Genes and Health

We Innovate Healthcare
Genes and Health
Genes and Health
Content
I
1
II
3
10
11
12
13
14
III
15
16
17
18
19
20
21
22
23
24
25
26
27
Molecular medicine:
genetics, genomics and
proteomics for
diagnosis and therapy
To d a y
To d a y
A
A
B
B
D
D
C
C
To m o r r o w
To m o r r o w
Diagnostics
Diagnostics
A
A
D
D
B
B
C
C
G
G
G
G
G
G
C
C
C
C
C
C
A
A
A
A
A
A
A
A
A
A
A
A
T
T
T
T
T
T
C
C
A
C
C
A
T
T
T
T
T
T
A
A
A
A
A
A
Pharmacogenomics:
genes and drug response
19
Proteomics:
seeing through the
undergrowth
31
Targets
for medicine
47
PCR:
an outstanding method
65
SNPs:
the great importance
of small differences
83
DNA chips:
choosy fish hooks
95
Basic conditions:
ethics, law and society
115
Prospects:
more knowledge for
medical science
127
A brief glossary
of terms
137
Genes and Health
If it were not for the great variability among individuals,

Medicine might be a Science, not an Art. This statement by Sir
William Osler (The Principles and Practice of Medicine, 1892) is
as topical now as it was over a hundred years ago. For it remains
true that drugs sometimes work as intended but sometimes do
not, that one patient will tolerate a drug but another will not and
that drugs sometimes have serious side effects. These differences are due at least in part to our genes, the genetic material that
makes each of us unique and that consequently makes each of us
react differently to drugs.
Genetics, genomics, proteomics and other branches of modern
biology can help us to understand the medical consequences
of these differences, and in fact have already led to the identification of many genetic factors that influence the action of drugs
whether this be by affecting the way in which the body deals
with a drug or by influencing the course of the disease concerned. And a new scientific discipline pharmacogenomics
that deals specifically with the relationships between our
genome and the effects of drugs has now appeared.
At the same time, increasing attention is now being paid to the
principal targets of drugs, namely proteins. Here again, a new
branch of science has appeared, namely proteomics, the study of
the totality of, and the complex interrelationships between, the
proteins of an organism. Thus, as well as learning more about
the genetic information that provides the blueprint for the production of proteins, we are building up an ever more detailed
picture of bodily function and malfunction at the molecular
level. Acquisition of an understanding of the interplay between
hereditary and nonhereditary factors in patients is an essential
step on the way to better targeted, more personalised therapy.
An important precondition for this has now been satisfied in
that for the first time in the history of medicine, diagnosis and
therapy are meeting on common ground. Thanks to the new
field of molecular diagnostics, both diagnosis and therapy are
now focused on the network of genes, proteins and other substances that exist in the human body. This is leading to the development of completely new ways of understanding, detecting,
preventing and specifically combating diseases. Applications of
molecular biology are in fact now leading to the development of
a new approach to diagnosis and therapy known as molecular
medicine.
Grouped around this term are a multiplicity of modern research

techniques and disciplines. These include, in equal measure,
pharmacogenomics, the search for new drug targets, proteome
research, the search for small but important genetic differences
known as SNPs, new techniques such as the PCR and DNA
chips, and bioethics.
At many events held over the past few years, Roche has attempted to cast light on current developments in medicine and to explain the scientific background and potential implications of
these developments. This publication is intended to supplement
that information and to introduce the reader to the most
important terms used in the new field of molecular medicine.
It can help to improve understanding of current developments
and can form a basis for the public debate that is being conducted at present about the uses of genetics and genomics in medicine.
Molecular medicine: genetics,

genomics and proteomics for
diagnosis and therapy
I
1
II
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
III
In its search for the causes of disease,

medicine has now advanced to the
molecular level. Genetics, genomics
and proteomics are opening the way
to a new and deeper understanding
of bodily processes and are providing
the tools for more precisely targeted
interventions when bodily function is
disturbed. For the first time in history,
diagnosis and therapy are meeting
on common ground.
27
It is a gradual revolution that has been going on for hundreds of

years and still has a long way to go. It is not possible to say exactly when it started. Perhaps it was in the 18th century, at a time
when cupping (mechanical leeching) was still considered to be
an effective remedy for headache, cancer and cholera. At a time
when the value of panaceas was taken as much for granted as the
belief in witchcraft, physicians and scientists began to turn medicine upside down. They gave a new meaning to the concept of
diagnosis and in so doing
Terms
took their craft closer to the
causes of disease. Even in
Molecular medicine the application of molecular biology
those days every medical
(in particular, genetics, genomics and proteomics) to medicine.
Genetics the study of inheritance; deals with the laws of intreatment was preceded by
heritance and the properties of genes.
an examination; for over
Genomics the study of the form, function and interactions of
2000 years an imbalance of
the genes of an organism.
Proteomics the study of the form, function and interactions
bodily humors had been the
of the proteins of an organism.
most common finding, and
bloodletting and cupping
were accordingly the most common forms of treatment. The diagnosis was generally made on the basis of a handful of symptoms and signs the art of diagnosis had no more than that to
offer. In the middle of the 18th century scientists such as the Italian anatomist Giovanni Battista Morgagni for the first time set
themselves the task of identifying the seat of a disease within the
body of their patients. They recognised that disturbances of
function can be correctly treated only if they are correctly understood and that an unmeasurable and undefinable imbalance of humors (liquids) was not adequate for that purpose.
Instead, they looked for tangible and testable causes of illness.
This relentless search for causes is the driving force of the gradual revolution in medicine that has been taking place since Morgagnis time: a shift away from symptom-based therapy towards
causally based therapy. To this end doctors and researchers are
delving ever deeper into the human body. Whereas Morgagni
continued to look into the solid components, and more specifically the organs, of the body, his colleague Marie-FranoisXavier 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 workCellularpathologiedrew attention
to the cells of which all organs of the body are composed. Later,
in the 20th century, increasing attention was paid to life processes within cells. All these efforts and discoveries are ultimately
directed towards a single goal, that of more precise medicine.

The objectives are to identify the causes of a disease, to consider the possibilities for intervention and then to provide effective
treatment. These objectives are more relevant now than ever
before, and on the way towards them we are at present taking
the next major step, that of replacing cellular pathology with
molecular medicine. Genetics, genomics and proteomics are
opening up totally new perspectives in diagnosis and therapy.
Accompanying the revolution: Morgagni and Virchow

Giovanni Battista Morgagni
(16821771) studied medicine
in Bologna and in 1715 was
appointed to the chair of anatomy at the University of Padua.
In 1761, while still at Padua, he
published his principal work De
sedibus et causis morborum per
anatomen indagatis (On the
seats and causes of disease investigated by anatomy). In this
work he departed from the
standard practice of the time by
concentrating not on the symptoms and signs of disease, but
on the location of disease within the organs of the body. In his
view, the pathological changes in organs that he demonstrated in many autopsies were the true causes of illness. This
view amounted to a rejection of the theory of humors
(humoralism) that since the time of Hippocrates had attributed disease to an imbalance between the four bodily liquids
blood, phlegm, yellow bile and black bile.
Rudolf Virchow (18211902)

studied medicine at the Berlin
Military Academy. After holding
a professorship in Wrzburg he
took up a chair of pathological
anatomy that had been created
especially for him at the University of Berlin. He was a political
activist, campaigning vigorously
for democracy and public provision of healthcare. The journal
Virchows Archiv, which he
established in 1847 together
with Benno E.H. Reinhardt, was
the organ of scientific pathology. In 1858, with his work Die Cellularpathologie in ihrer
Begrndung auf physiologische und pathologische Gewebelehre (Cellular pathology as based on physiological and
pathological histology), he founded a new theory of pathology in which the cells of the body were regarded as the sites
of origin of pathological changes.
Every disease is influenced by a larger or smaller

number of factors. These include on the one hand
environmental factors such as toxins, radiation,
infections, nutrition, age, stress and much more besides, and on
the other hand the genetic predisposition that causes our body
to react to the environment in a certain way. Small changes in
our genes can trigger, prevent, promote or alleviate diseases. The
same applies to external influences. Whether, when, and how severely a person falls ill is determined by the combination of all
these factors and proteins play a central role in mediating these
effects. They read and make working copies of the genes; they
carry out the instructions, while at the same time regulating the
Target: the molecular network of the cell
Molecular medicine: genetics, genomics and proteomics for diagnosis and therapy
action, of the genes; they receive signals from the environment,

pass these on and incorporate them into the molecular network
of the cell. It is precisely in this interplay of environment, genes
and proteins (as well as a variety of other equally important substances that differ from case to case) that drugs exert their effects. They act directly on the molecules that make up our body
The changing role of biology in medicine
chemical
synthesis
improvement
by computer
target molecule
potential
drugs
binding assay
biological
target search
rational
drug design
high-throughputscreening
The tasks of biology in medicine have changed greatly in the

past few decades. Over this period the initiative for innovations has passed increasingly from chemistry to biology.
Molecular medicine is an expression of this change.
Evaluation As recently as the 1970s the principal task of
biologists in medicine was to evaluate, i.e. to test the effectiveness of, new substances produced by chemists.
Targets As knowledge of the molecular basis of diseases
increased, biology was able to provide new targets for drug
development. These targets formed, and still form, the basis
for the search for new medicines by chemical synthesis.
Rational drug design Rational drug development arose as
a result of increasing knowledge of the structure, i.e. the form,
of proteins. The idea was that drugs would be designed by
10
best candidate
efficacy studies
biological
evaluation
computer so as to interact in a highly specific way with a certain part of a target molecule; only afterwards would they be
produced by chemists. This approach has not yet brought the
success that was hoped for.
High-throughput screening New, automated methods
then made it possible to test large numbers of substances in
hundreds or even thousands of miniaturised assays for certain biological, chemical and physical properties. Experiments
such as binding to a target, which previously had to be performed individually in molecular biology laboratories, could
now be performed in this way. The desired properties of suitable molecules can then be improved in a further step. Highthroughput screening has already resulted in the development of a number of particularly effective drugs.
and in this sense are themselves an important environmental

influence. The more we know about the actions of molecules in
our body, the more effectively we are able to intervene when
these actions become disordered.
z Every newly discovered molecule that plays a role in the development of a disease constitutes a potential target for drugs.
For example, in the past few decades biologists have discovered more and more oncogenes, i.e. cancer-promoting gene
variants. Many anticancer agents act by restoring the correct
function of the products of these genes (mostly proteins).
z Knowledge of the structure, i.e. the three-dimensional form,
of a target molecule makes it possible to decide in advance
whether a given substance has any potential for use as a drug.
Though it has yet to notch up many successes, computerbased rational drug design can greatly reduce the number of
substances selected for further development.
z If the genetic preconditions for a disease are known, a patients individual risk can be determined and appropriate
preventive action taken. Sickle-cell anemia is an example of
this. In this condition, an inherited modification of a certain
component of the gene for hemoglobin, the red blood pigment, results in production of an altered protein that changes
shape when the oxygen supply is inadequate. Under these
conditions the red blood cells assume the form of a sickle,
clump together and block the blood vessels. Carriers of this
trait therefore need to avoid great heights and changes in air
pressure (e.g. in aeroplanes), among other things.
z Many diseases are amenable to intervention at the gene level.
For example, genes can be turned on or off by drugs, and
one day it may even be possible to replace genes completely
by means of gene therapy. It is precisely in this latter field,
however, that further intensive research is required. In many
cases e.g. severe hereditary diseases due to mutation of a
single gene or a small number of genes gene therapy,
along with the stem cell therapy, offers the only hope of
genuine cure.
z Drugs do not always have the same effects. The effect of a
given drug can be too strong, too weak or absent altogether
in people with the same symptoms. Moreover, adverse effects
are always likely to occur. Our genes are at least partly responsible for these too; the discipline of pharmacogenetics
investigates these relationships and attempts to foresee, and
ultimately forestall, such problems.
11
Genetics, genomics and proteomics thus provide

medicine with a variety of new ways of intervening in the development and progression of diseases. Nevertheless, intervention has not become easier, for the
deeper medicine looks into life processes, the more complex are
the things it sees. The humoral pathology of Hippocrates distinguished between four humors; Morgagni extended the search
for the seat of diseases to a couple of dozen organs; Bichat concerned himself with a few hundred bodily tissues; Virchow
directed attention to the bodys cells, of which there are about
100 million million; and each of these cells contains an enormous number of nucleic acids, proteins, sugars, fats and other
organic and inorganic substances. And in addition to all this is
the far less measurable influence of external factors.
Nevertheless, the effort is worthwhile. For in the past, methods
of combating complex diseases were based largely on trial and
error, precisely because such disorders are not caused by a simple infection or gene mutation, but rather arise as a result of a
combination of external and internal, predisposing and protective, and variable or unchangeable influences. This is true of
most of the major diseases that afflict people in industrialised
countries, e.g. cancer, Alzheimers disease, diabetes and cardiovascular disease. Every ray of light that genetics, genomics and
proteomics cast on the factors that contribute to these diseases
helps in the fight against them.
A multiplicity of possible
causes
This is because disease-inducing environmental

influences can generally be modified where as
our genetic makeup generally cannot. Among the
risk factors that contribute to the development of disease, our
genetic predisposition is a constant. And this makes it all the
more important for us to learn more about, and where possible
to limit, its influence. In the 1980s scientists succeeded in identifying the genetic basis of a number of severe hereditary diseases brought about by a single defective gene. These include
Huntingtons disease (Huntingtons chorea), cystic fibrosis (mucoviscidosis) and hemophilia. More refined methods now allow
scientists to investigate the genetic causes of more complex diseases in which various genes can exert positive or negative
influences.
z Monogenic diseases such as Huntingtons disease, cystic fibrosis and hemophilia follow the classical (mendelian) laws of
The central importance

of genes
12
Monogenic hereditary diseases: cystic fibrosis
cell membrane
nucleotidebinding
region
regulatory
region
chloride
Defect on chromosome 7.
Impact of genetic defect on salt transport.
The first disease-causing genes to be discovered in the

1980s were associated with hereditary diseases due to specific mutations in individual genes. Such monogenic diseases
can be classified on the basis of their pattern of inheritance:
Autosomal recessive inheritance The altered gene must be
inherited from both the father and the mother for the disease
to occur. Carriers of only one altered gene are healthy, but
can transmit the disease to their children. An example of this
is cystic fibrosis (also known as mucoviscidosis), which is
due to a defect on chromosome 7. In this disease salt transport is disturbed in certain mucosal cells, causing the mucus
of the respiratory tract, digestive tract and other organs to be
extremely viscid. This results in frequent infections and inflammation. Whereas only half a century ago most cystic
fibrosis sufferers died during childhood, those born today can
expect to live 40 to 50 years thanks to specific (symptomatic)
treatment and diet.
Autosomal dominant inheritance In this case a single
altered gene is sufficient to cause the disease to appear, and
the disease is transmitted to fifty percent of the children of
sufferers. Huntingtons disease (Huntingtons chorea) is an
example of such a disease. In this condition a defect in the
Huntington gene on chromosome 4 causes production of an
incorrectly formed protein known as amyloid (a similar
change is seen in Alzheimers disease and Creutzfeldt-Jakob
disease). This leads firstly to motor disorders and later to

mental deterioration. The disease generally appears between
the age of 30 and 40 years and progresses in all cases to
death after 5 to 20 years. Physiotherapy and diet can slow
progression. Since the responsible gene was found in the
mid-1990s, scientists have been working to develop suitable
drugs, however as yet none has been approved for use.
Sex-linked inheritance In this case the responsible gene lies
on a sex chromosome (gonosome), generally the X chromosome. The presence of one unaffected copy is sufficient to
suppress the disease, therefore men are far more commonly
affected than women, since they lack a second X chromosome. Hemophilia is such a disease. This is due to an inherited deficiency of a blood coagulation factor, as a result of
which the affected persons blood coagulates very slowly or
not at all. Untreated hemophiliacs therefore die young of
internal and external bleeding. Previously, hemophiliacs were
treated with blood transfusions, and even today the missing
coagulation factor is obtained mostly from donor blood.
Unfortunately, diseases can be transmitted from the donor to
the recipient in this way, though less so than with transfusion
of whole blood. Now that the responsible gene has been
identified, however, the missing factor can be produced using
recombinant DNA technology, thus eliminating the risk of
transmitting other diseases.
13
inheritance. The pattern of occurrence and non-occurrence

of such diseases within affected families is determined by
whether only one or both copies of the gene in question need
to be altered for the disease to occur. In such cases the
responsible genes are relatively easy to identify via studies
comparing the genetic material of affected and unaffected
members of a family.
Genetic testing can be used to advise prospective parents of
the risk of having a baby that will be affected by a heritable
disease, such as cystic fibrosis.
z By contrast, the pattern of inheritance of polygenic diseases,
which include type 2 diabetes and most types of cancer, is not
so simple, since many genes are involved. Most such diseases
tend to cluster (occur with increased frequency) in certain
families, but not in such a way that the precise distribution of
affected and unaffected individuals can be predicted. This requires larger studies to identify the various genes that influence the disease to a greater or lesser extent. This task is rendered even more difficult by the fact that in this case genes
that predispose to the disease can overlap with genes that
protect against it. Hopes have therefore been placed in the
study of single nucleotide polymorphisms, or SNPs (pronounced snips), i.e. changes in single subunits of the
genome. The presence of such single nucleotide substitutions in important sections of a gene can have profound
effects on the function of the corresponding gene product.
The finding of an increased frequency of certain SNPs in association with a disease indicates that the genes concerned
play an important role in the disease in question.
z It is not always easy to separate environmental influences
from genetic influences, especially since the environment can
influence the behaviour of our genes. Twin studies and adoption studies are helpful here. Monozygotic (identical) twins
brought up in different families have an identical genome but
are subject to different environmental influences, while dizygotic (fraternal) twins brought up in the same family are
subject to essentially identical environmental influences and
have a similar, but not identical, genome. Finally, adopted
children share essentially the same environment with, but are
genetically quite different from, their stepbrothers and stepsisters.
z Our genetic makeup also exerts a decisive influence on our
predisposition to disease. Where genes that play a role in the
14
The complex interplay of genetic factors in disease

Now that the findings of molecular biology are being applied
to medicine it has become clear that very few diseases have
simple genetic causes. In the vast majority of cases many different genes exert a greater or a lesser influence both on the
disease itself and on each others action.
1. Complex diseases are multifactorial in origin. Both endogenous, in particular genetic, and exogenous factors
play important roles, and the relative influence of these two
types of factor can vary. Moreover, factors such as diet,
environment and behaviour can influence the bodys reactions independently of the actual causes of disease.
2. Genes can protect against or predispose to the development of complex diseases. The combined influences of
protective and predisposing factors result in an overall risk.
It is therefore entirely possible for a person to have a large
number of disease-promoting gene variants and to
transmit these to their offspring without themselves ever
becoming ill, provided only that they also have a sufficient
number of protective gene variants. In the age of molecular medicine, terms such as healthy, ill, normal and
abnormal are therefore no longer easy to define.
3. All risk factors including genetic factors for a complex
disease can be either categorical or quantitative. A gene
variant is said to act categorically if a certain disease can
occur only in its presence. By contrast, quantitatively acting gene variants act additively (or else their effects are
multiplied) up to a critical point at which disease occurs.
4. Complex diseases are polygenic, i.e. they result from the
action of a number of different genes. The contributions of
the individual genes are difficult to determine and can vary
enormously, especially as genes influence each others
action.
5. Another characteristic of complex diseases is genetic heterogeneity: since a number of genes can be jointly responsible for the occurrence of a disease, different combinations of genes can result in the same clinical picture.
This complex interplay of influences means that the genetic
environment
genes
nutrition
disease
lifestyle
causes of complex diseases can generally be determined only

via large, statistically complex studies and on the basis of a
deep understanding of the molecular processes that take
place in cells. Only now that the findings of genetics, genomics, proteomics and bioinformatics are being applied to
medicine has this become possible. In the case of complex
diseases, however, even tests based on techniques of molecular medicine can at best indicate only an approximate risk
that an individual will develop a particular disease. Absolute
assertions cannot be made. In the future, genetic testing may
increasingly be used to guide patients and healthcare providers in designing optimal treatment strategies based on
patients genetic variations.
For example, pharmacogenetic research is already underway
to provide physicians with a better understanding of the
influence of genetic variation on an individuals response to
medication.
development of a disease (or, for example, in intolerance of a

drug or failure of a drug to exert its expected effects) are
known, an individuals risk of developing that disease can to
some extent be determined by appropriate genetic tests.
Knowledge of his or her predisposition to a certain disease
allows the individual concerned to take appropriate precautions and to modify his or her lifestyle accordingly and if
15
necessary to take preventive medicines. Early prevention is

therefore one of the potential applications of molecular
medicine. Given, however, that most diseases result from the
combined action of a large number of genetic and environmental factors and that predisposing and protective genes
can overlap, such tests can only ever indicate a greater or lesser probability that an individual will develop a disease.
The importance of looking for the causes of disease has not changed at all since Morgagni propounded his organ-based pathology: only if we
truly understand a disease can we treat it correctly. Nowadays,
Diagnosis and therapy

look each other in the eye
Triple influence of genes: hepatitis C
1a, 1b,
2a, 2b,
3a
1b, 2a,
2b, 2c,
3a
4
4
5a
Over the past few years it has become increasingly clear that
even infectious diseases are subject to a complex set of
genetic influences. This is true not just in regard to the
causative pathogens whose genetic background is often
well known but also in regard to the host. Genetic differences between individuals make some people more, and
others less, susceptible to particular infections. In addition,
our genes influence the way our body deals with all types of
drug, including anti-infective agents. In particular, drugs for
use against viruses, which evolve rapidly, often show unsatisfactory efficacy and troublesome side effects. Therefore, if
better drugs are to be developed, both the genome of the
pathogen and that of the patient must be taken into account.
An example of an infectious disease that shows this kind of
16
complexity is hepatitis C,
which is expected to become
far more common over the
next few decades. Untreated,
this disease leads to cirrhosis
of the liver in about 20% of
those infected and to liver
1b
cancer in a smaller proportion
2a
of patients. The responsible
1b,
pathogen, hepatitis C virus
1b, 3a
6
(HCV), occurs in at least six
different types, plus subtypes,
3b
that occur with different frequency in different parts of the
1b, 3a
world (see figure). The existence of these variants has a
major influence on the effectiveness of standard therapies.
For example, interferon, the
most important agent for use in this disease, is relatively
ineffective against HCV type 1, the predominant type in
Europe and North America. Moreover, this type of HCV
generally causes more severe disease than other types.
In addition, standard interferon preparations usually cause
severe side effects, all the more so because they generally
have to be taken three times weekly. Administration at such
short intervals is necessary because interferon is broken
down very rapidly in the body and therefore acts for only a
few hours. Patients are thus obliged to put up with a constant ebb and flow of side effects. Improved drugs such as
pegylated interferon have been available for some time
now. Used in combination with ribavirin, pegylated interferon significantly increases the efficacy of treatment.
however, scientists are searching at other sites, namely in the

genes and proteins of our cells in other words, at the sites
where drugs act. For medicine, this is a major advance in that for
the first time in history diagnosis and therapy are, so to speak,
looking each other in the eye. For the first time it is possible to
determine the causes of a disease on the basis of a patients genetic predisposition, to predict the effect of drugs on a disease
on the basis of the molecular characteristics of the drugs concerned, and finally to choose therapy that is optimal for the
individual patient. Knowledge of the molecular level of the disease process thus opens up completely new approaches to treatment: new targets, new strategies, early prevention and better
understanding of the effects and side effects of drugs.
Molecular structure of Pegasys with (right) and without (left) PEG coating
One new way of improving therapy is therefore to increase

the period of time during which interferon remains in the
body. The product Pegasys works in this way. In this medicine the interferon is enclosed within a coating made up of
a branched molecule known as polyethylene glycol (PEG).
This delays breakdown of the interferon, with the result that
the drug only has to be taken once weekly. This results in
greater efficacy and fewer side effects.
Genetic factors are important in hepatitis C not only in relation to virus type and drug metabolism, but also in that they
partly determine the clinical course of the disease. As mentioned above, infections with HCV type 1 are generally more
severe than those with other types of the virus. Nevertheless, infection with any HCV type can show an acute or
chronic course, be mild or severe, and undergo spontaneous cure or lead to liver cancer, the direction taken by the
disease in an individual being largely determined by that
individuals genome. At present, however, the genes
responsible for these differences are largely unknown.
These genes are therefore another important object of HCV
research, since depending on the genetic predisposition of
the patient (and the virus), treatment may be necessary or
unnecessary, a particular drug may be suitable or unsuitable, and cure may be possible or unlikely. Therefore, the
more is known about the relationships between the genome
of the pathogen and that of the host, the more specific will
be the drugs that can be developed for use in this disease.
17
At the same time, these new methods and discoveries are a continuation of the revolution that began several hundred years ago
less emphasis on signs and symptoms, more investigation of
causes. And despite all the progress that has been made, the possibilities offered by medicine remain limited: the interactions
between our genes are so complex, our body is so adaptable, and
the influence of our environment and lifestyle is so great that we
cannot expect to find definitive answers and one-hundred-percent effective therapies in the very near future. Rather, genetics,
genomics and proteomics will first help doctors to avoid ineffective or even dangerous therapies after all, there is no such
thing as a panacea. But even this is a major step forward.
References
Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations.
Pharmacogenomics 1: 23-26, 2001
Bundesministerium fr Bildung und Forschung (ed.): Das nationale Genomforschungsnetz.
Bonn, 2003
Geschftsstelle des Wissenschaftlichen Koordinierungskomitees des Deutschen Humangenomprojekts (ed.): Das Humangenomprojekt 1st and 2nd edition
Healthnet-Services GmbH: (M)Eine Geschichte der Pathologie, Teil 1-7:
http://hns.pvs-bw.de/mod.php?mod=userpage&page_id=30
18
Pharmacogenomics:
genes and drug response
To d a y
To m o r r o w
Diagnostics
A drug may work well in one person,

but poorly or not at all in another.
One person may tolerate a drug well,
whereas another develops side
effects. This fact is as well known as
it is unfortunate. These individual
differences are largely due to our
genome, the genetic blueprint that
makes each of us unique. Thanks to
new knowledge and techniques,
medicine is now able to take greater
account of these differences thus
leading to the development of more
effective, safer and better tolerated
drugs.
B
C
People are all different thats obvious at first glance. There are
extra-long beds, creams for sensitive skin, height-adjustable seat
belts, tailored shirts and three dozen standard shoe sizes. Each
of us has different strengths and weaknesses, abilities and needs.
Environmental factors, chance and above all the small differences in our genomes make each of us unique. But if one person
finds a standard bed to be too short and another finds a standard
shoe size too big, why should we assume that everyone responds
to drugs in the same way?
Terms
In fact, it has been known for
some time that the efficacy
Pharmacogenetics describes the influence of genes on the
and tolerability of drugs
efficacy and side effects of drugs.
Pharmacogenomics studies interactions between drugs and
vary from one person to the
the genome.
next. Thus, some patients
Pharmacokinetics investigates the uptake, conversion and
need a lot more or a lot less
breakdown of drugs in the body over time. Environmental factors,
diet and genetic predisposition all play a role.
of a given drug than most
Pharmacodynamics deals with the influence of genes on the
people; side effects keep ocinteractions between drugs and their molecular targets.
curring unexpectedly; and
sometimes a drug that is usually highly effective does not work at all. Our uniqueness is
reflected in our bodys response to drugs. Because of this, personalised medicine has emerged as a hot new topic of discussion. Future drugs, it is hoped, will be better adapted to our genetic diversity and dissimilar life circumstances and will be
more efficient, more specific and safer. And they will be supported by a battery of fast, simple genetic tests that will enable
doctors to select the right drug for their patients specific needs.
First example
Herceptin
20
A first example of a genetically specific drug has

already been introduced in the form of Herceptin, which is used in the treatment of breast cancer. Herceptin is only effective in women with a genetic defect
which results in the overproduction of a molecule known as the
HER2 receptor. When present in excessive numbers on the surface of certain breast cells, these receptors promote cellular
growth, leading to tumours. Herceptin is directed against the
receptor; it therefore only helps women who have an increased
number of copies of the relevant gene. In all other women this
highly specific drug is much less effective. Herceptin can therefore be used only in conjunction with a suitable genetic test.
Three such tests are currently available: First, the receptors can
be visualised on the surface of tumour cells with the help of
specific antibodies linked to a dye. Second, there is a gene test

known as FISH which directly detects the genetic change in
question. Recently a third test based on the polymerase chain reaction (see chapter about PCR) has become available at least
for research purposes. Using this technique investigators can
copy the relevant DNA section in the laboratory, thus revealing
if the dangerous genetic change is present.
This development has given rise to much hope and anxiety: Expectations range from the attainment of perfectly personalised
drugs to progress for the rich only, and the boundaries between
the feasible and the conceivable, the possible and the necessary,
science and fiction, quickly become blurred. Whether drugs will
ever become as individualised as tailored shirts is more than
dubious, but even a few off-the-peg sizes and somewhat more
variety would be a huge step forward. The engine driving this
progress is genome research.
As early as 1958 the German pediatrician Friedrich Vogel suspected that our genes play an
important role in determining our response to
drugs. He even proposed a name for the branch of science that
investigates this phenomenon: pharmacogenetics, the study of
the influence of genes on drug effects. The new approach quickly led to the first application of genetics in medicine. Over 100
relevant genes are now known, and many more will follow as scientists rapidly refine the field of pharmacogenetics with the help
Pharmacogenetics
as a research discipline
Pharmacogenomics: genes and drug response
21
of methods and knowledge gained from unravelling the human

genome. Meanwhile the subject of enquiry has shifted from individual genes and their effects to the interplay between drugs
and the genome pharmacogenomics.
The significance of this new branch of science is far greater than
the small change in the name implies. We now know that our
genome influences the effects of drugs in at least three ways and
until a few years ago only one of them had really been considered.
z Pharmacokinetics describes the metabolism of drugs, i.e.
their uptake, conversion and breakdown in the body. In some
people, for example, a drug fails even before it reaches its site
of action. Their body takes up the molecule very slowly or
sometimes not at all. In other individuals, conversion of the
drug (for example to remove a protective molecular cap)
proceeds sluggishly. And a third group of patients breaks the
drug down too quickly or too slowly. If a drug is broken down
too rapidly, taken up too slowly or converted too slowly, its
effects will not be felt. Conversely, if a drug is broken down
or excreted too slowly, its effects may be magnified. It then
remains too long in the body, and the risk of side effects
increases sharply. These differences are due not only to
environmental factors and diet but also to peoples genetic
makeup. This is because specific proteins in our body are
responsible for metabolising drugs. And the blueprint for
those proteins resides in our genes. Hence, small differences
in the genomes of patients can result in pharmacokinetic differences. The discipline described by Vogel dealing with the
relationship between genes and drug metabolism is therefore
now regarded as forming part of pharmacokinetics.
z Pharmacodynamics, by contrast, describes the interaction
between drugs and their molecular targets. In the classic case
this relates to the etiology of a disease, i.e. its underlying molecular causes. Usually the activity of an endogenous protein
is impaired. The shape of such proteins is genetically determined. Small differences in our genome can therefore significantly alter the structure of these proteins. And since drugs
are usually highly sensitive to such differences after all, a
drug is supposed to act on a very specific target molecule so
as to have as few side effects as possible they may become
ineffective if the target molecule is altered.
z Palliative drugs, i.e. drugs that bring relief, constitute the
third and most complex pharmacodynamic way in which
genes can interfere with the activity of drugs. Palliative drugs
22
When drugs do not work: pharmacogenetics
effect:
target molecule(s)
etiological or palliative
potential side effects
conversion:
not always
necessary
cleavage of a
protective group
uptake:
e.g. via specific
receptors or channels
breakdown:
cleavage of
protective groups
often by modification
and/or cleavage
new compounds
may be formed
heart
gut
excretion:
often possible only after
conversion or breakdown
: (inactive) drug
: (inactive) drug in the bloodstream
: protective groups
: active drug
ureter
: cleaved, water-soluble
degradation product
Our genome influences the effects of drugs at at least three

levels:
z The metabolism of a drug i.e. its uptake, conversion or
breakdown may be prevented, slowed or accelerated.
Pharmacokinetics investigates the causes of these phenomena.
z Modification of the target molecule can directly weaken
or prevent the action of a drug. Pharmacodynamics describes these underlying etiological differences.
If a drug does not act on the cause of a disease but
rather on its manifestations, many genes may be involved in and interfere with its effects. These palliative
differences are also the subject of pharmacodynamic
research.
do not act directly on the cause of a disease, but rather on its

symptoms. Analgesics, for example, usually do not influence
the cause of pain but merely the perception of pain in the
brain. Nevertheless, such drugs often successfully counter
the cause if (as in painful cramps) the cause is directly related to the symptoms (i.e. the pain itself causes the cramps).
The ways in which such drugs counteract a disease or relieve
its symptoms are usually highly complex, and the genetic
reasons for why a drug might not work as desired can be just
as varied.
23
Genes have been found for each of these three

areas of pharmacogenetics. The vast majority of
them act at the pharmacokinetic level. This research area is already being applied to the routine development
of new drugs. Particularly in the early phases of clinical testing,
subjects are already screened for specific metabolic characteristics. This is generally done in order to obtain a picture of pharmacokinetic differences in the population that is as representative as possible. In this way entire drug families can be classified
according to their pharmacokinetic properties, thus shedding
light on many problems.
For example, an enzyme group known as cytochrome P450 plays
an important role in the metabolism of many drugs. One important function of P450 proteins is to make water-insoluble
substances soluble in order to facilitate their excretion. Over a
third of all drugs and exogenous substances in the body fall into
this category. We now know that because of genetic differences
the P450 machinery works more sluggishly in some people than
in others. In such people, known as poor metabolisers, the molecular disposal of drugs proceeds more slowly.
Another characteristic of the P450 family of enzymes is medically important: They convert fat-soluble molecules to make them
soluble in water. Bonds in the target substances are broken, new
bonds are forged and additional molecular groups are attached.
In other words, a new molecule with entirely new properties is
created. In the worst-case scenario, a harmless substance may be
transformed into a carcinogenic toxin. Many side effects of drugs
are due to the work of members of the P450 enzyme family and
because the enzymes activity can vary from one person to the
next, such side effects tend to occur sporadically.
Prominent example:
cytochrome P450
Hurdles for drugs: genetic polymorphisms

Genetic differences between individuals strongly influence
the function of the corresponding gene products (usually
proteins) and in this way affect the activity of drugs in the
body. Some of these genetic variants, known as polymorphisms, have been identified. They influence peoples response to drug therapies at the pharmacokinetic or pharmacodynamic level.
1. Pharmacokinetics: Genes coding for enzymes involved in the metabolism of drugs form the largest group
of known pharmacogenetic factors. Fluctuations in their
activity can slow or accelerate the uptake, conversion or
24
excretion of drugs. As a result, the drugs do not remain

in the body long enough to be effective or remain in the
body too long so that the risk of dangerous side effects
increases.
2. Pharmacodynamics: Some genes have also been
found that are directly responsible for the structure of
the target molecule, influence the associated signalling
pathway or interfere with some other metabolic pathway
involved in the manifestations of a disease. The interactions between such genes and drugs can be highly
complex.
Personalised medicine: the objectives of pharmacogenomics

Dosage Are the patients individual pharmacokinetic factors known, and can the dose of a drug be adjusted upward
or downward to suit his/her metabolism? This would help
ensure that the drug works while at the same time reducing
its side effects.
Efficacy If it is already known in the initial stage of treatment which drug is likely to work for a patient, a lot of trial and
error would be saved. Valuable time would be gained for the
treatment, unnecessary expense would be reduced and the
patience of the doctor and patient alike would be spared.
Safety Can adverse reactions to a specific drug be predicted? Can the patient be switched to another drug that
he/she is more likely to tolerate? Are there alternatives? Can
provisions be made in advance for medical supportive measures?
Prevention If the cause of the disease is related to genetic factors, the disease can be diagnosed early by means of
tests and possibly avoided by initiating specific measures such
as diet or exercise.
As with all proteins in the body, this variable activity ultimately depends on our genome, which,
in turn, is affected by external factors, including
drugs. Cytochrome P450 is a good example of this phenomenon
as well. The 50-plus genes that code for this family of enzymes
can be activated or suppressed by drugs. In this way drugs may
affect each other, intensifying or cancelling each others effects,
even if they have completely unrelated targets.
Thus, the genome and drugs form a complex network of dependencies and uncertainties, of effects and side effects, which
ultimately makes our bodys individual response to drugs unique, especially when pharmacodynamic factors are also taken
into consideration. Pharmacogenomics can elucidate this interplay and help doctors find out whether, at what dose and with
what risk a drug can be administered to a given patient before
problems arise in the first place.
Genome and environment

act together
The old slogan one size fits all still applies to

most drugs. If a drug fails to work in too many patients, if its activity fluctuates too strongly or if its side effects are
too common or too severe, it is not granted regulatory approval.
This means a huge loss for the manufacturer and one fewer
promising treatment for patients. The failure of a drug in this respect can be due to at least three reasons:
z Are there problems with the pharmacokinetics, i.e. the uptake, conversion or breakdown of the drug, and if so, what
can be done about it?
Old hurdles for new drugs
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 patients disease due to various, possibly unknown,
causes, and if so does the drug act on just one of them?
Distinguishing between these three possibilities has proved very
difficult and often impossible. The consequence is that the development of new drugs is ultimately limited to the detriment
of patients, who have only a limited range of therapeutic options
available:
z New classes of drugs that exhibit especially marked genetic
dependencies are not further pursued.
z Important target molecules are not considered because of
their variable structure, although perhaps this very changeability may cause disease.
z If it is not possible to distinguish between the causes of a disease, often only palliative drugs will produce a uniform effect, since they act on the effects rather than on the cause of
the disease. However, it is usually much safer and more effective to eliminate the underlying molecular cause.
To understand the molecular causes of a disease

or a drugs failure to have an effect, we therefore
need to look at the genes. We first have to identify those sections of the genome that are involved in causing the
disease or in a drugs activity and metabolism. However, because
the metabolic processes on which drugs act are extremely complicated, this is no small task. A promising solution is to look for
single nucleotide polymorphisms, or SNPs for short. These variations, which are spread more or less randomly throughout the
genome, are thought to determine our individual genetic differences to a large degree. Depending on the position of a SNP
within a gene, the corresponding gene product is more or less
strongly affected. An enzyme, for example, may be impaired,
destroyed or improved with corresponding implications for
drugs that interact with that enzyme. If specific SNPs are repeatedly associated with a disease or with specific side effects or
drug failures, it can be assumed that the genes concerned have
something to do with the observed disturbance.
SNPs yield the first

evidence
26
Once such SNPs have been found, the associated

genes have to be identified. Until a few years ago
this meant tedious searching and sequencing. Today, however,
this procedure can be bypassed. Thanks to the Human Genome
Project, which sequenced the entire human genome, the relevant data are already available. The search for the identity and
function of a gene is also straightforward, since researchers can
conduct a computer search for available data or comparable
genes. Even the associated gene products (usually proteins) and
their functions can be pinpointed quickly and easily in globally
linked databases.
The work becomes more difficult when the gene product in
question is unknown or has only been investigated in another
context. Many proteins occur in numerous variants, exist in a
number of complexes or assume different functions in different
cells. In such cases protein biochemists are consulted; they investigate the molecule in detail for its function and role in the
body. After all, in order for drug development and treatment to
take a pharmacogenetically active gene variant into account, the
causes and effects have to be identified beyond a doubt. Otherwise there is no gain over the old trial-and-error approach.
Databases to the rescue
The final link in the pharmacogenetic research

chain is to search for the presence of the gene
variant in a patient. Conducting this search within a reasonable period of time was thought to be impossible just
a few years ago. Our genome contains at least 30,000 40,000
genes and over three billion building blocks and a SNP is a substitution of just one of those building blocks at a very specific
site within an individual gene. Identifying this site requires tens
of thousands of individual experiments. Today all these experiments fit on a single tiny silicon chip.
Measuring just one centimeter square, a single DNA chip does
the work of a large laboratory. Using modern techniques, scientists can array one hundred to several hundred thousand short
DNA fragments or even entire genes on the chip. The DNA
fragments, known as oligonucleotides, are like anglers fishing
for highly specific genome segments in a solution. The key to
how this works is that the oligonucleotides are arrayed on the
chip in the form of single-strand DNA segments. If a genome
section having the same sequence is present in the test solution,
it binds to the fragment on the chip to form double-stranded
10,000 experiments at
once: DNA chips
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 whats needed for fast, simple and affordable genetic testing.
Thus the requirements for more individualised

drug therapy have been met at least as far as the
technical side is concerned. Though many approaches are still in
the trial phase and problems are likely to be encountered, applications are already emerging, for example a DNA chip that
recognises the various pharmacogenetically important gene
variants of the P450 enzyme family. This chip is the worlds first
commercial pharmacogenomic product. It can be used, for example, to screen test subjects during the process of drug development (even though at present such screening isnt usually
based on a gene test, but on physiological investigations). More
specific diagnoses will follow, and in a later step it is expected
that it will be possible to predict intolerance reactions and treatment failures for new and existing drugs. This will enable doctors and patients to resort to other drugs or at least to prepare
for expected problems.
Tailor-made drugs may be a long time in coming. But even these
rather modest goals are beset by obstacles. As with all new medical developments, there are ethical and legal reservations as well
as financial misgivings and still unresolved scientific and technical issues. In addition, applications of modern genetic and genomic research have consistently raised public misgivings, and
particularly in this area pharmacogenomics is in need of a rethink: The essential requirement for more personalised medicine is unobstructed access to genetic data. To be of use to
patients, findings must be made available to doctors and pharmacists. In any case, the prescription of a drug with significant
pharmacogenetic properties reveals the patients diagnosis anyway. And we should also realise that this is already the case with
many drugs today. Ethical, legal and societal implications of genetics and genomics in medicine as well as the technical and economic requisites are dealt with in more detail in the chapter on
basic conditions.
Potential and obstacles
28
References
Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations.
Pharmacogenomics 1: 23-26, 2001
Lindpaintner K: Herausforderungen und Verheiungen einer individuell zugeschnittenen
Behandlung komplexer Krankheiten. Roche, 2000
Frobse R, Albrecht H: Die ganz persnliche Pille. DIE ZEIT, 15/2002
Ma MK et al.: Genetic basis of drug metabolism. Am J Health Syst Pharm 59(21): 2061-2069,
2002
Kroll W, Hartwig W: Pharmakogenomik. Nachrichten aus der Chemie 50, March 2002
Lifescience.de: Pharmacogenomik Die Suche nach den idealen Pillen.
http://www.lifescience.de/ratgeber/mitte/index2.html
Human Genome Project Website: http://www.ornl.gov/
Human Genome Project Pharmacogenomics:
http://www.ornl.gov/hgmis/medicine/pharma.html
29
Proteomics:
seeing through the undergrowth
Every cell in our body contains at

least 100,000 different proteins, and
every cell type contains a different
set of proteins. Proteins form a vast
and highly complex network: they
construct and break down molecules, they transport, store and mobilise substances, they allow cells to
communicate with each other, they
give and receive orders, they keep
cells alive and can program cells to
die. It is precisely in this network that
drugs act, and only now are we
beginning to understand how, where,
when and why they act. Proteomics
can help us to see through this
molecular undergrowth.
The two Australian scientists who coined the term proteome at

a conference in Siena, Italy in 1994 were probably hoping that
use of this term would raise the profile of their field of study,
protein research. For they coined this term in deliberate analogy to the term genome, the principal focus of biological research in the 1990s. At that time the project aimed at decoding
the human genome was proceeding at breakneck speed with the
aid of unprecedented financial, technical and organisational
backing. One by one, yeasts,
threadworms and even huTerms
man beings revealed their
Genome the (largely unchangeable) totality of the genes of
genetic makeup. By contrast,
an organism.
proteins, the major products
Proteome the (in most cases constantly changing) totality
of all these genomes, were
Genomics the study of the form, function and interactions
largely ignored at that time,
of the genes of an organism.
Proteomics the study of the form, function and interactions
being regarded as a subject
for basic research by scientists interested only in the
acquisition of knowledge for its own sake. Since then, this situation has changed completely.
For a number of years now, scientists conducting basic research
have not been the only ones wanting to find out exactly what
cells produce on the basis of their genome. For proteins bring
about the vital processes that take place in an organism and are
therefore the most important target for strategies e.g. those
based on the use of drugs aimed at interfering with these
processes. However, whereas a cell can only ever have one genome, a cells proteome, that is to say the totality of its proteins,
is highly variable. In theory, each cells proteome is different at
every point in time and at every different site within the cell,
since unlike its genetic material, a cells proteins are being constantly produced, broken down, altered, moved around, bound
and separated. The task of bringing light into this tangled undergrowth is an even greater scientific challenge than that of
decoding the human genome, but at the same time it will accelerate progress in gene and genome research.
The importance of proteins lies in the multiplicity of the tasks that they perform. They play a central role in almost all the processes involved in the
life of an organism or viewed on a smaller scale a cell:
Diverse structures
and functions
32
z Structural proteins are responsible for the form and shape of
cells. They form the structural framework of the cell and a

large part of the outer envelope of the cell. Bodily structures
such as tendons and hairs are made of protein. Structural
proteins account for most of the protein in our body.
z Metabolic proteins, or enzymes, are responsible for the constant synthesis, rearrangement and breakdown of all the substances that are required by, or formed within, the body; they
also provide the energy required for these processes. Even
minor disturbances of the complex interplay between these
proteins can result in serious diseases.
z Signalling proteins are responsible for communication within (and to some extent also outside of) the body. These include hormones and intracellular messenger substances.
Many medicines act by interfering with signalling pathways
within the body.
z Regulatory proteins control the processes that take place
within an organism, including correct transcription of DNA,
the genetic material.
In addition, proteins perform a variety of other tasks, e.g. as antibodies in the immune system, oxygen transporters in the blood
and motors in muscle. The complex interplay between all the
proteins of the body is as fascinating as it is impenetrable. It is estimated that each type of cell in our body contains about 100,000
different proteins, whereas our genome contains only 30,000 to
40,000 genes and moreover is the same in all cells. Because of the
Diverse and changeable: the structure of proteins
}
}
primary structure
secondary
structure
tertiary structure
A chain of up to twenty different amino acids (primary structure the variable regions are indicated by the squares of different colours) arranges itself into three-dimensional structures. Among these, helical and planar regions are particularly
common. The position of these secondary structures in relation to one another determines the shape of the protein, i.e.
its tertiary structure. Often, a number of proteins form functional complexes with quaternary structures; only when
arranged in this way can they perform their intended functions. When purifying proteins, it is extremely difficult to retain
such protein complexes in their original form.
quaternary
structure
Proteomics: seeing through the undergrowth
33
amount and complexity of the relevant data and knowledge already available or yet to be obtained, attempts to intervene in the
world of proteins e.g. by means of drugs have until now been
based largely on the principle of trial and error. The new discipline of proteomics aims to change this.
As a first step, many laboratories around the

world are working to develop as complete a list as
possible of all the proteins that occur in the human body. This list is analogous to the sequence
of the human genome insofar as it provides no answers by itself,
but forms an immensely important basis for further research.
Just as gene researchers sifted through the DNA sequence to
identify genes and regulatory elements, protein researchers can
refer to this protein catalogue when conducting their experiments. At the same time, comparison of the genome with the
proteome facilitates the search for new genes and thereby provides genome researchers with an additional tool for interpreting their results.
In the next step, the proteome is considered in relation to its
time and place of occurrence and above all in relation to the external influences that act upon it. The appearance or disappearance of proteins in the course of an illness or in response to
administration of drugs has already been a subject of investigation for decades, however the search for such changes can now
be conducted in far more systematic fashion with the aid of proteomics. For example, a differential protein expression analysis,
i.e. a comparison between the proteome of a healthy subject and
that of a patient with a given disease under the same conditions,
can in theory identify and precisely characterise all the differences between the proteome of these two individuals and make
them visible at a glance. In short, proteome research can help to
identify the causes and effects of diseases, and of the treatment
of diseases, more rapidly, more simply and more precisely.
Protein catalogues
to provide order
and perspective
Proteome research thus constitutes an important

link between various fields and disciplines of
medicine:
z Diagnosis. The effects of genetic changes first become manifest at the level of proteins. In many cases the question of
whether a known mutation actually brings about an illness
Proteome research as a
link between disciplines
34
or is counterbalanced by other factors can be answered only

by means of protein testing. Proteins are therefore more suitable than genes for use as diagnostic markers of complex diseases.
z Therapy. In the vast majority of cases, medicines do not alter
the genome. Even when they influence the expression of the
genome, they do so by inducing changes in the proteome.
Such changes therefore indicate whether, and if so to what
extent, an administered drug exerts an effect. A snapshot of
a patients proteome can thus help the doctor to adapt treatment to that patients individual requirements.
z Research. One of the objectives of proteomics is to identify
metabolic and signalling pathways that play a role in the development of diseases. Each newly discovered protein in such
a pathway is a potential target for new drugs. Proteomics is
expected to provide a major boost for drug discovery.
z Development. The more is known about a target molecule,
the more simply and rapidly can a drug that acts specifically
on that molecule be developed. Knowledge of the proteome
can also provide information on potential problems and side
effects of a drug before they occur in clinical trials. For example, certain marker proteins can indicate that a cell has
been exposed to a substance that is toxic to them. Toxic effects of new drugs can thus sometimes be predicted at an early stage of drug development.
Given its huge importance, protein research has

always occupied a central role in the search for
new diagnostic methods, treatments and techniques. Most of the methods used today were known long before
the term proteomics was coined in Siena. Proteomics is now extending this branch of biology by making it possible to acquire
and analyse huge amounts of data in a minimal amount of time
just as in genome research. And this possibility opens up new
applications for proteomics.
In classical protein experiments, a single protein is isolated, its
identity, form and function are determined, the gene that codes
for it may be identified and, finally, the location of the protein
within the cell is determined. In proteomics the procedure is exactly the same in principle except that thousands of proteins
are analysed, identified and quantified (i.e. the amount present
within a cell determined) simultaneously. To achieve this, the
Strategy: standardisation
and automation
35
clinical sample
tissue
protein separation
(2D gel electrophoresis)
isoelectric point (pl)
molecular weight (kDa)
blood
tissue
protein extraction
detail
spot picking
(each spot contains
a protein)
bioinformatics
protein identification by
database search
mass spectrometry
(of each spot)
Proteome analysis using 2D gel electrophoresis and

MALDI-TOF mass spectrometry. Proteins are extracted from clinical material such as blood or tissue and
separated by 2D gel electrophoresis. The resulting

protein spots are then cut out of the gel and the
proteins are identified by mass spectrometry.
various steps in the process are performed largely automatically

and in parallel, and the results obtained are analysed using powerful computers and specially developed software programs.
The first step is always that of separating the mixture of proteins in a sample. The most important
method used to achieve this both in classical
protein analysis and in proteomics is two-dimensional (2-D)
gel electrophoresis. This technique has been used routinely since
the 1980s, and was in fact the subject of the conference in Siena
at which the term proteomics was first used.
In 2-D gel electrophoresis, the proteins in a sample are applied
to a rectangular piece of synthetic gel. Within this gel the proteins are separated firstly according to their charge and then at
Separation by charge
and size
36
Two-dimensional gel electrophoresis: an important separation method

control
Alzheimers disease
molecular weight
SDS-Page
IEF
Production of glial fibrillary acidic protein (GFAP)

is increased in Alzheimers disease.
In two-dimensional gel electrophoresis, proteins are separated according to their charge and size. The first step is
separation by charge. This is achieved by means of isoelectric focusing (IEF) using a polyacrylamide gel to which
a pH gradient is applied for this purpose. When an electric
field is applied to the gel, proteins migrate along the gradient for as long as they possess a net charge. Once a pro-
tein reaches its isoelectric point, i.e. once its net charge is
zero, it stops.
In the next step, the separated proteins are further sorted
according to size. This occurs at a right angle to the direction of the first separation, i.e. in a second dimension. The
detergent sodium dodecyl sulphate (SDS) is added for this
purpose. The molecules of this substance bind to the proteins to an extent that depends upon the size of the protein molecules. Once again, an electric field is applied,
however in this step the rate of migration is determined by
the charge of the SDS, which in turn is determined by the
size of the protein molecules. This type of separation is
known as SDS-PAGE, or sodium dodecyl sulphate-polyacrylamide gel electrophoresis.
Proteins that are very similar (e.g. modified forms of the
same molecule) are often extremely difficult to separate in
conventional gels. In such cases use is made of narrowrange gels with an extremely gentle gradient within the pH
range being examined. In this way even minimal differences
in charge can be detected.
a right angle to the direction of separation in the first step

according to their size. The proteins are then rendered visible,
resulting in a complex pattern of spots in which each spot represents a specific protein. The larger the spot, the more of the
protein concerned was present in the sample as in a map in
which larger towns are indicated by larger spots. The pattern of
spots, which is referred to as a proteome map, indicates both the
identity and the amount of the various proteins present. Under
the same conditions a given protein will always be found at the
same place; the map thus provides direct information on whether, and if so in what quantity, a particular protein is present in a
sample.
If the protein of interest is found, the spot containing it is cut
out of the gel. In the classical technique this is done by hand,
whereas in proteomics this task is performed by robots. In order
to permit comprehensive proteome comparisons and highly
precise analyses, however, it is desirable that all the proteins in a
sample be analysed. Many of these molecules are present in such
minute quantities that they cannot be rendered visible in the gel
using conventional staining methods. Often, however, it is precisely such proteins that are of interest as potential targets for
37
drugs. Moreover, many proteins occur in a number of isoforms,

i.e. slightly different variants that lie extremely close together in
the two-dimensional gel. In order to permit identification of
these less common variants as well, two-dimensional gel electrophoresis is generally complemented by other separation
methods.
For example, components of the sample to be

analysed can be separated in advance by filtration
or centrifugation. Also available are various techniques by means of which proteins can be presorted according
to various characteristics e.g. charge, size, shape or binding behaviour. And the 2-D gel electrophoresis itself can be further refined by limiting both of the separation factors that it employs,
namely charge and size, to a certain range within which resolution is particularly high. And for the task of excising the spots,
Roche researchers have now developed a grid-shaped tool that
cuts the gel into 6000 tiny pieces, each of which can be analysed
automatically.
In the next step, the proteins to be identified are cut into pieces.
Just as genome researchers split DNA into more easily sequenced fragments using nucleases as molecular scissors, protein
researchers use proteases, another kind of enzyme, to cleave
amino acid chains at precisely defined points. This results in a
mixture of variably sized protein fragments known as peptides.
Since proteases act in highly specific fashion the enzyme trypsin, for example, always cleaves the chain after the amino acids
arginine or lysine the peptide mixture obtained is specific for
each protein/protease pair.
Presorting increases
resolution
In order to identify the cleaved protein, the peptide mixture is fed into a mass spectrometer. As its
name suggests, this device is able to measure the
mass (and thus the weight) of molecules. A peptide mixture to
be analysed is embedded in a carrier material a matrix and
subjected to a laser impulse. The matrix transfers the energy of
the laser to the peptides, which are thereby ionised, i.e. charged,
and vaporised. The charged peptides are now accelerated by a
powerful electric field and fly through a flight tube. Small peptides fly faster than large peptides. The time that they take to
reach the end of the tube therefore indicates the size of the pro-
A fingerprint for every

protein
38
The tiny difference: peptide fingerprinting
UV laser
powerful
electric field
laser beam
+
++
+ +
+
amplifier
data analysis
in computer
matrix peptides
ionisation
paffle plates flight tube

(vacuum)
acceleration
In MALDI-TOF MS (matrix-assisted laser desorption ionisation time-of-flight mass spectrometry), the protein sample to be investigated is digested with a specific protease
and the resulting peptide mixture is embedded in a matrix
in a mass spectrometer. The energy of a laser impulse
applied to the matrix is transferred to the peptides, which
are thereby ionised and vaporised. The charged peptides
time of flight
detector
measurement
analysis
enter a flight tube along the length of which a powerful

electric field is applied. The peptides fly in the direction of
the electric field until they reach a detector at the end of the
tube the lighter molecules arriving first, the heavier molecules last. The time taken by the peptides to fly through
the tube is measured.
tein fragments. This technique bears the rather cumbersome

name of MALDI-TOF MS (matrix-assisted laser desorption
ionisation time-of-flight mass spectrometry).
The result of this measurement is like a fingerprint of the protein under investigation: a distinctive spectrum on the basis of
which the molecule can be identified beyond doubt. In fact, protein researchers actually refer to this process as fingerprinting
and compare their work with forensic science in which fingerprints are used to identify criminals. And just as police compare
fingerprints obtained at the scene of a crime with those in their
files, protein researchers search their databases for the protein
that fits the spectrum they have obtained. To do this, they do not
even need to have tested the proteins concerned, since nowadays
all the proteins in a database can be exposed to a given protease
39
MALDI-TOF spectra with corresponding proteins
1500 2000 2500
3000
3500
4000
4500 5000
1500 2000 2500
3000
3500
4000
4500 5000
Signal broadening caused by carbon isotope
1500
1500 2000 2500
3000
3500
4000
4500 5000
1500 2000 2500
3000
3500
4000
4500 5000
MALDI-TOF MS yields a spectrum a fingerprint that is

specific for the digestant and the protein present in the
sample. Each peak represents a signal of a certain strength
obtained at a certain time. The fingerprint can be reproduced at will and can be calculated in virtual fashion for the
proteins in a database. The spectrum thus obtained can be
used for direct identification of the source protein.
2000
2500
3000
3500
4000
Nevertheless, simple database comparisons of

this kind account for only a small part of protein
research. Only if the protein under investigation
is not found in one of the many databases, and is therefore considered to be possibly unknown, does the really exciting part of
the task begin: the work to identify the form and function of the
40
5000
A peculiarity of mass spectra is the broadening of the signal that occurs as a result of the natural occurrence to an
extent of about 1% - of the heavy carbon isotope 13C. With
larger peptides there is a high probability that at least one
carbon atom in the molecule will have this greater mass.
The presence of this isotope broadens the signal, and the
resulting strong, broad signal can obscure a weaker signal.
Nevertheless, the resolution of the technique can be considerably increased by use of mathematical methods and
in particular by use of extremely sensitive spectrometers.
The accuracy of modern MALDI-TOF mass spectrometers
is rated at about 10 ppm (parts per million).
and subjected to mass spectrometry in virtual form in a computer. To remain with the analogy of forensic science, this is equivalent to a suspects fingerprint being worked out from his photograph a possibility that would presumably make every police
officer in the world green with envy!
Form and function

go together
4500
13C
new molecule in the hope of deriving some medical benefit from

it. For every newly discovered protein is potentially a new target
within the molecular network of our body.
Determination of the form, i.e. the amino acid sequence and the
external shape, of the protein is easier to automate than determination of the function of the protein.Here again,the sequence
of amino acid units can be determined in a mass spectrometer;
in this case the peptides are removed from the end of the protein
and the individual units are determined one by one. Once the
amino acid sequence of the protein is known, conclusions can
be drawn as to the shape of the whole molecule though only to
a very limited extent. For even though an amino acid chain
basically arranges itself into a shape, the shaping of proteins
within cells often occurs via highly complex pathways. Many
proteins are assembled with the aid of enzymes that hold on to
a part of the polypeptide chain while the rest is being formed. In
the absence of these molecular chaperones the units orientate
themselves incorrectly. This process of protein folding is an
important area of research, including medical research. For
example, incorrectly folded proteins are responsible for the
feared brain degeneration of Creutzfeldt-Jakob disease. In this
case there appears to be a chain reaction in which incorrectly
folded proteins are able to induce pathological changes in the
structure of correctly folded proteins.
Until a few years ago determination of the spatial

configuration of a correctly folded protein was a
task worthy of a doctorate. Now, however, proteome research has largely automated this aspect of science and
thereby made it possible to undertake a comprehensive structural analysis of the human proteome. This applies in particular to
x-ray crystallography, the most important method for investigating the shape of small to medium-sized proteins. In this technique proteins are cultured en masse in host organisms, mostly
bacteria or yeasts, and then crystallised. Previously, culture of
crystals was regarded almost as an art form, mastered only by a
few specialists with golden hands. Nowadays it is performed automatically, the only task left to scientists being that of selecting
suitable specimens from the mass of cultured crystals. The tiny
crystals at least a twentieth of a millimeter of edge length is required are then bombarded with x-rays. On their way through
the crystal the x-rays are deflected (diffracted) by the protein
Automated structural
analysis
41
Function follows form: the enigma of protein structure

The structure of a protein determines its function. Thus,
muscle proteins are fibrous, membrane channels are tubular and enzymes are mostly rounded with one or more
depressions into which their substrate fits. Shown left is the
drug Herceptin, an antibody which is able to bind via the
yellow-highlighted regions to a protein that causes breast
cancer. Knowledge of the structure of the target protein is
important, since it permits the development of custommade drugs and thus obviates much trial and error.
Although the shape of a protein is ultimately determined by
its amino acid sequence, it has so far proved almost impossible to predict the structure of a given protein.
molecules, and from the pattern of this diffraction scientists can

work out the three-dimensional form of the molecule.
By far the most difficult part of the analysis of a

new protein, however, is identification of its function, i.e. its place within the molecular network of
the body. Here again, though, technological advances are providing more and more possibilities. These include protein chips:
millimeter-sized silicon wafers on which thousands of different
proteins are placed. If the new protein binds to one of the molecules on the chip, it can be assumed that it also binds to this
molecule in the living cell. Proteins, however, are notoriously
unfaithful: the unions they enter into are generally neither exclusive nor enduring. At another time, at another place, under
different circumstances they will gladly try out something else
with other proteins, with other molecules or even with DNA.
Within each and every cell of our body, therefore, is a complex
network of interactions and conditions which the human brain
cannot take in at a single glance. Fortunately, computers can
perform this task for us.
Bioinformatics is the science that attempts to bring order into
this chaos. It also creates an important link between genetics,
genomics and proteomics, since it combines the data from all
Fishing for the right

partner
42
three of these disciplines. Which gene goes with which protein?

Under what circumstances is this protein formed, and why?
What genetic signals order the production of this protein?
Which proteins help regulate such an order? And where are the
genes for these proteins located? Genes and their products,
proteins, are inseparable. This has long been known to science,
and proteomics is therefore a necessary extension of genome
research.
Cooperation between these disciplines is correspondingly close. Proteomics generally blends

seamlessly with genetic and genomic experiments, whose conclusions it checks and extends. Transcriptome
research, which deals with messenger RNA, the working copies
of our genetic material, is likewise often combined with experiments conducted at the level of proteins. In order to promote
this cooperation and especially in order to consolidate the efforts being undertaken in proteome research throughout the
world, proteome researchers have now formed a worldwide
body the Human Proteome Organisation, abbreviated as
HUPO analogous to that formed to promote genome research.
This organisation aims to undertake a variety of tasks:
z Awareness. As compared with genome research, proteome
research is still scarcely known to the general public. Its tasks
and objectives, and especially its importance, therefore need
to be publicised. The ultimate objective of HUPO in this
regard is to achieve acceptance of the proposition that the
human proteome merits at least as much support and as
many resources as the Human Genome Project.
z Coordination. In its latter stages, the decoding of the human
genome developed into a much-publicised race between the
publicly financed Human Genome Project and the private
company Celera Genomics. This unexpected competition
certainly resulted in the objective of the project being
achieved earlier than had been expected. On the other hand,
it also resulted in a lot of work being duplicated a massive
waste of research capacity that HUPO hopes will not be
repeated in the case of proteomic research.
z Protein catalogue. By analogy with the sequencing of the human genome, HUPO aims to develop a comprehensive protein
catalogue with the aid of which, for example, potential genes
identified in the Human Genome Project can be assessed.
HUPO - Human Proteome

Organisation
43
First successes: new signalling pathways in cancer
Dozens of signalling pathways are involved in the development of the various forms of cancer. Each newly discovered
pathway provides further potential targets for medical intervention. Proteomics can help in the elucidation of such signalling pathways.
control
ICE3
ras-transformed
mouse lymphocytes
prevents the cell from undergoing transformation into a cancer cell. In order to be able to exert this control function, ICE3
is split in healthy cells by an enzyme known as granzyme B
to form apopain. The 2-D gel of the cancer cells shows a
large amount of unsplit, i.e. inactive, ICE3. From a separate
experiment with gene chips it is known that the genetic
instructions for the production of granzyme B are absent
from the cancer cells. From this combination of findings the
signalling pathway that operates here can be deduced:
Signalling pathway diagram
ICE3
HMG2
ras
BTF3a
The 2D gel on the left shows a subset of the proteome of a
normal cell; that on the right the same subset of the proteome of a cancer cell. The differences are readily apparent.
- A protein referred to as ICE3 is present in greater quantity in the cancer cell.
- The amount of the high mobility group protein HMG2
is reduced in the cancer cell.
- The transcription factor BTF3a is formed in greater
amounts in the cancer cell.
ICE3 plays an important role in programmed cell death, or
apoptosis. This suicide of body cells occurs when the
genetic material of a cell is severely damaged; in this way it
44
granzyme B
apopain
apoptose
uncontrolled
cell
growth
The two other proteins whose amounts are altered in the

cancer cells likewise play a role in cancer. HMG2 binds to
deformed DNA. This protein appears to have largely disappeared from the cytoplasm of the cancer cells; this is evidence of genetic damage to the cells.
BTF3a is likewise a DNA-binding protein, however its function is to ensure correct transcription of genes. BTF3a had
previously been shown to be present in increased amounts
only in intestinal cancer cells. This protein therefore has
potential for use as a tumour marker.
Proteomics is nevertheless confronted by problems. One of the most important of these is a direct consequence of the central characteristic of
the proteome, namely its complexity. The total number of proteins in the human body is now known to be many times greater
than the number of genes. There are estimated to be about
100,000 different proteins per cell type in our body; some of
these are present in almost all cells, whereas others occur only in
a small number of cells. The Roche database, for example, currently contains over 150,000 mass spectra, corresponding to
around 4500 proteins. From this it is clear that a great deal of
work remains to be done. It is also likely that many proteins, and
in particular many modifications, i.e. subsequent alterations to
proteins, will prove to be extremely rare. Unlike the situation
with the human genome, it will be difficult ever to claim to have
catalogued the human protein completely. In fact, this will in all
likelihood be impossible, since failure to find any new proteins
in a particular period most certainly does not mean that no proteins remain to be discovered (for example, our body is constantly producing new antibodies in response to antigens). This
applies even more to protein interactions with other proteins
and with other components of the body such as genes. Only
potent interactions of this kind are easy to find.
Another problem that confronts proteome research is already
well known to its big brother, genome research: the problem of
patenting. As with genome research, it is to be expected that
courts of law will eventually determine what can be protected,
and how. For as long as uncertainty prevails in this area, however, every publication constitutes a risk for a researching company. The fact that this problem is well known from genome
research will at least prove helpful in this regard.
The enormous challenge

of proteomics
References
Langen H: Proteomics as a new field in biology: applications and potentials. Presentation at Roche Roundtable on Genetics and Genomics, May 2000
Screening Trends in Drug Discovery: Future Trends in Proteomics. Interview with Hanno Langen in:
Screening, 2/2001
Brauckmann B: Protein analysis helps in the evaluation of new drugs. Roche Facets No. 14, 2000
Human Proteome Organisation Website: http://www.hupo.org/
45
Targets for medicine
Without targets there can be no

drugs. Small wonder, then, that targets are the most sought-after and
fought-over objects in medical
research. New methods make the
search for targets faster, safer and
more effective and thereby lead
to advances in medicine, since every
new target is another ray of hope
in the fight against diseases.
They are the stars of medicine. Everywhere they are sought after, pursued and adored. Whenever they are found, they become
an object of feverish research. Scarcely any other area of biological research is being pursued with such financial backing and
intellectual effort as the search for targets for new drugs. This
is only to be expected, since it is in drug targets that pharmaceutical research has placed its greatest hopes for new, safer,
more efficient and more effective therapies.
In its broadest sense, a pharmaceutical target is any molecular
site within the body that is potentially susceptible to attack by
drugs. Most such targets are proteins, though other biomolecules such as DNA, RNA, sugars and fats also have potential as
targets for drug action. Common to all such molecules is their
key role in metabolic processes and thus their importance for
bodily function and malfunction. Expressed in another way,
targets often play a role in the development of disease. And that
is what makes them such interesting objects of research.
Because of the multiplicity of their functions and

properties, proteins are by far the most important targets for drugs in the body. They play a
role in the development and progression of almost all diseases.
Since their correct function is directly related to their form, one
of the fundamental requirements of a drug is that it be able to
distinguish between the correct and incorrect forms of a target
molecule. Disease can also be caused by an excess or deficiency
of a protein, or by its occurrence at the wrong time or in the
wrong place. Since proteins, even when in the body, participate
in a wide variety of chemical reactions and interactions, it is
relatively easy to influence their actions by means of drugs. Far
more difficult a task is to specifically influence only a certain
action of a certain protein. Almost all currently used drugs
influence the molecular network of the body at the level of
proteins.
Targets in the narrow

sense: proteins
DNA, the substance that bears our genetic information, controls the vast majority of bodily processes and lays down the framework for the
bodys reactions to the environment. Many diseases are due to
genes, though in most cases external factors also play a role. Malfunction of a gene can be due to an alteration in the sequence of
Important cause
of disease: DNA
48
its building blocks. Alternatively, the instructions issued by a

gene may be too strong or too weak, or may be carried out at the
wrong time or in the wrong place. Medically important genes
are often referred to as target genes. Since substances that directly influence DNA are difficult to find, most drugs act either
on gene products or on molecules that regulate, serve or process
genes and most such molecules are likewise proteins.
RNA has a number of functions in the body,

though its full role has yet to be elucidated. One of
its principal functions is to act as a blueprint for
the translation of genes into their products. Messenger RNA
(mRNA), the form of RNA that has this function, is singlestranded and can be blocked by RNA with the complementary
nucleotide sequence (antisense RNA). This possibility is already
being exploited in research as a rapid and rapidly reversible
means of switching off particular genes. RNA molecules can also
act as enzymes, the best-known example of this function being
ribosomes. These complex structures consisting of RNA and proteins are responsible for the synthesis of proteins. RNA molecules
also play a key role in the processing of mRNA. And finally, the
recent discovery of small nuclear RNA (snRNA), the role of which
is still largely unclear, has opened up an exciting new field of
research. RNA forms a far greater variety of structures than does
DNA and in this sense resembles proteins. This makes it more
accessible to drugs than are genes. So far, however, little is known
of the role of RNA in disease.
Much more research

required: RNA
Other substances that occur in the body are also

potential targets for drug action:
z Sugars are found, among other places, on cell
surfaces, where they serve as markers and permit mutual
recognition. They can assume many different forms and are
being intensively investigated at present.
z Fats not only form a large part of cell membranes, but also
serve as hormones, antioxidants and much more besides.
They are relatively small molecules that can assume very different forms.
z All metabolites, that is to say the starting substances, intermediate products and end-products of our metabolism, are
theoretically susceptible to influence by drugs.
Mostly minor roles:

sugars, fats, etc.
49
insulin
extracellular
intracellular
p91
p60PIK
*
*
PI 3'kinase
p85 p 110
??
ISSK
*
*
IRS-1
ISK
RAS
dynamin
Shc
SHPTP2
(SYP)
SOS
B-RAF
GRB 2
*14-3-3
RAF-1
Jun
Fos
p70s6k
??
MAPK
*
p90Rsk
p70s6k
GSK3
*
PP-1
phosphorylase
kinase
inactivation
glucose
transport
glycogen
synthase
activation
CRE's
PLO
*
*
protein
synthesis
PKC
PKC
40S
cGI-PDE
(type III)
translocation
PKa
MEK
NFr B translocation
transcription
factor
activation
GLUT4
translocation
??
glycogen
deposition
Molecules that play key roles in metabolism are potential drug targets.
With the exception of sugars, such molecules offer medicine

only very nonspecific targets for drug action. Moreover, most of
them are formed and broken down very rapidly in the body and
play only a minor role in metabolism as compared with proteins, in particular. Under some circumstances, however, it can
be useful to bind a specific intermediate product of an undesirable metabolic pathway and thereby block production of the
end-product of that pathway. Even the rapid turnover of such
targets can be advantageous if, for example, the action of a drug
needs to be very rapid in onset and brief.
If diseases are to be combated more effectively,

targets must play a central role after all, every
form of medical therapy is directed against some kind of target.
Two basic possibilities suggest themselves in this respect: either
existing therapeutic methods can be improved, or completely
new methods of treatment can be developed.
Improve or invent
50
y
NSF
SNAP
RABs
ARFs
Ga, y
SNARE`s
7TM
In innovative improvement, the target remains essentially the

same, but attempts are now made to influence it by different
means, i.e. by developing new drugs. To this end, the new drug
must be better adapted to the molecular structure of its target.
The development of improved therapies therefore requires a
sound knowledge of the target concerned.
By contrast, speculative target research aims to develop completely new methods of treatment, i.e. to find new molecular targets for drugs. As compared with innovative improvement, this
approach calls for greater financial investment and more extensive scientific input while at the same time having a greater risk
of failure: since no pharmaceutical experience is available with
them, presently unknown targets may, after prolonged and expensive research, prove unsuitable or too difficult to influence.
On the other hand, since agents developed in this way have the
potential to bring major advances in therapy, research of this
kind can have huge consequences both for manufacturers and
for patients.
Most of the hopes placed in new targets are located in the field
of speculative research. However, it is becoming increasingly
difficult to find suitable targets, since most of the molecules that
play a role in the development of the major diseases that afflict
mankind are probably already known. The economic and scien-
Great risk, great benefit: speculative target research

The search for completely new targets has a high failure
rate. In many cases a great deal of intensive research has
to be conducted before it becomes clear whether a newly
discovered target molecule has anything at all to do with a
disease, and even then the importance and general role of
the molecule in the body's metabolism has to be elucidated.
If a potential target is found to play only a minor role in the
genesis of disease, active agents developed to act on it are
unlikely to be of therapeutic value; if, on the other hand, a
potential target is found to play a major role in other important metabolic processes, the undesirable function is likely
to be difficult to block precisely and specifically.
At the same time, however, speculative target research has
the potential to deliver considerably greater therapeutic
advances than can be achieved by simple improvement of
familiar routes of attack. For example, previously unknown
targets can:
z provide a better, or simply a different, way of interfering
z
z
in a disease process and in this way lead to the development of new, and possibly more effective or better
tolerated, drugs;
be closer to the molecular cause of a disease than
presently known targets and thereby provide a more
specific and surer method of attack;
play a role in production of the symptoms, rather than
the cause, of a disease and therefore also have potential for use as a therapeutic target in other diseases with
similar symptoms;
be used as targets in diseases that have hitherto been
regarded as essentially untreatable due to a lack of any
suitable target; these include various types of cancer,
since cancer is often due to a variety of causes that need
to be treated individually.
51
tific outlay required for the development of completely novel

therapies is therefore becoming ever greater.
The profusion of techniques now used to find and

evaluate new drug targets are derived from a wide
variety of scientific disciplines including proteomics and genomics, genetics, molecular biology, chemistry,
physics and even informatics and information technology. In
addition, scientists make use of results obtained in other areas
of research.
The various scientific approaches pursue different goals and to
some extent build on each others findings. Research in this field
includes not just the search for new targets, but also analysis of
the function of these targets, evaluation of the potential usefulness of targets and the search for suitable sites of attack within
targets. For example, some techniques are directed exclusively
toward finding new biomolecules without regard to the biological function of these; others are used to elucidate the molecular
basis of diseases; others to investigate the properties and functions of molecules that play a role in disease; and yet others to
investigate ways of influencing newly discovered biomolecules
and ultimately to find optimal remedies, i.e. drugs, for diseases.
Most work, however, is conducted at a number of different levels of target research.
Successful search
at many levels
The area of drug research that has been most

important historically deals with the most important group of targets, namely proteins. Originally, most such work was aimed at identifying the sites and
mechanisms of action of new drugs manufactured by chemists.
To this end protein biochemists observed the properties of proteins that reacted with active substances and the interactions
that resulted from these reactions. As this field of research revealed more and more about the molecular processes that take
place in our body, speculation began about the molecular basis
of diseases and this provided drug research with new targets.
Today, protein biochemistry is as much concerned with the
search for new targets as it is with the evaluation of targets and
the choice of suitable agents. This situation has arisen because
it is now possible thanks to rapid progress in unravelling the
structure of proteins (e.g. by means of x-ray crystallography
Proteomics and classical

protein biochemistry
52
proteomics
databases
SNPs
gene chips
PCR
transg. mice
unknown
molecule
e.g. protein
proteomics
databases
structural
investigation
protein
structure
possibly
prediction of
structure
databases
SNPs
gene chips
PCR
transg. mice
genetic
background
associated
gene
transcript
regulation
proteomics
databases
gene chips
HTS
properties
physical
chemical
biological
proteomics
databases
SNPs
gene chips
transg. mice
proteomics
databases
SNPs
gene chips
transg. mice
HTS
function
e.g. enzyme,
hormone,
structural
protein
Modern target finding and evaluation is supported by

a wide variety of methods. The techniques employed
vary greatly in technical complexity and cost and often
molecular
environment
associated
metabolic pathway
complexes with
other proteins
when, where,
how, how much
proteomics
databases
SNPs
gene chips
transg. mice
proteomics
databases
transg. mice
HTS
selection
drug
finding
attack which
target where?
chemical
synthesis
test physical,
chemical and
biological
properties
databases
transg. mice
drug
evaluation
preclinical and
clinical studies
marketing
authorisation
procedure
yield useful findings only when their results are considered together. Most techniques are used in more
or less modified forms at various levels.
and mass spectroscopy; see chapter on proteomics) to make

predictions as to the required properties of new drugs. As far as
the search for targets is concerned, great hopes have been raised
by the discipline of proteomics, which aims to catalogue and
study the entire complement of proteins of an organism.
Already, a large number of previously unknown proteins and
thus potential drug targets have been discovered, their composition and form elucidated and their function described.
A similarly systematic method of searching for

new targets can now be conducted in the complete absence of any biological system: many drug
targets are now being searched for and evaluated in the worldwide network of research computers. Interconnected databases
are a veritable treasure-trove of information that can obviate
the need for many lengthy and costly experiments. If, for example, the composition, that is to say the amino acid sequence, of
an unknown protein is known, comparison with known mole-
Databases and the Human

Genome Project
53
cules of similar structure provides important information on

the likely properties, and thus also on the function, of the protein concerned. One of the largest sources of data in this field is
the sequence of the human genome, which is now known
thanks to the work of the Human Genome Project. A frantic
search for genes relevant to disease, and thus for targets for new
drugs, is now being conducted among the three billion items of
genetic information that make up this treasure-trove. Computerisation of biological research has now developed into an independent scientific discipline and career path known as bioinformatics.
Single nucleotide polymorphisms, or SNPs (pronounced snips) have become increasingly important in recent years. These randomly occurring variations in single DNA subunits are transmitted from
generation to generation and are considered to be responsible
for many medically important phenomena (see chapter on
SNPs) including intolerance, side effects and variations in the
effectiveness of drugs. They also play a role in the development
of many diseases. Thus, the consistent occurrence of specific
SNPs in patients with certain signs or symptoms suggests that
these SNPs interfere with the function of disease-related genes.
Those genes can then serve as potential targets for drugs. As
with proteomics and genomics, the systematic search for SNPs
is now giving rise to extensive databases.
SNPs single nucleotide

polymorphisms
The basis for another important technique used

in target research was established by the computer industry: several hundred thousand different
DNA fragments can now be accommodated on a glass or synthetic chip measuring just 1.5 cm square. This makes it possible,
for example, to find genes containing segments with an absolutely specific sequence (see chapter on DNA chips). DNA
chips can also be used to investigate the enormous variety of
messenger RNAs present in a cell, i.e. the transcriptome of a
cell. In a simple experiment of this type it is possible to determine whether a gene is transcribed at all in a given tissue under
a given set of conditions. In this way the transcriptome too can
provide information on potential targets for drugs; moreover,
mRNA itself contains possible targets.
DNA chips, the transcriptome and protein chips
54
Virtual laboratory: bioinformatics

For a long time now, the work that researchers do in their
computers has been at least as important as the work they
do in their laboratories. Modern experiments, especially if
automated, yield an unimaginable amount of data that could
not possibly be analysed without the aid of powerful computers and specially developed computer programs. Bioinformatics, the application of modern information technology
to biological research, has thus developed into an independent scientific discipline with a broad range of applications,
including the following:
z Sequence analysis is the original and core activity of
bioinformatics. For example, special programs have
been used to sift through the three billion building
blocks that make up our genome in order to identify
genes and regulatory elements. Such programs can also
perform sequence analysis of proteins and translation in
both directions (from gene sequence to protein and
vice-versa).
z DNA chip experiments provide amounts of data that
previously would have been acquired only after years of

work in major research institutions. In order to permit
evaluation of such data within a reasonable time, bioinformatics specialists are constantly developing new
programs and databases.
Prediction of the structure of proteins is a relatively new
and controversial subfield of bioinformatics. Modern supercomputers attempt to predict the three-dimensional
form of proteins on the basis of their amino acid sequence.
Use of databases and networking of data and laboratories are basic prerequisites for smooth performance and
evaluation of experiments. Known information must be
accessible everywhere at all times and must be available
in a form that is compatible with radically different computers, operating systems and programs.
In theory, the DNA chip technique is also suitable for the investigation of proteins, since in principle any molecule can be investigated on the silicon surface of chips. For this purpose, however, proteins like DNA have to be attached to, or ideally
synthesised on, the substrate, and this is vastly more difficult to
achieve with sensitive, highly complex molecules such as proteins than with a relatively stable and structurally simple molecule such as DNA.
Vanishingly small amounts of DNA can be rendered visible by means of the polymerase chain
reaction (PCR). With the aid of the DNA-extending enzyme polymerase, a single DNA molecule can be copied in
a chain reaction (amplified) as many times as desired and thus
made available for investigation. This technique has made it possible, among other things, to detect the presence of minute
amounts of the genetic material of HIV, the AIDS virus, and in
this way identify variants of this deadly infectious disease. If a
drug is to be used against one of the few, but highly variable,
products of viral RNA, PCR can be used to identify the specific
variant of this target molecule present in a patient and then
select the most appropriate drug. In future PCR will therefore
PCR polymerase chain

reaction
55
High-throughput screening system.
become of great value for investigating and classifying targets

and for determining the exact concentrations of these in patients
(see chapter on PCR).
High-throughput screening is an important aid

in the search for suitable drugs for a given target.
In this technique thousands of chemical substances (for example, modifications of a computer-designed
molecule intended to fit into the binding site of a protein variant that causes disease) are automatically tested for certain
properties (such as binding to the intended target). The data obtained are analysed by computer and candidate substances undergo a further round of improvement. This continues until a
molecule with optimal characteristics has been found. Highthroughput screening now performs reliably, rapidly and cheaply a task that only a few years ago required an enormous amount
of work on the part of hundreds of biologists, chemists and
physicists.
High-throughput
screening (HTS)
56
Animal models
What use is a target about which nothing is

known? Among the bewildering amount of data
provided by genomics and proteomics are many hitherto unknown biomolecules, however these are not (yet) targets. In order to qualify as targets, such molecules must not only play an
important role in the body, but also make a significant contribution to the genesis of disease. Unfortunately, however, the
function of an unknown gene or protein is not easy to discover,
especially as it can vary with environmental conditions and with
time and place. In complex diseases such as cancer, Alzheimers
disease and diabetes it is even more difficult to determine the
role and importance of a potential target in the molecular
processes of disease especially as such diseases can appear early or late, occur in more or less severe form, have many causes
and show different signs and symptoms.
Since it is impossible to identify and characterise all the factors
that can influence a complex disease, researchers make use of
animal models to investigate such diseases. The effects of specific changes especially gene variants on the health of animals
are observed, since the environmental factors that operate in
such models are known and can be controlled. Different animals
have proved useful for investigating particular diseases and addressing specific questions. For example, a large part of our
knowledge of human embryonic development and the molecular
basis of cancer was obtained via studies on fruit flies and threadworms, since it was in these animals that the genes that play a crucial role in these processes were first found and investigated. In
the case of applied research
into how disease develops, the
most important model is the
mouse.
Depiction of a target molecule with bound drug.
57
Humans and mice share 99 percent of their genes.

This means that Homo sapiens and Mus musculus
each possess only about 300 genes that are not
present in the other. Scientific findings obtained in mice can
therefore be extrapolated with relative ease to humans. The
mouse is thus an almost ideal experimental animal, especially as
it is easy to rear and breed in a laboratory. For decades, therefore, mice have played a central role in research into diseases.
Moreover, in the year 2002 the Mouse Genome Project was
brought to a successful conclusion, permitting direct comparison between the human and the mouse genome.
Of considerable importance for the role of mice in modern
medicine were experiments conducted in the early 1980s in
which foreign material was for the first time successfully inserted into the germ line of mice. This can now be done with any human gene. Animals into which a small amount of foreign genetic material has been inserted are known as transgenic. They are
particularly valuable for medical research purposes because the
effects of genes on the development of disease can be directly
studied in them. In transgenic mice in contrast to cell cultures
or cell extracts, for example the entire molecular environment
of a disease, i.e. the totality of factors that can promote or inhibit
the development of a disease, is present. The effects of medical
interventions are therefore very similar to those that actually occur in the human body.
It is also possible to selectively remove certain genes from the
genome of a mouse. Observation of these knockout mice (as
opposed to knock-in mice with additional genes) then reveals
the consequences of absence of the gene concerned, and this in
turn permits conclusions as to the function of the gene. These
two techniques (knock-in and knock-out) can be combined so
as to replace a gene with a different one. If the second gene is a
variant of the first, the effects of the variation can be directly
observed.
Transgenic mice are used above all to investigate the molecular
basis of diseases, and this work is resulting in the discovery of
more and more extremely important targets. Genetically modified laboratory animals are also being used to study targets
discovered by other means and, not least, to evaluate potential
drugs. In fact, experiments on transgenic mice can often replace
those on other animals such as monkeys, and even studies in humans, in the early phases of drug development.
Transgenic mice as a
model
58
Important objects of research: transgenic mice

Transgenic mice are among the most important tools of
molecular medicine. These animals have been genetically
modified either by insertion of additional mostly human
genes into their genome or by selective removal of certain
of their genes. Genes can also be altered or else removed
and then replaced by others. All such types of transgenic
mice are excellent models for investigating the influence of
genes and environmental factors on the development and
progression of diseases. The breeding and use of animals
for pharmaceutical research are strictly regulated in all
European countries and in most cases are under the control of independent committees that include representatives
of animal protection groups.
Two techniques are used to insert foreign genes into animals:
1) The technique of microinjection into the pronucleus of
fertilised ova was introduced in the early 1980s and has
now been performed successfully in a number of animal
species. A solution containing many copies of the desired gene is injected into the maternal or paternal
pronucleus (undeveloped cell nucleus) of a fertilised
ovum. Copies of the gene are randomly incorporated
into the genome of the resulting zygote. The ovum is

then implanted into the uterus of a female animal. About
a quarter of the offspring produced in this way contain
the desired gene in their genome and subsequently
transmit it to their own offspring. This method can be
used only to add genes to an animals genome.
2) By contrast, the technique of microinjection into the blastocyst can be used to specifically modify or eliminate an
animal's own genes. To date the only mammal in which
this technique has been performed successfully is the
mouse. This technique employs embryonic mouse stem
cells i.e. cells that still have the potential to develop
into any kind of cell whose genome has been altered
in the desired way by means of recombinant technology.
These cells are injected into a mouse blastocyst (multicellular stage of a developing embryo). The embryo is
then reimplanted into the mother, where it develops into
a chimeric mouse, only a proportion of whose cells possess the desired genetic modification. Those offspring
of such chimeric mice whose ovaries have developed
from the inserted stem cells will bear the altered genes
in all their cells.
Controlled mutagenesis: blastocyst injection
Microinjection of DNA
DNA
embryonic stem cells
unfertilised ovum
cell injection into

the blastocyst
DNAmicroinjection
donor blastocyst
fertilised ovum
reimplantation in mouse
prepared for pregnancy
reimplantation
mouse prepared
for pregnancy
chimeric mouse
breeding
transgenic offspring
offspring with the desired mutation
59
The existence of so many technical possibilities

does not make the search for targets in any way
routine. All hopefulness notwithstanding, it remains true that a newly discovered biomolecule is by no means
necessarily also a suitable target for drug research. As mentioned
above, decades of research have in any case already resulted in
the discovery of a great many targets. This applies in particular
to molecules that are relatively common, long-lived and stable
and that appear to play a major role in important diseases molecules, in other words, for which the risk of failure in drug research is relatively small.
The days of the gold rush in speculative target finding are therefore over. To continue with this metaphor, the large nuggets have
all been found, the claims have been pegged out, whatever gold
remains is buried quite deep below ground, and how much remains is more or less known already. The scientists of the Human Genome Project found that our genome contains only
about 30 000 genes less than a third the number that had been
hoped for: after all, even a lowly threadworm has 20 000 genes.
On the other hand, it is now known that these genes can be transcribed and translated into proteins in many different ways
and every type of cell in our body has at least 100 000 different
proteins, which is good news for target finding. Nevertheless,
this number includes isoforms the in many cases only slightly
differing variants of a given protein and protein complexes,
which in some cases can form and disintegrate rapidly. Estimates of the number of actual targets among proteins therefore
range from a few hundred to a few thousand i.e. not all that
many, whatever the exact figure.
The search becomes more

difficult
Nevertheless, the search is worthwhile, because

we still lack causal therapeutic options for use
against far too many of the major diseases that
afflict mankind, including cancer, cardiovascular disease, Alzheimers disease and diabetes. Even in the case of infectious diseases, medicine is at a very early stage in terms of the possibilities it has to offer precisely because molecular methods have
now made it possible to attack many pathogens selectively. Also
unsatisfactorily treated to date are many diseases which only a
few decades ago had not been investigated at all or else were incorrectly classified, such as allergies and autoimmune diseases.
And finally, many rare diseases have become objects of drug re-
Weighing risks against

benefits
60
search now that the existence of global markets and public funding has made the development of drugs for use against them economically viable for private companies.
The decision as to whether, when and how a potential target
should be further developed depends on a number of factors.
First among these is medical benefit: what benefit will a new
drug bring to victims of the disease it is intended to treat? In the
case of treatments for previously untreatable diseases (or forms
of disease) the answer is obvious, however even benefits such as
substantially improved tolerability, greater potency or avoidance
of side effects can justify the development of a new drug.
Another important factor influencing the decision for or against
a drug target is a scientific analysis of anticipated expenditure
versus risks. In other words, what is the probability that an agent
for use against the target can actually be developed, and what expenditure will be required in order to develop such an agent? Before these questions can be answered, certain details of the structure and function of the target must already be known after all,
it is pointless to direct efforts and resources to targets that are
difficult to influence if the therapeutic objective might be more
easily achievable by influencing other targets.
A third important factor influencing the evaluation of drug targets is economics. Every privately
owned pharmaceutical company has to finance
its research spending with income obtained from the sale of its
products without money there can be no research. Though
this is fundamentally true of any company that develops new
products, some particular considerations apply to the field of
medicine.
z Probably the most important of these see above is a duty
of care towards patients. The high degree of responsibility
borne by the healthcare industry in this regard inevitably influences economic decision-making. This gives rise to conflicts whenever a medically (or socially or politically) correct
decision is clearly economically incorrect in other words,
when it is clear from the outset that the cost of developing a
medically worthwhile drug can never be recovered via sales
of the drug. Some such conflicts can be resolved via orphan
disease programmes in which the state provides financial
and legal support for research into particularly rare diseases
and the development of drugs for use against them.
Drug research must also

pay for itself
61
z Another peculiarity of privately funded pharmaceutical re-
search is the composition of the electoral college that participates in the decision-making process. This includes not
just the company itself in the form of its management, research, finance and marketing directors, employees, investors
and financial analysts, but also patients in the form of patient
associations, legislators in the form of regulatory authorities,
and the public in the form of the press and a variety of associations.
z An additional peculiarity of the pharmaceutical industry is
the sheer unpredictability of biology. All scientific and technical advances notwithstanding, interventions in biological
systems are always subject to a high degree of uncertainty.
Evaluation of targets is therefore often difficult. Even the fact
that the total number of potential targets is unknown is important in that it has a substantial impact on the economic
significance of each and every target. A newly discovered biomolecule can turn out to be useless after prolonged research,
while years later it may suddenly acquire considerable value.
The same is true of new drugs: whatever precautions are taken, a new drug may prove to be unsuitable or even dangerous;
conversely, a completely new set of indications for a drug may
suddenly be found. Since the biological system that forms the
subject of medicine is the human body, such risks and uncertainties are of the greatest importance and the cost of researching and developing new drugs is correspondingly high.
z Undoubtedly one of the most-discussed aspects of the economics of pharmaceutical research is the question of patents.
While patent protection is essential for the survival of any
economic sector that undertakes research and development,
biological patents are subject to particular problems. Prominent among these is the fact that the distinction between a
(nonpatentable) discovery and a (patentable) invention is often difficult to make in biology. It is also argued that overgenerous patent protection can inhibit further research, and
thus medical progress. This view is propounded not just as
always by competitors, but also by a variety of community
groups, politicians and basic researchers representing to
some extent conflicting interests. After some heated initial
discussions and a certain amount of legal toing and froing,
the legal and practical arrangements that have now been
made, though still somewhat woolly, provide companies
with a sensible degree of security for forward planning.
62
research director
medical
profession
financial analysts
target selection
customer
advocates
market group
drug hunter
regulatory agency
Many different constituencies with partly overlapping and conflicting interests influence target selection in pharmaceutical research.
The past few decades have seen a significant shift

in the pattern of decisions on whether or not to
proceed with research on a given target. This is because the development of new drugs has become riskier due to ever-increasing technical costs and a shortage of potentially good targets.
Nevertheless, what ultimately determines a companys success
or failure is the ratio of risk to financial return.
Thus, a venture with a high risk coupled with a low likelihood of
(financial) success is to be avoided. Unfortunately, however, neither of these factors is easy to determine there are many examples of important drugs developed at great expense that
failed as a result of unforeseen problems, while conversely, the
success of many blockbusters (drugs with sales of over a billion
US dollars per year) was not anticipated at the time of their
development. Large amounts of money are therefore spent nowadays generating financial projections that attempt to take account of all conceivable risks and forms of success.
None of these considerations affects the fundamental value of
targets in medicine: without new targets, genuine progress is
difficult to achieve. It is therefore with good reason that the
word target, more than any other, embodies the hope for better,
more rapidly acting, better tolerated and individualised therapies, since every new target is at the same time a diagnostic tool
Between risk and return
63
Good fortune: unexpected blockbusters

active ingredient
(product)
position in
drug class
indication
original sales
forecast
Tamoxifen (Novaldex)
1st
breast cancer
100,000
Captopril (Capoten)
1st
hypertension, heart failure
USD 20 million
Cimetidine (Tagamet)
1st
Fluoxetine (Prozac)
Atorvostatin (Lipitor)
peptic ulcer
700,000
nd
depression
th
hypercholesterolemia
2
5
that can improve our understanding of the disease process in patients. We must therefore hope that the pop stars of medicine
continue to make headlines.
References
Lindpaintner K: Pharmacogenetics and the future of medical practice: conceptual considerations. Pharmacogenomics 1: 2326, 2001
Knowles J, Gromo G: Target selection in drug discovery. Nature Rev, Vol 2, January 2003
Brauckmann B: From basic research with genetically modified mice to new forms of medical
therapy. Roche Facets No. 14, May 2000
Ebeling M: Mit eigener Bio-IT am Forschungspuls. Internal Roche publication
Human Genome Project Website: http://www.ornl.gov/hgmis/
Geschftsstelle des Wissenschaftlichen Koordinierungskomitees des Deutschen
Humangenomprojekts (ed.): Das Humangenomprojekt 1st and 2nd edition
64
PCR: an outstanding method
Scarcely any invention has altered

biological science so radically in
such a short period as the polymerase chain reaction, or PCR. With this
technique, minute amounts of DNA
can be replicated very rapidly and
thereby amplified to such an extent
that the DNA becomes easy to
detect, study and use for any given
purpose. The potential of this technique in medicine has long been
known, and ever more applications
are being developed. Wherever
genes provide clues to the cause or
natural history of a disease, PCR is
the method of choice.
Long car journeys can sometimes be a godsend. Driving along a

monotonous stretch of dark road one April weekend in 1983,
American chemist Kary Mullis was struck by an idea that was
later to earn him the Nobel Prize: the principle of the polymerase chain reaction. Among the instruments and glassware of
his laboratory Mullis might never have had the most momentous and far-reaching idea of his life.
Within a few years PCR short for polymerase chain reaction
took the worlds biological laboratories by storm. By the mid1980s the technique was
used for the first time to diTerms
agnose a disease, when researchers identified the gene
DNA deoxyribonucleic acid; the chemical substance of our
genes
for sickle cell anemia. At
RNA ribonucleic acid; the chemical substance that makes
about the same time the
up the working copies of genes (mRNA), among other things
Nucleic acids a chemical term that covers both DNA and
method was introduced inRNA; nucleic acids are molecules consisting of long chains of
to forensic medicine. The
nucleotides linked together
polymerase chain reaction
Nucleotides the building blocks of DNA; they comprise
the four bases adenine, thymine, cytosine and guanine (A, T, C,
reaped the highest scientific
G; in RNA thymine is replaced by uracil [U]), a sugar and at
honour for its inventor in
least one phosphate group; without the phosphate group
record time: In 1993, just
these building blocks are referred to as nucleosides
Sequence the order of the nucleotides in DNA (DNA seten years after his historical
quence) or RNA (RNA sequence)
car journey, Kary Mullis rePrimer a short DNA fragment with a defined sequence that
ceived the Nobel Prize for
serves as an extension point for polymerases
Polymerases enzymes that link individual nucleotides toChemistry. The reason for
gether to form long DNA or RNA chains
this extraordinary success is
Hybridisation (annealing) the joining of two complementary DNA (or RNA) strands to form a double strand
that the technique provided
Complementary DNA The building blocks of DNA and
a solution to one of the most
RNA form specific pairings. Two strands whose building blocks
pressing problems facing
form a sequence of perfect pairings are able to form a stable
double strand and are referred to as complementary strands
biology at the time the
replication of DNA.
In the PCR procedure trace amounts of DNA can

be quickly and repeatedly copied to produce a
quantity sufficient to investigate using conventional laboratory
methods. In this way, for example, it is possible to sequence the
DNA, i.e. determine the order of its building blocks. Theoretically, a single DNA molecule is sufficient. PCR is therefore one
of the most sensitive biological techniques ever devised. Given
these capabilities, Mulliss method ultimately ushered in the age
of genomics. From the Human Genome Project to the search for
targets to the development of gene tests, there are few areas of
Rapid DNA cloning
66
genetic research today that do not depend on PCR. Only with

the advent of increasingly sensitive DNA chips in recent years
has PCR faced any notable competition (see chapter on DNA
chips). But even then it is often necessary to first copy, or amplify, the DNA of interest. For this reason PCR and DNA chips
often go hand in hand.
Simple and effective: the PCR principle
denaturing
polymerase
Zielsequenz
extension
Polymerase-Kettenreaktion: PCR
DNS
getrennte
DNS-Strnge
(denaturiert)
hybridisation
1
2
repeat cycles
The polymerase chain reaction serves to copy DNA. It uses

repeated cycles, each of which consists of three steps:
1. The reaction solution containing DNA molecules (to be
copied), polymerases (which copy the DNA), primers
(which serve as starting DNA) and nucleotides (which
are attached to the primers) is heated to 95C. This causes the two complementary strands to separate, a process known as denaturing or melting.
2. Lowering the temperature to 55C causes the primers
to bind to the DNA, a process known as hybridisation
or annealing. The resulting bonds are stable only if the
primer and DNA segment are complementary, i.e. if the
base pairs of the primer and DNA segment match. The
polymerases then begin to attach additional complementary nucleotides at these sites, thus strengthening
the bonding between the primers and the DNA.
3. Extension: The temperature is again increased, this time
to 72C. This is the ideal working temperature for the
polymerases used, which add further nucleotides to the
developing DNA strand. At the same time, any loose
bonds that have formed between the primers and DNA
segments that are not fully complementary are broken.
Each time these three steps are repeated the number of
copied DNA molecules doubles. After 20 cycles about a million molecules are cloned from a single segment of doublestranded DNA.
The temperatures and duration of the individual steps
described above refer to the most commonly used protocol.
A number of modifications have been introduced that give
better results to meet specific requirements.
4
5
67
The basic PCR principle is simple. As the name

implies, it is a chain reaction: One DNA molecule
is used to produce two copies, then four, then
eight and so forth. This continuous doubling is accomplished by
specific proteins known as polymerases, enzymes that are able
to string together individual DNA building blocks to form long
molecular strands. To do their job polymerases require a supply
of DNA building blocks, i.e. the nucleotides consisting of the
four bases adenine (A), thymine (T), cytosine (C) and guanine
(G). They also need a small fragment of DNA, known as the
primer, to which they attach the building blocks as well as a
longer DNA molecule to serve as a template for constructing the
new strand. If these three ingredients are supplied, the enzymes
will construct exact copies of the templates (see box on page 67).
This process is important, for example, when DNA polymerases double the genetic material during cell division. Besides
DNA polymerases there are also RNA polymerases that string
together RNA building blocks to form molecular strands. They
are mainly involved in making mRNA, the working copies of
genes.
Copies of copies
of copies
These enzymes can be used in the PCR to copy

any nucleic acid segment of interest. Usually this
is DNA; if RNA needs to be copied, it is usually first transcribed
into DNA with the help of the enzyme reverse transcriptase a
method known as reverse transcription PCR (RT-PCR). For the
copying procedure only a small fragment of the DNA section of
interest needs to be identified. This then serves as a template for
producing the primers that initiate the reaction. It is then possible to clone DNA whose sequence is unknown. This is one of the
methods major advantages.
Genes are commonly flanked by similar stretches of nucleic acid.
Once identified, these patterns can be used to clone unknown
genes a method that has supplanted the technique of molecular cloning in which DNA fragments are tediously copied in bacteria or other host organisms. With the PCR method this goal
can be achieved faster, more easily and above all in vitro, i.e. in
the test-tube. Moreover, known sections of long DNA molecules, e.g. of chromosomes, can be used in PCR to scout further
into unknown areas.
Copying unknown DNA
68
Soon after its discovery the PCR method was refined in several ways. One of the first modifications of the original protocol concerned the polymerases used.
Like all enzymes, polymerases function best at the body temperature of the organism in which they originate 37C in the
case of polymerases isolated from humans. Below this temperature the enzymes activity declines steeply, above this temperature it is quickly destroyed. In PCR, however, the two strands of
the DNA molecule must be separated in order to permit the primers to anneal to them. This is done by raising the temperature
to around 95. At such temperatures the polymerases of the vast
majority of organisms are permanently destroyed. As a result,
new enzyme had to be added in the first reaction step of each cycle a time-consuming and expensive proposition.
A solution was found in hot
springs. Certain microorganisms thrive in such hot
pools under the most inhospitable conditions, at temperatures that can reach 100C
and in some cases in the presence of extreme salt or acid
concentrations. The polymerases of these organisms
are adapted to high temperatures and are therefore
ideal for use in PCR. Today
Hot spring: Thermus aquaticus. Nearly all PCR techthe polymerases used in
niques in the world now use the Taq polymerases isolatnearly all PCR methods the
ed from the bacterium Thermus aquaticus, a microorworld over are derived from
ganism that dwells in hot springs at about 70C. The first
such microorganisms. This
T. aquaticus strain from which polymerases for PCR were
prominent bacterium goes
obtained was found in Yellowstone National Park in the
by the name of Thermus USA.
aquaticus, and its heat-stable
polymerase, called Taq polymerase, supports an entire industry. The organism was originally discovered in a 70C spring near Great Fountain Geyser in Yellowstone National Park in the USA. Employees of Cetus, who
Kary Mullis was working for at the time of his discovery, isolated the first samples from the hot spring and then cultivated in
the laboratory one of the most useful bacterial strains known today. Meanwhile Thermus aquaticus has been found in similar
hot springs all over the world.
Help from hot springs
69
The introduction of Taq polymerase has certainly not been the only modification to the PCR
method. This was helped by the fact that Mullis
published his discovery relatively early though not without some difficulty. Both Science and Nature, the two most
renowned scientific journals, failed to recognise the significance
of PCR and rejected the paper describing the method. Moreover,
despite global patent protection, the use of the PCR technique is
still free and unrestricted for basic researchers thanks to Roche,
which owns the rights to the method. In 1991 Roche obtained
an exclusive license from Mulliss former employer Cetus for
300 million dollars. Scientists from all over the world have modified the PCR method in many ways and adapted it for routine
diagnostic testing and molecular research. At the same time,
more and more new applications are emerging.
Further developments
around the world
In the 1990s biology was faced with one overriding preoccupation: the unravelling of the genome.
Thanks to huge technical and organisational efforts, first viruses and bacteria, then yeasts, plants and animals
relinquished the secrets of their genetic material. This accomplishment would have been unthinkable without PCR, which
made it possible to prepare large amounts of DNA within a short
time. The simple cloning of DNA has therefore remained one of
the main uses of the method. Thus PCR is used whenever the exact sequence of DNA building blocks needs to be determined:
e.g. in other genome sequencing projects, in gene research, in the
investigation of genomic changes, in the search for targets, etc.
An important topic in the field of genomics today is SNPs (pronounced snips), single nucleotide changes in the genome which
appear to account for a large proportion of the genetic differences between individuals (see chapter on SNPs). Among
other things, SNPs are responsible for disease susceptibility and
for differences in the way patients respond to drugs. In order to
detect such hereditary and often widespread variations, scientists have to sequence the genome of many different people in
parallel. Genes with SNPs are also potential targets for new
drugs. PCR therefore plays a key role in this important area of
drug research.
Forerunner of genomics:
DNA sequencing
70
adenine
thymine
guanine
cytosine
DNA, our genetic material.
When PCR is used only for detecting a specific

DNA segment, the method is referred to as qualitative PCR. Usually the standard protocol is
used. Qualitative PCR is an extremely sensitive method which is
theoretically able to detect a single DNA molecule in a sample
solution. In many cases specific genes are copied in this way, e.g.
in order to identify pathological changes. As mentioned earlier,
the first gene identified by PCR was the gene responsible for
sickle cell anemia. Countless other gene tests have meanwhile
been devised. Qualitative PCR is also used around the world in
forensic medicine to identify individuals. Usually individual regions of the genome are amplified and examined. However, although these regions differ between people, they reveal nothing
about the traits or character of the person in question.
PCR can of course be used to detect not only human genes but
also genes of bacteria and viruses. One of the most important
medical applications of the classical PCR method is therefore
the detection of pathogens. Here PCR is replacing immunological methods, in which antibodies against a pathogen are used to
identify the pathogen in a patients blood. Antibodies are not
detectable until several weeks after the onset of an infection,
whereas PCR is able to detect the DNA or RNA of the pathogens
much more quickly. Moreover, antibodies can remain in the
bloodstream long after an infectious disease has resolved.
Hence, only qualitative PCR can determine whether an infection
Sensitive determination:
qualitative PCR
71
has been eradicated, whether

it is chronic (and might therefore progress unnoticed)
and whether the individual
has been reinfected with a
different but related pathogen. Many viruses contain
RNA rather than DNA. In
such cases the viral genome
has to be transcribed before
PCR is performed, and RTHuman immunodeficiency virus.
PCR is therefore used.
Sometimes it is also necessary to detect pathogens outside the body. Fortunately, the PCR
method can detect the DNA of microorganisms in any sample,
whether of body fluids, foodstuffs or drinking water. PCR is
therefore used in all these areas. One of the most urgent problems PCR is helping to solve is to determine if donated blood is
contaminated. Blood banks are one of the major transmission
sources of hepatitis C, for example, and sometimes of HIV. Fast,
simple and above all inexpensive testing is essential and PCR
ideally meets all these criteria.
Quantitative PCR provides additional information beyond mere detection of DNA. It indicates
not just whether a specific DNA segment is present in a sample, but also how much of it is there. This information is required in a number of applications ranging from medical diagnostic testing through target searches to basic research.
Consequently, although quantitative PCR was not described until the 1990s, the method already exists in a number of variants
and protocols to meet a broad range of requirements. Theoretically it is possible to calculate the amount of DNA originally
present in a sample directly from the amount found at the end
of a PCR run. If, for example, there were not one but two double strands at the start of the reaction, exactly twice as much will
be present after each cycle. However, this simple approach
founders on the fact that conditions for the polymerases are not
optimal at the start or end of PCR. At the start the performance
of the enzymes is limited by the small amount of template present, while in the final cycles the enzymes activity declines as a
More than just yes or no:

quantitative PCR
72
result of continuous temperature changes. Moreover, in these

later cycles the amount of available nucleotides falls and the
newly formed templates increasingly bind to each other rather
than to the primers. The effects of all these factors vary greatly
depending on small differences in the reaction conditions (temperature, duration of the steps, concentrations of the reagents,
etc.). In practical terms it is therefore impossible to draw direct
conclusions about the number of molecules in the original sample from the amount of DNA present at the end of PCR.
Instead, researchers have developed various methods that determine the number of new DNA molecules formed in the reaction, i.e. after each cycle (see box on p. 74). Because this approach
affords continuous observation of the reaction, it is referred to
as real-time PCR. Ultimately such experiments involve conjugating the new DNA copies (but not the primers or free DNA
building blocks) to a dye, thus making it possible to determine
the quantity of template.
Target research
Quantitative PCR is used, for example, to help

search for and evaluate targets, i.e. the sites in the
body at which new drugs can act. This primarily relates to the
discovery of new genes, a task in which PCR is basically used as
a DNA copying tool. The same applies to already known genes
that come in a number of variants and that are fairly widespread
in the population (polymorphisms; see chapter on SNPs). However, for a gene truly to be a target for new drugs, its products
must be involved in the development or progress of a disease.
The common occurrence of specific gene variants in affected
individuals can only serve as an initial signpost. The question
ultimately is not whether a specific form of a gene is present or
not, but whether observed variations i.e. changes in a gene
sequence, multiple occurrences of a gene or its absence really
have different effects in healthy and ill people.
To investigate this a procedure known as target validation we
need to consider the genes products rather than the gene itself.
Gene products are usually proteins. Protein research is therefore
devising increasingly sophisticated methods to detect, identify
and assay its subjects of enquiry (see chapter on proteomics).
But the latter task, quantity determination, cannot be satisfactorily performed with available proteomic methods. Furthermore, proteins often differ markedly in their life cycle and
activity. As a result, the quantity of a protein provides only lim-
73
Shining examples: quantitative real-time PCR

Many applications require the amount of DNA originally
present in a PCR sample to be determined. Many techniques are available for calculating the number of DNA copies formed during the individual steps of the PCR procedure and thus for deriving the quantity originally present in
the sample. They are usually based on the middle, or exponential, phase of the PCR in which the amount of DNA
template is approximately doubled in each cycle.
z Example: competitive PCR This method is now largely
of historical significance only. It was one of the first
quantitative PCR methods developed. In addition to the
template of interest, another DNA template having a
very similar sequence was added to the same reaction
vessel. Both DNA strands were then cloned simultaneously under identical conditions. The amount of template and the amount of competitive DNA formed provided at least a rough estimate of the amount of DNA
present in the original sample.
z Example: real-time PCR Most of the quantitative PCR
methods in use today are based on a 1922 discovery by
the American Russ Higuchi, who used the dye ethidium
bromide (EtBr). Embedded in double-stranded DNA,
EtBr fluoresces when stimulated by light. The observed
fluorescence therefore indicates the amount of DNA
formed and does so at any given time during the PCR
reaction, hence the name real-time PCR. In this method
parallel runs are performed with the same known quantity of DNA and a comparative curve is plotted under
identical conditions. This presupposes that the sequence
of the DNA to be copied is known. Moreover, it is not
possible to distinguish directly between the correctly
formed product and primers that have annealed to form
forward
primer
forward
primer
probe
probe
a double strand. Nevertheless, the principle is still used,

though usually with other dyes that specifically interact
only with the desired DNA product. The experiments
have been simplified by the introduction of special
equipment such as the Roche LightCycler, which automates the entire procedure, heating and cooling the
solutions, stimulating the dye to fluoresce and continuously monitoring the fluorescence. Special computer
programs help to analyse the data.
Example: TaqMan probes: One way to measure only the
desired DNA product during PCR is to use TaqMan
probes, short DNA fragments that anneal to a middle
region of the template DNA (see below). The probes
bear a reporter dye (R) at one end and a quencher (Q)
at the other. Quenchers are molecules that quench the
fluorescence of dyes in their proximity. The polymerases
in the PCR solution are able to break down the TaqMan
probes during the doubling of the DNA template. In so
doing they free the reporter dye, which then migrates
away from the influence of the quencher. Hence the
fluorescence of the dye is measurable only if the polymerase has in fact copied the desired DNA strand. Each
freed molecule of reporter dye represents a DNA strand
that has been formed. TaqMan probes can therefore be
used to measure the amount of DNA formed at any
given time.
R
probe
forward
primer
74
ited information about the expression of the corresponding

gene. A far more sensitive method is available in the form of
quantitative PCR, which measures not the proteins themselves
but rather the working copies of the corresponding genes, the
mRNA.
In accordance with the principle of RT-PCR, the mRNA is reverse-transcribed into DNA. Its original quantity is then determined by quantitative real-time PCR. This provides an informative picture of how vigorously a gene is being transcribed and
in what form, since in many cases one and the same gene can
give rise to different products at different sites in the body. The
initial working copies of such genes are cut and stitched back
together by the cells in various ways. Consequently, the products
can have markedly different properties.
Example: STEP
One method of determining the quantity and

nature of working copies of genes in various samples is single target expression profiling, or STEP. In this technique attention is focused on a specific gene (a target), for which
an expression profile is prepared, i.e. its expression is measured
in various tissues of healthy and/or ill persons (see box on p. 74).
When the results are entered in a diagram it can be readily seen
in which areas of the body the gene in question is particularly
active or inactive. A diagram of this kind is therefore referred to
as a body map. A specific tissue from different people can also
be examined. Several hundred individual values can be meas-
Quantitative analysis of human DNA using the LightCycler.
75
PCR and target evaluation: STEP

Single Target Expression Profiling, or STEP for short, provides an overview of the expression of a specific gene in various samples. In this method the number of working copies
of a specific gene is measured by quantitative PCR. The
samples can come from different tissues of an individual or
from tissues of many different donors. By comparing the
profiles in healthy and ill people it is possible to determine,
for example, whether a gene in question really is associated
with the development or progression of a disease.
colon normal
level
colon cancer
The values obtained are plotted and compared. The example below is a comparative diagram showing the expression
profile of various tissues, including samples from the colon
of a healthy subject and of a patient with cancer of the colon
(circles). The gene in question was suspected of being more
active in the presence of colon cancer. However, the STEP
procedure showed that the colon-specific gene is equally
active in healthy and ill individuals.
ured and plotted. In many cases this reveals whether a target really is associated with a disease.
In addition to target evaluation, the STEP method like other
quantitative PCR techniques is also used in other areas of biological science, e.g. in the development of model systems (for
example new cell lines) and in basic research. The establishment
of whether, when and how a gene is expressed provides important information on the role of the corresponding gene product
in the bodys molecular network. This approach has shed light
on many biological processes, e.g. the bodys response to external factors such as administration of drugs.
Another important application of quantitative

PCR is in molecular diagnosis, i.e. the diagnosis
of diseases based on molecular findings rather
than on physiological symptoms. In this connection the diagnosis of viral diseases is an area that is gaining increasing importance. For simple diagnostic testing, i.e. to determine if a
pathogen is present in the patients body, qualitative PCR is sufficient. However, to follow the progress of a disease and to help
Applications in the field

of infectious diseases
76
choose the right treatment doctors often need to know the actual concentration of pathogens present. PCR is one of the few
techniques available today that is able to measure the pathogen
load. This is an important parameter, for example, in viral infections, which often follow a chronic course and produce no
clinical effects for some time, despite infection and in some
cases ongoing physical damage. In this case the viral load in
the bloodstream can provide an indication of how the disease
is progressing.
In addition, quantitative determinations serve to monitor the
success of treatment. If a drug works in a patient, his/her pathogen load will decline sharply. However, some viruses change so
rapidly that they cannot be completely eradicated by the drugs
used: they become resistant to them. This often occurs, for
example, in viral hepatitis C infection. The hepatitis C virus
(HCV) often causes chronic inflammation of the liver, leading
to liver cirrhosis or even cancer in a substantial proportion of
those affected. Damage to the liver often accumulates over
decades without being directly detectable. As mentioned earlier,
qualitative PCR has been used for some years now for diagnosing HCV infection. But the quantitative form of the technique
opens up whole new perspectives for the treatment of the disease. With the help of this method it is possible to monitor the
success of treatment as well as determine how rapidly the disease
is progressing, if at all. Unlike with conventional methods, it is
also possible to ascertain whether the disease has been eradicated or has become chronic.
Hepatitis C viruses.
77
Example: HIV
Another important application in which quantitative PCR is used in the field of infectious diseases is AIDS. Those infected with the causative agent of the disease, the human immunodeficiency virus (HIV), have to take a
cocktail of usually three drugs indefinitely, this being the only
way to keep the virus in check, if only for a while. HIV mutates
extremely rapidly, quickly becoming resistant to drugs. Viruses
that are resistant to all three drugs at the same time sometimes
even occur, requiring a new cocktail to be used.
In order that this moment, known as a viral breakthrough,
should not pass unnoticed rapid proliferation of the viruses
could cause the disease to flare up again the quantity of viral
particles in the blood must be measured periodically. A rise in
the viral load indicates that the drugs being used are losing their
efficacy. Quantitative PCR permits such monitoring and helps
doctors adjust the treatment optimally.
Genetic factors are always involved in the development of cancer. Their contribution varies greatly depending on the type of cancer. Genes not
only help to determine progression of the disease but can also
have a substantial influence on the effectiveness of the available
treatments. Identifying the genes that play a role in the development of cancer is therefore an important step towards improving treatment. Both qualitative and quantitative PCR play a crucial role in the fight against cancer. PCR can identify genes that
have been implicated in the development of cancer. Often the
genes exist in a number of variants with significantly different
effects. One example is the gene known as p53, whose product is
a central monitor of cellular division. If the function of this
monitor is disrupted in a cell, the cell can become cancerous relatively easily. Variants of p53 and similar genes can be detected
by qualitative PCR, giving doctors and patients an indication of
their personal risk of developing cancer or if the patient already has cancer how aggressively it can be expected to progress. Because multiple changes have usually accumulated before cancer actually develops, a reliable test must examine a large
number of gene variants. For this reason DNA chips are being
increasingly used to screen people for genetic changes (see chapter on DNA chips).
Applications
in cancer therapy
78
Meanwhile, quantitative PCR also is gaining importance in the fight against cancer. This is the
case where a cancer has a genetic basis but is not
due to an altered gene. In some cancers the genes that control
cellular division are intact but have been switched off. This can
occur through a process known as promoter methylation. In the
DNA region containing the start information for reading the
downstream gene (the promoter) the cell attaches small molecules (methyl groups) to specific building blocks of the DNA
(the cytosine bases). As a result, polymerases that normally read
the genes and produce working copies of them are no longer
able to dock to the start region. The gene therefore remains silent and no gene product is formed.
Only in recent years has it emerged that this mechanism of promoter methylation shuts down vital genes in many cancer cells.
Methylation status is therefore of crucial importance because
it provides information on the chance of a tumour becoming
malignant and giving rise to metastases. A simple and reliable
method used for detecting these crucial DNA changes is methylation-specific PCR (MSP). In this relatively new technique cellular DNA is first treated with sodium bisulphite, which converts normal cytosine to the RNA building block uracil but
leaves methylated cytosine intact. This results in different products depending on the methylation status of the DNA (see box).
Specific primers for those products are used in the subsequent
PCR procedure. In this way it can be determined if the original
DNA was methylated or not.
Example:
promoter methylation
A host of other
applications
New applications for PCR are still emerging,

particularly in the field of medicine. The search
for genetic predispositions to diseases is an especially important area of research. In many cases the onset of a
disease can be prevented or at least delayed by lifestyle modification or the taking of medications. One example of this is osteoporosis, a loss of bone density that is especially common in
postmenopausal women. Because the disease tends to run in families, it is clear that genetic factors are involved in its development and progression. An intensive search for the genes involved is currently under way and has already produced some
results. Several promising candidate genes have been identified
that appear to be involved in osteoporosis. In this context PCR
is not only an aid in the search for the culprit genes, but also of-
79
fers the possibility of identifying the responsible gene variants

in patients.
Similarly, PCR is helping in the investigation and diagnosis of a
growing number of diseases. It has also long been a standard
method in all laboratories that carry out research on or with nucleic acids. Even competing techniques such as DNA chips often
require amplification of DNA by means of PCR as an essential
preliminary step.
Identifying inactivated genes: methylation-specfic PCR (MSP)
unmethylated DNA
methylated DNA
CH3 CH3
ATC
CGCGTCG
CH3 CH3
ATC
ATC
primer 2
CH3 CH3
CH3
CGCGTCG
GCGCAGC
ATC
primer 1
ATC
primer 2
GCGCAGC
primer 2
no product
3.
PCR product
only with primer 1
In many tumours important genes that control cell growth

are switched off by methylation of the promoter region.
These changes can be detected by means of methylationspecific PCR (MSP).
In MSP all the normal cytosines (C) of the original DNA are
converted to the RNA building block uracil. The methylated
80
CH3
CGCGTCG
primer 1
2.
UGUGTUG
ACACAAC
ATC
NaHSO3
1.
UGUGTUG
CH3
CGCGTCG
ACACAAC
primer 1
no product
PCR
PCR product
only with primer 2
cytosines, by contrast, remain unchanged (1). The subsequent PCR procedure then uses specific primers for the
various products formed (2). Hence, either the original
methylated or the unmethylated DNA is copied. The original DNA was therefore methylated or not (3) depending on
the primer used to obtain a product.
References
Mullis KB, Faloona FA: Specific synthesis of DNA in vitro via a polymerase-catalyzed chain
reaction. Methods Enzymol 155: 335350, 1987
Higuchi R, Dollinger G, Walsh PS, Griffith R: Simultaneous amplification and detection of
specific DNA sequences. Bio/Technology 10: 413417, 1992
Wilfingseder D, Stoiber H: Quantifizierung von PCR-Produktmengen durch real-time PCRVerfahren. Antibiotika Monitor, Heft 1/2/2002
Reidhaar-Olson JF, Hammer J: The impact of genomics tools on target discovery. Curr Drug
Discovery, April 2001
Brock TD: Life at high temperatures. Bacteriology 303: Procaryotic Microbiology, 1994
http://www.bact.wisc.edu/Bact303/b27
Mullis KB: The polymerase chain reaction. Nobel Lecture, December 8, 1993
http://www.nobel.se/chemistry/laureates/1993/mullis-lecture.html
Kary Mullis Website: http://www.karymullis.com
Deutsches Hepatitis C Forum e.V. Homepage Qualitativer HCV-RNA Nachweis:
http://hepatitis-c.de/pcr1.htm
81
SNPs: the great importance

of small differences
G
G
G
G
G
G
Single nucleotide polymorphisms,

or SNPs, the tiny differences
between individual genomes, make
each of us unique. At the same time,
however, they are partly responsible
for individual differences in the
effectiveness and tolerability of
drugs. SNPs have therefore become
one of the most important objects
of medical research, especially
as they could also provide clues to
new targets.
C
C
C
C
C
C
A
A
A
A
A
A
A
A
A
A
A
A
T
T
T
T
T
T
C
C
A
C
C
A
T
T
T
T
T
T
A
A
A
A
A
A
Small genetic differences come in many varieties, and the vast

majority probably have no consequences at all. Some influence
characteristics that are medically irrelevant (for instance, they
may contribute to us having our mothers mouth but our fathers
chin). Others are potentially more relevant to health (for instance, they may contribute to us being as restless as our grandmother, or to putting on weight
Terms
at as early an age as did
our grandfather). Sometimes
SNPs single nucleotide polymorphisms differences in
these small differences can
individual building blocks (base pairs) of DNA that are distributed randomly over the genome and passed from generation
have more significant conseto generation.
quences (for instance, they
Genotype the alternative forms (alleles) of a gene present
may increase our risk of dein an individual; generally there is a maximum of two one from
veloping a disease or lower
the father and one from the mother.
Phenotype the constitution of a living creature that results
the likelihood that well refrom the interaction of genotype and environmental influences.
spond to a medicine). Very
Phenotype refers both to medically irrelevant characteristics
rarely mostly in the case of
and to diseases.
Pharmacogenetics the branch of science concerned
typical inheritable diseases
with the influence of genetic variation on the effectiveness and
their mere presence or abside effects of drugs.
sence determines whether a
Targets the molecules proteins or small organic molecules upon which drugs act in our body.
family member will develop
a disease or remain healthy.
Most of these genetic differences between individuals consist of single nucleotide polymorphisms, or SNPs (pronounced snips). SNPs are randomly distributed variations of the building blocks of our genome that
make each of us genetically unique. They contribute to family resemblance with regard both to external features and to the risk of
developing certain disorders. In medicine, SNPs have become important parameters in the search for new, safer, more effective and
better tolerated drugs aimed at providing more targeted therapy.
99.9 percent of the human genome is identical in
all individuals. On average, however, one in every
500 to 1000 base pairs of our genome differs from
the one found in the majority of people. These randomly
occurring changes are passed from generation to generation and
account for a high proportion of the DNA differences between
us. It is estimated that between three and six million such
variations lie hidden in our genome. When such a variation is
present in at least one percent of the population or to be more
precise, of a particular population, i.e. an ethnic group it is
referred to as a single nucleotide polymorphism, or SNP.
SNPs: common,
hereditary, stable
84
The small difference: SNPs

G
G
G
G
G
G
DNA sequence
C
C
C
C
C
C
A
A
A
A
A
A
A
A
A
A
A
A
T
T
T
T
T
T
C
C
A
C
C
A
T
T
T
T
T
T
A
A
A
A
A
A
DNA sequences differing by a

single nucleotide
Single nucleotide polymorphisms, or SNPs, are differences

in individual building blocks (nucleotides) of DNA that are
distributed randomly over the genome. They can occur in
any position within or outside of genes and accordingly can
have very different effects. SNPs present within the proteinencoding regions of a gene may result in incorporation of
an alternative amino acid in the protein for which the gene
serves as the blueprint, or template. Depending on where
this occurs within the protein and to what extent the alternative amino acid differs from the normally incorporated
one, such an amino acid exchange can have a profound
influence on the function of the protein.
M
M
SNP C
SNP A
individuals with
different SNPs
In the following theoretical example, replacement of a cytidine (C) with a guanine (G) in the gene results in formation
of an amino acid with completely different properties:
instead of the large, basic amino acid glutamine (Gln), the
small, neutral amino acid glycine (Gly) is formed.
SNP
unchanged
Gene
AAG-CGA-ATT-AGG AAG-GGA-ATT-AGG
Protein Lys -Gln -Ile -Arg Lys -Gly -Ile -Arg
Depending on where it occurs, such a variation can have very

different effects. SNPs are therefore subdivided into four groups
on the basis of their site of occurrence:
z rSNPs (random SNPs): Only about ten percent of our
genome is made up of genes. The great majority of SNPs are
therefore located in what we currently view as silent regions
of our genome. These SNPs are extremely unlikely to have
any perceivable effect on our phenotype, or constitution. In
science, many of them are used as markers in the mapping of
genes within the genome.
z gSNPs (gene-associated SNPs): Many SNPs are situated
alongside genes or in introns, the regions of a gene that do
not code for a gene product, i.e. do not form part of the template for a protein. The fact that they are inherited with these
genes makes gSNPs useful for the study of associations between the gene (and its variants) and certain phenotypes.
Mapping gSNPs may be functionally relevant if they influence important control elements of the gene and thereby decrease or increase transcription of a gene.
SNPs: the great importance of small differences
85
The less common, the more important: SNP groups
chromosome
gene
Genes are arranged in different ways on chromosomes. They

generally consist of a promotor region (pr), in which the control elements of the gene are located; exons (ex), the segments of the gene that are translated into the corresponding
protein (the gene product); and introns (in), segments located
between exons that are not part of the protein template. SNPs
occur in all regions of the genome and are subdivided into four
groups on the basis of their site of occurrence:
ex in
pr
rSNPs (random SNPs) are situated at a distance from

genes and are used mostly as aids to mapping. They are
relatively common (estimated at 36 million in humans).
gSNPs (gene-associated SNPs) are situated alongside

genes or in introns. They too are used mostly for gene mapping; they can also influence the control of gene activation.
gSNPs are less common than rSNPs (estimated at less
than 1 million).
cSNPs (coding SNPs) are situated in exons and often influence the function of the corresponding gene product
(estimated at about 100,000).
pSNPs (phenotype-relevant SNPs) are gSNPs or cSNPs
that influence the constitution of the organism. These are
the most important type of SNP from the point of view of
medicine, but are the least common type (probably no
more than 10,000).
z cSNPs (coding SNPs): Exons are the coding regions of a gene,
i.e. the sequences of a gene that are translated into the gene
product the protein. SNPs that are present in exons can
have a major influence on the function of the protein concerned if they result in incorporation of an alternative amino
acid.
z pSNPs (phenotype-relevant SNPs): Both gSNPs and cSNPs
can influence a persons phenotype: the former primarily via
the amount, and the latter usually via the form, of the protein for which the gene codes. pSNPs are the most important
type of SNP from the point of view of medicine. They form
one of the foundations of pharmacogenetics, the branch of
science concerned with the influence of gene variation on the
effectiveness and tolerability of drugs.
SNP analysis is now performed on a number of mature technology platforms with very high accuracy and reproducibility.
86
The following examples show how important

SNPs are in relation to medical practice. The effects of these sequence variations can be divided
into two categories that are of relevance to medicine: on the one
hand, interference with the action of drugs, or, to be more precise, with the way in which the body deals with drugs; and on the
other hand, involvement in the development and progression of
diseases.
Variations of the first category may help explain why drugs work
more or less well in some people and why they may cause undesirable side effects only in certain people. Knowledge of the position and effects of SNPs should therefore be considered in the
development of better and more targeted forms of treatment.
Variations of the second category are important reasons why individuals differ in terms of their susceptibility to certain diseases
despite living in similar environmental conditions.
Medical effect at
two levels
Gene variants can affect the action of drugs in

various ways, including:
z Absorption. The uptake of a drug into the
body can be disturbed, with a corresponding reduction in the effect of the drug.
z Activation. Many drugs work only after being converted into
a different form in the body, e.g. by removal of a protective
molecule. The function of the enzymes that catalyse such reactions may be impaired by SNPs in the gene concerned.
z Distribution. To work properly, a drug must reach its site of
action at the correct concentration. Since many proteins play
a role in this transport, SNPs can have a major influence on
drug distribution. If delivered at inappropriate (either too
high or too low) concentration to the target, or if delivered
to the wrong site, drugs can have undesirable effects.
z Breakdown and elimination. As in the example of the P450
family of enzymes (see below), foreign substances (e.g. medicines) commonly need to be chemically altered (broken
down) in the body so that they can be eliminated. If this happens too rapidly, the effectiveness of the drug may be reduced, whereas if breakdown is too slow, the drug remains in
the body for too long and reaches excessive levels that increase the likelihood of undesirable effects. Moreover, a
harmless molecule may be converted into one with dangerous properties, e.g. one that promotes cancer.
SNPs relevant to drug

response
(pharmacogenetics)
87
Breakdown four times faster: cyp2c19 and omeprazole
100
80
cure rate in %
The cyp2c19 gene of the cytochrome P450

gene family plays an important role in the
metabolism of fat-soluble drugs. Forexample,
the proton pump inhibitor omeprazole works
up to four times less well in individuals with
a certain variant of this gene (mut) than
in those with the most common genotype
(wt). This is because the drug is broken
down faster in individuals with the mut genotype. In this group of patients the cure rate
can accordingly be improved by giving
higher doses of the drug.
gastric ulcer
peptic ulcer
60
40
20
0
mut/mut
wt/mut
cyp2c19 genotype
z Molecular target. SNPs in the gene that codes for the target
molecule of a drug can directly interfere with the action of

the drug on its target. The beta2-adrenergic receptor and the
anti-asthma drug albuterol provide an example of this (see
below).
SNPs assume particular importance whenever

they are associated with the effectiveness or tolerability of medicines. The cytochrome P450
proteins are of great importance in the elimination of drugs
from the body (metabolism). Many of the P450 proteins occur
in a number of variants (based on SNPs). Some of these variants
have clearly altered function. Depending on which variant is
present, the way the body treats certain drugs, and thus how it
responds to them, can vary substantially. cyp2c19 is a member of
the P450 family and is one of the proteins responsible for ensuring that fat-soluble substances are rendered water-soluble in
the liver and thereby made available for elimination from the
body. Such substances include many drugs, e.g. omeprazole,
which is used to treat stomach (peptic) ulcers. At least two dozen
different SNPs are now known to exist within the cyp2c19 gene
and about 50 more are known to be located in its immediate
cyp2c19 in
stomach ulcers
88
wt/wt
vicinity. Some of these variants are known to have considerable

influence on the function of the enzyme encoded by the gene.
Thus, some individuals break down omeprazole four times
faster than others, with the result that standard doses of this
normally very potent drug bring scarcely any benefit in these individuals. Foreknowledge of these variations can be of great
clinical value: if a persons genetic status in this respect is known,
the dose of the drug can be adjusted accordingly at the outset.
Another example of pharmacogenetically important SNPs is the gene for the beta2-adrenergic receptor. This molecule performs a number of
functions in the body (see box). Activation of it in the lungs
relaxes the smooth muscle of the airways. Some anti-asthma
Beta2-adrenergic
receptors
Triple role: beta2-adrenergic receptors

16
27
NH2
164
cell membrane
COOH
The gene for the beta2-adrenergic receptor has at least

three medically important variants. One of them is of pharmacogenetic relevance, the two others signal increased
risk for disease or disease prognosis.
z The antiasthmatic agent albuterol works well only in patients in whom the amino acid arginine is present at
position 16 of the receptor, whereas it is less effective
in patients with a variant of the gene that results in the
presence of glycine at thisposition (pharmacogenetics).
z The genotype that codes for position 164 of the recep-
tor is significantly associated with an increased risk of

developing end-stage heart failure in the aftermath of a
major myocardial infarction. Thus, it has been suggested that bearers of a certain amino acid substitution
at this position should be followed particularly closely
and may ultimately require heart transplantation (disease prognosis/outcome).
A genetic variant that influences position 27 of the
beta2-adrenergic receptor appears to play a role in the
development of obesity (disease risk).
89
medicines therefore aim to activate this receptor. The presence

of a certain SNP in the gene for this receptor can greatly reduce
the effectiveness of the anti-asthma drug albuterol.
While we generally view the role of SNPs as being

to affect the relative risk of contracting a disease,
in extreme cases a single SNP may actually cause
a disease. In sickle cell anemia, for instance, substitution of a single nucleotide in the gene for hemoglobin, the
oxygen-carrying red blood pigment, results in synthesis of an altered protein which under certain chemical conditions takes on
an abnormal shape. This causes the affected red blood cells to assume the shape of a sickle (hence the name of the disease),
clump together and potentially block small blood vessels, leading to tissue death and extreme pain.
The great majority of diseases that afflict mankind arise as a result of a very complex, much more balanced interplay between
a number of genetic, environmental and life-style factors. In
these diseases SNPs can account for an increased (or decreased)
likelihood of developing the disease. Genetic testing for SNPs
may thus help in the evaluation of an individuals risk of developing a certain disease. Though these tests can only indicate a
somewhat higher or lower risk, they may provide prognostic information that allows the person concerned to make more informed decisions about preventive measures such as lifestyle
changes and about more targeted medical follow-up (early
recognition of any recurrence of the disease).
SNPs can also provide information about the molecular basis of
disease. The finding of an association between certain SNPs and
a particular disease suggests that the gene associated with those
SNPs may play a role in the development of that disease. In this
way new disease-relevant genes, and thus new targets for drugs,
can be discovered (see chapter on targets).
SNPs relevant to the

development and
progression of diseases
SNPs are therefore important at all levels of drug

research and development, from investigation of
the molecular basis of a disease through the
search for and evaluation of new targets to the clinical testing
and regulatory approval of new drugs. Knowledge of the distribution and effects of SNPs is already having a noticeable impact
at the latter two levels. For example, the fact that some people
Drug studies and

regulatory approval
90
break down certain drugs more or less rapidly is already being

considered to some extent in the testing of new drugs, and such
individual differences in drug metabolism are likely to be taken
increasingly into account as our knowledge of the distribution
and effects of SNPs grows.
More ambitious are attempts to develop drugs for use in specific patient groups. Where it is known in advance that a substance
is most effective in bearers of a certain gene variant, the drug
concerned can be tested specifically in that target population.
This, of course, assumes the availability of simple and rapid
methods of testing for the presence of that gene variant in trial
participants. The occurrence of adverse events can be reduced in
a similar way: if SNPs associated with the adverse event have
been identified, patients can be tested for them. Those at risk can
be offered an alternative drug or an appropriately adjusted dose.
In this application the study of SNPs does not lead to the discovery of new medicines, but to diagnostic possibilities that permit better targeted and safer use of existing forms of treatment.
Newly discovered SNPs may of course also be relevant to new
drugs for which regulatory approval is being sought. They may
prove useful in reducing the incidence of side effects or in enhancing response rates (efficacy) if these issues arise in the
course of clinical trials.
Given the likely medical importance of SNPs, it

was important to develop the knowledge base required for finding them and to spur on the development of increasingly high-throughput analytical platforms
and technologies. The task of finding and then evaluating several million variants among the three billion base pairs that make
up the human genome would have been completely impossible
as recently as the early 1990s. The Human Genome Project and
the technologies that were developed as spin-offs from it made
an important contribution to our ability to carry out an extensive examination of our genome for SNPs. For SNP research,
therefore, increasing automation and miniaturisation of biological methods were of pivotal importance.
Paving the way for

new methods
Although SNP research on a large scale has therefore been possible only for a relatively short period of time, a large number of these variations in the human
The SNP Consortium
91
Smaller, faster, automated: methods of studying SNPs

Techniques developed or refined in the past few years permit systematic study of SNPs. Automation and miniaturisation, in particular, have now made it possible to perform in
a single step thousands of experiments which only a few
decades ago would have required years of work in major
research institutions. Evaluation of the enormous quantities
of data gathered in this way has now become the subject
of a special branch of science known as bioinformatics.
Three stages of research can be distinguished:
1. Search for SNPs
A major part of the search for SNPs is now conducted in silico, i.e. by computer. The results of laboratory experiments
are stored in databases which are then compared in computers and scoured for SNPs. Special computer programs
are used as aids in the search; these also ensure that SNPs
are distinguished from sequencing errors. Since individual
SNPs can be very rare, it is still necessary to sequence the
genome of many additional volunteers in order to obtain
comprehensive data. These data are anonymised. The most
important technique used in DNA sequencing is the polymerase chain reaction, or PCR (see chapter on PCR).
2. Evaluation of SNPs
Databases and special computer programs are also used to
identify associations between SNPs and genes. Essential in
SNP analysis by computer.
92
the search for associations between SNPs and diseases or

pharmacogenetic problems are association studies. These
are used to determine whether certain SNPs prevail in individuals with a particular disease. Another important method
is that of in-vitro assays to investigate the properties of proteins. These can rapidly determine whether the presence of
a certain SNP in a gene results in altered function of the
corresponding gene product.
3. SNP tests for patients
The final link in the chain of applications of SNP research is
the development of tests to determine the presence of
SNPs in patients. It is not yet clear which of the various
techniques being developed for this purpose will prove to
be both economically viable and scientifically satisfactory
and therefore become established as the norm. A line
array, for example, can detect 60 variants on a membrane.
Far greater numbers can be detected using a SNP chip. In
this technique, several thousand SNPs can be accommodated and simultaneously tested on a DNA chip. Another
method uses MALDI-TOF (see chapter on proteomics) to
investigate SNP-specific oligonucleotides. Also used are a
variety of other methods that employ various techniques
including PCR and nanotechnology.
genome are already known. This is due in large measure to the

work of The SNP Consortium (TSC), an initiative that was set
up in 1999 by ten major pharmaceutical companies, the UKbased Wellcome Trust (the worlds largest medical research
charity) and five leading academic research centres with the aim
of drawing up a comprehensive SNP map of the human genome.
TSC set itself a two-year deadline for the task of identifying a total of 300,000 SNPs in the human genome and mapping at least
half of these, i.e. determining at least their approximate position
in the genome.
The various partners in this uniquely ambitious, privately funded project contributed over 50 million US dollars to TSC on
condition that the results it obtained be published immediately
and made freely available to the scientific community. The database that was built up a comprehensive SNP map of the human genome is now available free of charge to anyone at any
time and because there are no patents is not subject to any
licensing agreements.
In November 2001 The SNP Consortium published its final data
set: 1.7 million SNPs had been found, 1.5 million of these
mapped and 1.3 million allocated to a specific position in the (at
that time still provisional) sequence of the human genome. TSC
was thus an outstanding success, and its results provide a solid
basis for further SNP research.
Based on this body of knowledge, efforts are now

being directed towards the discovery of medically relevant SNPs. This work can in principle be
pursued via two different approaches:
z In association studies, SNPs in candidate genes (genes which
may plausibly be related to the biomedical issue at hand) are
examined for statistically significant associations with disease incidence, prevalence, outcome or drug response. This is
a technically demanding but important area of research. The
overall function of SNP association studies is to establish a
link between the mere mapping of individual differences in
our genome and the biological and in some cases medical
significance of these differences.
z Whole genome scanning, i.e. the systematic and unbiased
analysis of the entire genome for SNPs that show a statistically significant association with a phenotype, without preconceived notions about likely candidate genes, is the next
Current areas of SNP

research
93
logical, but very challenging, step in the use of SNPs. This approach has the distinct advantage that it is not limited to the
still small number of genes about which we know enough to
declare them candidate genes. At present, however, such tests
are still too costly and time-consuming for routine use, and
not a single example has been published. Also, because of a
number of unresolved issues, the feasibility of genome-wide
SNP association studies remains quite controversial.
From the perspective of patients, SNPs are a major step towards a more personalised approach to
treatment that takes genetic differences between
people into account. From the perspective of the pharmaceutical industry, SNPs have become important parameters to be
considered in drug research and development. SNPs have the
potential to lead to the discovery of new targets and thus eventually to the development of new and better drugs, and they bear
the promise of leading to more effective, safer and better tolerated forms of treatment. Our understanding of the importance
of these small but potentially crucial differences is growing all
the time. Finding those that matter for healthcare has therefore
become an important new aspect of pharmaceutical research
and development.
SNPs opening new

doors to better healthcare
References
Foernzler D: SNPs kleine Unterschiede mit groer Wirkung. BioWorld, June 2000
Stoneking M: Single nucleotide polymorphisms: from the evolutionary past Nature
409: 821-822, 2001
Chakravarti A: Single nucleotide polymorphisms: to a future of genetic medicine.
Nature 409: 822-823, 2001
The SNP Consortium Website: http://www.ncbi.nlm.nih.gov/SNP
TSC data on the CYP2C19 gene:
http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?locusId=1557
Holden AL: The SNP Consortium: summary of a private consortium effort to develop
an applied map of the human genome. BioTechniques 32: S22-S26, 2002
Weiner MP, Hudson TJ: Introduction to SNPs: discovery of markers for disease.
BioTechniques 32: S4-S13, 2002
Abraham J, Wilson DE: Roche scientists exceed expectations of genetic discoveries
more than 18,000 mouse SNPs identified in 27 months. Roche press release, Palo
Alto and Pleasanton, Calif., 5th November 2002
94
DNA chips: choosy fish hooks
Very rapid, very sensitive and very

safe these are the requirements of
a good investigative method.
Biochips satisfy these requirements.
DNA chips, the most important type
of biochip, are presently being developed at breakneck speed: already
tiny, they are becoming ever smaller,
ever more sensitive and ever safer.
At the same time, new applications
are being developed that could give
DNA chips a central role in medical
diagnosis.
When scientists fish in murky waters, it is not necessarily a bad

sign. After all, only rarely does nature provide clear solutions.
For example, a cell extract contains, in at most slightly presorted form, the entire inner life of thousands or even millions of
cells in the form of a generally colourless, opaque, thick fluid. All
that matters is that from such murky solutions scientists be able
to draw clear conclusions. And nowadays they are aided in this
task by fishing lines whose properties would make ordinary
fishermen green with envy:
Terms
fast, accurate and capable of
catching enormous numbers
Biochip a solid substrate (e.g. glass or plastic) upon which
of different types of fish at
biomolecules are anchored.
DNA chip biochip with single-stranded DNA as the probe.
the same time. These fishing
GeneChip a widely used DNA chip developed by the US
lines are known as biochips,
company Affymetrix.
and there can be little doubt
DNA deoxyribonucleic acid; the chemical substance of which
our genetic material consists.
that a bright future awaits
RNA ribonucleic acid; the chemical substance of which,
them. In medicine, at least,
among other things, working copies of genes (mRNA) consist.
cDNA complementary DNA; DNA transcribed enzymatically
they are in the process of
from RNA (mostly mRNA).
turning research, diagnosis
Nucleic acids generic chemical term for DNA and RNA;
and therapy upside down.
chain-shaped molecules whose individual building blocks are
bases/nucleotides.
Biochips are among the most
Oligonucleotides short nucleic acid chains composed of at
important instruments used
most a dozen building blocks (nucleotides).
in the miniaturisation and
Genes functional segments of our genetic material that serve
mostly as blueprints for the synthesis of proteins.
automation of biology. In
Genome the totality of the genes of an organism.
most applications their task
is to recognise and bind to
specific molecules in a solution like fishing lines set up to catch
only one kind of fish, but to do so with a high degree of reliability. The hooks used for this purpose are molecular probes attached to a substrate surface barely the size of a thumbnail. It is
this surface that gave biochips their name: it consists of plastic
or glass and is similar to the silicon chips used in the computer
industry. In principle, any substance that interacts with components of our cells can serve as a molecular probe.
In the type of biochip that is most important at

present, the molecular probe that is attached to
the chip is DNA. In future, DNA chips are likely
to serve the most varied of purposes ranging from
basic research in biology through diagnosis of disease to water
ecology. Equally as varied as their potential applications are the
shape, size and method of manufacture of DNA chips. Despite
The most important

biochips at present:
DNA chips
96
image of hybridised DNA array
DNA chip
fluorescently labelled
RNA (probe)
multiple probes for a single gene
50m
these differences, almost all DNA chips exploit the same biological principle, that of hybridisation.
The four bases that are the building blocks of DNA always pair
with the same partner. Our genetic material therefore consists
of two strands of DNA arranged in the form of a twisted rope
ladder, or double helix. The two strands are complementary in
the sense that the sequence of one can be deduced from that of
the other. This joining together, or hybridisation, of two nucleic acid chains to form a double-stranded structure has been exploited by biologists for decades. For example, labelled short
segments of single-stranded DNA (oligonucleotides) can be used
to search for the presence in our genome of oligonucleotides
with the complementary base sequence.
DNA chips do basically the same thing. DNA fragments tethered
to the chip bind to complementary base sequences in the solution being studied. The difference is that DNA chips make it
possible to perform many such experiments at once: millions of
copies of each of several hundred thousand different oligonucleotides can now be accommodated on a chip measuring just
one square centimeter. Conversely, such a chip can be used to
search for tens of thousands of different DNA segments in a solution and it is precisely this possibility that forms the basis of
entirely new applications in biological research and medicine.
97
fluorescent
labelling
probe
5'
3'
T
target DNA
T
5'
G
3'
hybridisation
G
C
The most important application of DNA chips is

in the search for, and the study of, genes. In this
application, as in the classical oligonucleotide experiment, short segments of DNA are used to help identify
longer genes. There are two basic ways in which this can be done:
either the genes are attached to the chip and incubated with a solution of a labelled oligonucleotide, or else the oligonucleotide
is attached to the chip and the genes are placed in the solution.
Of these two methods, the former was developed first, whereas
the latter is used more commonly nowadays because it permits
the performance of more experiments on a single chip.
The developers of both types of chip were in any case faced with
the same two problems, namely how to get the DNA onto the
chip and how to know when two matching bases have found
each other. In both cases a variety of approaches have been tried,
and it is not yet clear which techniques will win out. A large
number of companies are currently offering competing methods of tackling scientific tasks that are in some cases identical
but in other cases different. It may be that a number of different
techniques will survive.
Field of application:
gene research
98
Common example: the GeneChip
A
A
One of the most commonly used DNA chips at present is

Affymetrixs GeneChip. This is an oligonucleotide chip in
which the short strands of DNA are synthesised in situ, i.e.
on the chip, by means of photolithography (see above):
a. Reactive sites on the chip surface are blocked by photosensitive protector groups (small squares).
b. An opaque mask covers the greater part of the chip;
the beam of light therefore removes only those protective groups that are situated in a certain region.
c. The chip is incubated with a solution containing one of
the four nucleosides adenosine, thymidine, cytidine or
guanosine (A, T, C or G), which likewise bear protector groups; the nucleosides react with the previously
unmasked regions of the chip.
d.f. The cycle is repeated with another mask; this gives rise
to regions with different oligonucleotides of known sequence.
Such a GeneChip can be used to examine different types of
G
G
AG
A T CG
AGAC
CGTC
T AG T
DNA solution. Commonly performed experiments include

genome studies and, in particular, gene expression profiles.
These are used to identify those genes that are actually
expressed in a given cell type or tissue.
For this purpose the mRNA the working copy of genes
in a cell extract is transcribed into cDNA. This process is
known as reverse transcription, as opposed to transcription, which is the synthesis of RNA on the basis of DNA.
Precisely this can occur in a second step, since in many
experiments the cDNA that is formed is transcribed back
into cRNA. In one of these steps a label is introduced; commonly used for this purpose is, for example, the molecule
biotin. The cRNA (or cDNA) is then cut into smaller pieces
and placed on the chip, where it hybridises with
the oligonucleotides. Measurement of fluorescence then
shows how much of the label is bound at what sites on the
chip and thus what quantity of the mRNA of interest was
present in the cell extract.
One of the most important DNA chip technologies was developed by the Californian company
Affymetrix. The name of this companys best-known product,
GeneChip, is often used synonymously with the term DNA chip
to refer to any such product. (In addition to DNA chip and
gene chip, other terms including microarray, genome chip
and gene array are in common use.) Affymetrix manufactures
its GeneChips using the principle of photolithography, just as in
the manufacture of computer chips. In this technique a light
source, special masks and photosensitive protector molecules
are used to deposit billions of oligonucleotides with (at present)
up to 700,000 different base sequences alongside each other in
tiny cells (spots) on a chip (see box on page 100).
The example of GeneChip
99
total RNA
biotin-labelled
cRNA
cDNA
reverse
transcription
AAAA
in vitro
transcription
B
B
AAAA
B
AAAA
B
fragmentation
GeneChip
expression
array
B
B
fragmented,
biotin-labelled
cRNA
hybridisation
B
wash and
stain
scan and
quantitate
B
B
Such a GeneChip is then incubated with a solution containing

the DNA of interest, which has previously been labelled with a
fluorescent dye. Whether given oligonucleotides on the chip
have hybridised with DNA in the solution is apparent from the
positions on the chip at which fluorescent dye is present at the
end of the experiment. For this purpose the individual positions
on the chip are read with a scanner. The readings are analysed by
computer with the aid of specially developed programs.
100
Competing techniques:
design and function of DNA chips
The tasks performed using DNA chips are many and varied, and the design of such chips is correspondingly
diverse. Affymetrixs GeneChip is a commonly used DNA
chip, however various other other manufacturers are
offering a variety of techniques aimed at winning over
customers. Points of difference include not just chip design,
but also the way in which the experiments are performed
and the way in which the results are analysed.
z Probe material The most commonly used DNA chips
use short oligonucleotide chains as probes, however
RNA, cDNA, genes and even whole chromosomes can
be attached to chips.
z Manufacture DNA can be attached to chips in various
ways. These include photolithography, a technique borrowed from the computer chip industry. Other techniques include application by pipette, dropping and
electronic methods, e.g. in a manner similar to the
operation of an inkjet printer.
z Target molecules The probe and target molecules are
dependent on each other. Depending on what type of
target molecule is present on it, a chip may be suitable
for the study of other types of DNA, such as oligonucleotides, RNA, cDNA, genes, chromosomes or whole
genomes.
Reaction Hybridisation is not the only reaction that can
occur on a DNA chip. DNA molecules can also be
bound by ligases or via chemical or photochemical reactions. Another possibility is to combine PCR (polymerase chain reaction, see chapter on PCR) with a chip.
Detection Different reactions on the chip require
different methods of detection, and hybridisation can be
detected in various ways. In addition to fluorescence,
mass spectrometry (MS, see chapter on proteomics),
in particular, and also conductivity and electronic
methods, can be used for detection.
Analysis DNA chip experiments generate enormous
quantities of data that would be impossible to evaluate
without computer assistance. Of importance in this regard are not just suitably sophisticated programs, but
also automatic control of experiments, image analysis,
databases, Internet links and platforms and visualisation
of results.
Though GeneChips and other oligonucleotide

chips are the most commonly used type of biochip at present, a variety of other molecules are used on biochips. In the case of DNA chips, not only oligonucleotides and
genes, but also RNA, cDNA and even whole chromosomes can
be used. Depending on the problem to be addressed and the solution to be examined, chips can be either individually chosen
or specially made. One-off products are considerably more expensive than more or less standard products.
In addition, many attempts are being made at present to produce protein chips with a performance similar to that of DNA
chips. As compared with DNA, proteins are vastly more difficult
to produce in the required quantities and at constant quality.
Protein chips are therefore still very expensive. Attachment to
the chip is also problematic in that many proteins need a great
deal of freedom of movement in order to function correctly. In
addition, assessment of the diverse interactions that can occur
between proteins and other substances is difficult and time-consuming. Given, however, that proteins occupy a central place in
Variety on a chip
101
drug research, protein chips are regarded as an important tool,

especially in proteomics (see chapter on proteomics).
DNA chips have also found broad application in

drug development. In fact, medicine is currently
one of the most important and exciting, though
by no means the only, field of application of these tiny chips.
Many different variants of them are used in almost all branches
of biological science. Their outstanding feature in almost all
these applications is their ability to analyse genes rapidly and
simply. The enormous quantities of data collected in the Human
Genome Project and similar undertakings form the basis for the
evaluation of DNA chip experiments. When only small amounts
of the DNA (or RNA) of interest are available, it is generally still
necessary to amplify the nucleic acids first by means of PCR.
These two techniques are therefore often used in conjunction
(see chapter on PCR).
Growing number
of applications
The first field of application of DNA chips was in

basic research in biology. In this, unlike many
other, fields, use of DNA has long been routine.
Since in this field the chips are often used to address new questions, basic research also leads to the development of new techniques and opens up new fields of application.
Among other uses in basic research, DNA chips have been and
are used to map genomes, to find genes and control elements
and to search the genomes of different organisms for points in
common. Now, however, their role has been extended far beyond these uses: now that the sequence of the human genome is
known, they are being used to investigate the tasks and functions
of genes.
An important instrument for such investigations is gene expression analysis. In this, attention is focused not on the gene itself,
but on the working copies of a gene that are produced in a given
type of cell. These molecules, which are known as messenger
RNA (mRNA), act as intermediaries between the genome and
the life processes of the cell. Their primary role is as blueprints
for the synthesis of proteins. DNA chips now permit rapid and
simple generation of gene expression profiles in which the activity of thousands of genes is determined simultaneously. This
method, which is known as MEP (microarray-based expression
Now routine:
basic research in biology
102
profiling), can be used to answer important questions such as:

Which genes are expressed in which cells? When and under what
conditions does gene expression occur? Which genes are active
in diseases? And how is gene expression affected by administration of medicines?
The results of such experiments provide important insights into
the molecular processes that take place within cells. They also
provide evidence of the role of certain genes in the genesis, progression and treatment of diseases. Medicine thus becomes a
new field of application of DNA chips.
DNA chips long ago became a standard tool for

use in research into diseases, especially as they
permit analysis of almost complete genomes in a
single experiment. Basic research and applied science often
overlap to some extent here, however applications of DNA
chips, and in particular gene expression analysis, are becoming
increasingly important in all other areas of medicine, e.g.:
z Genetic causes of disease. Our genome plays at least a contributory role in the genesis of the great majority of diseases.
Discovering which genes play a role in which diseases and
how genes interact in diseases requires detailed observation
of many DNA segments simultaneously a task for which
DNA chips are well suited.
z Hereditary diseases and genetic predisposition. Where disease-relevant genes are known, DNA chips can make it possible to test patients for genetic susceptibility to the disease
concerned. In complex diseases such as cancer and Alzheimers disease a number of genes and environmental factors
are generally involved. DNA chips can help people who are
genetically predisposed to myocardial infarction avoid additional risk factors such as smoking, an unbalanced diet and
lack of exercise.
z Diagnosis. The causes of diseases can also be determined reliably with the aid of DNA chips. For example, different causes
can often bring about the same signs and symptoms, and if
these causes are genetic in nature they can be distinguished by
means of DNA chips. This is exemplified by various types of
cancer which, though also subject to external influences,
almost always result from genetic defects. Knowledge of what
genetic alteration is present in a patient can have a crucial
influence on what treatment is required. Another example of
A time of upheaval:
DNA chips in medicine
103
the use of DNA chips for diagnostic purposes is in infectious

diseases. Here they can be used to identify pathogens. Examples of both these diagnostic applications of DNA chips are given below.
z Therapy. Our genetic predisposition has considerable influence on the efficacy and tolerability of medicines. This is due
mostly to small differences in our genome known as single
nucleotide polymorphisms, or SNPs (pronounced snips)
(see chapters on pharmacogenomics and SNPs). DNA chips
can be used to detect these differences rapidly and reliably,
and in this way can provide doctors with crucial information
to assist them in choosing the most appropriate treatment
for a particular patient. Also, it is only with the aid of such
techniques that novel medicines that take account of individual differences in the way our body reacts to drugs can be
developed. DNA chips are therefore set to play a major role
in the development of personalised medicine.
DNA chips also have potential for use in consumer protection. An example of this is in the
field of green gene technology, i.e. the use of gene
technology in agriculture. The fact that in most industrialised
countries a proportion of the population takes a sceptical view
of this technology has led to the introduction of a variety of regulations including compulsory labelling. Given, however, that
the vast majority of foods produced in this way do not differ visibly from conventional products, it is often only by means of an
examination of the genome of the plant concerned that adherence to such regulations can be effectively checked. DNA chips
are well suited for use in such tests.
Checking up on green
gene technology
Ecology is a broad field of application for DNA

chips. For example, it is often necessary to distinguish between fairly closely related animal species
in a body of water in order to assess the condition
of the ecosystem concerned. This is because the presence of one
species may indicate a clean, but that of the other a polluted,
environment. Up to now, this task has often required painstaking and detailed work with a magnifying glass or even a microscope, since many species are scarcely distinguishable from their
close relatives on the basis of their appearance to the naked eye.
Chip instead of a
magnifying glass:
ecology and taxonomy
104
DNA chips can make such distinctions more rapidly, more simply and above all more reliably. This ability creates applications
for DNA chips in all situations in which closely related species
need to be studied. These include ecology, taxonomy, anthropology and research into evolution.
DNA chips could also have a bright future in forensic medicine. Prominent in this field of application is the ability of DNA chips to detect differences between individual genomes and thereby to identify
people. This can be required for identification of victims and already plays a crucial role in the search for and conviction of criminals. Many countries, e.g. the Netherlands, also allow their
police to use the genetic profile of people they are seeking in order to draw conclusions as to the external appearance of the person concerned. So far this applies mostly to determination of
sex, however the discovery of more genes could make it possible
also to determine a persons hair colour, eye colour and ethnic
origin. Though at present such forensic tests are performed almost exclusively using PCR methods, DNA chips can play a useful complementary role in many such tasks and in future may be
able to perform such tasks more rapidly and simply than PCR,
thereby supplanting it in this application.
Sure identification:
forensic medicine
Examples now exist of all these fields of application of DNA chips. In most cases commercial
products are already available, though in some
applications DNA chips are still in the developmental phase. As
mentioned above, most interest is focused on basic research in
biology and on medical research and drug development. In the
latter field, use of DNA chips could initiate a change of direction
towards a more personalised medicine that exploits the small
but significant genetic differences that exist between people in
order to develop new, more effective and safer drugs, especially
for specific subpopulations. DNA chips that provide high resolution at a low price form the basis for the kind of rapid and simple genetic test that is essential for personalised medicine. They
are thus important not just for finding genes responsible for
diseases, but also for the development of new drugs, for correct
diagnosis and for the choice of the most appropriate treatment
for the individual patient.
Focus on medical
applications
105
Another medical application of DNA chips is their use in infectious diseases. In this application attention is focused not just on
the genes of the patient, but more specifically on those of the pathogens. Many viruses (e.g. HIV), in particular, along with various other kinds of pathogen, develop resistance to important
drugs extraordinarily rapidly via mutations in their genome.
DNA chips can be used to examine the genome of such pathogens rapidly and reliably so that treatment that is optimal for the
individual patient can be chosen.
One of the earliest examples of the use of DNA

chips in medicine is in the treatment of AIDS.
Human immunodeficiency virus (HIV), the pathogen of this disease, has an extraordinary ability to undergo change. Each of the small number of components
of the virus can change so radically from one generation to the
next that drugs rapidly become quite ineffective against the
virus. In order to keep the virus in check despite this drug resistance, infected people have to take combinations of various
drugs. For a long time the only way of finding out which variant
of the virus was present and therefore which drugs would be
effective in a given patient was by trial and error.
The year 1996 saw the introduction of a DNA chip-based test by
means of which the variants of a certain HIV gene present in an
individual could be detected and drug resistances could accordingly be predicted. The intention was to make it possible for
doctors to prescribe drugs which they knew to be effective
against the HIV variant present in the individual patient, thereby avoiding a lengthy period of trial and error. As it turned out,
this chip did not find a place in medical practice, and in fact sequencing by means of PCR has now become the most important
method used for this kind of molecular diagnosis (see chapter
on PCR). Nevertheless, in the past few years more chips have
been developed to support AIDS therapy. These are designed to
examine as many as possible of the important regions of the
viral genome and thus to make the test applicable to other
categories of drug. In future they could play at least an important complementary role to PCR in this application.
DNA chips designed to identify viruses are also being developed
for other infectious diseases. An example is the hepatitis C virus,
which occurs in at least six different variants, each of which
requires a different form of treatment (see chapter on molecu-
Leading the way

in medicine:
DNA chips in AIDS
106
lar medicine). DNA chips could here again, along with PCR
become the most important means of distinguishing between
these variants. A more recent development is a DNA chip to detect human papillomavirus (HPV). The two dozen variants of
this virus cause genital warts, which are generally similar to the
more familiar common wart and are similarly harmless. Nevertheless, three of the twenty or more variants of HPV can cause
cervical cancer in women. Precise identification of the particular variant present in affected women is therefore of great importance. Up to now the condition of the neck of the womb has
been assessed by means of a diagnostic smear so that any altered
tissue can be removed. In extreme cases the entire uterus needs
to be removed. A DNA chip now makes it possible to identify papillomavirus present in the patients blood and thus to estimate
the risk of cancer more precisely. Women at high risk can thus
adjust their family planning to their increased risk.
The ability of DNA chips to detect all the variants

of a number of genes simultaneously could make
them an important instrument for the investigation, diagnosis and treatment of cancer. The first products designed for use in this field have already been used successfully,
and a large number of new chips are now in the developmental
phase. It is becoming increasingly clear that use of DNA chips
could greatly improve the survival chances of patients with
many types of cancer, since it permits more precise adaptation
of treatment to factors influencing the disease in the individual
patient. However, successful use in this application presupposes
the availability of more specific treatment options.
When a tumour arises, the body increasingly loses control over
the ability of the affected cells to undergo cell division. The cells
change, the affected tissue starts to grow in an uncontrolled way
and measures taken by the body to check the growth of the cells
become less and less effective. Ultimately the cancer cells break
away from their tissue of origin to form metastases, i.e. secondary growths in other parts of the body. For the process to advance to this stage, a whole series of control mechanisms have to
be switched off, and this occurs mostly via changes in certain
genes. Almost a hundred such oncogenes are now known and
new ones are still being discovered. The products of these genes
generally occupy important positions in signalling pathways
that regulate cell growth and division. Just as important a role,
Important application:
cancer diagnosis
107
Policing life and death: p53
genetic damage
other signals
mdm2
+
activation and
synthesis
autoregulatory
loop
-
DNA repair
direct
involvement
p53
indirect
involvement
other
genes
other
functions
+
bax etc.
apoptosis
p21
cell cycle held

in G1 phase
The tumour suppressor gene p53 is one of the most important genes involved in the development of cancer. Its gene
product (P53, written with a capital P) plays a central role
in the growth and division of somatic cells. It is active especially when genetic damage is present in the cell, but it can
also be activated by external signals. It is known to have at
least three functions:
1. Control of the cell cycle. Cells divide in a regular pattern
of events known as the cell cycle. If the genetic material
of a cell is damaged, P53 holds the cycle in the G1 phase
a sort of resting phase in order to permit repair of
the DNA. The signal for this is transmitted via a number
of proteins including P21.
2. Apoptosis. If the damage to a cells DNA is too great,
P53 induces the cell to commit suicide. This process,
known as apoptosis, stops occurring in cancer cells,
with the result that they replicate in an uncontrolled
fashion. P53 appears to be able to initiate apoptosis in
various ways, including induction of the bax gene.
108
3. DNA repair. There is considerable evidence that P53 not

only allows the cell time to repair its DNA, but also plays
an active role in this process. Here again, it acts indirectly via p21 and other genes, however certain properties of P53 suggest that it also plays a direct role in the
repair process.
Because of its central role in the control of cell growth and
division, p53 is an important target for cancer therapies.
Attempts have been made, for example, to restore the function of altered P53 and to stimulate synthesis of this protein.
The latter objective can be achieved, for example, via the
mdm2 gene, which together with p53 participates in an
autoregulatory loop: P53 stimulates formation of Mdm2,
which in turn inhibits P53. In healthy cells this cycle stops
excessive amounts of P53 from preventing normal division
of cells. If the function of p53 is impaired, but not abolished,
by genetic changes, drugs that act against Mdm2 may be
useful, since in such cases more P53 is required in order to
keep cells under control.
The decision as to whether such a drug will be useful therefore depends on the genetic profile of the patient concerned. Techniques that can identify each of the variants of
p53 and of its genetic environment are therefore an important precondition for specific, rapidly-acting and effective
therapy. The p53 GeneChip offers this possibility. On this
chip are several thousand short DNA segments with which
the p53 gene variants can be identified. The illustration
shows the fluorescence pattern that results from such a
test.
however, is played by genes with the opposite effect, i.e. genes

whose function is to limit cell proliferation. Malfunction or
complete loss of such tumour suppressor genes opens the way
to the development of cancer.
The most important tumour suppressor is the

protein P53. This molecule polices the growth of
cells and can even force cells to commit suicide if
their genetic material is too severely damaged (see box on page
108). The importance of this protein for cancer therapy is apparent from the fact that its function is disturbed in more than
half of all human cancers. A whole series of drugs work by
attempting to restore correct functioning of this tumour suppressor. Depending on the reason why P53 is no longer functioning correctly, different drugs may be required. Therefore, if
the best treatment for a patient is to be found quickly, the genetic
variants present in that patient must be ascertained. And that
task can be made easier with the aid of a specially designed DNA
chip.
Chips designed to detect many other oncogenes and tumour
suppressor genes are being developed at present, and some are already on the market. The aim of all such developments is to draw
a genetic profile of the patient so as to assist doctors in the task
of deciding which form of treatment is likely to be of benefit, and
which not, in that particular patient. And this applies not just to
the choice of the right drug: since in many cases there is no sure
The forces of law

and order in cells
109
means of determining how dangerous a particular ulcer is, many

unnecessary operations are performed, while conversely many
ulcers adjudged to be harmless are later found to be malignant.
In these situations DNA chips can considerably increase the accuracy of diagnosis.
In fact, DNA chips can help not just when an ulcer

has already been found. Cancer arises mostly as a
result of an accumulation over decades of mutations that occur either randomly or as a result of radiation or
toxins. Nevertheless, the likelihood of developing a certain type of
cancer differs between individuals even between individuals
whose lifestyle and environmental circumstances are identical
because each individual has inherited a certain pattern of genetic
changes. Knowledge of this genetic predisposition can therefore
be very important: people who have inherited an increased risk
of developing skin cancer, for example, should be more rigorous
than others about avoiding exposure to sunlight and should undergo regular medical checkups. In future it will probably be
possible to test for many such predispositions using DNA chips.
Disease prevention is thus another potential application of this
technology.
Use in preventive
medicine
From the discipline of pharmacogenomics comes

another current example of successful use of a
DNA chip. This area of research is concerned with
the interactions between our genes and drugs (see
chapter on pharmacogenomics). It is based above all on the observation that the effectiveness of drugs varies greatly and that
in some individuals drugs have dangerous side effects. Most
such differences in reaction to drugs are at least partially due to
differences in our genes. If the genetic causes of such differences
can be ascertained, treatment can be adjusted accordingly. It
may even be possible to develop special drugs for people with
certain genetic characteristics. Such drugs would be expected to
act more specifically and thus be safer and more effective.
Cytochrome P450, for example, is important for the efficacy and
tolerability of many drugs. This is a family of enzymes whose
task it is to render water-insoluble substances including many
drugs water-soluble. Above all, molecules are prepared for excretion from the body in this way (see also chapter on pharma-
First pharmacogenomic
product:
the AmpliChip CYP450
110
cogenomics). As the funcThe AmpliChip CYP450

tion of cytochrome P450 enzymes varies from person to
The AmpliChip CYP 450
arrived on the market
person, drugs are broken
in 2003. The scientific
down more rapidly in some
basis of this chip is
individuals than in others
formed by pharmacoand their action in the body
genomic data on the
influence of the cytovaries accordingly.
chrome P450 gene famAt least 50 separate genes
ily on the efficacy and
and hundreds of gene varitolerability of drugs.
The AmpliChip CYP450 is able to identify the most imporants are now known to code
tant variants of two important members of this group of
for this family of enzymes,
genes.
and recently it has become
possible to detect the most
important variants of two
important members of this group of genes by means of a DNA
chip. The AmpliChip CYP450, which was developed jointly by
the healthcare company Roche and Affymetrix, manufacturer of
the GeneChip, is one of the first products developed on the
basis of pharmacogenomic knowledge to become commercially
available. The basis for the development of this product is
knowledge of the influence of cytochrome P450 on the metabolism of drugs.
As in the case of their use in relation to cytochrome P450, DNA chips will in future find uses
in many areas of diagnosis namely wherever
genes play a role in the genesis or development of a disease.
Four such areas can be identified:
z the metabolism i.e. the absorption, conversion and breakdown of drugs; the genes of the cytochrome P450 family of
enzymes fall into this category;
z the action of different genotypes in cancer, i.e. the genetic
changes that play a role in cancer; the variants of p53 fall into
this category;
z a group of genes that play a role in the reaction of the body
to infection;
z genes that influence individual susceptibility to pathogens;
the extremely rare cases in which a gene variant can prevent
HIV infection are a well known example of this.
In all these areas DNA chips can permit precise diagnosis of the
genetic basis of a disease. With increasing knowledge of the
Outlook: use of DNA chips

for diagnosis
111
genes concerned and of the molecular basis of disease, such

chips will in future contribute to earlier detection, more effective treatment and possibly even prevention of diseases.
Before this prospect can become reality, however, a number of
obstacles have to be overcome. For one thing, the vast majority
of presently available DNA chip-based tests are too expensive for
routine use. Also, the potential for individual further development of the method, for example in public or private research
institutes, is severely limited by patents. Furthermore, the
method still suffers from technical difficulties such as the question of whether the RNA used as a marker is measurable with a
sufficient degree of accuracy in blood, the test liquid that has
generally been employed to date. This is a precondition for what
appears at present to be the most promising medical application
of DNA chips, namely gene expression analysis.
It is nevertheless to be expected that in the next few years DNA
chips will assume an important role in general, and especially in
medicine.
Pharmacogenetics: DNA chips in diagnosis
There are at least four areas in which DNA chips have a potential role in
medical diagnosis: drug metabolism, genotypes of cancer, infection-related genes and susceptibility to pathogens. The most important genes
currently known to play a role in these areas are indicated.
112
References
Erlich H: Diagnostic applications of genomics. Talk given at Roche R&D Media Day,
Munich/Penzberg, April 2002
Pedrocchi M: Multiprobe array systems for the analysis of human genes
Certa U et al.: Biosensors in biomedical research: development and applications of gene
chips. Chimia 53: 5761, 1999
Sinclair B: Everything's great when it sits on a chip. The Scientist 13[11]:18, May 24, 1999
Baron D: Genomics und Proteomics mit Gen-Chips und Protein-Arrays. Pharmazeutische
Zeitung 31/2001
DNA Microarray Website of Leming, Shi: http://www.gene-chips.com/
Affymetrix Website: http://www.affymetrix.com
InformationsSekretariat Biotechnologie Website: Genexpressionsanalyse
http://www.i-s-b.org/wissen/broschuere/chip.htm
113
Basic conditions:
ethics, law and society
Every innovation has consequences,

and every opportunity holds risks.
And ambition always precedes ability.
These generalisations apply also to
personalised medicine, the road
to which involves not just medicine
and science, but also ethics, law
and society. Controversies await us,
but eventually there can be benefits
for all.
In the 1990s the biosciences displaced physics and chemistry

from the forefront of scientific innovation and change. Genetics, genomics, proteomics and related disciplines are generating
impulses that have already had a significant impact on our daily lives and will do so even more in the future. At the same time,
however, this new-found prominence has thrust the biosciences
into the limelight of public attention: as with all technical and
scientific innovations, their results have become topics of heated debate. At the centre of this debate are the changes presently taking place in clinical medicine the field in which the practical applications of current research are most clearly palpable
at present. The potential medical applications of these new disciplines are vast, but so too is the responsibility imposed by the
use of such techniques in humans.
In fact, the promise associated with molecular medicine can be
realised only if certain basic conditions are satisfied. This applies to many areas of public, and also to some important areas
of private, life. Political, economic and legal questions are important in this regard, as are ethical, societal and cultural considerations. Also, given the highly emotional debate currently
raging about the use of genetics and genomics in medicine,
there is a need to clear away a number of misconceptions which
at present are standing in the way of a more objective assessment of the value of these new technologies. The differing interests of the various parties involved give rise to a complex area
of tension, somewhere in the middle of which a broad social
consensus needs to be found. Only then can the great opportunities provided by a molecular understanding of diseases be
turned to practical benefit.
The first, and most important, of the various interest groups involved is that of patients. Without
their acceptance, no change is possible. Their expectations of medicine are based on the elementary need for
preservation or restoration of individual health a need that
medicine has traditionally sought to satisfy with conventional
remedies and is now attempting to satisfy with the fruits of molecular research. Newly acquired understanding of the molecular basis of diseases is being used to combat diseases earlier
and more effectively and in some cases to prevent them from
occurring in the first place. Molecular biology thus has the
potential to be of great practical benefit to patients.
Help and protection:

the interests of patients
116
Equally important to individual patients, however, is the price of

such progress not just in economic, but also in social and
moral, terms. And it is precisely this last point that concerns
many patients: they want to be sure that new understanding of

diseases and the new possibilities in medicine that result from
this understanding will not lead to injustices. The availability of
such options should not be limited to the rich, nor should anybody be restricted in terms of their choice of occupation, find
their individual freedom limited or be socially stigmatised on
the basis of test results.
The aim must therefore be to exploit the opportunities, while identifying and limiting the risks,
that arise from newly acquired medical knowledge. At the core of more personalised medicine, for example, is
the need to know and analyse the individual characteristics of
patients. Only on the basis of this knowledge is it possible, for
example, to develop medicines whose use is based on genetic
criteria and which therefore are markedly more effective and
better tolerated in certain patients than are currently used medicines (see chapter on pharmacogenomics). At the same time,
however, this knowledge inevitably reveals to the medical staff
involved, to the dispensing pharmacy, to the patients healthcare
insurance fund, etc. an intimate detail of the genome of the
patient concerned, namely the genetic variation upon which
prescription of the drug in question is based. There is nothing
fundamentally new about this: for example, the fact that a person gets regular prescriptions for insulin identifies that person
unmistakeably as a diabetic. Many other forms of treatment
likewise reveal peoples diseases.
Exploiting opportunities,
limiting risks
Basic conditions: ethics, law and society
117
The mere fact that information on the diseases present in an individual has to be acquired and to some extent passed on to third
parties is therefore no great problem in itself; much more prob-
lematic in this regard is the nature and quality of the information that is passed on, and in particular the possible predictive
value of the results of genetic tests. In practice, however, any predictions of this kind will in the vast majority of cases be limited
to a statement as to whether a particular patient is likely to respond to a particular drug or group of drugs. More detailed or
specific information is generally not to be expected from such
tests.
The situation is different, however, in the case of tests that can
determine a persons likelihood of developing certain rare
hereditary diseases long before the actual onset of illness. This
predictive ability is most certainly highly desirable in the case of
diseases for which drugs will in future be developed that can not
only cure the disease, but also prevent, or at least delay the onset of, the disease in individuals identified as being at risk.
Preventive medicine should, must and will become increasingly
important in the future. Nevertheless, in some cases, the implications of genetic testing for the patient as well as for family
members may be profound as for example in some rare familial
diseases, such as Huntingtons disease, cystic fibrosis or hemophilia.
In these cases, individuals and their families may perceive the resultant information as stigmatising.
118
Therefore, if the individual patient is to be convinced of the value of more personalised medicine
with all its implications, a series of conditions must be satisfied:
z New, better targeted drugs should represent a definite therapeutic advance, especially in the case of diseases whose treatment has hitherto been unsatisfactory or nonexistent. Personalised medicines should be more effective and better
tolerated than presently available drugs. These two requirements apply mostly to research.
z New treatments must be marketed at a realistic price, i.e. a
price that is reasonable and appropriate to the medical value
of the drug concerned, and must be available to patients via
their regular healthcare services.
z People must be protected against any form of discrimination
arising from genetic or molecular tests especially as the
terms health and illness have no absolute meaning (see
box). The legal basis for such protection must be established
by politicians in representation of society as a whole.
Varied requirements
Freedom of choice: focus on dignity

The debate about the ethical aspects of the future of medicine centres around two questions: preservation of the dignity of patients and protection from discrimination. The following points, among others, need to be considered:
z Freedom of choice. Patients must retain sovereignty
over their physical and mental wellbeing. This includes the
right to decide whether to receive a certain form of treatment or to undergo a certain diagnostic investigation, i.e.
the right to know or not to know. An available medical option may neither be forced upon, nor withheld from, a patient. However, such choices can only be free if the patient
has first received extensive information and non-directive
counselling.
z Limits to freedom of choice. The sovereignty of an individual ends where that of another begins. Since genetic
factors are inherited, the results of genetic tests are always
relevant to the close relatives of the person tested. The
question of whether, and if so in what cases, relatives
should have a say in a patients decision to undergo a
certain genetic test therefore needs to be clarified. In practice, however, this problem is probably limited almost exclusively to a small number of classical hereditary diseases,
and even in them it probably arises only exceptionally. In

all other medical situations genetic testing adds little to
what is already known from a familys medical history.
z Patients who are not competent to make decisions. Cases in which a patient is unable to assess the
arguments for and against undergoing a particular medical treatment are already covered by clear rules. Nevertheless, the possibility of prenatal diagnosis and therapy
is new in this respect. A legal framework for the performance of prenatal procedures needs to be established on
the basis of a social consensus.
z Discrimination. When told they have a certain disease,
the vast majority of people still feel a sense of stigmatisation and suffer anxiety about a possible loss of personal
freedom. A major rethink is required here. It is clear from
the complexity of our genome that terms such as health,
illness, normality and abnormality, have no absolute
meaning at the molecular level. Every human being carries
both promoting and protecting factors for all diseases.
Therefore, in almost every case, all that a molecular diagnostic test can do is indicate disease probabilities.
See also Roche Charter on Genetics.
119
z The individuals freedom of choice must be preserved in all
cases. This applies not only to choice of treatment but also,

and in particular, to diagnosis, i.e. to the right to know or not
to know. This requirement applies to the public and also to
politicians as representatives of the public.
z Out of these requirements comes another requirement: suc-
cessful introduction of new medical techniques will depend

to a large extent on considerably better education of the public regarding the potential, and also the limits, of genetics in
medicine.
The requirements of individual patients for customised medicine differ from the interests of society as a whole. In many cases the interests of
these two sides appear, at least at first glance, to be in direct conflict, e.g. in relation to the sharing of burdens and risks between
individual patients and society. Like the question of confidentiality of patient data, however, this is not a fundamentally new
issue, and in most industrialised countries the problem is regulated by means of a greater or lesser degree of redistribution of
medical costs from the individual to the public purse. In terms
of economics, therefore, molecular medicine is simply a therapeutic innovation whose costs and benefits have to be weighed
against each other and whose impact on the balance between
public and private funding of medicine needs to be considered.
More problematic, however, is the question of how acquired data
should be handled, i.e. how patients can be protected against
misuse of their personal genetic information, in particular. Current data protection legislation is mostly aimed at minimising
the amount of data that are collected, stored, distributed and
Society as a whole:
the public interest
120
analysed. Pharmacogenetically guided therapies will require the

collection of additional data. This will necessarily result in the
gathering of sensitive information about individual patients,
and this information will then be revealed indirectly to a broader circle of people via the medicines that are prescribed for those
patients.
The task of directing the use of sensitive personal data into the
appropriate channels falls to politicians in their role as representatives of society as a whole and the many special interest
groups within it. They have to mediate between conflicting requirements and interests and work to build the broad social consensus that will be required if the new possibilities offered by
molecular medicine are to be exploited in a responsible fashion.
In the crossfire of conflicting interests in healthcare, criticism is often levelled at industry and
commerce. This is due above all to the fact that
companies operating in the healthcare sector seek to make a
profit out of health, a treasured asset both of individuals and of
society. However, this sweeping criticism overlooks the central
role played by commercial enterprises in satisfying the requirements of the various interest groups. After all, the pharmaceutical and diagnostics industry strives to meet the need of patients
for better targeted and safer treatments, and at the same time
makes an important contribution to human society and culture
via the research that it undertakes. And because it is highly innovative, it has to accept a high degree of responsibility.
Nevertheless, the changes that are about to occur in medicine
certainly raise questions for the affected industrial sectors and
do so precisely because of the special responsibility that these
sectors bear towards patients. Because whether and at what price
a medicine can eventually be offered for sale depends on a num-
Between economics and

responsibility: commerce
121
ber of very complex factors. In the first place, the process of developing, testing and obtaining marketing authorisation for new
drugs is set to become increasingly protracted and expensive.
The research that underpins the development of new drugs is
also expensive. Even though the drug regulatory authorities of
many countries are presently working on ways of simplifying
drug registration procedures, at least for particularly important
and innovative new drugs, it is becoming increasingly difficult
for companies to recoup the development costs of new drugs.
Pharmaceutical companies are also subject to multiple constraints in relation to the pricing of new medicines firstly as a
result of their need to recoup the development costs of the drugs
concerned, and secondly as a result of the price controls that are
imposed in many countries. Therefore, if the development of
more personalised forms of therapy for necessarily smaller target
groups is to be made economically viable, it may be necessary to
adopt new patterns of thinking and devise new forms of cooperation between industry and healthcare funding authorities.
A key role in the restructuring of medicine that is

taking place at present is played by research. It
lays the foundations for more individualised prevention and treatment of diseases, and is therefore subject to specific demands. Of central importance in this
regard is the safety of treatments and the correctness of the diagnoses on which they are based. And the central problem here
is the unpredictability of nature.
New treatments need to be more effective or better tolerated
than presently available treatments. At least in certain diseases,
this objective can certainly be achieved with the aid of pharmacogenetics and molecular diagnosis. More of a danger are the
unrealistic expectations that can arise as a result of overenthusiasm for the new possibilities.It must be remembered that molecular medicine is no magic wand. Side effects of drugs will still
exist in the future, as will fluctuations in effect. And the disappointment felt by patients whose pharmacogenetic profile indicates that the only available form of treatment for their disease
is likely to be ineffective in them has plenty of potential to create new conflicts. Therefore, in order not to arouse unrealisable
hopes in the first place, medical science needs to be cautious in
what it says about its possibilities, which are and always will be
limited.
Research:
laying foundations and
informing the public
122
The same applies to diagnostics. The common perception that

the results of genetic tests, in particular, are immutable and invariably correct in terms of what they predict creates a whole series of problems even though this perception is incorrect in the
vast majority of cases. The fact that these tests, like all biological
test methods, are subject to error and that their predictive value
is limited by the current (always unsatisfactory) state of scien-
tific knowledge is axiomatic to scientists, but by no means obvious to patients. Here again, frank and objective education of the
public is required.
Genetic testing and molecular diagnostics therefore can and must be accompanied by better education of the public. Education alone, however, will not convince the public of the value and purpose of the new possibilities
in medicine. Of more use in this regard are instruments that
allow to define their potential risks and benefits.
Such instruments have in fact already been available for a long
time. This too illustrates the point that molecular and pharmacogenetic tests do not differ fundamentally from the standard
techniques used in medicine today: they too can be assessed in
terms of parameters such as sensitivity and specificity that provide an objective measure of the informational content of test
results.
For example, a genetic test whose purpose is to spare patients
from a life-threatening side effect of an important drug must
satisfy the following criteria:
1. Specificity: The test must reliably indicate that a person
found to have the observed gene variant(s) will develop the
side effect in question. The less specific the test (i.e. the high-
Probability, not certainty
123
er the proportion of bearers of the gene variant who do not

develop the side effect), the more patients will be unnecessarily deprived of the important drug.
2. Sensitivity: The test must identify as high a proportion as
possible of people who will develop the side effect. Patients
who develop the dangerous side effect even though their test
result was negative have not been helped by the test.
Most genetic tests are likely to have an informational content

similar to that of conventional laboratory tests (e.g. plasma cholesterol level). In the vast majority of cases, therefore, genetic
tests should be regarded and used in exactly the same way as
conventional tests. Like these, they can indicate probabilities
that may be quantifiable to some extent, but they cannot provide
certainty. Only exceptionally, e.g. in rare, classical hereditary
diseases such as Huntingtons chorea, do genetic tests yield results with a predictive value of almost a hundred percent.
However, even an objective evaluation of tests on the basis of
their informational content can serve only as an aid to the decision on whether a particular molecular or genetic diagnostic test
should be performed in a particular patient. Ultimately, such
decisions must be made by patients themselves in consultation
with experienced physicians, and the welfare of the patient must
always be the primary consideration. Only in this way is it possible to ensure that advances in medicine bring benefit and do
not cause harm.
124
References
Schreiber H-P: Humangenomforschung, Gentechnik und Gesellschaft
Schreiber H-P: Human-Genom-Forschung und die Notwendigkeit eines sozialvertrglichen 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: 835838,
2002
Genetics in Discovery and Development. Roches Ethical Principles including Roche
Charter on Genetics. Roche, 2000
Roche Biomarker Program: Economic impact of genetics, 2004. Internal document
125
Prospects:
more knowledge for medical science
Medicine is changing. New techniques and knowledge derived from

genetics, genomics and proteomics
permit a deeper understanding of the
molecular causes of disease. Increasingly, medicine will focus on differences between patients and become
more personalised. It will become
ever clearer that no two illnesses are
the same and that even useful forms
of treatment are not effective in all
patients.
Medical science is in the grip of change. Genomics, proteomics

and other branches of molecular biology are generating a stream
of new findings, and modern technology has introduced techniques of miniaturisation, automation and parallelism into research and development. The entirely new field of molecular
diagnostics promises to have a lasting impact on therapeutic
practice. And medical science is increasingly realising that apparently identical clinical pictures can have entirely different
underlying causes requiring personalised treatment.
These developments also have a commercial side, pitting established pharmaceutical companies against young biotech firms in
a race to discover suitable target molecules and new drug compounds. At the same time, the development of new drugs up to
the stage of regulatory approval is becoming lengthier and more
expensive. Traditional drug research is growing riskier in economic terms, and it is becoming more difficult for it to contribute
to genuinely significant innovations. On top of this, despite
some minor successes, the options available for treating many of
the major common diseases remain unsatisfactory. A period of
radical change is imminent.
Behind this upheaval is the recognition that no

two illnesses are the same. It is now clear that with
the exception of a handful of hereditary diseases
and some severe infections, very few human diseases have a simple or even a single cause. And even in the exceptions just mentioned, which include, for example, cystic fibrosis and hemophilia as well as tuberculosis and AIDS, the severity of the
symptoms varies so much from one patient to the next that a
clinical picture of some complexity has to be assumed. After
decades of genetic research and several years of genomic investigation, we now know that a patients genetic predisposition
plays a significant role in the progression of almost all illnesses.
In the case of infections, another factor can be the variable genetic makeup of the pathogens involved.
These findings are neither new nor surprising. And yet they confront medical science with a daunting problem. Until now the
principle of one disease one treatment has essentially held
sway unchallenged. But if no two illnesses are the same, many of
the treatments used must be wrong or at least inappropriate.
Though fine diagnostic distinctions have always been a driving
force of medical progress, the sheer amount of newly acquired
No two illnesses
are the same
128
knowledge is now enormous. This calls for a rethink in many

cases. The indications for existing drugs will become narrower,
and the discovery of new drugs will be all the more crucial. And
distinguishing between subtle variants of a disease instead of
general clinical pictures will require a new molecular diagnostic
approach.
In fact, for the first time in the history of medical

science diagnostics and therapy are meeting on
common ground at the molecular level. Whereas drug therapies have always acted on the molecular network of
our bodies, the diagnoses upon which those therapies are based
have generally been made on the basis of physiological factors as
evidenced by physical signs and symptoms. Thus, the physician
observes the consequences of a disease in order to treat its causes an approach that fails to do justice to the true complexity of
the vast majority of diseases.
New biological test methods have now made it possible to routinely determine the actual molecular causes of diseases at the
patients bedside. The newness of this development is indicated
by the word routine. Of course, the molecular background of
diseases has long been investigated in patients. But until now
these research methods have been too expensive and elaborate
for routine use. Today that has all changed thanks to the widespread availability of PCR, especially its quantitative form, and
increasingly sophisticated DNA chips. In the coming years these
techniques will complement, and in some cases even supplant,
conventional diagnostics in many areas of medicine a process
that is expected to build up a huge momentum of its own: the
more widely PCR and DNA chips are used, the cheaper and
more varied they will become; this in turn will accelerate their
spread.
Molecular diagnostics is therefore playing a key role in the rapid
advances now taking place in medicine. It has a broad range of
applications:
z New therapies: The use of new drugs that take account of the
complex causes of diseases calls for differentiated molecularbased diagnostics.
z Screening: Screening tests are already an important instrument in healthcare. Cancers, abnormal blood sugar, abnormal bone density and abnormal blood pressure are all currently screened for. As our knowledge deepens, more such
Consequence:
a new role for diagnostics
Prospects: more knowledge for medical science
129
target
sequence
DNA strand
double helix
coiled
DNA
chromosome
PCR makes it possible to amplify specific DNA sequences.
tests will be developed to cover other common diseases such

as osteoarthritis, schizophrenia and epilepsy and infectious
diseases such as hepatitis C and tuberculosis. Molecular diagnostics based on PCR, DNA chips and other techniques has
the potential to significantly increase the accuracy, and thus
the acceptance, of screening tests. This, in turn, will drive forward their development and encourage widespread use of
them.
z Gene tests: Genetic predisposition is an important factor in
the prognosis and treatment of diseases. Gene tests already
make routine use of PCR and to an increasing extent DNA
chips as well.
z Hygiene: A marginal aspect, but one that promises to gain
importance in the coming years, is the search for and identification of pathogens in hospitals. Again, modern techniques
such as PCR and DNA chips are the fastest methods for detecting the presence of pathogens before they can spread.
The recognition that diseases can have entirely

different causes despite producing the same
symptoms is not new. What is new is the molecular biological understanding that now makes it possible to examine the genetic differences between individual patients and
the effects of these differences on treatment. In other words, no
No two treatments
are the same
130
two treatments are the same. A drug might be right for one patient but wrong for another, even though both patients have the
same illness, because drugs can vary in their efficacy and tolerability in different individuals. The field of pharmacogenetics
has been investigating the reasons underlying this phenomenon
for over a hundred years now, but only recently have molecular
genetic techniques made it possible to apply these insights to
clinical medicine. Pharmacogenetics is now threatening to upset the second half of the dogma of one disease one treatment.
In future the choice of the right treatment will depend not just
on the disease diagnosed, but also on the way in which each patients body deals with the drugs in question. To make this kind
of choice possible, two closely related factors need to be taken
into account:
z Genetic factors: Pharmacogenetics is concerned with the relationship between the gene variations and the bodys
response to drugs. Genetic differences can cause drugs to be
absorbed, metabolised or excreted too rapidly or too slowly.
Or they can prevent sufficient drug from reaching the target
site. Or they can give rise to adverse or even dangerous side
effects. Ruling out such genetically caused uncertainties
relating to the efficacy and safety of drugs will be one of the
major challenges facing pharmaceutical researchers in the
coming decades.
z Environmental factors: External factors are at least as important as genetic factors in determining the efficacy and safety
of drugs. Prominent among these factors is diet. Elements of
our diet can interact with drugs, accelerating or preventing
their uptake and affecting their excretion and utilisation. The
same applies to interactions between different drugs, which
can enhance or reduce each others effects and exacerbate
each others side effects. External stress factors such as physical and mental fitness, environmental toxins, radiation,
temperature and so forth can also influence the efficacy and
safety of drugs. In practice, the environmental influences to
which a patient is exposed cannot be exhaustively determined; also, they vary over time though this means that they
can be influenced. This is not true of gene variants. It is therefore all the more imperative to recognise how environmental
factors influence the way the body interacts with drugs.
131
If future therapies are to be based on genetic factors, medicine will inevitably become more personalised. However, the term personal in this
context does not mean that at some time in the
future patients will have their own tailor-made therapy. Rather,
it means that a far broader range of therapeutic options will be
offered from which doctors can select the one most suited to
their individual patients. Of course, such choices are already
available, at least for some diseases, however the number of such
choices will increase and so too hopefully will the success of
therapy. As an inevitable consequence of this development, the
target groups for drugs will become smaller. The indications for
new drugs will be determined not only by the molecular causes
of the diseases being treated, but also by the pharmacogenetic
profile of the individual patients. This is unexplored territory in
pharmacology.
In future, therefore, patients will be able to expect that a drug
that is prescribed for them is more likely to be truly suited to
them than at present. The effects of almost all currently used
drugs can vary to a greater or lesser extent, and in extreme cases
lack of efficacy is even the rule (see box). The safety of many currently used drugs is similarly unsatisfactory; for example, some
three patients per thousand die each year from severe side effects
of major drugs. This figure needs to be reduced, since even the
occasional occurrence of severe side effects can be acceptable
only if the disease concerned is relatively rare and unresearched
and therapeutic options and experience are correspondingly
limited.
Consequence:
the personalisation of
medicine
Great fluctuations: efficacy of drugs
drug group
poor efficacy
AT2 antagonists
10 25%
SSRIs
10 25%
ACE inhibitors
10 30%
Beta-blockers
15 25%
Tricyclic antidepressants 20 50%
132
Statins
30 70%
Beta 2 antagonists
40 70%
The efficacy of drugs is often unsatisfactory and may fluctuate in the extreme. The table gives the incidence of poor
efficacy for several major drug groups.
Angiotensin II (AT2) antagonists, like ACE (angiotensinconverting enzyme) inhibitors, are antihypertensive drugs
which are widely used in the treatment of heart failure. The
same applies to beta-blockers. Selective serotonin reuptake inhibitors (SSRIs) are used as psychotropic drugs,
especially in obsessive-compulsive disorders and, like tricyclic antidepressants, for depression. Statins, also known
as HMG-CoA reductase inhibitors, lower cholesterol levels.
Beta 2 antagonists are important antiasthma drugs.
Doctors responsibilities will grow accordingly. They will have to

deal with entirely new diagnostic resources, a considerably expanded range of therapies and as is already evident from the
growth of the Internet far better informed and more self-confident patients.
This means that the demands made on medical

science will increase. One innovation gives rise to
another. Personalised therapies require individual diagnoses. Molecular diagnoses call for differentiated
therapy. And both aspects, diagnosis and therapy, depend on
rapidly expanding technological possibilities. In fact, a synthesis is taking place at the moment: research and development,
diagnosis and therapy, information and prevention are evolving together. The key to successful healthcare lies in integrated
medicine.
If the new possibilities of medical science really are to bring
about progress, they must mesh smoothly. The concept of diagnosis will need to be extended beyond symptoms and clinical
findings to include the molecular underpinnings of diseases and
their treatment. Also to be considered is the hitherto relatively
undeveloped field of prevention, which in most cases is still limited to fresh air and a healthy diet. Testing, i.e. diagnosis, of genetic predisposition will play a far greater role here in future. It
will also make it possible to provide patients with more specific
counselling such as is already available, for example, in relation to high serum cholesterol levels.
Treatment follows seamlessly. The earlier a disorder is discovered, the easier it is to treat a long-recognised fact that can take
on new relevance in connection with the possibilities of early
molecular diagnosis. This is especially true when specific diagnosis is matched by a corresponding range of personalised
therapies. Progress will be achieved only if both sides move forward together.
The interplay of these developments requires a high degree of
coordination, information and above all cooperation. At the
same time, these developments will give rise to new ethical,
societal and legal challenges and issues.
Consequence:
integrated healthcare
133
For the pharmaceutical industry, these developments impose the need for a continuous rethink.
A new order will prevail in the healthcare market,
where the changes are already in full swing. New
strategies, alliances and competitions are emerging:
1. Integration of diagnosis and treatment: The more finely differences between individuals are distinguished and considered, the more difficult it is to separate these two poles. Close
cooperation is required here: drugs whose prescription depends upon pharmacogenomic considerations will be prescribable only if a corresponding means of testing is available.
Thus, a specific genetic variation first has to be identified in
the patient so that a drug geared to this variation can be sensibly used. And because the development of diagnostic tests
and therapy are to some extent interdependent, companies
with expertise in both these areas will find themselves at an advantage. Expertise therefore needs to be gathered together
either within a single company or else by means of close alliances between companies. The traditional boundaries between
diagnostics and therapy will therefore largely disappear.
2. Greater development risk: The fact that the available options
for treating most of the major common diseases are still unsatisfactory means one thing above all: Pharmaceutical companies will have to be more willing to take the risks associated with the development of new drugs with new mechanisms
of action. Certainly, the future will continue to hold the occasional surprise, as when a well-established drug is found to
possess previously unsuspected beneficial properties. But for
the most part, medical progress will depend on the exploration of new avenues particularly via new target molecules, which are already the most hotly contested objects in
medical research. Above all, new diagnoses, new targets and
new drug groups mean considerably stepped-up research
and development efforts with an undiminished risk of failure. Nevertheless, the effort may well be worthwhile. Successful developments that address unmet medical needs have
a huge sales potential.
3. Smaller target groups: The advent of more personalised
medical care inevitably means that a new drug can only be
sensibly used in a limited number of patients. This limits sales
possibilities, thus making it more difficult to recover the costs
of research and development. However, the development of
such drugs also has advantages. For example, drugs developed
Consequence:
upheaval in the pharmaceutical industry
134
in this way are more effective thanks to their targeted activity.

This should reduce the risk of failure in later stages of development while increasing acceptance among patients and
thereby reducing the number of patients who stop treatment.
The actual investment-to-yield ratio can be very attractive.
4. Competition by biotechnology: Young biotechnology companies backed by considerable venture capital are invading
the market for new drugs developed on the basis of molecular findings. Typically these companies develop drugs characterised by a high risk but huge sales potential. The pioneers
of the sector, e.g. the American companies Amgen and Genentech, have long been on an equal competitive footing with
the traditional pharmaceutical companies, which for their
part have almost without exception risen to the new challenges either by establishing their own biotechnology departments or entering into alliances or acquiring promising
innovative companies in the sector.
5. Increased demands: New opportunities bring new responsibilities. In the not-too-distant future, pharmacogenetic data
will certainly form part of the data required by health and
drug regulatory authorities. In addition, after a period of adjustment, patients are likely to become more demanding in
terms of the efficacy and safety of the drugs they take.
Internationally active healthcare companies will

not escape this trend. On the contrary, active participation in this process of change is fundamental to their survival, whereby the term change does not imply a revolution, but
rather a systematic evolution towards more informative investigations and more effective and safer drugs. The fact that many
years of laborious and detailed research work are required before personalised diagnostic tests and drugs can be developed
is evidence enough of the evolutionary nature of this change.
Also, in many cases a distinction will have to be made between
what is feasible and what is reasonable, desirable and economically sound. For instance, the size of a patient group above which
the development of drugs specifically for it becomes economically viable still cannot be predicted at least, not until new
forms of cooperation between society and industry, such as orphan disease programmes for particularly rare diseases, have
been set up.
High hurdles, high goal
135
Nevertheless, progress is opening up far-reaching new opportunities for medicine at the scientific and technical levels. Personalised diagnosis and treatment promise to be substantially more
effective with substantially fewer side effects. At the same time
they can tackle the causes of diseases whose treatment has until
now been only symptomatic and often inadequate. Notwithstanding all the commercial and ethical imponderables, in a certain sense the new possibilities also impose a moral obligation
to apply the new findings of molecular medicine for the practical benefit of patients.
136
A brief glossary of terms
fi
Amino acids the chemical building blocks of proteins. They

consist of a constant region, containing an amino group and an
acid group, and a variable region. At least 20 different amino
acids occur in nature.
Antibodies Y-shaped proteins that are part of our immune system. They bind to foreign substances in the body and thereby
mark them for destruction.
Antioxidants molecules that trap dangerous, highly reactive
oxygen compounds in the body and thereby render them harmless. Vitamins C and E, for example, are antioxidants.
Apoptosis programmed cell death. Cells whose genetic material is irrevocably damaged or altered commit suicide in order
to protect the rest of the body from the effects of the genetic
alteration.
Autosomal dominant inheritance an inherited characteristic that is expressed if it is present on either one of the two
autosomes of a particular kind.
Autosomal recessive inheritance an inherited characteristic that is expressed only if it is present on both autosomes
of a particular kind.
Autosomes chromosomes not involved in sex determination.
Humans have two of each kind, inherited from the mother and
the father respectively. Altogether there are 44 autosomes
(twice 22).
Bases chemical substances that have a basic (alkaline) action.
The bases of DNA are the fundamental building blocks of
the genome: adenine (A), thymine (T), guanine (G) and cytosine (C). When present on two strands of DNA, the bases join
to form stable pairs. In nature, base pairs form only between A
and T and between G and C. In RNA, thymine is replaced by
uracil, which likewise pairs with adenine.
Biochip a solid substrate (e.g. glass or plastic) upon which biomolecules are anchored.
Bioinformatics the in most cases computer-assisted analysis
of biological data by special databases, applications and programs.
cDNA complementary DNA; DNA transcribed enzymatically
from RNA (mostly mRNA).
Cell the smallest independently viable unit of an organism.
Chromosomes tightly packed DNA strands with associated
proteins that are present in the cell nucleus and that function
as bearers of genetic information. The human genome consists
of 23 chromosome pairs (22 autosomes and one of the two
sex chromosomes X and Y).
Complementary DNA The building blocks of DNA and
RNA form specific pairs. Two strands whose building blocks
form a sequence of perfect pairs are able to form a stable double strand and are referred to as complementary strands.
Denaturing a process induced by heat or chemicals in which
a biomolecule (e.g. DNA, RNA or a protein) loses its
natural form.
DNA deoxyribonucleic acid, the chemical substance of which
our genetic material consists.
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
DNA chip a biochip with single-stranded DNA as the probe.

Enzyme a biological catalyst, generally a protein, that can
accelerate and combine certain chemical reactions.
Exon a sequence of a gene that acts as a direct blueprint for
a gene product.
Gene array a special type of DNA chip.
GeneChip a widely used DNA chip developed by the US
company Affymetrix.
Genes functional segments of our genetic material that serve
mostly as blueprints for the synthesis of proteins.
Genetics the study of inheritance; deals with the laws of inheritance and the properties of genes, including the transmission of specific variants of a gene from one generation to the
next.
Genome the totality of the genetic material ( genes) of an
organism.
Genome chip a special kind of DNA chip.
Genomics the systematic study of the form, function and interactions of the genes that comprise the human genome.
Genotype the alternative forms (alleles) of a gene present in
an individual; generally there is a maximum of two one from
the father and one from the mother.
High-throughput screening a highly automated method of
identifying potential drugs in chemical libraries.
Hybridisation the joining of two complementary DNA (or
RNA) strands to form a double strand.
Intron a sequence of DNA situated between the exons of a
gene that is cut out of the corresponding mRNA before
this is translated into the gene product.
Metabolism the transformation of chemical substances in the
body or within a cell.
Microarray a widely-used synonym for DNA chip.
mRNA messenger RNA, the working copy of a gene that
acts as a blueprint for the synthesis of proteins. Unlike
DNA, it is able to leave the cell nucleus.
Nucleic acids generic chemical term for DNA and RNA;
chain-shaped molecules whose individual building blocks are
bases ( nucleotides).
Nucleotides the building blocks of DNA and RNA; they
comprise the four bases adenine, thymine, cytosine and guanine (A, T, C, G; in RNA thymine is replaced by uracil [U]),
a sugar and at least one phosphate group; without the phosphate group these building blocks are referred to as nucleosides.
Oligonucleotides short nucleic acid chains composed of at
most a few dozen building blocks ( nucleotides).
Oncogene a gene that plays a role in the development of
cancer.
Pharmacodynamics the study of the interactions between
drugs and their molecular targets.
Pharmacogenetics describes the influence of gene variations
in individuals on the efficacy and side effects of drugs.
Pharmacogenomics studies interactions between drugs and
the genome.
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
A brief glossary of terms
137
Pharmacokinetics the study of the uptake, conversion and

breakdown of drugs in the body over time. Environmental factors, diet and genetic predisposition all play a role.
Phenotype the constitution of a living creature that results from
its genotype and environmental influences.
Polymerase chain reaction (PCR) a technique for rapid
copying (amplification) of even minute amounts of DNA.
Polymerases
enzymes that link individual nucleotides
together to form long DNA or RNA chains.
Polymorphism existence in more than one form; in genetics, a region of DNA in which differences in the sequence of
building blocks occur in a relatively large number of people.
Primer a short DNA fragment with a defined sequence that
serves as an attachment and extension point for polymerases.
Promoter a region of DNA immediately before a gene that
contains the starting information for transcription of that gene.
Protein a molecule consisting of a chain of amino acids.
Because of the variety of their building blocks, proteins can differ greatly in form and function.
Proteome the totality of the proteins of an organism.
Proteomics the study of the form, function and interactions of
all the proteins of a tissue or organism.
Rational drug design computer-assisted design of new
drugs.
RNA ribonucleic acid; the chemical substance of which, among
other things, working copies of genes ( mRNA) consist.
Sequence the order of the nucleotides in
DNA (DNA
sequence) or RNA (RNA sequence).
Sex-linked inheritance an inherited characteristic transmitted via one of the two sex chromosomes (X or Y).
SNPs single nucleotide polymorphisms differences in individual building blocks (base pairs) of DNA that are distributed
randomly over the genome and passed from generation to
generation.
Targets the molecules, mostly proteins, upon which drugs
act in our body.
Template in molecular biology, mostly a fragment of DNA
that acts as a chemical template for polymerases.
Transgenic animals animals containing genes derived from
other species.
Tumour suppressor a molecule which, when functioning correctly, prevents cancer from developing.
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
138
fi
fi
fi
Cover picture
Conceptual computer artwork. Part of a DNA molecule, chromosomes, a DNA
autoradiogram and the triplets of nucleotide bases that code for amino acids
in a protein.
Source: Mehau Kulyk, Science Photo Library
Published by
F. Hoffmann-La Roche Ltd
Corporate Communications
CH-4070 Basel, Switzerland
2007
Third edition
All trademarks mentioned enjoy legal protection.
Any part of this work may be reproduced, but the source should be cited in full.
This brochure is published in German (original language) and English.
English translation: David Playfair

Layout:
Atelier Urs & Thomas Dillier, Basel
Printers:
Gremper AG, Basel
7000632-2
140

Genes and Health - Roche

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Genes and Health - Roche

Hochgeladen von

Copyright:

Verfügbare Formate

We Innovate Healthcare

Genes and Health

Genes and Health

If it were not for the great variability among individuals,

Grouped around this term are a multiplicity of modern research

Molecular medicine: genetics,

In its search for the causes of disease,

It is a gradual revolution that has been going on for hundreds of

directed towards a single goal, that of more precise medicine.

Accompanying the revolution: Morgagni and Virchow

Rudolf Virchow (18211902)

Every disease is influenced by a larger or smaller

Target: the molecular network of the cell

action, of the genes; they receive signals from the environment,

The changing role of biology in medicine

The tasks of biology in medicine have changed greatly in the

and in this sense are themselves an important environmental

Genetics, genomics and proteomics thus provide

This is because disease-inducing environmental

The central importance

Monogenic hereditary diseases: cystic fibrosis

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

inheritance. The pattern of occurrence and non-occurrence

The complex interplay of genetic factors in disease

causes of complex diseases can generally be determined only

development of a disease (or, for example, in intolerance of a

necessary to take preventive medicines. Early prevention is

Diagnosis and therapy

Triple influence of genes: hepatitis C

however, scientists are searching at other sites, namely in the

One new way of improving therapy is therefore to increase

A drug may work well in one person,

A first example of a genetically specific drug has

specific antibodies linked to a dye. Second, there is a gene test

Pharmacogenomics: genes and drug response

of methods and knowledge gained from unravelling the human

When drugs do not work: pharmacogenetics

Our genome influences the effects of drugs at at least three

do not act directly on the cause of a disease, but rather on its

Pharmacogenomics: genes and drug response

Genes have been found for each of these three

Hurdles for drugs: genetic polymorphisms

excretion of drugs. As a result, the drugs do not remain

Personalised medicine: the objectives of pharmacogenomics

Genome and environment

The old slogan one size fits all still applies to

Old hurdles for new drugs

Pharmacogenomics: genes and drug response

z Are there too many differences in the target molecule be-

tween individuals, and is it possible to target drugs at more

To understand the molecular causes of a disease

SNPs yield the first

Once such SNPs have been found, the associated

Databases to the rescue

The final link in the pharmacogenetic research

Pharmacogenomics: genes and drug response

DNA. The binding process generates a fluorescent pulse which

Thus the requirements for more individualised

Potential and obstacles

Pharmacogenomics: genes and drug response