Researchers in the US have developed a new way of making synthetic antibodies (“synbodies”) that offers a cheap, high throughput alternative to traditional methods, and may also open the door to new diagnostic tools and treatments.
Compared to conventional methods that start with a pathogen, injects it into an animal and then waits to see what kind of natural antibodies develop, the Arizona State University researchers’ method works “backwards”: they synthesized antibodies from random peptides and then looked for the proteins they might have a high affinity for.
You can read about how they developed the new method in a paper published 19 May online in the journal PLoS ONE.
Antibodies are how the immune system defends the organism from attack by disease pathogens. These highly specific molecules recognize, target and neutralize each type of invader. And when a new pathogen comes along, the immune system attempts to develop a new antibody to deal with it, sometimes successfully (the organism is sick for a while then recovers), and sometimes unsuccessfully (the organism dies).
Antibodies are useful for research too, but they are time-consuming, difficult and expensive to make, since we have to wait for a lab animal’s immune system to make it. The process starts with a target protein, which is then injected into the animal whose immune system hopefully responds by making antibodies. The antibodies, or the cells that produce them, are then extracted.
Stephen Albert Johnston and Chris Diehnelt of Arizona State University’s Biodesign’s Center for Innovations in Medicine, and colleagues, developed a way to make antibodies that works in reverse: you take amino acid sequences (peptides) and link them together to make a synthetic antibody or “synbody” that binds to one or more molecules in the human “proteome”, the vast repository of protein molecules that occur in human beings.
Johnston told the media that instead of starting with the target protein, they “turn the whole process on its head, making the antibody chemically, then finding out what it’s an antibody to”.
The researchers said synbodies have the potential not only to target proteins of disease-causing pathogens, but also to be a useful tool for research into new diagnostics and treatments.
They started with a 20-unit random sequence of amino acids, and linked them up to make a peptide, rather like stringing beads onto a necklace.
They then took two peptide “necklaces” and joined them together using a “chemical scaffold” to make a binding molecule or “ligand”. When they screened it against potential target proteins they found it was highly attracted to AKT-1.
What was remarkable about this method was that the two halves of the synbody (the two peptide necklaces) had low affinity to AKT-1, suggesting that two peptide chains with weak attraction to a given protein can combine to make a synbody with strong binding properties.
The raw material for the synbody came from a library of 10,000 peptides, each comprising a randomly assembled sequence of amino acids.
“The randomness turned out to be the key to all of this, because a random sequence has more flexibility and degrees of freedom than life sequences do,” said Johnston, explaining that each peptide “necklace” of amino acids is able to find 2 or 3 points of contact with just about any protein.
“When two such peptides are combined to form a synbody, a high-affinity ligand is produced, displaying specificity for a given protein,” he added.
The researchers said throughput is currently limited by the number of proteins you can arrange on an array slide, but technological improvements mean that will also change and speed up.
They reckon it will cost around $1 billion dollars over ten years to make ligands that match all 30,000 proteins in the human proteome using traditional methods.
Johnston said he was “too impatient” to wait for this, and the price tag is too high. The synbody approach appears to offer a much faster, and cheaper way to get there.
Another advantage of synbodies over biologically produced antibodies, said the researchers, is that they remain stable over time, which makes them more attractive for use in diagnostic tools.
“Discovery of High-Affinity Protein Binding Ligands – Backwards.”
Chris W. Diehnelt, Miti Shah, Nidhi Gupta, Paul E. Belcher, Matthew P Greving, Phillip Stafford, Stephen Albert Johnston.
PLoS ONE, 5(5): e10728, published online 19 May 2010.
Source: Arizona State University.
Written by: Catharine Paddock, PhD