Mass Spectrometry-driven Discovery of Natural Products

Fungi are categorized as one of the five eukaryotic kingdoms, members of which fungal species occupy a unique biological niche and coexist in fierce competition with other organisms, and are vast resources of bioactive compounds targeting both prokaryotic and eukaryotic cells. Many secondary metabolites from fungi have been used as drugs or as inspiration for the development of drugs for the past century. Additionally, certain species produce compounds that have industrial applications, including flavorings, pigments, and other food additives. On the other hand, fungal infection of crops and subsequent contamination by mycotoxins continue to be a health threat. Moreover, the rise in resistance to antibiotics, cancer chemotherapeutics and pesticides is a major threat to modern health care and agriculture, and without effective antibiotics, many routine surgical procedures would become more difficult than before. Fortunately, with the increasing availability of the whole-genome information and the application of detailed bioinformatics analysis, the real microbial “secondary metabolites landscape” has been revealed to be much larger than previously anticipated. In line with this recent genomic studies of species in Aspergillus section Flavi and Nigri have demonstrated that this is also the case for the many species in these groups of fungi even though they are already relatively well studied. Because of this huge hidden potential of biosynthetic gene clusters and their related secondary metabolites, methods for detection and analysis of both industrial valuable compounds and mycotoxins are vital. The aim of this PhD project has been to investigate the chemical potential of species in filamentous fungi, mainly focusing on species from genus Aspergillus, as well as an understudied Penicillium species. Analysis and the discovery of fungal natural products has been investigated based on the targeted analysis, investigation of biosynthetic intermediates by LC-HRMS, and molecular networking strategies, followed by the isolation and identification of significant promising compounds.

A targeted analysis approach was used to accelerate the identification and structural elucidation of newly identified natural products. The understudied species of Penicillium astrolabium was investigated for the production of the small peptide asperphenamate with antitumor activity. Initial one strain-many compounds (OSMAC) approach and targeted HRMS/MS analysis revealed two previously described and two novel asperphenamate analogues. By supplementing several proteogenic and non-proteogenic parasubstituted L-phenylalanine analogues, the biosynthesis of 22 new analogues was observed. Furthermore, the new analogues could readily be characterized by HRMS/MS, altogether demonstrating an unusual substrate uptake flexibility by the two non-ribosomal peptide synthetases (NRPSs) involved in asperphenamate biosynthesis.

Another goal of this PhD project has been to speed up the coupling of important bioactive compounds to their precursors and to explore their biosynthetic pathways. Thus, one part of the study focused on investigation of the nature of the initial precursor involved in the biosynthesis of yanuthone X from Aspergillus niger. A previous study characterized the gene cluster for yanuthone D, and demonstrated that yanuthones originates from the polyketide 6-MSA, encoded by YanA in A. niger. Surprisingly, a second series of yanuthones were discovered, including yanuthone X1 and X2, that are biosynthesized from an unknown precursor. We hypothesized that shikimic acid could be the source of the C6 core scaffold involved in biosynthesis of yanuthone X1 and its analogues. Indeed, feeding experiments with 13C labeled shikimic acid and subsequent HRMS analysis of A. niger and a yanAΔ mutant strains, revealed incorporation of the labeled shikimic acid into yanuthone X1 and X2. This result supported the hypothesis that yanuthone X1 originates from the shikimic acid pathway. Subsequent analysis by molecular networking using HRMS/MS data uncovered additional yanuthone X analogues. Altogether, the combination of isotope labeling and molecular networking proved very effective for detection of biosynthetic precursors and of two structurally related meroterpenoids, that both rely on several of the same tailoring enzymes for the late part of their biosynthesis. In another study, the compound calipyridone A was discovered from the rare fungus A. californicus. Next, bioinformatics analysis and gene deletion experiments were used to establish the function of each gene included in the predicted gene cluster, altogether allowing for characterization of the pathway leading to biosynthesis of calipyridone A. Importantly, calipyridone A is the first microbial 2-pyridone metabolite that was formed without a ring expansion reaction catalyzed by P450, like Calipyridone B and C were demonstrated to have antiviral activities against SARS-Covid 19.

As a major element of this PhD study, MS/MS based strategies were employed to conduct a comprehensive exploration of the secondary metabolite diversity in Aspergillus section Flavi, aiming at the discovery of "hidden" natural products especially from less investigated newly described species. More specifically, extracts from 23 taxonomically related fungal species in Aspergillus section Flavi were analyzed and dereplicated using both an in-house MS/MS library, as well as molecular networking, leading to the mapping of several known compound classes, as well as the discovery of new cyclopiazonic acid, fumifungin and tenuazonic acid producers. Additionally, this study revealed series-specific secondary metabolites, including a compound with a precursor mass of m/z 693.32, that could only be detected from A. caelatus, A. pseudotamarii and A. pseudocaelatus belonging to series Kitamyces. Subsequently, A. caelatus was chosen for the isolation and structural elucidation of novel bioactive compounds and putative chemical markers of m/z 693.32. Further investigation by feature-based molecular networking revealed additional structural isomers, leading to the characterization and isolation of six new compounds, named asperazine D-H and aspergillicin H. The asperazines D-H are dimers of two ditryptophenaline units, however the fact that different linkages are seen compared to other asperazines, indicates that a unique cytochrome P450s is required for their in A. caelatus. Finally, bioactivity-based molecular networking was used to characterize the bioactive compounds from a newly described species, A. sulphureoviridis. This led to the targeted isolation of six compounds including two new compounds named cladobotrin VII and cladobotrin VIII, along with tentative identification of other likely antibacterial compounds, that could however not be characterized due to very low production. Among the isolated compounds, F9775B showed anti-Bacillus subtilis and Dickeya solani activity with the IC₅₀ of 89.4 and 59.3 µg/mL respectively. Cladobotrin VIII showed anti-B. subtilis activity with IC50 of 95 µg/mL. Additionally, A. sulphureoviridis and B. subtilis were co-cultured and MALDI-TOF was applied to map the distribution of the produced compounds during the co-cultivation. This revealed the distribution and upregulation of several antifungal lipopeptides produced by B. subtilis, including surfactins, iturins, fengycins, and what appears to be an unidentified class of compounds.

Overall, the research of this PhD project has discovered several classes of bioactive fungal secondary metabolites, revealed the biosynthetic origins of several compounds, and highlighted the chemical diversity of Aspergillus section Flavi using different analytical approaches. This has included targeted mass spectrometry analyses, feeding experiments with isotope-labeled compounds, as well as molecular networking strategies. The applied approaches and obtained results have contributed new insights into the mass-spectrometry-based discovery of fungal secondary metabolites.