Volume 86 Issue 22 | pp. 18-23
Issue Date: June 2, 2008

Cover Stories

Harnessing Helices

Chemical braces hold peptides in place, heralding a potential new class of therapeutics
Department: Science & Technology
Face To Face
A stabilized α-helix (pink) binds to a key portion of the HIV capsid (blue), as shown in an NMR structure. A hydrocarbon side-chain cross-link (yellow) helps this helix enter cells.
Credit: Courtesy of Shibani Bhattacharya

TECHNOLOGY for reinforcing the α-helix, a familiar protein motif throughout biology, could lead to a new class of peptide-based drugs. Because of their particular blend of chemical properties, stabilized helices may work against disease targets that have traditionally been out of reach.

Protein-embedded α-helices mediate key protein handshakes in cancer, HIV, and other diseases. But actually using an α-helix as a drug has proven tricky. So-called stapled α-helices, boasting sturdy cross-links between nonnatural amino acid side chains, just might change that. This class of stabilized peptides can regulate signaling pathways to subvert cancer. They also appear to overcome several of the usual problems that have hampered the development of peptide drugs. The technology has reached the point where a start-up company based on peptide-stapling technology is looking ahead to the clinic. Meanwhile, basic research is ongoing to unravel some of the biology behind how these strapped-down helices work and what else they might be capable of doing.

Gregory L. Verdine is spearheading those research efforts. The Harvard University chemical biologist pioneered the development of stapled peptides and is the founder of Aileron Therapeutics, the Cambridge, Mass.-based start-up company working to develop stapled-peptide drugs.

In Verdine's eyes, traditional small-molecule and protein drugs have drawbacks that stapled peptides might address. For instance, he says, most antibodies and protein therapeutics can be used only to target proteins on the cell's outer surface. Stapled peptides can enter cells and therefore can address protein-protein interactions inside cells, which for the most part haven't been accessible with protein drugs.

IN ADDITION, because of their larger surface area relative to small molecules, stapled helices can make contacts with other proteins over a large, relatively flat surface area. This is how many protein-protein interactions take place. In contrast, small molecules tend to bind best to well-defined pockets on protein targets.

"So many proteins have been disease targets for years, but they've been difficult to address because of a lack of chemical entities to disrupt intracellular protein-protein interactions," says Loren D. Walensky, a chemical biologist at the Dana-Farber Cancer Institute, in Boston, who also works with stapled peptides.

Dana-Farber chemical biologist James E. Bradner, a collaborator of Verdine's, agrees. "It's very frustrating to have to tell a patient, 'I know how your cancer works, but we don't have a drug that targets this problem,'" he says.

It's not immediately obvious that an α-helix could be a way to resolve that dilemma. Although α-helices do play a role in protein-protein interactions inside cells, for the most part they function as part of a larger protein structure. Trying to yank one key helix out of a full-length protein hasn't generally been productive from a drug development perspective. In addition to degrading quickly in the bloodstream, any peptide by itself tends to be a floppy chain of amino acids in solution. When a floppy helix binds to a target, it loses some freedom of motion, which is energetically costly. Plus, like full-length protein therapeutics, the vast majority of peptides can't enter cells.

But according to Paramjit S. Arora of New York University, he and other groups of researchers, including Verdine and his colleagues, are eagerly exploring a loophole. "If you lock peptides into a helical state, you can potentially address every one of these problems," Arora says.

The lock on Verdine's stapled peptide helices is a hydrocarbon cross-link with a carbon-carbon double bond. This lock, or staple, preorganizes a flexible peptide into the shape of an α-helix. That way, the energetic cost when the peptide binds to its target is not as high.

This first generation of stapled peptides ranges from about 12 to 35 amino acids in length. Their sequences are substituted with a pair of nonnatural amino acids, first synthesized by the Verdine group, featuring olefin-tipped hydrocarbon side chains. A catalyst acts as the stapler, merging the two double bonds on the side chains into one in a reaction known as olefin metathesis. Robert H. Grubbs at California Institute of Technology, a pioneer in metathesis technology, was the first to apply this reaction to peptide side-chain cross-linking (Angew. Chem. Int. Ed. 1998, 37, 3281).

Stapling "shows how developments in one area—organometallic chemistry—can then lead to advances in a much broader field of biological chemistry," comments Andrew D. Hamilton of Yale University, who works with mimics of α-helices.

Since helices act as go-betweens in many biological processes, researchers have long been trying to stabilize them in various ways. Two key properties of stapled peptides—their resistance to protease enzymes and their ability to penetrate cells—could help them on the long road to a peptide drug.

The Verdine team's earliest work on stapled peptides showed that they resist protease activity (J. Am. Chem. Soc. 2000, 122, 5891). Proteases rapidly degrade peptides in the bloodstream. So when it comes to half-life in the blood, "you're lucky if a typical peptide lasts for three minutes," says Tomi K. Sawyer, chief scientific officer for Aileron.

Locking In
The olefin metathesis reaction unites two pendant double-bond side chains (orange and aqua) on a flexible peptide (left), leaving one double-bond "staple" that keeps the peptide in an α-helical shape (right).
Credit: Eric Smith/Dana-Farber Cancer Institute
Locking In
The olefin metathesis reaction unites two pendant double-bond side chains (orange and aqua) on a flexible peptide (left), leaving one double-bond "staple" that keeps the peptide in an α-helical shape (right).
Credit: Eric Smith/Dana-Farber Cancer Institute

A PROTEASE can clip only peptides that have unfolded enough to fit into the enzyme's active site. "By locking peptides in a helical state, what you're asking the protease to do is unravel the peptide and then cut it, and that costs energy," Arora says.

Stapling a peptide essentially makes a large ring that spans the peptide backbone and the two nonnatural side chains, comments peptide expert Victor J. Hruby of the University of Arizona. Making cyclic peptides and peptide mimics is an established strategy for warding off proteases, one that's been followed in established peptide-mimicking drugs, he says.

If stapled helices' resistance to proteases didn't come as much of a shock, what was actually surprising was that stapled peptides could cross cell membranes. In conjunction with the late Stanley J. Korsmeyer at the Dana-Farber Cancer Institute, Verdine and Walensky, then a postdoctoral fellow in Korsmeyer's lab, showed that was possible when they demonstrated stapled peptides' first biomedical application—blocking the growth of leukemia cells in mice (Science 2004, 305, 1466).

Last month, another research team used stapling technology to induce their peptide of interest to enter cells. The researchers, led by Asim K. Debnath, a molecular-modeling and drug-design expert at the New York Blood Center, and David Cowburn at the New York Structural Biology Center, used a stabilized α-helix to block HIV infection in cell cultures (J. Mol. Biol. 2008, 378, 565).

Debnath's research seeks to disrupt the HIV capsid, which helps HIV mature into an infectious virus. He was interested in exploring a promising capsid-binding peptide identified by another research group. Unfortunately, the peptide could not enter cells, a crucial property for being able to prevent HIV's assembly and maturation. Debnath reasoned that if the peptide could somehow be made to enter cells, it might be useful against HIV. He examined other options for carrying the peptide into cells and eventually he settled on stapling. With another group's X-ray crystal structure of the peptide-capsid complex in hand, Debnath says, he realized he could tailor his stapling site to be far away from the binding interaction.

Sure enough, the very first stapled peptide Debnath's team developed could enter cells and was effective against HIV in cell culture. What's more, the team conducted NMR studies that demonstrated that the staple doesn't dramatically affect the way the peptide helix binds to its target. Their NMR structure, which the team is now using to design peptides that will bind more tightly to the capsid, is the first atomic-scale structure of a stapled peptide binding to its target (J. Biol. Chem., DOI: 10.1074/jbc.C800048200).

IN THE NMR STRUCTURE, the stapled helix looks like a normal α-helix, and as expected, the staple is situated away from the key protein-protein interaction with the HIV capsid. "The fact that the stapled peptide does exactly what it's supposed to do is important," Cowburn says. "This kind of structural validation is a key issue for stapling technology going forward."

Stapling alone does not lead to cell-permeable, functional peptides in every case, however. For instance, Verdine's team recently made a library of stapled peptides that they hoped would interfere with a crucial interaction in cancer. Their first set of peptides was incapable of penetrating cells, and the team had to make changes to the peptides' amino acid sequences to fix the problem (J. Am. Chem. Soc. 2007, 129, 2456).

The science of cell-penetrating peptides is "a difficult field to make sense of," with literature reports that are often contradictory, comments Samuel H. Gellman of the University of Wisconsin, Madison, who has worked with both helix mimics and positively charged cell-penetrating peptides. Walensky's group is closely examining how a peptide's properties, such as charge and structure, influence cell penetration. In α-helices, the polar backbone gets tied up by hydrogen bonding to itself, rather than to water molecules around it. That and other factors, such as the "greasiness" of the staple itself, may make it easier for stapled peptides to penetrate membranes, Walensky says.

Arora and Hamilton tell C&EN that in some cases their helix mimics can also penetrate cells. A variety of factors contribute to cell-penetrating ability, Walensky says. "We're still learning the rules."

Aileron is also examining stapled peptides' properties in depth. According to Sawyer, the company is focused on preclinical testing of two stapled peptides and intends to submit an Investigational New Drug Application to the Food & Drug Administration sometime in 2009. These peptides are key α-helices from the proteins BID (BH3 interacting domain death agonist) and BIM (BCL-2 interacting mediator of cell death), which play central roles in turning on cell death.

Aileron's goal is to use these peptides to treat patients with cancers such as leukemia, lymphoma, and certain solid tumors, Sawyer says. At April's American Association for Cancer Research meeting, Aileron reported that injected doses of stapled peptides have a half-life of about three hours in rats, a 30- to 60-fold improvement over the nonstapled versions of those same peptides. They also provided evidence that stapled peptides don't trigger a mouse's immune response, something that can be a concern for peptide drug candidates.

The available information doesn't yet answer another question that comes up in peptide drug development, which is whether the drug might be given to patients as a pill. Sawyer and Verdine say that although they are investigating alternatives to injections, Aileron's current focus is on cancer therapeutics, an area where giving patients a drug by injection is not unusual. In terms of reaching the market, "that lowers the barrier for a peptide," Verdine adds.

With promising results in preclinical testing, the ability to quickly make adjustments to stapled peptides and scale up their production becomes important. Advances in peptide synthesis technology have enabled researchers to scale up production of stapled peptides.

According to Walensky, the synthesis he uses to make libraries on a small scale is a slightly modified version of a standard solid-phase peptide synthesis. He notes that it takes extra time to allow the nonnatural olefin-containing amino acids to couple to the growing peptide chain. In general, it takes three to four weeks' time to make a purified set of compounds that are ready for the first round of biological screening, Walensky says.

Aileron is now addressing large-scale production, Sawyer says. The company is currently making kilogram-scale batches of olefin-containing amino acids and multigram scale batches of stapled peptides. They plan to scale up further when they finish making modifications to the structures of the BID and BIM stapled helices.

As successful precedents for making helical peptides on a commercial scale, Verdine, Sawyer, and other sources point to Roche's Fuzeon, a helical peptide used to treat HIV (C&EN, March 14, 2005, page 17), and Amylin Pharmaceuticals and Eli Lilly & Co.'s Byetta, a helical peptide that is a type 2 diabetes therapy. Stephen B. H. Kent, a peptide synthesis expert at the University of Chicago, says that the cost of making peptides on a large scale has dropped significantly over the past 20 years. He credits the price drop to advances in solid-phase peptide synthesis as well as the introduction of high-resolution mass spectrometry for quality control. "The peptide synthesis community is applying all their newest tools to developing peptides as therapeutics," he says.

Stapling "is elegant work that is also potentially very general," Hamilton comments.

APART FROM their work with Aileron, Verdine and Walensky have academic collaborations that they hope will reveal just how versatile stapled helices can be. For example, with Nika N. Danial at Dana-Farber, they have used a stapled peptide to simulate a full-length protein's role in insulin secretion, a result that might have implications in diabetes research (Nat. Med. 2008, 14, 144).

Verdine tells C&EN that he is collaborating with several experts to apply stapling to other key protein-protein interactions in cancer. For instance, Verdine and Bradner are working on a stapled peptide that they say potently inhibits a cancer-causing transcription factor, a protein that helps form the machinery that binds DNA and turns genes on and off. This class of proteins was long thought to be untargetable by drugs, Verdine says.

"This has evolved into a major academic research program," Verdine says. Academic researchers, he adds, can play a key role in the drug development process because they are free to explore potential drug targets where the biological rationale is more speculative. "Stapled peptides are allowing us to define that biology, and that may lower the risk of pursuing new targets in the future," he says.

Chemical & Engineering News
ISSN 0009-2347
Copyright © American Chemical Society