Clever Approach May Provide New Clues to Drug Design
Circumventing a long-standing problem in drug design, scientists have developed a novel way to identify a new class of protein building blocks that could serve as valuable leads for drug development.
The new method, called mirror-image phage display, represents an important advance in the rapidly growing field of drug-design. It will also offer new insights into the structure and function of important proteins. The method, described in the March 29 issue of Science , was developed in the laboratory of Peter S. Kim of the Howard Hughes Medical Institute (HHMI) and the Whitehead Institute for Biomedical Research.
The technique, developed by Kim, first author Ton Schumacher, and their colleagues, takes advantage of the fact that all proteins and enzymes can function in two mirror-image forms. Proteins that occur in nature are made out of L-amino acids. Their alter-personalities, made of oppositely configured D-amino acids, are exact mirror images of the natural forms and have the opposite orientation.
The L- and the D-forms are like a pair of baseball gloves with identical catching abilities but opposite handedness. However, the D-forms are different from the naturally occurring L-forms in one key respect: While the L-forms are degraded by digestive enzymes and elicit a vigorous response from the immune system which makes them poor drug candidates, the D-form is neither biodegraded nor processed by the human immune system. Peptides that are not degraded can turn out to be useful drugs cyclosporin, a drug used to suppress the immune systems of kidney transplant patients, is one example.
"Our approach was to take advantage of these mirror image properties to find D-forms that bind to natural proteins," said Kim. Understanding the value of this approach, said Kim, requires a quick course in drug design. All drugs work by binding to target or disease proteins, and drug design experts believe that knowing the precise structure of the "target" protein--the lock--will allow them to find the right agent--a key that can turn off the protein's function.
Until recently, searching for this key has been a guessing game involving a trial-and-error search of suspected peptides. In recent years, scientists worldwide have created a number of collections, or libraries, of genetically encoded protein fragments. These libraries and the development of sophisticated techniques to screen them has made the search for key peptides easier and more efficient. This searching method, however, continues to suffer from a major drawback. "Any peptide that we identify by this method will be a naturally occurring one, made out of L-amino acids, and therefore makes a lousy drug," said Kim. "On the other hand, simply making a D-form of this peptide would be like trying to fit a left-handed baseball glove to a right hand."
The method developed by the Whitehead-HHMI team circumvents this problem, allowing researchers to use the tremendous advantage of genetic libraries while still avoiding the disadvantages of L-amino acid peptides. In this method, the scientists begin by chemically synthesizing D-forms of the target proteins and using them to screen the peptide libraries. The result of this screening provides researchers with the L-form of a peptide that binds well to the D-form of the target protein. The researchers then make the mirror-image D-form of the L-peptide, which binds to the naturally occuring protein--the protein that was the initial target. "So what we have done is to identify peptides made out of D-amino acids that bind to natural protein targets," said Schumacher. The scientists used this approach to identify a peptide that interacts with the SH3 domain, a protein module that is found in a variety of important molecules inside cells.
Kim said that currently their ability to synthesize D-forms of proteins is limited to proteins that contain about 100 or less amino acids (a D-form of the HIV protease, which consists of 99 amino acids, is one example). Kim anticipates that as the ability to synthesize proteins improves, so will the utility of the mirror-image phage display method. He adds that this method can also be adapted to DNA and RNA libraries which also could yield new leads of biological and medical importance.