Correlative metabologenomics of 110 fungi reveals metabolite–gene cluster pairs

Caesar, Lindsay K.; Butun, Fatma Ayaloglu; Robey, Matthew T.; Ayon, Navid J.; Gupta, Raveena; Dainko, David; Bok, Jin Woo; Nickles, Grant; Stankey, Robert J.; Johnson, Don H.; Mead, David A.; Cank, Kristóf B.; Earp, Cody E.; Raja, Huzefa A.; Oberlies, Nicholas H.; Keller, Nancy P.; Kelleher, Neil L.

doi:10.1038/s41589-023-01276-8

Cited by 28 publications

(21 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to the increased demand for novel drugs, efforts that systematically link fungal SM bioactivity to BGCs are urgently needed. Various methods of correlating SMs to their corresponding BGCs have been undertaken such as targeted genome mining, heterologous expression, metabologenomics with gene cluster family (GCF) networking and correlation-based scoring 39 , and feature-based correlation methods utilizing genome- metabolome ontologies in bacterial species 40 . While these methods are continuing to improve, SM discovery is still far ahead of the characterization of the BGCs responsible for their biosynthesis.…”

Section: Resultsmentioning

confidence: 99%

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Riedling,

Walker,

Rokas

2023

Preprint

View full text Add to dashboard Cite

Fungal secondary metabolites (SMs) play a significant role in the diversity of ecological communities, niches, and lifestyles in the fungal kingdom. Many fungal SMs have medically and industrially important properties including antifungal, antibacterial, and antitumor activity, and a single metabolite can display multiple types of bioactivities. The genes necessary for fungal SM biosynthesis are typically found in a single genomic region forming biosynthetic gene clusters (BGCs). However, whether fungal SM bioactivity can be predicted from specific attributes of genes in BGCs remains an open question. We adapted previously used machine learning models for predicting SM bioactivity from bacterial BGC data to fungal BGC data. We trained our models to predict antibacterial, antifungal, and cytotoxic/antitumor bioactivity on two datasets: 1) fungal BGCs (dataset comprised of 314 BGCs), and 2) fungal (314 BGCs) and bacterial BGCs (1,003 BGCs); the second dataset was our control since a previous study using just the bacterial BGC data yielded prediction accuracies as high as 80%. We found that the models trained only on fungal BGCs had balanced accuracies between 51-68%, whereas training on bacterial and fungal BGCs yielded balanced accuracies between 61-74%. The lower accuracy of the predictions from fungal data likely stems from the small number of BGCs and SMs with known bioactivity; this lack of data currently limits the application of machine learning approaches in studying fungal secondary metabolism. However, our data also suggest that machine learning approaches trained on bacterial and fungal data can predict SM bioactivity with good accuracy. With more than 15,000 characterized fungal SMs, millions of putative BGCs present in fungal genomes, and increased demand for novel drugs, efforts that systematically link fungal SM bioactivity to BGCs are urgently needed.

show abstract

Section: Resultsmentioning

confidence: 99%

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Riedling,

Walker,

Rokas

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Therefore, GCFs with the same core genes often produce related families of SMs (35). Several GCFs were identified as sharing a common biosynthetic core with other GCFs suggesting that these GCFs were not entirely resolved (9). In these instances, we collapsed GCFs with identical cores into a composite GCF that we call a 'superfamily' and refer to original GCFs as 'Clans' to clarify their distinct accessory gene complements.…”

Section: Identifying Core Genes In Highly Conserved Ics Gcfsmentioning

confidence: 99%

“…We dereplicated ICS BGCs into GCFs (Supplementary Figure S1) (8)(9)(10) to clarify the biosynthetic potential that these clusters represent. Of the 3,800 ICS BGCs, 92.4% could be categorized into GCFs (Supplementary Table S6), while the remaining clusters were found only in a single genome.…”

Section: Ics Biosynthetic Diversity Across Taxonomic Ranksmentioning

confidence: 99%

“…Current genome-mining algorithms that predict SMproducing biosynthetic gene clusters (BGCs) in fungal genomes utilize the contiguous physical arrangement of genes to enable BGC predictions (5)(6)(7). This approach has successfully revealed millions of putative BGCs, many of which remain uncharacterized (8,9). However, these algorithms rely on prior knowledge of chemical-class defining "backbone" synthases/synthetases (e.g., nonribosomal synthetases, polyketide synthases, terpene synthases/cyclases), biasing results towards well-characterized classes of BGCs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mining for a New Class of Fungal Natural Products: The Evolution, Diversity, and Distribution of Isocyanide Synthase Biosynthetic Gene Clusters

Nickles

Oestereicher

Keller

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

The products of non-canonical isocyanide synthase (ICS) biosynthetic gene clusters (BGCs) have notable bioactivities that mediate pathogenesis, microbial competition, and metal-homeostasis through metal-associated chemistry. We sought to enable research into this class of compounds by characterizing the biosynthetic potential and evolutionary history of these BGCs across the Fungal Kingdom. We developed the first genome-mining pipeline to identify ICS BGCs, locating 3,800 ICS BGCs in 3,300 genomes. Genes in these clusters share promoter motifs and are maintained in contiguous groupings by natural selection. ICS BGCs are not evenly distributed across fungi, with evidence of gene-family expansions in several Ascomycete families. We show that the ICS dit1/2 gene cluster family (GCF), which was thought to only exist in yeast, is present in ~30% of all Ascomycetes, including many filamentous fungi. The evolutionary history of the dit GCF is marked by deep divergences and phylogenetic incompatibilities that raise questions about convergent evolution and suggest selection or horizontal gene transfers have shaped the evolution of this cluster in some yeast and dimorphic fungi. Our results create a roadmap for future research into ICS BGCs. We developed a website (https://isocyanides.fungi.wisc.edu/) that facilitates the exploration, filtering, and downloading of all identified fungal ICS BGCs and GCFs.

show abstract

“…Genomics and metabolomics studies, coupled with databases cataloging natural products and their corresponding biosynthetic gene clusters, have emerged as crucial components of this transformative phase. The affordability of these cutting-edge tools has increased their usage, enabling a wider community of researchers to incorporate them into their scientific pursuits. − Furthermore, innovative strategies that integrate multiple layers of information have emerged, offering new avenues for efficient natural products discovery.…”

mentioning

confidence: 99%

Integrating Taxonomic and Chemical Diversity of Mangrove-Associated Ascomycetes to Discover or Repurpose Bioactive Natural Products

Martínez-Aldino,

Rivera-Chávez,

Morales-Jiménez

2023

J. Nat. Prod.

View full text Add to dashboard Cite

Natural product reisolation is a bottleneck when discovering new bioactive chemical entities from nature. To overcome this issue, multi-informative approaches integrating several layers of data have been applied with promising results. In this study, integration of taxonomy, nontargeted metabolomics, and bioactivity information resulted in the selection of Scytalidium sp. IQ-074 and Diaporthe sp. IQ-053 to isolate new natural products active against hPTP1B1–400 and repurpose others as antibiotics. Strain IQ-074 was selected based on the hypothesis that investigating poorly studied and highly metabolic taxa could lead to the isolation of new chemical entities. A chemical investigation of IQ-074 resulted in the isolation of papyracillic acid A (14), 7-deoxypapyracillic acid A (15a and 15b), and linear polyketides scytalpolyols A–D (16–19). Compound 17 inhibited hPTP1B1–400 with a half-maximal inhibitory concentration of 27.0 ± 1.7 μM. Diaporthe sp. IQ-053 was selected based on its antibacterial properties against pathogenic strains. Its chemical investigation yielded dothiorelones A (20) and I (21), cytosporones B (22) and C (23), pestalotiopsone B (24), and diaporthalasin (25). Compounds 22 and 25 inhibited the growth of Staphylococcus aureus and Staphylococcus epidermidis 42R and moderately inhibited the growth of Acinetobacter baumannii A564, a pandrug-resistant bacterium.

show abstract

Correlative metabologenomics of 110 fungi reveals metabolite–gene cluster pairs

Cited by 28 publications

References 60 publications

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Predicting fungal secondary metabolite activity from biosynthetic gene cluster data using machine learning

Mining for a New Class of Fungal Natural Products: The Evolution, Diversity, and Distribution of Isocyanide Synthase Biosynthetic Gene Clusters

Integrating Taxonomic and Chemical Diversity of Mangrove-Associated Ascomycetes to Discover or Repurpose Bioactive Natural Products

Contact Info

Product

Resources

About