Introduction

Enargite (Cu3AsS4) is an important industrial copper ore that is common in porphyry ores and deep epithermal “high sulphidation” deposits and is associated with other major copper sulfides such as chalcopyrite (CuFeS2) and chalcocite (Cu2S) [16]. Recent studies showed that high recovery yields of copper can be rapidly obtained from enargite using hydrometallurgical processes, such as alkaline sodium sulfide leaching and pressure leaching [24, 29]. However, because of the large amount of sodium sulfide employed and special equipment requirements, these hydrometallurgical processes are not relevant for industrial application. In addition, there is currently no economical method to separate enargite from other copper sulfides, such as chalcopyrite, in industrial flotation processes when it occurs as an impurity [25]. Therefore, conventional pyrometallurgical technologies would inevitably cause environmental problems by emission of poisonous gas and dusts when processing arsenic-bearing copper sulfides.

Bioleaching offers an alternative to conventional technologies for enargite processing. Enargite is recalcitrant as its dissolution rate is very slow in acidic solution [30, 31]. Acidithiobacillus ferrooxidans is only able to bioleach 3% of the copper from enargite in 130 days at 25 °C [31]. However, bioleaching with extremely thermoacidophilic (optimal growth T m ≥ 60 °C, pH ≤3) archaea belonging to the phylum Crenarchaeota could be a promising alternative for the extraction of copper from enargite. This is predicated based on two process features: accelerated bioleaching rate and improved recovery yield resulting from the use of high incubation temperatures and minimization of surface passivation [27, 37]. However, both Cu2+ and As5+ ions accumulate in the leachate to high concentration during biomining of arsenic-bearing copper sulfide ores and would challenge the survival and proliferation of extreme thermoacidophiles [41]. These organisms generally have a significantly lower minimal inhibitory concentration (MIC) of heavy metals/metalloids as compared with bioleaching mesophiles and moderate thermoacidophiles [23, 40]. Therefore, the availability of simultaneous resistance towards both copper and arsenic constitutes a desirable trait in extreme thermoacidophiles.

A variant of M. sedula with Cu2+ resistance up to 200 mM, named CuR1, was recently isolated in our laboratory through adaptive laboratory evolution (ALE) [18]. The supranormal cupric resistant trait of CuR1 indicates that it is eligible for the isolation of arsenate and copper cross-resistant strains through ALE. Previous studies on arsenic resistance have shown that the typical ars operons, such as arsRCB and arsRCDAB, are broadly distributed in biomining-relevant acidophilic bacteria [14, 15, 17]. However, none of these well-characterized arsenic resistance mechanisms were observed in the genomes of extremely thermoacidophilic archaea. In their absence, it is plausible that novel arsenic resistance mechanisms might exist in lithoautotrophic thermoacidophilic archaea.

Microbial consortia rather than pure cultures have been adopted to extract basic (copper, zinc) and precious (gold, nickle) metals from sulfide minerals in the biomining industry or at bench-scale studies [5, 9,10,11, 26]. Previous studies have shown that bioleaching consortia consisting of acidophiles belonging to different genera or species exhibited more dramatic bioleaching capacity due to higher Fe2+/sulfur oxidation activity, and organic carbon source utilization relative to single pure cultures [22, 26]. In addition, succession of microbial community structure was observed when acclimating biomining consortia to high pulp density of chalcopyrite at bench-scale for two years or continuously processing gold-bearing arsenopyrite concentrate in stirred tanks at commercial scale for around two decades [5, 13, 26, 39]. Consistent with this idea, it has also been proposed that genetic mutation might enhance the growth and resistance of bioleaching species, therefore, altering the microbial community structure during bioleaching [26]. However, experimental data supporting this hypothesis are unavailable, which probably is ascribed to the lacking of quantitative techniques to differentiate these evolved strains from their closely related parental strains when they coexisted in the same consortium. If one of the strains had a unique genetic marker a quantitative PCR-based technique would be able to provide relevant data to support this hypothesis. The M. sedula copA mutant, a strain with a pyrE gene insertion in copA gene (Msed_0490), was the first genetically engineered bioleaching extremely thermoacidophile [18]. The copper resistance of the M. sedula copA mutant was reduced from 76 to 40 mM Cu2+ as compared with the parental strain M. sedula DSM 5348 [18]. The copper-resistant M. sedula CuR1, a derivative of M. sedula DSM 5348 isolated through ALE, contained mutations in genome and showed increased copper resistance to 200 mM Cu2+ [19]. Therefore, either the M. sedula CuR1 or its derivatives isolated in this study can be adopted to construct a defined consortium together with the M. sedula copA mutant to simulate the adaptation to elevated pulp density of copper sulfide minerals (such as chalcopyrite and enargite).

The purpose of this study was to isolate and characterize novel M. sedula strains with cross-resistance to high levels of both arsenate and copper using an ALE process as described in our previous study [20]. Heterotrophic growth with 0.2% (w/v) tryptone as the sole energy substrate of the evolved strains under high concentrations of Cu2+ or As5+ was compared. Genetic changes were evaluated by genome resequencing. Additionally, the effects of introduction an isolated cross-resistant strain M. sedula ARS50-2 to a defined consortium (consisted of a naturally occurring M. hakonensis HO 1-1 and a genetically engineered M. sedula copA mutant) on the enargite bioleaching in batch cultures at different PD were studied. Finally, the successions of microbial community structure of these defined consortia were quantitatively analyzed by qPCR. The findings from this study provide insights into arsenate resistance mechanisms in arsenate and copper cross-resistant strains of M. sedula, and highlight the importance of genetic mutations on modulating the microbial community structure succession and its function during the bioleaching of copper sulfide ores.

Materials and methods

Archaeal strains and cultivation

Archaeal strains used in this study included Metallosphaera sedula DSM 5348T (wild type), the Cu2+ resistant strain M. sedula CuR1 [18], the Cu2+ sensitive strain M. sedula copA mutant [18], newly derived Cu2+ and As5+ cross-resistant strains M. sedula ARS50-1 and ARS50-2 [2], and M. hakonensis HO 1-1 [12]. Basal salts medium (BSM) [3] was adjusted to pH 2.0 using sulfuric acid, and supplemented with 0.2 or 0.1% (w/v) tryptone for heterotrophic growth of M. sedula strains and M. hakonensis HO 1-1, respectively, as described previously [18]. To test the Cu2+ and As5+ resistance capacities, CuSO4 or KH2AsO4 (Sigma-Aldrich, USA) was added to the medium. The cultures were incubated at 75 °C in either glass screw-cap flasks with aeration in orbital baths or in glass screw-cap test tubes, which were placed in rotary drum agitator that was mounted in incubators as described previously [18].

Mineral components

Mineralogical components of enargite concentrate used in this study was identical with that described in a previous study: enargite (60%), pyrite (FeS2) (30%), nowackiite (Cu6Zn3As4S12) (5%) and balance quartz, as indicated by X-ray diffraction (XRD) analyses [1]. The particle size of this concentrate was superfine, 84.1% of which was less than 30 μm in diameter. The main chemical composition of the concentrate was (w/w): 27.66% Cu, 9.59% As, 14.01% Fe, 39.75% S and 0.82% Zn.

Isolation of arsenate resistant M. sedula strains

The Cu2+ resistant M. sedula strain CuR1 was inoculated from a frozen permanent into 5 mL of medium for heterotrophic growth in glass screw-cap test tubes. Planktonic growth was monitored by the optical density at the wavelength of 540 nm. Mid-exponential phase cultures were repeatedly subcultured into fresh heterotrophic growth medium supplemented with a gradually increasing concentration of KH2AsO4. After several weeks of passage, resistant cells grew heterotrophically in the presence of 50 mM As5+ in the medium and two clonal isolates were recovered using solid complex medium. Culture plates were prepared using basal salts medium (adjusted to pH 3.0 using sulfuric acid) mixed with 0.6% (w/v) gelrite (Kelco) and supplemented with 0.2% (w/v) tryptone. Plates were incubated at 75 °C for 5 days. The Cu2+ and As5+ cross-resistant capacities of two clonal isolates were tested and compared with their ancestral strain M. sedula DSM 5348T (wild type) and parental strain M. sedula CuR1, respectively.

Genome resequencing and nucleotide sequence accession numbers

Genomic DNA isolation, sequencing and sequence assembly were carried out as described previously [2]. The nucleotide sequences have been deposited with GenBank under accession no. CP012172 and CP012173 for re-sequenced M. sedula ARS50-1 and M. sedula ARS50-2, respectively [2].

Bioleaching experiments

Prior to inoculation with cells, 1 and 3% (w/v) enargite concentrate were, respectively, added to 50 mL BSM medium adjusted to pH 2.0 with sulfuric acid in 250 mL screw-cap glass flasks and incubated at 75 °C with agitation (175 rpm) for 24 h to wash out the residual floatation reagents in the minerals. The enargite residue was collected by centrifugation at 3000 rpm for 5 min, suspended and transferred into 250 mL flasks containing 50 mL of fresh BSM medium (pH 2.0) supplemented with 0.05% (w/v) tryptone. A total number of 1.8 × 109 cells of either pure Metallosphaera strains or mixed Metallosphaera spp. cultures (consortia) were then added to the flasks.

Two consortia were constructed for enargite bioleaching. Consortium A consisted of M. sedula copA mutant and M. hakonensis HO 1-1, while consortium B contained M. sedula copA mutant, M. hakonensis HO 1-1 and M. sedula ARS50-2. The copper sensitive M. sedula copA mutant was used to construct the defined consortia instead of the ancestral wild type M. sedula because it contains a pyrE gene insertion in copA gene (Msed_0490). This unique genetic marker can be screened for and quantitatively measured by qPCR. M. hakonensis HO 1-1 is an extremely thermoacidophile with high sulfur oxidation capacity and has the potential for bioleaching refractory chalcopyrite [6, 32]. M. hakonensis HO 1-1 was included in consortium to consolidate the sulfur oxidation capacity. M. sedula ARS50-2 rather than M. sedula ARS50-1 was adopted to construct consortium B because it showed a slightly higher growth rate when grown heterotrophically in the presence of 140 mM Cu2+ or 75 mM As5+. The adaptive evolution, genetic differences and metal/metalloid resistance capacities of the cross-resistant M. sedula ARS50-2 and the copper sensitive M. sedula copA mutant were compared (Fig. S1). All experiments were conducted in duplicates and abiotic controls for each pulp density were included under the same conditions. All samples were incubated at 75 °C in glass screw-cap flasks with agitation in orbital baths. Evaporated water was compensated daily by addition of distilled water based on weight loss. One milliliter supernatant samples were taken at 3 days interval for analysis of the ionic concentrations and the planktonic cell number.

Analytical methods

Planktonic cell numbers were counted using a Thoma counting chamber under light microscope. The pH of the leachate was measured using a pH meter (Fisher Scientific, Model AB30). The Eh value of leachate was determined using a platinum electrode with an Ag/AgCl reference electrode (American Marine Pinpoint ORP Monitor). The concentration of Fe2+/Fe3+ was quantified using the Ferrozine assay [34]. The concentration of Cu2+ was determined spectrophotometrically with diethyldithiocarbamate (DDTC) at a wavelength of 440 nm [38]. The total concentrations of extracellular arsenic ions were quantified by inductively coupled plasma mass spectrometry (ICP-MS) as described previously [28]. Briefly, culture samples were clarified by centrifugation. Samples of the resulting supernatants were analyzed by ICP-MS using an Agilent ICP-MS 7500cx. A certified arsenate reference standard was used for sample normalization. All reported values are averages from duplicate samples.

Microbial community structure analysis by qPCR

Leachate samples (4 mL) were centrifuged at 2500 rpm for 20 s to remove solid particulates. The supernatants were then transferred to separate tubes and centrifuged at 10,000g for 5 min to pellet the cells. Genomic DNA was prepared from the pellets using a Wizard® Genomic DNA Purification Kit (Promega). Strain-specific PCR primer pairs for qPCR were designed (Table 1) and validated by PCR (one cycle of 95 °C for 5 min, and then 31 cycles of 95 °C for 15 s, 64 °C for 30 s, and 72 °C for 30 s) and agarose gel electrophoresis. The qPCR was performed using an iCycler iQ Real-time PCR detection system (Bio-Rad Laboratories, Inc.). Procedures for the detection of the specificity of primers and for qPCR were as described [42]. The reaction mixture contained 12.5 μl SYBR® Green Real-Time PCR Master Mix (Toyobo Co., Ltd., Osaka, Japan) which contains SYBR Green I dye, Taq DNA polymerase, dNTP and MgCl2; and 1 μl of 10 μM solution of sense/anti-sense primer, 0.2 µl or 1 µl of genomic DNA. H2O was added into the reaction mixture to a final volume of 25 µl. The qPCR program was identical with the aforementioned PCR program except for the cycle number increased to 40. After the completion of each run, melting curves for amplicons were measured by raising the temperature 0.5 °C from 55 to 95 °C while monitoring fluorescence. The cell abundance of M. sedula ARS50-2 in consortium B was calculated by subtracting the cell abundance of M. sedula copA mutant from the cell abundance of total M. sedula strains.

Table 1 Strain specific qPCR primer pairs for Metallosphaera spp. based on the copA gene
Full size table

Results and discussion

Phenotypic response to copper and arsenate

The heterotrophic growth of wild type, M. sedula DSM 5348 (WT-Mse), the copper-resistant strain CuR1, and the newly isolated copper and arsenate resistant strains M. sedula ARS50-1 and M. sedula ARS50-2 were examined in the presence of 140 mM Cu2+ or 75 mM As5+ (Fig. 1). All strains grew to stationary phase in the absence of added copper and arsenate in approximately 2.5 days (Fig. 1a). M. sedula CuR1, M. sedula ARS50-1 and M. sedula ARS50-2 grew in the presence of 140 mM Cu2+ with no apparent growth lag (Fig. 1b). This indicated that the high copper resistance of M. sedula ARS50-1 and M. sedula ARS50-2 was inherited from their parental strain, CuR1. Of the strains tested, only M. sedula ARS50-1 and M. sedula ARS50-2 grew well in the presence of 75 mM As5+ (Fig. 1c). Under identical conditions, no growth was observed for either the wild type or CuR1. The MIC for As5+ for CuR1 was found to be 30 mM when grown in flasks. This is the first report on the isolation of high level arsenic and copper cross-resistance for an extremely thermoacidophilic member of the archaea. Importantly, this trait was heritable. M. sedula ARS50-1 and M. sedula ARS50-2 retained the ability to grow well in the presence of 75 mM As5+ despite passaging them three times in medium without As5+. This finding also excluded the possibility that arsenate resistance was stress-induced form of metalloid resistance. Finally, both the Fe2+ and sulfur oxidation capacities of M. sedula ARS50-1 and M. sedula ARS50-2 remained almost identical with that of the wild type M. sedula and the parental strain CuR1. This might indicate that the genes involved in Fe2+ and sulfur oxidation pathways were not affected by mutations during the adaptive laboratory evolution process.

Fig. 1
figure 1

Heterotrophic growth curves of different M. sedula strains in the absence of extra Cu2+/As5+ (a), in the presence of 140 mM Cu2+ (b) or 75 mM As5+ (c). WT-Mse (M. sedula DSM 5348T), CuR1 (M. sedula CuR1), ARS50-1 (M. sedula ARS50-1), ARS50-2 (M. sedula ARS50-2)

Full size image

Genome resequencing of cross-resistant M. sedula stains

The significant phenotypic difference between CuR1 and its arsenate resistant derivatives, M. sedula ARS50-1 and ARS50-2, under high concentration of As5+ might have arisen due to mutation(s) accumulated during the adaptive process and their identities might provide insights into the mechanism of arsenate resistance. Genome resequencing resulted in a sequence depth of 240-fold for both M. sedula ARS50-1 and M. sedula ARS50-2 (Table 2). Genome resequencing data showed that M. sedula ARS50-1 and M. sedula ARS50-2 gained 12 and 13 mutations, respectively, with five shared mutations in both strains during the adaptive process. This period consisted of 28 generations (cell divisions) occurring through nine consecutive culture passages. This resulted in a mutation rate of (2.20 ± 0.11) × 10−7 mutations per cell division. This rate was comparable to the mutation rate for the CuR1 as described previously [19].

Table 2 Summary of sequencing coverage and mutations
Full size table

Both genomes had mutations in both coding regions and intergenic regions relative to their parental strain, CuR1 (Table 2). Only insertions and non-synonymous mutations evident in coding regions were further considered because these were most likely to alter gene functions. A tabulated summary of these mutations with detailed information including mutation position in the genome, gene annotation, mutation type, gene/protein length and its proximity to identified domains is presented in Table 3. Two non-synonymous mutations were observed in Msed_0837, a pseudogene, in the genome of M. sedula ARS50-2, which might encode a transposase as indicated by BLAST analysis. A non-synonymous mutation led to the change of Gly163 to Glu163 in Msed_1139, which located in the conserved domain CsaX_III_U. Currently, the physiological role of Msed_1139 remains unknown. A non-synonymous mutation changing Asp31 to Asn31 (acidic to neutral) was mapped to the extracellular region of Msed_1340, which encoded a small membrane protein containing two transmembrane helices, but had no known function or conserved domains. This mutation might change the physiological function of Msed_1340 to some extent. Msed_1407, which encoded a small cytoplasmic protein with the physiological function unknown, underwent a non-synonymous mutation in both strains consisting of a change from Cys10 to Tyr10 at the N-terminal. Msed_1846, which encoded an S-adenosylmethionine (SAM) dependent FkbM family methyltransferase in M. sedula, underwent a 1-nt insertion that truncated the gene sequence by 15 nucleotides. Consequently, the fragment constituted by the 13 amino acid residues at the C-terminal was substituted by another 8 amino acid residues. The effects of this C-terminal event on the physiological function of Msed_1846 was unknown, although it lied outside the conserved domain. The FkbM methyltransferase from Streptomyces tacrolimus was involved in methylating the C-31 hydroxyl group of 31-O-demethyl-FK506 and FK520 to enable synthesis of macrocyclic polyketides [21]. However, as the genes needed for biosynthesis of the macrolactone ring of FK506 were absent in the genome of M. sedula [4], it was unlikely that Msed_1846 involved in polyketide synthesis. The SAM-dependent methyltransferase involved in the methylation of various molecules, including proteins, DNA and secondary metabolites [35]. Whether the Msed_1846 was relevant to arsenate resistance remained to be identified. The hypothetical protein Msed_1892 contained a non-synonymous change of Asn23 to Asp23. This protein contained a LanC-like superfamily domain. However, the mutation was outside of this region. As the natural amino acid was not conserved among homologs in other extremely thermoacidophilic species, this mutation might not affect the physiological role of Msed_1892. The uncharacterized membrane-associated protein Msed_2066 contained a non-synonymous change from His83 to Pro83 in the genome of M. sedula ARS50-1. Since this amino acid residue was not highly conserved among the homologs of Msed_2066 from extremely thermoacidophiles, this mutation might not change the physiological function. This gene contained a conserved DedA domain and recent studies showed that DedA membrane proteins from E. coli involved in cell division, membrane homeostasis, pH/PMF homeostasis and temperature sensitivity [8]. In the genome of M. sedula ARS50-2 there was a 1 nt insertion of T that truncated Msed_2066 from 600 nt to 228 nt. This mutation probably inactivated the gene. However, no significant difference in arsenate resistance was observed between M. sedula ARS50-1 and M. sedula ARS50-2 apparently excluding a role for Msed_2066 in arsenate resistance (Fig. 1c).

Table 3 Summary of mutations in M. sedula ARS50-1 and M. sedula ARS50-2 genome as compared with parental strain CuR1
Full size table

Enargite concentrate bioleaching by pure culture of Metallosphaera spp

Bioleaching capacities of each pure culture of the Metallosphaera strains were evaluated by processing enargite concentrate at PD of 1% (w/v) (Fig. 2). Different bioleaching behaviors were observed between M. hakonensis HO 1-1 and these two M. sedula strains.

Fig. 2
figure 2

Changes of planktonic cell number (a), pH value (b), Eh values (c), Fe2+ ion (d), Fe3+ ion (e), and copper recovery yield (f) of enargite bioleached by pure cultures of Metallosphaera spp. at PD of 1% (w/v)

Full size image

The planktonic cell numbers of the M. sedula copA mutant and M. sedula ARS50-2 increased rapidly after inoculation and reached the plateau (around 6 × 108 cells/mL) at the 9th day followed by a decrease after the 15th day (Fig. 2a). However, the cell density of M. hakonensis HO 1-1 in leachate remained at low level (around 4 × 107 cells/mL) during the first 9 days. After this period its cell density increased to 2 × 108 cells/mL and remained at this level throughout the duration of the experiment (Fig. 2a). The survival of these extremely thermoacidophiles indicated that they can effectively utilize the sulfur and/or ferrous iron dissolved from enargite concentrate for growth [Eqs. (1)–(5) [33, 36]. The pH values of the leachate of M. sedula copA mutant and M. sedula ARS50-2 decreased steadily after inoculation (Fig. 2b). This indicated that sulfur was efficiently mobilized from enargite and pyrite, and was further oxidized to sulfate and generated protons by these extreme thermoacidophiles (Fig. 2b) [Eqs. (3) and (5)]. However, the pH of the leachate of M. hakonensis HO1-1 remained largely unchanged before the 21st day, thereafter it decreased. Only a slight decrease of pH was observed for the abiotic control (Fig. 2b). The Eh of the leachate of abiotic control remained around 340 mV (using an Ag/AgCl electrode as a reference) (Fig. 2c). The Eh of M. hakonensis HO 1-1 remained around 340 mV till the 21st day, then increased up to terminal value of 482 mV. On the contrary, the Eh of both M. sedula copA mutant and M. sedula ARS50-2 increased rapidly to around 540 mV and then remained largely unchanged (Fig. 2c). The concentration of Fe2+ in leachate remained at low level throughout the bioleaching process in the presence of extremely thermoacidophile (Fig. 2d). The Fe3+ concentration in leachate of M. sedula ARS50-2 and M. sedula copA mutant increased drastically since the 3rd day after inoculation and reached a plateau on the 18th and 21st day, respectively (Fig. 2e). This indicated that M. sedula strains can effectively dissolve pyrite, the only iron-bearing mineral in the enargite concentrate, to obtain the energy for survival through ferrous iron oxidation and generate the ferric iron for the enargite and pyrite leach [Eqs. (2) and (3)]. On the contrary, the Fe3+ concentrations of M. hakonensis HO1-1 remained at low level (almost 0 mM before the 24th day), which was comparable with that in the abiotic control (Fig. 2e). This should be attributed to the deficiency of Fe2+ oxidation ability for M. hakonensis HO1-1. This assumption was validated by comparison its ferrous iron oxidation capacity with the other two M. sedula strains with the ferrous iron as the sole energy substrate (Fig. S2). The inability of M. hakonensis HO1-1 to oxidize ferrous iron to Fe3+ would lead to the unavailable of sufficient ferrous and sulfur, the energy substrates for the cellular metabolism, that originally released from enargite and pyrite [(Eqs. (1)–(3)]. Therefore, fewer planktonic cells were sustained and an accordingly higher pH value of leachate together with much lower Eh was observed during the bioleaching by M. hakonensis HO1-1. Consistent with the concentration of Fe3+ in leachate, moderately high copper recovery yields were achieved by pure cultures of M. sedula copA mutant (61.73%) and M. sedula ARS50-2 (59.93%) (Fig. 2f). Only 11.30% and 14.95% copper was recovered from enargite concentrate for the abiotic control and M. hakonensis HO1-1, respectively.

$$2{\text{FeS}}_{2} + 2{\text{H}}_{2} {\text{O}} + {\text{O}}_{2} \to 2{\text{Fe}}^{2 + } + 4{\text{SO}}_{4}^{2 - } + 4{\text{H}}^{ + } ,$$
(1)
$${\text{FeS}}_{2} + 8{\text{H}}_{2} {\text{O}} + 14{\text{Fe}}^{3 + } \to 15{\text{Fe}}^{2 + } + 2{\text{SO}}_{4}^{2 - } + 16{\text{H}}^{ + } ,$$
(2)
$${\text{Cu}}_{3} {\text{AsS}}_{4} + 9{\text{Fe}}^{3 + } + 2{\text{H}}_{2} {\text{O}} \to 3{\text{Cu}}^{2 + } + 9{\text{Fe}}^{2 + } + 4{\text{S}}^{0} + {\text{AsO}}_{2}^{ - } + 4H^{ + } ,$$
(3)
$$4{\text{Fe}}^{2 + } + 4{\text{H}}^{ + } + {\text{O}}_{2} \to 4{\text{Fe}}^{3 + } + 2{\text{H}}_{2} {\text{O,}}$$
(4)
$${\text{S}}^{0} + 2{\text{H}}_{2} {\text{O}} + {\text{O}}_{2} \to {\text{SO}}_{4}^{2 - } + 2{\text{H}}^{ + } .$$
(5)

Enargite concentrate bioleaching by consortia

By comparison of the bioleaching behaviors between consortium A (consisted of M. sedula copA mutant and M. hakonensis HO 1-1) and consortium B (consisted of M. sedula copA mutant, M. hakonensis HO 1-1 and M. sedula ARS50-2) at the PD of 1 and 3% (w/v), it was obvious that introducing M. sedula ARS50-2 into consortium A improved the enargite concentrate bioleaching, especially at the PD of 3% (w/v) (Fig. 3).

Fig. 3
figure 3

Changes of planktonic cell number (a), pH value (b), Eh values (c), Fe2+ ion (d), Fe3+ ion (e), and copper recovery yield (f) of enargite bioleached by consortia consisted of Metallosphaera spp. at PD of 1 and 3% (w/v)

Full size image

The planktonic cell numbers of consortium A and consortium B increased dramatically to their respective plateaus 15 days after the initiation of bioleaching, without a lag phase was observed (Fig. 3a). The planktonic cell number of consortium B was always higher than that of consortium A at each pulp density, which might indicate that more energy substrates, such as ferrous iron and sulfur, were mobilized from the enargite and pyrite for the proliferation of the cultures in the presence of M. sedula ARS50-2. Significant increase in the planktonic cell number of each consortium was observed at the PD 3% (w/v) on the 21st and 24th day as compared to that at the PD of 1%.

The pH values of leachate of both consortia significantly decreased compared with the abiotic control starting on the 3rd day after inoculation (Fig. 3b). The pH value of consortium B was slightly lower than that of consortium A throughout the experiment at the PD of 1% (w/v) (Fig. 3b). This observation might indicate that the sulfur oxidation activity of consortium B was only slightly higher than that of the consortium A at this pulp density. It was worth to mention that the pH values of both consortia were much lower than that of each of these pure cultures from the 9th day after the initiation of the bioleaching at PD of 1% (w/v) by inoculating identical amount of cells, either pure culture or mixed cultures (Fig. S3a). This indicated that the M. hakonensis HO 1-1 improved the sulfur oxidation activities of the consortia when it coexisted with M. sedula copA and/or M. sedula ARS 50-2. Therefore, the defined consortia constructed in this study were robust regarding on the sulfur oxidation capacity in spite of few extremely thermoacidophilic strains adopted. The pH of consortium A at the PD of 3% (w/v) remained much higher than that at the PD of 1% (w/v). This indicated that sulfur oxidation activity of consortium A was inhibited by the increased pulp density of minerals. However, the pH of consortium B at 3% (w/v) PD was much lower than that of consortium A; this difference continued to increase over time. This indicated that consortium B probably had a higher sulfur oxidation capacity.

The Eh of the leachate of consortium A and consortium B were very similar, and remained almost unchanged at around 540 mV (vs. Ag/AgCl) after the 6th day at the PD of 1% (w/v) (Fig. 3c). Both the leachate of consortium A and consortium B had much lower Eh at PD of 3% (w/v). The Eh of consortium B was obviously higher than that of consortium A after the 9th day at PD of 3% (w/v). This observation indicated that the ferrous iron oxidation capacity of consortium A, which only contained a metal sensitive Fe2+ oxidizer M. sedula copA mutant, was significantly inhibited; however, a compromised inhibitory effect was observed for consortium B, which contained another Fe2+ oxidizer, the cross-resistant M. sedula ARS50-2, as compared with consortium A. This assumption was further confirmed by the increase of ferrous iron and the decrease of ferric iron accompanied with the elevated PD (Fig. 3d, e). At the PD of 3%, the concentration of Fe2+ in both consortia remained much higher than that of each consortium at the PD of 1% (w/v) (Fig. 3d). The concentration of Fe2+ in leachate of consortium A and consortium B increased slightly in the first 12 days, and peaked at 18th and 15th day, respectively, then started to decrease (Fig. 3d). The concentration of Fe2+ in consortium A was higher than that of consortium B after the 18th day. The concentration of Fe3+ in leachate of consortium A and consortium B at PD of 1% (w/v) increased drastically after the 3rd day, peaked at the 15th day, and started to decrease gradually after the 18th day (Fig. 3e). The concentration of Fe3+ of consortium B increased greatly since the 6th day, and reached 16.48 mM at the 27th day at the PD of 3% (w/v) (Fig. 3e); while the concentration of Fe3+ in consortium A only increased slightly after inoculation, and reached 5.31 mM at the end of the bioleaching (Fig. 3e). These data indicate that consortium B had much higher Fe2+ oxidation capacity as compared with consortium A at the PD of 3%.

Consistent with the comparable concentration of Fe3+ generated in leachate by each consortium, moderately high copper recovery yields were achieved from enargite concentrate by consortium A (61.52%) and consortium B (66.83%) at the PD of 1% (w/v), respectively, after 27 days of bioleaching (Fig. 3f). Compared with the copper recovery yield achieved by each of these ferrous iron oxidizing M. sedula strains, only the consortium B had a slightly higher copper recovery yield at the PD of 1%(w/v) (Fig. S3b). Nevertheless, this indicates that the sulfur oxidizing M. hakonensis HO 1-1 has the potential to consolidate the bioleaching process together with these M. sedula strains. With the increase of PD to 3% (w/v), moderately high copper recovery yield (61.57%) was achieved by consortium B; however, only 35.28% copper was recovered by consortium A.

Previous study has shown that mineral particles, especially at elevated pulp density, impose mechanical damage to bioleaching acidophile via attrition in stirred reactors [7]. The mechanical damage might contribute to the decreased copper recovery yield of each consortium at the PD of 3% (w/v) as compared with that at the PD of 1%(w/v) in this study. However, the mechanical damage alone cannot explain why there were significant differences between consortium A and consortium B during bioleaching at PD of 3% regarding on the planktonic cell number, pH value, Eh, Fe2+/Fe3+ concentration and copper recovery yield (Fig. 3a–f). Instead, these differences were probably correlated with the Cu2+ concentration in leachate (Fig. 4). The Cu2+ ion in leachate of consortium A increased steadily to 39.17 mM in the first 12 days, and remained almost unchanged until the 21st day, and then increased slightly to 46 mM till the end of bioleaching at the PD of 3% (w/v). The Cu2+ ion that mobilized from enargite concentrate accumulated to inhibitory concentration to M. sedula copA mutant (with a MIC of 40 mM), the only ferrous iron oxidizer in consortium A, and therefore inhibited the ferrous oxidation capacity of this consortium since the 12th day as reflected by the increased Fe2+ concentration together with the low Fe3+ concentration and Eh since the 12th day (Fig. 3c–e) [18]. It was undoubtedly that the activity of M. sedula copA mutant in consortium B was inhibited further at the PD of 3% (w/v) because much higher concentration of Cu2+ ion was accumulated in leachate after the 15th day (Fig. 4). Since the other ferrous iron oxidizer, M. sedula ARS50-2, in consortium B has a much higher MIC for Cu2+ ion (more than 140 mM), only a compromised inhibitory effect was observed in consortium B during the bioleaching at the PD of 3% (w/v) (Fig. 3c–e).

Fig. 4
figure 4

Changes of Cu2+ concentration in leachate during enargite bioleaching at PD of 1 and 3% (w/v)

Full size image

The arsenic ions in leachate of each consortium at the 12th and 21st day at each PD were analyzed to ascertain if it also affected the bioleaching processes at elevated PD (Fig. S4). The arsenic ions in leachate of consortium A and consortium B at the PD of 1 and 3% (w/v) was low at the 12th and 21st day, which was not proportionated to the concentration of Cu2+ during enargite bioleaching. A similar observation was showed in a previous study during enargite bioleaching by extremely thermoacidophilic Acidianus brierleyi, ascribing to the formation of scorodite (FeAsO4·2H2O) [37]. Therefore, the low arsenic concentration in leachate in the presence of extremely thermoacidophiles in this study probably also resulted from the precipitation of scorodite. The arsenic concentration was much less than the MIC of these M. sedula strain (30 mM for M. sedula copA mutant and 75 mM for M. sedula ARS50-2). Therefore, the arsenic ions in leachate did not affect the activity of both consortia at each PD.

Microbial community structure analysis by qPCR

The succession of microbial community structure during the bioleaching of enargite at PD of 1 and 3% (w/v) was quantitatively analyzed by qPCR based on the strain-specific primer pairs (Fig. S5). By comparison the microbial community succession between consortium A and consortium B at the PD of 1 and 3% (w/v), it was obvious that introducing M. sedula ARS50-2 into consortium A decreased the viability of M. sedula copA mutant and increased the viability of M. hakonensis HO1-1 in consortium (Fig. 5).

Fig. 5
figure 5

Microbial community succession during the bioleaching of enargite with extremely thermoacidophilic consortia [a consortium A at PD 1% (w/v); b consortium B at PD 1% (w/v); c consortium A at PD 3% (w/v); d consortium B at PD 3% (w/v)]

Full size image

M. sedula copA mutant was the predominant species in consortium A during the bioleaching of 1% (w/v) enargite (Fig. 5a), which was consistent with the observations that the arsenic ions and Cu2+ in leachate were not inhibitive to it (Fig. 4 and Fig. S4). M. hakonensis HO 1-1 was found with low proportion in leachate throughout the bioleaching process, possibly due to its deficiency in utilizing Fe2+ to obtain the energy for cell growth. M. sedula ARS50-2 substituted M. sedula copA mutant as the major species in consortium B at PD of 1% (w/v) (Fig. 5b). The cell abundance of M. sedula ARS50-2 increased drastically to the maximum at the 9th day and maintained as the major species since then. M. sedula copA mutant was also abundant, but with much lower proportion of the consortium population as compared with that in consortium A.

M. sedula copA mutant in consortium A was still the predominant species throughout the bioleaching process at PD of 3% (w/v) (Fig. 5c). Increased viability of M. sedula copA mutant was detected during the first 9 days, thereafter it decreased significantly and remained at low level, which was probably ascribed to the inhibitory effect imposed by the dissolved Cu2+ in leachate. The cell density of M. hakonensis HO 1-1 increased significantly at the 15th day as compared with that at the 9th day. The proportion of M. hakonensis HO 1-1 in consortium A increased steadily since the 15th day, with a high ratio (46.50%) was detected at the 27th day (Fig. 5c). M. sedula copA mutant was substituted by M. sedula ARS50-2 as a minor species after this cross-resistant strain had been introduced into consortium A when initiating bioleaching at the PD of 3% (w/v) (Fig. 5d). Its abundance decreased further since the 15th day. M. sedula ARS50-2 remained as the major species till the 21st day during enargite bioleaching at 3% (w/v). M. hakonensis HO 1-1 was almost undetectable during the first 9 days, then its abundance increased significantly after the 15th day to become the predominant species in the consortium at the end of the bioleaching trial (Fig. 5d). This might be ascribed to the reduced sulfur compounds accumulated during enargite bioleaching, which might be not effectively utilized by these M. sedula strains, thus provided substrates for the chemolithoautotrophic growth of M. hakonensis HO 1-1.

Introduction of cross-resistant M. sedula ARS50-2 to the consortium A consolidated the bioleaching capacity probably by modulating the microbial community structure. The total viability of M. sedula strains together with M. hakonensis HO1-1 in consortium B was higher than that in consortium A since the 15th day at the PD of 1% (w/v), which might confer higher Fe2+ and sulfur oxidation capacities and, therefore, led to the slightly higher copper recovery yield (Fig. 3f). With the increase of PD to 3% (w/v), the bioleaching capacity of consortium A was significantly inhibited ascribing to the decrease of viability of M. sedula copA mutant imposed by Cu2+, with only 35.28% copper recovered; however, the bioleaching capacity of consortium B that containing the cross-resistant M. sedula ARS50-2 was just slightly inhibited, with 61.57% copper recovery was achieved.

Taken together, the increased Cu2+ resistance conferred by mutations that occurred during ALE was critical for M. sedula ARS50-2 to decrease the viability of M. sedula copA mutant and increase the viability of M. hakonensis HO1-1 when co-existing in consortium, especially at elevated pulp density. The data in this study are insightful for the succession of microbial community structure usually observed during the consecutive adaptation or application of bioleaching consortium to process sulfide minerals at high pulp density [26, 39]. Owing to the mutations that occurred during the adaptation/application processes, additional novel strains with new traits (such as higher heavy metal resistance or higher growth rate) could be evolved under these selection pressures that were characteristic of high concentrations of heavy metal/metalloid ions in the bioleaching industry. Consequently, the changing microbial community structure altered the bioleaching capacity of the consortium resulting in a consolidated process.

Conclusion

In this study Cu2+ and As5+ cross-resistant strains of the extreme thermoacidophile M. sedula strains were generated using ALE and mutations conferring As5+ resistance were identified. Introduction of the cross-resistant M. sedula ARS50-2 to a consortium consisted of Metallosphaera spp. modulated the microbial community structure and enhanced the bioleaching of refractory enargite. Differences in viability of the M. sedula copA mutant and M. sedula ARS50-2 correlated with their copper resistance capacity when coexisting in consortium for enargite bioleaching. This study implied that genetic mutations occurred spontaneously might confer new traits (such as Cu2+ resistance) to the evolved strains during bioleaching processes, and therefore modulated the microbial community structure and bioleaching capacity.