Chemical clues to infection: A pilot study on the differential secondary metabolite production during the life cycle of selected Cordyceps species

1 Chemical clues to infection: A pilot study on the differential secondary metabolite production during the life cycle of selected Cordyceps species Esteban Charria-Girón1,2* , Rita Toshe1* , Artit Khonsanit3, Noppol Kobmoo4, Papichaya Kwanthong4, Tatiana E. Gorelik5,6,7 , Janet Jennifer Luangsa-ard4, Sherif S. Ebada1,8 , Marc Stadler1,9 1 Department Microbial Drugs, Helmholtz Centre for Infection Research, Inhoffenstraße 7, 38124 Braunschweig, Germany 2 Bioinformatics Group, Wageningen University & Research, Droevendaalsesteeg 1, 6708 PB Wageningen, Netherlands 3 67 Moo 9 Ban Huai, Muang Nong Hang, Kuchinarai, Kalasin, Thailand 4 Plant Microbe Interaction Research Team, BIOTEC, National Science and Technology Development Agency (NSTDA) 111 Thailand Science Park, Phahonyothin Road, Khlong Nueng, Khlong Luang, Pathum Thani, 12120, Thailand 5 Department Structure and Function of Proteins, Helmholtz Centre for Infection Research (HZI), 38124 Braunschweig, Germany 6 Department Microbial Natural Products, Helmholtz-Institute for Pharmaceutical Research Saarland (HIPS), 66123 Saarbrücken, Germany 7 Ernst Ruska-Centre for Microscopy and Spectroscopy with Electrons (ER-C), Forschungszentrum Jülich, 52425 Jülich, Germany 8 Department of Pharmacognosy, Faculty of Pharmacy, Ain Shams University, 11566 Cairo, Egypt 9 Institute of Microbiology, Technische Universität Braunschweig, Spielmannstraße 7, 38106 Braunschweig, Germany Corresponding authors: Esteban Charria-Girón ([email protected]); Sherif S. Ebada ([email protected]); Marc Stadler ([email protected]) Copyright: © Esteban Charria-Girón et al. This is an open access article distributed under terms of the Creative Commons Attribution License (Attribution 4.0 International – CC BY 4.0). Research Article Abstract Cordyceps species are widespread entomopathogens and promising biocontrol agents that produce diverse secondary metabolites, yet the roles of these molecules during the infection process remain unclear. To interpret how fungal chemistry contributes to host colonization, we compared the metabolomes and virulence traits of two strains of phylogenetically distinct Cordyceps species (C. javanica and C. blackwelliae) and assessed their effects on beet armyworms (Spodoptera exigua). Virulence assays revealed species-dependent pathogenicity, with C. javanica showing the highest virulence. Combining untargeted metabolomics, feature-based molecular networking (FBMN), 3D electron-diffraction crystallography and comprehensive 1D/2D NMR, we gained insights into their metabolomic traits. For instance, C. javanica displayed notable beauveriolide diversity, including three previously undescribed derivatives (1–3), while C. blackwelliae produced mainly diketopiperazines in vitro. The FBMN results revealed putative beauveriolide analogs in the C. blackwelliae extracts, unlike the cadaver analysis, revealing beauvericins in infected corpses. Remarkably, the crude extracts obtained from authentic insect cadavers contained beauveriolides and beauvericins, providing in vivo chemical evidences of their production during infection for the first time. Moreover, bioassays with purified compounds showed that insecticidal activity cannot be attributed across all beauveriolides but depends on amino-acid composition, implying multifunctional roles beyond direct toxicity. Altogether, these results reveal context-dependent metabolic reprogramming and species-specific chemical strategies in entomopathogenic fungi, with implications for microbial ecology, host specificity, and the rational development of fungal biocontrol agents. The results of this study also give rise to the need for more intensified study on the chemical composition of the insect cadavers that are colonized by other entomopathogens. Academic editor: Irina Druzhinina Received: 20 September 2025 Accepted: 1 December 2025 Published: 12 December 2025 Citation: Charria-Girón E, Toshe R, Khonsanit A, Kobmoo N, Kwanthong P, Gorelik TE, Luangsa-ard JJ, Ebada SS, Stadler M (2025) Chemical clues to infection: A pilot study on the differential secondary metabolite production during the life cycle of selected Cordyceps species. IMA Fungus 16: e172651. https://doi. org/10.3897/imafungus.16.172651 IMA Fungus 16: e172651 (2025) DOI: 10.3897/imafungus.16.172651 * These authors contributed equaly. 2 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Key words: Cordycipitaceae, Entomopathogenic fungi, host-colonization, metabolomics, virulence Introduction The genus Cordyceps (Hypocreales) includes fungi renowned for their ability to parasitize insects, a unique trait defining their role as entomopathogens. Historically, fungi associated with insects and arachnids featuring conspicuous stromata were classified in this genus (Shrestha et al. 2014). However, successive taxonomic revisions based on molecular data have led to their reclassification into various families within the Hypocreales order with up to 23 different genera currently accepted within the Cordycipitaceae (Sung et al. 2007; Kepler et al. 2017; Shrestha et al. 2017; Wang et al. 2020). These revisions have significantly broadened our understanding of the diversity within these families, providing a robust foundation for the study of their ecology and derived applications. Preliminary correlations based on phylogenetic information could be found between certain genera and host/ substrate specificity, suggesting their ability to colonize and effectively target specific insect hosts (Kobmoo et al. 2023; Khonsanit et al. 2024). This unique parasitic behavior has been exploited to develop specific biocontrol agents in agriculture. Moreover, species of Cordyceps and Ophiocordyceps are of significant value in traditional Chinese medicine, where both the fungus and the dead insect host are consumed together. This duo mirrors the philosophical yin and yang concept, representing the balance between life and death and natural dualities that together restore harmony (Paterson 2008). These combinations have long been employed to treat illnesses, including respiratory and liver disorders, cardiovascular diseases, low libido, hyperlipidemia, and chronic kidney disease (Paterson 2008). In fact, currently the global market for these applications within traditional Chinese medicine is estimated to be worth up to USD 1.5 billion (Niego et al. 2023). Cordyceps species are known as prolific producers of bioactive secondary metabolites. These compounds might be involved in suppressing host defenses or manipulating host physiology, yet their exact role in the infection process remains poorly understood (Yin et al. 2020; Sun et al. 2022; Kobmoo et al. 2024). Despite the potential involvement of these metabolites in virulence, this aspect of Cordyceps chemical biology is still largely understudied. In contrast, these compounds might prove useful in the development of therapeutic treatments due to their multifaceted roles and diverse biological properties (Helaly et al. 2019; Toshe et al. 2024). The current pilot study describes the metabolome and virulence traits of C. javanica and C. blackwelliae, two strains of insect-associated fungi that were selected for their distinct ecological roles and potential as biocontrol agents. By combining state-of-the-art metabolomics, classical chemical screening, and virulence assays, we aimed to identify key metabolites involved in the infection process and evaluate their suitability for pest control applications. 3 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Materials and methods General experimental procedures Optical rotation was measured at 20 °C on an MCP-150 polarimeter (Anton-Paar Opto Tec GmbH, Seelze, Germany). UV/Vis spectra were recorded on a UV-2450 spectrophotometer (Shimadzu, Kyoto, Japan). NMR spectra were acquired using an Avance III 500 spectrometer (Bruker, Billerica, MA, USA; 1H NMR at 500 MHz, 13C NMR at 125 MHz), with samples in DMSO-d6. High-resolution electrospray ionization mass spectra (HR-ESI-MS) were acquired using an Agilent 1200 Infinity Series HPLC-UV system (Agilent Technologies, Santa Clara, CA, USA) with a C18 Acquity UPLC BEH column (50 × 2.1 mm, 1.7 μm; Waters, Milford, MA, USA). The mobile phases were: solvent A (H2O + 0.1% formic acid) and solvent B (acetonitrile + 0.1% formic acid). The gradient started at 5% B for 0.5 min, increasing to 100% B over 19.5 min, holding at 100% B for 5 min, with flow rate of 0.6 mL/min, with UV/Vis detection at 190–600 nm. The system was connected to a time-of-flight mass spectrometer (ESI-TOF-MS, maXis, Bruker, Billerica, MA, USA; scan range 100–2500 m/z, rate 2 Hz, capillary voltage 4500 V, dry temperature 200 °C). Fungal material Cordyceps blackwelliae (BCC37653) was isolated from a lepidopteran pupa found on the underside of a leaf in Khao Yai National Park, while C. javanica BCC 79245 (TBRC 7259) and BCC 82944 (TBRC 7262) were isolated from lepidopteran larvae in Nam Nao National Park and Wang Takrai Waterfalls, respectively. Pure cultures of C. blackwelliae and C. javanica have been previously deposited in the BIOTEC culture collection (BCC) (Mongkolsamrit et al. 2018; Helaly et al. 2019). Phylogenetic analyses based on ITS and tef sequences (GenBank accession numbers: ITS = PP709053, tef = PP735442) confirmed that BCC 37653 was nested with the type species of C. blackwelliae TBRC 7257 (Mongkolsamrit et al. 2018). Virulence assays Fungal virulence was evaluated against beet armyworm (Spodoptera exigua, Lepidoptera), using an established protocol at BIOTEC. The representative strains were grown on PDA at 25 °C until sporulation (ranging from one to four weeks). Spores were collected into 1 mL sterile water, counted, and then adjusted to a concentration of 108 spores/mL. For each assay, beet armyworms were injected with 3 µL of spore suspension (n = 30 insects per strain, divided into three independent replicates of 10 individuals each). Mortality was monitored daily for 7 days and categorized as total mortality (TM), any dead insect, or mycelium-associated mortality (MM), dead insects showing visible external mycelial outgrowth. Mortality rate was calculated per replicate as the number of dead insects (TM or MM) divided by the total number of insects in each replicate. To evaluate the virulence of purified compounds against beet armyworm larvae, beauveriolides I (4) and M (5) were selected based on availability, and cytochalasin D served as a positive control. Compounds were dissolved in DMSO and serially diluted to give concentrations ranging from 40 to 0.625 µg/mL. 4 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Larvae were injected with 0.5 µL of each dilution (n = 6 per concentration), maintained at 25 °C for one week, and monitored daily to assess mortality. Saline (8.5% NaCl) and DMSO served as negative controls. Metabolomics studies To evaluate the production of secondary metabolites, C. blackwelliae and C. javanica were cultured in yeast malt liquid medium (YM 6.3; 10 g/L malt extract, 4 g/L D-glucose, 4 g/L yeast extract, pH 6.3 before autoclaving). Strains were first grown on YM agar and after sufficient growth, five plugs (7 mm diameter) were transferred into flasks containing 100 mL of SMYA medium and incubated at 25 °C with shaking at 220 rpm for 7 days. For screening, 3 mL aliquots of each seed culture were added to 500 mL flasks containing 200 mL of production medium and incubated at 23 °C and 140 rpm. Glucose levels were monitored daily, and cultures were harvested three days post-glucose depletion. Supernatant and mycelium were separated by vacuum filtration and processed separately. The supernatant was extracted twice with ethyl acetate, the combined organic phases were evaporated under reduced pressure. The mycelia were soaked in acetone, ultrasonicated, filtered, and the obtained acetone solution was evaporated, dispersed in water, and processed in a similar manner to the supernatant. After the detection of diverse molecules, we focused on exploring beauveriolide diversity. Cordyceps javanica exhibited the highest yield and chemical diversity. Therefore, both C. javanica strains were cultured in five additional liquid media (ZM ½, Q6 ½, GG1, Supermalt, and MMK) and two solid media (BRFT and V+YES). Cultivation and extraction for the solid media followed the same procedures used for YM and liquid cultures, with solid cultures treated in the same manner as mycelia from liquid media as previously described (Helaly et al. 2019; Toshe et al. 2024). To investigate the presence of secondary metabolites in insect host corpses, a tissue segment was excised from the interface between the insect body and the emerging fruiting body or visible mycelia. The tissue was extracted twice with 1 mL of acetone:methanol (1:1) by vigorous shaking followed by ultrasonication. Combined extracts were filtered, dried under reduced pressure, and reconstituted in 50 µL of DMSO for analysis. All crude extracts were analyzed using ultrahigh-performance liquid chromatography coupled with diode array detection and ion mobility tandem mass spectrometry (UHPLC-DAD-IM-MS/MS) under previously established instrumental setting conditions and processed following our established protocols (Pfütze et al. 2024; Harms et al. 2024). Scaled-up cultivation of C. javanica and purification of beauveriolides Cultivation of C. javanica BCC 82944 was scaled up in GG1 (4 L). Mycelial and supernatant extracts showed similar LC-MS profiles and were combined (20 g). A liquid-liquid pre-separation was performed using n-heptane, methanol, and water. A 9:1 methanol-to-water solution was mixed with an equal volume of n-heptane in a separatory funnel, resulting in two immiscible phases and a precipitate in the methanol-water phase. The precipitate was collected by filtration (Fraction I, 6.0 g), then methanol-water phase corresponded to Fraction II (12.6 g), and the n-heptane phase to Fraction III (1.4 g). 5 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Fraction I was purified using preparative HPLC on a Büchi Pure C-850 FlashPrep system (Büchi Labortechnik GmbH; Essen, Germany) using a Gemini C18 column (250 × 50 mm, 10 μm; Phenomenex®, Torrance, CA, USA). The mobile phase consisted of deionized water with 0.1% formic acid (solvent A) and acetonitrile (MeCN) with 0.1% formic acid (solvent B) at pH 2.5. The flow rate was set at 30 mL/min, and UV detection was recorded at 210, 254, 300, and 350 nm. For each run, 300 mg aliquot was injected (10 runs) and the separation was performed using a gradient from 30% to 55% B in 12 minutes, 55% to 85% B in 30 minutes, from 85% to 100% B in 10 minutes and ended with an isocratic elution at 100% B for 10 min. A total of six compounds were obtained: 3 (1.2 mg, tR = 35 min), 5 (5.3 mg, tR = 42 min), 1 (3.2 mg, tR = 45 min), 4 (14.1 mg, tR = 47 min), 6 (6.3 mg, tR = 56 min) and 2 (1.7 mg, tR = 54 min). Fractions II was purified using the same system and column but with the following gradient: 40% to 100% in 60 minutes and ended with isocratic elution at 100% B for 10 min. This procedure afforded 7 (6.9 mg, tR = 62 min), 4 (712.7 mg, tR = 42 min) and 6 (19.9 mg, tR = 56 min). Scaled-up cultivation of C. blackwelliae and purification of diketopiperazines To investigate the metabolites of C. blackwelliae BCC 37653, a scale-up cultivation was conducted on YM medium. Five mycelial plugs of fully-grown YMA plates were transferred to 500-mL flasks containing 200 mL of YM medium (10 L in total), which were incubated at 23 °C and 140 rpm. Glucose was monitored daily, and cultures were harvested three days after glucose depletion. Mycelia and supernatant were extracted following the same protocol used for the C. javanica scale-up, yielding 200 and 300 mg of crude extract for the mycelia and supernatant, respectively. Crude extracts were purified on a Gilson PLC 2250 preparative HPLC system (Gilson, Middleton, WI, USA) using a Gemini C18 column (250 × 50 mm, 10 μm; Phenomenex®, Torrance, CA, USA). Mobile phases and detection parameters were the same as above despite using the following gradient: from 35% B to 100% B in 75 minutes and ended with isocratic elution at 100% B for 10 minutes. This resulted in the isolation of compound 10 (3.2 mg, tR = 26 min), 9 (2.8 mg, tR = 27 min), 8 (5.9 mg, tR = 28 min), 12 (0.48 mg, tR = 47 min) and 11 (0.57 mg, tR = 52 min). Beauveriolide Q (1): White powder; UV-Vis (MeOH): λmax 225; HR-ESI-MS: m/z 527.3226 [M+H]+ (calcd. 527.3228 for C29H43N4O5+), 549.3046 [M+Na]+ (calcd. 549.6574 for C29H42N4NaO5+); tR = 11.64 min (LR-ESI-MS); C29H42N4O5 (526.32 g mol−1). Beauveriolide R (2): White powder; UV-Vis (MeOH): λmax 226; NMR data (1H: 500 MHz, 13C: 125 MHz, DMSO-d6) see Table 1; HR-ESI-MS: m/z 555.3540 [M+H]+ (calcd. 555.3541 for C31H47N4O5+), 577.3359 [M+Na]+ (calcd. 577.7105 for C31H46N4NaO5+); tR = 13.19 min (LR-ESI-MS); C31H46N4O5 (554.35 g mol−1). Beauveriolide S (3): Colorless oil; UV-Vis (MeOH): λmax 225; HR-ESI-MS: m/z 412.2809 [M+H]+ (calcd. 412.2806 for C21H38N3O5+), 434.2625 [M+Na]+ (calcd. 434.5254 for C21H37N3NaO5+); tR = 9.78 min (LR-ESI-MS); C21H37N3O5 (411.27 g mol−1). 6 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Single crystal structure via 3D electron diffraction of beauveriolide Q Dry material was suspended in n-hexane and a drop of the suspension was placed onto a holey-carbon-coated copper TEM grid. Grids were air-dried and excess solvent soaked up using filter paper. Grids were transferred into a TEM at room temperature and cooled to liquid nitrogen temperature directly in the TEM vacuum. Electron diffraction data were collected using a Thermo Fisher GLACIOS TEM operating at 200 kV employing the EPU-D (Thermo Fisher) Table 1. 1D (1H and 13C) NMR data of 2. pos. δC a type δH b (multi, J[Hz]) HDA 1 170.2, CO 2 35.4, CH2α 2.31 dd (14.0, 8.8) β 2.44 dd (14.0, 4.4) 3 75.8, CH 4.84 dt (10.2, 5.3) 4 34.8, CH 2.07 (m) 5 30.8, CH2α 1.02 m; β 1.33 m 6 26.5, CH2α 1.16 m; β 1.31 m 7 29.0, CH21.23 overlapped 8 31.2, CH21.23 overlapped 9 22.1, CH21.24 overlapped 10 14.0, CH30.84 t (6.9) 11 15.5, CH30.80 d (6.9) Trp 1 171.4, CO 2 56.0, CH 4.20 q (7.7) 3 25.7, CH2α 3.03 dd (14.5, 8.0) β 3.11 dd (14.5, 7.3) 4 109.8, C 4a 127.1, C 5 118.3, CH 7.50 d (7.5) 6 118.4, CH 6.97 t (7.5) 7 121.0, CH 7.06 t (7.5) 8 111.4, CH 7.32 d (7.5) 8a 136.1, C 9-NH– 10.86 d (2.5) 10 123.6, CH 7.11 d (2.5) NH– 8.42 d (7.6) Ala 1 170.8, CO 2 48.4, CH 3.90 p (6.9) 3 15.6, CH31.12 d (6.9) NH– 8.33 d (7.4) Leu 1 169.6, CO 2 52.0, CH 4.39 q (8.0) 3 40.8, CH21.43 t (7.4) 4 24.4, CH 1.49 m 5 22.0, CH30.85 d (6.5) 6 22.2, CH30.87 d (6.5) NH– 7.19 d (9.2) Measured in DMSO-d6 at a 125 MHz for 13C and b 500 MHz for 1H. 7 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection module. Diffraction patterns were recorded in nanodiffraction mode with an effective probe beam size on the sample of 1 micron. Data were recorded in MRC format and converted to 16-bit TIFFs using the MRC2TIFF (10.5281/zenodo.7936068). TIFF sequences were then processed in PETS2 (Palatinus et al. 2019). The best-performing dataset was selected for structure analysis. Unit cell parameters were: a = 16.5864 Å, b = 5.1703 Å, c = 17.3397 Å, α = 90.000°, β = 94.011°, γ = 90.000°, with a volume of 1483 Å, corresponding to a monoclinic unit cell. Extinctions corresponding to a 21 screw axis along the b-axis were detected, with no further extinctions observed, indicating the P21 space group (No. 4). The expected molecular volume was 719.74 ų (Hofmann 2002), corresponding to two molecules in the unit cell. The structure was solved using SHELXD and kinematically refined with SHELXL (Sheldrick 2015). The absolute configuration was determined during the dynamical refinement in JANA (Palatinus et al. 2015; Petrǐcěk et al. 2023). Further crystallographic details are provided in the Suppl. material 1. Antimicrobial and cytotoxic assays Serial dilution assays were performed in 96-well microtiter plates to determine the minimum inhibitory concentrations (MICs) against a panel of yeasts, filamentous fungi, and bacteria, as well as the half-maximal inhibitory concentrations (IC50) against human cell lines, following our previously described protocol (Charria-Girón et al. 2023). Results and discussion Virulence and chemical traits from studied Cordyceps strains The virulence of three Cordyceps strains was evaluated against the beet armyworm (Spodoptera exigua, Lepidoptera). Cordyceps blackwelliae and C. javanica strains were originally derived from a Lepidoptera pupae and larvae, respectively. Two types of insect mortality were considered: (TM) total mortality encompassing all deceased individuals, and (MM) mycelium-associated mortality representing dead insects exhibiting visible mycelia. This distinction was made to identify distinct virulence strategies. Entomopathogens are known to produce secondary metabolites, particularly mycotoxins (Schrank and Vainstein 2010; Cito et al. 2016), inside their host to cause death. Fungal growth and toxin production must be carefully balanced, as premature host death can prevent accumulation of nutrients required for development and sporulation (Boomsma et al. 2014). After insects are colonized and dead, fungi sporulate and disperse, processes that require favorable conditions and lytic activity to breach the cuticle from within (Charnley 2003). Therefore, visible mycelial outgrowth might indicate moderate toxin production combined with tissue-degrading enzymatic activity. Conversely, the absence of external mycelia suggests rapid, systemic killing driven by metabolites acting at the cellular level. From the two C. javanica strains, BCC 82944 was the most virulent, achieving 100% TM by day 3, while BCC 79245 caused an average TM of 80%, with both causing MM only around 20 to 40% (Fig. 1A). Cordyceps blackwelliae displayed an intermediate virulence with TM up to around 90% by day 3, but failed to produce visible mycelial 8 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection outgrowth even after seven days (Fig. 1B). The evaluated strains caused high TM but differed in external mycelial development, with C. javanica seemingly producing more visible mycelia than C. blackwelliae. Observations of early host death suggest the production of active and host-targeted molecules, while differences in mycelial development likely reflect species-specific differences in metabolite repertoires and infection strategies. Building on the hypothesis that secondary metabolites contribute to host infection, we decided to investigate their metabolome. However, it is noteworthy that metabolites produced during infection may differ from those observed under axenic cultivation. We cultured each fungus in yeast malt (YM) liquid medium until glucose was depleted. The obtained crude extracts were analyzed using ultrahigh-performance liquid chromatography coupled with diode array detection and ion mobility tandem mass spectrometry (UHPLC-DAD-IM-MS/ MS), and raw data were pre-processed with MetaboScape (Charria-Girón et al. 2023). Features detected in blanks were excluded, and the resulting features were dereplicated against our in-house spectral library. While C. javanica was expected to produce beauveriolide-type cyclodepsipeptides and asperfurans (Helaly et al. 2019), C. blackwelliae was to the best of our knowledge never studied for secondary metabolite production. Figure 1. Virulence of Cordyceps spp., C. blackwelliae BCC 37653, C. javanica BCC 79245, and BCC 82944 against the beet armyworm (Spodoptera exigua, Lepidoptera) observed over a 7-day period. A Relative survival considering total mortality (TM). B Relative survival considering mycelium-associated mortality (MM). Error bars denote the standard deviation across three independent biological replicates. C From left to right: larva turned into pupa, deceased insect without visible mycelia, and deceased insect with visible mycelia. 9 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection To expand our understanding of the chemical space produced by both Cordyceps species, natural product classes were predicted de novo from MS/MS spectra using CANOPUS (Dührkop et al. 2021; Valencia-Revelo et al. 2025). Hierarchical clustering analysis integrating CANOPUS annotations revealed clear differences in the metabolome of C. blackwelliae and C. javanica (Fig. 2A). While alkaloidand fatty-acid-like features were the most abundant in extracts from C. blackwelliae, compounds classified as amino acids and peptides, including beauveriolides, were the major metabolites in C. javanica. Beauveriolides I, Jb, and L, annotated by comparison with authentic standards, were not detected in C. blackwelliae. Our feature-based molecular networking (FBMN) analysis revealed a broader diversity of beauveriolides beyond annotations in both species. Cordyceps blackwelliae produced distinct analogs that clustered together within the same molecular family (MF) as the annotated beauveriolides. This suggests that, although the metabolome of C. blackwelliae is more diversified than that of C. javanica, beauveriolides production is conserved between the two species but with distinct amino acid rearrangements. Manual inspection supported by ModiFinder (Shahneh et al. 2024), suggested the presence of tyrosine-containing analogs in C. blackwelliae, exemplified by two features with neutral masses of 579.3588 and 607.3499 Da, which likely denote the substitution of alanine with tyrosine in beauveriolides I and L, respectively. The detection of beauvericins exclusively in C. blackwelliae suggests a broader chemical arsenal than C. javanica, which intriguingly correlates with the distinct virulence traits observed. Beauveriolides, which have been mainly characterized in Beauveria bassiana, are known to play a role in the early pathogenesis (Kim et al. 2024; Apirajkamol et al. 2025). The conservation of these metabolites in both Cordyceps species supports their likely involvement during infection, whereas the broader metabolome differentiation possibly reflects species-specific virulence mechanisms or host specialization strategies, also considering the phylogenetic distance between both taxa. Targeted Isolation and Structure Elucidation of Compounds We hypothesize that beauveriolides are the authentic metabolites prevalent during infection, because diketopiperazines are broadly produced across various fungal lineages and are not exclusive to entomopathogens, making their role in infection less likely (Ma et al. 2016). Since both C. javanica strains produced beauveriolides in higher titers than C. blackwelliae, we systematically evaluated beauveriolide production in several media for BCC 82944 and BCC 79245. Notably, C. javanica BCC 82944 exhibited a broader diversity of beauveriolides, including features with previously unreported molecular formulas that suggested putatively novel congeners, and which could be obtained from the GG1 medium in higher yields. Although C. blackwelliae predominantly produced diketopiperazines in vitro, we scaled up its cultivation in YM medium to comprehensively profile its secondary metabolome. Chemical investigation of the crude extracts afforded seven beauveriolide congeners including three previously undescribed from C. javanica, while the crude extract of C. blackwelliae afforded three diketopiperazine and two β-carboline derivatives (Fig. 3). 16 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Figure 6. A Evaluation of the virulence of beauveriolides I (4) and M (5) produced by C. javanica against the beet armyworm (Spodoptera exigua, Lepidoptera) over a 7-day period, considering the relative survival rate based on total mortality (TM). Cytochalasin D served as the positive control, while saline solution (8.5% NaCl) and DMSO were used as negative controls. B HPLC-UV/Vis chromatograms (210 nm) of the crude extracts obtained from the insect host corpses from C. javanica strains MY11508 and MY10919, and C. blackwelliae strain MY04954 herbarium material. Isolated metabolites are indicated by bold compound numbers. Putative beauveriolide derivatives are labeled as BV followed by their corresponding molecular weight. Similarly, detected beauvericins are labeled each with their name and molecular weight. corpses provides direct in-host chemical evidence that secondary metabolites are produced during natural infection, and bioassays with purified compounds showed that insecticidal activity is highly structure-dependent, with single amino-acid substitutions switching beauveriolides from inactive to lethal. Together, these results support a model in which metabolic 17 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection reprogramming, together with structural diversification of cyclodepsipeptides, triggers species-specific virulence strategies. This chemical plasticity has practical implications for biocontrol development, identifying active effector molecules and safer biocontrol agents, as well as advancing our understanding of host specificity and fungal evolution. Moving forward, broader taxonomic sampling and integration of genomic/transcriptomic data, together with spatially resolved metabolite mapping, will be essential to link biosynthetic regulation to ecological function and to translate these insights into ecologically informed biocontrol solutions. Last but not least, the current study also points toward another interesting phenomenon that goes back to the use of “cordicipitoid” ascomycetes like Cordyceps and Ophiocordyceps in Traditional Asian Medicine, where the “herbal” drugs traditionally used consisted of both, the fungal stroma and the insect that the fungus had colonized. From reports in the literature, we can be fairly sure that the metabolite profiles of stromata are very different from those encountered in the cultures. Once we know better what the insect carcasses contain, it may be possible to interpret the data on the efficacy. For instance, stromata of Ophiocordyceps sinensis and other species are sold on the market even as OTC drugs, but they are mainly known to contain cordycepin as an active principle (cf. Liu et al. 2015). The mycelial cultures of these fungi are known to contain different major metabolites, but the current study is the first to investigate the host carcasses that are also supposed to be part of the TCM drugs (see discussion in Paterson 2008; Kinjo and Zang 2001). Acknowledgments The authors wish to thank Dr. Kirsten Harmrolfs, Esther Surges and Aileen Golasch for recording the NMR spectra as well as HR-MS data, Ulrike Beutling for the metabolomics analyses, Wera Collisi for conducting the bioassays, Silke Reinecke for her advisory assistance, Wasana Noisripoom for acquiring the sequencing data of the studied strains, Nopporn Chutiwitoonchai for conducting the antiviral screening, and Natalia Andrea Llanos-López for carrying out the nematode assay. Additional information Conflict of interest The authors have declared that no competing interests exist. Ethical statement No ethical statement was reported. Use of AI No use of AI was reported. Adherence to national and international regulations All the fungal strains used in this study have been legally obtained, respecting the Convention on Biological Diversity (Rio Convention). 18 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Funding This work benefited from the European Union’s Horizon 2020 Research and Innovation Staff Exchange program (MSCA-RISE grant no. 101008129, acronym: MYCOBIOMICS), the Southeast Asia–Europe Joint Funding Scheme (SEA–Europe Grant number JFS20ST-127, Acronym: Antiviralfun), and the Grant no. P21-50844, from the National Science and Technology Development Agency (NSTDA). Esteban Charria-Girón was supported by the DAAD stipend 57694196 Forschungsstipendien für Doktorandinnen und Doktoranden (2024). Rita Toshe received a grant from the Bischöfliche Stiftung Cusanuswerk. The Alexander von Humboldt (AvH) Foundation is gratefully acknowledged for granting Sherif S. Ebada the Georg-Forster Fellowship for Experienced Researchers Stipend (Ref 3.4-1222288-EGY-GF-E). Author contributions E.C.-G. and S.S.E. designed the research; E.C.G., R.T., A.K., P.K., T.E.G., and S.S.E conducted experimental procedures and analyzed data; E.C.G and S.S.E. wrote the first draft. M.S. and J.J.L. acquired funding and coordinated the project. All authors have read and agreed to the published version of the manuscript. Author ORCIDs Esteban Charria-Girón https://orcid.org/0000-0002-3823-1416 Rita Toshe https://orcid.org/0009-0005-9068-5430 Artit Khonsanit https://orcid.org/0009-0009-0931-4288 Noppol Kobmoo https://orcid.org/0000-0002-3261-9970 Tatiana E. Gorelik https://orcid.org/0000-0002-0911-2039 Janet Jennifer Luangsa-ard https://orcid.org/0000-0001-6801-2145 Sherif S. Ebada https://orcid.org/0000-0002-2753-0031 Marc Stadler https://orcid.org/0000-0002-7284-8671 Data availability All of the data that support the findings of this study are available in the main text or Supplementary Information. References Apirajkamol N, Mainali B, Taylor PW et al. (2025) Characterisation of the pathogenicity of Beauveria sp. and Metarhizium sp. fungi against the fall armyworm, Spodoptera frugiperda (Lepidoptera: Noctuidae). Agriculture 15: 170. https://doi.org/10.3390/agriculture15020170 Bérubé C, Barbeau X, Cardinal S et al. (2017) Interfacial supramolecular biomimetic epoxidation catalysed by cyclic dipeptides. Supramolecular Chemistry 29: 330–349. https://doi.org/10.1080/10610278.2016.1236197 Boomsma JJ, Jensen AB, Meyling NV et al. (2014) Evolutionary interaction networks of insect pathogenic fungi. Annual Review of Entomology 59: 467–485. https://doi. org/10.1146/annurev-ento-011613-162054 Charnley AK (2003) Fungal pathogens of insects: Cuticle degrading enzymes and toxins. Advances in Botanical Research 40: 241–321. https://doi.org/10.1016/S00652296(05)40006-3 19 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Charria-Girón E, Marín-Felix Y, Beutling U et al. (2023) Metabolomics insights into the polyketide-lactones produced by Diaporthe caliensis sp. nov., an endophyte of the medicinal plant Otoba gracilipes. Microbiology Spectrum 11(6): e0274323. https://doi.org/10.1128/spectrum.02743-23 Cito A, Barzanti GP, Strangi A et al. (2016) Cuticle-degrading proteases and toxins as virulence markers of Beauveria bassiana (Balsamo) Vuillemin. Journal of Basic Microbiology 56: 941–948. https://doi.org/10.1002/jobm.201600022 Dhrkop K, Nothias LF, Fleischauer M et al. (2021) Systematic classification of unknown metabolites using high-resolution fragmentation mass spectra. Nature Biotechnology 39: 462−471. https://doi.org/10.1038/s41587-020-0740-8 Faini F, Catillo M, Torres R (1978) A new β-carboline alkaloid from Vestia lycioides. Phytochemistry 17: 338. https://doi.org/10.1016/S0031-9422(00)94190-6 Gemmi M, Mugnaioli E, Gorelik TE et al. (2019) P. 3D electron iffraction: The nanocrystallography revolution. ACS Central Science 5: 1315–1329. https://doi.org/10.1021/ acscentsci.9b00394 Gurung K, Šimek P, Jegorov Jr A et al. (2024) Structure and absolute configuration of natural fungal product beauveriolide I, isolated from Cordyceps javanica, determined by 3D electron diffraction. Acta Crystallographica Section C 80: 56–61. https://doi. org/10.1107/S2053229624001359 Harms K, Charria-Girón E, Stchigel AM et al. (2024) Reaping the chemical diversity of Morinagamyces vermicularis using feature-based molecular networking. Journal of Natural Products 87: 2335–2342. https://doi.org/10.1021/acs.jnatprod.4c00654 Helaly SE, Kuephadungphan W, Phainuphong P et al. (2019) Pigmentosins from Gibellula sp. as antibiofilm agents and a new glycosylated asperfuran from Cordyceps javanica. Beilstein Journal of Organic Chemistry 15: 2968–2981. https://doi.org/10.3762/ bjoc.15.293 Hiroshi K, Satoshi O, Takashi T et al. (2010) JP2010126454A. https://worldwide.espacenet.com/patent/search/family/042225663/publication/JP2010126454A?q=pn%3DJP2010126454A Hofmann DWM (2002) Fast estimation of crystal densities. Acta Crystallographic Section B 58: 489−493. https://doi.org/10.1107/S0108768101021814 Jegorov A, Paizs B, Kuzma M et al. (2004) Extraribosomal cyclic tetradepsipeptides beauveriolides: profiling and modeling the fragmentation pathways. Journal of Mass Spectrometry 39: 949–960. https://doi.org/10.1002/jms.674 Jegorov A, Sedmera P, Matha V et al. (1994) Beauveriolides L and La from Beauveria tenella and Paecilomyces fumosoroseus. Phytochemistry 37: 1301–1303. https:// doi.org/10.1016/S0031-9422(00)90402-3 Kepler RM, Luangsa-Ard JJ, Hywel-Jones NL et al. (2017) A phylogenetically-based nomenclature for Cordycipitaceae (Hypocreales). IMA Fungus 8: 335–353. https://doi. org/10.5598/imafungus.2017.08.02.08 Khonsanit A, Thanakitpipattana D, Mongkolsamrit S et al. (2024) A phylogenetic assessment of Akanthomyces sensu lato in Cordycipitaceae (Hypocreales, Sordariomycetes): introduction of new genera, and the resurrection of Lecanicillium. Fungal Systematics and Evolution 14: 271–305. https://doi.org/10.3114/fuse.2024.14.17 Kim JC, Hwang IM, Kim HM et al. (2024) Rapid analysis of insecticidal metabolites from the entomopathogenic fungus Beauveria bassiana 331R using UPLC-Q-Orbitrap MS. Mycotoxin Research 40: 123–132. https://doi.org/10.1007/s12550-023-00509-y 20 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Kinjo N, Zang M (2001) Morphological and phylogenetic studies on Cordyceps sinensis distributed in southwestern China. Mycoscience 42(6): 567–574. https://doi. org/10.1007/BF02460956 Klar PB, Krysiak Y, Xu H et al. (2023) Accurate structure models and absolute configuration determination using dynamical effects in continuous-rotation 3D electron diffraction data. Nature Chemistry 15: 848–855. https://doi.org/10.1038/s41557-02301186-1 Kobmoo N, Mongkolsamrit S, Khonsanit A et al. (2024) Integrative taxonomy of Metarhizium anisopliae species complex, based on phylogenomics combined with morphometrics, metabolomics, and virulence data. IMA Fungus 15: 30. https://doi. org/10.1186/s43008-024-00154-9 Kobmoo N, Tasanathai K, Araújo JPM et al. (2023) New mycoparasitic species in the genera Niveomyces and Pseudoniveomyces gen. nov. (Hypocreales: Cordycipitaceae), with sporothrix-like asexual morphs, from Thailand. Fungal Systematics and Evolution 12: 91–110. https://doi.org/10.3114/fuse.2023.12.07 Kuzma M, Jegorov A, Kacer P et al. (2001) Sequencing of new beauveriolides by high-performance liquid chromatography and mass spectrometry. Journal of Mass Spectrometry 36: 1108–1115. https://doi.org/10.1002/jms.213 Liu Y, Wang J, Wang W et al. (2015) The chemical constituents and pharmacological actions of Cordyceps sinensis. Evidence‐Based Complementary and Alternative Medicine 2015(1): 575063. https://doi.org/10.1155/2015/575063 Ma YM, Liang XA, Kong Y et al. (2016) Structural diversity and biological activities of indole diketopiperazine alkaloids from fungi. Journal of Agricultural and Food Chemistry 64: 6659–6671. https://doi.org/10.1021/acs.jafc.6b01772 Mongkolsamrit S, Noisripoom W, Thanakitpipattana D et al. (2018) Disentangling cryptic species with isaria-like morphs in Cordycipitaceae. Mycologia 110: 230–257. https://doi.org/10.1080/00275514.2018.1446651 Namatame I, Tomoda H, Ishibashi S et al. (2004) Antiatherogenic activity of fungal beauveriolides, inhibitors of lipid droplet accumulation in macrophages. Proceedings of the National Academy of Sciences 101: 737–742. https://doi.org/10.1073/ pnas.0307757100 Naveen B, Mudiraj A, Khamushavalli G et al. (2016) Concise total synthesis of water soluble metatacarboline A, C, D, E and F and its anticancer activity. European Journal of Medicinal Chemistry 113: 167–178. https://doi.org/10.1016/j.ejmech.2016.02.033 Niego AGT, Lambert C, Mortimer PE et al. (2023) The contribution of fungi to the global economy. Fungal Diversity 121: 95–137. https://doi.org/10.1007/s13225-02300520-9 Palatinus L, Brázda P, Jelínek M et al. (2019) Specifics of the data processing of precession electron diffraction tomography data and their implementation in the program PETS2.0. Acta Crystallographica Section B 75: 512–522. https://doi.org/10.1107/ S2052520619007534 Palatinus L, Petrǐcěk V, Corre CA (2015) Structure refinement using precession electron diffraction tomography and dynamical diffraction: theory and implementation. Acta Crystallographica Section A 71: 235−244. https://doi.org/10.1107/ S2053273315001266 Paterson RR (2008) Cordyceps – A traditional Chinese medicine and another fungal therapeutic biofactory? Phytochemistry 69: 1469–1495. https://doi.org/10.1016/j. phytochem.2008.01.027 21 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Petrǐcěk V, Palatinus L, Plasǐl J et al. (2023) Jana2020 – a new version of the crystallographic computing system Jana. Zeitschrift fr Kristallographie - Crystalline Materials 238: 271−282. https://doi.org/10.1515/zkri-2023-0005 Pfütze S, Charria-Girón E, Schulzke E et al. (2024) Depicting the chemical diversity of bioactive meroterpenoids produced by the largest organism on earth. Angewandte Chemie International Edition 63: e202318505. https://doi.org/10.1002/anie.202318505 Schrank A, Vainstein MH (2010) Metarhizium anisopliae enzymes and toxins. Toxicon 56: 1267–1274. https://doi.org/10.1016/j.toxicon.2010.03.008 Shahneh MRZ, Strobel M, Vitale GA et al. (2024) ModiFinder: Tandem mass spectral alignment enables structural modification site localization. Journal of the American Society of Mass Spectrometry 35: 2564–2578. https://doi.org/10.1021/ jasms.4c00061 Sheldrick GM (2015) Crystal structure refinement with SHELXL. Acta Crystallographica section C 71: 3−8. https://doi.org/10.1107/S2053229614024218 Shi T, Li YJ, Wang ZM et al. (2023) New pyrroline isolated from antarctic krill-derived actinomycetes Nocardiopsis sp. LX-1 combining with molecular networking. Marine Drugs 21: 127. https://doi.org/10.3390/md21020127 Shrestha B, Sung GH, Sung JM (2017) Current nomenclatural changes in Cordyceps sensu lato and its multidisciplinary impacts. Mycology 8: 293–302. https://doi.org/10.1 080/21501203.2017.1386242 Shrestha B, Tanaka E, Han JG et al. (2014) A brief chronicle of the genus Cordyceps Fr., the oldest valid genus in Cordycipitaceae (Hypocreales, Ascomycota). Mycobiology 42: 93–99. https://doi.org/10.5941/MYCO.2014.42.2.93 Simon G, Bérubé C, Voyer N et al. (2019) Anti-biofilm and anti-adherence properties of novel cyclic dipeptides against oral pathogens. Bioorganic & Medicinal Chemistry 27: 2323–2331. https://doi.org/10.1016/j.bmc.2018.11.042 Sun Y, Hong S, Chen H et al. (2022) Production of helvolic acid in Metarhizium contributes to fungal infection of insects by bacteriostatic inhibition of the host cuticular microbiomes. Microbiology Spectrum 10: e0262022. https://doi.org/10.1128/spectrum.02620-22 Sung GH, Sung J-M, Hywel-Jones NL et al. (2007) A multi-gene phylogeny of Clavicipitaceae (Ascomycota, Fungi): Identification of localized incongruence using a combinational bootstrap approach. Molecular Phylogenetics and Evolution 44: 1204–1223. https://doi.org/10.1016/j.ympev.2007.03.011 Tehan R, Mattos D, Arai T et al. (2025) Paraisariamides, cycloheptapeptide toxins from entomopathogenic fungi that inhibit total protein synthesis. ChemRxiv. https://doi. org/10.26434/chemrxiv-2025-x4f35 Toshe R, Charria-Girón E, Khonsanit A et al. (2024) Bioprospection of tenellins produced by the entomopathogenic fungus Beauveria neobassiana. Journal of Fungi 10: 69. https://doi.org/10.3390/jof10010069 Valencia-Revelo D, Charria-Girón E, Schmidt K et al. (2025) Unconventional semi-solid cultivation enhances cytochalasins production by the Colombian fungus Xylaria sp. CM-UDEA-H199. BMC Biotechnology 25: 57. https://doi.org/10.1186/s12896-02500978-2 Wang YB, Wang Y, Fan Q et al. (2020) Multigene phylogeny of the family Cordycipitaceae (Hypocreales): new taxa and the new systematic position of the Chinese cordycipitoid fungus Paecilomyces hepiali. Fungal Diversisty 103: 1–46. https://doi.org/10.1007/ s13225-020-00457-3 22 IMA Fungus 16: e172651 (2025), DOI: 10.3897/imafungus.16.172651 Esteban Charria-Girón et al.: Cordyceps chemical traits of infection Xu YJ, Luo F, Gao Q et al. (2015) Metabolomics reveals insect metabolic responses associated with fungal infection. Analytical and Bioanalytical Chemistry 407: 4815–4821. https://doi.org/10.1007/s00216-015-8648-8 Yin Y, Chen B, Song S et al. (2020) Production of diverse beauveriolide analogs in closely related fungi: a rare case of fungal chemodiversity. mSphere 5: e00667-20. https://doi.org/10.1128/msphere.00667-20 Supplementary material 1 Additional information Authors: Esteban Charria-Girón, Rita Toshe, Artit Khonsanit, Noppol Kobmoo, Papichaya Kwanthong, Tatiana E. Gorelik, Janet Jennifer Luangsa-ard, Sherif S. Ebada, Marc Stadler Data type: docx Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited. Link: https://doi.org/10.3897/imafungus.16.172651.suppl1