Insights Into Divergent Mechanisms of Pyrite Oxidation by Indigenous Fungi From Mine Tailings

ABSTRACT

Fungi can oxidise pyrite, a process which has potential implications for metal bioleaching and acid mine drainage formation. However, this capacity has been demonstrated primarily using only a few model fungal strains. Diverse fungal communities are known to inhabit pyrite‐rich mine tailings but how these indigenous fungi may mediate pyrite oxidation and by what mechanisms remain largely unexplored. Here, we isolated 51 pyrite‐oxidizing fungal strains spanning 28 genera from three mine tailings areas in South China. From this native library, we selected three potent pyrite oxidisers, Pithomyces cynodontis DC37, Penicillium janthinellum ZJS52, and Aspergillus niger DBS124, for in‐depth mechanistic characterisation. These indigenous strains solubilised 1.2%–1.8% of the total iron from pyrite in fungus‐pyrite co‐culture systems, exceeding values reported for model fungi under comparable conditions. They also acidified the medium to a pH as low as 2.15, providing dual lines of evidence for their robust oxidative capacity. Further analyses revealed distinct surface erosion features and changes in iron (Fe) and sulfur (S) speciation, confirming fungal‐mediated pyrite oxidation. All three strains secreted organic acids (e.g., oxalic and citric acids) and showed peroxidase activity, but they exhibited divergent patterns. By combining hybrid genome sequencing (Illumina and Nanopore) with time‐course transcriptomics, we uncovered the genetic basis for these divergent patterns. Genes coding for lignin and manganese peroxidases in P. cynodontis DC37 were co‐induced by reactive oxygen species and metal ions released during pyrite oxidation. Furthermore, A. niger DBS124 employed an oxalic‐acid‐based mechanism upon pyrite exposure, driven by coordinated upregulation of central carbon metabolism and oxalate biosynthesis. Collectively, our findings highlight the underexplored biotechnological potential of mine‐tailings fungi and establish a molecular foundation for understanding their roles in Fe‐ and S‐cycling within mining environments and for developing fungal‐mediated bioleaching technologies.

To elucidate the mechanisms of fungal pyrite oxidation, a large‐scale isolation of indigenous pyrite‐oxidizing fungi from mine tailings was conducted, and three focal strains were subjected to integrated multi‐omics and molecular analysis in fungus‐pyrite co‐culture systems, revealing their divergent oxidative strategies.

1. Introduction

The acidification of heavy metal mine tailings represents a significant environmental challenge (Park et al. 2019). Pyrite, as the most globally abundant metal sulfide mineral, is widely recognised as a major contributor to acid mine drainage (AMD) (Chandra and Gerson 2010; Fan et al. 2017). Once exposed to air, pyrite can be oxidised through the combined action of oxygen and water, producing sulfuric acid and releasing heavy metal ions from tailings. Microbially accelerated oxidation of pyrite and its underlying mechanisms have long been a central research focus in the field of mine pollution and remediation worldwide.

Since the 1980s, numerous studies have reported on the oxidation of metal sulfides by pure cultures of acidophilic bacteria or archaea. Key genera involved include Acidithiobacillus, Leptospirillum, and Ferroplasma, whose mechanisms have been well characterised (Dopson et al. 2004; Hutchins et al. 1986; Lizama and Suzuki 1988). In these prokaryotes, oxidation of metal sulfides is primarily driven by the generation of Fe³⁺, which serves as a chemical oxidant, together with possible direct enzymatic interactions with the mineral surface (Baker and Banfield 2003). However, research on fungal‐mediated sulfide mineral oxidation remains limited, despite the advantages of fungi such as their adaptability to a wide pH range and their high metal tolerance (Valix and Loon 2003; Wu and Ting 2006; Xu et al. 2014).

Several studies have demonstrated the ability of fungi to oxidise metal sulfide minerals. Research has primarily focused on species from the Aspergillus and Penicillium genera, typically isolated from decaying woody substrates and mine tailings (Cecchi et al. 2019; Yang et al. 2019). A smaller number of investigations have included lignin‐degrading fungi from decayed wood (e.g., Phanerochaete chrysosporium and Trichoderma harzianum) or species from oil‐contaminated soils (e.g., Exophiala spinifera and Fusarium oxysporum). Studies of the underlying oxidation mechanisms have largely relied upon traditional methods, such as monitoring pH changes and analysing fungal metabolites. Among these, the mechanism by which Phanerochaete chrysosporium oxidises pyrite has been relatively well characterised, involving a combination of extracellular enzymes (e.g., lignin peroxidase, manganese peroxidase, and glyoxal oxidase), H₂O₂, Fe³⁺, and organic acids (Ofori‐Sarpong et al. 2011; Ofori‐Sarpong et al. 2013; Yang et al. 2018). Beyond these well‐characterised fungi, numerous other congeneric species possess analogous biochemical pathways, suggesting a broader, yet underexplored, potential for sulfide mineral oxidation (Li et al. 2026). Moreover, with advances in sequencing technology, recent research has employed techniques such as second‐generation de novo transcriptome sequencing to monitor the expression of organic acid and metal resistance‐related genes in the fungus Trichoderma asperellum during bioleaching (Nkuna and Matambo 2024).

Notwithstanding, our current understanding of fungal‐mediated pyrite oxidation is largely confined to studies of single strains and discrete metabolic pathways. These insights are often dissociated from the systematic, multi‐scale perspective needed to integrate molecular, physiological and genomic mechanisms. Even when molecular approaches are employed, the prevalent reliance on short‐read sequencing often results in fragmented genome assemblies and fails to resolve complex genomic regions. Furthermore, while heavy metal mining environments are rich in metal sulfides, research on most fungi isolated from such habitats (e.g., Aspergillus niger, Acidea extrema, Penicillium simplicissimum, Apiotrichum loubieri) has emphasised their metal resistance and molecular adaptations, rather than their capacity to oxidise metal sulfides (Hujslová et al. 2017; Kumari et al. 2015). Consequently, it remains unclear as to whether these indigenous fungi can oxidise metal sulfides, or whether their oxidation mechanisms resemble those previously reported, and what specific molecular pathways may be involved in the process.

To address these knowledge gaps, we established a fungus‐pyrite co‐culture system to investigate the oxidation mechanisms of mine‐derived fungi through multi‐scale analyses. The specific aims of this study were: (1) to isolate fungi from three distinct mine tailings in South China and select the most efficient pyrite‐oxidizing strain for in‐depth mechanistic study; (2) to characterise the mineral surface changes driven by fungal oxidation using surface‐sensitive mineralogical analyses; (3) to elucidate the underlying chemical mechanisms via systematic biochemical assays of metabolites and enzymes; and (4) to decipher the molecular genetic basis via hybrid genome sequencing (Illumina and Nanopore) and time‐course transcriptomics to identify key genes involved in fungal pyrite oxidation. This study therefore aimed to systematically reveal the extent of fungal pyrite bio‐oxidation in mine tailings, thereby bridging a critical knowledge gap and potentially establishing a molecular‐level theoretical foundation for developing fungi‐based leaching technologies.

2. Materials and Methods

2.1. Soil Sample Collection

From August to September 2021, a total of 34 acidic soil sampling sites (pH 2–7) were selected from three tailings areas: Dabaoshan in Shaoguan, Guangdong Province; Dachang Copper Mine Reservoir in Hechi, Guangxi Zhuang Autonomous Region; and Zijinshan Gold‐Copper Mine in Fujian Province. The three mine sites differed in geographical location, major mineral types (Dabaoshan polymetallic tailings; Zijinshan copper mine tailings; Dachang polymetallic tailings), and mining history. A total of 102 surface soil samples were collected (three parallel samples per site) at a depth of 0–20 cm. They were placed in sterile 50 mL centrifuge tubes, transported to the laboratory on ice, and stored at 4°C for subsequent fungal isolation. Throughout the sampling process, spades were cleaned to minimise anthropogenic contamination and cross‐sample pollution.

2.2. Isolation, Culture, and Identification of Fungi

Soil samples were suspended in sterile water, vortexed and allowed to stand for clarification. The microbial suspension from each sample was spread onto Potato Dextrose Agar, Malt Extract Agar, Czapek's Agar, and Martin's agar plates (Martin 1950; Samson and Frisvad 2004). All media were supplemented with 50 μg/mL streptomycin and akamycin to inhibit bacterial growth. Plates were incubated at 30°C for 1–3 weeks, and emerging colonies were sub‐cultured to obtain pure strains. Genomic DNA was extracted with a commercial kit, and the ITS region was amplified via PCR with primers ITS1/ITS4 (for the detailed PCR system and primers see Supporting Information and Methods S1); PCR products were sent for sequencing, and fungal species were identified by sequence homology alignment against the NCBI GenBank database.

2.3. Bio‐Oxidation of Pyrite

Experiments on pyrite powder bio‐oxidation were conducted to screen for metal sulfide‐oxidizing fungi, with three independent biological replicates for all assays to ensure reproducibility. Three 5 mm mycelial plugs from actively growing colonies were inoculated into 150 mL of culture medium containing 50 mM pyrite (FeS₂) and incubated at 30°C with shaking at 150 rpm (abiotic controls were set as uninoculated medium; its composition is shown in Table S1). As Fe³⁺ can oxidise pyrite, the heat‐sensitive FeSO₄·7H₂O was filter‐sterilised (0.22 μm) rather than being autoclaved in order to prevent the conversion of Fe²⁺ to Fe³⁺. Pyrite powder was pretreated by deoxygenation and autoclaving before use (see Supporting Information and Methods S2 for details). During cultivation, 10 mL of samples were collected aseptically every two days to monitor pH and total iron concentrations.

Pyrite slice bio‐oxidation experiments were conducted to investigate mechanisms of fungal pyrite surface oxidation. Prior to these experiments, bulk pyrite was cut, ground, polished into thin slices and sterilised (see Supporting Information and Methods S2 for details). The purity of the pyrite material was confirmed by XRD (Figure S1), with an initial Fe/S atomic ratio of 0.524 determined by EDS (Table S2). To minimise the contribution of fungal extracellular peroxidases to pyrite bio‐oxidation, a sterile cellophane membrane was placed between the fungus and the polished pyrite slice during co‐culture on the solid medium. Owing to its size‐selective permeability, the membrane was expected to prevent the transport of extracellular enzymes whilst allowing diffusible fungal metabolites to reach the mineral surface. All experimental and control set‐ups were prepared in triplicate and incubated at 30°C for 20 days, after which the pyrite slices were collected for mineral surface analysis.

2.4. Determination of pH and Total Dissolved Iron Concentrations

The pH of the culture broth was measured using a pH meter (Sartorius PB‐10). Total dissolved iron concentrations were analysed by acetylene‐air flame atomic absorption spectrophotometer (Shimadzu AA‐7000). Fungal‐dissolution of iron (Ofori‐Sarpong et al. 2011) was estimated in wt.% as shown in Equation (1), where V is the volume of solution, C is the concentration of free iron cations, P _m is the percentage of the constituent in the pyrite and W is the mass of pyrite used. The medium contained FeSO₄·7H₂O as a component of the trace element solution, providing iron for fungal growth (Boon and Heijnen 1998; Yang et al. 2018). This 0.05 wt.% iron contributed by FeSO₄·7H₄O in the medium was subtracted from the total fungal dissolution of iron to obtain the iron derived from fungal pyrite oxidation. A fungal strain was considered capable of oxidizing metal sulfides if the changes in pH and total iron concentration in its culture were significantly different from those of the blank control, as determined by Tukey's HSD test (p < 0.05).

$$\begin{array}{c}\text{Fungal dissolution of iron},\mathrm{FDI}\left(\mathrm{wt}.\%\right)\hfill \\ =\frac{V\left(\mathrm{mL}\right)\times C\left(\mathrm{\mu g}/\mathrm{mL}\right)}{P_m\left(\%\right)\times W\left(\mathrm{g}\right)}\times 100\%-0.05\%\hfill \end{array}$$ (1)

2.5. Mineralogical Surface Analysis of Pyrite by X‐Ray Diffraction, Scanning Electron Microscopy, and X‐Ray Photoelectron Spectroscopy

Secondary precipitates formed in the pyrite reaction system were identified using a benchtop X‐ray diffractometer (Rigaku MiniFlex‐600). The analysis was conducted with a Cu target X‐ray tube operated at 40 kV and 15 mA, scanning over a 2θ range: 10°–80°.

The surface morphology and elemental composition of pyrite samples were examined using a Hitachi SU8010 field emission scanning electron microscope equipped with an energy‐dispersive X‐ray spectroscopy system. Measurements were performed at accelerating voltages of 1.5 kV or 15 kV.

The chemical states of surface elements were determined using X‐ray photoelectron spectroscopy (Thermo Fisher K‐Alpha) with a monochromated Al Kα source (1486.6 eV). All measurements were carried out under an ultra‐high vacuum of 5.2 × 10⁻⁸ mbar with a 400 μm X‐ray spot size. Survey scans (0–1350 eV) were acquired at a pass energy of 100 eV, and high‐resolution scans of C 1s, Fe 2p, S 2p, and O 1s regions were collected at 30 eV pass energy. Spectra were calibrated to the C 1s peak at 284.8 eV and fitted using Thermo Avantage software (version 5.9931). Survey scans confirmed the presence of C, Fe, S and O in samples (Figure S2), and high‐resolution core‐level spectra were deconvoluted to identify chemical shifts. Figures S3 and S4 present the high‐resolution S 2p and O 1s XPS spectra of pyrite surfaces.

2.6. Determination of Metabolite Concentrations

The concentrations of key fungal metabolites were quantified. Hydrogen peroxide (H₂O₂) was measured spectrophotometrically (Beers and Sizer 1952). Briefly, 5 mL of sample was mixed sequentially with 4.0 mL of 2 mol/L H₂SO₄ and 10.0 mL of ammonium metavanadate solution, diluted to 25 mL with deionised water, and incubated for 5 min before reading the absorbance at 460 nm. Major low‐molecular‐weight organic acids (citric, malic, gluconic, and oxalic acids) were analysed by high‐performance liquid chromatography (HPLC; Shimadzu LC‐20AB system) using a Rezex ROA‐Organic Acid H⁺ (8%) column (Kelebek et al. 2009; Wang et al. 2007). The HPLC separation employed 0.005 M H₂SO₄ as the mobile phase at 40°C with detection at 210 nm (DAD detector), a 20 μL injection volume and a 20 min run time (Kelebek et al. 2009; Wang et al. 2007).

2.7. Determination of Peroxidase Activity Secreted by Fungi

The activities of lignin peroxidase (LiP), manganese peroxidase (MnP), and glyoxal oxidase (GLOX) were determined using commercial assay kits. MnP and LiP activities were measured via spectrophotometric methods based on substrate oxidation, and GLOX activity was quantified using an enzyme‐linked immunosorbent assay kit. Enzyme activities were calculated following the manufacturer's instructions (see Supporting Information and Methods S3 for details).

2.8. Whole‐Genome Sequencing

Whole‐genome sequencing of the three fungal strains was performed using Nanopore and Illumina MiSeq platforms. The long reads were assembled with Canu (v2.2) and error‐corrected through two rounds of polishing using Illumina short reads with Pilon (v1.23) (Hu et al. 2020; Koren et al. 2017). Scaffolding was conducted with SSPACE (v1.1) and BLASR (Boetzer et al. 2011). Genome completeness was assessed using BUSCO (v5.4.3) (Manni et al. 2021). Gene prediction was carried out with GeneMark (v4.33), and tRNA, rRNA, and tmRNA genes were identified using tRNAscan‐SE (v2.0.11), Barrnap (v0.9), and Aragorn (v1.2.41), respectively. Predicted genes were functionally annotated against the KEGG, GO, NCBI NR, eggNOG, Swiss‐Prot, and Pfam databases. A circular genome map was generated using the ChiPlot online platform (https://www.chiplot.online/). GC content and GC skew were calculated using a sliding window of 5000 bp.

2.9. Transcriptome Analysis

Fungal biomass from pyrite‐containing and pyrite‐free co‐culture experiments was collected at days 0, 2, 4, 6, 8, 12, and 16 for transcriptome profiling. Total RNA was extracted and assessed for quality using a NanoDrop 2000 spectrophotometer and an Agilent 2100 Bioanalyzer. Sequencing libraries were prepared from poly(A)‐enriched mRNA by fragmentation, double‐stranded cDNA synthesis, end repair, adapter ligation and PCR amplification. Library quality was verified prior to Illumina paired‐end (150 bp) sequencing. Raw reads were processed to remove adapter sequences, low‐quality reads (Q30 score below 90%), and ribosomal RNA using fastp and SortMeRNA (Chen et al. 2018; Kopylova et al. 2012). Clean reads were aligned to the respective reference genomes using HiSat2 (Kim et al. 2019), and gene expression was quantified by featureCounts and normalised as Transcripts per Kilobase Million (Liao et al. 2014). Differentially expressed genes were identified by comparing control and pyrite‐treated samples, using thresholds of |log₂FC| ≥ 1 and p‐value < 0.05. Functional enrichment analysis of KEGG pathways and GO terms was performed with TBtools (Chen et al. 2020).

2.10. Statistical Analysis

Statistical analyses were performed using R (v4.5.0) and data were visualised with R (ggplot2 package) and Origin 2018 (v9.5). Tukey's HSD test and t‐tests were conducted in R with the stats package (v4.5.0), respectively. Differential gene expression analysis was conducted with DESeq2 package (v1.48.1) to obtain adjusted p‐values (Benjamini‐Hochberg false discovery rate (FDR) method) and log₂fold changes.

3. Results

3.1. Identification of Three Fungi With Strong Pyrite Oxidation Capacity

A total of 125 fungal strains were isolated from three mine tailing sites in South China (Figure S5), of which 51 showed potential metal sulfide oxidation capacity (Figure S6). Pithomyces cynodontis DC37, Penicillium janthinellum ZJS52, and Aspergillus niger DBS124 showed the highest pyrite oxidation capacity among the isolated strains in their respective mine tailings. The final pH values of the culture media were as follows: P. cynodontis DC37, pH 3.1; P. janthinellum ZJS52, pH 2.19; A. niger DBS124, pH 1.86 (Figure 1A–C, Tables S3 and S4). Given that these three strains originated from different isolated sources, belonged to distinct genera, and showed considerably different final culture pH values (Table S3), they may vary in their pyrite oxidation capacities and possibly in their underlying oxidation mechanisms. The three strains were therefore selected for further analysis.

FIGURE 1.

Pyrite oxidation activities of the three focal fungal strains. (A–C) Colony morphology of the three isolates obtained in this study. (D, E) Temporal changes in the pH (D) and total dissolved iron (E) in the culture systems of these strains with or without pyrite powder supplementation. Lowercase letters indicate significant differences among treatments at the same sampling time point (Tukey's HSD test, p < 0.05). (F) Maximum values of Fungal Dissolution of Iron (FDI) for the three focal fungal isolates and the reference strain Phanerochaete chrysosporium ME446 (Ofori‐Sarpong et al. 2013) during the incubation period. Lowercase letters indicate significant differences among the three focal isolates (Tukey's HSD test, p < 0.05). Data are shown as mean ± standard deviation (SD); n = 3 per strain. Note that the FDI value for P. chrysosporium ME446 represents a single data point and was therefore excluded from the statistical analysis.

The temporal profiles for pH and total iron concentrations during liquid cultivation revealed that all three focal fungal strains rapidly acidified the medium, with the pH dropping to approximately 2.5 within four days (Figure 1D). Once the pH fell below approximately 2.5, free iron concentrations rose sharply in all systems, peaking between days 12 and 16 (Figure 1E). Among the three strains, A. niger DBS124 (53.9 μg/mL) exhibited the highest iron leaching efficiency, followed by P. janthinellum ZJS52 (37.1 μg/mL) and P. cynodontis DC37 (36.2 μg/mL). Over the 16‐day co‐culture period, the fungi showed increasing iron dissolution capacities, with FDI values of 1.24 wt.% for P. cynodontis DC37, 1.28 wt.% for P. janthinellum ZJS52, and 1.88 wt.% for A. niger DBS124 (Figure 1F). Based upon the observed process dynamics (rapid acidification, peak iron release, and subsequent stabilisation), and the sampling days (2, 4, 6, 8, 12, 16), adjacent time points were grouped into an initial phase (Days 2 and 4), an intermediate phase (Days 6 and 8), and a terminal phase (Days 12 and 16).

3.2. Effects of Physical Fungal Contact and Cellophane‐Separated Co‐Culture on the Pyrite Surface

SEM‐EDS analysis revealed erosion features on the pyrite surface, including grooves and etching pits, caused by fungal co‐cultured with pyrite in physical contact conditions. Localised changes in atomic composition were also observed (Figure 2A–C and Table S2). Among the three strains, P. janthinellum ZJS52 caused the most extensive erosion, with grooves up to 44.2 μm wide and dense hyphal attachment. It was accompanied by a high surface oxygen content (69.59%), indicating strong oxidative activity. P. cynodontis DC37 produced linear grooves with localised oxidation (25.3 μm wide, 42.68% oxygen content) and accumulated carbon‐rich residues, suggesting the presence of embedded organic substances. A. niger DBS124 formed only minor etching pits, approximately 10 μm in width.

FIGURE 2.

SEM and XPS characterisation of pyrite surfaces following fungal incubation. (A–F) SEM images of polished pyrite surface after 20 days of co‐culture with three focal fungal strains. (G) Comparison of etching pit diameters on pyrite surfaces after co‐culture with three focal fungi under physical contact and cellophane‐separated conditions (t‐tests, *p < 0.05, ***p < 0.001, n = 3). (H–K) High‐resolution XPS Fe 2p spectra of polished pyrite after 20‐day cultivation with (H) sterile culture control, (I) Pithomyces cynodontis DC37, (J) Penicillium janthinellum ZJS52, and (K) Aspergillus niger DBS124.

When a cellophane membrane was placed between the fungi and the pyrite slice to prevent physical contact, the density and size of these oxidation features were substantially reduced for all three strains (Figure 2D–F). The most pronounced reduction observed was for P. cynodontis DC37 (Figure 2G and S7), indicating a strong dependence on contact‐dependent mechanisms, likely involving extracellular peroxidases. In contrast, the cellophane membrane had minimal effect on A. niger DBS124, suggesting that its oxidation mechanism depends largely on diffusible metabolites, which can pass through the cellophane membrane.

XPS analysis revealed changes in the surface valency states of pyrite after 20 days of co‐cultivation with each fungal strain. The intensity of the characteristic Fe(II)‐S 2p_1/2 peak increased from 18.9% to a range of 24.5%–26.5%, whilst the total spectral contribution of Fe³⁺ oxides (at 708.8 and 710.9 eV) decreased from 29.8% to a range of 21.8%–26.8% (Figure 2H–K and Table S5). This indicates the dissolution of naturally formed surface oxides, exposing the underlying Fe(II)‐S matrix of pyrite.

Corresponding changes were observed in the S 2p spectra (Figure S3). The intensity of the main FeS₂ peak (162.5 eV) increased from 56.5% to 68.9%–74.8% (Table S5), further supporting the removal of surface oxides and exposure of fresh pyrite. Concurrently, the prominent sulfate peak (tentatively assigned to ferric sulfate at 168.68 eV) diminished markedly.

3.3. Organic Acid and Peroxidase Production During Fungal‐Pyrite Bio‐Oxidation

The three fungal strains exhibited distinct organic acid secretion patterns (Figure 3D,G). Aspergillus niger DBS124 predominantly produced oxalic acid, which peaked at 257.6 mg/L on day 6, alongside an earlier peak of citric acid (10.8 mg/L on Day 2). In contrast, P. janthinellum ZJS52 primarily secreted citric acid, reaching 12.48 mg/L on day 4, with only minor oxalic acid detected. For P. cynodontis DC37, only citric acid was detected, and its concentration was the lowest among the three strains. The rapid acid accumulation, especially during the initial phase (Days 2 and 4), directly drove a drop in system pH, thereby enhancing the efficiency of pyrite iron leaching.

FIGURE 3.

Genomic and metabolic insights into organic acid production and gene expression in response to pyrite. (A–C) Circular genome maps of the fungal strains. Different colours represent different scaffolds, which are labelled with different numbers. The inner rings show GC content and GC skew at corresponding genomic positions. (D) Temporal variations of oxalic acid concentration in fungal‐pyrite co‐culture systems. (E) Pyrite alters the expression of key oxalic acid metabolism genes in A. niger DBS124 (day 2). Statistical significance (DESeq2 with Benjamini–Hochberg adjusted p‐values): pyruvate carboxylase, p _adj = 0.299; oxaloacetate hydrolase, p _adj = 0.0148. (F) Metabolic pathways for oxalic and citric acid production in A. niger (Kobayashi et al. 2014). CE, citrate exporter; CS, citrate synthase; OAH, oxaloacetate hydrolase; PC, pyruvate carboxylase. (G) Peak citric acid concentrations for A. niger DBS124 (day 2), P. janthinellum ZJS52 (day 4), and P. cynodontis DC37 (day 6) during 16‐day co‐culture with pyrite. (H) Effect of pyrite on the expression of key genes involved in the citric acid metabolism and secretion pathway in P. janthinellum ZJS52. Statistical significance (DESeq2): pyruvate carboxylase (day 2, p _adj = 0.491), citrate synthase (day 4, p _adj = 0.0413), and citrate exporter (day 2, p _adj = 0.0015). For all panels, data are presented as mean ± SD (n = 3) Asterisks indicate statistical significance: *p _adj < 0.05, **p _adj < 0.01.

To assess enzymatic activity during pyrite oxidation, we monitored the activities of LiP, MnP, and GLOX, along with the concentration of H₂O₂ (Figures 4A,C, S8A, and S8B). P. cynodontis DC37 and P. janthinellum ZJS52 exhibited strong peroxidase activities in the initial and intermediate phases. Their LiP and MnP activities peaked at 12.76 and 1.56 U/L for P. cynodontis DC37, and 16.02 and 0.91 U/L for P. janthinellum ZJS52, respectively. In contrast, A. niger DBS124 showed minimal activities (LiP, 0.91 U/L; MnP, 0.22 U/L). All three strains produced comparable levels of GLOX, peaking at approximately 2.00 U/L during the intermediate phase. During this phase, P. cynodontis DC37 also reached its maximum H₂O₂ concentration (38.05 μg/mL). However, H₂O₂ concentrations varied considerably between the strains, suggesting that GLOX activity alone was not the primary determinant of H₂O₂ accumulation.

FIGURE 4.

Peroxidase activity and gene expression in P. cynodontis DC37 during pyrite co‐culture. Temporal profiles of (A) Manganese peroxidase activity, (C) Lignin peroxidase activity in fungal‐pyrite co‐culture systems, alongside the effect of pyrite on the expression of the corresponding (B) MnP and (D) LiP encoding genes in the fungus P. cynodontis DC37. Statistical significance (DESeq2): MnP (day 2, p _adj = 1.24E−13); LiP (day2, p _adj = 9.24E−4) Data are presented as mean ± SD (n = 3). Asterisks indicate significance levels: ***p _adj < 0.001.

3.4. Genomic Features of the Fungal Strains

Whole‐genome sequencing of the three focal fungal strains was performed using Illumina and Nanopore platforms, yielding high‐quality genome assemblies. Circular genome maps (Figure 3A–C) provide an overview of the genomic architecture of each strain. The assembled genome of P. janthinellum ZJS52 is approximately 36.6 Mb and comprises 19 chromosomes. P. cynodontis DC37 has a genome size of about 35.5 Mb with 13 chromosomes, whereas A. niger DBS124 possesses a relatively compact genome of approximately 34.2 Mb organised into 9 chromosomes. The GC contents of the three genomes are comparable, ranging from 49.14% to 52.24%, with no obvious large‐scale anomalies, indicating overall genomic stability. Genome completeness exceeded 98% for all three strains, and the number of predicted protein‐coding genes ranged from 11,118 to 12,799 per strain (Table S6). Functional annotation using KEGG, GO, NR, eggNOG, Swiss‐Prot, and Pfam databases successfully assigned putative functions to more than 95% of the genes in each strain (Table S7).

Genomic analysis successfully identified key genes for the metabolism and secretion of the detected organic acids in all fungi (Table S8). For citric acid, which was secreted by all three strains, genes encoding enzymes for its metabolism and transport were identified in each genome. In A. niger DBS124, a strong organic acid producer, genes for the oxalic acid biosynthesis pathway were annotated, including key enzymes such as pyruvate carboxylase (PC) and oxaloacetate hydrolase (OAH) (Figure 3F). This genetic repertoire provides a molecular basis for its pronounced acidification capacity.

The genome of P. cynodontis DC37, which exhibited strong extracellular peroxidase activity, contained annotated genes for LiP, MnP, and GLOX (Table S9), consistent with the results for its enzymatic activity. In contrast, despite detecting MnP and LiP activities in cultures of P. janthinellum ZJS52 and A. niger DBS124, no homologous genes for these enzymes were annotated in their genomes.

3.5. Transcriptomic Responses to Fungal Pyrite Oxidation

High‐quality transcriptomic sequencing (yielding ≥ 6.00 Gb clean data per sample with Q30 > 92.17%; Table S10) enabled a time‐course analysis. This revealed that pyrite exposure was associated with significant transcriptional changes in all three fungal strains. KEGG and GO enrichment analyses of differentially expressed genes (DEGs) revealed phase‐specific metabolic adaptations during pyrite bio‐oxidation (Figure S9 and Table S11). During the initial phase (Days 2 and 4), a concerted induction of metabolic pathways, including carbohydrate, amino acid, nucleotide, organic acid and sulfur metabolism, indicated a comprehensive metabolic activation. During the intermediate phase (Days 6 and 8), accelerated pyrite oxidation triggered a distinct oxidative stress response, marked by the upregulation of genes encoding antioxidant enzymes (e.g., superoxide dismutase, glutathione, catalase), which enriched in peroxidase activity (GO:0004601) and the peroxisome pathway (KEGG:04146). By the mid‐to‐terminal phase (Days 6–16), the fungi further adjusted genes involved in sulfur metabolism (KEGG:00920; GO:0006790) and iron homeostasis, demonstrating a continued adaptation to dynamic changes in environmental iron and sulfur concentrations.

Consistent with their acidification capacities, pyrite addition upregulated the expression of genes involved in organic acid synthesis and secretion. This upregulation was more pronounced in strains with a higher inherent acidogenic capacity, P. janthinellum ZJS52 and A. niger DBS124 (Figure S10). As the most effective acid‐producing and pyrite‐oxidizing strain, A. niger DBS124 functioned primarily via substantial oxalic acid secretion (Figure 3D). During the initial phase (Day 2), genes encoding oxaloacetate hydrolase were significantly upregulated (Figure 3E). Later, during the initial‐to‐intermediate phase (Days 4 and 8), genes involved in citric acid production also showed elevated expression (Figures 3F,H and S10). These results suggested that pyrite enhances the production of oxalic and citric acids, particularly during the initial phase when a sharp pH drop occurred. Furthermore, consistent with genomic annotation, the expression of genes encoding LiP and MnP was significantly induced by pyrite on Day 2 (Figure 4B,D), supporting activation of this oxidative peroxidase pathway during pyrite exposure.

4. Discussion

4.1. Mine Tailings‐Derived Fungi as a Widespread Yet Underexplored Resource for Pyrite Oxidation

Current research on fungal oxidation of metal sulfides has involved fungi from two phyla, Ascomycota and Basidiomycota, spanning a total of seven genera: Aspergillus, Penicillium, Trametes, Exophiala, Fusarium, Phanerochaete, and Trichoderma (Table S12; Cecchi et al. 2019; Yang et al. 2019). The isolation sources of these fungi are diverse, including decayed wood, industrially contaminated soils, and mine tailings (Cecchi et al. 2019; Etemadzadeh et al. 2016; Hein et al. 2023). Among the seven reported genera, only two were isolated from mine tailings. To our knowledge, no study has systematically isolated fungi from mine tailings and evaluated their metal sulfide oxidation potential on a large scale.

We conducted a large‐scale assessment of pyrite‐oxidizing capacity in fungi isolated from mine tailings. Among 125 fungal isolates, 51 significantly oxidised pyrite. These active isolates span three phyla and comprise 28 genera, with phylum Mucoromycota reported here for the first time in metal sulfide oxidation. This expands the known generic repertoire from seven to 28, including 25 genera newly documented for this trait. Notably, 15 of these genera have no reported association with any known mechanisms underlying metal sulfide oxidation (Table S3). Their activity was consistently accompanied by marked acidification of the culture medium.

A literature survey based on the taxonomic identities of these isolates revealed that these fungi, or their close relatives, have been reported to possess mechanisms potentially involved in pyrite oxidation (Table S3). This observation may be explained by the fact that phylogenetically related species often share similar metabolic pathways. The active isolates include organic acid‐producing genera such as Aspergillus and Penicillium, which are known to secrete organic acids, with some species producing oxalic acid (Han et al. 2007; Jiao et al. 2022). They also include fungi involved in lignocellulose decomposition, such as Marasmius cladophyllus and Trichoderma harzianum, which can employ organic acids and lignin‐modifying peroxidases (Janusz et al. 2017; Table S3).

To our knowledge, 18 of these 51 active isolates have not been reported to possess any known metal sulfide oxidation mechanisms (Table S3). The discovery of this capacity in less studied genera such as Pithomyces, Catenulostroma, and Ochrocladosporium suggested the possibility of previously unreported mechanisms. Notably, Pithomyces cynodontis did not exhibit strong medium acidification ability but showed the highest pyrite oxidation capacity among all the isolates. This implies that P. cynodontis may employ distinct pyrite‐oxidizing mechanisms beyond the common organic acid secretion pathway prevalent among mine tailings‐derived fungi. These newly identified fungal taxa with metal sulfide oxidation capacity provide a promising direction for future research into the diversity of fungal pyrite oxidation pathways.

4.2. Differences in Pyrite Bio‐Oxidation Capabilities and Mechanisms Between Three Focal Fungal Strains

The strains P. janthinellum ZJS52 and A. niger DBS124 belong to genera known for metal leaching activity. These two genera are recognised for secreting organic acids such as citrate, gluconate, and oxalate, and are commonly used in recovering heavy and valuable metals from solid wastes (Rasoulnia and Mousavi 2016; Yang et al. 2019). In contrast, P. cynodontis DC37 has not been well‐documented in this context.

During the 16‐day co‐culture with pyrite, P. cynodontis DC37, P. janthinellum ZJS52, and A. niger DBS124 released 1.24, 1.28, and 1.88 wt.% of iron from pyrite, respectively (Figure 1F). All three strains showed values comparable to or higher than that of Phanerochaete chrysosporium (1.20 wt.%), a well‐studied fungus in this field (Ofori‐Sarpong et al. 2011). In the initial phase (Days 2 and 4), each strain acidified the medium to below pH 3 through the secretion of organic acids such as citric and oxalic acid (Figure 1D). The resulting acidic environment dissolved insoluble Fe(III) passivation layers on pyrite, potentially promoting further oxidation (Caldeira et al. 2003; Dong et al. 2023; Srichandan et al. 2019). Furthermore, extracellular enzyme activities including MnP, LiP, and GLOX were detected in the culture systems of all strains (Figures 4A,C and S8A).

Although the types of oxidation mechanisms identified in this study were largely similar across the three fungal strains, they appeared to differ in their primary reliance on these mechanisms. P. cynodontis DC37 and P. janthinellum ZJS52 may oxidise pyrite primarily through the secretion of extracellular oxidative enzymes, with MnP and LiP potentially playing key roles. In contrast, A. niger DBS124 appeared to facilitate pyrite oxidation mainly via the production of oxalic acid, suggesting a metabolism‐mediated dissolution process. This functional divergence was further supported by chemical and molecular biological analyses under the test conditions.

4.3. Peroxidase Activities and Gene Expression During Pyrite Oxidation

LiP and MnP are heme‐containing oxidoreductases which utilise H₂O₂ as an electron acceptor and have been implicated in fungal oxidation of pyrite. They are primarily found in lignin‐oxidizing fungi, such as white‐rot fungi (Janusz et al. 2017; Ofori‐Sarpong et al. 2013). To our knowledge, however, no previous studies have reported the secretion of these peroxidases by the three focal strains examined here. We detected considerable LiP‐like and MnP‐like activities in the culture systems of P. cynodontis DC37 and P. janthinellum ZJS52 (Figure 4A,C). Furthermore, the corresponding genes were successfully annotated in the genome of P. cynodontis (Table S9), providing the first evidence of these specific peroxidase genes and their activities within the genus Pithomyces. However, the absence of homologous genes in P. janthinellum ZJS52, despite detectable enzyme activities, warrants further consideration. This discrepancy may be explained by at least two possibilities. Firstly, this strain may harbour MnP or LiP homologues with low sequence similarity to known reference sequences. Alternatively, the observed activity could be attributed to novel, uncharacterised enzymes that perform similar oxidative functions but share little sequence identity with known peroxidases. Secondly, the substrates used in the kit assays (veratryl alcohol for LiP and 2,6‐dimethoxyphenol (DMP) for MnP) are not exclusively specific to LiP and MnP. They can also be oxidised by other enzymes, such as aryl‐alcohol oxidase (AAO) and lytic polysaccharide monooxygenases (LPMOs), respectively (Breslmayr et al. 2018; Kumar and Goswami 2006). Therefore, the detected activities may not solely reflect the activity of canonical peroxidases. At present, both interpretations remain speculative. Future work should include MnP and LiP activity validation using complementary methods (e.g., capillary electrophoresis enzyme assays or nanostructure‐initiator mass spectrometry (Deng et al. 2018; Kudo et al. 2017)) and explore alternative approaches for identifying potential peroxidase‐encoding sequences, such as using home‐built profile hidden Markov models or structural homology searches.

We also examined the potential regulatory factors that may influence peroxidase gene expression in P. cynodontis DC37. Previous studies have shown that elevated levels of reactive oxygen species (ROS), particularly H₂O₂, can stimulate the expression of LiP and MnP encoding genes in white‐rot fungi (Belinky et al. 2003; Wiberth et al. 2019). During the intermediate phase (Days 6 and 8) of co‐cultivation, accelerated pyrite oxidation was associated with a coordinated induction of oxidative stress response genes in P. cynodontis DC37 (Table S11). This response likely originated from both exogenous ROS generated during pyrite oxidation (Tan et al. 2024; Wang et al. 2024; Zhao et al. 2022) and endogenous ROS produced as a byproduct of enhanced mitochondrial electron transport (Raha and Robinson 2001). Consistent with significant ROS generation, we also observed upregulation of genes involved in the repair of oxidative damage to DNA, lipids, and proteins. The presence of pyrite was associated with increased expression of genes encoding LiP and MnP in P. cynodontis DC37 (Figure 4B,D). The transcriptional response was consistent with the activation of ROS‐responsive regulatory pathways that upregulate these peroxidases as part of the fungal antioxidant defence under mineral‐induced oxidative stress.

4.4. Enhanced Oxalic Acid Secretion in A. niger DBS124 Associated With Pyrite‐Induced Metabolic Activation

The prominent bio‐oxidation capacity of A. niger DBS124 appears to be closely associated with its dominant secretion of oxalic acid (Figure 3D). Compared to citric and gluconic acids, oxalic acid is a stronger organic acid and contributed to the most acidic environment among the three strains (Figure 1D), which may promote pyrite oxidation (Tu et al. 2017). Moreover, oxalic acid can, in principle, be oxidised by MnP to generate hydrogen peroxide, potentially further contributing to pyrite oxidation (Kuan and Tien 1993; Urzúa et al. 1998).

The presence of pyrite was associated with enhanced oxalic acid secretion by A. niger DBS124, which aligned with pyrite‐induced changes in its central metabolism. During the initial phase (Days 2 and 4) of pyrite co‐cultivation, enrichment analyses revealed a coordinated upregulation of diverse metabolic pathways in all three fungi (Figure S9). This observation was consistent with previous reports that pyrite can stimulate microbial metabolic activity (Latorre et al. 2016; Zhang et al. 2024; Zhou et al. 2023). Exposure to pyrite was accompanied by increased transcriptional expression of genes encoding key enzymes in the oxalic acid biosynthetic pathway, specifically oxaloacetate hydrolase (Figure 3E,F). Additionally, pyrite exposure was associated with enhanced fungal energy metabolism and respiratory activity, as supported by both our data and previous studies (Zhang et al. 2024; Zhou et al. 2023), which may lead to elevated ATP production and overall metabolic turnover. Collectively, our results suggested that pyrite‐driven changes in energy metabolism, along with the upregulation of oxalic acid biosynthetic genes, may contribute to the increased oxalic acid secretion observed in A. niger DBS124.

In summary, using multiple methods, this study identified three fungal strains with strong pyrite oxidation capacity and revealed that they employed divergent mechanisms. However, several limitations should be acknowledged. Firstly, the kit used to measure MnP and LiP may lack strict enzyme specificity, and further experiments and gene annotation are needed to confirm these results. Secondly, the relative contributions of organic acid secretion versus peroxidase activity to pyrite oxidation were not distinguished or quantified in this study. Future work using genetic manipulation (e.g., knockout of genes involved in oxalic acid biosynthesis or peroxidase expression) or biochemical reconstitution with purified components will be required to quantitatively assess the contribution of each mechanism.

5. Conclusions

Our results suggested that the fungi isolated represent an underexplored resource for sulfide mineral oxidation. All three focal strains appeared to oxidise pyrite through acidification and peroxidase activity, with divergent primary mechanisms. A. niger DBS124 appeared to rely predominantly on oxalic acid secretion, whereas P. cynodontis DC37 and P. janthinellum ZJS52 showed evidence for greater MnP‐ and LiP‐like activities, respectively. The enhanced production of oxalic acid by A. niger DBS124, which likely contributes to its superior pyrite oxidation capacity, was consistent with pyrite‐associated upregulation of both central metabolic pathways and genes involved in specific oxalic acid biosynthesis. Additionally, we provided genetic evidence of LiP and MnP in P. cynodontis. Their expression appeared to be co‐induced by reactive oxygen species during pyrite oxidation, which may reflect an adaptive strategy linking oxidative stress defence to enzymatic oxidation. Collectively, our findings help to clarify how these fungi adapt to and oxidise pyrite in mine tailings and offer insight into the mechanistic basis for their potential application in bioleaching and bioremediation (Figure 5).

FIGURE 5.

Conceptual diagram of the mechanism of pyrite oxidation by fungi from mine tailings in South China. Arrows (with fungal species names) denote the primary pyrite oxidation mechanism employed by each fungus.

Author Contributions

Yu‐xiang Li: conceptualization, methodology, software, data curation, writing – original draft. Yan‐ying Li: software, data curation. Xiao‐lin Luo: methodology. Guan‐xiong Wu: methodology. Jia‐wei Li: methodology. Jia‐xin Wan: formal analysis, investigation. Xuan‐xuan Liu: formal analysis, investigation. Jie‐Liang Liang: conceptualization, writing – review and editing, funding acquisition, supervision. Kai‐tian Yang: methodology. Jin‐tian Li: writing – review and editing, funding acquisition, supervision. Zhuo‐hui Wu: writing – review and editing.

Funding

This work was supported by the National Natural Science Foundation of China (32470001 and 42177009), Guangdong Special Support Program for Young Top‐Notch Talent (2024TQ08A986), National Key Research and Development Program of China (2023YFC3207300).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Composition of the nitrogen‐limited liquid medium.

Table S2: Energy dispersive X‐ray spectroscopy (EDS) analysis of atomic composition at selected surface points on pyrite.

Table S3: Bioleaching capacity and pH in the system after co‐cultivation of pyrite powder with 51 metal‐sulfide‐oxidizing fungi isolated from the south China mine tailings.

Table S4: Species annotation and colony morphological descriptions of the three representative fungal strains.

Table S5: Comparison of XPS peak areas for pyrite (Fe(II) 2p_1/2 and S) and the oxidised iron layer (peaks between Fe(II) 2p and Fe(II) 2p_3/2).

Table S6: Genome assembly and annotation statistics of three fungal strains.

Table S7: Parameters and summary statistics of functional annotation for the three fungal genomes.

Table S8: Summary of annotated key genes involved in organic acid metabolism pathways in the three fungal strains.

Table S9: Annotation details of extracellular enzyme genes potentially involved in pyrite oxidation in P. cynodontis DC37.

Table S10: Summary of RNA‐seq data output and quality metrics.

Table S11: Adaptation‐related differentially expressed genes (DEGs) in three representative fungi in response to pyrite stress.

Table S12: Fungal species reported to oxidise metal sulfides.

Figure S1: The X‐ray diffractogram of pyrite in this study matches well with the standard pattern (JCPDS PDF#42‐1340).

Figure S2: Full‐range XPS survey spectra of pyrite before and after 20 days of incubation with three fungal strains.

Figure S3: High‐resolution XPS S 2p spectra of polished pyrite after 20‐day cultivation with (A) sterile culture control, (B) P. cynodontis DC37, (C) P. janthinellum ZJS52, and (D) A. niger DBS124. Peak assignments as follows: S²⁻ (162.00 and 162.95 eV), S₂ ²⁻ (162.57 and 163.76 eV), S_n ²⁻/S⁰ (164.62 and 165.71 eV) and SO₄ ²⁻ (168.68 and 169.91 eV) (Tu et al. 2017; Yang et al. 2018).

Figure S4: High‐resolution XPS O 1s spectra of polished pyrite after 20‐day cultivation with (A) sterile culture control, (B) P. cynodontis DC37, (C) P. janthinellum ZJS52, and (D) A. niger DBS124. The peak at 530.2 eV corresponds to O²⁻ in iron oxides (e.g., Fe₂O₃), the peak at 531.6 eV is attributed to OH⁻ in hydroxides such as Fe(OH)₃ and FeOOH, and the peak at 532.6 eV is associated with adsorbed H₂O. It should be noted that the adsorbed water signal is highly dependent on sample transfer conditions and should be considered qualitative rather than quantitative.

Figure S5: Location and field views of the three studied mine tailing sites. (A) Map of China showing the distribution of the sampling sites. (B) Representative field photographs of the sampling sites. From left to right: Lutang Copper Mine (Guangxi Zhuang Autonomous Region), Dabaoshan Mine (Guangdong Province), and Zijin Mountain Mine (Fujian Province).

Figure S6: Colony morphology of a subset of fungal isolates from three tailings sites. The image shows colonies of 33 out of the 51 total fungal isolates obtained from Lutang Copper Mine (Guangxi Zhuang Autonomous Region), Dabaoshan Mine (Guangdong Province), and Zijin Mountain Mine (Fujian Province) grown on PDA plates.

Figure S7: SEM images (A–F) of polished pyrite surface after 20 days co‐culture with three representative fungal strains. The diameter of each pit was defined as the longest distance along its short axis, and the three largest such diameters are presented.

Figure S8: Temporal profiles of (A) Glyoxal oxidase activity and (B) hydrogen peroxide concentration in fungal‐pyrite co‐culture systems. Data are presented as mean ± standard deviation (n = 3).

Figure S9: Functional enrichment analysis of differentially expressed genes (DEGs) in three fungal strains across three phases during pyrite bio‐oxidation. Strains include: (A) P. cynodontis DC37, (B) P. janthinellum ZJS52, and (C) A. niger DBS124. The colour intensity of each bubble represents the statistical significance (−log₁₀(p‐value)) of the enriched KEGG pathway or GO term, while the bubble size corresponds to the enrichment score.

Figure S10: Expression patterns of genes involved in the production and secretion of citrate acid in response to pyrite over the cultivation period in the three fungal strains. Asterisks indicate statistical significance (DESeq2 with Benjamini–Hochberg adjusted p‐values): *p _adj < 0.05, **p _adj < 0.01, ***p _adj < 0.001.

Data Availability Statement

The genome assemblies and transcriptomic sequencing data analysed in this study have been deposited in the National Genomics Data Center, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences. The data are publicly available at https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA062661 under the accession number PRJCA062661.