Iron regulatory pathways differentially expressed during Madurella mycetomatis grain development in Galleria mellonella

Abstract

Mycetoma is a chronic granulomatous infection of the subcutaneous tissue, most often caused by the fungal pathogen Madurella mycetomatis. Characteristic of the infection is the formation of grains. However, knowledge of the function and formation of the grain is limited. Here, we use a Galleria mellonella larvae infection model and transcriptomic profiling to identify processes associated with M. mycetomatis grain formation. Larvae were infected with M. mycetomatis and, after 4, 24, 72 and 168 h post-inoculation, RNA was extracted from larval content and sequenced. We found that 3498 G. mellonella and 136 M. mycetomatis genes were differentially expressed during infection. In particular, genes encoding proteins related to iron transport were highly expressed by both G. mellonella (transferrin and ferritin) and M. mycetomatis (SidA, SidD and SidI). LC-MS/MS analysis of M. mycetomatis cultured under iron-limiting conditions revealed the presence of SidA and SidD orthologs, and concurrent RP-HPLC and LC-MS identified a singly charged, putative siderophore in culture supernatant. Furthermore, we show that M. mycetomatis can obtain iron from holoferritin. Thus, our results highlight the importance of iron acquisition pathways during grain formation, suggesting potential avenues for development of new diagnostic and therapeutic strategies for mycetoma.

Fig. 1. Experimental workflow, burden of infection on G. mellonella larvae, and quality of transcriptomics data.

A Schematic illustration of the time course experiment design and transcriptomic assays used for high-throughput RNA profiling. Total RNA was extracted at five time points (three biological replicates each at each time point). At point 0 h, RNA was extracted from healthy G. mellonella larvae (host) and cultured M. mycetomatis strain (mm55) (pathogen) to serve as non-infected controls. After the inoculation of G. mellonella larvae with M. mycetomatis, RNA was extracted at 4, 24, 72, and 168 h post-infection (Left panel). Measure of the growth of M. mycetomatis cultured pathogen on YANAAI agar and YNAAI agar containing 500 M of the iron chelator 2′2-bipiridyl to study the impact of the FeCl₂, FeCl₃, apo- and holo-transferrin, and apo- and holoferritin on the growth (Middle panel). Impact of iron on the infection in the model. Grain development was monitored for 10 days, and different measures were performed. Total RNA was extracted at time 0, 4, 24, and 72 h, and RNA-Seq performed (Right panel). B To evaluate the burden of infection in the host, the melanisation of the hemolymph, the number of grains, and the grains size were measured and plotted in violin plots for each time point. In each of the violins, the box indicates the interquartile range (first quartile(Q1), third quartile(Q3), and the median value). The highest values are noted at t = 72 h (the maximum burden of infection). Non-infected larvae were included as controls. The data shown is an example based on observations from three evaluated biological replicates. The difference in the total number of grains or the total grain size observed on the respective time points was determined using the Mann–Whitney U test in GraphPad Prism 8. A p value > 0.05 was deemed significant (Methods). C Grain formation in the host was visualized 40 times magnified using H&E staining and light microscopy. Yellow arrows point towards the grain inside the capsule. Based on the characteristics of the grain development observed at each time point, the grains were defined as early, pre-mature, mature, and late grains. In the early grains, the cement material is not formed, and hemocytes are present between hyphae. In the pre-mature grains, hemolymph is forming, but still, individual hemocytes can be noted within the forming cement material. In the mature grain, the cement material is completely formed, and a capsule surrounds the grain. In the late grain, you see the capsule disappearing and an influx of hemocytes towards the grain. D t-SNE clustering: Host and pathogen RNA-Seq libraries mapped to G. mellonella genome assembly ASM364042v2. The replicates at each time point are shown as colored dots. The t-SNE plot shows the clustering of the RNA-Seq libraries mapped to the host genome. E Pathogen transcripts detected in host RNA-Seq after infection (three biological replicates were used at each time point, total n = 15). The line plot shows the number of pathogen reads detected in the host RNA-Seq data after infection, the largest number of the pathogen transcripts detected at time 72 h (mature grain) (lines are colored by time point). Data in the line plot are represented as mapped sequenced reads ± SEM. F RNA-Seq vs. LQ-ssCAGE mapped sequence reads correlation. The correlation matrix shows the Spearman correlation of the RNA-Seq and LQ-ssCAGE reads between replicates. Source data are provided as a Source Data file (B–E). Source data are available in the Gene Expression Omnibus (GEO) under accession numbers GSE213329 and GSE213332(F).

Fig. 2. Hierarchical clustering of the host top DEGs and gene set enrichment analysis.

A Heatmaps showing the hierarchical clustering of the top 50 DEGs of G. mellonella for four consecutive changes (Methods). Columns represent the biological replicates (3 replicates per time point), and rows represent the host gene symbols. Gene names are provided (Supplementary Data 5). Replicates are colored and grouped per time point. 0 h replicates are highlighted with yellow rectangles. Heatmap cells are colored by Z-score scale (−2 to 2). B Dotplots showing enriched pathways in biological processes and molecular functions in the host during infection. Each pair of plots shows enriched pathways of the top DEGs at each consecutive point. Each pathway is represented as a dot. The dots are colored by the p-adjusted value as computed by the gseGO function of the clusterProfiler R package (pvalueCutoff parameter of gseGO was set to <0.05). For testing differentially expressed genes, F-statistic, the associated P value, adj P value were corrected using Benjamini–Hochberg multiple testing correction (Statistics and reproducibility). The count represents the number of genes that belong to a given gene set, and the GeneRatio represents the count/setSize. setSize is the total number of genes in the gene set. Each pathway corresponds to a Gene Ontology (GO) term. Source data are available in the Gene Expression Omnibus (GEO) under accession number GSE213329.

Fig. 3. Pathogen reads in infected host and gene set enrichment analysis.

A Heatmaps of the top 20 DEG cultured pathogen vs. pathogen reads in infected host 4 h. Full annotation of the genes in each of the heatmaps in (Supplementary Data3). B GSEA of the top 20 DEG (For testing differentially expressed genes, F-statistic, the associated P value, and the adj. P value were corrected using Benjamini–Hochberg multiple testing correction.) pathogen genes and enriched GO terms (gseGEO parameter pvalueCutoff = 0.05). For testing differentially expressed genes, F-statistic, the associated P value, adj. P value were corrected using Benjamini–Hochberg multiple testing correction (Statistics and reproducibility). Source data are available in the Gene Expression Omnibus (GEO) under accession number GSE213329.

Fig. 4. Host and pathogen transcriptomic changes before and after the development of Eumycetoma grain.

A The responses of host DEGs are shown in four patterns (Gradual, early, early long, and late response). Y-axis represents the Z-score of gene abundance, and the X-axis shows the time point and grain development stage. The gradual, early response and early long response show sub-patterns (down and up) shown as dashed or continuous lines, respectively. Full annotation of the genes in each response group in (Supplementary Data 6). B Expression of some of the host DEGs corresponding to the patterns in (A). Y-axis shows the count per million (CPM), and the X-axis shows the time point (n = 15, 3 biological replicates per time point). Data are presented as mean values ± SEM. C Transcriptomic response patterns of the pathogen DEGs. Only three patterns are detected. As in (A), Y and X axes represent the Z-score gene abundance and grain development stage, respectively. Full annotation of the genes in each response group is in Supplementary Data 8. D Expression of some of the pathogen DEGs corresponding to the patterns in (C) (n = 15, 3 biological replicates per time point). Data are presented as mean values ± SEM. We computed a two-way ANOVA test. Source data for (A, C) are available in the Gene Expression Omnibus (GEO) under accession number GSE213329. Source data for (B, D) are provided as a Source Data file.

Fig. 5. Iron regulation in G. mellonella and siderophore biosynthesis pathways in M. mycetomatis.

A PPI of Ferritin genes in Drosophila melanogaster adapted from G. Xiao et al., with a focus on transferrin (Tsf1), Ferritin 1 Heavy Chain Homologue (Fer1HCH), Ferritin 2 Light Chain Homologue (Fer2LCH), Iron-regulatory protein 1A and 1B (Irp-1A and Irp-1B). Expression levels of respective G. mellonella homologs during infection are shown in the bar plot. Aconitate hydratase encoding genes LOC113511518, LOC113516537, LOC113522652, and LOC113510586 are differentially expressed during infection (three biological replicates were used at each time point, total n = 15). Data are presented as mean values ± SEM. B Generated hairpin loops in LOC113510017 and LOC113510018 transcripts, containing the conserved CAGUGU sequence characteristic for Iron-Responsive Elements (IREs). Expression of ferritin and transferrin homologous genes in G. mellonella based on LQ-ssCAGE (n = 15, 3 biological replicates per time point). In the barplot, the error bar represents the standard error of the mean (SEM). C Siderophore biosynthesis pathway adapted from Gründlinger et al.. The pathway starts with mevalonate converted to anhydromevalonyl-CoA by SidI and SidH. A. fumigatus uses three hydroxamate-type siderophores for iron uptake. The extracellular triacetylfusarinine C (TAFC), hyphal ferricrocin (FC), and conidial hydroxyferricrocin (HFC). N⁵-hydroxyornithine is generated from ornithine by SidA. The SidG gene has no homologs gene in M. mycetomatis, and the gene marked as (?) is not characterized yet in A. fumigatus. In the PPI, SidF and SidL are annotated to the same gene in M. Mycetomatis (N(6)-hydroxylysine O-acetyltransferase). No significant similarity was found for SidG in M. mycetomatis. All siderophore genes except SidG are differentially expressed in RNA-Seq data, as shown in the bar plot (n = 15, 3 biological replicates per time point). Data are presented as mean values ± SEM. We computed a two-way ANOVA test; the data in the bar chart are represented as expression ± SEM. Source data for the bar plots are provided as a Source Data file.

Fig. 6. Growth of M. mycetomatis measured by the diameter of the mycelium.

A Growth of M. mycetomatis on YNAAI agar supplemented with PBS, FeCl₂, FeCl₃, apo- and holo-transferrin, and apo- and holoferritin. Two biological replicates were performed. Data presented as mean value, showing both individual data points. B Growth of M. mycetomatis on YNAAI agar containing 500 M of the iron chelator 2′2-bipiridyl, and supplemented with PBS, FeCl₂, FeCl₃, apo- and holo-transferrin, and apo- and holoferritin. Two biological replicates were performed. Data presented as mean value, showing both individual data points. Source data are provided as a Source Data file.

Fig. 7. Validation of M. mycetomatis siderophore biosynthesis.

A RP-HPLC analysis identified the time-dependent presence of three secreted metabolites (P1–P3) in untreated culture supernatants at A254 nm. P2 also showed weak A440 nm. B Post-ferration RP-HPLC revealed that P1 shifted to co-elute with P2 (increased A 254 nm and a 10-fold increase in 440 nm). C Spectrum scanning confirmed increased A440 + 20 nm, specifically following ferration for P2. D High-resolution LC-MS analysis of unferrated P1 (Rt 22.2 min) revealed a singly charged, putative siderophore with m/z = 855.2671 (inset) and a compound with m/z 822. E Since the species with m/z = 855.2671 disappeared upon Fe³⁺ addition, it represents a hitherto unidentified M. mycetomatis siderophore. Compound with m/z 822 is also evident in P2 (ferrated), and the nature of the species with m/z 769.3940 (inset) remains to be revealed. Source data are available via the PRIDE partner repository with the dataset identifier PXD063449.

Fig. 8. Effect of iron on the infection in G. mellonella larvae.

A Toxicity of 332 µM EDDA, and 25 µM FeCl₂ and FeCl₃ in G. mellonella. B Survival of M. mycetomatis-infected larvae upon exposure to 332 µM EDDA, 25 µM FeCl₂, and 25 µM FeCl₃. C M. mycetomatis grain count in G. mellonella larvae upon exposure to 332 µM EDDA, 25 µM FeCl₂, and 25 µM FeCl₃. Data are presented as mean values ± SD, determined based on five biological replicates. D M. mycetomatis Grain size in G. mellonella larvae upon exposure to 332 µM EDDA, 25 µM FeCl₂, and 25 µM FeCl₃. Data are presented as mean values ± SD, determined based on five biological replicates. E Expression of G. mellonella genes related to iron regulation during infection upon exposure to 332 µM EDDA, 25 µM FeCl₂, and 25 µM FeCl₃ (n = 15, 3 biological replicates per time point). Data are presented as mean values ± SEM. We computed a two-way ANOVA test. F Expression of M. mycetomatis genes related to iron regulation during infection in G. mellonella larvae upon exposure to 332 µM EDDA, 25 µM FeCl₂, and 25 µM FeCl₃ (n = 15, 3 biological replicates per time point). Data are presented as mean values ± SEM. G Grocott stain of a M. mycetomatis grain in G. mellonella 72 h after infection and exposure to FeCl₃. The arrows indicate the hyphae growing outside of the grain. The data shown are an example based on observations from five evaluated biological replicates. Source data for (A–F) are provided as a Source Data file.

Fig. 9. Proposed model of Iron regulation in G. mellonella during infection with M. mycetomatis “Created in BioRender. Konings, M. (2025) https://BioRender.com/t19drer”.

A In healthy larvae, in the absence of iron (Fe²⁺), the IRP1 and IRP2 form a complex, which can bind to Iron-Responsive Elements (IRE) in the 5′ prime UTR region of the ferritin subunits coding transcripts LOC13510017 and LOC13510018, preventing translation of the mRNA to form ferritin. In the presence of iron (Fe²⁺), 4Fe-4S is formed, binding to IRP1. The IRP1/4Fe-4S complex prevents binding to the IRE, enabling translation of both LOC13510017 and LOC13510018, forming ferritin. Ferritin will bind Fe²⁺ for transportation and storage. B During infection, increased expression of aconitate hydratase results in increased formation of the 4Fe-4S complex, binding of IRP1, and thus the translation of both subunits of ferritin for the binding of Fe²⁺, creating an iron-derived environment. In response to the limited available iron, the fungus shows significantly increased expression of the siderophore-related biosynthesis pathway for the production of siderophores needed for sequestering Fe²⁺, which is essential for the survival of the fungus.