Where do drugs come from? Most clinical molecules are either produced by chemists in a laboratory, or naturally in living organisms. While synthetic chemistry is a pipeline to drug discovery, nature-made molecules continue to have an important role as drug templates. Certain soil bacteria called filamentous actinomycetes are particularly apt at producing biologically active compounds. Astoundingly, around two-thirds of clinically used antibiotics are derived from their products. We recognize this unmet demand for new pharmaceuticals to battle drug-resistant infections and other hard-to-treat maladies. Exploring natural sources for unique chemistry to inspire new drugs remains crucial. 

 

Rare actinomycetes are interesting producers of new therapeutic molecules. They earn their 'rare' classification because they are less studied than their famous drug-producing cousins, the Streptomycetes. Researchers like us continue to explore many actinomycetes for useful molecules to treat various diseases, from bacterial infections to cancer and more. Recently, our research teams led by Professors Joshua Blodgett at Washington University in St. Louis and Shugeng Cao at the University of Hawai’i delved deeper into the bioactive molecules produced by the rare actinomycete Lentzea flaviverrucosa. 

 

For any living organism to produce a drug-like compound, it must harbor specific instructions within its genes. Sets of genes called biosynthetic gene clusters work together to create these molecules. We chose to study L. flaviverrucosa in particular because a genetic marker suggested it could produce piperazate (Piz). Piz is an uncommon amino acid. It is not used to build proteins but rather is a chemical building block in several natural drug-like molecules. Piz is considered a 'privileged scaffold' by chemists, and molecules containing Piz often have biological activity. Privileged scaffolds have molecular features that can target many diverse receptors in biological systems, making them extremely interesting and versatile for drug development. Presence of Piz can indicate bioactivity, and L. flaviverrucosa genes suggest the production of Piz and of other molecules that incorporate Piz. So, our research team investigated the strain to determine what those molecules would be and the potential bioactivity. 

 

To do this, we engaged in targeted metabolomics, using a sensitive mass-spectrometer  machine to carefully sift through all molecules produced by L. flaviverrucosa to reveal only ones that contained Piz. Our device measured and shattered thousands of L. flaviverrucosa-produced molecules based on their mass-to-charge ratio and intrinsic fragmentation habits, or the way they broke apart. Our group programmed the machine to only flag molecules that "break" with a specific signal that indicates Piz may be present. We can then back-track to and use those fragment patterns to identify the larger parent molecules: like solving a puzzle to an unknown product. Once we put these pieces together, the larger molecules can be isolated for purification and additional chemistry solves their structures. 

 

Throughout this process, we discovered two new cyclopeptide molecules that did indeed include Piz: Petrichorin A and B. While Petrichorin B looks highly similar to a few known actinomycete products, Petrichorin A is new to science. The major difference we detected in these molecules is their dimer structures. Dimers are larger molecules made of two smaller molecules, called monomers, bound together. Petrichorin B is characterized as a homodimer, made of two identical halves. For example, if molecule X were made into a homodimer, it would look like X-X, vaguely resembling a dumbbell. Petrichorin A is different because it is a heterodimer, meaning each half of the molecule is different. Its halves, or monomers, are not the same, so X-Y symbolizes it. To our knowledge, natural heterodimers are much rarer than homodimers. 

 

The petrichorins are scientifically unique, yet share many characteristics with some previously discovered molecules and homodimers. Because most of those similar molecules are cytotoxic against cancer cells, we tested the two new petrichorins for this activity. After testing the petrichorins against lymphoma, fibrosarcoma, ovarian, and prostate cancer cell lines, we observed that heterodimeric Petrichorin A (X-Y) was most effective at killing these cells. Petrichorin B (X-X) was also fairly active. However, monomeric Petrichorin C (in our example, X alone) was far less active than A or B in the same assays. 

 

These bioactivity trends neatly illustrate the concept of drug multivalency, where dimers are sometimes more effective than monomers. Multivalency, or a several-pronged attack on biological targets, has been long acknowledged in drug-like molecules. Though heterodimers have been overlooked compared to homodimers. Our results infer that Petrichorin A's heterodimeric structure, coupled with strong biological activity, is a promising discovery. We anticipate this will open an exciting new avenue for exploring natural Piz-heterodimers in the hunt for new drugs.