What the Parts Can Teach Us About the Whole

Molecular biology may be a great legacy of the 20th century, but with the human genome mapped, researchers now need a bigger-picture view.

Abstract

Systems biology tackles a challenge of grand proportions: Can we ever know all the pathways of disease? With astonishing speed, this new brand of science is ‘well on its way,’ says one researcher, to producing major breakthroughs in science and medicine that could lead to lower-cost treatments and better diagnosis.

This year, the Nobel Prize in chemistry went to Roger Kornberg, PhD, of Stanford University School of Medicine, for his work in explaining the mechanisms of molecular transcription — otherwise known as the copying of DNA into RNA.

Kornberg’s research, which shed light on how cells express information in the human genome, was first published in 2001. As far as some in the emerging field of systems biology are concerned, however, the intervening five years might as well have been a lifetime.

“What we’re trying to do sits at the intersection of biology, math, and physical sciences,” says John Aitchison, PhD, associate director of the Institute for Systems Biology. Researchers essentially attempt to simulate complex biological systems with the help of computational models.

PHOTOGRAPH BY RICK DAHMS

“As the Human Genome Project neared completion, the need for this field became apparent,” says Karin Rodland, PhD, a researcher at Pacific Northwest National Laboratory (PNNL). A U.S. Department of Energy facility in Richland, Wash., PNNL is among a small number of its kind that are developing a new type of science known as systems biology. “[Systems biology] is starting one step beyond genomics and going three or four steps beyond,” Rodland says. “The rate of change is incredibly fast and getting faster.”

So what exactly is this curiosity called systems biology? Even at the Seattle-based Institute for Systems Biology (ISB), the answer is not quite clear. “You might get a different definition of systems biology from different people at ISB,” says John Aitchison, PhD, the institute’s associate director.

For Aitchison, systems biology involves the integration of biology, technology, and computation. “What we try to do sits at the intersection of biology, math, and physical sciences,” he says. Researchers essentially attempt to simulate complex biological systems with the help of two different types of computational models. “You can look at molecules and try to quantify them as much as possible, but you inevitably run into technological problems.”

Top-down modeling is a means of understanding different global data sets and what these sets contain, such as networks of interaction among different types of molecules, genes, or cells. Bottom-up modeling involves building relationships with smaller models and working within the dynamics of small networks. “These, you can simulate,” Aitchison says.

The Human Genome Project started with DNA, building what Karin Rodland, PhD, a researcher at Pacific Northwest National Laboratory, calls a “storage catalog” — essentially, a repository of all information stored in the body. “Systems biology started with the parts catalog and then went to the parts list (RNA), and then to the inventory of what’s there right now (proteins),” she says.

“I would argue that the two (models) go hand-in-hand,” Aitchison believes, with “the biology driving technology.” Technology leads to new types of data, and analysis of the data leads to new questions in biology. It is all part of a cycle, says Aitchison, with each aspect defining and pushing the next.

“I think of it as an attempt to understand a whole system,” says Douglas Sheeley, ScD, program manager for the National Center of Research Resources’ division for biomedical technology, in Bethesda, Md. NCRR is part of the National Institutes of Health. “It’s about truly understanding how a whole system works,” he says, and if not the entire system, then at least broad segments to see how individual parts interact with one another.

“A lot of effort goes into defining what you are looking at so you can ask the right questions,” Sheeley says. “What are the constants of the biology? What kind of analytical tools do you need?

As an example, he notes that there are about 6,000 different proteins that could be expressed in yeast. “That’s a daunting technological challenge.”

The PNNL has been awarded a Biomedical Technology Research grant from NCRR. These centers, primarily in academic settings, serve two purposes: development of cutting-edge technologies critical to biomedical research, and access to those technologies for the research community through service, training, and dissemination, including technology transfer. The latter is more difficult to achieve, according to Sheeley.

“In some cases, the technology is something that can’t move, and you’ve simply got to provide the broadest access possible,” Sheeley says. Otherwise, dissemination could mean replicating the software and methodology elsewhere.

In its role as a supporter of the development of new biomedical research tools, NCRR is funding the National Resource for Cell Analysis and Modeling at the University of Connecticut Health Center, in Farmington, Conn. The program is developing software called the Virtual Cell, an Internet-based application that helps scientists convert biological information into mathematical equations, constructing testable models.

Other grantees are searching for clues about how proteins and carbohydrates interact, and are conducting research on biological markers for diseases and effective drug therapies.

LOFTY GOALS

Sheeley divides systems biology into three components: computational functions, analytical functions, and biological expertise. The analytical part could include microarrays, mass spectrometry, flow cytometry-based methods, or other high-throughput techniques so that researchers can examine many things at once. “The constraints of all three of these domains influence each other,” he says.

The computational part of systems biology serves not only as an organizing tool, but also as a means of modeling the systems in order to test hypotheses. The actual biology then comes into play.

Sheeley believes that research centers doing systems biology need to bring aboard individuals who are proficient in all three areas — not always an easy task. The NIH and others are investing in “cross training” at graduate schools, looking for more brainpower in bioinformatics.

According to Rodland, systems biology is opening the door for a new specialty called computational biology. “It’s creating an entirely new breed of scientist,” she says. Computational biologists work closely with traditional biologists in the laboratory, but to apply their new advances to human biology, Rodland says, more clinicians need to learn about the field. NIH is investing in an infrastructure for clinical trials resulting from advances in systems biology.

What distinguishes systems biology from other forms of research is the attempt to capture environmental stimuli — for example, a disease state — on a number of different levels.

“We’re trying to figure out how [each phenotype] relates to various underlying network levels,” says Aitchison, who also holds adjunct professorships at the University of Alberta and the University of British Columbia, along with a faculty affiliation at the University of Washington. “If you can understand how network phenotypes can lead to a biological state, you should be able to understand how troubled networks lead to a disease state.”

Systems biologists, therefore, try to relate the phenotype to molecules associated with a particular phenomenon by examining complex networks. “It requires really new ways of thinking about biology,” Aitchison says. “The ultimate goal is a mathematical model of biological systems,” he says. “That’s a very lofty goal, but there are a lot of potential benefits along the way.”

Some of this activity is way out on the cutting edge. “Systems biology, it’s fair to say, is in its infancy,” Aitchison says.

“It’s definitely an emerging field,” Rodland states, but she seizes on Aitchison’s analogy to express the view that systems biology is well on its way to producing some major breakthroughs in science and medicine. “Maybe it’s no longer a toddler and more of a preschooler now,” she jokes.

MORE THAN THE SUM OF ITS PARTS

An example of the potential of systems biology is detailed in a recent issue of the Proceedings of the National Academy of Sciences (Ippolito 2006). Federally funded researchers at the Center for Genome Sciences at Washington University, in St. Louis, applied DNA microarray technology to measure how metabolites and metabolite byproducts might affect metabolism in cancer cells.

According to the research, the test’s developer was able to uncover evidence that certain aggressive neuroendocrine tumors respond to three specific neurotransmitters. This led to the discovery that a previously untried combination of three existing drugs showed promise in fighting prostate tumors in mice.

The study also found that advanced computing algorithms sorted through genetic profiles of some 400 human cancers to find that genes present in neuroendocrine tumors also were prevalent in some other types of cancer. This, according to the researchers, gives hope that the drug combination could have wide application. Because the drugs already are on the market, this systems biology-enabled research could lead to new, lower-cost treatments.

Aitchison says that systems biology also could be useful in early diagnostics because microarrays can display so much information, thanks to advances in computing technology. “These data can be used for network discovery, and then to understand how this information can be used in interventions,” he says.

According to Rodland, systems biology combines different pieces of data to predict the likelihood of an individual being stricken with type 2 diabetes or developing a tumor, for example. “Is it imprinted in the gene?” she asks. “Will such a condition be of concern to the next generation?” Advances might lead to diagnostic tests for biologic markers that can predict various disease states.

Rodland believes that systems biology requires a shift away from the traditional reductionist research approach, because the science of genetics has opened up the potential for more holistic studies. “It’s clear, if you’re taking a look at the whole organism, that the whole is a lot more complicated than the sum of the parts,” she says.

“It’s not just a catch phrase to call it ‘21st century biology’,” Rodland says. Molecular biology may be a great legacy of the 20th century, but today’s researchers need more of the big-picture viewpoint. “We know that the reductionist method isn’t working,” Rodland states.

“A major part of what systems biology tries to do is map out the networks,” Rodland adds, “and map out the flow of information.” Put in layman’s terms, she says, “What systems biology is trying to do is come up with a troubleshooting manual to predict the consequences of messing with the system.”

For example, pancreatic cancer is hard to detect because its symptoms are similar to those of pancreatitis or even indigestion. “Usually when it’s discovered, it’s late in the game,” she says. Similarly, there is no good early detection method for ovarian cancer, and patients are often vague in reporting symptoms, such as complaining of abdominal pain or gastrointestinal problems.

Researchers are now applying systems biology when looking for biomarkers that might lead to simple tests to detect these specific types of cancer, or to determine if an individual is genetically predisposed to these conditions. Such tests could also give clinicians better tools to select the right treatments to attack a tumor, or be part of an early stage of a biotech drug discovery.

“This field always seems tantalizingly close to a payoff, but we’re not there yet,” says Sheeley.

NEW FRONTIER

Although diagnostic testing might represent the near-term promise of advanced technology, Aitchison refers to this application more as a spinoff of global biology than true systems biology. He also believes that genomics is more an example of global biology than systems biology.

“The genome provides only a blueprint because it tells us nothing about life,” Aitchison says. “Global approaches are like high-throughput stamp collecting. Systems biology tries to analyze the interaction.”

The Human Genome Project started with DNA, building what Rodland calls a “storage catalog” —essentially, a repository of all information stored in the body. “Systems biology started with the parts catalog and then went to the parts list (RNA), and then to the inventory of what’s there right now (proteins),” Rodland explains.

Investigators working on the next generation of scientific research are looking to see which of the parts are actually expressed in an organ or a system.

“The new frontier looks at the activity of the inventory,” Rodland says. Are the proteins active, and if so, how? What is the consequence of proteins being active? Will the cell grow, reproduce, or die? Was the cell damaged? Ordinarily, damaged cells should die, thanks to a built-in suicide mechanism.

“If you don’t turn on the [cell’s] suicide mechanism, you get cancer, says Rodland.

Aitchison says that the ISB’s mantra is “P-4,” which stands for predictive, preventive, personalized, and participatory medicine. By understanding conditions, researchers can personalize treatments.

“As technology advances and becomes cheaper, individuals could do their own testing,” Aitchison surmises.

This is where the ethics get tricky. Privacy advocates have rung the alarm about the potential for genomic testing to become a means for health insurers to deny coverage to patients predisposed for serious conditions. “It’s something that policy makers eventually will have to address,” Aitchison stresses.

The field is wide open, the questions innumerable, the possibilities intriguing. The future promises to be fun.