Nature is the ultimate inspiration for innovators, including those who would emulate biology to realize applications in the life sciences. Indeed, some of the most intrepid innovators are those who emulate biology to help them design medicines. One such innovator is Dietrich Stephan, PhD, the CEO and founder of NeuBase Therapeutics and the chairman of the board at Peptilogics. He is a big believer in learning from nature to replace lost functions.
“The parts of the human machine have evolved over millennia through trial and error,” Stephan says. “Who are we to deviate from that and create suboptimal solutions? In biomimicry, I’m fascinated by the starting point, which must be the natural molecule, [as well as everything that follows and contributes to] a very conscious decision of what to change and why.”
Just as the best repair strategy for cars is to use parts from an original equipment manufacturer, the best drug design strategy is to avoid “corrupting the evolution” that has gone into refining natural molecules. Pursuing this drug design strategy requires an understanding of exactly how all the molecular “parts” work together. In the absence of such precise insights, drug designers commonly opt for the next best strategy: throwing a random assortment of molecules at the body to find stopgap remedies.
Antibodies that bind to pathogens or malignant cells. Viral envelops that deliver mRNA to specified targets. These are outstanding examples of biomimicry in therapeutic design. “It’s interesting,” Stephan declares, “to see where we start and where we end with these new modalities, and whether they all vector toward the natural human machinery or not.”
Mimicking DNA and peptides
“At Peptilogics, we are making short, naturally occurring, synthetic peptides,” Stephan points out. “We have created a deep learning platform that can quickly make massive scales of peptides to see how we can rescue function with these naturally occurring parts. We’re training the machine to learn from nature to predict function, and to help us create a replacement part made of the same substance as the original part.”
A different approach is followed at NeuBase. The company intends to drug the human genome by taking inspiration from the double helix itself. Using DNA’s inherent self-complementarity as the basis for drug design, NeuBase “engineers in” accessory properties for therapeutic effect. Stephan says, “You need to turn those dials carefully to make sure you don’t mess with the core activity you need.”
Preventing natural degradation
The most common criticism that confronted Peptilogics’ drugs at the outstart was the natural degradation of synthetic peptides in circulation. “We were concerned,” Stephan recalls, “but empirically we saw that when these peptides are in circulation, they associate with other circulating proteins and peptides and stabilize. We have shown half-lives of seven hours in circulation with short, naturally occurring synthetic peptides. They survive long enough to function.”
Classically, the approach adopted to prevent peptidase-mediated degradation has been to replace natural levorotatory building blocks in peptides with dextrorotatory stereoisomers.
“We have not had to engineer in any alternative amino acids or other chemical modifications to stabilize peptides,” Stephan asserts. “Other naturally occurring therapeutic proteins, such as monoclonal antibodies and enzyme replacements, don’t use substituted amino acids to stabilize them. There is a world where you can use natural amino acids in therapeutics and take advantage of the body’s own way of processing them.”
Nucleic acid–based therapeutics are a different story. RNA and DNA are not meant to exist outside cells, so it raises a red flag if they course through the circulation. “As an industry, we have had to go to great lengths to protect mRNAs until they get into the cytoplasm where they can be translated,” Stephan points out. “As a DNA-mimetic company, one of the properties we had to ‘engineer in’ is stability against exo- and endonucleases. We had to use a composite chemical structure that is not recognized by nucleases.”
Sending medicinal messages
Nanoparticle-mediated cross-kingdom intercellular communication is pervasive in nature, including among species that have cohabitated within humans over millions of years. Senda Biosciences, a preclinical company and an offshoot of a Flagship Pioneering enterprise, is the first to access the chemical addressing code of natural nanoparticles to enable programming “to the cell” (in contrast to “within the cell” programing that genome editing enables), thereby unlocking the ability to comprehensively program medicines.
Senda has compiled 75,000 molecular components from natural nanoparticles into an atlas that is poised to complement, if not replace, state-of-the-art synthetic lipid nanoparticle–mediated drug delivery. “We are building on existing technology and bringing a new component to the ‘natural programming language,’” says Guillaume Pfefer, PhD, CEO at Senda and CEO Partner at Flagship Pioneering.
Utilizing the natural programing language of information molecules like DNA, mRNA, siRNA, gene editors, and coating peptides that act within cells to regulate function, was the first step toward programable medicines. “Too few of these abundant information molecules have been translated into medicines,” Pfefer observes. “We are sitting on part of the equation for programmable medicines, but we have not solved the equation yet.”
Whereas understanding the genetic code enabled its reprogramming within cells, a crucial piece of the puzzle is sending information-carrying molecules to their destinations. “The problem is, we don’t know how to send the right molecule to the right cell,” Pfefer admits. “We have solved the issue of programming ‘within the cell,’ but we have not solved programing the message ‘to the cell.’ If we solve both, we can comprehensively program medicines.”
To find a solution to the latter problem, Senda’s team probed the natural messengers that allow intercellular communication across species in the six kingdoms of life—protists, bacteria, fungi, archaea, animals, and plants. “Everywhere in nature, including within us, you see intercellular cross-kingdom communication, which involves nano-sized particles that are coded chemically to direct communication to specific cells, to dose cells safely and repeatedly,” Pfefer emphasizes. “Doesn’t that ring like something we would like to use for medicine?”
Natural nanoparticles are composed of lipids, proteins, carbohydrates, and other diverse molecules with species-specific structural and functional features. “This rich compositional diversity allows for varied functions of natural nanoparticles,” Pfefer elaborates. “It acts as a rich reservoir from which we can mine components for our Senda Atlas.”
Therapeutic information molecules like mRNA can be stabilized within nanoparticles and activated once the assembly penetrates targeted cells. Combining components in its atlas, Senda is creating nanoparticles to drive specific outcomes such as tissue or cell tropism, increased potency, or the ability to repeat doses.
The ability to program precisely depends on the ability to program comprehensively. “You will not expect parts of a sports car to come in pieces that miraculously assemble for a fine performance,” Pfefer explains. “Similarly, these information molecules act very precisely. They need to be combined in a very comprehensive way with programmable nanoparticles. So, we have created an mRNA engine tapping into the natural programing language of genetic codes.”
With the development of “SendRNAs” that combine RNA and nanoparticle components, Senda’s technology already exceeds the performance of mRNA vaccines for COVID-19 in preclinical models. “We not only generate protection against the disease in this model, but we also cut viral transmission for the very first time,” Pfefer claims. This is because SendRNAs generate both systemic and mucosal immune responses. The technology can optimize and program nanoparticles with an mRNA encoding for the SARS-CoV-2 spike protein. The technology is also surpassing gene editing efficacies in preclinical models, including the ability to dose repeatedly that has historically limited the safety of gene therapies.
By exploring how plants inject toxic nanoparticles into invasive pathogens and convey solar energy into human cells through diet, and how commensal and pathogenic bacteria transport DNA into human cells to replicate and survive, one appreciates how nature has bridged communication gaps between living kingdoms, and indeed between the living and nonliving worlds.
“We can transfect mRNA into circulating immune cells in a nonhuman primate at double-digit levels,” Pfefer says. Senda has been able to activate B and T lymphocytes within germinal centers of lymph nodes using SendRNAs at a level that is five-log higher than what other products have achieved in a similar model—with up to 60% less mRNA. Currently, the company is exploring applications in vaccines, chimeric antigen receptor T cells, and treatments for local gastrointestinal infections via oral administration that limits systemic exposure.
To find the right nanoparticle composition for targeted messaging, Senda is exploiting artificial intelligence. The platform uses 20 different combinations of nanoparticles, at present, to optimize SendRNAs for specific outcomes, such as delivery to the spleen with no liver expression.
“Our platform has demonstrated the ability to program SendRNA medicines that can access historically difficult-to-reach organs such as the lung and pancreas,” Pfefer declares. “We can even program Senda nanoparticles to reach the brain via the intranasal route.”
Next-generation anti-infectives
Small-molecule peptidomimetics can at times be a better therapeutic option than a synthetic replica of a natural peptide. Small-molecule anti-infectives that mimic natural antimicrobial peptides are being designed by Maxwell Biosciences. These anti-infectives, which are called Claromers, represent a way to cope with the global rise in resistance against antibiotics and antivirals. Claromers can target specific membrane vulnerabilities in a range of viral, bacterial, and fungal pathogens.
The company’s lead candidate, a Claromer that mimics the function of human cathelicidin antimicrobial peptide (LL-37), treats chronic rhinosinusitis—a severe and chronic sinus infection caused by combinations of fungi and bacteria. Preclinical in vitro data shows that LL-37 is effective against all chronic rhinosinusitis–related pathogens reported to date.
“LL-37 is expressed everywhere in the body, especially in mucosal tissues,” says Joshua McClure, Maxwell’s founder and CEO. “It serves as the first antimicrobial barrier that the innate immune system offers against all pathogens coming in through the air, food, or water.” McClure came across this peptide serendipitously when he discovered that members of his family lacked LL-37 and had an increased susceptibility to rare infections.
Although changes in lifestyle can boost LL-37, the adoption of such modifications is extremely difficult. “That’s where pharmaceuticals come in,” McClure observes. “Everyone with a depressed immune system needs anti-infectives. What if we could come up with an armored small molecule that would mimic this peptide, without its weaknesses? Many pathogens release enzymes that break down LL-37.”
Fortuitously, Annelise Barron, PhD, a scientist at Stanford University, had already developed and patented a small-molecule mimic of LL-37. McClure brought her on board as scientific co-founder at Maxwell, and they spent the next seven years in pre-IND studies.
Since COVID-19, this LL-37-mimicking small molecule has received abundant attention and funding. Researchers at the National Institute of Allergy and Infectious Diseases led by Anthony S. Fauci, MD, did a Syrian hamster study and demonstrated the compound’s pan-coronavirus and pan-influenza efficacy. Other laboratories have shown that several enveloped viruses are vulnerable to LL-37.
McClure has confidence in the small-molecule mimics of peptides because they can avoid being degraded by proteases. This ability helps keep dosages low, reducing risks and facilitating regulatory approvals. “We’ve hung the functional side chains from the nitrogen on the backbone instead of the carbon,” McClure points out. “So, the proteases don’t even recognize it.”
Conclusion
If we think about the design of medicines in broad terms, we are more likely to accept that all medicines—inorganic small compounds, organic biomolecules, composite agents, and so on—mimic nature in some fashion, even if we are unsure of how, exactly, the medicine’s mechanistic paths should be mapped.
Nonetheless, Peptilogics’ Stephan suggests, “It might be worth thinking about what a biomimic isn’t” and considering the effects on the biomimic’s function.
In any case, observes Maxwell’s McClure, biomimicry is the future for anti-infectives in biotech. “Antibiotics are dead,” he declares. “With machine learning, you’ll see a lot of peptidomimetics.”
Biomimetics requires an ability to observe nature and move beyond awe to integrate natural designs into therapeutic agents. Done well, the approach may be our best shot at circumventing existing roadblocks of biocompatibility, adverse reactions, toxicity, and long-term efficacy in therapeutic design. “Ultimately,” Senda’s Pfefer emphasizes, “it’s all about survival.”