Harnessing Molecular and Bioactivity Network Analysis to Prioritize Antibacterial Compound Isolation From Ant-Associated Fungi

Abstract

Introduction: Antimicrobial resistance is a global public health problem that requires the development of new bioactive compounds. In this context, metabolomic analyses can expedite the research of fungal metabolites as a valuable resource.

Objectives: To investigate the metabolic profiles and isolate antibacterial compounds from micromycetes associated with Mexican cloud forest ants by utilizing network analysis of their chemical and bioactivity data.

Material and methods: 248 fungal strains isolated from six ant's species, soil of their anthills, and soil of the surroundings were evaluated for their in vitro inhibition growth of extensively drug-resistant Acinetobacter baumannii and hypervirulent Klebsiella pneumoniae; subsequently, their metabolites were dereplicated and analyzed by molecular networking and compound activity mapping from spectrometric data. Prioritization of some fungi for isolation of their major constituents was performed, and their structures were established by spectroscopic and spectrometric analysis and their bioactivity determined.

Results: From the fungal collection, 15 secondary metabolites (1-15) were dereplicated, and 10 compounds (16-25), including the new (E)-tridec-7-ene-3,5,6,10-tetraol (25), were isolated from Ascomycetes of Trichoderma, Cladosporium, and Clonostachys genera. Compounds 16-18 stood out for being bioactive. This study is the first report of antibacterial activity against A. baumannii for the tricyclic pyridin-2-ones deoxy-PF1140 (16) and PF1140 (17), with minimum inhibitory concentration of 50 μg/mL.

Conclusion: Network analysis and dereplication proved effective in bioprospecting for antibacterial compounds, offering valuable insights into the chemical diversity of cloud forest soil fungi and their potential applications. Moreover, this study broadens the knowledge of fungal secondary metabolites linked to leafcutter, fire, and warrior ants.

Keywords: Atta; Solenopsis; ant; anthill; antibacterial compound; bioactivity network; cloud forest fungi; molecular network.

FIGURE 1

Location of the collection points (A). Distribution of bioprospected morphotypes according to ant species and isolation point. Satellite image powered by Esri (https://www.arcgis.com/home/item.html?id=10df2279f9684e4a9f6a7f08febac2a9) Ants' images were taken and modified from AntWeb v8.106.1 (https://www.antweb.org) (B). Comparison of the antibacterial effect of morphotypes, one line connects the effect of the same morphotype (C). Top 10 values against Acinetobacter baumannii in a bar chart (positive control of colistin in water at 20 μg/mL). Asterisks mark morphotypes associated with ants (D). Violin plot of A. baumannii inhibition according to morphotype origin. The median is drawn with a continuous line and the quartiles with dotted lines. The medians per group do not differ statistically (Kruskal–Wallis test, p = 0.13; E).

FIGURE 2

Molecular networks of bioprospected morphotypes. Exclusive nodes according to the isolation point of their morphotype are colored. Networks with three or more nodes and a single color are enclosed with a dotted line (A). Solitary nodes represent 35.8% of the chemical diversity under study (B). The point chart shows m/z ratio of the precursor ions of all nodes in ascending order (C). Venn diagram of 3036 nodes forming the molecular networks, by morphotype origin and in parentheses the number of sampling points (D).

FIGURE 3

Compound activity mapping against Acinetobacter baumannii . This network has a clustering score > 0.1 (available value: −1 to 1) and an activity score > 0.25 (available value: 0–3). The blue square modes correspond to the morphotypes of interest. Isolated compounds were marked (ion detected [2M+H]⁺).

FIGURE 4

Dereplicated compounds and their associated nodes and molecular networks. The color of the nodes retains the code of Figure 2, each one contains the m/z ratio of the precursor ion, and the width of edges encodes the cosine of similarity. Dereplicated nodes are joined to the directly related ones by means of black edges. The image of the morphotypes corresponds to PD agar.

FIGURE 5

Morphotypes under chemical study and their isolation point. The morphotypes were grown on PD agar (A). Isolated compounds and their associated node in molecular networks. The color of the nodes retains the code of Figure 2, each node contains the m/z ratio of the precursor ion, and the edges vary in width according to the cosine similarity. In certain cases, the cosine value is indicated (B).

FIGURE 6

Key 2D NMR correlations of compound 25. Spectra available in Section S19.

FIGURE 7

Inhibition of Acinetobacter baumannii growth by 16 and 17 at different concentrations. Compounds were dissolved in DMSO. Positive control of colistin in water at 20 μg/mL inhibited growth by 99.57 ± 0.50%. Statistically significant differences are indicated (Student's t‐test, p < 0.05).