus how it responds to them, can vary substantially. cyp2c19 is a member of the P450 family and is one of the proteins responsible for ensuring that fat-soluble substances are rendered water-soluble in the liver and thereby made available for elimination from the body. Such substances include many drugs, e.g. omeprazole, which is used to treat stomach (peptic) ulcers. At least two dozen different SNPs are now known to exist within the cyp2c19 gene and about 50 more are known to be located in its immediate
cyp2c19 in stomach ulcers
88
wt/wt
vicinity. Some of these variants are known to have considerable influence on the function of the enzyme encoded by the gene. Thus, some individuals break down omeprazole four times faster than others, with the result that standard doses of this normally very potent drug bring scarcely any benefit in these individuals. Foreknowledge of these variations can be of great clinical value: if a person’s genetic status in this respect is known, the dose of the drug can be adjusted accordingly at the outset.
Another example of pharmacogenetically important SNPs is the gene for the beta2-adrenergic receptor. This molecule performs a number of functions in the body (see box). Activation of it in the lungs relaxes the smooth muscle of the airways. Some anti-asthma
Beta2-adrenergic receptors
Triple role: beta2-adrenergic receptors 16
27
NH2
164
cell membrane
COOH
The gene for the beta2-adrenergic receptor has at least three medically important variants. One of them is of pharmacogenetic relevance, the two others signal increased risk for disease or disease prognosis. z The antiasthmatic agent albuterol works well only in patients in whom the amino acid arginine is present at position 16 of the receptor, whereas it is less effective in patients with a variant of the gene that results in the presence of glycine at thisposition (pharmacogenetics). z The genotype that codes for position 164 of the recep-
z
tor is significantly associated with an increased risk of developing end-stage heart failure in the aftermath of a major myocardial infarction. Thus, it has been suggested that bearers of a certain amino acid substitution at this position should be followed particularly closely and may ultimately require heart transplantation (disease prognosis/outcome). A genetic variant that influences position 27 of the beta2-adrenergic receptor appears to play a role in the development of obesity (disease risk).
SNPs: the great importance of small differences
89
medicines therefore aim to activate this receptor. The presence of a certain SNP in the gene for this receptor can greatly reduce the effectiveness of the anti-asthma drug albuterol.
While we generally view the role of SNPs as being to affect the relative risk of contracting a disease, in extreme cases a single SNP may actually cause a disease. In sickle cell anemia, for instance, substitution of a single nucleotide in the gene for hemoglobin, the oxygen-carrying red blood pigment, results in synthesis of an altered protein which under certain chemical conditions takes on an abnormal shape. This causes the affected red blood cells to assume the shape of a sickle (hence the name of the disease), clump together and potentially block small blood vessels, leading to tissue death and extreme pain. The great majority of diseases that afflict mankind arise as a result of a very complex, much more balanced interplay between a number of genetic, environmental and life-style factors. In these diseases SNPs can account for an increased (or decreased) likelihood of developing the disease. Genetic testing for SNPs may thus help in the evaluation of an individual’s risk of developing a certain disease. Though these tests can only indicate a somewhat higher or lower risk, they may provide prognostic information that allows the person concerned to make more informed decisions about preventive measures such as lifestyle changes and about more targeted medical follow-up (early recognition of any recurrence of the disease). SNPs can also provide information about the molecular basis of disease. The finding of an association between certain SNPs and a particular disease suggests that the gene associated with those SNPs may play a role in the development of that disease. In this way new disease-relevant genes, and thus new targets for drugs, can be discovered (see chapter on targets).
SNPs relevant to the development and progression of diseases
SNPs are therefore important at all levels of drug research and development, from investigation of the molecular basis of a disease through the search for and evaluation of new targets to the clinical testing and regulatory approval of new drugs. Knowledge of the distribution and effects of SNPs is already having a noticeable impact at the latter two levels. For example, the fact that some people
Drug studies and regulatory approval
90
break down certain drugs more or less rapidly is already being considered to some extent in the testing of new drugs, and such individual differences in drug metabolism are likely to be taken increasingly into account as our knowledge of the distribution and effects of SNPs grows. More ambitious are attempts to develop drugs for use in specific patient groups. Where it is known in advance that a substance is most effective in bearers of a certain gene variant, the drug concerned can be tested specifically in that target population. This, of course, assumes the availability of simple and rapid methods of testing for the presence of that gene variant in trial participants. The occurrence of adverse events can be reduced in a similar way: if SNPs associated with the adverse event have been identified, patients can be tested for them. Those at risk can be offered an alternative drug or an appropriately adjusted dose. In this application the study of SNPs does not lead to the discovery of new medicines, but to diagnostic possibilities that permit better targeted and safer use of existing forms of treatment. Newly discovered SNPs may of course also be relevant to new drugs for which regulatory approval is being sought. They may prove useful in reducing the incidence of side effects or in enhancing response rates (efficacy) if these issues arise in the course of clinical trials.
Given the likely medical importance of SNPs, it was important to develop the knowledge base required for finding them and to spur on the development of increasingly high-throughput analytical platforms and technologies. The task of finding and then evaluating several million variants among the three billion base pairs that make up the human genome would have been completely impossible as recently as the early 1990s. The Human Genome Project and the technologies that were developed as spin-offs from it made an important contribution to our ability to carry out an extensive examination of our genome for SNPs. For SNP research, therefore, increasing automation and miniaturisation of biological methods were of pivotal importance.
Paving the way for new methods
Although SNP research on a large scale has therefore been possible only for a relatively short period of time, a large number of these variations in the human
The SNP Consortium
SNPs: the great importance of small differences
91
Smaller, faster, automated: methods of studying SNPs Techniques developed or refined in the past few years permit systematic study of SNPs. Automation and miniaturisation, in particular, have now made it possible to perform in a single step thousands of experiments which only a few decades ago would have required years of work in major research institutions. Evaluation of the enormous quantities of data gathered in this way has now become the subject of a special branch of science known as bioinformatics. Three stages of research can be distinguished: 1. Search for SNPs A major part of the search for SNPs is now conducted in silico, i.e. by computer. The results of laboratory experiments are stored in databases which are then compared in computers and scoured for SNPs. Special computer programs are used as aids in the search; these also ensure that SNPs are distinguished from sequencing errors. Since individual SNPs can be very rare, it is still necessary to sequence the genome of many additional volunteers in order to obtain comprehensive data. These data are anonymised. The most important technique used in DNA sequencing is the polymerase chain reaction, or PCR (see chapter on PCR). 2. Evaluation of SNPs Databases and special computer programs are also used to identify associations between SNPs and genes. Essential in
SNP analysis by computer.
92
the search for associations between SNPs and diseases or pharmacogenetic problems are association studies. These are used to determine whether certain SNPs prevail in individuals with a particular disease. Another important method is that of in-vitro assays to investigate the properties of proteins. These can rapidly determine whether the presence of a certain SNP in a gene results in altered function of the corresponding gene product. 3. SNP tests for patients The final link in the chain of applications of SNP research is the development of tests to determine the presence of SNPs in patients. It is not yet clear which of the various techniques being developed for this purpose will prove to be both economically viable and scientifically satisfactory and therefore become established as the norm. A ‘line array’, for example, can detect 60 variants on a membrane. Far greater numbers can be detected using a ‘SNP chip’. In this technique, several thousand SNPs can be accommodated and simultaneously tested on a DNA chip. Another method uses MALDI-TOF (see chapter on proteomics) to investigate SNP-specific oligonucleotides. Also used are a variety of other methods that employ various techniques including PCR and nanotechnology.
genome are already known. This is due in large measure to the work of ‘The SNP Consortium’ (TSC), an initiative that was set up in 1999 by ten major pharmaceutical companies, the UKbased Wellcome Trust (the world’s largest medical research charity) and five leading academic research centres with the aim of drawing up a comprehensive SNP map of the human genome. TSC set itself a two-year deadline for the task of identifying a total of 300,000 SNPs in the human genome and mapping at least half of these, i.e. determining at least their approximate position in the genome. The various partners in this uniquely ambitious, privately funded project contributed over 50 million US dollars to TSC on condition that the results it obtained be published immediately and made freely available to the scientific community. The database that was built up – a comprehensive SNP map of the human genome – is now available free of charge to anyone at any time and – because there are no patents – is not subject to any licensing agreements. In November 2001 The SNP Consortium published its final data set: 1.7 million SNPs had been found, 1.5 million of these mapped and 1.3 million allocated to a specific position in the (at that time still provisional) sequence of the human genome. TSC was thus an outstanding success, and its results provide a solid basis for further SNP research.
Based on this body of knowledge, efforts are now being directed towards the discovery of medically relevant SNPs. This work can in principle be pursued via two different approaches: z In association studies, SNPs in candidate genes (genes which may plausibly be related to the biomedical issue at hand) are examined for statistically significant associations with disease incidence, prevalence, outcome or drug response. This is a technically demanding but important area of research. The overall function of SNP association studies is to establish a link between the mere mapping of individual differences in our genome and the biological – and in some cases medical – significance of these differences. z Whole genome scanning, i.e. the systematic and unbiased analysis of the entire genome for SNPs that show a statistically significant association with a phenotype, without preconceived notions about likely candidate genes, is the next
Current areas of SNP research
SNPs: the great importance of small differences
93
logical, but very challenging, step in the use of SNPs. This approach has the distinct advantage that it is not limited to the still small number of genes about which we know enough to declare them candidate genes. At present, however, such tests are still too costly and time-consuming for routine use, and not a single example has been published. Also, because of a number of unresolved issues, the feasibility of genome-wide SNP association studies remains quite controversial.
From the perspective of patients, SNPs are a major step towards a more personalised approach to treatment that takes genetic differences between people into account. From the perspective of the pharmaceutical industry, SNPs have become important parameters to be considered in drug research and development. SNPs have the potential to lead to the discovery of new targets and thus eventually to the development of new and better drugs, and they bear the promise of leading to more effective, safer and better tolerated forms of treatment. Our understanding of the importance of these small but potentially crucial differences is growing all the time. Finding those that matter for healthcare has therefore become an important new aspect of pharmaceutical research and development.
SNPs – opening new doors to better healthcare
References Foernzler D: SNPs – kleine Unterschiede mit großer Wirkung. BioWorld, June 2000 Stoneking M: Single nucleotide polymorphisms: from the evolutionary past … Nature 409: 821-822, 2001 Chakravarti A: Single nucleotide polymorphisms: … to a future of genetic medicine. Nature 409: 822-823, 2001 The SNP Consortium – Website: http://www.ncbi.nlm.nih.gov/SNP TSC data on the CYP2C19 gene: http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?locusId=1557 Holden AL: The SNP Consortium: summary of a private consortium effort to develop an applied map of the human genome. BioTechniques 32: S22-S26, 2002 Weiner MP, Hudson TJ: Introduction to SNPs: discovery of markers for disease. BioTechniques 32: S4-S13, 2002 Abraham J, Wilson DE: Roche scientists exceed expectations of genetic discoveries – more than 18,000 mouse SNPs identified in 27 months. Roche press release, Palo Alto and Pleasanton, Calif., 5th November 2002
94
DNA chips: choosy fish hooks
Very rapid, very sensitive and very safe – these are the requirements of a good investigative method. Biochips satisfy these requirements. DNA chips, the most important type of biochip, are presently being developed at breakneck speed: already tiny, they are becoming ever smaller, ever more sensitive and ever safer. At the same time, new applications are being developed that could give DNA chips a central role in medical diagnosis.
When scientists fish in murky waters, it is not necessarily a bad sign. After all, only rarely does nature provide clear solutions. For example, a cell extract contains, in at most slightly presorted form, the entire inner life of thousands or even millions of cells in the form of a generally colourless, opaque, thick fluid. All that matters is that from such murky solutions scientists be able to draw clear conclusions. And nowadays they are aided in this task by ‘fishing lines’ whose properties would make ordinary fishermen green with envy: Terms fast, accurate and capable of catching enormous numbers Biochip a solid substrate (e.g. glass or plastic) upon which of different types of fish at biomolecules are anchored. DNA chip biochip with single-stranded DNA as the probe. the same time. These fishing GeneChip a widely used DNA chip developed by the US lines are known as biochips, company Affymetrix. and there can be little doubt DNA deoxyribonucleic acid; the chemical substance of which our genetic material consists. that a bright future awaits RNA ribonucleic acid; the chemical substance of which, them. In medicine, at least, among other things, working copies of genes (mRNA) consist. cDNA complementary DNA; DNA transcribed enzymatically they are in the process of from RNA (mostly mRNA). turning research, diagnosis Nucleic acids generic chemical term for DNA and RNA; and therapy upside down. chain-shaped molecules whose individual building blocks are bases/nucleotides. Biochips are among the most Oligonucleotides short nucleic acid chains composed of at important instruments used most a dozen building blocks (nucleotides). in the miniaturisation and Genes functional segments of our genetic material that serve mostly as blueprints for the synthesis of proteins. automation of biology. In Genome the totality of the genes of an organism. most applications their task is to recognise and bind to specific molecules in a solution – like fishing lines set up to catch only one kind of fish, but to do so with a high degree of reliability. The ‘hooks’ used for this purpose are molecular probes attached to a substrate surface barely the size of a thumbnail. It is this surface that gave biochips their name: it consists of plastic or glass and is similar to the silicon chips used in the computer industry. In principle, any substance that interacts with components of our cells can serve as a molecular probe.
In the type of biochip that is most important at present, the molecular probe that is attached to the chip is DNA. In future, DNA chips are likely to serve the most varied of purposes ranging from basic research in biology through diagnosis of disease to water ecology. Equally as varied as their potential applications are the shape, size and method of manufacture of DNA chips. Despite
The most important biochips at present: DNA chips
96
image of hybridised DNA array
DNA chip
fluorescently labelled RNA (probe)
multiple probes for a single gene
50µm
these differences, almost all DNA chips exploit the same biological principle, that of hybridisation. The four bases that are the building blocks of DNA always pair with the same ‘partner’. Our genetic material therefore consists of two strands of DNA arranged in the form of a twisted rope ladder, or double helix. The two strands are complementary in the sense that the sequence of one can be deduced from that of the other. This joining together, or ‘hybridisation’, of two nucleic acid chains to form a double-stranded structure has been exploited by biologists for decades. For example, labelled short segments of single-stranded DNA (oligonucleotides) can be used to search for the presence in our genome of oligonucleotides with the complementary base sequence. DNA chips do basically the same thing. DNA fragments tethered to the chip bind to complementary base sequences in the solution being studied. The difference is that DNA chips make it possible to perform many such experiments at once: millions of copies of each of several hundred thousand different oligonucleotides can now be accommodated on a chip measuring just one square centimeter. Conversely, such a chip can be used to search for tens of thousands of different DNA segments in a solution – and it is precisely this possibility that forms the basis of entirely new applications in biological research and medicine.
DNA chips: choosy fish hooks
97
fluorescent labelling probe
5'
3' T
C
A
T
A
A
G
T
G
C
T
A
A
T
target DNA T
5'
G
A
T
G
A
T
T
C
A
G
C
C
C
G 3'
hybridisation
T
G
T
A
C
T
A
A
A
T
G
A
T
T
G C
T
C
G
T
A
A
G
C
A
T
C
C
The most important application of DNA chips is in the search for, and the study of, genes. In this application, as in the classical oligonucleotide experiment, short segments of DNA are used to help identify longer genes. There are two basic ways in which this can be done: either the genes are attached to the chip and incubated with a solution of a labelled oligonucleotide, or else the oligonucleotide is attached to the chip and the genes are placed in the solution. Of these two methods, the former was developed first, whereas the latter is used more commonly nowadays because it permits the performance of more experiments on a single chip. The developers of both types of chip were in any case faced with the same two problems, namely how to get the DNA onto the chip and how to know when two matching bases have found each other. In both cases a variety of approaches have been tried, and it is not yet clear which techniques will win out. A large number of companies are currently offering competing methods of tackling scientific tasks that are in some cases identical but in other cases different. It may be that a number of different techniques will survive.
Field of application: gene research
98
G
Common example: the GeneChip
a
b
c
e
d
A
f
G
A
A A
One of the most commonly used DNA chips at present is Affymetrix’s GeneChip. This is an oligonucleotide chip in which the short strands of DNA are synthesised in situ, i.e. on the chip, by means of photolithography (see above): a. Reactive sites on the chip surface are blocked by photosensitive protector groups (small squares). b. An opaque mask covers the greater part of the chip; the beam of light therefore removes only those protective groups that are situated in a certain region. c. The chip is incubated with a solution containing one of the four nucleosides adenosine, thymidine, cytidine or guanosine (A, T, C or G), which likewise bear protector groups; the nucleosides react with the previously unmasked regions of the chip. d.–f. The cycle is repeated with another mask; this gives rise to regions with different oligonucleotides of known sequence. Such a GeneChip can be used to examine different types of
A
G
G G AG
A T CG AGAC CGTC T AG T
DNA solution. Commonly performed experiments include genome studies and, in particular, gene expression profiles. These are used to identify those genes that are actually expressed in a given cell type or tissue. For this purpose the mRNA – the working copy of genes – in a cell extract is transcribed into cDNA. This process is known as ‘reverse transcription’, as opposed to transcription, which is the synthesis of RNA on the basis of DNA. Precisely this can occur in a second step, since in many experiments the cDNA that is formed is transcribed back into cRNA. In one of these steps a label is introduced; commonly used for this purpose is, for example, the molecule biotin. The cRNA (or cDNA) is then cut into smaller pieces and placed on the chip, where it hybridises with the oligonucleotides. Measurement of fluorescence then shows how much of the label is bound at what sites on the chip – and thus what quantity of the mRNA of interest was present in the cell extract.
One of the most important DNA chip technologies was developed by the Californian company Affymetrix. The name of this company’s best-known product, GeneChip, is often used synonymously with the term DNA chip to refer to any such product. (In addition to ‘DNA chip’ and ‘gene chip’, other terms including ‘microarray’, ‘genome chip’ and ‘gene array’ are in common use.) Affymetrix manufactures its GeneChips using the principle of photolithography, just as in the manufacture of computer chips. In this technique a light source, special masks and photosensitive protector molecules are used to deposit billions of oligonucleotides with (at present) up to 700,000 different base sequences alongside each other in tiny cells (spots) on a chip (see box on page 100).
The example of GeneChip
DNA chips: choosy fish hooks
99
total RNA
biotin-labelled cRNA
cDNA reverse transcription
AAAA
in vitro transcription
B B
AAAA B
AAAA
B fragmentation GeneChip expression array
B B fragmented, biotin-labelled cRNA
hybridisation
B
B
B
B
B
B wash and stain
B
B
scan and quantitate
B B
Such a GeneChip is then incubated with a solution containing the DNA of interest, which has previously been labelled with a fluorescent dye. Whether given oligonucleotides on the chip have hybridised with DNA in the solution is apparent from the positions on the chip at which fluorescent dye is present at the end of the experiment. For this purpose the individual positions on the chip are read with a scanner. The readings are analysed by computer with the aid of specially developed programs.
100
Competing techniques: design and function of DNA chips The tasks performed using DNA chips are many and varied, and the design of such chips is correspondingly diverse. Affymetrix’s GeneChip is a commonly used DNA chip, however various other other manufacturers are offering a variety of techniques aimed at winning over customers. Points of difference include not just chip design, but also the way in which the experiments are performed and the way in which the results are analysed. z Probe material The most commonly used DNA chips use short oligonucleotide chains as probes, however RNA, cDNA, genes and even whole chromosomes can be attached to chips. z Manufacture DNA can be attached to chips in various ways. These include photolithography, a technique borrowed from the computer chip industry. Other techniques include application by pipette, dropping and electronic methods, e.g. in a manner similar to the operation of an inkjet printer. z Target molecules The probe and target molecules are dependent on each other. Depending on what type of target molecule is present on it, a chip may be suitable
z
z
z
for the study of other types of DNA, such as oligonucleotides, RNA, cDNA, genes, chromosomes or whole genomes. Reaction Hybridisation is not the only reaction that can occur on a DNA chip. DNA molecules can also be bound by ligases or via chemical or photochemical reactions. Another possibility is to combine PCR (polymerase chain reaction, see chapter on PCR) with a chip. Detection Different reactions on the chip require different methods of detection, and hybridisation can be detected in various ways. In addition to fluorescence, mass spectrometry (MS, see chapter on proteomics), in particular, and also conductivity and electronic methods, can be used for detection. Analysis DNA chip experiments generate enormous quantities of data that would be impossible to evaluate without computer assistance. Of importance in this regard are not just suitably sophisticated programs, but also automatic control of experiments, image analysis, databases, Internet links and platforms and visualisation of results.
Though GeneChips and other oligonucleotide chips are the most commonly used type of biochip at present, a variety of other molecules are used on biochips. In the case of DNA chips, not only oligonucleotides and genes, but also RNA, cDNA and even whole chromosomes can be used. Depending on the problem to be addressed and the solution to be examined, chips can be either individually chosen or specially made. One-off products are considerably more expensive than more or less standard products. In addition, many attempts are being made at present to produce protein chips with a performance similar to that of DNA chips. As compared with DNA, proteins are vastly more difficult to produce in the required quantities and at constant quality. Protein chips are therefore still very expensive. Attachment to the chip is also problematic in that many proteins need a great deal of freedom of movement in order to function correctly. In addition, assessment of the diverse interactions that can occur between proteins and other substances is difficult and time-consuming. Given, however, that proteins occupy a central place in
Variety on a chip
DNA chips: choosy fish hooks
101
drug research, protein chips are regarded as an important tool, especially in proteomics (see chapter on proteomics).
DNA chips have also found broad application in drug development. In fact, medicine is currently one of the most important and exciting, though by no means the only, field of application of these tiny chips. Many different variants of them are used in almost all branches of biological science. Their outstanding feature in almost all these applications is their ability to analyse genes rapidly and simply. The enormous quantities of data collected in the Human Genome Project and similar undertakings form the basis for the evaluation of DNA chip experiments. When only small amounts of the DNA (or RNA) of interest are available, it is generally still necessary to amplify the nucleic acids first by means of PCR. These two techniques are therefore often used in conjunction (see chapter on PCR).
Growing number of applications
The first field of application of DNA chips was in basic research in biology. In this, unlike many other, fields, use of DNA has long been routine. Since in this field the chips are often used to address new questions, basic research also leads to the development of new techniques and opens up new fields of application. Among other uses in basic research, DNA chips have been and are used to map genomes, to find genes and control elements and to search the genomes of different organisms for points in common. Now, however, their role has been extended far beyond these uses: now that the sequence of the human genome is known, they are being used to investigate the tasks and functions of genes. An important instrument for such investigations is gene expression analysis. In this, attention is focused not on the gene itself, but on the working copies of a gene that are produced in a given type of cell. These molecules, which are known as messenger RNA (mRNA), act as intermediaries between the genome and the life processes of the cell. Their primary role is as blueprints for the synthesis of proteins. DNA chips now permit rapid and simple generation of gene expression profiles in which the activity of thousands of genes is determined simultaneously. This method, which is known as MEP (microarray-based expression
Now routine: basic research in biology
102
profiling), can be used to answer important questions such as: Which genes are expressed in which cells? When and under what conditions does gene expression occur? Which genes are active in diseases? And how is gene expression affected by administration of medicines? The results of such experiments provide important insights into the molecular processes that take place within cells. They also provide evidence of the role of certain genes in the genesis, progression and treatment of diseases. Medicine thus becomes a new field of application of DNA chips.
DNA chips long ago became a standard tool for use in research into diseases, especially as they permit analysis of almost complete genomes in a single experiment. Basic research and applied science often overlap to some extent here, however applications of DNA chips, and in particular gene expression analysis, are becoming increasingly important in all other areas of medicine, e.g.: z Genetic causes of disease. Our genome plays at least a contributory role in the genesis of the great majority of diseases. Discovering which genes play a role in which diseases and how genes interact in diseases requires detailed observation of many DNA segments simultaneously – a task for which DNA chips are well suited. z Hereditary diseases and genetic predisposition. Where disease-relevant genes are known, DNA chips can make it possible to test patients for genetic susceptibility to the disease concerned. In complex diseases such as cancer and Alzheimer’s disease a number of genes and environmental factors are generally involved. DNA chips can help people who are genetically predisposed to myocardial infarction avoid additional risk factors such as smoking, an unbalanced diet and lack of exercise. z Diagnosis. The causes of diseases can also be determined reliably with the aid of DNA chips. For example, different causes can often bring about the same signs and symptoms, and if these causes are genetic in nature they can be distinguished by means of DNA chips. This is exemplified by various types of cancer which, though also subject to external influences, almost always result from genetic defects. Knowledge of what genetic alteration is present in a patient can have a crucial influence on what treatment is required. Another example of
A time of upheaval: DNA chips in medicine
DNA chips: choosy fish hooks
103
the use of DNA chips for diagnostic purposes is in infectious diseases. Here they can be used to identify pathogens. Examples of both these diagnostic applications of DNA chips are given below. z Therapy. Our genetic predisposition has considerable influence on the efficacy and tolerability of medicines. This is due mostly to small differences in our genome known as ‘single nucleotide polymorphisms’, or SNPs (pronounced ‘snips’) (see chapters on pharmacogenomics and SNPs). DNA chips can be used to detect these differences rapidly and reliably, and in this way can provide doctors with crucial information to assist them in choosing the most appropriate treatment for a particular patient. Also, it is only with the aid of such techniques that novel medicines that take account of individual differences in the way our body reacts to drugs can be developed. DNA chips are therefore set to play a major role in the development of personalised medicine.
DNA chips also have potential for use in consumer protection. An example of this is in the field of ‘green gene technology’, i.e. the use of gene technology in agriculture. The fact that in most industrialised countries a proportion of the population takes a sceptical view of this technology has led to the introduction of a variety of regulations including compulsory labelling. Given, however, that the vast majority of foods produced in this way do not differ visibly from conventional products, it is often only by means of an examination of the genome of the plant concerned that adherence to such regulations can be effectively checked. DNA chips are well suited for use in such tests.
Checking up on green gene technology
Ecology is a broad field of application for DNA chips. For example, it is often necessary to distinguish between fairly closely related animal species in a body of water in order to assess the condition of the ecosystem concerned. This is because the presence of one species may indicate a clean, but that of the other a polluted, environment. Up to now, this task has often required painstaking and detailed work with a magnifying glass or even a microscope, since many species are scarcely distinguishable from their close relatives on the basis of their appearance to the naked eye.
Chip instead of a magnifying glass: ecology and taxonomy
104
DNA chips can make such distinctions more rapidly, more simply and above all more reliably. This ability creates applications for DNA chips in all situations in which closely related species need to be studied. These include ecology, taxonomy, anthropology and research into evolution.
DNA chips could also have a bright future in forensic medicine. Prominent in this field of application is the ability of DNA chips to detect differences between individual genomes and thereby to identify people. This can be required for identification of victims and already plays a crucial role in the search for and conviction of criminals. Many countries, e.g. the Netherlands, also allow their police to use the genetic profile of people they are seeking in order to draw conclusions as to the external appearance of the person concerned. So far this applies mostly to determination of sex, however the discovery of more genes could make it possible also to determine a person’s hair colour, eye colour and ethnic origin. Though at present such forensic tests are performed almost exclusively using PCR methods, DNA chips can play a useful complementary role in many such tasks and in future may be able to perform such tasks more rapidly and simply than PCR, thereby supplanting it in this application.
Sure identification: forensic medicine
Examples now exist of all these fields of application of DNA chips. In most cases commercial products are already available, though in some applications DNA chips are still in the developmental phase. As mentioned above, most interest is focused on basic research in biology and on medical research and drug development. In the latter field, use of DNA chips could initiate a change of direction towards a more personalised medicine that exploits the small but significant genetic differences that exist between people in order to develop new, more effective and safer drugs, especially for specific subpopulations. DNA chips that provide high resolution at a low price form the basis for the kind of rapid and simple genetic test that is essential for personalised medicine. They are thus important not just for finding genes responsible for diseases, but also for the development of new drugs, for correct diagnosis and for the choice of the most appropriate treatment for the individual patient.
Focus on medical applications
DNA chips: choosy fish hooks
105
Another medical application of DNA chips is their use in infectious diseases. In this application attention is focused not just on the genes of the patient, but more specifically on those of the pathogens. Many viruses (e.g. HIV), in particular, along with various other kinds of pathogen, develop resistance to important drugs extraordinarily rapidly via mutations in their genome. DNA chips can be used to examine the genome of such pathogens rapidly and reliably so that treatment that is optimal for the individual patient can be chosen.
One of the earliest examples of the use of DNA chips in medicine is in the treatment of AIDS. Human immunodeficiency virus (HIV), the pathogen of this disease, has an extraordinary ability to undergo change. Each of the small number of components of the virus can change so radically from one generation to the next that drugs rapidly become quite ineffective against the virus. In order to keep the virus in check despite this drug resistance, infected people have to take combinations of various drugs. For a long time the only way of finding out which variant of the virus was present – and therefore which drugs would be effective – in a given patient was by trial and error. The year 1996 saw the introduction of a DNA chip-based test by means of which the variants of a certain HIV gene present in an individual could be detected and drug resistances could accordingly be predicted. The intention was to make it possible for doctors to prescribe drugs which they knew to be effective against the HIV variant present in the individual patient, thereby avoiding a lengthy period of trial and error. As it turned out, this chip did not find a place in medical practice, and in fact sequencing by means of PCR has now become the most important method used for this kind of molecular diagnosis (see chapter on PCR). Nevertheless, in the past few years more chips have been developed to support AIDS therapy. These are designed to examine as many as possible of the important regions of the viral genome and thus to make the test applicable to other categories of drug. In future they could play at least an important complementary role to PCR in this application. DNA chips designed to identify viruses are also being developed for other infectious diseases. An example is the hepatitis C virus, which occurs in at least six different variants, each of which requires a different form of treatment (see chapter on molecu-
Leading the way in medicine: DNA chips in AIDS
106
lar medicine). DNA chips could – here again, along with PCR – become the most important means of distinguishing between these variants. A more recent development is a DNA chip to detect human papillomavirus (HPV). The two dozen variants of this virus cause genital warts, which are generally similar to the more familiar common wart and are similarly harmless. Nevertheless, three of the twenty or more variants of HPV can cause cervical cancer in women. Precise identification of the particular variant present in affected women is therefore of great importance. Up to now the condition of the neck of the womb has been assessed by means of a diagnostic smear so that any altered tissue can be removed. In extreme cases the entire uterus needs to be removed. A DNA chip now makes it possible to identify papillomavirus present in the patient’s blood and thus to estimate the risk of cancer more precisely. Women at high risk can thus adjust their family planning to their increased risk.
The ability of DNA chips to detect all the variants of a number of genes simultaneously could make them an important instrument for the investigation, diagnosis and treatment of cancer. The first products designed for use in this field have already been used successfully, and a large number of new chips are now in the developmental phase. It is becoming increasingly clear that use of DNA chips could greatly improve the survival chances of patients with many types of cancer, since it permits more precise adaptation of treatment to factors influencing the disease in the individual patient. However, successful use in this application presupposes the availability of more specific treatment options. When a tumour arises, the body increasingly loses control over the ability of the affected cells to undergo cell division. The cells change, the affected tissue starts to grow in an uncontrolled way and measures taken by the body to check the growth of the cells become less and less effective. Ultimately the cancer cells break away from their tissue of origin to form metastases, i.e. secondary growths in other parts of the body. For the process to advance to this stage, a whole series of control mechanisms have to be switched off, and this occurs mostly via changes in certain genes. Almost a hundred such ‘oncogenes’ are now known and new ones are still being discovered. The products of these genes generally occupy important positions in signalling pathways that regulate cell growth and division. Just as important a role,
Important application: cancer diagnosis
DNA chips: choosy fish hooks
107
Policing life and death: p53
• genetic damage • other signals mdm2
+ activation and synthesis
autoregulatory loop -
DNA repair
direct involvement
p53
indirect involvement other genes
+
other functions
+ bax etc.
apoptosis
p21
cell cycle held in G1 phase
The tumour suppressor gene p53 is one of the most important genes involved in the development of cancer. Its gene product (P53, written with a capital ‘P’) plays a central role in the growth and division of somatic cells. It is active especially when genetic damage is present in the cell, but it can also be activated by external signals. It is known to have at least three functions: 1. Control of the cell cycle. Cells divide in a regular pattern of events known as the cell cycle. If the genetic material of a cell is damaged, P53 holds the cycle in the G1 phase – a sort of resting phase – in order to permit repair of the DNA. The signal for this is transmitted via a number of proteins including P21. 2. Apoptosis. If the damage to a cell’s DNA is too great, P53 induces the cell to ‘commit suicide’. This process, known as ‘apoptosis’, stops occurring in cancer cells, with the result that they replicate in an uncontrolled fashion. P53 appears to be able to initiate apoptosis in various ways, including induction of the bax gene.
108
3. DNA repair. There is considerable evidence that P53 not only allows the cell time to repair its DNA, but also plays an active role in this process. Here again, it acts indirectly via p21 and other genes, however certain properties of P53 suggest that it also plays a direct role in the repair process. Because of its central role in the control of cell growth and division, p53 is an important target for cancer therapies. Attempts have been made, for example, to restore the function of altered P53 and to stimulate synthesis of this protein. The latter objective can be achieved, for example, via the mdm2 gene, which together with p53 participates in an ‘autoregulatory loop’: P53 stimulates formation of Mdm2, which in turn inhibits P53. In healthy cells this cycle stops excessive amounts of P53 from preventing normal division of cells. If the function of p53 is impaired, but not abolished, by genetic changes, drugs that act against Mdm2 may be useful, since in such cases more P53 is required in order to keep cells under control.
The decision as to whether such a drug will be useful therefore depends on the genetic profile of the patient concerned. Techniques that can identify each of the variants of p53 and of its genetic environment are therefore an important precondition for specific, rapidly-acting and effective therapy. The p53 GeneChip offers this possibility. On this chip are several thousand short DNA segments with which the p53 gene variants can be identified. The illustration shows the fluorescence pattern that results from such a test.
however, is played by genes with the opposite effect, i.e. genes whose function is to limit cell proliferation. Malfunction or complete loss of such ‘tumour suppressor genes’ opens the way to the development of cancer.
The most important tumour suppressor is the protein P53. This molecule polices the growth of cells and can even force cells to ‘commit suicide’ if their genetic material is too severely damaged (see box on page 108). The importance of this protein for cancer therapy is apparent from the fact that its function is disturbed in more than half of all human cancers. A whole series of drugs work by attempting to restore correct functioning of this tumour suppressor. Depending on the reason why P53 is no longer functioning correctly, different drugs may be required. Therefore, if the best treatment for a patient is to be found quickly, the genetic variants present in that patient must be ascertained. And that task can be made easier with the aid of a specially designed DNA chip. Chips designed to detect many other oncogenes and tumour suppressor genes are being developed at present, and some are already on the market. The aim of all such developments is to draw a genetic profile of the patient so as to assist doctors in the task of deciding which form of treatment is likely to be of benefit, and which not, in that particular patient. And this applies not just to the choice of the right drug: since in many cases there is no sure
The forces of law and order in cells
DNA chips: choosy fish hooks
109
means of determining how dangerous a particular ulcer is, many unnecessary operations are performed, while conversely many ulcers adjudged to be harmless are later found to be malignant. In these situations DNA chips can considerably increase the accuracy of diagnosis.
In fact, DNA chips can help not just when an ulcer has already been found. Cancer arises mostly as a result of an accumulation over decades of mutations that occur either randomly or as a result of radiation or toxins. Nevertheless, the likelihood of developing a certain type of cancer differs between individuals – even between individuals whose lifestyle and environmental circumstances are identical – because each individual has inherited a certain pattern of genetic changes. Knowledge of this genetic predisposition can therefore be very important: people who have inherited an increased risk of developing skin cancer, for example, should be more rigorous than others about avoiding exposure to sunlight and should undergo regular medical checkups. In future it will probably be possible to test for many such predispositions using DNA chips. Disease prevention is thus another potential application of this technology.
Use in preventive medicine
From the discipline of pharmacogenomics comes another current example of successful use of a DNA chip. This area of research is concerned with the interactions between our genes and drugs (see chapter on pharmacogenomics). It is based above all on the observation that the effectiveness of drugs varies greatly and that in some individuals drugs have dangerous side effects. Most such differences in reaction to drugs are at least partially due to differences in our genes. If the genetic causes of such differences can be ascertained, treatment can be adjusted accordingly. It may even be possible to develop special drugs for people with certain genetic characteristics. Such drugs would be expected to act more specifically and thus be safer and more effective. Cytochrome P450, for example, is important for the efficacy and tolerability of many drugs. This is a family of enzymes whose task it is to render water-insoluble substances – including many drugs – water-soluble. Above all, molecules are prepared for excretion from the body in this way (see also chapter on pharma-
First pharmacogenomic product: the AmpliChip CYP450
110
cogenomics). As the funcThe AmpliChip CYP450 tion of cytochrome P450 enzymes varies from person to The AmpliChip CYP 450 arrived on the market person, drugs are broken in 2003. The scientific down more rapidly in some basis of this chip is individuals than in others formed by pharmacoand their action in the body genomic data on the influence of the cytovaries accordingly. chrome P450 gene famAt least 50 separate genes ily on the efficacy and and hundreds of gene varitolerability of drugs. The AmpliChip CYP450 is able to identify the most imporants are now known to code tant variants of two important members of this group of for this family of enzymes, genes. and recently it has become possible to detect the most important variants of two important members of this group of genes by means of a DNA chip. The AmpliChip CYP450, which was developed jointly by the healthcare company Roche and Affymetrix, manufacturer of the GeneChip, is one of the first products developed on the basis of pharmacogenomic knowledge to become commercially available. The basis for the development of this product is knowledge of the influence of cytochrome P450 on the metabolism of drugs.
As in the case of their use in relation to cytochrome P450, DNA chips will in future find uses in many areas of diagnosis – namely wherever genes play a role in the genesis or development of a disease. Four such areas can be identified: z the metabolism – i.e. the absorption, conversion and breakdown – of drugs; the genes of the cytochrome P450 family of enzymes fall into this category; z the action of different genotypes in cancer, i.e. the genetic changes that play a role in cancer; the variants of p53 fall into this category; z a group of genes that play a role in the reaction of the body to infection; z genes that influence individual susceptibility to pathogens; the extremely rare cases in which a gene variant can prevent HIV infection are a well known example of this. In all these areas DNA chips can permit precise diagnosis of the genetic basis of a disease. With increasing knowledge of the
Outlook: use of DNA chips for diagnosis
DNA chips: choosy fish hooks
111
genes concerned and of the molecular basis of disease, such chips will in future contribute to earlier detection, more effective treatment and possibly even prevention of diseases. Before this prospect can become reality, however, a number of obstacles have to be overcome. For one thing, the vast majority of presently available DNA chip-based tests are too expensive for routine use. Also, the potential for individual further development of the method, for example in public or private research institutes, is severely limited by patents. Furthermore, the method still suffers from technical difficulties such as the question of whether the RNA used as a marker is measurable with a sufficient degree of accuracy in blood, the test liquid that has generally been employed to date. This is a precondition for what appears at present to be the most promising medical application of DNA chips, namely gene expression analysis. It is nevertheless to be expected that in the next few years DNA chips will assume an important role in general, and especially in medicine.
Pharmacogenetics: DNA chips in diagnosis
There are at least four areas in which DNA chips have a potential role in medical diagnosis: drug metabolism, genotypes of cancer, infection-related genes and susceptibility to pathogens. The most important genes currently known to play a role in these areas are indicated.
112
References Erlich H: Diagnostic applications of genomics. Talk given at Roche R&D Media Day, Munich/Penzberg, April 2002 Pedrocchi M: Multiprobe array systems for the analysis of human genes Certa U et al.: Biosensors in biomedical research: development and applications of gene chips. Chimia 53: 57–61, 1999 Sinclair B: Everything's great when it sits on a chip. The Scientist 13[11]:18, May 24, 1999 Baron D: Genomics und Proteomics mit Gen-Chips und Protein-Arrays. Pharmazeutische Zeitung 31/2001 DNA Microarray – Website of Leming, Shi: http://www.gene-chips.com/ Affymetrix – Website: http://www.affymetrix.com InformationsSekretariat Biotechnologie – Website: Genexpressionsanalyse http://www.i-s-b.org/wissen/broschuere/chip.htm
DNA chips: choosy fish hooks
113
Basic conditions: ethics, law and society
Every innovation has consequences, and every opportunity holds risks. And ambition always precedes ability. These generalisations apply also to personalised medicine, the road to which involves not just medicine and science, but also ethics, law and society. Controversies await us, but eventually there can be benefits for all.
In the 1990s the biosciences displaced physics and chemistry from the forefront of scientific innovation and change. Genetics, genomics, proteomics and related disciplines are generating impulses that have already had a significant impact on our daily lives and will do so even more in the future. At the same time, however, this new-found prominence has thrust the biosciences into the limelight of public attention: as with all technical and scientific innovations, their results have become topics of heated debate. At the centre of this debate are the changes presently taking place in clinical medicine – the field in which the practical applications of current research are most clearly palpable at present. The potential medical applications of these new disciplines are vast, but so too is the responsibility imposed by the use of such techniques in humans. In fact, the promise associated with molecular medicine can be realised only if certain basic conditions are satisfied. This applies to many areas of public, and also to some important areas of private, life. Political, economic and legal questions are important in this regard, as are ethical, societal and cultural considerations. Also, given the highly emotional debate currently raging about the use of genetics and genomics in medicine, there is a need to clear away a number of misconceptions which at present are standing in the way of a more objective assessment of the value of these new technologies. The differing interests of the various parties involved give rise to a complex area of tension, somewhere in the middle of which a broad social consensus needs to be found. Only then can the great opportunities provided by a molecular understanding of diseases be turned to practical benefit.
The first, and most important, of the various interest groups involved is that of patients. Without their acceptance, no change is possible. Their expectations of medicine are based on the elementary need for preservation or restoration of individual health – a need that medicine has traditionally sought to satisfy with conventional remedies and is now attempting to satisfy with the fruits of molecular research. Newly acquired understanding of the molecular basis of diseases is being used to combat diseases earlier and more effectively and in some cases to prevent them from occurring in the first place. Molecular biology thus has the potential to be of great practical benefit to patients.
Help and protection: the interests of patients
116
Equally important to individual patients, however, is the price of such progress – not just in economic, but also in social and moral, terms. And it is precisely this last point that concerns
many patients: they want to be sure that new understanding of diseases and the new possibilities in medicine that result from this understanding will not lead to injustices. The availability of such options should not be limited to the rich, nor should anybody be restricted in terms of their choice of occupation, find their individual freedom limited or be socially stigmatised on the basis of test results.
The aim must therefore be to exploit the opportunities, while identifying and limiting the risks, that arise from newly acquired medical knowledge. At the core of more personalised medicine, for example, is the need to know and analyse the individual characteristics of patients. Only on the basis of this knowledge is it possible, for example, to develop medicines whose use is based on genetic criteria and which therefore are markedly more effective and better tolerated in certain patients than are currently used medicines (see chapter on pharmacogenomics). At the same time, however, this knowledge inevitably reveals – to the medical staff involved, to the dispensing pharmacy, to the patient’s healthcare insurance fund, etc. – an intimate detail of the genome of the patient concerned, namely the genetic variation upon which prescription of the drug in question is based. There is nothing fundamentally new about this: for example, the fact that a person gets regular prescriptions for insulin identifies that person unmistakeably as a diabetic. Many other forms of treatment likewise reveal people’s diseases.
Exploiting opportunities, limiting risks
Basic conditions: ethics, law and society
117
The mere fact that information on the diseases present in an individual has to be acquired and to some extent passed on to third parties is therefore no great problem in itself; much more prob-
lematic in this regard is the nature and quality of the information that is passed on, and in particular the possible predictive value of the results of genetic tests. In practice, however, any predictions of this kind will in the vast majority of cases be limited to a statement as to whether a particular patient is likely to respond to a particular drug or group of drugs. More detailed or specific information is generally not to be expected from such tests. The situation is different, however, in the case of tests that can determine a person’s likelihood of developing certain rare hereditary diseases long before the actual onset of illness. This predictive ability is most certainly highly desirable in the case of diseases for which drugs will in future be developed that can not only cure the disease, but also prevent, or at least delay the onset of, the disease in individuals identified as being at risk. Preventive medicine should, must and will become increasingly important in the future. Nevertheless, in some cases, the implications of genetic testing for the patient as well as for family members may be profound as for example in some rare familial diseases, such as Huntington’s disease, cystic fibrosis or hemophilia. In these cases, individuals and their families may perceive the resultant information as stigmatising.
118
Therefore, if the individual patient is to be convinced of the value of more personalised medicine with all its implications, a series of conditions must be satisfied: z New, better targeted drugs should represent a definite therapeutic advance, especially in the case of diseases whose treatment has hitherto been unsatisfactory or nonexistent. Personalised medicines should be more effective and better tolerated than presently available drugs. These two requirements apply mostly to research. z New treatments must be marketed at a realistic price, i.e. a price that is reasonable and appropriate to the medical value of the drug concerned, and must be available to patients via their regular healthcare services. z People must be protected against any form of discrimination arising from genetic or molecular tests – especially as the terms ‘health’ and ‘illness’ have no absolute meaning (see box). The legal basis for such protection must be established by politicians in representation of society as a whole.
Varied requirements
Freedom of choice: focus on dignity The debate about the ethical aspects of the future of medicine centres around two questions: preservation of the dignity of patients and protection from discrimination. The following points, among others, need to be considered: z Freedom of choice. Patients must retain sovereignty over their physical and mental wellbeing. This includes the right to decide whether to receive a certain form of treatment or to undergo a certain diagnostic investigation, i.e. the right to know or not to know. An available medical option may neither be forced upon, nor withheld from, a patient. However, such choices can only be free if the patient has first received extensive information and non-directive counselling. z Limits to freedom of choice. The sovereignty of an individual ends where that of another begins. Since genetic factors are inherited, the results of genetic tests are always relevant to the close relatives of the person tested. The question of whether, and if so in what cases, relatives should have a say in a patient’s decision to undergo a certain genetic test therefore needs to be clarified. In practice, however, this problem is probably limited almost exclusively to a small number of classical hereditary diseases,
and even in them it probably arises only exceptionally. In all other medical situations genetic testing adds little to what is already known from a family’s medical history. z Patients who are not competent to make decisions. Cases in which a patient is unable to assess the arguments for and against undergoing a particular medical treatment are already covered by clear rules. Nevertheless, the possibility of prenatal diagnosis and therapy is new in this respect. A legal framework for the performance of prenatal procedures needs to be established on the basis of a social consensus. z Discrimination. When told they have a certain disease, the vast majority of people still feel a sense of stigmatisation and suffer anxiety about a possible loss of personal freedom. A major rethink is required here. It is clear from the complexity of our genome that terms such as ‘health’, ‘illness’, ‘normality’ and ‘abnormality’, have no absolute meaning at the molecular level. Every human being carries both promoting and protecting factors for all diseases. Therefore, in almost every case, all that a molecular diagnostic test can do is indicate disease probabilities. See also ‘Roche Charter on Genetics’.
Basic conditions: ethics, law and society
119
z The individual’s freedom of choice must be preserved in all
cases. This applies not only to choice of treatment but also, and in particular, to diagnosis, i.e. to the right to know or not to know. This requirement applies to the public and also to politicians as representatives of the public.
z Out of these requirements comes another requirement: suc-
cessful introduction of new medical techniques will depend to a large extent on considerably better education of the public regarding the potential, and also the limits, of genetics in medicine.
The requirements of individual patients for customised medicine differ from the interests of society as a whole. In many cases the interests of these two sides appear, at least at first glance, to be in direct conflict, e.g. in relation to the sharing of burdens and risks between individual patients and society. Like the question of confidentiality of patient data, however, this is not a fundamentally new issue, and in most industrialised countries the problem is regulated by means of a greater or lesser degree of redistribution of medical costs from the individual to the public purse. In terms of economics, therefore, molecular medicine is simply a therapeutic innovation whose costs and benefits have to be weighed against each other and whose impact on the balance between public and private funding of medicine needs to be considered. More problematic, however, is the question of how acquired data should be handled, i.e. how patients can be protected against misuse of their personal genetic information, in particular. Current data protection legislation is mostly aimed at minimising the amount of data that are collected, stored, distributed and
Society as a whole: the public interest
120
analysed. Pharmacogenetically guided therapies will require the collection of additional data. This will necessarily result in the gathering of sensitive information about individual patients, and this information will then be revealed indirectly to a broader circle of people via the medicines that are prescribed for those patients. The task of directing the use of sensitive personal data into the appropriate channels falls to politicians in their role as representatives of society as a whole and the many special interest groups within it. They have to mediate between conflicting requirements and interests and work to build the broad social consensus that will be required if the new possibilities offered by molecular medicine are to be exploited in a responsible fashion.
In the crossfire of conflicting interests in healthcare, criticism is often levelled at industry and commerce. This is due above all to the fact that companies operating in the healthcare sector seek to make a profit out of health, a treasured asset both of individuals and of society. However, this sweeping criticism overlooks the central role played by commercial enterprises in satisfying the requirements of the various interest groups. After all, the pharmaceutical and diagnostics industry strives to meet the need of patients for better targeted and safer treatments, and at the same time makes an important contribution to human society and culture via the research that it undertakes. And because it is highly innovative, it has to accept a high degree of responsibility. Nevertheless, the changes that are about to occur in medicine certainly raise questions for the affected industrial sectors – and do so precisely because of the special responsibility that these sectors bear towards patients. Because whether and at what price a medicine can eventually be offered for sale depends on a num-
Between economics and responsibility: commerce
Basic conditions: ethics, law and society
121
ber of very complex factors. In the first place, the process of developing, testing and obtaining marketing authorisation for new drugs is set to become increasingly protracted and expensive. The research that underpins the development of new drugs is also expensive. Even though the drug regulatory authorities of many countries are presently working on ways of simplifying drug registration procedures, at least for particularly important and innovative new drugs, it is becoming increasingly difficult for companies to recoup the development costs of new drugs. Pharmaceutical companies are also subject to multiple constraints in relation to the pricing of new medicines – firstly as a result of their need to recoup the development costs of the drugs concerned, and secondly as a result of the price controls that are imposed in many countries. Therefore, if the development of more personalised forms of therapy for necessarily smaller target groups is to be made economically viable, it may be necessary to adopt new patterns of thinking and devise new forms of cooperation between industry and healthcare funding authorities.
A key role in the restructuring of medicine that is taking place at present is played by research. It lays the foundations for more individualised prevention and treatment of diseases, and is therefore subject to specific demands. Of central importance in this regard is the safety of treatments and the correctness of the diagnoses on which they are based. And the central problem here is the unpredictability of nature. New treatments need to be more effective or better tolerated than presently available treatments. At least in certain diseases, this objective can certainly be achieved with the aid of pharmacogenetics and molecular diagnosis. More of a danger are the unrealistic expectations that can arise as a result of overenthusiasm for the new possibilities.It must be remembered that molecular medicine is no magic wand. Side effects of drugs will still exist in the future, as will fluctuations in effect. And the disappointment felt by patients whose pharmacogenetic profile indicates that the only available form of treatment for their disease is likely to be ineffective in them has plenty of potential to create new conflicts. Therefore, in order not to arouse unrealisable hopes in the first place, medical science needs to be cautious in what it says about its possibilities, which are and always will be limited.
Research: laying foundations and informing the public
122
The same applies to diagnostics. The common perception that the results of genetic tests, in particular, are immutable and invariably correct in terms of what they predict creates a whole series of problems – even though this perception is incorrect in the vast majority of cases. The fact that these tests, like all biological test methods, are subject to error and that their predictive value is limited by the current (always unsatisfactory) state of scien-
tific knowledge is axiomatic to scientists, but by no means obvious to patients. Here again, frank and objective education of the public is required.
Genetic testing and molecular diagnostics therefore can and must be accompanied by better education of the public. Education alone, however, will not convince the public of the value and purpose of the new possibilities in medicine. Of more use in this regard are instruments that allow to define their potential risks and benefits. Such instruments have in fact already been available for a long time. This too illustrates the point that molecular and pharmacogenetic tests do not differ fundamentally from the standard techniques used in medicine today: they too can be assessed in terms of parameters such as sensitivity and specificity that provide an objective measure of the informational content of test results. For example, a genetic test whose purpose is to spare patients from a life-threatening side effect of an important drug must satisfy the following criteria: 1. Specificity: The test must reliably indicate that a person found to have the observed gene variant(s) will develop the side effect in question. The less specific the test (i.e. the high-
Probability, not certainty
Basic conditions: ethics, law and society
123
er the proportion of bearers of the gene variant who do not develop the side effect), the more patients will be unnecessarily deprived of the important drug. 2. Sensitivity: The test must identify as high a proportion as possible of people who will develop the side effect. Patients who develop the dangerous side effect even though their test result was negative have not been helped by the test.
Most genetic tests are likely to have an informational content similar to that of conventional laboratory tests (e.g. plasma cholesterol level). In the vast majority of cases, therefore, genetic tests should be regarded and used in exactly the same way as conventional tests. Like these, they can indicate probabilities that may be quantifiable to some extent, but they cannot provide certainty. Only exceptionally, e.g. in rare, classical hereditary diseases such as Huntington’s chorea, do genetic tests yield results with a predictive value of almost a hundred percent. However, even an objective evaluation of tests on the basis of their informational content can serve only as an aid to the decision on whether a particular molecular or genetic diagnostic test should be performed in a particular patient. Ultimately, such decisions must be made by patients themselves in consultation with experienced physicians, and the welfare of the patient must always be the primary consideration. Only in this way is it possible to ensure that advances in medicine bring benefit and do not cause harm.
124
References Schreiber H-P: Humangenomforschung, Gentechnik und Gesellschaft Schreiber H-P: Human-Genom-Forschung und die Notwendigkeit eines sozialverträglichen Umgangs mit genetischem Wissen Lindpaintner K: The importance of being modest, or: How good is good enough? – Reflections on the pharmacogenetics of abacavir. Pharmacogenomics 3: 835–838, 2002 Genetics in Discovery and Development. Roche’s Ethical Principles including Roche Charter on Genetics. Roche, 2000 Roche Biomarker Program: Economic impact of genetics, 2004. Internal document
Basic conditions: ethics, law and society
125
Prospects: more knowledge for medical science
Medicine is changing. New techniques and knowledge derived from genetics, genomics and proteomics permit a deeper understanding of the molecular causes of disease. Increasingly, medicine will focus on differences between patients and become more personalised. It will become ever clearer that no two illnesses are the same and that even useful forms of treatment are not effective in all patients.
Medical science is in the grip of change. Genomics, proteomics and other branches of molecular biology are generating a stream of new findings, and modern technology has introduced techniques of miniaturisation, automation and parallelism into research and development. The entirely new field of molecular diagnostics promises to have a lasting impact on therapeutic practice. And medical science is increasingly realising that apparently identical clinical pictures can have entirely different underlying causes requiring personalised treatment. These developments also have a commercial side, pitting established pharmaceutical companies against young biotech firms in a race to discover suitable target molecules and new drug compounds. At the same time, the development of new drugs up to the stage of regulatory approval is becoming lengthier and more expensive. Traditional drug research is growing riskier in economic terms, and it is becoming more difficult for it to contribute to genuinely significant innovations. On top of this, despite some minor successes, the options available for treating many of the major common diseases remain unsatisfactory. A period of radical change is imminent.
Behind this upheaval is the recognition that no two illnesses are the same. It is now clear that with the exception of a handful of hereditary diseases and some severe infections, very few human diseases have a simple or even a single cause. And even in the exceptions just mentioned, which include, for example, cystic fibrosis and hemophilia as well as tuberculosis and AIDS, the severity of the symptoms varies so much from one patient to the next that a clinical picture of some complexity has to be assumed. After decades of genetic research and several years of genomic investigation, we now know that a patient’s genetic predisposition plays a significant role in the progression of almost all illnesses. In the case of infections, another factor can be the variable genetic makeup of the pathogens involved. These findings are neither new nor surprising. And yet they confront medical science with a daunting problem. Until now the principle of ‘one disease – one treatment’ has essentially held sway unchallenged. But if no two illnesses are the same, many of the treatments used must be wrong or at least inappropriate. Though fine diagnostic distinctions have always been a driving force of medical progress, the sheer amount of newly acquired
No two illnesses are the same
128
knowledge is now enormous. This calls for a rethink in many cases. The indications for existing drugs will become narrower, and the discovery of new drugs will be all the more crucial. And distinguishing between subtle variants of a disease instead of general clinical pictures will require a new molecular diagnostic approach.
In fact, for the first time in the history of medical science diagnostics and therapy are meeting on common ground – at the molecular level. Whereas drug therapies have always acted on the molecular network of our bodies, the diagnoses upon which those therapies are based have generally been made on the basis of physiological factors as evidenced by physical signs and symptoms. Thus, the physician observes the consequences of a disease in order to treat its causes – an approach that fails to do justice to the true complexity of the vast majority of diseases. New biological test methods have now made it possible to routinely determine the actual molecular causes of diseases at the patient’s bedside. The newness of this development is indicated by the word ‘routine’. Of course, the molecular background of diseases has long been investigated in patients. But until now these research methods have been too expensive and elaborate for routine use. Today that has all changed thanks to the widespread availability of PCR, especially its quantitative form, and increasingly sophisticated DNA chips. In the coming years these techniques will complement, and in some cases even supplant, conventional diagnostics in many areas of medicine – a process that is expected to build up a huge momentum of its own: the more widely PCR and DNA chips are used, the cheaper and more varied they will become; this in turn will accelerate their spread. Molecular diagnostics is therefore playing a key role in the rapid advances now taking place in medicine. It has a broad range of applications: z New therapies: The use of new drugs that take account of the complex causes of diseases calls for differentiated molecularbased diagnostics. z Screening: Screening tests are already an important instrument in healthcare. Cancers, abnormal blood sugar, abnormal bone density and abnormal blood pressure are all currently screened for. As our knowledge deepens, more such
Consequence: a new role for diagnostics
Prospects: more knowledge for medical science
129
target sequence DNA strand double helix coiled DNA
chromosome
PCR makes it possible to amplify specific DNA sequences.
tests will be developed to cover other common diseases such as osteoarthritis, schizophrenia and epilepsy and infectious diseases such as hepatitis C and tuberculosis. Molecular diagnostics based on PCR, DNA chips and other techniques has the potential to significantly increase the accuracy, and thus the acceptance, of screening tests. This, in turn, will drive forward their development and encourage widespread use of them. z Gene tests: Genetic predisposition is an important factor in the prognosis and treatment of diseases. Gene tests already make routine use of PCR and to an increasing extent DNA chips as well. z Hygiene: A marginal aspect, but one that promises to gain importance in the coming years, is the search for and identification of pathogens in hospitals. Again, modern techniques such as PCR and DNA chips are the fastest methods for detecting the presence of pathogens before they can spread.
The recognition that diseases can have entirely different causes despite producing the same symptoms is not new. What is new is the molecular biological understanding that now makes it possible to examine the genetic differences between individual patients and the effects of these differences on treatment. In other words, no
No two treatments are the same
130
two treatments are the same. A drug might be right for one patient but wrong for another, even though both patients have the same illness, because drugs can vary in their efficacy and tolerability in different individuals. The field of pharmacogenetics has been investigating the reasons underlying this phenomenon for over a hundred years now, but only recently have molecular genetic techniques made it possible to apply these insights to clinical medicine. Pharmacogenetics is now threatening to upset the second half of the dogma of ‘one disease – one treatment’. In future the choice of the right treatment will depend not just on the disease diagnosed, but also on the way in which each patient’s body deals with the drugs in question. To make this kind of choice possible, two closely related factors need to be taken into account: z Genetic factors: Pharmacogenetics is concerned with the relationship between the gene variations and the body’s response to drugs. Genetic differences can cause drugs to be absorbed, metabolised or excreted too rapidly or too slowly. Or they can prevent sufficient drug from reaching the target site. Or they can give rise to adverse or even dangerous side effects. Ruling out such genetically caused uncertainties relating to the efficacy and safety of drugs will be one of the major challenges facing pharmaceutical researchers in the coming decades. z Environmental factors: External factors are at least as important as genetic factors in determining the efficacy and safety of drugs. Prominent among these factors is diet. Elements of our diet can interact with drugs, accelerating or preventing their uptake and affecting their excretion and utilisation. The same applies to interactions between different drugs, which can enhance or reduce each other’s effects and exacerbate each other’s side effects. External stress factors such as physical and mental fitness, environmental toxins, radiation, temperature and so forth can also influence the efficacy and safety of drugs. In practice, the environmental influences to which a patient is exposed cannot be exhaustively determined; also, they vary over time – though this means that they can be influenced. This is not true of gene variants. It is therefore all the more imperative to recognise how environmental factors influence the way the body interacts with drugs.
Prospects: more knowledge for medical science
131
If future therapies are to be based on genetic factors, medicine will inevitably become more personalised. However, the term ‘personal’ in this context does not mean that at some time in the future patients will have their own tailor-made therapy. Rather, it means that a far broader range of therapeutic options will be offered from which doctors can select the one most suited to their individual patients. Of course, such choices are already available, at least for some diseases, however the number of such choices will increase and so too – hopefully – will the success of therapy. As an inevitable consequence of this development, the target groups for drugs will become smaller. The indications for new drugs will be determined not only by the molecular causes of the diseases being treated, but also by the pharmacogenetic profile of the individual patients. This is unexplored territory in pharmacology. In future, therefore, patients will be able to expect that a drug that is prescribed for them is more likely to be truly suited to them than at present. The effects of almost all currently used drugs can vary to a greater or lesser extent, and in extreme cases lack of efficacy is even the rule (see box). The safety of many currently used drugs is similarly unsatisfactory; for example, some three patients per thousand die each year from severe side effects of major drugs. This figure needs to be reduced, since even the occasional occurrence of severe side effects can be acceptable only if the disease concerned is relatively rare and unresearched and therapeutic options and experience are correspondingly limited.
Consequence: the personalisation of medicine
Great fluctuations: efficacy of drugs
drug group
poor efficacy
AT2 antagonists
10 – 25%
SSRIs
10 – 25%
ACE inhibitors
10 – 30%
Beta-blockers
15 – 25%
Tricyclic antidepressants 20 – 50%
132
Statins
30 – 70%
Beta 2 antagonists
40 – 70%
The efficacy of drugs is often unsatisfactory and may fluctuate in the extreme. The table gives the incidence of poor efficacy for several major drug groups. Angiotensin II (AT2) antagonists, like ACE (angiotensinconverting enzyme) inhibitors, are antihypertensive drugs which are widely used in the treatment of heart failure. The same applies to beta-blockers. Selective serotonin reuptake inhibitors (SSRIs) are used as psychotropic drugs, especially in obsessive-compulsive disorders and, like tricyclic antidepressants, for depression. Statins, also known as HMG-CoA reductase inhibitors, lower cholesterol levels. Beta 2 antagonists are important antiasthma drugs.
Doctor’s responsibilities will grow accordingly. They will have to deal with entirely new diagnostic resources, a considerably expanded range of therapies and – as is already evident from the growth of the Internet – far better informed and more self-confident patients.
This means that the demands made on medical science will increase. One innovation gives rise to another. Personalised therapies require individual diagnoses. Molecular diagnoses call for differentiated therapy. And both aspects, diagnosis and therapy, depend on rapidly expanding technological possibilities. In fact, a synthesis is taking place at the moment: research and development, diagnosis and therapy, information and prevention are evolving together. The key to successful healthcare lies in integrated medicine. If the new possibilities of medical science really are to bring about progress, they must mesh smoothly. The concept of diagnosis will need to be extended beyond symptoms and clinical findings to include the molecular underpinnings of diseases and their treatment. Also to be considered is the hitherto relatively undeveloped field of prevention, which in most cases is still limited to fresh air and a healthy diet. Testing, i.e. diagnosis, of genetic predisposition will play a far greater role here in future. It will also make it possible to provide patients with more specific counselling – such as is already available, for example, in relation to high serum cholesterol levels. Treatment follows seamlessly. The earlier a disorder is discovered, the easier it is to treat – a long-recognised fact that can take on new relevance in connection with the possibilities of early molecular diagnosis. This is especially true when specific diagnosis is matched by a corresponding range of personalised therapies. Progress will be achieved only if both sides move forward together. The interplay of these developments requires a high degree of coordination, information and above all cooperation. At the same time, these developments will give rise to new ethical, societal and legal challenges and issues.
Consequence: integrated healthcare
Prospects: more knowledge for medical science
133
For the pharmaceutical industry, these developments impose the need for a continuous rethink. A new order will prevail in the healthcare market, where the changes are already in full swing. New strategies, alliances and competitions are emerging: 1. Integration of diagnosis and treatment: The more finely differences between individuals are distinguished and considered, the more difficult it is to separate these two poles. Close cooperation is required here: drugs whose prescription depends upon pharmacogenomic considerations will be prescribable only if a corresponding means of testing is available. Thus, a specific genetic variation first has to be identified in the patient so that a drug geared to this variation can be sensibly used. And because the development of diagnostic tests and therapy are to some extent interdependent, companies with expertise in both these areas will find themselves at an advantage. Expertise therefore needs to be gathered together either within a single company or else by means of close alliances between companies. The traditional boundaries between diagnostics and therapy will therefore largely disappear. 2. Greater development risk: The fact that the available options for treating most of the major common diseases are still unsatisfactory means one thing above all: Pharmaceutical companies will have to be more willing to take the risks associated with the development of new drugs with new mechanisms of action. Certainly, the future will continue to hold the occasional surprise, as when a well-established drug is found to possess previously unsuspected beneficial properties. But for the most part, medical progress will depend on the exploration of new avenues – particularly via new target molecules, which are already the most hotly contested objects in medical research. Above all, new diagnoses, new targets and new drug groups mean considerably stepped-up research and development efforts with an undiminished risk of failure. Nevertheless, the effort may well be worthwhile. Successful developments that address unmet medical needs have a huge sales potential. 3. Smaller target groups: The advent of more personalised medical care inevitably means that a new drug can only be sensibly used in a limited number of patients. This limits sales possibilities, thus making it more difficult to recover the costs of research and development. However, the development of such drugs also has advantages. For example, drugs developed
Consequence: upheaval in the pharmaceutical industry
134
in this way are more effective thanks to their targeted activity. This should reduce the risk of failure in later stages of development while increasing acceptance among patients and thereby reducing the number of patients who stop treatment. The actual investment-to-yield ratio can be very attractive. 4. Competition by biotechnology: Young biotechnology companies backed by considerable venture capital are invading the market for new drugs developed on the basis of molecular findings. Typically these companies develop drugs characterised by a high risk but huge sales potential. The pioneers of the sector, e.g. the American companies Amgen and Genentech, have long been on an equal competitive footing with the traditional pharmaceutical companies, which for their part have almost without exception risen to the new challenges – either by establishing their own biotechnology departments or entering into alliances or acquiring promising innovative companies in the sector. 5. Increased demands: New opportunities bring new responsibilities. In the not-too-distant future, pharmacogenetic data will certainly form part of the data required by health and drug regulatory authorities. In addition, after a period of adjustment, patients are likely to become more demanding in terms of the efficacy and safety of the drugs they take.
Internationally active healthcare companies will not escape this trend. On the contrary, active participation in this process of change is fundamental to their survival, whereby the term ‘change’ does not imply a revolution, but rather a systematic evolution towards more informative investigations and more effective and safer drugs. The fact that many years of laborious and detailed research work are required before personalised diagnostic tests and drugs can be developed is evidence enough of the evolutionary nature of this change. Also, in many cases a distinction will have to be made between what is feasible and what is reasonable, desirable and economically sound. For instance, the size of a patient group above which the development of drugs specifically for it becomes economically viable still cannot be predicted – at least, not until new forms of cooperation between society and industry, such as ‘orphan disease’ programmes for particularly rare diseases, have been set up.
High hurdles, high goal
Prospects: more knowledge for medical science
135
Nevertheless, progress is opening up far-reaching new opportunities for medicine at the scientific and technical levels. Personalised diagnosis and treatment promise to be substantially more effective with substantially fewer side effects. At the same time they can tackle the causes of diseases whose treatment has until now been only symptomatic and often inadequate. Notwithstanding all the commercial and ethical imponderables, in a certain sense the new possibilities also impose a moral obligation to apply the new findings of molecular medicine for the practical benefit of patients.
136
A brief glossary of terms
fi
Amino acids the chemical building blocks of proteins. They consist of a constant region, containing an amino group and an acid group, and a variable region. At least 20 different amino acids occur in nature. Antibodies Y-shaped proteins that are part of our immune system. They bind to foreign substances in the body and thereby mark them for destruction. Antioxidants molecules that trap dangerous, highly reactive oxygen compounds in the body and thereby render them harmless. Vitamins C and E, for example, are antioxidants. Apoptosis ‘programmed cell death’. Cells whose genetic material is irrevocably damaged or altered ‘commit suicide’ in order to protect the rest of the body from the effects of the genetic alteration. Autosomal dominant inheritance an inherited characteristic that is expressed if it is present on either one of the two autosomes of a particular kind. Autosomal recessive inheritance an inherited characteristic that is expressed only if it is present on both autosomes of a particular kind. Autosomes chromosomes not involved in sex determination. Humans have two of each kind, inherited from the mother and the father respectively. Altogether there are 44 autosomes (twice 22). Bases chemical substances that have a basic (alkaline) action. The bases of DNA are the fundamental building blocks of the genome: adenine (A), thymine (T), guanine (G) and cytosine (C). When present on two strands of DNA, the bases join to form stable pairs. In nature, base pairs form only between A and T and between G and C. In RNA, thymine is replaced by uracil, which likewise pairs with adenine. Biochip a solid substrate (e.g. glass or plastic) upon which biomolecules are anchored. Bioinformatics the in most cases computer-assisted analysis of biological data by special databases, applications and programs. cDNA complementary DNA; DNA transcribed enzymatically from RNA (mostly mRNA). Cell the smallest independently viable unit of an organism. Chromosomes tightly packed DNA strands with associated proteins that are present in the cell nucleus and that function as bearers of genetic information. The human genome consists of 23 chromosome pairs (22 autosomes and one of the two sex chromosomes X and Y). Complementary DNA The building blocks of DNA and RNA form specific pairs. Two strands whose building blocks form a sequence of perfect pairs are able to form a stable double strand and are referred to as complementary strands. Denaturing a process induced by heat or chemicals in which a biomolecule (e.g. DNA, RNA or a protein) loses its natural form. DNA deoxyribonucleic acid, the chemical substance of which our genetic material consists.
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
DNA chip a biochip with single-stranded DNA as the probe. Enzyme a biological catalyst, generally a protein, that can accelerate and combine certain chemical reactions. Exon a sequence of a gene that acts as a direct blueprint for a gene product. Gene array a special type of DNA chip. GeneChip a widely used DNA chip developed by the US company Affymetrix. Genes functional segments of our genetic material that serve mostly as blueprints for the synthesis of proteins. Genetics the study of inheritance; deals with the laws of inheritance and the properties of genes, including the transmission of specific variants of a gene from one generation to the next. Genome the totality of the genetic material ( genes) of an organism. Genome chip a special kind of DNA chip. Genomics the systematic study of the form, function and interactions of the genes that comprise the human genome. Genotype the alternative forms (alleles) of a gene present in an individual; generally there is a maximum of two – one from the father and one from the mother. High-throughput screening a highly automated method of identifying potential drugs in chemical libraries. Hybridisation the joining of two complementary DNA (or RNA) strands to form a double strand. Intron a sequence of DNA situated between the exons of a gene that is cut out of the corresponding mRNA before this is translated into the gene product. Metabolism the transformation of chemical substances in the body or within a cell. Microarray a widely-used synonym for DNA chip. mRNA messenger RNA, the working copy of a gene that acts as a blueprint for the synthesis of proteins. Unlike DNA, it is able to leave the cell nucleus. Nucleic acids generic chemical term for DNA and RNA; chain-shaped molecules whose individual building blocks are bases ( nucleotides). Nucleotides the building blocks of DNA and RNA; they comprise the four bases adenine, thymine, cytosine and guanine (A, T, C, G; in RNA thymine is replaced by uracil [U]), a sugar and at least one phosphate group; without the phosphate group these building blocks are referred to as nucleosides. Oligonucleotides short nucleic acid chains composed of at most a few dozen building blocks ( nucleotides). Oncogene a gene that plays a role in the development of cancer. Pharmacodynamics the study of the interactions between drugs and their molecular targets. Pharmacogenetics describes the influence of gene variations in individuals on the efficacy and side effects of drugs. Pharmacogenomics studies interactions between drugs and the genome.
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi fi
fi
fi
fi fi
fi fi
fi
fi fi
fi
fi
fi fi fi
fi
fi
fi
A brief glossary of terms
137
Pharmacokinetics the study of the uptake, conversion and breakdown of drugs in the body over time. Environmental factors, diet and genetic predisposition all play a role. Phenotype the constitution of a living creature that results from its genotype and environmental influences. Polymerase chain reaction (PCR) a technique for rapid copying (amplification) of even minute amounts of DNA. Polymerases enzymes that link individual nucleotides together to form long DNA or RNA chains. Polymorphism existence in more than one form; in genetics, a region of DNA in which differences in the sequence of building blocks occur in a relatively large number of people. Primer a short DNA fragment with a defined sequence that serves as an attachment and extension point for polymerases. Promoter a region of DNA immediately before a gene that contains the starting information for transcription of that gene. Protein a molecule consisting of a chain of amino acids. Because of the variety of their building blocks, proteins can differ greatly in form and function. Proteome the totality of the proteins of an organism. Proteomics the study of the form, function and interactions of all the proteins of a tissue or organism. Rational drug design computer-assisted design of new drugs. RNA ribonucleic acid; the chemical substance of which, among other things, working copies of genes ( mRNA) consist. Sequence the order of the nucleotides in DNA (DNA sequence) or RNA (RNA sequence). Sex-linked inheritance an inherited characteristic transmitted via one of the two sex chromosomes (X or Y). SNPs single nucleotide polymorphisms – differences in individual building blocks (base pairs) of DNA that are distributed randomly over the genome and passed from generation to generation. Targets the molecules, mostly proteins, upon which drugs act in our body. Template in molecular biology, mostly a fragment of DNA that acts as a chemical template for polymerases. Transgenic animals animals containing genes derived from other species. Tumour suppressor a molecule which, when functioning correctly, prevents cancer from developing.
fi
fi
fi
fi fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
138
fi fi
fi
Cover picture Conceptual computer artwork. Part of a DNA molecule, chromosomes, a DNA autoradiogram and the triplets of nucleotide bases that code for amino acids in a protein. Source: Mehau Kulyk, Science Photo Library©
Published by F. Hoffmann-La Roche Ltd Corporate Communications CH-4070 Basel, Switzerland
© 2007 Third edition All trademarks mentioned enjoy legal protection. Any part of this work may be reproduced, but the source should be cited in full. This brochure is published in German (original language) and English.
English translation: David Playfair Layout:
Atelier Urs & Thomas Dillier, Basel
Printers:
Gremper AG, Basel
7000632-2
140