Genes and Health.pdf

Document Sample
Genes and Health.pdf Powered By Docstoc
					                   We Innovate Healthcare

Genes and Health
Genes and Health


                                             1                 2

                                                                                                        Molecular medicine:
                3    4    5    6        7        8        9        10   11   12       13   14

                15   16   17   18      19        20       21       22   23       24   25   26
                                                                                                        genetics, genomics and
                                                                                                        proteomics for
                                                                                                        diagnosis and therapy      7
                                    To d a y
                                    To d a y

                                             A                                                 B
                                             A                                                 B



                                    To m o r r o w                 Diagnostics
                                    To m o r r o w                 Diagnostics


                                                                                                        genes and drug response   19

                                                                                                        seeing through the
                                                                                                        undergrowth               31

                                                                                                        for medicine              47

                                                                                                        an outstanding method     65

                                   G         C        A            A     T        C        T       A
                                   G         C        A            A     T        C        T       A
                                   G         C        A            A     T        A        T       A
                                   G         C        A            A     T        C        T       A
                                                                                                        the great importance
                                                                                                        of small differences      83
DNA chips:
choosy fish hooks          95

Basic conditions:
ethics, law and society   115

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 differen-
             ces 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 identifi-
             cation 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 con-
             cerned. 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 pro-
             duction 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 sub-
             stances that exist in the human body. This is leading to the de-
             velopment 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

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 attempt-
ed to cast light on current developments in medicine and to ex-
plain 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 conduc-
ted at present about the uses of genetics and genomics in medi-

Molecular medicine: genetics,
genomics and proteomics for
diagnosis and therapy

                                                                    1             2

                                            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

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.
                           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 ex-
                           actly 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 med-
                           icine upside down. They gave a new meaning to the concept of
                                                                    diagnosis and in so doing
    Terms                                                           took their craft closer to the
    Molecular medicine the application of molecular biology
                                                                    causes of disease. Even in
    (in particular, genetics, genomics and proteomics) to medicine. those days every medical
    Genetics the study of inheritance; deals with the laws of in-   treatment was preceded by
    heritance and the properties of genes.
    Genomics the study of the form, function and interactions of    an examination; for over
    the genes of an organism.                                       2000 years an ‘imbalance of
    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 di-
                           agnosis was generally made on the basis of a handful of symp-
                           toms and signs – the art of diagnosis had no more than that to
                           offer. In the middle of the 18th century scientists such as the Ital-
                           ian 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 un-
                           derstood – and that an unmeasurable and undefinable imbal-
                           ance 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 grad-
                           ual revolution in medicine that has been taking place since Mor-
                           gagni’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 specif-
                           ically the organs, of the body, his colleague Marie-François-
                           Xavier 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 process-
                           es within cells. All these efforts and discoveries are ultimately

                      directed towards a single goal, that of more precise medicine.
                      The objectives are to identify the causes of a disease, to consid-
                      er 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          Rudolf Virchow (1821–1902)
                                  (1682–1771) studied medicine        studied medicine at the Berlin
                                  in Bologna and in 1715 was          Military Academy. After holding
                                  appointed to the chair of ana-      a professorship in Würzburg he
                                  tomy at the University of Padua.    took up a chair of pathological
                                  In 1761, while still at Padua, he   anatomy that had been created
                                  published his principal work De     especially for him at the Univer-
                                  sedibus et causis morborum per      sity of Berlin. He was a political
                                  anatomen indagatis (‘On the         activist, campaigning vigorously
                                  seats and causes of disease in-     for democracy and public pro-
                                  vestigated by anatomy’). In this    vision of healthcare. The journal
                                  work he departed from the           “Virchows Archiv”, which he
                                  standard practice of the time by    established in 1847 together
                                  concentrating not on the symp-      with Benno E.H. Reinhardt, was
                                  toms and signs of disease, but      the organ of scientific pathol-
   on the location of disease within the organs of the body. In his   ogy. In 1858, with his work Die Cellularpathologie in ihrer
   view, the pathological changes in organs that he demon-            Begründung auf physiologische und pathologische Gewebe-
   strated in many autopsies were the true causes of illness. This    lehre (‘Cellular pathology as based on physiological and
   view amounted to a rejection of the theory of humors               pathological histology’), he founded a new theory of pathol-
   (humoralism) that since the time of Hippocrates had attrib-        ogy in which the cells of the body were regarded as the sites
   uted disease to an imbalance between the four bodily liquids       of origin of pathological changes.
   blood, phlegm, yellow bile and black bile.

Target: the molecular net-            Every disease is influenced by a larger or smaller
work of the cell                      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 se-
                      verely 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

                                          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 sub-
                        stances that differ from case to case) that drugs exert their ef-
                        fects. They act directly on the molecules that make up our body

     The changing role of biology in medicine

                                                                                           by computer

      target molecule              potential                  binding assay             best candidate               efficacy studies

         biological                 rational                high-throughput-                                          biological
       target search              drug design                   screening                                             evaluation

     The tasks of biology in medicine have changed greatly in the          computer so as to interact in a highly specific way with a cer-
     past few decades. Over this period the initiative for inno-           tain part of a target molecule; only afterwards would they be
     vations has passed increasingly from chemistry to biology.            produced by chemists. This approach has not yet brought the
     Molecular medicine is an expression of this change.                   success that was hoped for.
     Evaluation As recently as the 1970s the principal task of             High-throughput screening New, automated methods
     biologists in medicine was to evaluate, i.e. to test the effective-   then made it possible to test large numbers of substances in
     ness of, new substances produced by chemists.                         hundreds or even thousands of miniaturised assays for cer-
     Targets As knowledge of the molecular basis of diseases               tain biological, chemical and physical properties. Experiments
     increased, biology was able to provide new targets for drug           such as binding to a target, which previously had to be per-
     development. These targets formed, and still form, the basis          formed individually in molecular biology laboratories, could
     for the search for new medicines by chemical synthesis.               now be performed in this way. The desired properties of suit-
     Rational drug design Rational drug development arose as               able molecules can then be improved in a further step. High-
     a result of increasing knowledge of the structure, i.e. the form,     throughput screening has already resulted in the develop-
     of proteins. The idea was that drugs would be designed by             ment 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 deve-
    lopment of a disease constitutes a potential target for drugs.
    For example, in the past few decades biologists have discov-
    ered 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, computer-
    based 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 pa-
    tient’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 pig-
    ment, 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 res-
    ponsible 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
A multiplicity of possible      Genetics, genomics and proteomics thus provide
causes                          medicine with a variety of new ways of interven-
                                ing in the development and progression of dis-
                eases. 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 distin-
                guished between four humors; Morgagni extended the search
                for the seat of diseases to a couple of dozen organs; Bichat con-
                cerned 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 enor-
                mous 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 sim-
                ple infection or gene mutation, but rather arise as a result of a
                combination of external and internal, predisposing and protec-
                tive, 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 cardio-
                vascular 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.

The central importance           This is because disease-inducing environmental
of genes                         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 iden-
                tifying the genetic basis of a number of severe hereditary dis-
                eases brought about by a single defective gene. These include
                Huntington’s disease (Huntington’s chorea), cystic fibrosis (mu-
                coviscidosis) and hemophilia. More refined methods now allow
                scientists to investigate the genetic causes of more complex dis-
                eases in which various genes can exert positive or negative
                z Monogenic diseases such as Huntington’s disease, cystic fibro-
                    sis and hemophilia follow the classical (mendelian) laws of

Monogenic hereditary diseases: cystic fibrosis

                                                                       cell membrane


                                                                             regulatory                             chloride

Defect on chromosome 7.                                            Impact of genetic defect on salt transport.

The first disease-causing genes to be discovered in the            disease). This leads firstly to motor disorders and later to
1980s were associated with hereditary diseases due to spe-         mental deterioration. The disease generally appears between
cific mutations in individual genes. Such monogenic diseases       the age of 30 and 40 years and progresses in all cases to
can be classified on the basis of their pattern of inheritance:    death after 5 to 20 years. Physiotherapy and diet can slow
Autosomal recessive inheritance – The altered gene must be         progression. Since the responsible gene was found in the
inherited from both the father and the mother for the disease      mid-1990s, scientists have been working to develop suitable
to occur. Carriers of only one altered gene are healthy, but       drugs, however as yet none has been approved for use.
can transmit the disease to their children. An example of this     Sex-linked inheritance – In this case the responsible gene lies
is cystic fibrosis (also known as mucoviscidosis), which is        on a sex chromosome (gonosome), generally the X chromo-
due to a defect on chromosome 7. In this disease salt trans-       some. The presence of one unaffected copy is sufficient to
port is disturbed in certain mucosal cells, causing the mucus      suppress the disease, therefore men are far more commonly
of the respiratory tract, digestive tract and other organs to be   affected than women, since they lack a second X chromo-
extremely viscid. This results in frequent infections and in-      some. Hemophilia is such a disease. This is due to an inher-
flammation. Whereas only half a century ago most cystic            ited deficiency of a blood coagulation factor, as a result of
fibrosis sufferers died during childhood, those born today can     which the affected person’s blood coagulates very slowly or
expect to live 40 to 50 years thanks to specific (symptomatic)     not at all. Untreated hemophiliacs therefore die young of
treatment and diet.                                                internal and external bleeding. Previously, hemophiliacs were
Autosomal dominant inheritance – In this case a single             treated with blood transfusions, and even today the missing
altered gene is sufficient to cause the disease to appear, and     coagulation factor is obtained mostly from donor blood.
the disease is transmitted to fifty percent of the children of     Unfortunately, diseases can be transmitted from the donor to
sufferers. Huntington’s disease (Huntington’s chorea) is an        the recipient in this way, though less so than with transfusion
example of such a disease. In this condition a defect in the       of whole blood. Now that the responsible gene has been
Huntington gene on chromosome 4 causes production of an            identified, however, the missing factor can be produced using
incorrectly formed protein known as amyloid (a similar             recombinant DNA technology, thus eliminating the risk of
change is seen in Alzheimer’s disease and Creutzfeldt-Jakob        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 re-
       quires larger studies to identify the various genes that influ-
       ence the disease to a greater or lesser extent. This task is ren-
       dered 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 (pro-
       nounced ‘snips’), i.e. changes in single subunits of the
       genome. The presence of such single nucleotide substi-
       tutions 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 as-
       sociation 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 adop-
       tion studies are helpful here. Monozygotic (‘identical’) twins
       brought up in different families have an identical genome but
       are subject to different environmental influences, while dizy-
       gotic (‘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 step-
     z Our genetic makeup also exerts a decisive influence on our
       predisposition to disease. Where genes that play a role in the

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             environment
simple genetic causes. In the vast majority of cases many dif-
ferent 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 en-                                           genes
    dogenous, 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 reac-               nutrition
    tions independently of the actual causes of disease.
2. Genes can protect against or predispose to the develop-
    ment 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           lifestyle
    number of protective gene variants. In the age of molecu-
    lar medicine, terms such as ‘healthy’, ‘ill’, ‘normal’ and
    ‘abnormal’ are therefore no longer easy to define.               causes of complex diseases can generally be determined only
3. All risk factors — including genetic factors — for a complex      via large, statistically complex studies and on the basis of a
    disease can be either categorical or quantitative. A gene        deep understanding of the molecular processes that take
    variant is said to act categorically if a certain disease can    place in cells. Only now that the findings of genetics, ge-
    occur only in its presence. By contrast, quantitatively act-     nomics, proteomics and bioinformatics are being applied to
    ing gene variants act additively (or else their effects are      medicine has this become possible. In the case of complex
    multiplied) up to a critical point at which disease occurs.      diseases, however, even tests based on techniques of molec-
4. Complex diseases are polygenic, i.e. they result from the         ular medicine can at best indicate only an approximate risk
    action of a number of different genes. The contributions of      that an individual will develop a particular disease. Absolute
    the individual genes are difficult to determine and can vary     assertions cannot be made. In the future, genetic testing may
    enormously, especially as genes influence each other’s           increasingly be used to guide patients and healthcare pro-
    action.                                                          viders in designing optimal treatment strategies based on
5. Another characteristic of complex diseases is genetic het-        patient’s genetic variations.
    erogeneity: since a number of genes can be jointly respon-       For example, pharmacogenetic research is already underway
    sible for the occurrence of a disease, different combina-        to provide physicians with a better understanding of the
    tions of genes can result in the same clinical picture.          influence of genetic variation on an individual’s response to
This complex interplay of influences means that the genetic          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 precau-
                         tions 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 environ-
                            mental factors and that predisposing and protective genes
                            can overlap, such tests can only ever indicate a greater or less-
                            er probability that an individual will develop a disease.

Diagnosis and therapy                 The importance of looking for the causes of dis-
look each other in the eye            ease has not changed at all since Morgagni pro-
                                      pounded his organ-based pathology: only if we
                        truly understand a disease can we treat it correctly. Nowadays,

     Triple influence of genes: hepatitis C                                                            complexity is hepatitis C,
                                                                                                       which is expected to become
                                                                                                       far more common over the
                                                                                                       next few decades. Untreated,
                                                                                                       this disease leads to cirrhosis
                                                   1b, 2a,                                             of the liver in about 20% of
                                                   2b, 2c,                           1b                those infected and to liver
                  1a, 1b,                          3a                                                  cancer in a smaller proportion
                  2a, 2b,                                      4                                       of patients. The responsible
                  3a                                                              1b,                  pathogen, hepatitis C virus
                                                        4              1b, 3a      6                   (HCV), occurs in at least six
                                                                                  3b                   different types, plus subtypes,
                                                                                                       that occur with different fre-
                                                        5a                            1b, 3a           quency in different parts of the
                                                                                                       world (see figure). The exis-
                                                                                                       tence of these variants has a
                                                                                                       major influence on the effec-
                                                                                                       tiveness of standard therapies.
                                                                                                       For example, interferon, the
     Over the past few years it has become increasingly clear that       most important agent for use in this disease, is relatively
     even infectious diseases are subject to a complex set of            ineffective against HCV type 1, the predominant type in
     genetic influences. This is true not just in regard to the          Europe and North America. Moreover, this type of HCV
     causative pathogens – whose genetic background is often             generally causes more severe disease than other types.
     well known – but also in regard to the host. Genetic differ-        In addition, standard interferon preparations usually cause
     ences between individuals make some people more, and                severe side effects, all the more so because they generally
     others less, susceptible to particular infections. In addition,     have to be taken three times weekly. Administration at such
     our genes influence the way our body deals with all types of        short intervals is necessary because interferon is broken
     drug, including anti-infective agents. In particular, drugs for     down very rapidly in the body and therefore acts for only a
     use against viruses, which evolve rapidly, often show unsat-        few hours. Patients are thus obliged to put up with a con-
     isfactory efficacy and troublesome side effects. Therefore, if      stant ebb and flow of side effects. Improved drugs such as
     better drugs are to be developed, both the genome of the            pegylated interferon have been available for some time
     pathogen and that of the patient must be taken into account.        now. Used in combination with ribavirin, pegylated inter-
     An example of an infectious disease that shows this kind of         feron 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 ge-
                   netic predisposition, to predict the effect of drugs on a disease
                   on the basis of the molecular characteristics of the drugs con-
                   cerned, and finally to choose therapy that is optimal for the
                   individual patient. Knowledge of the molecular level of the dis-
                   ease process thus opens up completely new approaches to treat-
                   ment: 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        chronic course, be mild or severe, and undergo spontane-
the period of time during which interferon remains in the        ous cure or lead to liver cancer, the direction taken by the
body. The product Pegasys works in this way. In this medi-       disease in an individual being largely determined by that
cine the interferon is enclosed within a coating made up of      individual’s genome. At present, however, the genes
a branched molecule known as polyethylene glycol (PEG).          responsible for these differences are largely unknown.
This delays breakdown of the interferon, with the result that    These genes are therefore another important object of HCV
the drug only has to be taken once weekly. This results in       research, since depending on the genetic predisposition of
greater efficacy and fewer side effects.                         the patient (and the virus), treatment may be necessary or
Genetic factors are important in hepatitis C not only in rela-   unnecessary, a particular drug may be suitable or unsuit-
tion to virus type and drug metabolism, but also in that they    able, and cure may be possible or unlikely. Therefore, the
partly determine the clinical course of the disease. As men-     more is known about the relationships between the genome
tioned above, infections with HCV type 1 are generally more      of the pathogen and that of the host, the more specific will
severe than those with other types of the virus. Neverthe-       be the drugs that can be developed for use in this disease.
less, infection with any HCV type can show an acute or

                                    Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy             17
     At the same time, these new methods and discoveries are a con-
     tinuation 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 pos-
     sibilities 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-per-
     cent effective therapies in the very near future. Rather, genetics,
     genomics and proteomics will first help doctors to avoid inef-
     fective or even dangerous therapies – after all, there is no such
     thing as a panacea. But even this is a major step forward.


     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 Humangenom-
         projekts (ed.): Das Humangenomprojekt – 1st and 2nd edition
     Healthnet-Services GmbH: (M)Eine Geschichte der Pathologie, Teil 1-7:

genes and drug response

                                         To d a y

                                                  A                             B



                                         To m o r r o w       Diagnostics


A drug may work well in one person,
                                                          B                     D
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
                       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 differ-
                       ences 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
 Pharmacogenetics describes the influence of genes on the
                                                                  some time that the efficacy
 efficacy and side effects of drugs.                              and tolerability of drugs
 Pharmacogenomics studies interactions between drugs and          vary from one person to the
 the genome.
 Pharmacokinetics investigates the uptake, conversion and         next. Thus, some patients
 breakdown of drugs in the body over time. Environmental factors, need a lot more or a lot less
 diet and genetic predisposition all play a role.                 of a given drug than most
 Pharmacodynamics deals with the influence of genes on the
 interactions between drugs and their molecular targets.          people; side effects keep oc-
                                                                  curring unexpectedly; and
                                                                  sometimes a drug that is usu-
                       ally highly effective does not work at all. Our uniqueness is
                       reflected in our body’s response to drugs. Because of this, per-
                       sonalised medicine has emerged as a hot new topic of discus-
                       sion. Future drugs, it is hoped, will be better adapted to our ge-
                       netic diversity and dissimilar life circumstances and will be
                       more efficient, more specific and safer. And they will be sup-
                       ported by a battery of fast, simple genetic tests that will enable
                       doctors to select the right drug for their patients’ specific needs.

First example                     A first example of a genetically specific drug has
Herceptin                         already been introduced in the form of Hercep-
                                  tin, which is used in the treatment of breast can-
                   cer. 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 sur-
                   face 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 there-
                   fore 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 re-
                action (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: Ex-
                pectations 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.

Pharmacogenetics                As early as 1958 the German pediatrician Frie-
as a research discipline        drich 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 quick-
                ly led to the first application of genetics in medicine. Over 100
                relevant genes are now known, and many more will follow as sci-
                entists rapidly refine the field of pharmacogenetics with the help

                                                          Pharmacogenomics: genes and drug response   21
     of methods and knowledge gained from unravelling the human
     genome. Meanwhile the subject of enquiry has shifted from in-
     dividual 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 dif-
        ferences. 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 mo-
        lecular causes. Usually the activity of an endogenous protein
        is impaired. The shape of such proteins is genetically deter-
        mined. Small differences in our genome can therefore signi-
        ficantly 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

When drugs do not work: pharmacogenetics

                                                                           • target molecule(s)
                                                                           • etiological or palliative
                                                                           • potential side effects

                                                                                 • not always

                                                                                   cleavage of a
                                                                                   protective group

                                                                                 • e.g. via specific
                                                                                   receptors or channels

       breakdown:                                                                  cleavage of
       • often by modification                                                     protective groups
         and/or cleavage                                          heart                                                       gut
       • new compounds
         may be formed

                                             • often possible only after
                                               conversion or breakdown

                                                                                       : (inactive) drug

                                        ureter                                         : (inactive) drug in the bloodstream
                                                                                       : protective groups
                                                                                       : active drug
                                                                                       : cleaved, water-soluble
                                                                                         degradation product

Our genome influences the effects of drugs at at least three                       or prevent the action of a drug. Pharmacodynamics de-
levels:                                                                            scribes these underlying ‘etiological’ differences.
z The metabolism of a drug – i.e. its uptake, conversion or                 z      If a drug does not act on the cause of a disease but
    breakdown – may be prevented, slowed or accelerated.                           rather on its manifestations, many genes may be in-
    Pharmacokinetics investigates the causes of these phe-                         volved in and interfere with its effects. These ‘palliative’
    nomena.                                                                        differences are also the subject of pharmacodynamic
z Modification of the target molecule can directly weaken                          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 relat-
                             ed 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
Prominent example:                     Genes have been found for each of these three
cytochrome P450                        areas of pharmacogenetics. The vast majority of
                                       them act at the pharmacokinetic level. This re-
                       search 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 characteris-
                       tics. This is generally done in order to obtain a picture of phar-
                       macokinetic differences in the population that is as representa-
                       tive 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 im-
                       portant 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 mo-
                       lecular disposal of drugs proceeds more slowly.
                       Another characteristic of the P450 family of enzymes is medical-
                       ly 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.

     Hurdles for drugs: genetic polymorphisms

     Genetic differences between individuals strongly influence         excretion of drugs. As a result, the drugs do not remain
     the function of the corresponding gene products (usually           in the body long enough to be effective or remain in the
     proteins) and in this way affect the activity of drugs in the      body too long so that the risk of dangerous side effects
     body. Some of these genetic variants, known as polymor-            increases.
     phisms, have been identified. They influence people’s res-      2. Pharmacodynamics: Some genes have also been
     ponse to drug therapies at the pharmacokinetic or phar-            found that are directly responsible for the structure of
     macodynamic level.                                                 the target molecule, influence the associated signalling
     1. Pharmacokinetics: Genes coding for enzymes in-                  pathway or interfere with some other metabolic pathway
        volved in the metabolism of drugs form the largest group        involved in the manifestations of a disease. The inter-
        of known pharmacogenetic factors. Fluctuations in their         actions between such genes and drugs can be highly
        activity can slow or accelerate the uptake, conversion or       complex.

   Personalised medicine: the objectives of pharmacogenomics

   Dosage Are the patient’s individual pharmacokinetic fac-              Safety Can adverse reactions to a specific drug be pre-
   tors known, and can the dose of a drug be adjusted upward             dicted? Can the patient be switched to another drug that
   or downward to suit his/her metabolism? This would help               he/she is more likely to tolerate? Are there alternatives? Can
   ensure that the drug works while at the same time reducing            provisions be made in advance for medical supportive mea-
   its side effects.                                                     sures?
   Efficacy If it is already known in the initial stage of treat-        Prevention If the cause of the disease is related to genet-
   ment which drug is likely to work for a patient, a lot of trial and   ic factors, the disease can be diagnosed early by means of
   error would be saved. Valuable time would be gained for the           tests and possibly avoided by initiating specific measures such
   treatment, unnecessary expense would be reduced and the               as diet or exercise.
   patience of the doctor and patient alike would be spared.

Genome and environment                As with all proteins in the body, this variable ac-
act together                          tivity 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 de-
                       pendencies and uncertainties, of effects and side effects, which
                       ultimately makes our body’s individual response to drugs uni-
                       que, especially when pharmacodynamic factors are also taken
                       into consideration. Pharmacogenomics can elucidate this inter-
                       play 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.

Old hurdles for new drugs                The old slogan ‘one size fits all’ still applies to
                                         most drugs. If a drug fails to work in too many pa-
                       tients, 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 re-
                       spect can be due to at least three reasons:
                       z Are there problems with the pharmacokinetics, i.e. the up-
                          take, conversion or breakdown of the drug, and if so, what
                          can be done about it?

                                                                                   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 de-
                velopment of new drugs is ultimately limited – to the detriment
                of patients, who have only a limited range of therapeutic options
                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 change-
                    ability may cause disease.
                z If it is not possible to distinguish between the causes of a dis-
                    ease, often only palliative drugs will produce a uniform ef-
                    fect, since they act on the effects rather than on the cause of
                    the disease. However, it is usually much safer and more effec-
                    tive to eliminate the underlying molecular cause.

SNPs yield the first             To understand the molecular causes of a disease
evidence                         or a drug’s failure to have an effect, we therefore
                                 need to look at the genes. We first have to identi-
                fy 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 com-
                plicated, this is no small task. A promising solution is to look for
                single nucleotide polymorphisms, or SNPs for short. These vari-
                ations, which are spread more or less randomly throughout the
                genome, are thought to determine our individual genetic differ-
                ences 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 re-
                peatedly 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.

Databases to the rescue        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 rele-
               vant 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 in-
               vestigate 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. Other-
               wise there is no gain over the old trial-and-error approach.

10,000 experiments at         The final link in the pharmacogenetic research
once: DNA chips               chain is to search for the presence of the gene
                              variant in a patient. Conducting this search with-
               in 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 sub-
               stitution 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 experi-
               ments 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, scien-
               tists 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

                                                          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 ge-
               netic testing.

Potential and obstacles         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, appli-
               cations 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 ex-
               ample, to screen test subjects during the process of drug devel-
               opment (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 treat-
               ment failures for new and existing drugs. This will enable doc-
               tors 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 med-
               ical developments, there are ethical and legal reservations as well
               as financial misgivings and still unresolved scientific and tech-
               nical issues. In addition, applications of modern genetic and ge-
               nomic research have consistently raised public misgivings, and
               particularly in this area pharmacogenomics is in need of a re-
               think: The essential requirement for more personalised me-
               dicine is unobstructed access to genetic data. To be of use to
               patients, findings must be made available to doctors and phar-
               macists. In any case, the prescription of a drug with significant
               pharmacogenetic properties reveals the patient’s diagnosis any-
               way. And we should also realise that this is already the case with
               many drugs today. Ethical, legal and societal implications of ge-
               netics and genomics in medicine as well as the technical and eco-
               nomic requisites are dealt with in more detail in the chapter on
               basic conditions.


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,
Kroll W, Hartwig W: Pharmakogenomik. Nachrichten aus der Chemie 50, March 2002 Pharmacogenomik – Die Suche nach den idealen Pillen.
Human Genome Project Website:
Human Genome Project – Pharmacogenomics:

                                                            Pharmacogenomics: genes and drug response   29
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 molecu-
les, they transport, store and mobil-
ise 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 analo-
                        gy to the term ‘genome’, the principal focus of biological re-
                        search 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,
  Terms                                                       threadworms and even hu-
                                                              man beings revealed their
  Genome the (largely unchangeable) totality of the genes of
  an organism.
                                                              genetic makeup. By contrast,
  Proteome the (in most cases constantly changing) totality   proteins, the major products
  of the proteins of an organism.                             of all these genomes, were
  Genomics the study of the form, function and interactions
  of the genes of an organism.                                largely ignored at that time,
  Proteomics the study of the form, function and interactions being regarded as a subject
  of the proteins of an organism.
                                                              for basic research by scien-
                                                              tists interested only in the
                        acquisition of knowledge for its own sake. Since then, this situa-
                        tion 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 ge-
                        nome, 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 con-
                        stantly produced, broken down, altered, moved around, bound
                        and separated. The task of bringing light into this tangled un-
                        dergrowth is an even greater scientific challenge than that of
                        decoding the human genome, but at the same time it will accel-
                        erate progress in gene and genome research.

Diverse structures                The importance of proteins lies in the multiplic-
and functions                     ity of the tasks that they perform. They play a cen-
                                  tral role in almost all the processes involved in the
                   life of an organism or – viewed on a smaller scale – a cell:

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 con-
    stant synthesis, rearrangement and breakdown of all the sub-
    stances 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 with-
    in (and to some extent also outside of) the body. These in-
    clude 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 an-
tibodies 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 es-
timated 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
                                     A chain of up to twenty different amino acids (primary struc-
                                     ture – the variable regions are indicated by the squares of dif-
                                     ferent colours) arranges itself into three-dimensional struc-
       }        secondary            tures. Among these, helical and planar regions are particularly
                structure            common. The position of these secondary structures in rela-
                                     tion to one another determines the shape of the protein, i.e.
                                     its tertiary structure. Often, a number of proteins form func-
                tertiary structure
                                     tional complexes with quaternary structures; only when
                                     arranged in this way can they perform their intended func-
                                     tions. When purifying proteins, it is extremely difficult to retain
                                     such protein complexes in their original form.

                                                 Proteomics: seeing through the undergrowth                33
               amount and complexity of the relevant data and knowledge al-
               ready 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 disci-
               pline of proteomics aims to change this.

Protein catalogues             As a first step, many laboratories around the
to provide order               world are working to develop as complete a list as
and perspective                possible of all the proteins that occur in the hu-
                               man 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 experi-
               ments. At the same time, comparison of the genome with the
               proteome facilitates the search for new genes and thereby pro-
               vides genome researchers with an additional tool for interpret-
               ing 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 ex-
               ternal influences that act upon it. The appearance or disappear-
               ance of proteins in the course of an illness or in response to
               administration of drugs has already been a subject of investiga-
               tion for decades, however the search for such changes can now
               be conducted in far more systematic fashion with the aid of pro-
               teomics. 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 differ-
               ences 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.

Proteome research as a       Proteome research thus constitutes an important
link between disciplines     link between various fields and disciplines of
               z Diagnosis. The effects of genetic changes first become mani-
                 fest at the level of proteins. In many cases the question of
                 whether a known mutation actually brings about an illness

                 or is counterbalanced by other factors can be answered only
                 by means of protein testing. Proteins are therefore more suit-
                 able than genes for use as diagnostic markers of complex dis-
               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 treat-
                 ment 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 de-
                 velopment 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 ex-
                 ample, certain marker proteins can indicate that a cell has
                 been exposed to a substance that is toxic to them. Toxic ef-
                 fects of new drugs can thus sometimes be predicted at an ear-
                 ly stage of drug development.

Strategy: standardisation      Given its huge importance, protein research has
and automation                 always occupied a central role in the search for
                               new diagnostic methods, treatments and tech-
               niques. Most of the methods used today were known long before
               the term proteomics was coined in Siena. Proteomics is now ex-
               tending 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 ex-
               actly 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

                                                          Proteomics: seeing through the undergrowth   35
              clinical sample                          tissue                   protein separation
                                                  protein extraction           (2D gel electrophoresis)
         blood               tissue
                                                                                 isoelectric point (pl)

                                                                                                            molecular weight (kDa)

                                                                                            spot picking
                                                                                          (each spot contains
                                                                                              a protein)

           bioinformatics                       mass spectrometry
        protein identification by                  (of each spot)
           database search

Proteome analysis using 2D gel electrophoresis and        separated by 2D gel electrophoresis. The resulting
MALDI-TOF mass spectrometry. Proteins are extract-        protein spots are then cut out of the gel and the
ed from clinical material such as blood or tissue and     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 pow-
                     erful computers and specially developed software programs.

Separation by charge                The first step is always that of separating the mix-
and size                            ture 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 pro-
                     teins are separated firstly according to their charge and then – at

   Two-dimensional gel electrophoresis: an important separation method

                   control               Alzheimer’s disease
                                                                    tein reaches its isoelectric point, i.e. once its net charge is
                                                                    zero, it stops.
molecular weight

                                                                    In the next step, the separated proteins are further sorted

                                                                    according to size. This occurs at a right angle to the direc-
                                                                    tion 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 pro-
                                                                    teins to an extent that depends upon the size of the pro-
                                                                    tein molecules. Once again, an electric field is applied,
                                                                    however in this step the rate of migration is determined by
   Production of glial fibrillary acidic protein (GFAP)
                                                                    the charge of the SDS, which in turn is determined by the
   is increased in Alzheimer’s disease.
                                                                    size of the protein molecules. This type of separation is
                                                                    known as SDS-PAGE, or sodium dodecyl sulphate-poly-
   In two-dimensional gel electrophoresis, proteins are sepa-       acrylamide gel electrophoresis.
   rated according to their charge and size. The first step is      Proteins that are very similar (e.g. modified forms of the
   separation by charge. This is achieved by means of iso-          same molecule) are often extremely difficult to separate in
   electric focusing (IEF) using a polyacrylamide gel to which      conventional gels. In such cases use is made of narrow-
   a pH gradient is applied for this purpose. When an electric      range gels with an extremely gentle gradient within the pH
   field is applied to the gel, proteins migrate along the gradi-   range being examined. In this way even minimal differences
   ent for as long as they possess a net charge. Once a pro-        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 rep-
                             resents 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 wheth-
                             er, and if so in what quantity, a particular protein is present in a
                             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 pre-
                             cisely 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 elec-
                trophoresis is generally complemented by other separation

Presorting increases            For example, components of the sample to be
resolution                      analysed can be separated in advance by filtration
                                or centrifugation. Also available are various tech-
                niques by means of which proteins can be presorted according
                to various characteristics – e.g. charge, size, shape or binding be-
                haviour. And the 2-D gel electrophoresis itself can be further re-
                fined by limiting both of the separation factors that it employs,
                namely charge and size, to a certain range within which resolu-
                tion 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
                In the next step, the proteins to be identified are cut into pieces.
                Just as genome researchers split DNA into more easily sequen-
                ced 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 tryp-
                sin, for example, always cleaves the chain after the amino acids
                arginine or lysine – the peptide mixture obtained is specific for
                each protein/protease pair.

A fingerprint for every         In order to identify the cleaved protein, the pep-
protein                         tide 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 pep-
                tides fly faster than large peptides. The time that they take to
                reach the end of the tube therefore indicates the size of the pro-

The tiny difference: peptide fingerprinting

  UV laser

                                          electric field
                                      +                        -

    laser beam
                                                           +                                  + +
                                                       +                                       +


                                                                                                                    data analysis
                                                                                                                    in computer
                     matrix peptides          paffle plates flight tube                        detector

                         ionisation       acceleration             time of flight              measurement            analysis

In MALDI-TOF MS (matrix-assisted laser desorption ioni-                enter a flight tube along the length of which a powerful
sation time-of-flight mass spectrometry), the protein sam-             electric field is applied. The peptides fly in the direction of
ple to be investigated is digested with a specific protease            the electric field until they reach a detector at the end of the
and the resulting peptide mixture is embedded in a matrix              tube – the lighter molecules arriving first, the heavier mol-
in a mass spectrometer. The energy of a laser impulse                  ecules last. The time taken by the peptides to fly through
applied to the matrix is transferred to the peptides, which            the tube is measured.
are thereby ionised and vaporised. The charged peptides

                  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 pro-
                  tein under investigation: a distinctive spectrum on the basis of
                  which the molecule can be identified beyond doubt. In fact, pro-
                  tein researchers actually refer to this process as ‘fingerprinting’
                  and compare their work with forensic science in which finger-
                  prints 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                    Signal broadening caused by carbon isotope              13C

        1500 2000 2500   3000   3500   4000   4500 5000

        1500 2000 2500   3000   3500   4000   4500 5000

                                                                            1500   2000   2500   3000   3500   4000   4500   5000

                                                                      A peculiarity of mass spectra is the broadening of the sig-
        1500 2000 2500   3000   3500   4000   4500 5000               nal 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.
        1500 2000 2500   3000   3500   4000   4500 5000
                                                                      Nevertheless, the resolution of the technique can be con-
                                                                      siderably increased by use of mathematical methods and
     MALDI-TOF MS yields a spectrum – a fingerprint – that is         in particular by use of extremely sensitive spectrometers.
     specific for the digestant and the protein present in the        The accuracy of modern MALDI-TOF mass spectrometers
     sample. Each peak represents a signal of a certain strength      is rated at about 10 ppm (parts per million).
     obtained at a certain time. The fingerprint can be repro-
     duced 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.

                         and subjected to mass spectrometry in virtual form in a compu-
                         ter. To remain with the analogy of forensic science, this is equiv-
                         alent to a suspect’s fingerprint being worked out from his pho-
                         tograph – a possibility that would presumably make every police
                         officer in the world green with envy!

Form and function                       Nevertheless, simple database comparisons of
go together                             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 con-
                         sidered to be possibly unknown, does the really exciting part of
                         the task begin: the work to identify the form and function of the

              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 deter-
              mination 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.

Automated structural            Until a few years ago determination of the spatial
analysis                        configuration of a correctly folded protein was a
                                task worthy of a doctorate. Now, however, pro-
              teome research has largely automated this aspect of science and
              thereby made it possible to undertake a comprehensive structur-
              al analysis of the human proteome. This applies in particular to
              x-ray crystallography, the most important method for investi-
              gating the shape of small to medium-sized proteins. In this tech-
              nique 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 au-
              tomatically, 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 re-
              quired – are then bombarded with x-rays. On their way through
              the crystal the x-rays are deflected (diffracted) by the protein

                                                            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 tubu-
                                                          lar 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 custom-
                                                          made 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 impos-
                                                          sible 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.

Fishing for the right              By far the most difficult part of the analysis of a
partner                            new protein, however, is identification of its func-
                                   tion, i.e. its place within the molecular network of
                    the body. Here again, though, technological advances are pro-
                    viding 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 mol-
                    ecules 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 ex-
                    clusive 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

             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

HUPO - Human Proteome        Cooperation between these disciplines is corre-
Organisation                 spondingly close. Proteomics generally blends
                             seamlessly with genetic and genomic experi-
             ments, 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 expe-
             riments conducted at the level of proteins. In order to promote
             this cooperation and especially in order to consolidate the ef-
             forts 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 hu-
                man 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.

                                                         Proteomics: seeing through the undergrowth   43
     First successes: new signalling pathways in cancer

     Dozens of signalling pathways are involved in the develop-       prevents the cell from undergoing transformation into a can-
     ment of the various forms of cancer. Each newly discovered       cer cell. In order to be able to exert this control function, ICE3
     pathway provides further potential targets for medical inter-    is split in healthy cells by an enzyme known as granzyme B
     vention. Proteomics can help in the elucidation of such sig-     to form apopain. The 2-D gel of the cancer cells shows a
     nalling pathways.                                                large amount of unsplit, i.e. inactive, ICE3. From a separate
                                                                      experiment with gene chips it is known that the genetic
       control                           ras-transformed              instructions for the production of granzyme B are absent
                                 ICE3    mouse lymphocytes            from the cancer cells. From this combination of findings the
                                                                      signalling pathway that operates here can be deduced:

                                                                       Signalling pathway diagram


                                                                        ras        granzyme B           X

                                BTF3a                                                                apopain
     The 2D gel on the left shows a subset of the proteome of a                                     apoptose          growth
     normal cell; that on the right the same subset of the pro-
     teome of a cancer cell. The differences are readily apparent.
     - A protein referred to as ICE3 is present in greater quan-      The two other proteins whose amounts are altered in the
        tity in the cancer cell.                                      cancer cells likewise play a role in cancer. HMG2 binds to
     - The amount of the high mobility group protein HMG2             deformed DNA. This protein appears to have largely disap-
        is reduced in the cancer cell.                                peared from the cytoplasm of the cancer cells; this is evi-
     - The transcription factor BTF3a is formed in greater            dence of genetic damage to the cells.
        amounts in the cancer cell.                                   BTF3a is likewise a DNA-binding protein, however its func-
                                                                      tion is to ensure correct transcription of genes. BTF3a had
     ICE3 plays an important role in programmed cell death, or        previously been shown to be present in increased amounts
     apoptosis. This ‘suicide’ of body cells occurs when the          only in intestinal cancer cells. This protein therefore has
     genetic material of a cell is severely damaged; in this way it   potential for use as a tumour marker.

The enormous challenge       Proteomics is nevertheless confronted by prob-
of proteomics                lems. One of the most important of these is a di-
                             rect consequence of the central characteristic of
              the proteome, namely its complexity. The total number of pro-
              teins 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, cur-
              rently 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 pro-
              teins remain to be discovered (for example, our body is con-
              stantly 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, how-
              ever, every publication constitutes a risk for a researching com-
              pany. The fact that this problem is well known from genome
              research will at least prove helpful in this regard.


              Langen H: Proteomics as a new field in biology: applications and potentials. Presentation at Roche Round-
                 table 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:

                                                                          Proteomics: seeing through the undergrowth      45
Targets for medicine

Without targets there can be no
drugs. Small wonder, then, that tar-
gets 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 af-
               ter, pursued and adored. Whenever they are found, they become
               an object of feverish research. Scarcely any other area of biolog-
               ical 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 pharma-
               ceutical 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 biomole-
               cules 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.

Targets in the narrow          Because of the multiplicity of their functions and
sense: proteins                properties, proteins are by far the most impor-
                               tant 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

Important cause               DNA, the substance that bears our genetic infor-
of disease: DNA               mation, controls the vast majority of bodily pro-
                              cesses 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. Mal-
               function of a gene can be due to an alteration in the sequence of

               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 di-
               rectly 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.

Much more research              RNA has a number of functions in the body,
required: RNA                   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 single-
               stranded 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 pro-
               teins 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.

Mostly minor roles:          Other substances that occur in the body are also
sugars, fats, etc.           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 dif-
                 ferent forms.
               z All metabolites, that is to say the starting substances, inter-
                 mediate products and end-products of our metabolism, are
                 theoretically susceptible to influence by drugs.

                                                                                Targets for medicine   49

                             *                                        PI 3'kinase                    ??                          GLUT4
                                                                                                                                 translocation                     7TM
  p91                                                                 p85α p 110        ??                                                                    βy
                                                                                                   ISSK                                                NSF
                                                          *                                                                                            SNAP
                             *          IRS-1             *                                                                         cGI-PDE            RABs
              *                                                   dynamin
                                                                                                    RAS          PKa                (type III)         ARFs
                                                                                                                                    translocation      Ga, βy
                                                                                                                14-3-3                                 SNARE`s
                                                                         SOS                      B-RAF                   PKCβ
                                  SHPTP2              *       GRB 2                                        RAF-1
                                           *                                                                    *

                  NFr B translocation                     PKC ζ                                                 *
                                                40S                                          ??            MAPK

                                                                               *                                *                          glucose
                      protein                                                                              p90Rsk
                      transcription                                                                                    GSK3
                                                Fos                                          *
                                                                                        PP-1                           glycogen
   CRE's                                                                                     *                         activation

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 pro-
                                  teins, in particular. Under some circumstances, however, it can
                                  be useful to bind a specific intermediate product of an undesir-
                                  able 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.

Improve or invent                               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.

                   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 com-
                   pletely new methods of treatment, i.e. to find new molecular tar-
                   gets for drugs. As compared with innovative improvement, this
                   approach calls for greater financial investment and more exten-
                   sive 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 ex-
                   pensive 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          z provide a better, or simply a different, way of interfering
rate. In many cases a great deal of intensive research has            in a disease process and in this way lead to the de-
to be conducted before it becomes clear whether a newly               velopment of new, and possibly more effective or better
discovered target molecule has anything at all to do with a           tolerated, drugs;
disease, and even then the importance and general role of         z   be closer to the molecular cause of a disease than
the molecule in the body's metabolism has to be elucidated.           presently known targets and thereby provide a more
If a potential target is found to play only a minor role in the       specific and surer method of attack;
genesis of disease, active agents developed to act on it are      z   play a role in production of the symptoms, rather than
unlikely to be of therapeutic value; if, on the other hand, a         the cause, of a disease and therefore also have poten-
potential target is found to play a major role in other impor-        tial for use as a therapeutic target in other diseases with
tant metabolic processes, the undesirable function is likely          similar symptoms;
to be difficult to block precisely and specifically.              z   be used as targets in diseases that have hitherto been
At the same time, however, speculative target research has            regarded as essentially untreatable due to a lack of any
the potential to deliver considerably greater therapeutic             suitable target; these include various types of cancer,
advances than can be achieved by simple improvement of                since cancer is often due to a variety of causes that need
familiar routes of attack. For example, previously unknown            to be treated individually.
targets can:

                                                                                                        Targets for medicine        51
               tific outlay required for the development of completely novel
               therapies is therefore becoming ever greater.

Successful search              The profusion of techniques now used to find and
at many levels                 evaluate new drug targets are derived from a wide
                               variety of scientific disciplines including proteo-
               mics 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 useful-
               ness 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 biolog-
               ical function of these; others are used to elucidate the molecular
               basis of diseases; others to investigate the properties and func-
               tions 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 lev-
               els of target research.

Proteomics and classical      The area of drug research that has been most
protein biochemistry          important historically deals with the most im-
                              portant group of targets, namely proteins. Orig-
               inally, 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 pro-
               teins that reacted with active substances and the interactions
               that resulted from these reactions. As this field of research re-
               vealed 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     • proteomics      •                • proteomics   • proteomics         • proteomics          • proteomics      • proteomics       •
 • databases      • databases       • databases      • databases    • databases          • databases           • databases       • databases        • databases
 • SNPs           •                 • SNPs           •              • SNPs               • SNPs                • SNPs            •                  •
 • gene chips     •                 • gene chips     • gene chips   • gene chips         • gene chips          • gene chips      •                  •
 • PCR            •                 • PCR            •              •                    •                     •                 •                  •
 • transg. mice   •                 • transg. mice   •              • transg. mice       • transg. mice        • transg. mice    • transg. mice     • transg. mice
 •                •                 •                • HTS          •                    • HTS                 •                 • HTS              •

    unknown           investi-         genetic         properties      function            molecular             selection             drug              drug
    molecule           gation        background                                           environment                                finding          evaluation

 • e.g. protein   • protein         • associated     • physical     • e.g. enzyme,       • associated          • attack which    • chemical         • preclinical and
                    structure         gene           • chemical       hormone,             metabolic pathway     target where?     synthesis          clinical studies
                  • possibly        • transcript     • biological     structural         • complexes with                        • test physical,   • marketing
                                                                                           other proteins
                    prediction of   • regulation                      protein                                                      chemical and       authorisation
                                                                                         • when, where,
                    structure                                                              how, how much
                                                                                                                                   biological         procedure

Modern target finding and evaluation is supported by                                 yield useful findings only when their results are con-
a wide variety of methods. The techniques employed                                   sidered together. Most techniques are used – in more
vary greatly in technical complexity and cost and often                              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 com-
                              position and form elucidated and their function described.

Databases and the Human                     A similarly systematic method of searching for
Genome Project                              new targets can now be conducted in the com-
                                            plete absence of any biological system: many drug
                              targets are now being searched for and evaluated in the world-
                              wide 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 exam-
                              ple, the composition, that is to say the amino acid sequence, of
                              an unknown protein is known, comparison with known mole-

                                                                                                                                  Targets for medicine                   53
               cules of similar structure provides important information on
               the likely properties, and thus also on the function, of the pro-
               tein 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. Comput-
               erisation of biological research has now developed into an in-
               dependent scientific discipline and career path known as bioin-

SNPs – single nucleotide      Single nucleotide polymorphisms, or SNPs (pro-
polymorphisms                 nounced ‘snips’) have become increasingly im-
                              portant in recent years. These randomly occurr-
               ing 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.

DNA chips, the transcrip-      The basis for another important technique used
tome and protein chips         in target research was established by the comput-
                               er industry: several hundred thousand different
               DNA fragments can now be accommodated on a glass or syn-
               thetic chip measuring just 1.5 cm square. This makes it possible,
               for example, to find genes containing segments with an ab-
               solutely 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 deter-
               mine 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.

  Virtual laboratory: bioinformatics

  For a long time now, the work that researchers do in their       z DNA chip experiments provide amounts of data that
  computers has been at least as important as the work they            previously would have been acquired only after years of
  do in their laboratories. Modern experiments, especially if          work in major research institutions. In order to permit
  automated, yield an unimaginable amount of data that could           evaluation of such data within a reasonable time, bio-
  not possibly be analysed without the aid of powerful com-            informatics specialists are constantly developing new
  puters and specially developed computer programs. Bioin-             programs and databases.
  formatics, the application of modern information technology      z   Prediction of the structure of proteins is a relatively new
  to biological research, has thus developed into an indepen-          and controversial subfield of bioinformatics. Modern su-
  dent scientific discipline with a broad range of applications,       percomputers attempt to predict the three-dimensional
  including the following:                                             form of proteins on the basis of their amino acid se-
  z Sequence analysis is the original and core activity of             quence.
      bioinformatics. For example, special programs have           z   Use of databases and networking of data and laborato-
      been used to sift through the three billion building             ries are basic prerequisites for smooth performance and
      blocks that make up our genome in order to identify              evaluation of experiments. Known information must be
      genes and regulatory elements. Such programs can also            accessible everywhere at all times and must be available
      perform sequence analysis of proteins and translation in         in a form that is compatible with radically different com-
      both directions (from gene sequence to protein and               puters, operating systems and programs.

                     In theory, the DNA chip technique is also suitable for the inves-
                     tigation of proteins, since in principle any molecule can be in-
                     vestigated on the silicon surface of chips. For this purpose, how-
                     ever, 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 pro-
                     teins than with a relatively stable and structurally simple mole-
                     cule such as DNA.

PCR – polymerase chain              Vanishingly small amounts of DNA can be ren-
reaction                            dered visible by means of the polymerase chain
                                    reaction (PCR). With the aid of the DNA-extend-
                     ing 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 pos-
                     sible, 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

                                                                                                         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               ‘High-throughput screening’ is an important aid
screening (HTS)               in the search for suitable drugs for a given target.
                              In this technique thousands of chemical sub-
              stances (for example, modifications of a computer-designed
              molecule intended to fit into the binding site of a protein vari-
              ant that causes disease) are automatically tested for certain
              properties (such as binding to the intended target). The data ob-
              tained are analysed by computer and candidate substances un-
              dergo a further round of improvement. This continues until a
              molecule with optimal characteristics has been found. High-
              throughput screening now performs reliably, rapidly and cheap-
              ly a task that only a few years ago required an enormous amount
              of work on the part of hundreds of biologists, chemists and

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 un-
                known biomolecules, however these are not (yet) targets. In or-
                der to qualify as targets, such molecules must not only play an
                important role in the body, but also make a significant contri-
                bution 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 ear-
                ly 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 specif-
                ic 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 ad-
                dressing 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 thread-
                worms, since it was in these animals that the genes that play a cru-
                cial 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

                                                   Depiction of a target molecule with bound drug.

                                                                                Targets for medicine   57
Transgenic mice as a            Humans and mice share 99 percent of their genes.
model                           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, there-
               fore, 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 compari-
               son 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 insert-
               ed into the germ line of mice. This can now be done with any hu-
               man gene. Animals into which a small amount of foreign genet-
               ic 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 oc-
               cur 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
               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 mod-
               ified 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 hu-
               mans, in the early phases of drug development.

Important objects of research: transgenic mice

Transgenic mice are among the most important tools of             into the genome of the resulting zygote. The ovum is
molecular medicine. These animals have been genetically           then implanted into the uterus of a female animal. About
modified either by insertion of additional – mostly human –       a quarter of the offspring produced in this way contain
genes into their genome or by selective removal of certain        the desired gene in their genome and subsequently
of their genes. Genes can also be altered or else removed         transmit it to their own offspring. This method can be
and then replaced by others. All such types of transgenic         used only to add genes to an animal’s genome.
mice are excellent models for investigating the influence of   2) By contrast, the technique of microinjection into the blas-
genes and environmental factors on the development and            tocyst can be used to specifically modify or eliminate an
progression of diseases. The breeding and use of animals          animal's own genes. To date the only mammal in which
for pharmaceutical research are strictly regulated in all         this technique has been performed successfully is the
European countries and in most cases are under the con-           mouse. This technique employs embryonic mouse stem
trol of independent committees that include representatives       cells – i.e. cells that still have the potential to develop
of animal protection groups.                                      into any kind of cell – whose genome has been altered
Two techniques are used to insert foreign genes into ani-         in the desired way by means of recombinant technology.
mals:                                                             These cells are injected into a mouse blastocyst (multi-
1) The technique of microinjection into the pronucleus of         cellular stage of a developing embryo). The embryo is
    fertilised ova was introduced in the early 1980s and has      then reimplanted into the mother, where it develops into
    now been performed successfully in a number of animal         a chimeric mouse, only a proportion of whose cells pos-
    species. A solution containing many copies of the de-         sess the desired genetic modification. Those offspring
    sired gene is injected into the maternal or paternal          of such chimeric mice whose ovaries have developed
    pronucleus (undeveloped cell nucleus) of a fertilised         from the inserted stem cells will bear the altered genes
    ovum. Copies of the gene are randomly incorporated            in all their cells.

   Microinjection of DNA                                         Controlled mutagenesis: blastocyst injection
                                                                       embryonic stem cells

        unfertilised ovum                                                                          cell injection into
                                            DNA-                                                   the blastocyst
                                                               donor blastocyst
          fertilised ovum                                                                reimplantation in mouse
                                                                                         prepared for pregnancy
      mouse prepared
      for pregnancy

                                                                         chimeric mouse

                    transgenic offspring                                 offspring with the desired mutation

                                                                                                    Targets for medicine        59
The search becomes more        The existence of so many technical possibilities
difficult                      does not make the search for targets in any way
                               routine. All hopefulness notwithstanding, it re-
               mains 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 – mol-
               ecules, in other words, for which the risk of failure in drug re-
               search is relatively small.
               The days of the gold rush in speculative target finding are there-
               fore 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 re-
               mains is more or less known already. The scientists of the Hu-
               man 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 tran-
               scribed 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. Esti-
               mates 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.

Weighing risks against          Nevertheless, the search is worthwhile, because
benefits                        we still lack causal therapeutic options for use
                                against far too many of the major diseases that
               afflict mankind, including cancer, cardiovascular disease, Alz-
               heimer’s disease and diabetes. Even in the case of infectious dis-
               eases, medicine is at a very early stage in terms of the possibil-
               ities 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 in-
               correctly classified, such as allergies and autoimmune diseases.
               And finally, many rare diseases have become objects of drug re-

              search now that the existence of global markets and public fund-
              ing has made the development of drugs for use against them eco-
              nomically 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 ex-
              penditure will be required in order to develop such an agent? Be-
              fore these questions can be answered, certain details of the struc-
              ture 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.

Drug research must also       A third important factor influencing the evalua-
pay for itself                tion 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
              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 in-
                  fluences economic decision-making. This gives rise to con-
                  flicts 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.

                                                                              Targets for medicine   61
     z Another peculiarity of privately funded pharmaceutical re-
       search is the composition of the ‘electoral college’ that par-
       ticipates in the decision-making process. This includes not
       just the company itself in the form of its management, re-
       search, 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 asso-
     z An additional peculiarity of the pharmaceutical industry is
       the sheer unpredictability of biology. All scientific and tech-
       nical 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 im-
       portant in that it has a substantial impact on the economic
       significance of each and every target. A newly discovered bio-
       molecule 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 tak-
       en, 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 un-
       certainties are of the greatest importance and the cost of re-
       searching and developing new drugs is correspondingly high.
     z Undoubtedly one of the most-discussed aspects of the eco-
       nomics 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. Promi-
       nent among these is the fact that the distinction between a
       (nonpatentable) discovery and a (patentable) invention is of-
       ten difficult to make in biology. It is also argued that over-
       generous 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.

                                                  research director

                          financial analysts                               profession

                                                 target selection
                   market group                                                     advocates

                                          drug hunter           regulatory agency

               Many different constituencies with partly overlapping and conflicting in-
               terests influence target selection in pharmaceutical research.

Between risk and return        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 de-
               velopment of new drugs has become riskier due to ever-increas-
               ing 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, nei-
               ther of these factors is easy to determine – there are many ex-
               amples 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 now-
               adays generating financial projections that attempt to take ac-
               count 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 thera-
               pies, since every new target is at the same time a diagnostic tool

                                                                                                Targets for medicine   63
     Good fortune: unexpected blockbusters

     active ingredient             position in                 indication                                  original sales
     (product)                     drug class                                                              forecast

     Tamoxifen (Novaldex)          1st                         breast cancer                               £ 100,000

     Captopril (Capoten)           1st                         hypertension, heart failure                 USD 20 million

     Cimetidine (Tagamet)          1st                         peptic ulcer                                £ 700,000
     Fluoxetine (Prozac)           2                           depression                                  ?
     Atorvostatin (Lipitor)        5                           hypercholesterolemia                        ?

                      that can improve our understanding of the disease process in pa-
                      tients. We must therefore hope that the ‘pop stars’ of medicine
                      continue to make headlines.


                      Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considera-
                          tions. 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:
                      Geschäftsstelle des Wissenschaftlichen Koordinierungskomitees des Deutschen
                          Humangenomprojekts (ed.): Das Humangenomprojekt – 1st and 2nd edition

PCR: an outstanding method

Scarcely any invention has altered
biological science so radically in
such a short period as the polyme-
rase 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 tech-
nique 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 poly-
                      merase chain reaction. Among the instruments and glassware of
                      his laboratory Mullis might never have had the most momen-
                      tous 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 mid-
                                                                 1980s the technique was
                                                                 used for the first time to di-
                                                                 agnose a disease, when re-
 DNA deoxyribonucleic acid; the chemical substance of our        searchers identified the gene
 genes                                                           for sickle cell anemia. At
 RNA ribonucleic acid; the chemical substance that makes
 up the working copies of genes (mRNA), among other things       about the same time the
 Nucleic acids a chemical term that covers both DNA and          method was introduced in-
 RNA; nucleic acids are molecules consisting of long chains of
                                                                 to forensic medicine. The
 nucleotides linked together
 Nucleotides the building blocks of DNA; they comprise           polymerase chain reaction
 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
 least one phosphate group; without the phosphate group
                                                                 honour for its inventor in
 these building blocks are referred to as nucleosides            record time: In 1993, just
 Sequence the order of the nucleotides in DNA (DNA se-           ten years after his historical
 quence) or RNA (RNA sequence)
 Primer a short DNA fragment with a defined sequence that        car journey, Kary Mullis re-
 serves as an extension point for polymerases                    ceived the Nobel Prize for
 Polymerases enzymes that link individual nucleotides to-        Chemistry. The reason for
 gether to form long DNA or RNA chains
 Hybridisation (annealing) the joining of two comple-            this extraordinary success is
 mentary 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
 form a sequence of perfect pairings are able to form a stable   pressing problems facing
 double strand and are referred to as complementary strands      biology at the time – the
                                                                 replication of DNA.

Rapid DNA cloning                 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. Theoreti-
                  cally, 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

                      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 amp-
                      lify, the DNA of interest. For this reason PCR and DNA chips
                      often go hand in hand.

Simple and effective: the PCR principle

                                                                                                        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
                                                          denaturing                                        are attached to the primers) is heated to 95°C. This cau-
                                                                                                            ses the two complementary strands to separate, a pro-
                                                                                                            cess 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
                                                                       Polymerase-Kettenreaktion: PCR


                                                                                                            primer and DNA segment are complementary, i.e. if the
                                                      hybridisation                                         base pairs of the primer and DNA segment match. The
                                                                                                            polymerases then begin to attach additional comple-
                                                                                                            mentary nucleotides at these sites, thus strengthening
                            polymerase                                                                      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 mil-
                                                                                                        lion molecules are cloned from a single segment of double-
                                                                                                        stranded DNA.
                              repeat cycles …


                                                                                                        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.

                                                                                                                                  PCR: an outstanding method            67
Copies of copies              The basic PCR principle is simple. As the name
of copies                     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 poly-
               merases 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

Copying unknown DNA            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 possi-
               ble 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 molecu-
               lar cloning in which DNA fragments are tediously copied in bac-
               teria 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 mole-
               cules, e.g. of chromosomes, can be used in PCR to scout further
               into unknown areas.

Help from hot springs          Soon after its discovery the PCR method was re-
                               fined in several ways. One of the first modifica-
               tions of the original protocol concerned the polymerases used.
               Like all enzymes, polymerases function best at the body tem-
               perature of the organism in which they originate – 37°C in the
               case of polymerases isolated from humans. Below this tempera-
               ture the enzyme’s activity declines steeply, above this tempera-
               ture it is quickly destroyed. In PCR, however, the two strands of
               the DNA molecule must be separated in order to permit the pri-
               mers 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 cy-
               cle – a time-consuming and expensive proposition.
               A solution was found in hot
               springs. Certain microorga-
               nisms thrive in such hot
               pools under the most inhos-
               pitable conditions, at tempe-
               ratures that can reach 100°C
               and in some cases in the pre-
               sence of extreme salt or acid
               concentrations. The poly-
               merases of these organisms
               are adapted to high tem-
               peratures and are therefore
               ideal for use in PCR. Today
               the polymerases used in           Hot spring: Thermus aquaticus. Nearly all PCR tech-
               nearly all PCR methods the        niques in the world now use the Taq polymerases isolat-
                                                 ed from the bacterium Thermus aquaticus, a microor-
               world 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 poly-
               merase, supports an entire industry. The organism was original-
               ly discovered in a 70°C spring near Great Fountain Geyser in Yel-
               lowstone National Park in the USA. Employees of Cetus, who
               Kary Mullis was working for at the time of his discovery, isolat-
               ed the first samples from the hot spring and then cultivated in
               the laboratory one of the most useful bacterial strains known to-
               day. Meanwhile Thermus aquaticus has been found in similar
               hot springs all over the world.

                                                                           PCR: an outstanding method   69
Further developments           The introduction of Taq polymerase has certain-
around the world               ly not been the only modification to the PCR
                               method. This was helped by the fact that Mullis
               published his discovery relatively early – though not with-
               out 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 mod-
               ified 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.

Forerunner of genomics:         In the 1990s biology was faced with one overrid-
DNA sequencing                  ing preoccupation: the unravelling of the genome.
                                Thanks to huge technical and organisational ef-
               forts, first viruses and bacteria, then yeasts, plants and animals
               relinquished the secrets of their genetic material. This accom-
               plishment 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 ex-
               act 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 (pro-
               nounced ‘snips’), single nucleotide changes in the genome which
               appear to account for a large proportion of the genetic dif-
               ferences 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, scien-
               tists 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.


               DNA, our genetic material.

Sensitive determination:      When PCR is used only for detecting a specific
qualitative PCR               DNA segment, the method is referred to as qual-
                              itative 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 re-
               gions of the genome are amplified and examined. However, al-
               though 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 immunolog-
               ical 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

                                                                      PCR: an outstanding method   71
                                                   has been eradicated, whether
                                                   it is chronic (and might there-
                                                   fore progress unnoticed)
                                                   and whether the individual
                                                   has been reinfected with a
                                                   different but related path-
                                                   ogen. Many viruses contain
                                                   RNA rather than DNA. In
                                                   such cases the viral genome
                                                   has to be transcribed before
                                                   PCR is performed, and RT-
Human immunodeficiency virus.
                                                   PCR is therefore used.
                                                   Sometimes it is also neces-
               sary 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 prob-
               lems 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.

More than just yes or no:        Quantitative PCR provides additional informa-
quantitative PCR                 tion beyond mere detection of DNA. It indicates
                                 not just whether a specific DNA segment is pres-
                 ent in a sample, but also how much of it is there. This informa-
                 tion is required in a number of applications ranging from med-
                 ical diagnostic testing through target searches to basic research.
                 Consequently, although quantitative PCR was not described un-
                 til the 1990s, the method already exists in a number of variants
                 and protocols to meet a broad range of requirements. Theoreti-
                 cally 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 dou-
                 ble 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 pres-
                 ent, while in the final cycles the enzymes’ activity declines as a

                  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 (tem-
                  perature, 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 sam-
                  ple from the amount of DNA present at the end of PCR.
                  Instead, researchers have developed various methods that deter-
                  mine the number of new DNA molecules formed in the reac-
                  tion, 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 conju-
                  gating 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). How-
                  ever, 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 satisfac-
                  torily performed with available proteomic methods. Further-
                  more, 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              a double strand. Nevertheless, the principle is still used,
     present in a PCR sample to be determined. Many tech-                though usually with other dyes that specifically interact
     niques are available for calculating the number of DNA co-          only with the desired DNA product. The experiments
     pies formed during the individual steps of the PCR proce-           have been simplified by the introduction of special
     dure and thus for deriving the quantity originally present in       equipment such as the Roche LightCycler, which auto-
     the sample. They are usually based on the middle, or ex-            mates the entire procedure, heating and cooling the
     ponential, phase of the PCR in which the amount of DNA              solutions, stimulating the dye to fluoresce and contin-
     template is approximately doubled in each cycle.                    uously monitoring the fluorescence. Special computer
     z Example: competitive PCR – This method is now largely             programs help to analyse the data.
         of historical significance only. It was one of the first    z   Example: TaqMan probes: One way to measure only the
         quantitative PCR methods developed. In addition to the          desired DNA product during PCR is to use TaqMan
         template of interest, another DNA template having a             probes, short DNA fragments that anneal to a middle
         very similar sequence was added to the same reaction            region of the template DNA (see below). The probes
         vessel. Both DNA strands were then cloned simulta-              bear a reporter dye (R) at one end and a quencher (Q)
         neously under identical conditions. The amount of tem-          at the other. Quenchers are molecules that quench the
         plate and the amount of ‘competitive’ DNA formed pro-           fluorescence of dyes in their proximity. The polymerases
         vided at least a rough estimate of the amount of DNA            in the PCR solution are able to break down the TaqMan
         present in the original sample.                                 probes during the doubling of the DNA template. In so
     z Example: real-time PCR – Most of the quantitative PCR             doing they free the reporter dye, which then migrates
         methods in use today are based on a 1922 discovery by           away from the influence of the quencher. Hence the
         the American Russ Higuchi, who used the dye ethidium            fluorescence of the dye is measurable only if the poly-
         bromide (EtBr). Embedded in double-stranded DNA,                merase has in fact copied the desired DNA strand. Each
         EtBr fluoresces when stimulated by light. The observed          freed molecule of reporter dye represents a DNA strand
         fluorescence therefore indicates the amount of DNA              that has been formed. TaqMan probes can therefore be
         formed and does so at any given time during the PCR             used to measure the amount of DNA formed at any
         reaction, hence the name real-time PCR. In this method          given time.
         parallel runs are performed with the same known quan-
         tity 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

              primer                      R        probe        Q

             forward                           R
                                                    probe        Q

                                                                     probe         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
                In accordance with the principle of RT-PCR, the mRNA is re-
                verse-transcribed into DNA. Its original quantity is then deter-
                mined by quantitative real-time PCR. This provides an inform-
                ative 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 sam-
                ples is ‘single target expression profiling’, or STEP. In this tech-
                nique 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, pro-       The values obtained are plotted and compared. The exam-
     vides an overview of the expression of a specific gene in var-    ple below is a comparative diagram showing the expression
     ious samples. In this method the number of working copies         profile of various tissues, including samples from the colon
     of a specific gene is measured by quantitative PCR. The           of a healthy subject and of a patient with cancer of the colon
     samples can come from different tissues of an individual or       (circles). The gene in question was suspected of being more
     from tissues of many different donors. By comparing the           active in the presence of colon cancer. However, the STEP
     profiles in healthy and ill people it is possible to determine,   procedure showed that the colon-specific gene is equally
     for example, whether a gene in question really is associated      active in healthy and ill individuals.
     with the development or progression of a disease.

                colon cancer                                               colon normal

                        ured and plotted. In many cases this reveals whether a target re-
                        ally 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 bi-
                        ological 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 impor-
                        tant 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 exter-
                        nal factors such as administration of drugs.

Applications in the field              Another important application of quantitative
of infectious diseases                 PCR is in molecular diagnosis, i.e. the diagnosis
                                       of diseases based on molecular findings rather
                        than on physiological symptoms. In this connection the diag-
                        nosis of viral diseases is an area that is gaining increasing im-
                        portance. For simple diagnostic testing, i.e. to determine if a
                        pathogen is present in the patient’s body, qualitative PCR is suf-
                        ficient. However, to follow the progress of a disease and to help

choose the right treatment doctors often need to know the ac-
tual 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 in-
fections, 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 path-
ogen 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 diagnos-
ing HCV infection. But the quantitative form of the technique
opens up whole new perspectives for the treatment of the dis-
ease. 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 eradicat-
ed or has become chronic.

Hepatitis C viruses.

                                                      PCR: an outstanding method   77
Example: HIV                   Another important application in which quanti-
                               tative PCR is used in the field of infectious dis-
               eases is AIDS. Those infected with the causative agent of the dis-
               ease, 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.

Applications                    Genetic factors are always involved in the devel-
in cancer therapy               opment of cancer. Their contribution varies great-
                                ly 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 develop-
               ment of cancer is therefore an important step towards improv-
               ing treatment. Both qualitative and quantitative PCR play a cru-
               cial 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 rel-
               atively 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 al-
               ready has cancer – how aggressively it can be expected to pro-
               gress. Because multiple changes have usually accumulated be-
               fore 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 chap-
               ter on DNA chips).

Example:                          Meanwhile, quantitative PCR also is gaining im-
promoter methylation              portance 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 mole-
                  cules (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 si-
                  lent and no gene product is formed.
                  Only in recent years has it emerged that this mechanism of pro-
                  moter 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 methyl-
                  ation-specific PCR (MSP). In this relatively new technique cel-
                  lular DNA is first treated with sodium bisulphite, which con-
                  verts normal cytosine to the RNA building block uracil but
                  leaves methylated cytosine intact. This results in different prod-
                  ucts 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.

A host of other                    New applications for PCR are still emerging,
applications                       particularly in the field of medicine. The search
                                   for genetic predispositions to diseases is an espe-
                  cially important area of research. In many cases the onset of a
                  disease can be prevented or at least delayed by lifestyle modifi-
                  cation or the taking of medications. One example of this is os-
                  teoporosis, a loss of bone density that is especially common in
                  postmenopausal women. Because the disease tends to run in fa-
                  milies, it is clear that genetic factors are involved in its develop-
                  ment and progression. An intensive search for the genes in-
                  volved 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 nu-
                       cleic 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   CH3

                     CGCGTCG                            ATC               CGCGTCG                       ATC

                                                   1.         NaHSO3

                                                                          CH3 CH3   CH3
                     UGUGTUG                            ATC               CGCGTCG                        ATC

                                                              primer 1
                                                              primer 2
                                                                          CH3 CH3   CH3

                     UGUGTUG                            ATC               CGCGTCG                       ATC
                     ACACAAC                                              GCGCAGC
                        primer 1                                            primer 2
                      GCGCAGC              no product                     ACACAAC             no product
                       primer 2                                            primer 1

                                                   3.              PCR

                                        PCR product                          PCR product
                                   only with primer 1                        only with primer 2

     In many tumours important genes that control cell growth        cytosines, by contrast, remain unchanged (1). The subse-
     are switched off by methylation of the promoter region.         quent PCR procedure then uses specific primers for the
     These changes can be detected by means of methylation-          various products formed (2). Hence, either the original
     specific PCR (MSP).                                             methylated or the unmethylated DNA is copied. The origi-
     In MSP all the normal cytosines (C) of the original DNA are     nal DNA was therefore methylated or not (3) depending on
     converted to the RNA building block uracil. The methylated      the primer used to obtain a product.


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 PCR-
     Verfahren. 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
Mullis KB: The polymerase chain reaction. Nobel Lecture, December 8, 1993
Kary Mullis Website:
Deutsches Hepatitis C Forum e.V. Homepage – Qualitativer HCV-RNA Nachweis:

                                                                             PCR: an outstanding method   81
SNPs: the great importance
of small differences

                                       G   C   A   A   T   C   T   A
                                       G   C   A   A   T   C   T   A
                                       G   C   A   A   T   A   T   A
                                       G   C   A   A   T   C   T   A
                                       G   C   A   A   T   C   T   A
                                       G   C   A   A   T   A   T   A

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.
                       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 instan-
                       ce, they may contribute to us being as restless as our grandmoth-
                                                                  er, or to putting on weight
 Terms                                                            at as early an age as did
                                                                  our grandfather). Sometimes
 SNPs single nucleotide polymorphisms – differences in
 individual building blocks (base pairs) of DNA that are distri-  these small differences can
 buted randomly over the genome and passed from generation        have more significant conse-
 to generation.                                                   quences (for instance, they
 Genotype the alternative forms (alleles) of a gene present
 in an individual; generally there is a maximum of two – one from may increase our risk of de-
 the father and one from the mother.                              veloping a disease or lower
 Phenotype the constitution of a living creature that results     the likelihood that we’ll re-
 from the interaction of genotype and environmental influences.
 Phenotype refers both to medically irrelevant characteristics
                                                                  spond to a medicine). Very
 and to diseases.                                                 rarely – mostly in the case of
 Pharmacogenetics the branch of science concerned                 typical inheritable diseases –
 with the influence of genetic variation on the effectiveness and
                                                                  their mere presence or ab-
 side effects of drugs.
 Targets the molecules – proteins or small organic mole-          sence determines whether a
 cules – upon which drugs act in our body.                        family member will develop
                                                                  a disease or remain healthy.
                                                                  Most of these genetic differ-
                       ences between individuals consist of single nucleotide polymor-
                       phisms, or SNPs (pronounced ‘snips’). SNPs are randomly dis-
                       tributed variations of the building blocks of our genome that
                       make each of us genetically unique. They contribute to family re-
                       semblance with regard both to external features and to the risk of
                       developing certain disorders. In medicine, SNPs have become im-
                       portant parameters in the search for new, safer, more effective and
                       better tolerated drugs aimed at providing more targeted therapy.

SNPs: common,                      99.9 percent of the human genome is identical in
hereditary, stable                 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.

The small difference: SNPs

                                                 G   C   A   A   T   C   T   A
                                                                                                                   SNP C
                                                 G   C   A   A   T   C   T   A
                                                 G   C   A   A   T   A   T   A
                                                 G   C   A   A   T   C   T   A

                                                 G   C   A   A   T   C   T   A
                                                                                                                   SNP A
                                                 G   C   A   A   T   A   T   A

DNA sequence                               DNA sequences differing by a                                individuals with
                                           single nucleotide                                           different SNPs

Single nucleotide polymorphisms, or SNPs, are differences            In the following theoretical example, replacement of a cyti-
in individual building blocks (nucleotides) of DNA that are          dine (C) with a guanine (G) in the gene results in formation
distributed randomly over the genome. They can occur in              of an amino acid with completely different properties:
any position within or outside of genes and accordingly can          instead of the large, basic amino acid glutamine (Gln), the
have very different effects. SNPs present within the protein-        small, neutral amino acid glycine (Gly) is formed.
encoding 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 alter-
native amino acid differs from the normally incorporated                     unchanged            SNP
one, such an amino acid exchange can have a profound                 Gene    AAG-CGA-ATT-AGG › AAG-GGA-ATT-AGG
influence on the function of the protein.                            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 tem-
                      plate for a protein. The fact that they are inherited with these
                      genes makes gSNPs useful for the study of associations be-
                      tween the gene (and its variants) and certain phenotypes.
                      Mapping gSNPs may be functionally relevant if they influ-
                      ence important control elements of the gene and thereby de-
                      crease or increase transcription of a gene.

                                                                             SNPs: the great importance of small differences        85
     The less common, the more important: SNP groups


              pr           ex in

     Genes are arranged in different ways on chromosomes. They             gSNPs (gene-associated SNPs) are situated alongside
     generally consist of a promotor region (pr), in which the con-        genes or in introns. They too are used mostly for gene map-
     trol elements of the gene are located; exons (ex), the seg-           ping; they can also influence the control of gene activation.
     ments of the gene that are translated into the corresponding          gSNPs are less common than rSNPs (estimated at less
     protein (the gene product); and introns (in), segments located        than 1 million).
     between exons that are not part of the protein template. SNPs     ·   cSNPs (coding SNPs) are situated in exons and often in-
     occur in all regions of the genome and are subdivided into four       fluence the function of the corresponding gene product
     groups on the basis of their site of occurrence:                      (estimated at about 100,000).

                                                                           pSNPs (phenotype-relevant SNPs) are gSNPs or cSNPs

         rSNPs (random SNPs) are situated at a distance from               that influence the constitution of the organism. These are
         genes and are used mostly as aids to mapping. They are            the most important type of SNP from the point of view of
         relatively common (estimated at 3–6 million in humans).           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 con-
                          cerned if they result in incorporation of an alternative amino
                       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 pro-
                          tein 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 technol-
                       ogy platforms with very high accuracy and reproducibility.

Medical effect at               The following examples show how important
two levels                      SNPs are in relation to medical practice. The ef-
                                fects 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 pre-
                cise, with the way in which the body deals with drugs; and on the
                other hand, involvement in the development and progression of
                Variations of the first category may help explain why drugs work
                more or less well in some people and why they may cause unde-
                sirable side effects only in certain people. Knowledge of the po-
                sition 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 in-
                dividuals differ in terms of their susceptibility to certain diseases
                despite living in similar environmental conditions.

SNPs relevant to drug         Gene variants can affect the action of drugs in
response                      various ways, including:
(pharmacogenetics)            z Absorption. The uptake of a drug into the
                              body can be disturbed, with a corresponding re-
                              duction 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 re-
                  actions 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. med-
                  icines) commonly need to be chemically altered (broken
                  down) in the body so that they can be eliminated. If this hap-
                  pens too rapidly, the effectiveness of the drug may be re-
                  duced, whereas if breakdown is too slow, the drug remains in
                  the body for too long and reaches excessive levels that in-
                  crease the likelihood of undesirable effects. Moreover, a
                  harmless molecule may be converted into one with danger-
                  ous properties, e.g. one that promotes cancer.

                                                         SNPs: the great importance of small differences   87
     Breakdown four times faster: cyp2c19 and omeprazole

     The cyp2c19 gene of the cytochrome P450
     gene family plays an important role in the                        100
                                                                                 gastric ulcer
     metabolism of fat-soluble drugs. Forexample,
     the proton pump inhibitor omeprazole works                                  peptic ulcer
     up to four times less well in individuals with                    80
     a certain variant of this gene (‘mut’) than

                                                      cure rate in %
     in those with the most common genotype                            60
     (‘wt’). This is because the drug is broken
     down faster in individuals with the mut geno-
     type. In this group of patients the cure rate                     40
     can accordingly be improved by giving
     higher doses of the drug.

                                                                             mut/mut        wt/mut        wt/wt

                                                                                       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

cyp2c19 in                            SNPs assume particular importance whenever
stomach ulcers                        they are associated with the effectiveness or tol-
                                      erability 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 en-
                     suring 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

                      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 in-
                      dividuals. 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.

Beta2-adrenergic                    Another example of pharmacogenetically impor-
receptors                           tant SNPs is the gene for the beta2-adrenergic re-
                                    ceptor. 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

   Triple role: beta2-adrenergic receptors

                                                                    27         16


                                                             164                             cell membrane


   The gene for the beta2-adrenergic receptor has at least                   tor is significantly associated with an increased risk of
   three medically important variants. One of them is of phar-               developing end-stage heart failure in the aftermath of a
   macogenetic relevance, the two others signal increased                    major myocardial infarction. Thus, it has been sug-
   risk for disease or disease prognosis.                                    gested that bearers of a certain amino acid substitution
   z The antiasthmatic agent albuterol works well only in pa-                at this position should be followed particularly closely
       tients in whom the amino acid arginine is present at                  and may ultimately require heart transplantation (dis-
       position 16 of the receptor, whereas it is less effective             ease prognosis/outcome).
       in patients with a variant of the gene that results in the        z   A genetic variant that influences position 27 of the
       presence of glycine at thisposition (pharmacogenetics).               beta2-adrenergic receptor appears to play a role in the
   z The genotype that codes for position 164 of the recep-                  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.

SNPs relevant to the           While we generally view the role of SNPs as being
development and                to affect the relative risk of contracting a disease,
progression of diseases        in extreme cases a single SNP may actually cause
                               a disease. In sickle cell anemia, for instance, sub-
               stitution of a single nucleotide in the gene for hemoglobin, the
               oxygen-carrying red blood pigment, results in synthesis of an al-
               tered protein which under certain chemical conditions takes on
               an abnormal shape. This causes the affected red blood cells to as-
               sume the shape of a sickle (hence the name of the disease),
               clump together and potentially block small blood vessels, lead-
               ing to tissue death and extreme pain.
               The great majority of diseases that afflict mankind arise as a re-
               sult 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 devel-
               oping a certain disease. Though these tests can only indicate a
               somewhat higher or lower risk, they may provide prognostic in-
               formation that allows the person concerned to make more in-
               formed 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).

Drug studies and               SNPs are therefore important at all levels of drug
regulatory approval            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 distri-
               bution and effects of SNPs is already having a noticeable impact
               at the latter two levels. For example, the fact that some people

               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 specif-
               ic 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 dis-
               covery of new medicines, but to diagnostic possibilities that per-
               mit 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 en-
               hancing response rates (efficacy) if these issues arise in the
               course of clinical trials.

Paving the way for             Given the likely medical importance of SNPs, it
new methods                    was important to develop the knowledge base re-
                               quired for finding them and to spur on the devel-
               opment of increasingly high-throughput analytical platforms
               and technologies. The task of finding and then evaluating sever-
               al 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 exten-
               sive examination of our genome for SNPs. For SNP research,
               therefore, increasing automation and miniaturisation of biolog-
               ical methods were of pivotal importance.

The SNP Consortium           Although SNP research on a large scale has there-
                             fore been possible only for a relatively short peri-
               od of time, a large number of these variations in the human

                                                       SNPs: the great importance of small differences   91
     Smaller, faster, automated: methods of studying SNPs

     Techniques developed or refined in the past few years per-     the search for associations between SNPs and diseases or
     mit systematic study of SNPs. Automation and miniaturisa-      pharmacogenetic problems are association studies. These
     tion, in particular, have now made it possible to perform in   are used to determine whether certain SNPs prevail in indi-
     a single step thousands of experiments which only a few        viduals with a particular disease. Another important method
     decades ago would have required years of work in major         is that of in-vitro assays to investigate the properties of pro-
     research institutions. Evaluation of the enormous quantities   teins. These can rapidly determine whether the presence of
     of data gathered in this way has now become the subject        a certain SNP in a gene results in altered function of the
     of a special branch of science known as bioinformatics.        corresponding gene product.
     Three stages of research can be distinguished:
                                                                    3. SNP tests for patients
     1. Search for SNPs                                             The final link in the chain of applications of SNP research is
     A major part of the search for SNPs is now conducted in sil-   the development of tests to determine the presence of
     ico, i.e. by computer. The results of laboratory experiments   SNPs in patients. It is not yet clear which of the various
     are stored in databases which are then compared in com-        techniques being developed for this purpose will prove to
     puters and scoured for SNPs. Special computer programs         be both economically viable and scientifically satisfactory
     are used as aids in the search; these also ensure that SNPs    and therefore become established as the norm. A ‘line
     are distinguished from sequencing errors. Since individual     array’, for example, can detect 60 variants on a membrane.
     SNPs can be very rare, it is still necessary to sequence the   Far greater numbers can be detected using a ‘SNP chip’. In
     genome of many additional volunteers in order to obtain        this technique, several thousand SNPs can be accommo-
     comprehensive data. These data are anonymised. The most        dated and simultaneously tested on a DNA chip. Another
     important technique used in DNA sequencing is the poly-        method uses MALDI-TOF (see chapter on proteomics) to
     merase chain reaction, or PCR (see chapter on PCR).            investigate SNP-specific oligonucleotides. Also used are a
                                                                    variety of other methods that employ various techniques
     2. Evaluation of SNPs                                          including PCR and nanotechnology.
     Databases and special computer programs are also used to
     identify associations between SNPs and genes. Essential in

     SNP analysis by computer.

               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 UK-
               based 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 to-
               tal 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 fund-
               ed 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 data-
               base that was built up – a comprehensive SNP map of the hu-
               man 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.

Current areas of SNP           Based on this body of knowledge, efforts are now
research                       being directed towards the discovery of medical-
                               ly 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 dis-
                  ease 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 statisti-
                  cally significant association with a phenotype, without pre-
                  conceived notions about likely candidate genes, is the next

                                                        SNPs: the great importance of small differences   93
                    logical, but very challenging, step in the use of SNPs. This ap-
                    proach 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.

SNPs – opening new              From the perspective of patients, SNPs are a ma-
doors to better healthcare      jor step towards a more personalised approach to
                                treatment that takes genetic differences between
                people into account. From the perspective of the pharmaceuti-
                cal 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 even-
                tually to the development of new and better drugs, and they bear
                the promise of leading to more effective, safer and better toler-
                ated 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.


                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:
                TSC data on the CYP2C19 gene:
                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

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 devel-
oped 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
                       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 presort-
                       ed 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
 Biochip a solid substrate (e.g. glass or plastic) upon which
                                                              catching enormous numbers
 biomolecules are anchored.                                   of different types of fish at
 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
 company Affymetrix.                                          lines are known as biochips,
 DNA deoxyribonucleic acid; the chemical substance of which   and there can be little doubt
 our genetic material consists.                               that a bright future awaits
 RNA ribonucleic acid; the chemical substance of which,
 among other things, working copies of genes (mRNA) consist.  them. In medicine, at least,
 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;
 chain-shaped molecules whose individual building blocks are  and therapy upside down.
 bases/nucleotides.                                           Biochips are among the most
 Oligonucleotides short nucleic acid chains composed of at
 most a dozen building blocks (nucleotides).
                                                              important instruments used
 Genes functional segments of our genetic material that serve in the miniaturisation and
 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 reliabil-
                       ity. The ‘hooks’ used for this purpose are molecular probes at-
                       tached 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 compo-
                       nents of our cells can serve as a molecular probe.

The most important               In the type of biochip that is most important at
biochips at present:             present, the molecular probe that is attached to
DNA chips                        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

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 biolo-
              gical 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 nucle-
              ic acid chains to form a double-stranded structure has been ex-
              ploited 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 solu-
              tion 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 oligonu-
              cleotides 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 so-
              lution – 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
            probe        3'
                                T       A       C       T   A   A       G   T                   T
                                                                                C       G               A

           target DNA

                    T   G       A       T       G       A   T       T   C   A   G       C       A       T        C       C   G
              5'                                                                                                                 3'


                            T       A       C       T       A   A       G   T       C       G       T       A

             T      G       A       T       G       A       T   T       C   A       G       C       A       T        C       C    G

Field of application:               The most important application of DNA chips is
gene research                       in the search for, and the study of, genes. In this
                                    application, as in the classical oligonucleotide ex-
                    periment, 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 so-
                    lution 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 meth-
                    ods 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.

  Common example: the GeneChip

    a                    b                    c                    d                  e                   f

                                                      A                                   G    G              A T CG
                                                  A                                                G          AGAC
                                                           A                                  G               CGTC
                                                      A                A                      AG              T AG T

  One of the most commonly used DNA chips at present is            DNA solution. Commonly performed experiments include
  Affymetrix’s GeneChip. This is an oligonucleotide chip in        genome studies and, in particular, gene expression profiles.
  which the short strands of DNA are synthesised in situ, i.e.     These are used to identify those genes that are actually
  on the chip, by means of photolithography (see above):           expressed in a given cell type or tissue.
  a. Reactive sites on the chip surface are blocked by pho-        For this purpose the mRNA – the working copy of genes –
        tosensitive protector groups (small squares).              in a cell extract is transcribed into cDNA. This process is
  b. An opaque mask covers the greater part of the chip;           known as ‘reverse transcription’, as opposed to transcrip-
        the beam of light therefore removes only those pro-        tion, which is the synthesis of RNA on the basis of DNA.
        tective groups that are situated in a certain region.      Precisely this can occur in a second step, since in many
  c. The chip is incubated with a solution containing one of       experiments the cDNA that is formed is transcribed back
        the four nucleosides adenosine, thymidine, cytidine or     into cRNA. In one of these steps a label is introduced; com-
        guanosine (A, T, C or G), which likewise bear protec-      monly used for this purpose is, for example, the molecule
        tor groups; the nucleosides react with the previously      biotin. The cRNA (or cDNA) is then cut into smaller pieces
        unmasked regions of the chip.                              and placed on the chip, where it hybridises with
  d.–f. The cycle is repeated with another mask; this gives rise   the oligonucleotides. Measurement of fluorescence then
        to regions with different oligonucleotides of known se-    shows how much of the label is bound at what sites on the
        quence.                                                    chip – and thus what quantity of the mRNA of interest was
  Such a GeneChip can be used to examine different types of        present in the cell extract.

The example of GeneChip              One of the most important DNA chip technolo-
                                     gies 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).

                                                                                              DNA chips: choosy fish hooks        99
      total RNA                                     cDNA                               cRNA

                                   reverse                        in vitro               B
                    AAAA        transcription                  transcription

                     AAAA                                                                        B

                    AAAA                                                                 B


               array                                                                 B
                                                               biotin-labelled      B
       hybridisation                                                cRNA

                           B                           B         B

         B                     wash and         B                    scan and
                                 stain                               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.

   Competing techniques:
   design and function of DNA chips

   The tasks performed using DNA chips are many and var-               for the study of other types of DNA, such as oligonucle-
   ied, and the design of such chips is correspondingly                otides, RNA, cDNA, genes, chromosomes or whole
   diverse. Affymetrix’s GeneChip is a commonly used DNA               genomes.
   chip, however various other other manufacturers are             z   Reaction Hybridisation is not the only reaction that can
   offering a variety of techniques aimed at winning over              occur on a DNA chip. DNA molecules can also be
   customers. Points of difference include not just chip design,       bound by ligases or via chemical or photochemical reac-
   but also the way in which the experiments are performed             tions. Another possibility is to combine PCR (polyme-
   and the way in which the results are analysed.                      rase chain reaction, see chapter on PCR) with a chip.
   z Probe material The most commonly used DNA chips               z   Detection Different reactions on the chip require
       use short oligonucleotide chains as probes, however             different methods of detection, and hybridisation can be
       RNA, cDNA, genes and even whole chromosomes can                 detected in various ways. In addition to fluorescence,
       be attached to chips.                                           mass spectrometry (MS, see chapter on proteomics),
   z Manufacture DNA can be attached to chips in various               in particular, and also conductivity and electronic
       ways. These include photolithography, a technique bor-          methods, can be used for detection.
       rowed from the computer chip industry. Other tech-          z   Analysis DNA chip experiments generate enormous
       niques include application by pipette, dropping and             quantities of data that would be impossible to evaluate
       electronic methods, e.g. in a manner similar to the             without computer assistance. Of importance in this re-
       operation of an inkjet printer.                                 gard are not just suitably sophisticated programs, but
   z Target molecules The probe and target molecules are               also automatic control of experiments, image analysis,
       dependent on each other. Depending on what type of              databases, Internet links and platforms and visualisation
       target molecule is present on it, a chip may be suitable        of results.

Variety on a chip                   Though GeneChips and other oligonucleotide
                                    chips are the most commonly used type of bio-
                     chip at present, a variety of other molecules are used on bio-
                     chips. 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 so-
                     lution to be examined, chips can be either individually chosen
                     or specially made. One-off products are considerably more ex-
                     pensive than more or less standard products.
                     In addition, many attempts are being made at present to pro-
                     duce 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-con-
                     suming. Given, however, that proteins occupy a central place in

                                                                                            DNA chips: choosy fish hooks       101
                drug research, protein chips are regarded as an important tool,
                especially in proteomics (see chapter on proteomics).

Growing number                 DNA chips have also found broad application in
of applications                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).

Now routine:                    The first field of application of DNA chips was in
basic research in biology       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 ques-
                tions, basic research also leads to the development of new tech-
                niques 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 be-
                yond 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 expres-
                sion 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 ac-
                tivity of thousands of genes is determined simultaneously. This
                method, which is known as MEP (microarray-based expression

              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 administra-
              tion 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, pro-
              gression and treatment of diseases. Medicine thus becomes a
              new field of application of DNA chips.

A time of upheaval:           DNA chips long ago became a standard tool for
DNA chips in medicine         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 con-
                 tributory 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 dis-
                 ease-relevant genes are known, DNA chips can make it pos-
                 sible to test patients for genetic susceptibility to the disease
                 concerned. In complex diseases such as cancer and Alzhei-
                 mer’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 addi-
                 tional risk factors such as smoking, an unbalanced diet and
                 lack of exercise.
              z Diagnosis. The causes of diseases can also be determined reli-
                 ably 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

                                                                    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. Exam-
                ples of both these diagnostic applications of DNA chips are gi-
                ven below.
              z Therapy. Our genetic predisposition has considerable influ-
                ence 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 indivi-
                dual 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.

Checking up on green         DNA chips also have potential for use in con-
gene technology              sumer 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 reg-
              ulations including compulsory labelling. Given, however, that
              the vast majority of foods produced in this way do not differ vis-
              ibly from conventional products, it is often only by means of an
              examination of the genome of the plant concerned that adher-
              ence to such regulations can be effectively checked. DNA chips
              are well suited for use in such tests.

Chip instead of a             Ecology is a broad field of application for DNA
magnifying glass:             chips. For example, it is often necessary to distin-
ecology and taxonomy          guish 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 painstak-
              ing and detailed work with a magnifying glass or even a micro-
              scope, since many species are scarcely distinguishable from their
              close relatives on the basis of their appearance to the naked eye.

                DNA chips can make such distinctions more rapidly, more sim-
                ply 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, anthro-
                pology and research into evolution.

Sure identification:           DNA chips could also have a bright future in fo-
forensic medicine              rensic medicine. Prominent in this field of appli-
                               cation is the ability of DNA chips to detect dif-
                ferences between individual genomes and thereby to identify
                people. This can be required for identification of victims and al-
                ready plays a crucial role in the search for and conviction of cri-
                minals. Many countries, e.g. the Netherlands, also allow their
                police to use the genetic profile of people they are seeking in or-
                der to draw conclusions as to the external appearance of the per-
                son 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 al-
                most exclusively using PCR methods, DNA chips can play a use-
                ful 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.

Focus on medical                 Examples now exist of all these fields of applica-
applications                     tion 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 reso-
                lution at a low price form the basis for the kind of rapid and sim-
                ple 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.

                                                                      DNA chips: choosy fish hooks   105
              Another medical application of DNA chips is their use in infec-
              tious diseases. In this application attention is focused not just on
              the genes of the patient, but more specifically on those of the pa-
              thogens. Many viruses (e.g. HIV), in particular, along with var-
              ious 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 patho-
              gens rapidly and reliably so that treatment that is optimal for the
              individual patient can be chosen.

Leading the way               One of the earliest examples of the use of DNA
in medicine:                  chips in medicine is in the treatment of AIDS.
DNA chips in AIDS             Human immunodeficiency virus (HIV), the pa-
                              thogen of this disease, has an extraordinary abil-
              ity 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 re-
              sistance, 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 accord-
              ingly 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, there-
              by 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 se-
              quencing 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 im-
              portant 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-

               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 de-
               tect 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. Never-
               theless, three of the twenty or more variants of HPV can cause
               cervical cancer in women. Precise identification of the particu-
               lar variant present in affected women is therefore of great im-
               portance. 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 pa-
               pillomavirus 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.

Important application:         The ability of DNA chips to detect all the variants
cancer diagnosis               of a number of genes simultaneously could make
                               them an important instrument for the investiga-
               tion, diagnosis and treatment of cancer. The first products de-
               signed 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. second-
               ary growths in other parts of the body. For the process to ad-
               vance 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,

                                                                       DNA chips: choosy fish hooks   107
      Policing life and death: p53

                              • genetic damage
                              • other signals

                                                                            +                 mdm2

                                      activation and
           DNA repair                                               p53
                                         indirect                                                          functions
                                                                                  bax etc.                    apoptosis


                                      cell cycle held
                                       in G1 phase

  The tumour suppressor gene p53 is one of the most impor-            3. DNA repair. There is considerable evidence that P53 not
  tant genes involved in the development of cancer. Its gene              only allows the cell time to repair its DNA, but also plays
  product (P53, written with a capital ‘P’) plays a central role          an active role in this process. Here again, it acts indi-
  in the growth and division of somatic cells. It is active espe-         rectly via p21 and other genes, however certain proper-
  cially when genetic damage is present in the cell, but it can           ties of P53 suggest that it also plays a direct role in the
  also be activated by external signals. It is known to have at           repair process.
  least three functions:                                              Because of its central role in the control of cell growth and
  1. Control of the cell cycle. Cells divide in a regular pattern     division, p53 is an important target for cancer therapies.
      of events known as the cell cycle. If the genetic material      Attempts have been made, for example, to restore the func-
      of a cell is damaged, P53 holds the cycle in the G1 phase       tion of altered P53 and to stimulate synthesis of this protein.
      – a sort of resting phase – in order to permit repair of        The latter objective can be achieved, for example, via the
      the DNA. The signal for this is transmitted via a number        mdm2 gene, which together with p53 participates in an
      of proteins including P21.                                      ‘autoregulatory loop’: P53 stimulates formation of Mdm2,
  2. Apoptosis. If the damage to a cell’s DNA is too great,           which in turn inhibits P53. In healthy cells this cycle stops
      P53 induces the cell to ‘commit suicide’. This process,         excessive amounts of P53 from preventing normal division
      known as ‘apoptosis’, stops occurring in cancer cells,          of cells. If the function of p53 is impaired, but not abolished,
      with the result that they replicate in an uncontrolled          by genetic changes, drugs that act against Mdm2 may be
      fashion. P53 appears to be able to initiate apoptosis in        useful, since in such cases more P53 is required in order to
      various ways, including induction of the bax gene.              keep cells under control.

                                                    The decision as to whether such a drug will be useful there-
                                                    fore depends on the genetic profile of the patient con-
                                                    cerned. Techniques that can identify each of the variants of
                                                    p53 and of its genetic environment are therefore an impor-
                                                    tant 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

                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 forces of law              The most important tumour suppressor is the
and order in cells             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 ap-
                parent 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 sup-
                pressor. Depending on the reason why P53 is no longer func-
                tioning 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
                Chips designed to detect many other oncogenes and tumour
                suppressor genes are being developed at present, and some are al-
                ready 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

                                                                            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 ac-
               curacy of diagnosis.

Use in preventive              In fact, DNA chips can help not just when an ulcer
medicine                       has already been found. Cancer arises mostly as a
                               result of an accumulation over decades of mu-
               tations 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 un-
               dergo 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

First pharmacogenomic           From the discipline of pharmacogenomics comes
product:                        another current example of successful use of a
the AmpliChip CYP450            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 ob-
               servation 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 ex-
               cretion from the body in this way (see also chapter on pharma-

               cogenomics). As the func-
               tion of cytochrome P450 en-        The AmpliChip CYP450
               zymes varies from person to                                         The AmpliChip CYP 450
               person, drugs are broken                                            arrived on the market
                                                                                   in 2003. The scientific
               down more rapidly in some
                                                                                   basis of this chip is
               individuals than in others                                          formed by pharmaco-
               and their action in the body                                        genomic data on the
               varies accordingly.                                                 influence of the cyto-
                                                                                   chrome P450 gene fam-
               At least 50 separate genes                                          ily on the efficacy and
               and hundreds of gene vari-                                          tolerability of drugs.
               ants are now known to code         The AmpliChip CYP450 is able to identify the most impor-
                                                  tant variants of two important members of this group of
               for this family of enzymes,
               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 metabo-
               lism of drugs.

Outlook: use of DNA chips      As in the case of their use in relation to cyto-
for diagnosis                  chrome 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 break-
                  down – 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

                                                                            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 effec-
      tive 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 develop-
      ment 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 ques-
      tion 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

          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-re-
      lated genes and susceptibility to pathogens. The most important genes
      currently known to play a role in these areas are indicated.


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:
Affymetrix – Website:
InformationsSekretariat Biotechnologie – Website: Genexpressionsanalyse

                                                                            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. Genet-
                ics, genomics, proteomics and related disciplines are generating
                impulses that have already had a significant impact on our dai-
                ly 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 heat-
                ed debate. At the centre of this debate are the changes present-
                ly taking place in clinical medicine – the field in which the prac-
                tical applications of current research are most clearly palpable
                at present. The potential medical applications of these new dis-
                ciplines 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 ap-
                plies to many areas of public, and also to some important areas
                of private, life. Political, economic and legal questions are im-
                portant in this regard, as are ethical, societal and cultural con-
                siderations. 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 assess-
                ment of the value of these new technologies. The differing in-
                terests 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 opportu-
                nities provided by a molecular understanding of diseases be
                turned to practical benefit.

Help and protection:           The first, and most important, of the various in-
the interests of patients      terest groups involved is that of patients. Without
                               their acceptance, no change is possible. Their ex-
                pectations 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 mo-
                lecular research. Newly acquired understanding of the molecu-
                lar 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.

                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 any-
                body be restricted in terms of their choice of occupation, find
                their individual freedom limited or be socially stigmatised on
                the basis of test results.

Exploiting opportunities,      The aim must therefore be to exploit the oppor-
limiting risks                 tunities, while identifying and limiting the risks,
                               that arise from newly acquired medical knowl-
                edge. 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 med-
                icines (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 per-
                son gets regular prescriptions for insulin identifies that person
                unmistakeably as a diabetic. Many other forms of treatment
                likewise reveal people’s diseases.

                                                             Basic conditions: ethics, law and society   117
      The mere fact that information on the diseases present in an in-
      dividual 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 informa-
      tion that is passed on, and in particular the possible predictive
      value of the results of genetic tests. In practice, however, any pre-
      dictions of this kind will in the vast majority of cases be limited
      to a statement as to whether a particular patient is likely to re-
      spond to a particular drug or group of drugs. More detailed or
      specific information is generally not to be expected from such
      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 on-
      set 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 impli-
      cations 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 hemo-
      In these cases, individuals and their families may perceive the re-
      sultant information as stigmatising.

Varied requirements                   Therefore, if the individual patient is to be con-
                                      vinced of the value of more personalised medicine
                         with all its implications, a series of conditions must be satis-
                       z New, better targeted drugs should represent a definite thera-
                         peutic advance, especially in the case of diseases whose treat-
                         ment has hitherto been unsatisfactory or nonexistent. Per-
                         sonalised medicines should be more effective and better
                         tolerated than presently available drugs. These two require-
                         ments 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.

   Freedom of choice: focus on dignity

   The debate about the ethical aspects of the future of medi-              and even in them it probably arises only exceptionally. In
   cine centres around two questions: preservation of the dig-              all other medical situations genetic testing adds little to
   nity of patients and protection from discrimination. The follow-         what is already known from a family’s medical history.
   ing points, among others, need to be considered:                      z Patients who are not competent to make deci-
   z Freedom of choice. Patients must retain sovereignty                    sions. Cases in which a patient is unable to assess the
        over their physical and mental wellbeing. This includes the         arguments for and against undergoing a particular med-
        right to decide whether to receive a certain form of treat-         ical treatment are already covered by clear rules. Never-
        ment or to undergo a certain diagnostic investigation, i.e.         theless, the possibility of prenatal diagnosis and therapy
        the right to know or not to know. An available medical op-          is new in this respect. A legal framework for the perform-
        tion may neither be forced upon, nor withheld from, a pa-           ance of prenatal procedures needs to be established on
        tient. However, such choices can only be free if the patient        the basis of a social consensus.
        has first received extensive information and non-directive       z Discrimination. When told they have a certain disease,
        counselling.                                                        the vast majority of people still feel a sense of stigmatisa-
   z Limits to freedom of choice. The sovereignty of an in-                 tion and suffer anxiety about a possible loss of personal
        dividual ends where that of another begins. Since genetic           freedom. A major rethink is required here. It is clear from
        factors are inherited, the results of genetic tests are always      the complexity of our genome that terms such as ‘health’,
        relevant to the close relatives of the person tested. The           ‘illness’, ‘normality’ and ‘abnormality’, have no absolute
        question of whether, and if so in what cases, relatives             meaning at the molecular level. Every human being carries
        should have a say in a patient’s decision to undergo a              both promoting and protecting factors for all diseases.
        certain genetic test therefore needs to be clarified. In prac-      Therefore, in almost every case, all that a molecular diag-
        tice, however, this problem is probably limited almost ex-          nostic test can do is indicate disease probabilities.
        clusively to a small number of classical hereditary diseases,    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 pub-
                  lic regarding the potential, and also the limits, of genetics in

Society as a whole:              The requirements of individual patients for cus-
the public interest              tomised medicine differ from the interests of so-
                                 ciety as a whole. In many cases the interests of
               these two sides appear, at least at first glance, to be in direct con-
               flict, e.g. in relation to the sharing of burdens and risks between
               individual patients and society. Like the question of confiden-
               tiality of patient data, however, this is not a fundamentally new
               issue, and in most industrialised countries the problem is regu-
               lated 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 thera-
               peutic 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. Cur-
               rent data protection legislation is mostly aimed at minimising
               the amount of data that are collected, stored, distributed and

               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 broad-
               er circle of people via the medicines that are prescribed for those
               The task of directing the use of sensitive personal data into the
               appropriate channels falls to politicians in their role as repre-
               sentatives of society as a whole and the many special interest
               groups within it. They have to mediate between conflicting re-
               quirements and interests and work to build the broad social con-
               sensus that will be required if the new possibilities offered by
               molecular medicine are to be exploited in a responsible fashion.

Between economics and          In the crossfire of conflicting interests in health-
responsibility: commerce       care, 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 require-
               ments of the various interest groups. After all, the pharmaceuti-
               cal 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 in-
               novative, 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-

                                                             Basic conditions: ethics, law and society   121
               ber of very complex factors. In the first place, the process of de-
               veloping, 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 con-
               straints 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 cooper-
               ation between industry and healthcare funding authorities.

Research:                       A key role in the restructuring of medicine that is
laying foundations and          taking place at present is played by research. It
informing the public            lays the foundations for more individualised pre-
                                vention and treatment of diseases, and is there-
               fore subject to specific demands. Of central importance in this
               regard is the safety of treatments and the correctness of the di-
               agnoses 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 pharma-
               cogenetics and molecular diagnosis. More of a danger are the
               unrealistic expectations that can arise as a result of overenthusi-
               asm for the new possibilities.It must be remembered that molec-
               ular medicine is no magic wand. Side effects of drugs will still
               exist in the future, as will fluctuations in effect. And the disap-
               pointment felt by patients whose pharmacogenetic profile indi-
               cates that the only available form of treatment for their disease
               is likely to be ineffective in them has plenty of potential to cre-
               ate 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

                 The same applies to diagnostics. The common perception that
                 the results of genetic tests, in particular, are immutable and in-
                 variably correct in terms of what they predict creates a whole se-
                 ries 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 obvi-
                 ous to patients. Here again, frank and objective education of the
                 public is required.

Probability, not certainty        Genetic testing and molecular diagnostics there-
                                  fore can and must be accompanied by better edu-
                 cation of the public. Education alone, however, will not con-
                 vince 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 pharma-
                 cogenetic 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 pro-
                 vide an objective measure of the informational content of test
                 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-

                                                                 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 unneces-
         sarily 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 cho-
      lesterol 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 re-
      sults 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 deci-
      sion 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 pos-
      sible to ensure that advances in medicine bring benefit and do
      not cause harm.


Schreiber H-P: Humangenomforschung, Gentechnik und Gesellschaft
Schreiber H-P: Human-Genom-Forschung und die Notwendigkeit eines sozialverträg-
    lichen 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,
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
more knowledge for medical science

Medicine is changing. New tech-
niques and knowledge derived from
genetics, genomics and proteomics
permit a deeper understanding of the
molecular causes of disease. Increa-
singly, medicine will focus on diffe-
rences 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
               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 tech-
               niques of miniaturisation, automation and parallelism into re-
               search 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 ap-
               parently identical clinical pictures can have entirely different
               underlying causes requiring personalised treatment.
               These developments also have a commercial side, pitting estab-
               lished pharmaceutical companies against young biotech firms in
               a race to discover suitable target molecules and new drug com-
               pounds. 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 econo-
               mic 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.

No two illnesses               Behind this upheaval is the recognition that no
are the same                   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 sim-
               ple or even a single cause. And even in the exceptions just men-
               tioned, which include, for example, cystic fibrosis and hemo-
               philia 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 investi-
               gation, 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 ge-
               netic makeup of the pathogens involved.
               These findings are neither new nor surprising. And yet they con-
               front 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

                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

Consequence:                    In fact, for the first time in the history of medical
a new role for diagnostics      science diagnostics and therapy are meeting on
                                common ground – at the molecular level. Where-
                as 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 caus-
                es – 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 rou-
                tinely 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 wide-
                spread 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
                Molecular diagnostics is therefore playing a key role in the rapid
                advances now taking place in medicine. It has a broad range of
                z New therapies: The use of new drugs that take account of the
                    complex causes of diseases calls for differentiated molecular-
                    based diagnostics.
                z Screening: Screening tests are already an important instru-
                    ment in healthcare. Cancers, abnormal blood sugar, abnor-
                    mal bone density and abnormal blood pressure are all cur-
                    rently screened for. As our knowledge deepens, more such

                                                          Prospects: more knowledge for medical science   129
                                                  DNA strand

                                                         double helix


              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 diag-
                nostics 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 for-
                ward their development and encourage widespread use of
              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 identi-
                fication of pathogens in hospitals. Again, modern techniques
                such as PCR and DNA chips are the fastest methods for de-
                tecting the presence of pathogens before they can spread.

No two treatments             The recognition that diseases can have entirely
are the same                  different causes despite producing the same
                              symptoms is not new. What is new is the mole-
              cular biological understanding that now makes it possible to ex-
              amine the genetic differences between individual patients and
              the effects of these differences on treatment. In other words, no

two treatments are the same. A drug might be right for one pa-
tient but wrong for another, even though both patients have the
same illness, because drugs can vary in their efficacy and toler-
ability 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 up-
set 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 pa-
tient’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 re-
    lationship 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 impor-
    tant 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 phy-
    sical 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 deter-
    mined; also, they vary over time – though this means that they
    can be influenced. This is not true of gene variants. It is there-
    fore all the more imperative to recognise how environmental
    factors influence the way the body interacts with drugs.

                                          Prospects: more knowledge for medical science   131
Consequence:                             If future therapies are to be based on genetic fac-
the personalisation of                   tors, medicine will inevitably become more per-
medicine                                 sonalised. 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
                        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 cur-
                        rently 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

   Great fluctuations: efficacy of drugs

                                                           The efficacy of drugs is often unsatisfactory and may fluc-
      drug group               poor efficacy               tuate in the extreme. The table gives the incidence of poor
      AT2 antagonists          10 – 25%                    efficacy for several major drug groups.
                                                           Angiotensin II (AT2) antagonists, like ACE (angiotensin-
      SSRIs                    10 – 25%
                                                           converting enzyme) inhibitors, are antihypertensive drugs
      ACE inhibitors           10 – 30%                    which are widely used in the treatment of heart failure. The
      Beta-blockers            15 – 25%                    same applies to beta-blockers. Selective serotonin reup-
                                                           take inhibitors (SSRIs) are used as psychotropic drugs,
      Tricyclic antidepressants 20 – 50%                   especially in obsessive-compulsive disorders and, like tri-
      Statins                  30 – 70%                    cyclic antidepressants, for depression. Statins, also known
                                                           as HMG-CoA reductase inhibitors, lower cholesterol levels.
      Beta 2 antagonists       40 – 70%
                                                           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 ex-
               panded range of therapies and – as is already evident from the
               growth of the Internet – far better informed and more self-con-
               fident patients.

Consequence:                   This means that the demands made on medical
integrated healthcare          science will increase. One innovation gives rise to
                               another. Personalised therapies require indivi-
               dual diagnoses. Molecular diagnoses call for differentiated
               therapy. And both aspects, diagnosis and therapy, depend on
               rapidly expanding technological possibilities. In fact, a synthe-
               sis is taking place at the moment: research and development,
               diagnosis and therapy, information and prevention are evolv-
               ing together. The key to successful healthcare lies in integrated
               If the new possibilities of medical science really are to bring
               about progress, they must mesh smoothly. The concept of diag-
               nosis 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 lim-
               ited to fresh air and a healthy diet. Testing, i.e. diagnosis, of ge-
               netic 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 rela-
               tion to high serum cholesterol levels.
               Treatment follows seamlessly. The earlier a disorder is discov-
               ered, 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 diag-
               nosis is matched by a corresponding range of personalised
               therapies. Progress will be achieved only if both sides move for-
               ward 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.

                                                         Prospects: more knowledge for medical science   133
Consequence:                    For the pharmaceutical industry, these develop-
upheaval in the pharma-         ments impose the need for a continuous rethink.
ceutical industry               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 dif-
                  ferences between individuals are distinguished and consid-
                  ered, the more difficult it is to separate these two poles. Close
                  cooperation is required here: drugs whose prescription de-
                  pends upon pharmacogenomic considerations will be prescrib-
                  able 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 sen-
                  sibly 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 ad-
                  vantage. Expertise therefore needs to be gathered together
                  either within a single company or else by means of close allian-
                  ces 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 un-
                  satisfactory means one thing above all: Pharmaceutical com-
                  panies will have to be more willing to take the risks associat-
                  ed with the development of new drugs with new mechanisms
                  of action. Certainly, the future will continue to hold the oc-
                  casional 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 explo-
                  ration of new avenues – particularly via new target mole-
                  cules, 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 fail-
                  ure. Nevertheless, the effort may well be worthwhile. Suc-
                  cessful 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

                  in this way are more effective thanks to their targeted activity.
                  This should reduce the risk of failure in later stages of devel-
                  opment 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 com-
                  panies backed by considerable venture capital are invading
                  the market for new drugs developed on the basis of molecu-
                  lar findings. Typically these companies develop drugs char-
                  acterised by a high risk but huge sales potential. The pioneers
                  of the sector, e.g. the American companies Amgen and Gen-
                  entech, 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 chal-
                  lenges – either by establishing their own biotechnology de-
                  partments or entering into alliances or acquiring promising
                  innovative companies in the sector.
               5. Increased demands: New opportunities bring new responsi-
                  bilities. 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 ad-
                  justment, patients are likely to become more demanding in
                  terms of the efficacy and safety of the drugs they take.

High hurdles, high goal        Internationally active healthcare companies will
                               not escape this trend. On the contrary, active par-
               ticipation in this process of change is fundamental to their sur-
               vival, whereby the term ‘change’ does not imply a revolution, but
               rather a systematic evolution towards more informative investi-
               gations and more effective and safer drugs. The fact that many
               years of laborious and detailed research work are required be-
               fore 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 economi-
               cally sound. For instance, the size of a patient group above which
               the development of drugs specifically for it becomes economi-
               cally viable still cannot be predicted – at least, not until new
               forms of cooperation between society and industry, such as ‘or-
               phan disease’ programmes for particularly rare diseases, have
               been set up.

                                                        Prospects: more knowledge for medical science   135
      Nevertheless, progress is opening up far-reaching new opportu-
      nities for medicine at the scientific and technical levels. Person-
      alised 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. Notwith-
      standing all the commercial and ethical imponderables, in a cer-
      tain sense the new possibilities also impose a moral obligation
      to apply the new findings of molecular medicine for the practi-
      cal benefit of patients.

A brief glossary of terms

Amino acids the chemical building blocks of proteins. They                                         fi
                                                                      DNA chip a biochip with single-stranded DNA as the probe.
   consist of a constant region, containing an amino group and an                                                    fi
                                                                      Enzyme a biological catalyst, generally a protein, that can
   acid group, and a variable region. At least 20 different amino        accelerate and combine certain chemical reactions.
   acids occur in nature.                                                                      fi
                                                                      Exon a sequence of a gene that acts as a direct blueprint for
Antibodies Y-shaped proteins that are part of our immune sys-            a gene product.
   tem. They bind to foreign substances in the body and thereby                                         fi
                                                                      Gene array a special type of DNA chip.
   mark them for destruction.                                                                       fi
                                                                      GeneChip a widely used DNA chip developed by the US
Antioxidants molecules that trap dangerous, highly reactive              company Affymetrix.
   oxygen compounds in the body and thereby render them harm-         Genes functional segments of our genetic material that serve
   less. Vitamins C and E, for example, are antioxidants.                                                           fi
                                                                         mostly as blueprints for the synthesis of proteins.
Apoptosis ‘programmed cell death’. Cells whose genetic mate-          Genetics the study of inheritance; deals with the laws of inher-
   rial is irrevocably damaged or altered ‘commit suicide’ in order                                     fi
                                                                         itance and the properties of genes, including the transmis-
   to protect the rest of the body from the effects of the genetic       sion of specific variants of a gene from one generation to the
   alteration.                                                           next.
Autosomal dominant inheritance an inherited characteris-                                                                 fi
                                                                      Genome the totality of the genetic material ( genes) of an
   tic that is expressed if it is present on either one of the two       organism.
   fi  autosomes of a particular kind.                                                                       fi
                                                                      Genome chip a special kind of DNA chip.
Autosomal recessive inheritance an inherited characteris-             Genomics the systematic study of the form, function and inter-
   tic that is expressed only if it is present on both autosomes                        fi
                                                                         actions of the genes that comprise the human genome.
   of a particular kind.                                                                                                fi
                                                                      Genotype the alternative forms (alleles) of a gene present in
Autosomes chromosomes not involved in sex determination.                 an individual; generally there is a maximum of two – one from
   Humans have two of each kind, inherited from the mother and           the father and one from the mother.
   the father respectively. Altogether there are 44 autosomes         High-throughput screening a highly automated method of
   (twice 22).                                                           identifying potential drugs in chemical libraries.
Bases chemical substances that have a basic (alkaline) action.        Hybridisation the joining of two complementary DNA (or fi
   The bases of DNA are the fundamental building blocks of               fi  RNA) strands to form a double strand.
   the genome: adenine (A), thymine (T), guanine (G) and cyto-                                                               fi
                                                                      Intron a sequence of DNA situated between the exons of a
   sine (C). When present on two strands of DNA, the bases join          fi                                               fi
                                                                            gene that is cut out of the corresponding mRNA before
   to form stable pairs. In nature, base pairs form only between A       this is translated into the gene product.
   and T and between G and C. In RNA, thymine is replaced by          Metabolism the transformation of chemical substances in the
   uracil, which likewise pairs with adenine.                            body or within a cell.
Biochip a solid substrate (e.g. glass or plastic) upon which bio-     Microarray a widely-used synonym for DNA
   molecules are anchored.                                            mRNA messenger RNA, the working copy of a gene that    fi
Bioinformatics the in most cases computer-assisted analysis                                                          fi
                                                                         acts as a blueprint for the synthesis of proteins. Unlike
   of biological data by special databases, applications and pro-        fi  DNA, it is able to leave the cell nucleus.
   grams.                                                                                                           fi
                                                                      Nucleic acids generic chemical term for DNA and RNA;       fi
cDNA complementary DNA; DNA transcribed enzymatically                    chain-shaped molecules whose individual building blocks are
        fi                fi
   from RNA (mostly mRNA).                                               fi         fi
                                                                            bases ( nucleotides).
Cell the smallest independently viable unit of an organism.                                                     fi
                                                                      Nucleotides the building blocks of DNA and RNA; they   fi
Chromosomes tightly packed DNA strands with associated                                      fi
                                                                         comprise the four bases adenine, thymine, cytosine and gua-
   proteins that are present in the cell nucleus and that function                           fi
                                                                         nine (A, T, C, G; in RNA thymine is replaced by uracil [U]),
   as bearers of genetic information. The human genome consists          a sugar and at least one phosphate group; without the phos-
   of 23 chromosome pairs (22 autosomes and one of the two               phate group these building blocks are referred to as nucleosides.
   sex chromosomes X and Y).                                          Oligonucleotides short nucleic acid chains composed of at
Complementary DNA The building blocks of DNA and       fi                                                        fi
                                                                         most a few dozen building blocks ( nucleotides).
   fi  RNA form specific pairs. Two strands whose building blocks                       fi
                                                                      Oncogene a gene that plays a role in the development of
   form a sequence of perfect pairs are able to form a stable dou-       cancer.
   ble strand and are referred to as complementary strands.           Pharmacodynamics the study of the interactions between
Denaturing a process induced by heat or chemicals in which               drugs and their molecular targets.
                        fi        fi             fi
   a biomolecule (e.g. DNA, RNA or a protein) loses its               Pharmacogenetics describes the influence of gene variations
   natural form.                                                         in individuals on the efficacy and side effects of drugs.
DNA deoxyribonucleic acid, the chemical substance of which            Pharmacogenomics studies interactions between drugs and
   our genetic material consists.                                        the genome.

                                                                                                     A brief glossary of terms       137
Pharmacokinetics the study of the uptake, conversion and
   breakdown of drugs in the body over time. Environmental fac-
   tors, 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
Polymerases       fi   enzymes that link individual nucleotides
                           fi          fi
   together to form long DNA or RNA chains.
Polymorphism existence in more than one form; in genet-    fi
   ics, 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 polyme-
Promoter a region of DNA immediately before a gene that fi
   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 dif-
   fer 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
RNA ribonucleic acid; the chemical substance of which, among
                                     fi         fi
   other things, working copies of genes ( mRNA) consist.
Sequence the order of the nucleotides in            fi   DNA (DNA
   sequence) or RNA (RNA sequence).
Sex-linked inheritance an inherited characteristic transmit-
   ted via one of the two sex chromosomes (X or Y).
SNPs single nucleotide polymorphisms – differences in individ-
   ual building blocks (base pairs) of DNA that are distributed
   randomly over the genome and passed from generation to
Targets the molecules, mostly proteins, upon which drugs
   act in our body.
Template in molecular biology, mostly a fragment of DNA      fi
   that acts as a chemical template for polymerases.
Transgenic animals animals containing genes derived from
   other species.
Tumour suppressor a molecule which, when functioning cor-
   rectly, prevents cancer from developing.

      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


Shared By:
handongqp handongqp