Identification and Characterization of the Determinants of Copper Resistance in the Acidophilic Fungus Acidomyces richmondensis MEY-1 Using the CRISPR/Cas9 System

ABSTRACT

Copper (Cu) homeostasis has not been well documented in filamentous fungi, especially extremophiles. One of the main obstacles impeding their characterization is the lack of a powerful genome-editing tool. In this study, we applied a CRISPR/Cas9 system for efficient targeted gene disruption in the acidophilic fungus Acidomyces richmondensis MEY-1, formerly known as Bispora sp. strain MEY-1. Using this system, we investigated the basis of Cu tolerance in strain MEY-1. This strain has extremely high Cu tolerance among filamentous fungi, and the transcription factor ArAceA (A. richmondensis AceA) has been shown to be involved in this process. The ArAceA deletion mutant (ΔArAceA) exhibits specific growth defects at Cu concentrations of ≥10 mM and is transcriptionally more sensitive to Cu than the wild-type strain. In addition, the putative metallothionein ArCrdA was involved in Cu tolerance only under high Cu concentrations. MEY-1 has no Aspergillus nidulans CrpA homologs, which are targets of AceA-like transcription factors and play a role in Cu tolerance. Instead, we identified the Cu-transporting P-type ATPase ArYgA, homologous to A. nidulans YgA, which was involved in pigmentation rather than Cu tolerance. When the ΔArYgA mutant was grown on medium supplemented with Cu ions, the black color was completely restored. The lack of CrpA homologs in A. richmondensis MEY-1 and its high tolerance to Cu suggest that a novel Cu detoxification mechanism differing from the AceA-CrpA axis exists.

IMPORTANCE Filamentous fungi are widely distributed worldwide and play an important ecological role as decomposers. However, the mechanisms of their adaptability to various environments are not fully understood. Various extremely acidophilic filamentous fungi have been isolated from acidic mine drainage (AMD) with extremely low pH and high heavy metal and sulfate concentrations, including A. richmondensis. The lack of genetic engineering tools, particularly genome-editing tools, hinders the study of these acidophilic and heavy metal-resistant fungi at the molecular level. Here, we first applied a CRISPR/Cas9-mediated gene-editing system to A. richmondensis MEY-1. Using this system, we identified and characterized the determinants of Cu resistance in A. richmondensis MEY-1. The conserved roles of the Cu-binding transcription factor ArAceA in Cu tolerance and the Cu-transporting P-type ATPase ArYgA in the Cu-dependent production of pigment were confirmed. Our findings provide insights into the molecular basis of Cu tolerance in the acidophilic fungus A. richmondensis MEY-1. Furthermore, the CRISPR/Cas9 system used here would be a powerful tool for studies of the mechanisms of adaptability of acidophilic fungi to extreme environments.

INTRODUCTION

Copper (Cu) is an essential micronutrient for all life forms that serves as a cofactor for numerous enzymes (1). However, excess Cu can be detrimental. To maintain Cu homeostasis, all biological systems have developed sophisticated mechanisms to cope with Cu fluctuations in ecological niches, including Cu uptake, transport, storage, and detoxification, mediated by Cu-binding transcription factors (TFs), Cu importers/exporters, and Cu-binding metallothioneins (MTs) (2). Cu-binding TFs regulate the transcription of specific genes, such as Cu transporters for Cu uptake under Cu-insufficient conditions and Cu exporters for detoxification under excess-Cu conditions (3). MTs are a group of ubiquitous, small, cysteine-rich, metal-binding proteins with the ability to coordinate metal ions through metal-thiolate bonds and are important for protection against the toxicity of metals (4). The multiple cysteine residues in MTs are arranged in Cys-X-Cys or Cys-X₂-Cys repeats that coordinate metal binding (5).

In Saccharomyces cerevisiae, the Cu-binding TF Mac1p binds to the promoters of two high-affinity transporter genes, CTR1 and CTR3, and activates their transcription under Cu-insufficient conditions (6). However, the binding of excess Cu to Mac1p releases the protein from the CTR1 promoter, preventing further Cu uptake (7). Mac1p homologs have been identified in Trichoderma reesei (Tmac1), Aspergillus nidulans (AnMac1), as well as several human fungal pathogens, including Candida albicans (CaMac1), Cryptococcus neoformans (Cuf1), Aspergillus fumigatus (AfMac1), and Histoplasma capsulatum (Mac1), which function in a similar way and are necessary for virulence (8–15). Under excess-Cu conditions, another Cu-binding TF, Ace1p, activates Cu-detoxifying genes such as CUP1 and CRS5, encoding MTs, and SOD1, encoding superoxide dismutase (16–20). In C. neoformans, two MTs, Cmt1 and Cmt2, were shown to be important for Cu detoxification and fungal virulence; they are both regulated by the Cu-binding TF Cuf1 (9, 21). Similarly, in Candida glabrata, the Ace1p homolog Amt1 activates the MT-I and MT-II genes in response to Cu, where this activation is essential for high-level Cu resistance (22, 23). In C. albicans, the membrane-localized P-type ATPase CaCrp1 and two MTs, CaCup1 and CaCrd2, are involved in Cu resistance, with CaCrp1 having a more prominent role than CaCup1 and CaCrd2 (24, 25). Another P-type ATPase, CaCcc2, is involved in cuproenzyme biosynthesis but not Cu detoxification (26).

Previous studies have shown that filamentous fungi rely mainly on conserved P-type ATPases other than MTs for Cu detoxification. In A. nidulans, the CaCrp1 homolog CrpA (AN3117) is the major determinant of Cu resistance, and the metal-dependent induction of CrpA is under the control of the TF AceA, the ortholog of Ace1p (27). The A. nidulans genome encodes a second P-type ATPase, YgA (AN3624), with a high level of homology to CaCcc2, which is involved in conidial pigmentation (28). Similarly, in A. fumigatus, Aspergillus flavus, and Fusarium graminearum, AceA induces the expression of the P-type ATPase CrpA (and CrpB in A. flavus) for Cu detoxification under excess-Cu conditions (29–31). In Fusarium oxysporum, the A. nidulans CrpA homolog CrpF was also shown to be responsible for Cu tolerance, although the transcriptional regulation of crpF has yet to be clarified (32). The A. nidulans CrpA homolog PcpA of Penicillium janthinellum was also shown to be involved in Cu tolerance (33).

Filamentous fungi are widely distributed in various ecological niches, including acid mine drainage (AMD), with extremely low pH and high heavy metal and sulfate concentrations. As major contributors in AMD environments, fungi can adapt to low pH and high heavy metal concentrations. Various acidophilic and acid-tolerant fungi have been isolated, but only a few have been sequenced, including Acidomyces richmondensis, isolated from the AMD of the Richmond Mine at Iron Mountain in California (34–36) (BioProject accession no. PRJNA207869); Acidiella bohemica, isolated from the AMD of the Fankou Pb/Zn Mine in China (37) (BioProject accession no. PRJNA725650); and Acidothrix acidophila, isolated from extremely acidic soils (38) (https://mycocosm.jgi.doe.gov/Aciaci1). Genomic and transcriptomic analyses provide insight into the adaptation mechanisms of extremophilic fungi. However, genetic engineering of these acidophilic fungi is rarely reported, hampering efforts to understand their adaptation mechanisms at the molecular level. At present, only Agrobacterium tumefaciens-mediated transformation of the acidophilic fungi Acidea extrema and A. acidophila has been reported, and it is time-consuming (>28 days) and has a low transformation efficiency (39). In addition, there have been no reports on the gene disruption of these acidophilic fungi. Therefore, the development of efficient genetic tools for acidophilic fungi would help us to obtain a deeper mechanistic understanding of their adaptability to acidic and high heavy metal conditions in AMD environments. In this study, a CRISPR/Cas9-based genome-editing system developed for Myceliophthora thermophila (40) was used to knock out A. richmondensis MEY-1 genes via polyethylene glycol (PEG)-mediated protoplast transformation. Using this method, we identified genes involved in Cu resistance, including a major contributor, the Cu-binding TF ArAceA, and a minor contributor, the MT ArCrdA. The deletion of ArAceA resulted in a significant reduction in Cu resistance. The ΔArAceA mutant strain was transcriptionally more sensitive to Cu, whereas the ΔArCrdA strain displayed only slightly increased susceptibility to high Cu levels. Unexpectedly, different from the BFW strain, MEY-1 has no CrpA homologs, suggesting that a different mechanism mediates Cu resistance in strain MEY-1.

RESULTS

Reclassification of Bispora sp. MEY-1 as A. richmondensis MEY-1.

The acidophilic fungus Bispora sp. strain MEY-1 was first isolated from AMD in China (41). Its internal transcribed spacer (ITS) region (GenBank accession no. EU927406) was identical to that of the A. richmondensis BFW strain, and the 18S rRNA gene sequence obtained by whole-genome sequencing (GenBank accession no. OQ346270) shared 99.88% nucleotide identity with that from the A. richmondensis BFW strain. In addition, all published protein-coding gene sequences from Bispora sp. MEY-1 (GenBank accession no. KJ866875, HM003043, HM003044, HM003045, GU351880, GU074519, FJ492963, FJ212324, FJ472925, EU919724, MG520667, MK821250, FJ695140, and MN326865) share 99 to 100% identity with those from the A. richmondensis BFW strain (35). Based on this evidence, we propose to reclassify Bispora sp. MEY-1 as A. richmondensis MEY-1.

A. richmondensis MEY-1 has high-level Cu tolerance.

In addition to the lower pH, AMD usually contains high concentrations of heavy metals. Thus, the microbial residents must evolve to achieve high-level tolerance to these metals. Since A. richmondensis MEY-1 was isolated from AMD and Cu was one of the heavy metals in this environment, its Cu tolerance was determined. Compared to other filamentous fungi that could not grow at concentrations of 10 or 20 mM CuCl₂, the growth of A. richmondensis MEY-1 was only slightly affected by 30 mM CuCl₂ (Fig. 1), indicating that A. richmondensis MEY-1 is highly tolerant to Cu.

FIG 1.

Growth of A. richmondensis MEY-1 and other filamentous fungi under different copper (Cu) concentrations. Vogel’s minimal medium (MM) supplemented with 2% sucrose and 20 mM citrate (MMC) was used for A. richmondensis MEY-1; MM was used for other fungi. Humicola insolens Y1 and M. thermophila ATCC 42464 were incubated at 42°C for 4 days, Chaetomium strumarium CY-3 was incubated at 37°C for 4 days, and all other fungi were incubated at 28°C for 4 days. Detailed information about these fungal strains is shown in Table S1 in the supplemental material. A. jensenii, Aspergillus jensenii; P. citrinum, Penicillium citrinum; A. alternata, Alternaria alternata; S. commune, Schizophyllum commune; T. purpureogenus, Talaromyces purpureogenus; P. lilacinum, Purpureocillium lilacinum; P. chrysogenum, Penicillium chrysogenum; P. spadiceum, Porostereum spadiceum; I. lacteus, Irpex lacteus; D. gregaria, Dothiorella gregaria.

The AceA-like transcription factor ArAceA is involved in Cu detoxification.

According to homology searches using the A. fumigatus Cu detoxification transcription factor AfAceA (AFUB_036430), the putative TF M433DRAFT_149096 (named ArAceA here) was found in the A. richmondensis MEY-1 strain. Like AfAceA, F. graminearum AceA (FgAceA), AnAceA, AMT1 (42), and Cup2, ArAceA also has the N-terminal conserved signature, including a Zn module and a Cu regulatory domain, which are typical features of putative Cu detoxification TFs (29) (see Fig. S1A in the supplemental material). The Zn module of ArAceA (residues 2 to 41) contains a Zn-binding motif (Cys-X₂-Cys-X₈-Cys-X-His) and a conserved (R/K)GRP sequence motif, which are essential for minor groove site-specific binding (43). The Cu regulatory domain contains eight critical Cys residues arranged in one Cys-X₂-Cys and three Cys-X-Cys motifs, which cooperatively bind four Cu(I) ions into a tetracopper cluster (44, 45). Although the N terminus of ArAceA has high identity to those of its homologs, phylogenetic analysis using MEGA7 software (46) based on their full-length protein sequences showed that ArAceA clustered into a group far from the group in which the Aspergillus, Trichoderma, and Penicillium genera were clustered (Fig. S1B), reflecting their different evolutionary trajectories (shaped by different natural niches).

For the development of a rapid and efficient transformation method for A. richmondensis MEY-1, we tested Geneticin (G418) and hygromycin B, currently used as selectable markers for fungal transformations. We found that A. richmondensis MEY-1 is hypersensitive to Geneticin and hygromycin B. Conidial germination of the wild-type (WT) strain was completely inhibited on medium containing 5 μg/mL Geneticin or 4 μg/mL hygromycin B (Fig. S2). Thus, both antibiotics were used to select A. richmondensis MEY-1 transformants. To assess the role of ArAceA in Cu tolerance, we generated a single-knockout mutant for the ArAceA locus using the CRISPR/Cas9 system (Fig. S3) as well as a complemented strain, ArAceA^C. The ΔArAceA mutant, WT, and ArAceA^C strains were tested at different concentrations of various heavy metal ions. As shown in Fig. 2A, only the ΔArAceA mutant showed reduced resistance to Cu²⁺, exhibiting morphological defects at 10 mM CuCl₂ and total growth inhibition at 20 mM CuCl₂. This Cu-sensitive growth phenotype is probably due to the excessive accumulation of copper caused by a deficiency in copper export, which is similar to the ΔFgAceA mutant of F. graminearum (31). The reduced resistance to Cu²⁺ was rescued to WT levels by the introduction of the A. richmondensis MEY-1 ArAceA gene into the ΔArAceA strain, confirming the involvement of the TF ArAceA in Cu tolerance (Fig. 2A and B). Additionally, ArAceA was not involved in H₂O₂-induced oxidative stress or osmotic stress (Fig. 2C).

FIG 2.

ArAceA is involved in Cu detoxification. (A) Growth of the wild-type (WT) strain, the ΔArAceA mutant, and the complemented strain (ArAceA^C) under different concentrations of various heavy metal ions. (B) Plate growth diameters of the WT, ΔArAceA, and ArAceA^C strains under different concentrations of CuCl₂. Error bars represent standard deviations from three independent experiments. (C) Growth of the WT, ΔArAceA, and ArAceA^C strains under different concentrations of NaCl and H₂O₂. Equal numbers of arthroconidia (1 × 10⁴) were spotted onto MMC plates. Growth was observed after incubation for 4 days at 28°C.

Comparative transcriptional analysis of the WT strain and the ΔArAceA mutant exposed to Cu.

To gain genomic insight into the Cu tolerance mechanism and the role of ArAceA in Cu tolerance in A. richmondensis MEY-1, we treated the ΔArAceA mutant and WT strains with or without 15 mM CuCl₂ for 6 h. Subsequently, RNA was extracted and prepared for high-throughput RNA sequencing (RNA-Seq). Given that the A. richmondensis BFW strain displays high genome completeness (99.5%) (35) and that there is very high sequence identity between the two strains, the genome of strain BFW was used as the reference genome for RNA-Seq analysis. The percentage of reads mapped to the reference genome ranged from 93.61% to 94.86% (94.14% on average). Of the 11,177 protein-coding genes in strain BFW, 10,183 genes (91.1%) were detected in strain MEY-1. Pearson correlation analysis demonstrated that the biological replicates were reliable for all tested samples (Fig. S4). RNA-Seq data (Table S3) from the WT and ΔArAceA biological replicates were subjected to principal-component analysis (PCA); data from the same strains grown under the same growth conditions were clustered. Data from the WT strain were clustered, while data from the ΔArAceA mutant were dispersed (Fig. 3A). This indicated that 15 mM Cu²⁺ has little effect on the gene expression profile of the WT strain and that the deletion of ArAceA increased the sensitivity to Cu²⁺ exposure. Consistent with these observations, the number of differentially expressed genes (DEGs) between the WT strain with and without Cu exposure (WT_Cu versus WT, 45 upregulated and 4 downregulated) was much lower than that between the ΔArAceA mutant with and without Cu exposure (ΔArAceA_Cu vs. ΔArAceA, 89 upregulated and 197 downregulated) (Fig. 3B). In addition, there was a relatively large number of DEGs in ΔArAceA versus WT (248 upregulated and 49 downregulated) compared to ΔArAceA_Cu versus WT_Cu (87 upregulated and 90 downregulated) (Fig. 3C). Among the 248 upregulated genes in ΔArAceA versus WT, 100 genes were also among the 197 downregulated genes in ΔArAceA_Cu versus ΔArAceA (Fig. 3D; Table S4). Gene ontology (GO) enrichment analysis showed that the 297 DEGs in ΔArAceA versus WT were significantly enriched in various metabolic processes (Fig. S5 and Table S5), indicating that ArAceA has an additional role in regulating metabolism besides Cu resistance.

FIG 3.

The ΔArAceA mutant is more sensitive to Cu. (A) Principal-component analysis of RNA-Seq data from the WT and the ΔArAceA mutant grown in medium with or without 15 mM CuCl₂. (B) Venn diagram of differentially expressed genes (DEGs) induced by Cu in the WT strain and the ΔArAceA mutant. (C) Venn diagram of DEGs between the WT strain and the ΔArAceA mutant under normal and Cu conditions. (D) Normalized gene expression of DEGs in panels B and C. Rows are centered; unit variance scaling is applied to rows. Both rows and columns are clustered using the correlation distance and average linkage. Genes with a fold change of >2.0, a Q value of <0.05, and an FPKM value of >20 under at least one condition or in at least one strain were considered significantly differentially expressed.

In the WT strain, Cu significantly upregulated the expression of two membrane protein-coding genes, M433DRAFT_310551 and M433DRAFT_348929, whereas the upregulation was abolished in the ΔArAceA mutant (Fig. S6), indicating that both genes were candidate targets for ArAceA. M433DRAFT_310551 contains 327 amino acids (aa) with six predicted transmembrane helices (TMHs), while M433DRAFT_348929 contains 411 aa with seven TMHs. M433DRAFT_348929 (Tzn-1) shares 47.4% identity with the zinc transporter TZN-1 (NCU07621) of Neurospora crassa (47). To assess the roles of the M433DRAFT_310551 and tzn-1 genes in Cu tolerance, both genes were deleted (Fig. S7A, B, D, and E). No obvious difference in Cu tolerance was noted between the ΔM433DRAFT_310551 or the Δtzn-1 mutant and the WT strain (Fig. S7C and F).

A Cu-transporting ATPase is involved in pigmentation but not Cu resistance.

In Aspergillus and Fusarium species, the Cu²⁺-exporting ATPase CrpA and its regulator AceA are necessary for Cu resistance. Using BLASTP analysis, two A. nidulans CrpA orthologs, M433DRAFT_720 and M433DRAFT_149502, were identified in the A. richmondensis BFW strain, with identities of 41.3% and 42%, respectively. However, neither gene was detected in our MEY-1 strain using gene-specific primers, and there were no RNA-Seq reads mapping to these two genes. Comparative analysis of the gene organizations identified a specific genomic region comprising 77.8 kb that contained both M433DRAFT_720 and M433DRAFT_149502 in strain BFW (Fig. S8). Transcriptomic analysis of strain MEY-1 showed that there were no RNA-Seq reads mapping to any genes in this region, suggesting that this region is absent in strain MEY-1. The presence of this region in strain BFW is probably due to horizontal gene transfer (HGT) between eukaryotic species, as is the case with the large transposable element HEPHAESTUS (Hφ) recently found in certain fungal strains. Paecilomyces variotii strains containing Hφ are more tolerant to toxic metal/metalloid ions than Hφ-lacking strains (48). Notably, there were two other putative P-type ATPase-encoding genes (M433DRAFT_58219 and M433DRAFT_718) and three putative multicopper oxidase-encoding genes (M433DRAFT_149492, M433DRAFT_58180, and M433DRAFT_730) in this region (Fig. S8). We speculated that this region would lead to a greater tolerance of strain BFW to Cu, as A. flavus containing two homologs of A. nidulans CrpA tolerates higher concentrations of Cu than other Aspergillus spp. (30).

Another P-type ATPase, M433DRAFT_66060, shared 32.2% sequence identity with A. nidulans CrpA but had the highest sequence identity (58.2%) with A. nidulans YgA. Therefore, M433DRAFT_66060 is the putative ortholog of A. nidulans YgA, named ArYgA here. Bioinformatic analyses predicted the presence of distinctive domains described for other well-studied Cu-transporting ATPases (Fig. 4A). These domains include three GMxCxxC classical heavy metal-associated domains (HMAs) in the cytoplasmic N terminus, a phosphatase domain (TGES), a conserved Cu translocation motif (CPC) placed in the sixth transmembrane (TM) domain, an aspartyl kinase domain (DKTG) in a large cytoplasmic loop containing an aspartate residue transiently phosphorylated during the catalytic cycle (49), and a conserved domain for ATP binding and energy transduction (GDGINDSP) (50). The N-terminal Cu-binding domain is more similar to those of A. nidulans YgA and YgA homologs, including CtpA of A. fumigatus (51), BcCCC2 of Botrytis cinerea (52), and CLAP1 of Colletotrichum lindemuthianum (53).

FIG 4.

Functional analysis of the Cu-transporting ATPase ArYgA. (A) Sequence analysis of ArYgA and alignment of ArYgA with F. oxysporum CrpF (FOXG_03265), F. graminearum FgCrpA (FGRAMPH1_01T10037), A. flavus CrpA (AFLA_020960) and CrpB (AFLA_053470), C. albicans CaCrp1p (GenBank accession no. AAF78958), A. fumigatus AfCrpA (Afu3g12740) and CtpA (Afu4g12620), A. nidulans CrpA (AN3117) and YgA (AN3624), B. cinerea BcCcc2 (BC1G_10836), C. lindemuthianum CLAP1 (GenBank accession no. AF494193), and S. cerevisiae ScCcc2p (YDR270W). (B) Strategy for the construction of a ΔArYgA mutant via homologous recombination and diagnostic PCR of the ΔArYgA mutant using the primer pair YgA-in-F/YgA-in-R. The ΔArYgA mutant displayed a 2,709-bp product, while the WT displayed a 695-bp product. (C) Mutant characterization in MMC supplemented with the indicated concentrations of CuCl₂. Arthrospores (1 × 10⁴) were point inoculated onto MMC plates. Images of colonies were taken after 6 days of incubation at 28°C. M, molecular marker.

Transcriptomic data showed that the expression of ArYgA is not regulated by Cu or ArAceA (Table S3). To assess the role of the ArYgA gene in Cu tolerance, a knockout mutant (ΔArYgA) was generated using the CRISPR/Cas9 system (Fig. 4B). As shown in Fig. 4C, there was no clear difference in the growth rate or Cu resistance between the WT strain and the ΔArYgA mutant. However, the melanization of the mutant was severely impaired compared with that of the WT strain. The impairment of melanization was reversed when the ΔArYgA mutant was grown in medium supplemented with ≥1 mM CuCl₂ (Fig. 4C), suggesting that the defect in the melanization of the ΔArYgA mutant is likely caused by a deficiency in copper uptake. Consistent with the N-terminal sequence similarities, the impaired pigmentation of the ΔArYgA mutant and the reversal thereof by Cu addition were also observed in the ctpA mutant of A. fumigatus (51), the BcCcc2 mutant of B. cinerea (52), and the clap1 mutant of C. lindemuthianum (53), indicating the functional conservation of these homologous Cu-transporting ATPases.

The metallothionein ArCrdA participates in high-Cu resistance.

Previous studies have shown that MTs play a significant role in Cu tolerance (3, 54). In A. richmondensis, a single putative MT-like protein (M433DRAFT_300682 [ArCrdA]) was identified. The expression of ArCrdA was not regulated by Cu or ArAceA (Table S3). The 112 residues of ArCrdA share 60% identity (84% query coverage) with MT AnCrdA (AN7011) of A. nidulans, which is a possible ortholog of CaCrd2. Sequence alignment revealed the presence of conserved CxC motifs that resemble those found in other fungi (Fig. 5A). Previous studies showed that deletions of AnCrdA in A. nidulans and its homolog AfCrdA in A. fumigatus did not lead to a Cu-sensitive phenotype (27, 29). Similarly, the AnCrdA homolog FgCrdA in F. graminearum was not involved in Cu resistance (31). To assess the role of ArCrdA in Cu tolerance, we generated a single-knockout mutant for ArCrdA using the CRISPR/Cas9 system (Fig. 5B). The ΔArCrdA mutant and WT strains were tested at different Cu concentrations. As shown in Fig. 5C and D, the ΔArCrdA mutant displayed slightly greater susceptibility to Cu, especially at high Cu concentrations.

FIG 5.

Functional analysis of the metallothionein ArCrdA. (A) Sequence alignment of ArCrdA with A. nidulans AnCrdA (GenBank accession no. CBF79264), A. fumigatus AfCrdA (AFUB_098700), F. graminearum FgCrdA (FGRAMPH1_01T09281), and C. albicans CaCrd2p (GenBank accession no. AAF78959). Asterisks denote identical residues. Double and single dots denote conserved and semiconserved residues, respectively. CxC repeats are in light-blue boxes. (B) Strategy for the construction of a ΔArCrdA mutant via homologous recombination and diagnostic PCR of the ΔArCrdA mutant using primer pair crdA-up-F/crdA-down-R. The ΔArCrdA mutant displayed a 3,762-bp product, while the WT displayed a 1,700-bp product. (C) Fungal colonies from the WT strain and the ΔArCrdA mutant grown for 4 days at 28°C on MMC plates containing the indicated concentrations of CuCl₂. (D) Plate growth diameters of the WT strain and the ΔArCrdA mutant under different concentrations of CuCl₂. Arthrospores (1 × 10⁴) were point inoculated onto MMC plates. Error bars represent standard deviations from three independent experiments.

DISCUSSION

Fungi inhabiting AMD environments have attracted much attention due to their potential for decomposing organic carbon, absorbing heavy metals, and reducing AMD acidity. However, studies on fungi in AMD environments are limited mainly to morphological and community composition analyses of fungal species. The molecular basis for the adaptation of acidophilic fungi to AMD environments has rarely been reported, due mainly to the lack of an efficient targeted genomic editing system. Although an Agrobacterium-mediated transformation system has been developed for the acidophilic fungi A. extrema and A. acidophila, there have been no reports on gene disruption. Here, we showed that the CRISPR/Cas9 system developed for M. thermophila functions well in the acidophilic fungus A. richmondensis MEY-1. Using the CRISPR/Cas9 editing system, the efficiency of gene disruption was approximately 16.7 to 47.1% (see Table S6 in the supplemental material). Despite the relatively low disruption efficiency, this system is necessary to enable gene disruption in strain MEY-1. Considering that this CRISPR/Cas9 system has been modified to edit the genomes of Humicola insolens Y1 and Aspergillus niger N1 (55, 56), it has the potential for application to other nonmodel fungi, including extremophilic organisms.

In eukaryotes, physiological Cu levels are maintained by regulating the balance among Cu uptake, compartmentalization, and detoxification. In this study, we identified two operators of the Cu detoxification system of A. richmondensis MEY-1. The Cu-binding TF ArAceA has a more prominent role in Cu resistance; the MT ArCrdA plays a minor role. Although AceA homologs in various fungi were found to be involved mainly in Cu detoxification, there are minor differences in their functions. For example, the ΔAceA mutant of A. nidulans is sensitive to high Cd levels (27), and the ΔAfAceA mutant of A. fumigatus is sensitive to high Zn levels (29), whereas the ΔArAceA mutant of A. richmondensis MEY-1 is sensitive to Cu only (Fig. 2). In addition, the functions of MTs also differ in these fungi. The AfCrdA mutant of A. fumigatus and the CrdA mutant of A. nidulans are not sensitive to high Cu levels (27, 29), whereas the ΔArCrdA mutant of A. richmondensis MEY-1 is involved in resistance to higher Cu levels (Fig. 5). Different from these fungi in which MTs play no or minor roles in Cu tolerance, F. oxysporum MT1, although not a CrdA homolog, is essential for Cu, Cd, and Zn resistance (54). The different functions of AceA and CrdA homologs among these fungi are probably due to their evolutionary trajectories, as indicated by phylogenetic analyses of their AceA homologs (Fig. S1).

The TF AceA and its homologs exert their effects by activating the downstream target gene CrpA or its homologous genes. This AceA-CrpA axis is conserved in Aspergillus species and F. graminearum. In the A. richmondensis BFW strain, two CrpA homologs (M433DRAFT_720 and M433DRAFT_149502) were present. However, strain MEY-1 does not carry either gene. Both genes are located in a specific gene cluster in strain BFW (Fig. S8). Since the AMD biofilm where the A. richmondensis BFW strain was isolated is a complex community comprised of different fungi along with bacteria, archaea, and bacteriophages, interspecies or intraspecies HGT may occur. Actually, HGT occurs frequently among filamentous fungi, conferring selective advantages to the recipient fungal host. In addition, biofilms in water environments are thought to be hot spots for HGT (57). Taken together, this specific gene cluster was probably acquired by the BFW strain through HGT to adapt to the AMD environment. A Cu-transporting rather than a Cu-exporting ATPase (M433DRAFT_66060 [ArYgA]), homologous to A. nidulans YgA, was found and shown to have a conserved function in pigmentation but not Cu tolerance (Fig. 4). The absence of a conserved direct target of ArAceA in A. richmondensis MEY-1 implies that strain MEY-1 uses an alternative ArAceA target to cope with Cu toxification, which needs further investigation.

MATERIALS AND METHODS

Strains, media, and growth conditions.

A. richmondensis MEY-1 was used as the parental wild-type (WT) strain. For sporulation, A. richmondensis strains were grown on potato dextrose agar (PDA) medium supplemented with 20 mM citrate (PDAC) at 30°C for 10 to 15 days. Citrate was added to the medium separately to adjust the pH to <3.0. To assess mycelial growth, the WT MEY-1 strain and its corresponding mutants were grown on Vogel’s minimal medium (MM) (58) supplemented with 2% sucrose, 20 mM citrate (MMC), and different concentrations of CuCl₂ and incubated at 30°C for 7 days. For flask cultures, arthroconidia of A. richmondensis MEY-1 and its corresponding mutants were inoculated into 100 mL of MMC (with or without the addition of 15 mM CuCl₂) to a final concentration of 2 × 10⁵ arthroconidia/mL in a 250-mL Erlenmeyer flask. Selection medium (0.5× Vogel’s salts, 85 g/L sucrose, 1.5 g/L tryptone, 3 g/L yeast extract, 91 g/L sorbitol, and 7.5 g/L agarose) was used for screening transformants after protoplast transformation.

Other fungal strains used in this study are summarized in Table S1 in the supplemental material. The isolated fungal strains were identified based on morphology and ITS sequence analysis. The ITS region of each fungus was amplified by PCR with universal ITS primers and sequenced.

Construction of cassettes for CRISPR/Cas9-mediated genome editing.

The primers used in this study are listed in Table S2. For the deletion of M433DRAFT_149096 (ArAceA), a 638-bp region upstream of ArAceA (5′-flanking region, extending from positions +116 to +753 with respect to the ArAceA translation start point) was amplified using primer pair aceA-up-F/aceA-up-R, and a 658-bp region downstream of ArAceA (3′-flanking region, extending from positions +1022 to +1679 with respect to the ArAceA translation start point) of ArAceA was amplified using primer pair aceA-down-F/aceA-down-R. The promoter region of the gene (M433DRAFT_131430) encoding glyceraldehyde-3-phosphate dehydrogenase (Pgpd) was amplified from A. richmondensis MEY-1 genomic DNA with the primer pair Pgpd-F/Pgpd-R. The fragment containing the Geneticin resistance gene (neo) and its terminator was amplified using primer pair neo-F/neo-T-R. These fragments were assembled into the pJET1.2 cloning vector (Thermo Fisher Scientific, Waltham, MA, USA), which was amplified by primer pair pJET-F/pJET-R using a ClonExpress Ultra one-step cloning kit (Vazyme Biotech, Nanjing, China). The fragment containing 5′-Pgpd-neo-3′ was amplified using the primer pair aceA-up-F/aceA-down-R and used for protoplast transformation of the WT strain. For complementation, the fragment PaceA-ArAceA-TaceA containing the ArAceA promoter (PaceA), the ArAceA coding sequence, and the ArAceA terminator (TaceA) was amplified using primer pair PaceA-F/TaceA-R. The fragment containing the hygromycin B resistance gene (hph) and the terminator of the gpd gene (Tgpd) from H. insolens was amplified from pAg1-hyg-P-T (59) using the primer pair hph-F/Tgpd-R. The fragments PaceA-ArAceA-TaceA, A. richmondensis MEY-1 Pgpd, and hph-Tgpd were cloned into the pEASY-Blunt Simple cloning vector (TransGen Biotech, Beijing, China) using the ClonExpress Ultra one-step cloning kit. Next, the Pgpd-hph-Tgpd-PaceA-ArAceA-TaceA cassette was amplified using the primer pair Pgpd-F/TaceA-R and used for protoplast transformation of the ΔArAceA mutant. After being randomly integrated into the genome of the ΔArAceA mutant, transformants with hygromycin B resistance were randomly selected for growth assays.

For the deletion of M433DRAFT_348929 (tzn-1), the 5′- and 3′-flanking regions of tzn-1 (extending from positions +14 to +641 and +920 to +1561 with respect to the tzn-1 translation start point, respectively) were amplified using the primer pairs tzn1-up-F/tzn1-up-R and tzn1-down-F/tzn1-down-R, respectively. The fragment Pgpd-neo was amplified using the primer pair Pgpd-F/neo-T-R. After assembly into the pEASY-Blunt Simple cloning vector, 5′-Pgpd-neo-3′ was amplified using the primer pair tzn1-up-F/tzn1-down-R and used for protoplast transformation of the WT strain.

For the deletion of M433DRAFT_310551, the 5′- and 3′-flanking regions (extending from positions −495 to +161 and +424 to +1083 with respect to the M433DRAFT_310551 translation start point, respectively) were amplified using the primer pairs 310551-up-F/310551-up-R and 310551-down-F/310551-down-R, respectively. After assembly into the pEASY-Blunt Simple cloning vector with Pgpd-neo, 5′-Pgpd-neo-3′ was amplified using the primer pair 310551-up-F/310551-down-R and used for protoplast transformation of the WT strain.

For the deletion of M433DRAFT_300682 (ArCrdA), the 5′- and 3′-flanking regions of ArCrdA (extending from positions −694 to −24 and +387 to +1006 with respect to the ArCrdA translation start point, respectively) were amplified using the primer pairs crdA-up-F/crdA-up-R and crdA-down-F/crdA-down-R, respectively. After assembly into the pCE2 TA/Blunt-Zero cloning vector (Vazyme Biotech) with Pgpd-neo, 5′-Pgpd-neo-3′ was amplified using primer pair crdA-up-F/crdA-down-R and used for protoplast transformation of the WT strain.

For the deletion of M433DRAFT_66060 (ArYgA), the 5′- and 3′-flanking regions of ArYgA (extending from positions +543 to +1182 and +1639 to +2363 with respect to the ArYgA translation start point, respectively) were amplified using the primer pairs YgA-up-F/YgA-up-R and YgA-down-F/YgA-down-R, respectively. After assembly into the pEASY-Blunt Simple cloning vector with Pgpd-neo, 5′-Pgpd-neo-3′ was amplified using the primer pair YgA-up-F/YgA-down-R and used for protoplast transformation of the WT strain.

To select specific single guide RNAs (sgRNAs) targeting ArAceA, tzn-1, ArCrdA, M433DRAFT_310551, and ArYgA, we used the sgRNAcas9 tool (60) to identify sgRNA target sites with high scores. Next, a target-directed M. thermophila U6 promoter-driven sgRNA was created by overlapping PCR and cloned into the pJET1.2/blunt cloning vector, generating the corresponding plasmids U6p-ArAceA-sgRNA, U6p-tzn-1-sgRNA, U6p-ArCrdA-sgRNA, U6p-M433DRAFT_310551-sgRNA, and U6p-ArYgA-sgRNA. The Cas9 expression PCR cassette Ptef1-Cas9-TtprC was amplified from plasmid p0380-bar-Ptef1-Cas9-TtprC (40) using primer pair Ptef-cas-F/TtrpC-cas-R. All constructed plasmids were verified by sequencing.

PEG-mediated transformation of A. richmondensis MEY-1.

Arthroconidia were harvested and spread onto PDAC plates with cellophane. After incubation at 30°C for 30 h, mycelia were harvested and used to prepare the protoplasts. Protoplasts of A. richmondensis MEY-1 were prepared as previously described (55), with some modifications. Briefly, 5 mg/mL of lysing enzymes from Trichoderma harzianum (catalog no. L1412; Sigma-Aldrich, St. Louis, MO) dissolved in solution A (100 mM KH₂PO₄, 1.2 M sorbitol [pH 5.6]) was added to release the protoplasts from the mycelia. Protoplasts were collected by centrifugation at 2,000 rpm for 5 min and then washed twice with solution B (50 mM CaCl₂, 1 M sorbitol, 10 mM Tris-HCl [pH 7.5]). After being collected by centrifugation, protoplasts were resuspended to a final concentration of approximately 2 × 10⁶ protoplasts/mL in solution B. Next, 10 μL of a DNA solution and 50 μL of a PEG solution (25% PEG 6000, 50 mM CaCl₂, 10 mM Tris-HCl) were added to 200 μL of protoplasts, and the samples were mixed gently. After incubation on ice for 20 min, 2 mL of a PEG solution was added, and the samples were mixed gently. The mixture was then incubated at room temperature for 5 min and added to melted selection medium (approximately 45°C) supplemented with the appropriate antibiotic. Finally, the resultant mixtures were poured into petri dishes (9 cm in diameter).

For gene disruption by the CRISPR/Cas9 system, 10 μg of the Cas9 expression PCR cassette Ptef1-Cas9-TtprC from p0380-bar-Ptef1-Cas9-TtprC, the guide RNA expression PCR cassette, and the corresponding donor fragment were mixed at a molar concentration ratio of 1:1:1 and added to the fungal protoplasts. The transformants were grown for 4 to 6 days on selection medium at 30°C, with selection for neo resistance using Geneticin (5 μg/mL) and hph resistance using hygromycin B (4 μg/mL). The presence of the transgene was confirmed by PCR.

Medium shift experiments.

Arthroconidia were inoculated into 100 mL of liquid MMC and grown at 30°C at 200 rpm for 4 days. The mycelium was washed twice with sterilized water and then transferred to 100 mL of MMC, with or without the addition of 15 mM CuCl₂, for 6 h before RNA extraction.

RNA sequencing and transcription expression analyses.

After being harvested via vacuum filtration, mycelia were immediately homogenized in liquid nitrogen for total RNA extraction. RNA was extracted using the TRIzol method (Invitrogen, Waltham, MA, USA) and further purified using a Qiagen (Hilden, Germany) RNeasy minikit. RNA integrity was checked using agarose gel electrophoresis and the Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). Qualified RNA with an optical density at 260 nm (OD₂₆₀)/OD₂₈₀ of >2.0 and an RNA integrity number (RIN) of >8 was used for RNA-Seq, which was performed using the DNBSEQ platform at the Beijing Genomics Institute (BGI) (Shenzhen, China). Prior to read mapping, adaptors and low-quality reads were trimmed using SOAPnuke v1.4.0 (61). Filtered clean reads were aligned against predicted transcripts from the A. richmondensis BFW genome v1.0 (35) using HISAT2 v2.1.0 (62). The read counts were determined using RSEM v1.3.3 (63). The abundance of each transcript was calculated from fragments per kilobase of transcript per million mapped reads (FPKM) values (Table S3). Differential gene expression analysis was performed using the DESeq2 package (v1.5.1). Genes with a fold change of >2.0 (|log₂ ratio| of ≥1) and a Q value (adjusted P value) of <0.05 were considered significantly differentially expressed between conditions or strains. To discover significantly up- and downregulated genes, only genes with a relatively high transcript abundance (FPKM value of >20 in at least one strain) were considered for further analysis. Principal-component analysis (PCA) plots and clustered heat maps were generated using the ClustVis Web tool (https://biit.cs.ut.ee/clustvis).

Growth test.

For comparison of growth at different Cu concentrations among various filamentous fungi, a piece of mycelium from each fungus grown on PDA was inoculated into the middle of MM plates (MMC for A. richmondensis MEY-1). Photos were taken after culture at the optimal growth temperature for 4 or 6 days.

For comparison of growth between different A. richmondensis MEY-1 strains, arthroconidia were collected from the corresponding strains, filtered, and adjusted to 1 × 10⁷ arthroconidia/mL. Next, 1 μL of the suspension was plated onto MMC supplemented with different concentrations of CuCl₂, H₂O₂, or NaCl at 30°C for 4 or 6 days.

Data availability.

RNA-Seq data are available at the Gene Expression Omnibus under accession number GSE217610.