The Protein Universe

December 16, 2003

Tags: RNAProtein Function

CAMBRIDGE, Mass. — Like any other birth, the birth of a protein is a remarkable event. The fledgling molecule is a chain of amino acids linked together like a string of beads. It slips from the ribosome—its womb in the cell—and in a series of quick steps, folds into an exquisitely complex 3-D shape perfectly suited to its role in biology. As it jostles through the cell, it links up with other proteins to create the cellular machinery that produces life.

The story of life and its associated processes takes place within a vast universe of proteins and their interactions. Proteins are the cell’s workhorses, active in nearly everything cells do. They control cell structure, storage, signaling, movement, and defense. As enzymes, they control chemical reactions. As hormones, they control growth, development and even mood.

Usually, proteins rapidly become perfectly shaped entities ready for normal behavior. But in some cases, they fold into abnormal shapes, wreaking havoc in the body. Indeed, dysfunctional proteins are the root cause of most genetic diseases. They can lose their ability to recognize protein partners or fail to engage in necessary reactions. In these cases, they degrade and the cell might suffer from a loss of its vital activities.

Rogue, misfolded proteins can clump together and cause problems in the cell, as seen in patients with disabling illnesses like Alzheimer’s and Parkinson’s disease. Understanding how proteins function under both normal and diseased states is critically important to life sciences research.

In the wake of genomics, the science of proteins is a bountiful frontier. Genomics has provided the celebrated “blueprint of life,” but in and of itself, that blueprint is just a catalogue of anonymous gene sequences, explains Matthias Mann, a professor of bioinformatics who studies proteins at the University of Southern Denmark Scientists at research centers such as Whitehead Institute, Harvard University, the Scripps Research Institute and others now are assessing the biological function of these genes by studying the proteins they generate.

It is largely via protein research, scientists say, that the genome’s role in life, disease, and evolution will be elucidated. But studying entire proteomes—meaning the complete diversity of proteins in individuals—is a monumental challenge. The proteomes of higher species comprise tens of thousands of protein conformations and millions of untold molecular interactions. Nevertheless, scientists are taking on the challenge. But is this challenge beyond the scope of modern science?

“We will absolutely be able to do this,” says Whitehead Director Susan Lindquist. “There’s no question it’s going to take some time. But in the next decade we’re going to see some profound changes.”

Within the Fold

Lindquist is among the world’s leading experts on protein folding. In most cases, she emphasizes, protein folding is a vital part of normal biology. Her research has revealed the key role a group of proteins called “chaperones” plays in getting other proteins to fold correctly. Because they are dedicated to this task, protein chaperones are indispensable to maintaining good health. “Probably half of all diseases are caused by protein folding problems,” Lindquist says. “A few of the very big diseases are clearly induced by misfolding—Alzheimer’s, cystic fibrosis and Parkinson’s disease, for instance. It’s also clear that certain cancers develop because proteins don’t fold properly. They lose their regulatory functions—some proteins that are supposed to stop cell growth become disabled and others that stimulate cell growth take off their brakes.”

In a surprising finding, Lindquist’s research has shown that protein folding also plays an important evolutionary role. Much of her work in this area addresses a protein called Hsp90. This chaperone helps other proteins fold correctly when exposed to stressful conditions that might induce them to misfold in harmful ways. More important, it also helps a special class of very unstable proteins fold at normal temperatures. These proteins are known as “signal transducers.” They are meant to be unstable because that helps them to be highly sensitive to signals for growth control and development. By exposing fruit flies to stress or drugs, her group has been able to test the effects of extra demands on Hsp90 function. This causes flies to morph into a number of bizarre forms.

“We saw strange eyes, wings and legs in a few individuals,” Lindquist recalls. When the malformed insects were bred together and selected to have the same unusual development in subsequent generations, they eventually maintained their forms even when normal Hsp90 function was restored. “We deduced that taxing Hsp90 function reveals hidden variation in the genome,” Lindquist explains. Hsp90 normally suppresses genetic changes that can alter body shape. When Hsp90 levels are dramatically reduced—say, in response to overwhelming stress—these genetic changes occur and malformations begin to appear. Some scientists now attribute sudden evolutionary changes in the fossil record—so-called “punctuated” evolution—to the release of hidden genetic variations by corresponding drops in protein chaperones’ response to environmental stress.

In addition to using fruit flies, Lindquist performs many initial studies in yeast before turning to animal models. Like mammalian cells, yeast cells are eukaryotic, meaning they have a membrane-covered nucleus and many other biological features of “higher” cells. Prion proteins cause deadly diseases when they misfold in human brains, but when this occurs in yeast cells, Lindquist notes, the effects aren’t toxic.

While using yeast cells as a living test tube to study prion folding mechanisms, she and her postdoc Jiyan Ma, now an assistant professor of molecular and cellular biochemistry at Ohio State University, stumbled upon an interesting finding: When the prion proteins appeared in the interior of the cells, they sometimes folded into a highly unusual conformation known to be associated with human illnesses. Could this internalized prion also cause mammalian disease?

To find out, Lindquist and Ma generated a transgenic mouse that expressed the same interior prion. The mouse became sick, suggesting that some human neurological illnesses arise from a breakdown in the “quality control” machinery that normally degrades misfolded proteins from inside the cell.

“This mouse is showing us that when prions remain [inside the cell], they exist there in a highly toxic form,” Lindquist explains. “Does this finding have relevance for all prion diseases? I don’t know. But I do think the findings give us a new way to think about these diseases.”

What’s more, she adds, the research establishes something the scientists have suspected all along. “Protein folding problems are very ancient and the mechanisms cells use to cope with them are fundamental to biology,” Lindquist notes. “That means we can use simple organisms to gain important insights into much more complex ones.”

Proteins and Disease

In many cases, protein investigations drive the search for new ways to prevent disease and treat patients. For instance, Whitehead Member Harvey Lodish is examining the role that a protein hormone called erythropoietin, or Epo, plays in controlling red cell formation. Epo is secreted by the kidneys when the level of oxygen in the blood drops and more red blood cells are needed—as happens with severe bleeding. The hormone binds to a receptor, identified several years ago in the Lodish lab, that is located on the surface of bone marrow cells called erythroid progenitors. Typically, these cells die soon after they are formed. But in response to Epo binding, the progenitors undergo a series of divisions and differentiate into red blood cells. His work with mice has shown that the Epo receptor activates a number of complex signals that together prevent progenitor cells from dying and stimulate their division.

Recently, scientists discovered that administration of Epo also prevents the death of brain cells that normally follows a stroke or blow to the head. Lodish wants to know if the “anti-death” signals induced by Epo in nerve cells are similar to those in the red cell progenitors. This property of Epo, if harnessed, could lead to new methods to limit brain damage caused by neurodegenerative diseases.

Lodish’s research also may have an impact on treatment for patients undergoing bone marrow transplants for cancers. After the procedure, donor cells in the marrow can attack a patient’s organs and tissues, causing debilitating side effects. Scientists long have hoped to reduce the incidence of side effects by using the patient’s own hematopoietic stem cells, adult stem cells that produce all the blood and immune cells in the body. But the cells are rare and often located with other, cancerous cells. Producing them in sufficient quantities for therapy is challenging.

Lodish’s goal is to ease the burden of locating and purifying hematopoietic stem cells. Toward this end, he recently identified a surface protein called endoglin that is abundant on these stem cells and helps distinguish them from other cells in bone marrow and blood. Endoglin, also necessary for the growth of blood vessels, “flags” hematopoietic stem cells, making them easier to purify. The protein may help clinicians spot stem cells during cultivation, Lodish says, perhaps enabling them to produce enough for effective transplants.

The Emergence of Proteomics

Research such as this represents the cutting edge of an established, hypothesis-driven approach to protein chemistry. However, a new force—proteomics—also is beginning to shape the field’s future, propelled by advances in a high-throughput instrument called the mass spectrometer. This device, used by British researcher Sir J. J. Thomson in his 1897 discovery of the electron, measures the mass of individual molecules as they are converted to electrically charged particles called ions. Biologists use the instrument to sequence biomolecules, including proteins, and to identify their locations. This once was a cumbersome task: Sequences were read by painstakingly feeding molecules into a spectrometer one piece at a time. Today, scientists can identify proteins simply by matching amino acid fragments to nucleotide sequences contained in a variety of genomic databases.

Mass spectrometry has become so powerful, connections among literally thousands of molecular components can be assessed rapidly. With this systems-level view, scientists can obtain highly sophisticated perspectives on the cellular changes associated with disease. The technology has fostered the creation of a new field called systems biology, which combines the work of biologists, computer scientists, engineers, and other scientists to decipher the working relationship among proteins, genes, DNA and all the other biological elements in a cell that, when put together, produce living creatures. Growing numbers of Whitehead researchers are beginning to think deeply about systems biology, among them Member Richard Young.

In research published last fall in the journal Science, Young combined a new technique with microarray technology. He and his colleagues conducted their studies on baker’s yeast, which has a cellular structure similar to human cells. The technique they developed, with the aid of computer scientists and engineers, allows them to locate regulatory proteins across an entire genome, something never done before. The result is a picture of how the genome is regulated to produce a living cell.

Today, a host of companies are hopping on the proteomics/systems-biology bandwagon, looking for venture capital and funding. But researchers caution that it will take time for systems biology to yield any major new findings. The field is in its infancy, an era of massive data collection. High-throughput methods applied to genomics, proteomics, and other related fields are producing data at a rate that far exceeds analytical capabilities.

“Right now, systems biology is a very trendy thing; it’s very exciting and justifiably so,” says Whitehead’s Lindquist. “But it also provides a very broad overview. Sometimes you can’t see the forest through the trees, so you need to narrow your focus with hypothesis-based research. I think the real power will come from going back and forth between these two approaches.”

Prescription RNA

Whitehead scientists also are employing a new method to assess protein function that many believe could revolutionize biology. This technique, called RNA interference (RNAi), enables gene expression—and thereby protein chemistry—to be selectively controlled. In RNAi experiments, tiny molecules called “short interference RNAs” bind with the targeted messenger RNAs that carry a gene’s protein building instructions. The RNA binding effectively blocks the gene’s activity, so that targeted proteins are never formed. RNAi has the advantage of being fast and efficient – a sharp contrast with gene-knockout methods that can take months or even years to develop. “RNAi is extremely popular now,” Lodish says. “We use it routinely in our studies. You can use it to look at effects in whole animals or in cultured cells. You can also screen thousands of RNAi’s, but that’s very tedious; it will require robotics. But that’s a direction in which a number of our projects may go.”

Ultimately, scientists are confident that proteomic research will produce clinical advances. Especially promising opportunities lie in the diagnostic arena. The more scientists learn about how proteins function in healthy and diseased states, the better they get at identifying proteins that predict the course of a given illness. Such proteins, also called “biomarkers,” will someday be easily identified on the basis of clinical screening. Armed with proteomic data, clinicians will have better opportunities to select the best course of treatment.

Proteomics may enhance drug development efforts, but experts caution that at least five to 10 years will pass before the technology yields new, marketable products. Experimental methods are still under development, experts say, and the process of ushering new drugs through the Food and Drug Administration is time consuming.

Nevertheless, scientists are optimistic that new therapeutic opportunities await. For instance, it may someday be possible to develop drugs that repair deformed proteins as a means of curing a patient. These would represent a whole new class of drug targets. Most drugs on the market today inhibit enzymes involved in disease processes. Drugs that target folding problems could greatly expand the clinician’s arsenal. But designing protein-fixing medicines is challenging, in part because scientists often don’t know the actual 3-dimensional shape of a protein, let alone how to restore it to some target conformation. Revealing protein structure is difficult, particularly for large molecules that can be thousands of amino acids in length.

According to University of Southern Denmark’s Mann, misfolded proteins often exist in a flexible, random structure that is particularly hard to elucidate, especially for drug targeting. “Normally, we design small molecules that simply bind with proteins and inhibit what they do. Here, we’re asking for something completely different: We’re asking the drug to bind with the protein and change it from one structure to another. That’s a lot to ask. Nevertheless, many neurodegenerative diseases act by these folding pathways. So, this research is very cutting edge and has great promise for the future, even though it’s very speculative.”

Conformational changes also are the bane of computer modelers trying to develop complex models of cell behavior. These efforts require that biochemical processes be reduced to a series of mathematical equations, so that cell changes resulting from a given stimulus can be predicted and quantified.

The participation of computer scientists in these endeavors reflects the multidisciplinary nature of modern protein research. Today, computer scientists and other information technologists work side by side with biologists in the laboratory. In many instances, the academic lines are blurred—computer scientists become biologists and vice versa. It’s a marriage of necessity: Proteomics is generating data at a rate that far exceeds the analytical capacity of the mere mortal. Sophisticated computer algorithms are necessary to wade through it all. And the arrangement is by no means one sided. Just as information technology enables biology, so does the latter enable the former. Researchers now are using electronic circuits to manipulate DNA and control protein expression.

Lindquist also is wading into the information technology arena. In a recent project, she began to construct tiny wires for nanoscale electronics out of a protein derived from yeast. The self-assembling protein forms fibers that can be coated with mixtures of silver and gold. The resulting metal wires are as narrow as the span of a few dozen atoms. The most ancient structures in biology therefore serve as the backbone for some of the most advanced electronics currently in existence.

Clearly, proteins have much to teach us. And the future will provide many years of opportunity to study the lessons they offer.

Journalist Charlie Schmidt writes frequently about genomics and molecular biology. He is a recipient of the 2003 National Association of Science Writer’s Science in Society Journalism Award for magazine writing. He lives in Portland, Maine.


Communications and Public Affairs
Phone: 617-258-6851

Whitehead Institute is a world-renowned non-profit research institution dedicated to improving human health through basic biomedical research.
Wholly independent in its governance, finances, and research programs, Whitehead shares a close affiliation with Massachusetts Institute of Technology
through its faculty, who hold joint MIT appointments.

© Whitehead Institute for Biomedical Research              455 Main Street          Cambridge, MA 02142