Three metallothionein isoforms and sequestration of intracellular silver in the hyperaccumulator Amanita strobiliformis


Author for correspondence:
Pavel Kotrba
Tel: +420 220445134


  • Metallothioneins (MTs) are cysteine-rich peptides involved in heavy metal tolerance of many eukaryotes. Here, we examined their involvement in intracellular binding of silver (Ag) in the ectomycorrhizal fungus Amanita strobiliformis.
  • The Ag complexes and their peptide ligands were characterized using chromatography and mass spectrometry. The full-length coding sequences obtained from a cDNA library were used for complementation assays in yeast mutant strains. Abundance of respective transcripts in A. strobiliformis was measured by quantitative real-time reverse-transcribed polymerase chain reaction (qRT-PCR).
  • Ag-speciation analyses showed that intracellular Ag was in wild-grown fruit bodies and cultured extraradical mycelia of A. strobiliformis sequestered by metallothioneins. The determined sequence of the peptide facilitated isolation of three cDNA clones, AsMT1a, AsMT1b and AsMT1c. These encode isomorphic MTs consisting of 34 amino acid residues and sharing 82% identity. In mycelia the expression of AsMT1s is induced by Ag. All AsMT1s expressed in yeasts complemented hypersensitivity of mutants to cadmium (Cd) and copper (Cu) and formed Ag complexes. Only the Ag-AsMT1a complex was detected in the A. strobiliformis fruit body in which AsMT1a was the prevailing transcript.
  • The present study identified the existence of metallothionein isoforms in ectomycorrhizal fungi. We demonstrated that intracellular sequestration of Ag in fruit bodies and mycelia of hyperaccumulating A. strobiliformis is dominated by metallothioneins.


The natural capacity of macrofungi to accumulate heavy metals has sparked the interest of biologists for over 50 yr. Of particular interest are mycorrhizal fungi, which intimately link plants and soil and which often ensure healthy tree growth (Smith & Read, 2008). Many studies have shown that metal-resistant ectomycorrhizal (EM) fungi can ameliorate metal toxicity to the associated higher plants (Colpaert & Van Assche, 1993; Godbold et al., 1998; Adriaensen et al., 2006; Bellion et al., 2006; Krznaric et al., 2009; Jourand et al., 2010). By contrast, some EM fungi promote mobilization of metal ions in soils or even their extraction from minerals (Gadd, 2007). This may occur either directly by released organic acids and other exudates or indirectly by the accumulation of metal in a form that will became bioavailable during fruit body decay. The fungal capacities to mobilize, immobilize and/or accumulate metals are thus of interest as a result of their impact on the bioavailability and ecotoxicity of particular metal in soil matrices, with certain potential for bioremediation purposes.

Efforts made towards the understanding of metal tolerance in EM species suggested that these fungi sequester toxic metal species by employing mechanisms described for arbuscular mycorrhiza (AM), saprobic fungi and in other eukaryotes (Bellion et al., 2006; Gadd, 2007). Metallothioneins (MTs) and phytochelatins (PCs) comprise two major classes of metal-binding peptides found in many eukaryotic organisms. PCs are small peptides of general structure (γ-Glu-Cys)nX (PCn; = 2–11; X represents Gly, Ser, β-Ala, Glu, Gln or no residue). These are known for their pivotal role in the detoxification of heavy metals in plants and certain yeasts that possess the capacity to synthesize PCs using glutathione or its homologs as a substrate (Clemens & Simm, 2003; Clemens, 2006). PCs were first described in the Cd-exposed yeast Schizosaccharomyces pombe (Kondo et al., 1984) and PC2, PC3 and minor PC4, but not MTs, were recently reported as responsible for the sequestration of Cd in the EM Boletus edulis (Collin-Hansen et al., 2007). MTs are cysteine-rich peptides of distinct sizes ranging from 25 (fungal) to 84 (plant) amino acids, capable of high-affinity coordination of heavy metal ions via cysteine residues shared along the peptide sequence in Cys-X-Cys or Cys-Cys motifs (Binz & Kägi, 1999). While the role of plant MTs is generally attributed to the homeostasis of essential heavy metals (Clemens, 2006; Freisinger, 2008), in mammals MTs are associated with numerous cellular functions (Coyle et al., 2002). Several lines of evidence suggest that, as in mammals, fungal MTs are involved in the response to oxidative stress (Tamai et al., 1993; Tucker et al., 2004; González-Guerrero et al., 2007) and metal toxicity and as well as in the regulation of metal homeostasis. The latter is reflected by differential transcription of their genes during mycorrhiza development. Symbiosis-regulated MT genes were identified in the EM basidiomycete Pisolithus tinctorius (Voiblet et al., 2001) and the AM fungus Gigaspora margarita (Lanfranco et al., 2002).

Unlike mammalian 61- or 62-amino acid (AA) MTs, showing high sequence conservancy, fungal MTs produced in a response to metal stress encompass quite a diverse group of peptides. In the yeast Saccharomyces cerevisiae, a 53-AA CUP1 plays a primary role in detoxification of excess Cu (Butt & Ecker, 1987) and a 69-AA CRS5 is involved in homeostasis of Cu and Zn (Pagani et al., 2007). Only upon Cu exposure, but not in response to oxidative stress, does the ascomycete Neurospora crassa synthesize its 26-AA Cu MT (Kumar et al., 2005). Both Cu and oxidative stress induced transcription of the GintMT1 gene encoding 71-AA MT in the AM fungus Glomus intraradices (González-Guerrero et al., 2007). Small copper-binding cysteine-rich peptides, presumably MTs, were isolated from the EM fungi Laccaria laccata and Paxillus involutus (Howe et al., 1997). A gene encoding 34-AA metallothionein PiMT1, involved in the sequestration of Cd in P. involutus, was isolated and shown to confer higher Cu tolerance to the transgenic EM fungus Hebeloma cylindrosporum (Courbot et al., 2004; Bellion et al., 2007). Ramesh et al. (2009) recently showed that Cu in H. cylindrosporum induces expression of indigenous HcMT1 and HcMT2 encoding 59-AA and 57-AA MTs, which share only 40% identity. Transcription of HcMT2, but not of HcMT1, is also induced by Cd. While the induction of both MT and PC2 by Cd was reported in the aquatic hyphomycete Heliscus lugdunensis (Jaeckel et al., 2005), yeast Candida glabrata produces specifically nonhomologous 63-AA MT1 and 52-AA MT2 in the response to Cu. It also produces PCs in a response to the Cd stress (Mehra et al., 1988).

According to Brooks (1998), plants are considered as hyperaccumulating if they deposit in their organs at least 100-fold higher concentrations of a particular element than other (related) species growing over an underlying substrate with the same characteristics. The elemental hyperaccumulation in vascular plants may have several functions, including plant defense against herbivores and pathogens; at least some tests have demonstrated defense by hyperaccumulated As, Cd, Ni, Se and Zn in plants (Boyd, 2004, 2007). Despite the pronounced ability to accumulate various metals species, macrofungi generally do not exhibit hyperaccumulating phenotypes. However, we recently reported that the EM fungus Amanita strobiliformis has the capacity to hyperaccumulate biocidal Ag from the pristine environment (Borovička et al., 2007). It prompted our efforts to determine the molecules involved in intracellular binding of the hyperaccumulated metal. We isolated the intracellular Ag complexes of fruit bodies from their natural habitat and characterized the MT component of the complex as the 34-AA AsMT1a. We also cloned two other cDNAs, AsMT1b and AsMT1c, whose predicted products share high identity with AsMT1a. We further showed that expression of AsMT1s was induced by Ag in extraradical mycelia and that all three full-length cDNAs were able to complement Cu- and Cd-sensitive phenotypes and form Ag complexes in mutant S. cerevisiae.

Materials and Methods

Organisms, culture media and conditions

The young fruit bodies of A. strobiliformis (Paulet ex Vittad.) Bertill. containing 500–550 mg Ag kg−1 DW were collected from their natural habitat under Tilia trees at sedimentary bedrock (280 μg Ag kg−1 soil DW) in Prague-Klíčov, Czech Republic. The fruit bodies were cleared of substrate debris, washed with sterile distilled water and stored at −80°C. The fruit bodies to be used in RNA work were fixed by freeze-drying and stored at −80°C. The mycelial culture of A. strobiliformis was obtained from axenic explants (30 mm3) from the fresh breaking surface of the pileus, as described previously with Entoloma clypeatum (Gryndler et al., 2010). The A. strobiliformis collection has been deposited in the herbarium of the Mycological Department, National Museum, Prague (PRM 857486). To evaluate the effect of Ag, the mycelia colonies were grown at 25°C for 21 d on potato dextrose agar (Sigma-Aldrich, St. Louis, MO, USA) and transferred to liquid Melin–Norkrans (MNM) medium (Marx, 1969) with 1% (w/v) glucose as carbon source. After 16 wk of cultivation at 25°C (pregrowth period) mycelia were grown for an additional 14 d in the same media without or with AgNO3 added to a final concentration of 18.7 μM. The mycelium upon harvest was washed with 0.1% MgSO4 and stored at −80°C or freeze-dried.

DNA manipulations in E. coli DH5α were performed according to standard protocols (Sambrook & Russell, 2001). The S. cerevisiae strains used for complementation assays were DTY113 (MATα trp1-1 leu2-3,-112 gal1 ura3-50 cup1Δ61) (Tamai et al., 1993) and DTY168 (MATα his6 leu2-3,-112 ura3-52 ycf1::hisG) (Szczypka et al., 1994). The lithium acetate method (Gietz et al., 1995) was used to transform the host S. cerevisiae. Transformants were grown at 30°C on URA+-selective SD medium (0.7% (w/v) Difco yeast nitrogen base, 2% (w/v) glucose, 0.005% adenine hemisulfate, and 0.003% (w/v) of each of l-histidine, l-tryptophan and l-leucine). The mid-log cultures of S. cerevisiae transformants were adjusted to an optical density at 590 nm (OD590) of 0.02, and 3 μl of serial dilutions were spotted on SD medium plates without metal or supplemented with 100 μM CuSO4, 100 μM CdCl2 or 1–40 μM AgNO3. The growth of S. cerevisiae transformants in the Ag-containing liquid medium was initiated by its inoculation with mid-log cultures to the OD590 of 0.1. The optical densities of cultures were followed at 3–6 h intervals. The transformants used for isolation of Ag–AsMT1 complexes were grown in a medium containing 20 μM Ag for 44 h (final OD590 of 1.0–1.3).

Isolation of Ag complexes

The cell-free extracts were obtained from 80 g of A. strobiliformis (stipes and caps in natural proportion) or 2.3 g of the Ag-exposed mycelium. Tissue was ground in liquid N2 using a mortar and pestle and disintegrated tissue was extracted with 60 ml of 50 mM Hepes (pH 7.0). The cell debris was removed by centrifugation (20 000 g, 10 min, 4°C). Proteins of the fruit body extract were further precipitated with 80% (NH4)2SO4 and separated by centrifugation. The resulting supernatant was desalted by dialysis with a Spectra/Por 7 (Spectrum Labs., Rancho Domingues, CA, USA) tubing of molecular weight cutoff (MWCO) of 1000 Da against 50 mM Hepes (pH 7.0), and brought to a final volume of 600 μl by ultrafiltration with a Centricon YM-3 membrane (Millipore, Bedford, MA, USA). The size exclusion chromatography (SEC) was performed with a Superdex Peptide GL column (1 × 30 cm; GE Healthcare, Uppsala, Sweden), a BioLogic DuoFlow fast protein liquid chromatography (FPLC) system (Bio-Rad, Herkules, CA, USA) and 50 mM Hepes, 100 mM KNO3 (pH 7.0) as a mobile phase at flow rate of 0.5 ml min−1. Ribonuclease A (GE Healthcare), ubiquitin (Sigma-Aldrich), synthetic 2.1 kDa peptide and glutathione (Merck, Darmstadt, Germany) were used as molecular weight (MW) standards. Ag contents in aliquots of 0.5 ml fractions from SEC and in fractions obtained in each step of isolation procedure were determined by atomic absorption spectrometry (AAS; model Spectr AA300, Varian, Inc., Palo Alto, CA, USA). The tissue debris was extracted with 65% nitric acid for 16 h at room temperature and the Ag content of supernatant resulting from the 10 min centrifugation at 20 000 g was analyzed by AAS.

The cell-free extracts of Ag-exposed S. cerevisiae were prepared from 1 g wet weight of the cells harvested by centrifugation for 5 min at 4000 g  and 25°C. Separated cells were washed with 20 ml of fresh SD medium and resuspended at a density of 0.25 g (wet weight) ml–1 50 mM Hepes (pH 7.0). Subsequently, 0.5 g of glass beads (0.5 mm diameter) was added for cell disruption in a Mini-Beadbeater device (BioSpec Products, Inc., Bartlesville, OK, USA). This was conducted in three cycles of 1 min disintegration at the maximum speed and 2 min at 0°C. The glass beads and cell debris were separated from the cell extracts by centrifugation for 5 min at 20 000 g and 4°C. The supernatant was brought to a final volume of 500 μl by ultrafiltration with Microcon YM-3 (Millipore) and resolved by SEC as already described.

Reverse-phase chromatography and electrophoresis of the Ag complex ligands

The sulfhydryl-containing molecules from SEC fractions were fluorescent-labeled in reaction with a 7-fluorobenzofurazan-4-sulfonic acid (SBD-F) and resolved by reverse-phase high-performance liquid chromatography (RP-HPLC) or sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described previously (Borovička et al., 2010). Glutathione (Merck), synthetic PC2 and PC3 (Vidia Ltd, Vestec, Czech Rep.), rabbit liver MT1a (Enzo Life Sci., Inc., Plymouth Meeting, PA, USA), and Broad Range protein marker (Bio-Rad) were used as standards. The pooled SEC fractions 28–36 from mycelia (Fig. 6a) were 30-fold concentrated by ultrafiltration with Centricon YM-3 before labeling. The RP-HPLC analyses were performed with an Agilent 1200 HPLC with PEEK capillaries, an Agilent G1321A spectrofluorimetric detector (Agilent Technologies, Inc., Santa Clara, CA, USA) and an analytical 250 mm column (4 mm ID) packed with Separon SGX C8 (5 μm; Tessek Ltd, Prague, Czech Rep.). The acetonitrile proportion in H2O (both with 0.1% (v/v) trifluoroacetic acid) during elution was 5–25% (v/v) linear gradient from 0 to 20 min, 25–70% from 20 to 25 min, 70% from 25 to 33 min and 70–5% from 33 to 35 min. The separation by SDS-PAGE was conducted in 16% polyacrylamide gels in a Tris-Tricine buffer system with 6 M urea.

Separation and sequencing of ligands

The eligible fractions from SEC were pooled and concentrated by ultrafiltration with a Microcon YM-3 membrane (Millipore) to a volume of 500 μl, and tris(carboxyethyl)phosphine (TCEP) was added at 140 mM for 1 h reduction reaction at room temperature. The ligands were released from the complex by acidification with trifluoroacetic acid (TFA) added to a final concentration of 1% and resolved on a Superdex Peptide GL column (1 × 30 cm) with 1% TFA in water as a mobile phase at 0.5 ml min−1. Obtained 0.5 ml fractions were freeze-dried, lyophilized material was resolved in 80 μl of 50 mM Hepes (pH 7.0) and aliquots were screened for their Ag content by AAS.

The peptide fractions from FPLC were reduced by 50 mM TCEP and alkylated by 50 mM iodoacetamide (m/z values after subtraction of carboxamidomethyl group mass are reported). The mixture was aspirated to a ZipTip column (Millipore) equilibrated by subsequent washes with acetonitrile and 0.1% TFA in LC-MS grade water. After a wash of unbound reagents with 0.1% TFA, the peptides were released by solution of alpha-cyano-4-hydroxycinnamic acid (2 g l−1 in 80% acetonitrile). Mass spectra were recorded on a 4800 Plus MALDI TOF/TOF mass analyzer (AB Sciex, Foster City, CA, USA) equipped with a Nd:YAG laser (355 nm, firing rate 200 Hz). MS/MS was performed with 1 kV collision energy and the operating pressure of the collision cell set to 10−6 Torr. Further confirmation of the assigned sequence was achieved by comparison of MS/MS spectra of peptides subjected to methyl- and ethyl-esterification of glutamic and aspartic acid side chains and the C-terminus of peptides (Volf et al., 2002) or peptides with free sulfhydryl groups of cysteines (reduced with 100 mM TCEP solution). The sequence obtained by manual de novo analysis was named P1613.

Preparation of AsMT1 cDNA probes, cDNA library construction and screening

Total RNA was extracted from 50 mg of dried fungal tissues ground with mortar and pestle using an RNeasy Plant Mini Kit and RNase-Free DNase Set (Qiagen, Hilden, Germany) according to the manufacturer’s instructions, except that the better-performing buffer RLT was used. The first cDNA strand was reverse-transcribed from 2.5 μg total RNA using an Enhanced Avian HS RT-PCR kit (Sigma-Aldrich) with the anchored oligo-dT23 primer. The reaction for probe synthesis was prepared using the same RT-PCR kit with 4 μl of cDNA preparation as a template, degenerated primer KN1613 (5′-AAYGARGGIGIWSITGYAARTGYGG-3′ corresponding to the sequence NEGGSKC of P1613), anchored oligo-dT23 and AccuTaq DNA as recommended by the manufacturer. Resulting amplicons were cloned in a pGEM-T® Easy vector (Promega, Madison, WI, USA). DNA sequences were determined using T7 and SP6 primers, a CEQ 8000 DNA sequencer and a CEQ DTCS Kit (Beckman Coulter, Fullerton, CA, USA). The probes were generated by PCR amplification of clones harboring cDNAs corresponding to P1613 and its isoforms with the primer pair T7 and SP6 and a digoxigenin-dUTP (DIG) Labeling Mix (Roche Diagnostics, Mannheim, Germany).

An Oligotex mRNA Mini Kit (Qiagen) was used according to the manufacturer’s protocol to obtain 5 μg of A. strobiliformis mRNA from approx. 150 μg of total RNA. The λZAP-based cDNA library was constructed and amplified using a ZAP-cDNA® Gigapack® III Gold Cloning Kit (Stratagene, La Jolla, CA, USA) Following the manufacturer’s instructions. A total of 10 000 plaques of the λZAP cDNA library were transferred on to nylon membrane discs (Roche Diagnostics) and screened for AsMT1 cDNAs with the labeled probes following the user’s guide for plaque hybridization. The DIG-labeled probe was detected according to the protocol supplied with a DIG DNA Labeling and Detection Kit (Roche Diagnostics). The phages of positive plaques were amplified, and pBluescript-based phagemids were excised from λ DNA using an ExAssist A® helper phage and E. coli SOLR™ of the ZAP-cDNA® Synthesis Kit as recommended in the Stratagene instruction manual. The nucleotide sequences were determined on both strands with T3 and T7 primers as described earlier. Nucleotide sequences of AsMT1a, AsMT1b and AsMT1c were deposited in GenBank under accession numbers HQ439183, HQ439184 and HQ439185, respectively.

Vectors for functional expression of AsMT1 cDNAs in S. cerevisiae

The coding sequences plus 37 bp of the 3′-untranslated region of AsMT1 genes were amplified from their individual cDNAs using the bAmtI-F (5′-GCGGGATCCCAAACATCATGCACTCGAACGT-3′; start codon underlined) and the hAmt-R (5′-ACCGAAGCTTGGTIATAGCGCAGGCAC-3′) primers, universal to all three isoforms, and Accu Taq DNA polymerase (Sigma). The resulting fragments were flanked with BamHI and HindIII target sequences (italicized on primer sequences), which were used to insert the amplicons into the BamHI/HindIII-treated plasmid p426GPD (Mumberg et al., 1995). The resulting plasmids harboring coding sequences of AsMT1a, AsMT1b and AsMT1c were pGPD-AsMT1a, -AsMT1b and -AsMT1c, respectively. Transcription was controlled by the constitutive glyceraldehyde-3-phosphate dehydrogenase (GPD) promoter and the cytochrome-c oxidase gene CYC1 terminator of S. cerevisiae.

Quantitative real-time reverse-transcribed PCR (qRT-PCR)

The extraction of the total RNA and the first cDNA strand synthesis were performed as already described. The absolute quantification by qRT-PCR was done using a DyNAmo Flash SYBR Green qPCR Kit (Finnzymes, Espoo, Finland) in a 12 μl reaction volume with 0.35 μM gene-specific primers and 1.5 μl of diluted cDNA, produced from 90 ng of the total RNA, as a template. The primer sequences were 5′-CCAACGCCACTTGCTCCTGCCT-3′ and 5′-TGGTGATAGCGCAGGCACAGCAGTC-3′ for AsMT1a (114 bp amplicon); 5′-CAAACATCATGCACTCGAACGT TA-3′ and 5′-TGAACGACGCGTTTAGTGGGTGT-3′ for AsMT1b (125 bp amplicon); 5′-CAAACATCATGCACTCGAACGTTA-3′ and 5′-AACGACGCGTTTAGTTGGTGGCA-3′ for AsMT1c (123 bp amplicon); and 5′-CAAACATCATGCACTCGAACGT-3′ and 5′-GGTIATAGCGCAGGCAC-3′ for all three AsMT1s (151 bp amplicon). The reactions with cDNA from each of three independent extractions were performed in a MiniOpticon Real Time PCR System (Bio-Rad) in triplicate. Serial dilutions of plasmids pGPD-AsMT1a, -AsMT1b and -AsMT1c were used as the calibrator templates (Peirson et al., 2003).


Silver deposition form in A. strobiliformis fruit bodies

To inspect the Ag-containing cellular components, the disintegrated tissue of A. strobiliformis was extracted under the mild conditions at pH 7.0. Virtually all Ag was found in the fruit body cell-free extract. The treatment with excess 65% nitric acid revealed separated cell debris essentially free from H+-exchangeable Ag+. The analysis of the extract showed that all Ag was contained in a complex, which was retained by the YM-3 membrane of MWCO 3000 Da (data not shown). Further fractionation of the extract by ammonium sulfate (80% saturation) resulted in the precipitation of the major portion of cellular proteins, rendering 86% of the extracted Ag soluble. The SEC of desalted and concentrated soluble fraction revealed virtually all extracted Ag present in two peaks (Fig. 1). The majority of Ag (85%) was found to be associated with a MW fraction of 6.2 kDa and 15% of the metal was contained in a fraction of MW 3.3 kDa.

Figure 1.

Size exclusion chromatography (SEC) fractionation of the extract from the fruit body of Amanita strobiliformis. The extract was prepared at a neutral pH and the majority of cellular proteins were removed by ammonium sulfate precipitation. The elution maxima of the molecular weight (MW) standards are indicated by arrows.

Based on the thiophilic nature of Ag+ and the size of the Ag complexes, cysteine-rich peptides were considered likely ligands. We thus employed the sulfhydryl-specific SBD-F probe to inspect the presence of thiol-containing molecules in the Ag-containing fractions. Separation of labeled thiol compounds by the reverse-phase chromatography in a gradient mode showed a late triple peak cluster eluting from the C8 column with 24–25% acetonitrile (Fig. 2a). We did not detect glutathione, PC2 or PC3 as a complex component in any of the analyzed fractions. Considering the protein nature of the ligands, the labeled thiol compounds of the SEC fractions were also resolved using SDS-PAGE in the Tris-Tricine buffer system with 6 M urea under reducing conditions. As shown in Fig. 2(b), the fractions of the 6.2 and 3.3 kDa peaks from SEC contained the labeled peptides of MW << 6.1 kDa of the rabbit MT1a.

Figure 2.

Detection of thiol-containing peptides in silver (Ag) complexes labeled with fluorochrome 7-fluorobenzofurazan-4-sulfonic acid (SBD-F). (a) Reverse-phase high-performance liquid chromatography (HPLC) of extract subjected to size exclusion chromatography (SEC) fractionation. The inset chromatograms show the analysis of SEC fractions 28 and 30. The retention times of glutathione (GSH) and phytochelatins PC2 and PC3 are indicated by arrows. Asterisks identify peaks also present in the control blank reaction. (b) Electrophoretic analysis of fractions obtained from SEC of the extract. The protein molecular weight (MW) standard (M) is shown to the left. (c) Electrophoretic analysis of ligands released from Ag complexes by acid treatment and separated from the metal by SEC. MT1a denotes the SBD-labeled 6.1 kDa rabbit metallothionen 1a.

Metallothionein involved in the intracellular sequestration of fruit body Ag

We chose a hydrogen ion/Ag+ competition for metal-binding site(s) to obtain the Ag ligand molecules. The SEC fractions containing Ag were acidified to pH 0.95 with TFA, peptides were again separated by SEC (not shown), derivatized with SBD-F and resolved using SDS-PAGE (Fig. 2c). While the separated ligands showed the same electrophoretic pattern as observed with the native Ag complexes, the corresponding SEC fractions did not contain Ag, indicating that TFA treatment liberated ligands in their apo-form.

The mass spectra obtained with the fraction 31 showed the presence of MW peaks of 1000–3500 Da with a well separated 1613.6 Da peak (Fig. 3a). The amino acid sequence of this peptide (designated P1613) was determined by mass spectrometry as NEGGSC(Q/K)CGDSCGCGTH. The presence of C-X-C motifs, which is characteristic of MTs (Binz & Kägi, 1999), and substantial identity with cysteine-rich domains of plant and fungal MTs (Fig. 4a) strongly suggested that the P1316 fragment originated from MT.

Figure 3.

Abundance of AsMT1s and their corresponding mRNAs in Amanita strobiliformis. (a) Mass spectrum of peptide fragments isolated from Ag complexes (a part of the spectrum obtained with fraction 31 shown in Fig. 2c). The amino acid sequence of the peptide of m/z 1613.6 was determined by MALDI TOF/TOF. Also presented are AsMT1a fragments identified based on the transcript sequence (the C-terminal sequences are shown). (b) Abundance of AsMT1a, AsMT1b and AsMT1c transcripts in the fruit body tissue. AsMT1s, experimentally determined level of all AsMT1 mRNAs; AsMT1s*, calculated summation of data for individual AsMT1s. The data are means ± SE (= 3).

Figure 4.

Alignment of P1613 and AsMT1s with metallothioneins (MTs) from other organisms. (a) Similarity of P1316 to the best-scoring fragments of plant and fungal MTs identified by BLAST analysis. Their amino acid residues identical to those of P1316 are boxed. (b) Multiple alignment of fungal MT sequences. Only residues different from AsMT1a are shown for AsMT1b and AsMT1c. Fully and partially conserved cysteine residues are boxed with black and white backgrounds, respectively. The GenBank accession numbers are as follows: Arabidopsis thaliana MT1a, NM_100633; Brassica oleracea MT1, AF458412; Oryza sativa MT4c, NM_001073615; Agaricus bisporus, CAC85298; Hebeloma cylindrosporum HcMT1, EU049884; Gigaspora margarita GmarMT1, AJ4215217; Podospora anserina, CAA06385; Neurospora crassa, CAA26793; Colletotrichum gloeosporioides, AAA74033; Lentinula edodes, CO501612; Antrodia cinnamonea, DR031991; Ganoderma lucidum, ABP02008; Paxillus involutus, AAS19463; AsMT1a, HQ439183; AsMT1b, HQ439184; AsMT1c, HQ439185.

In order to amplify a portion of transcript corresponding to P1613, the degenerated primer KN1613 was designed as described in the Materials and Methods section. The RT-PCR with additional oligo-dT23 primer and the total RNA of A. strobiliformis as a template produced a cDNA fragment harboring the P1613 coding sequence. It was followed by a termination codon and 117 nucleotides of a downstream untranslated region (UTR). Two additional cDNA amplicons containing sequences coding the predicted P1613 homologs were also identified. Nucleotide sequencing showed that the specific residues of P1613 are substituted with Asp-15 or Ala-15 and Asn-17 (see Fig. 4b for details). Notable differences were apparent within downstream UTRs of amplified sequences (see GenBank accession numbers HQ439183, HQ439184 and HQ439185). Each amplicon was labeled and used as a probe to screen the A. strobiliformisλZAP cDNA library by plaque hybridization. It allowed us to isolate three individual cDNAs, designated AsMT1a (with P1613 coding sequence), AsMT1b and AsMT1c. The AsMT1s encoded 34-AA MTs, which share > 82% identity (Fig. 4b). All AsMTs contained three C-X-C motifs organized within a single domain, which is characteristic for some other fungal MTs. In addition, the CC motif was found in AsMT1b.

The analysis of mass spectra of purified peptide ligands revealed that the isolated Ag complex contained fragments of different lengths, which all originated from AsMT1a (Fig. 3a shows fragments of fraction 31 from SEC of the TFA-treated Ag complex). The AsMT1b- or AsMT1c-specific peptide fragments were not detected. The statement that the AsMT1a fragments exist in vivo would be speculative as the isolation of the apo-MT involved strong acid treatment, which may result in partial hydrolysis of AsMT1a. However, the appearance of a triple peak on the reverse-phase chromatogram (Fig. 2b) suggests that the native complex contains three functional AsMT1a fragments.

In order to compare the levels of transcription of genes encoding AsMT1a, AsMT1b, and AsMT1c in analyzed fruit bodies, we performed qRT-PCR analysis. The results showed that AsMT1a mRNA represented, in average, 75% of all AsMT1 transcripts (Fig. 3b), which supports the concept of AsMT1a being the major ligand involved in intracellular binding of Ag in frut bodies of A. strobiliformis.

Expression of AsMT1a, b and c in S. cerevisiae

In an attempt to document the function of AsMT1 isoforms as heavy metal-binding ligands, we expressed the individual coding sequences in two S. cerevisiae metal-sensitive mutant strains, and inspected their growth on media with and without metal supplements. The p426GPD-based pGPD-AsMT1a, -AsMT1b and -AsMT1c vectors were constructed, allowing constitutive expression of AsMT1a, AsMT1b, and AsMT1c, respectively. The detoxification of Cd in S. cerevisiae relies largely upon a ycf1-encoded ABC-type transporter involved in vacuolar sequestration of the bis(glutathionato)cadmium complex (Li et al., 1997). The disruption of ycf1 causes a Cd-sensitive phenotype of the DTY168 mutant strain, which is unable to grow in the presence of 100 μM Cd2+ (Szczypka et al., 1994; see also p426GPD-transformed DTY168 in Fig. 5a). The Cu-tolerance of S. cerevisiae depends on the sequestration of toxic concentrations of Cu by CUP1; which is a 53-AA MT induced by an excess of Cu (Butt & Ecker, 1987). The inactivation of the single-copy cup1 gene by the cup1Δ61 mutation in DTY113 thus prevents the growth of the strain in the presence of 100 μM Cu (Tamai et al., 1993; and p426GPD-transformed DTY168 in Fig. 5a). Fig. 5(a) shows that all three AsMT1 cDNAs complemented the ycf1-null and Δcup1 genotypes and protected the mutant hosts from the Cd and Cu toxicity, respectively.

Figure 5.

Functional expression of AsMT1s in Saccharomyces cerevisiae mutants. Indicated mutant strains were transformed with the empty p426GPD vector or with the same expression vector inserted with coding sequences of indicated AsMT1s. (a) Functional complementation of yeast mutants on selective media. Diluted yeast cultures were spotted on SD medium (0.7% (w/v) Difco yeast nitrogen base, 2% (w/v) glucose, 0.005% adenine hemisulfate, and 0.003% (w/v) of each of l-histidine, l-tryptophan and l-leucine) with or without metal supplement. (b) Sequestration of Ag in cup1Δ61 mutant. Size exclusion chromatography (SEC) fractionation of extract was under the same conditions as those used for fractionation of the extract from Amanita strobiliformis (Fig. 1). The cell-free extracts were prepared from cells exposed to 20 μM Ag in liquid SD medium.

Although the transcription of cup1 is also induced by Ag (Butt & Ecker, 1987), we did not observe any difference in Ag-tolerance with DTY113 transformed with an empty vector and its variants containing AsMT1a, b or c. Both the control and the cells transformed with AsMT1s showed a minimal inhibition concentration of 35 μM Ag on agar plates and 30 μM Ag in liquid media. In order to gain an insight into the function of heterologous AsMT1 peptides in Ag binding, we cultivated the DTY113 transformants in liquid media supplemented with a sublethal concentration of Ag (20 μM). The harvested cells were disintegrated and their cytoplasmic fraction was extracted under essentially the same conditions as used for the isolation of the Ag complex from A. strobiliformis. The SEC of the cell-free extracts from the pGPD-AsMT1 transformants, but not from the cells harboring p426GPD, revealed the presence of Ag complexes with MWs of approx. 6 and 3 kDa (Fig. 5b), resembling those isolated from A. strobiliformis (Fig. 1). Electrophoretic separation of corresponding fractions labeled with SBD-F further identified the peptides of estimated size 3 kDa as the complex ligands (data not shown). These data signify that all AsMT1s have the capacity to sequester Ag in S. cerevisiae.

Sequestration of Ag and induction of AsMT1 in A. strobiliformis mycelia

In order to obtain a further insight at the molecular level into the response to Ag of A. strobiliformis, we inspected the capacity of extraradical mycelia to produce AsMT1s by using SEC and qRT-PCR analyses. The mycelia subjected to a chronic 14 d exposure to 18.7 μM Ag in liquid MNM accumulated 224 μg g−1 of Ag. Nearly 64% of accumulated Ag was liberated to the extract. Our approach does not allow clear discrimination between the intracellular accumulation and a metabolism-independent uptake by biosorption at the cell wall. Nearly 68% of the extracted Ag was present in a fraction of 5.2 kDa (Fig. 6a). The remaining portion of the metal was associated with the protein fraction of MW > 20 kDa (column exclusion limit). This fraction, however, may have resulted from nonspecific binding of the metal released during extraction from the biosorption sites at the mycelial cell wall. When labeled with SBD-F, the ligands of 5.2 kDa showed the same electrophoretic mobility as the fruit body AsMT1 (Fig. 6b), thereby suggesting that AsMT1s are also Ag-sequestering ligands in the mycelia. Analysis of the gene expression by qRT-PCR revealed that Ag-exposed mycelium accumulates 17-fold more AsMT1 mRNA than the unexposed mycelium (Fig. 6c).

Figure 6.

Presence of AsMT1 in extraradical mycelium of Amanita strobiliformis. Mycelia grown in Melin–Norkrans (MNM) medium were subjected to 14 d exposure to 18.7 μM Ag (2 mg Ag l−1). (a) Size exclusion chromatography (SEC) fractionation of the cell-free extract. The elution maxima of the molecular weight (MW) standards are indicated by arrows. (b) Electrophoretic analysis of 7-fluorobenzofurazan-4-sulfonic acid (SBD)-labeled peptide contained in pooled SEC fractions 24–36 (AsMTmycel.). In corresponding SEC fractions from unexposed mycelia, SBD-labeled molecules were not detected (not shown). Rabbit metallothionen 1a (6.1 kDa MT1a) and the AsMT1 peptide of fraction 30 shown in Fig. 2(b) (AsMT1F30) were used as standards. (c) Abundance of AsMT1 transcripts in Ag-exposed mycelia (Ag+) and in mycelia cultured for the same period without Ag supplement (Ag−). The data are means ± SE (= 3).


Efficient metal uptake from soil, its translocation to above-ground tissues and competence to detoxify accumulated metal species are key cellular mechanisms underlying the (hyper)accumulating phenotypes of plants and fungi (Bellion et al., 2006; Krämer, 2010). It has been well documented that the concentrations of Ag in macrofungi from pristine habitats are significantly higher than those in underlying soils, which qualifies the macrofungi as effective accumulators of Ag (Borovička et al., 2010). However, a bibliographic search revealed that the hyperaccumulation thresholds of 100 and 300 mg kg−1 for EM and saprobic fungi, respectively, are only rarely exceeded. There are only two EM Amanita species, A. strobiliformis and A. solitaria, which can be regarded as Ag hyperaccumulators with the common Ag concentration in their fruit bodies ranging from 200 to 1200 mg kg−1 (Borovička et al., 2007). We used wild-grown fruit bodies containing 550 mg kg−1 of Ag and the Ag-exposed mycelial culture to show that sequestration of intracellular Ag in A. strobiliformis involves MT(s). Virtually all fruit body Ag was present in Ag complexes in which the presence of AsMT1a-encoded MT was confirmed at the protein level. We also describe two additional AsMT1b and AsMT1c transcripts of A. strobiliformis fruit body, whose predicted products are 85 and 82% identical to AsMT1a, respectively. Such a high degree of identity is characteristic of mammalian and plant MT isoforms, which may be distributed in various ratios in individual tissues (Coyle et al., 2002; Freisinger, 2008). This is the first report of the isolation of MT coding sequences from wild-grown and hyperaccumulating fungus and the first evidence of the presence of MT isoforms in EM fungi.

The sequence features of A. strobiliformis AsMT1s can be regarded as region-specific traits of known MTs from ascomycetes or basidiomycetes. The distribution of C-X-C motifs indicates a closer relationship of AsMT1s with MTs from the filamentous ascomycetes Podospora anserina, N. crassa and Colletotrichum gloeosporioides, compared with known MTs of EM fungi. On the other hand, the first cysteine of AsMT1s is preceded by a 13-AA domain lacking cysteine residues, which appears characteristic of MTs from EM basidiomycetes. It should also be noted that a single cysteine residue flanking C-X-C motifs of MTs from both phyla is absent in the AsMT1 sequences. Taken together, these results reveal a certain difficulty in assigning the sequence signature to fungal MTs. Ramesh et al. (2009) recently noted that the consensus sequence C-X(3,4,5)-C-X-C-X(3)-C-X-C attributed to the MTs from multicellular fungi (Binz & Kägi, 1999) is conserved only in MTs from filamentous ascomycetes. The authors also suggested that the short C-X-C-X(2,3)-C motif and the cysteine residue at the C-terminal end represent the sequence signature of fungal MTs. As shown here, however, the AsMT1 isoforms do not conform to the latter criterion of the C-terminal cysteine.

To demonstrate the roles of AsMT1s in metal detoxification, yeast complementation assays in yeasts were performed for this study. Expression of the AsMT1 genes provided tolerance to Cu- and Cd-sensitive yeasts, thereby confirming that the AsMT1 peptides may confer defense against heavy metal ions. Any yeast mutant specifically exhibiting the Ag-sensitive phenotype is not yet available. We thus resorted to the analysis of cell-free extracts from the Ag-exposed yeasts and showed that the presence of AsMT1 gene products results in the formation of Ag complexes resembling those identified in A. strobiliformis. However, the apparent sequestration of Ag by AsMT1s failed to improve the tolerance of the Δcup1 S. cerevisiae. Ecker et al. (1986) also noted that overproduction of CUP1, which has the capacity to bind Ag in vitro (Butt & Ecker, 1987), did not protect the Δcup1 from the Ag toxicity. It suggests that the natural Ag tolerance threshold in S. cerevisiae is independent of MTs. Although the increased accumulation of the AsMT1 transcripts and the formation of the corresponding Ag complex in the Ag-exposed mycelia of A. strobiliformis corroborate the idea that AsMT1s provide protection against Ag, additional mechanisms could be simultaneously involved in detoxification of the metal. It appears reasonable to assume that Ag will promote the production of reactive oxygen species in vivo and thus compromise the redox homeostasis (Schützendübel & Polle, 2002; Cortese-Krott et al., 2009). It has been well documented that besides the production of PiMT1, P. involutus responds to Cu and Cd by significantly increasing the intracellular activities of antioxidative defense-related enzymes (Bellion et al., 2007).

The information on the role of MTs in sequestration of Ag in fungi is quite limited. Both MT genes in C. glabrata and CUP1 of S. cerevisiae are induced by Ag, but there is no direct evidence on their function in detoxification of Ag (Ecker et al., 1986; Mehra et al., 1988). The possible involvement of MTs in sequestering Ag in 8–10 kDa complexes of saprobic Agaricus bisporus was suggested by Byrne & Tušek-Žnidarič (1990). Recently, we showed that the thiol-rich 3 kDa peptides, resembling those of A. strobiliformis, form 7 and 3.3 kDa Ag complexes in fruit bodies of Amanita submembranacea collected from the Ag-polluted site (Borovička et al., 2010). It thus appears that the capacity to produce Ag-sequestering MTs in Amanita spp. is not restricted to hyperaccumulating species. Our data suggest that all AsMT1 isoforms are equally well suited to sequester Ag, but the marked prevalence of the AsMT1a transcript and the failure to detect AsMT1b and AsMT1c in the Ag complex signify that sequestration of Ag in fruit bodies of A. strobiliformis is dominated by AsMT1a.

Altogether, the results presented here show that A. strobiliformis encodes MT isoforms and employs MTs to sequester intracellular Ag in its fruit bodies and mycelium. Further effort is needed to delineate the extent to which the expression of AsMT1s supports detoxification of Ag and to explain the mechanism used in translocation of Ag from mycelia to the fungus fruit body.


We thank Prof. Dennis J. Thiele (Duke University Medical Center, NC, USA) for the gift of strains DTY113 and DTY168. This work was funded by research project IAA600480801 from the Grant Agency of the Academy of Sciences of the Czech Rep. and 1M6837805002 from the Czech Ministry of Education, Youth and Sports. The authors also acknowledge support from individual Institutional Research Plans MSM6046137305, MSM0021620858, AV0Z10480505 and AV0Z50200510.