Banner by Troy Barlow

How to Build a Protein with Superpowers

by Eion Plenn

Organic Chemist Seth Horne grabs a 3D printed protein from his bookshelf, nestled right next to a dusty volume of the American Chemical Society and a picture of his family. The molecule looks like an alien orb, a sculpture comprised of turquoise and orange “squigglies.” It’s actually a bacterial protein Horne and his team used to study mimicry of protein structure, what would later become the foundation of his research at the University of Pittsburgh.

“You know, proteins have this shape. They have this property of adopting this folded shape,” Horne says. “We’re trying to exploit that to create ordered materials and to create materials that have interesting properties. Responsive properties.”

Proteins are crucial for life, and they have been the subject of many biomedical, food processing and biotech discoveries. It wasn’t until the early 1990s, right around the time Horne was starting his Chemistry degree at Texas A&M University, that new protein engineering technologies emerged. Discoveries in computational biology led to the de novo protein design, where now scientists can use a variety of tools to visualize, modify and synthesize protein structure.

In December 2018, Horne co-published a piece in ChemBioChem which discussed the discovery of a “disulfide-rich mini protein” mimic. Horne’s paper cites that these mimics are thought to have many therapeutic uses, including treatment of cancer, autoimmune diseases and chronic pain. Why? Because these mimics have a superpower: resistance to degradation by protease.

Our DNA codes for hundreds of proteases, “pac-man”-like enzymes which chomp up the proteins that enter our bloodstream. From a defensive standpoint, proteases are beneficial, guarding against foreign agents and pathogens. However, if you want to inject a protein into the blood with positive effects, such as an anti-tumor drug, there’s a downside – proteases will identify this protein as “foreign” and rapidly break it down.

Horne’s mini protein circumvents this problem. The “disulfide-rich mini protein” is comprised of an unnatural backbone, making it more resistant to breakdown by protease. In replacing natural residues with artificial ones, chemists are able to create a protein with a “heterogeneous‐backbone.” As a result, the protein is protected from protease.

But this is the easy part. The trick, Horne says, is making sure the artificial protein has the same properties as the original one. Biochemists have a name for these structures: foldamers, or proteins comprised of artificial amino acids which mimic the structure of a naturally occurring protein. The theory behind it all is quite simple: if the protein can retain its shape, then the protein should retain its function.

How do biochemists like Horne synthesize complicated protein structures in the lab? It starts with an amino acid. There are 20 naturally occurring amino acids, and each acts like a different “letter” in nature’s “alphabet.” These amino acids (letters) come together to form the primary structure of proteins (words), which can then go on to fold into secondary, tertiary and, in some cases, quaternary scaffolds (sentences and paragraphs).

But the amino acids Horne uses in his lab are unique. They cannot be found in any introductory biology textbook or the health foods store. There are also far more than 20 of them. Horne and his colleagues use man-made amino acids known as “β(beta)-amino acids.” β-amino acids are structural siblings of a natural amino acid, except the amine group has been “bumped” one carbon down the chain and is now at the beta position. β-amino acids are used to create foldamers mainly due to their versatility, making the perfect “substitutes” to a natural peptide backbone.

“A desired protein scaffold is like a white canvas,” Horne says. “What we have now in our research is a palette of different building blocks we can use to try to paint this natural scaffold. We start from a natural, or at least a known folded structure, and try to modify it.”

While the science of protein engineering may sound as simple as putting together a box of Legos, the process is arduous and complicated. The big problems, Horne says, become what kinds of modifications should biochemists use, and where should these modifications be placed so that the protein will still retain the fold. 

Regardless, there are many clinical applications to these protein mimics, a famous example being HIV anti-viral therapy. In the mid-90s, the drug company Fuzeon sought to develop synthetic fusion inhibitors – proteins which prevent circulating HIV in the blood from fusing with the host cell and infecting it. While it was thought in the pharmaceutical field that you couldn't make a protein-based drug “that big” – 30-40 amino acids long, to be precise –Fuzeon set out to do just that. 

Interestingly, the company’s downfall was also its greatest achievement. One injection of the medication would contain 100 mg of the synthetic peptide, which is an extremely high dosage considering the potency of the drug. The size of the injection was because the proteins were rapidly degraded in the body by protease. There was simply no way around a large dose.

Fuzeon may have crafted a new way of treating HIV, but the high dosage of the medication was unsustainable and costly from a synthetic chemistry standpoint.  Yet, in Horne’s eyes, Fuzeon’s bumpy history highlights the benefits of studying unnatural amino acids in the first place.

“Why then make more stable versions of [Fuzeon]?” Horne questions. “Well, if we can retain the native activity but improve the stability ten-fold, 100-fold, 10,000-fold, we can give a lot less and get the same functional effect.” 

Not only does the use of foldamers in pharmaceuticals become relevant to HIV treatment, but diagnostic agents as well. For instance, an oncologist may use a bioactive chemical to “hone in” on a tumor cell, sort of like a molecular magnifying glass. A protein that would achieve this function would need to be extremely stable in the body and resistant to protease. Horne claims that foldamers and their β-amino acids counterparts would be able to do just that. 

“If we don't show what’s possible in these kind of things in either a pre-clinical or an in vitro type of setting, why would anyone want to surmount those practical barriers to make them more accessible? The kinds of synthesis and design challenges that we solve is to show what's possible in these kinds of agents, in a way that will hopefully inspire future work.”