Hyperstable Peptides for 'On-Demand' Drugs

Scientists at the University of Washington’s Institute for Protein Design have shown it is possible to create small, hyperstable peptides that could provide the basis for developing powerful new drugs and diagnostic tests. 

“These super stable peptides provide an ideal molecular scaffold on which it should be possible to design ‘on demand’ a new generation of peptide-based drugs,” said UW Medicine protein engineering pioneer David Baker, who oversaw the research project.  He is a UW professor of biochemistry.

In a study, which appears in the journal Nature, the researchers demonstrate that not only is it possible to design peptides that fold into a wide variety of different conformations, but also that it is possible to incorporate functional groups of chemicals not normally found in peptides.  Both of these abilities could give designers even greater flexibility to create drugs that act on their molecular targets with high precision. Such drugs should not only be more potent but would also be less likely to have harmful side effects.

Most drugs work by binding to a key section of a protein in a way that alters how the protein functions, typically by stimulating or inhibiting the protein’s activity. For the binding to occur, the drug must fit into the target site on the protein as a key fits into a lock. How close the lock-and-key fit is can often determine how well the medication works.

Currently, most prescription drugs are either made of small molecules or much larger proteins.  Both classes of drugs have advantages and disadvantages.

Small molecule drugs, for example, tend to be easy to manufacture, tend to have a long shelf life, and are often easily absorbed. But they often don’t fit the targeted “lock” as selectively as could be hoped.  This imperfect fit can result in off-target binding and side-effects that diminish their effectiveness. Protein drugs, on the other hand, often fit their target receptors very well but they are difficult to manufacture, are more unstable, and lose their potency if they are not kept refrigerated.  Because of their size and instability, they need to be injected into patients.

Peptide drugs fall in between these two classes. They are small, so they have many of advantages of small molecule drugs. But they are made of a chain of amino acids, the same components that make up proteins, so they have the potential to achieve the precision of larger protein drugs.

The power of some tiny peptides can be observed in venomous creatures. A number of poisonous insects and sea creatures produce small peptide toxins. Those are some of the most potent pharmacologically active compounds known.  Their potency is among the reasons why medical scientists are interested in tapping into beneficial uses of peptides.

In the new study, Gaurav Bhardwaj, Vikram Khipple Mulligan, Christopher D. Bahl, senior fellows in the Baker lab, and their colleagues, developed computational methods that are now incorporated in the computer program called Rosetta.  These methods are being used to design peptides ranging from 18 to 47 amino acids in length in 16 different conformations, called topologies.

Originally developed by Baker and his earlier team, Rosetta uses advanced modelling algorithms to design new proteins by calculating the energies of the biochemical interactions within a protein, and between the protein and its surroundings. Because a protein will assume the shape in which the sum of these interaction energies is at its minimum, the program can calculate which shape a protein will most likely assume in nature.

The peptides were made hyperstable by designing them to have interior crosslinking structures, called disulfide bonds, which staple together different sections of the peptide.  Additional stabilization was secured by tying the two ends of the peptide chain together, a process called cyclization. The resulting constrained peptides were so stable that they were able to survive temperatures to 95 °C, nearly boiling. This survival feat would be impossible for antibody drugs.

The researchers also showed that the design of these peptides could include amino acids not normally found in proteins. Amino acids have a property called handedness or chirality. Two amino acids can be made of the same atoms but have different arrangements, just as our hands have the same number of fingers but have two mirror-image configurations, right and left. This handedness keeps the right hand from fitting properly into a left-handed glove and vice versa.

 In nature, perhaps due to a chance event billions of years ago, amino acids in living cells are all left-handed. Right-handed amino acids are very rare in naturally occurring proteins.  Nevetheless, the researchers were able to insert right-handed amino acids in their designed peptides.

 “Being able to include other types of amino acids allows us to create peptides with a much wider variety of conformations,” said Baker, “and  being able to use right-handed amino acids essentially doubles your palette.”

“By making it possible to create peptides that include ‘unnatural’ amino acids, this approach will allow researchers to explore peptide structures and function that have not been explored by nature through evolution,” Baker said.