Actinobacteria are common soil bacteria with an incredible talent for producing antibiotics that safeguard our ecosystems and help us battle dangerous pathogens. These versatile bacteria are responsible for producing approximately two-thirds of all known antibiotics. Their adaptability allows them to thrive in various ecosystems and forge intricate relationships with higher life forms. Recognizing this potential, we set out on a quest to unearth the hidden treasures within the world of Actinobacteria.

The journey begins

Our journey began with a discovery during the postdoc of my supervisor when she stumbled upon an Amycolatopsis strain isolated from the gut of fungus-growing termites by one of her colleagues (and co-author of this study) during a field trip. Little did she know, that this strain would become the starting point of our scientific journey and accompany her for many years of her scientific career.

Upon initial analyses, she quickly recognized the significant antimicrobial activity of this strain, particularly against the fungal weed of the termite's fungus garden, and embarked on a quest to identify the compounds responsible for this activity. The exploration of the strain’s metabolome led to the discovery of a novel group of macrolactams, that were named macrotermycins. However, the path to isolating these compounds was accompanied by challenges. Working in darkness and maintaining cool temperatures was necessary to overcome the intrinsic reactivity of macrotermycins. But it was worth it: the isolated compounds exhibited promising selective antibacterial and antifungal properties and the strain was suspected to act as defensive symbiont. 

Macrolactams are a fascinating and diverse group of natural products, each with distinct pharmacological properties, and as I began my Ph.D., my mission was clear: analyze the biosynthetic repertoire of this potent termite-associated Amycolatopsis strain  and unravel the mysteries surrounding the biosynthesis of the macrolactam compound class to enable their biotechnological exploration.

Improving Genomes Was Essential

However, detailed genomic analysis was tricky as back in the days the initial short-read Illumina sequencing run produced only a highly fragmented genome. While part of the putative macrolactam cluster was detectable, my first analysis uncovered puzzling inconsistencies in module number and composition, which were not in alignment with the characterized macrolactam scaffold.  This deficiency stemmed from the limitations of short-read sequencing during the project's beginning .  

Thus, one of my next tasks as PhD student was to improve the genome quality of our original termite strain to be able to exploit the biosynthetic pathway of macrotermycins. However, I soon realized that, due to the sheer size of these PKS-encoding gene clusters and the numerous repetitive gene sequences within them and other encoded clusters, it was absolutely necessary to utilize not only improved short-read sequencing techniques but also incorporate long-read sequencing technologies. Generating the hybrid genomes posed a technical hurdle for me as a chemical biologist, and I was fortunate to receive help from two of my colleagues. Together, we successfully achieved the required genome quality. This approach worked so well that we decided to extend our analysis to several other actinobacterial isolates from our strain collection. Although it was not initially planned, it ultimately proved to be crucial for the study's success.

Armed with several improved genomes, I revisited the macrolactam-encoding gene sequence and was finally able to reconstruct the complete version of the biosynthetic gene cluster  (BGC) in strain M39.

The Wide Distribution of Macrolactams

In the quest to expand the understanding of these intriguing compounds, I questioned myself how widely macrolactams are distributed across the bacterial world, and compiled a list of all known macrolactam scaffolds, and found strong variances in ring sizes and substitution patterns. However, not for all scaffolds the biosynthetic pathway was elucidated.

Thus, I pursued a targeted genome mining analysis for macrolactam-related clusters within  my newly sequenced strains and available genomes on NCBI, and was able to unveil that approximately 10% of strains within a given actinobacterial family harbored a BGC encoding macrolactam production. This insight provided strong evidence of the widespread distribution of macrolactams among Actinobacteria.

The Influence of PKS Domains on Ring Size

I then delved into the comparative analysis  of all macrolactam-related BGCs detected in Amycolatopsis during the genome mining approach.  A first key finding of my analysis was that the variations observed in macrolactam ring sizes depended on the presence or absence of crucial domains within the two terminal polyketide synthase (PKS) modules. These terminal modules contained two reducing (KR) domains, and mutations in these domains led to the observed variations in ring sizes of isolated macrolactams. 

Our Three Amycolatopsis Companions

To validate my computational pathway predictions, my subsequent studies focused on the chemical analysis of different species of Actinobacteria within the Amycolatopsis genus, which we anticipated to harbor macrolactam-type gene clusters. One of which belonged to those strains, whose genome we sequenced by chance. We were surprised to realize the phylogenetic proximity of these diverse Actinobacteria species, despite their origins in geographically distinct locations . These bacteria came also from entirely different ecological niches, reflecting the global abundance and colonization potential of different habitats.

1. Amycolatopsis sp. M39: This potent strain was isolated from the gut of termites and is the producer of macrotermycins.
2. Amycolatopsis saalfeldensis: This strain was isolated from a cave, demonstrating the adaptability of Actinobacteria even in the harshest conditions.
3. Amycolatopsis sp. PS_44_ISF1: The third species was recently isolated from the uropygial gland of birds, revealing the remarkable ability of Actinobacteria to colonize diverse hosts.

Our phylogenomic analysis unveiled that all three strains shared a common evolutionary lineage. 

Exploring the Metabolome

These three bacterial species were our companions on this scientific journey. As I delved deeper into the metabolic repertoire of all three strains, I detected a shared cluster of metabolites, which contained quasi-molecular ion features of macrotermycins primarily associated with Amycolatopsis sp. M39. However, I noticed additional nodes within this cluster, produced solely by A. saalfeldensis and Amycolatopsis sp.  PS_44_ISF1. This discovery, combined with genomic analysis, ultimately led me and my colleague to isolate a macrolactam known as ciromicin A, validating that these Amycolatopsis species were previously undiscovered macrolactam producers. One of my colleagues was even able to isolate novel and less abundant macrotermycin derivatives. However the structural assignments of macrotermycins was very challenging and only the comparative analysis of several isomers and calculations of chemical shift patterns allowed to narrow down and even correct the stereochemical assignment. Furthermore, after many cloning attempts and testing different hosts, I was able to heterologously express the macrolactam-relared BGCs from A. saalfeldensis poving it responsible for ciromicin A biosynthesis in this strain.

Future Prospects: A Wealth of Discovery Awaits

The outcomes of this research project emphasize the remarkable potential of Actinobacteria as reservoirs of novel compounds, particularly macrolactams. Our focused genome mining approach unveiled that the biosynthetic pathways responsible for macrolactam production are widely distributed among Actinobacteria. Simultaneously, this study reinforced the enduring importance of revisiting results, even after a significant period. Enhancing the quality of the genome was crucial for me to refine the pathway composition and, consequently, experimental setups related to molecular engineering. Likewise, it was essential to revisit the structural analysis of such a complex compound family once more structural data becomes available, allowing us to enhance our chemical and, consequently, biochemical understanding.

Copyright: Photo Saalfeld caves by Matthias Frank Schmidt kindly provided by Saalfelder Feengrotten und Tourismus GmbH.
Illustration of Pachycephala schlegelii adapted from kindly provided photo by Holger Teichman deposited at Macaulay Library at Cornell Lab (ML 204247051).
https://macaulaylibrary.org/asset/204247051
All other photos and illustrations were created by Elena Seibel.