Leroy Hood expounds the principles, practice
and future of
systems biology
Stephen L. Carney
Drug Discovery Today 2003, 8:436-438
Text only, + thumbnails, + full figures, PDF
Publications by
Stephen L. Carney
Leroy Hood, President, Institute for Systems
Biology
For the benefit of our readers who might not
be familiar with the idea, could you briefly define the concept of
systems biology?
Systems biology is the ability to look at
all of the elements in a biological system - by elements, I mean genes,
messenger
RNA, proteins, protein interactions and so
forth - and to measure their relationships to one another as the system
functions
in response to biological or genetic
perturbations. Then one can attempt to model the behaviour after
integrating the
different levels of information, either
graphically or mathematically, so that, ultimately, you will be able to
describe the
behaviour of the system given any kind of
perturbation. In the future, we will be able to redesign systems by
modification
or drugs to have completely new systems
properties. What distinguishes systems biology from the more classical
biology
of the past 35 years or so, which looked at
genes and proteins one at a time, is the attempt to look at all, or at
least most,
of the elements and their
interrelationships.
What, in your opinion, are the essential
elements that underpin and drive the development of a systems biology
department?
The key elements are: first, to create a
cross-disciplinary faculty environment where one has physicists,
computer
scientists, engineers, chemists, biologists
and mathematicians all working together. The challenge is to create a
language
that can be commonly shared to allow them to
communicate with one another with the aim of creating a focus on
particular
systems problems. The second challenge would
be to create an environment that has all of the high-throughput
platforms
that one needs for gathering biological
information. In genomics, that would be large-scale DNA sequencing, DNA
arrays
and large-scale genotyping. In proteomics,
that would be the ability to identify and quantify proteins, to look at
their
interactions and measure their
compartmentalization and chemical modification, and so on. Metabolomics
is the ability to
look at the small molecules that operate in
the context of systems. A third requirement would be the development of
new
global technologies and powerful new
computational tools for gathering, classifying, analyzing, integrating
and, ultimately,
modeling biological information. That is one
of the reasons you need a cross-disciplinary environment. In all cases,
the
systems biology itself should drive the
nature and strategy of tool development, be it technical or be it
computational.
Another element that is essential for doing
good systems biology is being able to partner effectively with academia
and,
perhaps equally importantly, with industry.
This is because one scientific entity is not going to be able to invent
all the
technologies or all the computational tools,
nor will they have access to all the biology that one can explore. So
you need
partnerships to be able to facilitate the
integration of technology, computation and biology. Another challenge is
how you
can then keep these new tools that have been
developed and mature them into high-throughput platforms that can be
used
to inexpensively generate enormous amounts
of data. Finally, there are issues of how you can integrate what I call
systems biology, which is hypothesis-driven,
iterative and integrative - a cyclical kind of process - with discovery
science,
which is identifying all the elements in a
transcriptome or the proteome of a particular cell type. Discovery
information is
useful for disease stratification
(classification) and provides the elements for beginning systems
biology. I would draw a
sharp distinction between performing array
analyses and doing systems biology; systems biology has to be
necessarily
integrative of many different types of
biological information. So, in the end, I think the challenge for
systems biology is how
the academic or industrial scientific
entities can effectively integrate technology with computation, biology
and, ultimately,
medicine.
If you were to highlight only one success
that has arisen from systems biology, what would it be and why?
Together with Eric Davidson at Caltech and
Hamid Bolouri here at the Institute for Systems Biology, we have used
systems approaches to define the gene
regulatory network that controls the development of the endoderm of the
sea
urchin. This is a 55 gene network. We know
the linkages and the interrelationships in some detail. From that
network you
can begin to make predictions about how you
could change development if you made modifications to the network. To
give
you one example of engineering the system,
Eric and his colleagues perturbed the network in a defined manner and
quite
predictably generated an organism with not
one gut, but two. This is a graphical illustration of how, if you
rationally
understand the gene regulatory networks that
control development, you can redesign those networks to get completely
new
systems properties. You can predict the
behaviour of a network, given a particular perturbation in the system.
This could
never have been done in a million years with
the classical, one-gene one-protein-at-a-time, approach to understanding
gene
regulatory networks.
What would you consider the role for systems
biology within the pharmaceutical industry? How would you
envisage this approach driving the discovery
of new pharmaceutical agents?
Systems approaches to human systems (immune,
cardiovascular, cancer, and so on) have the chance to enormously
facilitate the process of target selection
and the various aspects of pharmacogenomics; that is, the ability to
identify
adverse side effects or individuals that
genetically do not react effectively to drugs. My feeling is that
systems biology is
going to usher in a new kind of medicine,
which I call predictive, preventive and personalized medicine. The
predictive will
deal with, and be able to identify, hundreds
if not thousands of variant genes that may predispose, in particular
combinations, to various late-onset diseases
(e.g. heart disease, cancer, autoimmune disease, and so on). It would
therefore be possible to actually write out
a probabilistic health history for the individual. Disease prevention
will use
systems biology to place defective genes in
the context of the networks in which they operate. As I described with
the sea
urchin, it would then be possible to
circumvent the genetic limitations with newly designed drugs, or
modified proteins or
genes, and thus be able to say 'a prediction
for you is that you have a 70% chance of getting breast cancer when you
are
60 years of age, but if at the age of 40,
you start taking this drug, you need never get breast cancer, because we
can
circumvent whatever limitations the genes
might create.'
I'll say parenthetically that disease arises
either as a consequence of gene defects or a combination of gene defects
and/or
pathologic, environmental cues, and we can
learn to deal with the circumvention of pathologic, environmental cues
in
exactly the same way using systems biology.
My feeling is that systems biology will be a central strategy for
discovering,
initially, therapeutic drugs and,
eventually, preventive drugs.
In the future, the question is, how
effective big pharma will be in utilizing systems biology? It requires a
lot of integration
that they are not very well set up to do.
One possible alternative scenario is that a few successful systems
biology
companies could emerge and focus on the
early stage processes of drug target discovery and drug identification,
stratification of disease and
pharmacogenetics.
The outcome for drug discovery is that you
will probably identify targets that are significantly more complex
than have been the case in the past. How
likely is it that medicinal chemistry will be able to deliver selective
tools that can act upon multiple, possibly
diverse, targets?
I do not think that the targets will be more
complex, rather they will be more effectively chosen. I think what the
pharmaceutical industry has objected to is
the concept that any given disease, such as prostate cancer, is almost
certainly
stratified into a multiplicity of different
diseases caused by differing combinations of genetic and environmental
factors,
even if their initial phenotypic outcome is
similar. For example, prostate cancer may be considered as a disease
that can
be stratified on the basis of systems
approaches into a number of disorders that we will treat in different
ways. The tools
of discovery and the tools of systems
biology will be ideal for this process of stratification. By stratifying
disease entities
and putting them individually through
clinical trials, you increase enormously the probability that the drugs
undergoing trial
are actually going to work on a significant
fraction of patients. Consider a hypothetical disease in which there are
10
different genetic predispositions giving
similar phenotypes. Each of these predispositions accounts for 10% of
the total, and
you have one drug that works perfectly in
one of them. The resultant global efficacy of 10% would not appear to be
a
particularly effective clinical trial. With
the systems approach, we can stratify disease and look at the individual
types of
disease in the context of which drugs work.
Then we can conduct clinical trials on the patient population predicted
to be
responsive to the drug. In this case, the
success rate will be 100%. I think we can enormously increase the
efficiency of
defining drugs that are going to be
effective on a very high proportion of selective stratified populations.
Now the argument
that has to be made is that for each of
these approaches, we're going to cut down the total cost of creating the
drug
because the markets will be smaller than for
the current blockbuster drugs, and because the disease is going to be
highly
stratified and the market fragmented.
I think that is consistent with what we can
expect from systems biology. Microfluidics and nanotechnology are going
to
enormously decrease the cost of doing the
global and integrative studies that underlie both systems biology and
the
multiparameter analyses of predictive
medicine. For example, in the future we will sequence an individual
genome for under
US$1000, compared with today's cost of maybe
US$50 million or more. This will give you an idea of the scaling factors
that we can think about as we move these new
technologies first towards the discovery and then towards targeted ways
for designing drugs and ways to circumvent
gene defects and environmental cues.
With the exception of Eli Lilly's US$140m
investment in a Centre for Systems Biology in Singapore, there seems
little big pharma involvement at present -
do you think they are waiting for the results of the Lilly foray or do
you think the drug discovery industry is not
receptive to the concept of systems biology at the moment?
I think big pharma in general has a wait and
see attitude about new approaches, as they should. The pharma business
is
not to invent new approaches, but rather to
apply successful approaches. The key issue for them is to determine when
a
new approach should be applied. Systems
biology is a nascent discipline and the most interesting systems in
which we
have demonstrated its power are in microbes,
yeast and sea urchins, and not humans, to date. There is scepticism
about
systems biology, possibly not whether it's
going to be fruitful but how long it will take to get to a point where
it will be
useful to a drug company. But having said
that, many of the executives in the drug industry don't understand
systems
biology. They have no idea about its
potential but you've got to persuade the executives that it deserves the
investment. I
am on Lilly's Advisory Board and their
Centre for Systems Biology is a marvellously interesting experiment. I
think the
challenge for the pharmaceutical industry is
going to be the challenge of the integrations necessary for systems
biology;
that is, the integration of new technologies
with new computational tools with hypothesis-driven systems biology all
focused on medicine.
In the end, the important word in systems
biology is biology. The biology has to drive the development of the
computational
tools. It also has to drive the development
of the new technologies. It should also be noted that many academics are
sceptical of systems biology. Some sceptics
feel systems biology really isn't anything new beyond the integrative
physiology that has been practiced for
years. This was exactly the same argument we heard in 1985 and 1986 from
the
National Institutes of Health when they said
they didn't need a Human Genome Project; they were already spending
US$300,000,000 a year on genetics and that
was the same as the Human Genome Project. At the time, they didn't
understand how profoundly different those
things really were.
In retrospect we can see that the arguments
of doing things the same old way have always been a barrier to advancing
new opportunities. The pharmaceutical
industry widely believes that genomics really hasn't advanced their
cause very
much. The reason for that is totally
understandable. Genomics alone is a single dimension of information.
Systems biology
looks at many dimensions of biological
information. Genomics alone is good for classification; it isn't
necessarily good for
understanding biology and ultimately
providing powerful approaches to drug discovery. I think many in pharma
and
academia do not understand the difference
between a one-dimensional view of information that you see with DNA
arrays
and the global, integrative, iterative and
hypothesis-driven approach of systems biology.
Freeman Dyson said that if you really want
to transform the field of science, invent a new technology. You
clearly have a history in this with DNA and
protein sequencers and synthesizers - what would you see as being
a key enabling technology in the next 10
years?
I think without a doubt it's the synthesis
of microfluidics, microelectronics and nanotechnology for integrating,
automating
and creating high-throughput production of
tools for data production and analysis. It is nanotech we will use to
sequence
individual DNA molecules; it is nanotech we
will use to measure RNAs and protein concentrations and protein
interactions.
Microfluidics is the way nanotechnology
communicates with the outside world and, in the future, the two will be
beautifully
integrated together. My very strong feeling,
is that every analytic technique in biology and medicine will be
transformed by
microfluidics and nanotechnology.
What would you see as being the next
milestone achievement that might be delivered by a systems biology
approach?
I think it's going to be a series of
incremental steps, rather than one big achievement. It will require the
development of
novel technologies, powerful new
computational tools, as well as their integrated application to the
problems of biology and
medicine. One of the biggest challenges in
systems biology is how to develop computational tools for integrating
the many
different types of biological information,
DNA, RNA, protein, protein interactions and the phenotypes, that are
required by
systems biology approaches. The ultimate aim
will be to develop graphically integrated models that can be converted
into
mathematical descriptions of systems. I
think that these computational challenges are some of the biggest we
face at the
Institute for Systems Biology. We are
incrementally moving towards success, but it isn't going to be one big
development
like the DNA sequencer; it's going to be an
incremental series of integrative plug-in algorithms to digital and
network
platforms. I think developing the
computational tools to handle, analyze integrate and model information
is, however, one of
the very biggest challenges of the day.
Leroy Hood
The Institute for Systems Biology
1441 North 34th Street
Seattle, WA 98103-8904, USA
tel: +1 206 732 1200
fax: +1 206 732 1299
e-mail: [EMAIL PROTECTED]
Leroy Hood is recognized as one of the
world's leading scientists in molecular biotechnology and genomics. In
2000, Hood
co-founded, and is currently President of,
the Institute for Systems Biology in Seattle
(http://www.systemsbiology.org),
which pioneers systems approaches to biology
and medicine. He earned an MD from Johns Hopkins University in 1964
and a PhD in Biochemistry from the
California Institute of Technology in 1968. His professional career
began at Caltech
where he and his colleagues pioneered four
instruments, sequencers and synthesizers for DNA and protein. Hood was
also
one of the first advocates of, and a key
player in, the Human Genome Project. In 1992, he moved to the University
of
Washington to create the cross-disciplinary
Department of Molecular Biotechnology as the William Gates III Professor
of
Biomedical Science. He has played a role in
founding numerous biotechnology companies, including Amgen, Applied
Biosystems, Systemix, Darwin, Rosetta and
MacroGenics. In a distinguished career, Hood has been honoured many
times, most notably receiving the 1987
Lasker Award for his studies on the mechanism of immune diversity and
the 2002
Kyoto Prize in Advanced Technology. He has a
life-long commitment to making science accessible and understandable to
the general public, especially children. One
of his foremost goals is bringing hands-on, inquiry-based science to
K-12
classrooms.
Copyright
� 2003 Elsevier Science Ltd. All rights
reserved.
_______________________________________________
www.e-laser.org
[EMAIL PROTECTED]
