Out of the nearly infinite number of molecules created by nature and those that can be synthesized by medicinal chemists in the lab, what makes a molecule a drug? And critically, where are the richest caches of “drug-like” molecules?
Let’s start with the definition of “drug-like.” Doctors keep it simple and weigh benefit (does it help the patient’s condition) versus risk (are the side effects tolerable), maximizing the former and minimizing the latter. Medicinal chemists, on the other hand, study a litany of physicochemical properties and create rules that aim to predict and define “drug-like” molecules. The validity and utility of this approach is highlighted in a recent study by drug hunters at Novartis, who tracked how these properties changed over the course of lead optimization and were associated with the characteristics that doctors and patients care about, namely benefit and risk.
Intriguingly, it appears that there is one class of molecules that often breaks these human-derived rules and yet still delivers value to patients: natural products. These are molecules derived from nature and represent the oldest medicines on Earth. Natural products also bust another key trend: ROI. Despite several decades of waning interest from biopharma, natural products represent 30% of the new small molecule drugs registered between 2000 and 2020, as calculated by David Newman and Gordon Cragg, which represents a significantly outsized success relative to investment.
This all begs the question: what is it about natural products that allows them to defy the conventional wisdom of both medicinal chemists and biopharma executives? I believe that it is the wisdom of nature.
Returning to the idea of what defines a molecule as “drug-like,” one can think of it at the cellular/molecular level: a drug enters an organism, interacts with the existing biological systems and effects change in those systems, generally through interactions with proteins. All of these existing biological elements – the receptors, enzymes, transporters, carriers, etc. (i.e. the drug targets) – have evolved to utilize the chemical substances existing in the natural world. Across evolution there are recurring motifs that govern these essential interactions. As argued in a recent paper by Robert Young, Paul Leeson and several other leading authorities on drug-like properties from pharma and academia, The Time and Place for Nature in Drug Discovery, drugs that more closely mimic natural products have conserved recognition elements and properties associated with them. Said another way, these drugs are taking advantage of nature’s solution, interacting with targets in an evolutionarily-optimized manner.
These benefits also extend to pseudo-natural products (PNPs), which are not formed in nature but contain natural product-like fragments. The work by Young and others builds on a hypothesis originally suggested in the landmark 1997 paper by Chris Lipinski, that natural products exhibit a wider range of acceptable “drug-like” characteristics than synthetic, non-natural drugs. This emphasizes that nature does not have to play by the medicinal chemist’s rulebook, expanding an already enormous cache of “drug-like” natural product molecules. At Enveda, we recognize that we have only just begun to grasp how to utilize the wisdom of nature to improve drug discovery.
We have built a natural product-based drug discovery platform that can help patients in need of novel treatments by providing chemical solutions to the most difficult biological problems.
In my next blog, I will describe how our platform uses computational metabolomics and machine learning to solve many of the challenges associated with natural product-based drug discovery. We are decoding natural intelligence using its artificial counterpart.