Multimetal bioremediation and biomining by a combination of new aquatic strains of Mucor hiemalis

Here we describe a unique microbial biotechnology for simultaneous bioremediation and biomining of twelve ionic metals overcoming the obstacles of multimetal toxicity to microbes. After a thorough search of key microorganisms in microbiomes of many sulfidic springs in Bavaria (Germany) over an area of 200 km2, we found three new strains EH8, EH10 and EH11 of Mucor hiemalis physiologically compatible and capable of multimetal-remediation and enrichment. We combined the multimetal-resistance, hyper-accumulation and elicitation power of EH8, EH10 and EH11 to develop a novel biotechnology for simultaneous removal, fractionation and enrichment of metal ions. As a first step we showed the intracellular fixing and deposition of mercury as nanospheres in EH8’s sporangiospores. Scanning Electron Microscopy-Energy-Dispersive X-Ray analysis revealed binding and precipitation of other applied metal ions as spherical nano-particles (~50–100 nm) at the outer electro-negative cellwall-surface of EH8, EH10 and EH11 sporangiospores. Microbiomes, germinated spores and dead insoluble cellwalls of these strains removed >81–99% of applied Al, Cd, Co, Cr, Cu, Hg, Ni, Pb, U, and Zn simultaneously and furthermore enriched precious Ag, Au and Ti from water all within 48 h, demonstrating the potential of new biotechnologies for safe-guarding our environment from metal pollution and concentrating precious diluted, ionic metals.

Increasing global demand and use of various metals/metalloids leads to their accumulation in the environment and thereby simultaneously increases the risk of multimetal-toxicity to various organisms. In higher organisms accumulated heavy metals can damage nerves, liver, bones, etc., as well as block functional groups of enzymes and cause cancer 1 . This metal accumulation is mainly due to various industrial processes such as; metal production, mining, electroplating, nuclear power generation, municipality waste incineration and by manufacture of a myriad of metal-containing electronic components 2 . Toxic metals/metalloids may even be mixed with recalcitrant and/or radioactive organic pollutants, e.g. Cu and biphenyl 3 , which can increase the severity of metal toxicity and problems of remediation. Controlling this pollution now represents a large challenge for our society. The removal and recycling of toxic metals or metalloids, particularly from water phase, deserves our special attention because metal-contaminated materials like soil, sediment and gaseous emissions can be washed-off with water, so that ultimately metal contaminants are highly concentrated in, e.g., waste water (nuclear, metallurgical, mining) and landfill leachates. Several technologies have been described previously for the removal of toxic metals or metalloids from water, e.g. flocculation 4 , adsorption 4 , bioaccumulation by live microbial cells 5 , immobilization 6,7 and reverse osmosis 4 . Existing technologies for metal removal from aqueous phase are either too costly, e.g. by reverse osmosis 4 , or they create excessive additional waste.
In contrast to ex-situ physical-chemical methods, in-situ natural enrichments of metal ions may take place due to physical-chemical and/or biological processes. Natural hydrothermal conditions, e.g. at temperatures above 60 °C, can enhance oxidative-reductive complexation and precipitation processes of metal ions in water, but have high energy requirements. Therefore, there is an increasing interest in biotechnology for the development of low-cost, low energy-consumption and effective bioremediation 8 and biomining 9 . Ex-situ or in-situ metal/metalloid removal by microbial biosorption 2 and enzyme-mediated metal precipitation at cell surfaces 10 could be cheap alternatives for treating metal-contaminated wastewater and/or water purification. So far, only a limited number of natural microorganisms have been found to show biosorption properties for a few toxic metals/metalloids, whereby they required very stringent culture conditions, e.g. concerning C-and N-source 11 and Fungal inhibition/toxicity assay. In fungal inhibition/toxicity assays 34 the strains EH8, EH10 and EH11 were not mutually antagonistic (Fig. 2). No visible demarcation line between the dual culture mycelium fronts of EH8 and EH10/EH11 was found, suggesting they are physiologically tolerant to each other ( Fig. 2A.1 www.nature.com/scientificreports www.nature.com/scientificreports/ dual cultures of aquatic Mucor hiemalis, e.g. EH5 28 and EH10, and EH10 and EH12, reacted antagonistically to one another, as shown by e.g. formation of demarcation lines between mycelium fronts, release of oily drops at the hyphal tips and color changes in response to the presence of antagonistic strains ( Fig. 2A.2-A.3). Purified viable germinating sporangiospores of aquatic M. hiemalis strains showed, as expected, strong presence of chitin that could be recognized and visualized as tiny anti-chitin stain spheres on the cell walls, as detected by fluorescence microscopy of anti-chitin fluorescein-labeled antibody (Fig. S2).
Zeta-potential of cell surface, hyper-metal accumulation and precipitation. Acid-base titration of the aquatic M. hiemalis sporangiospores did not show any distinct pK-values in the investigated pH range suggesting no domination of weak acids or bases ( Fig. 2A). The zeta-potential of spore cell surfaces varied depending The brushlike morphology of EH11's in situ microbiome fixed on rock (F, C.1) from the methane-emitting salty sulfidic environment of Künzing spring and of EH8's ex-situ grown microbiome fixed on expanded clay spheres (C.2) is visible. EH11 (C.3, sporangium S with sporangiophore P) showed spring-like hyphal morphology (C.4 and C.5) due to adaptation to bubbling methane. However, the spring-like hyphal morphology of EH11 disappeared after further cultivations on solid malt extract-agar medium.
www.nature.com/scientificreports www.nature.com/scientificreports/ on nutrient/mineral contents of the incubation medium. Surprisingly, a groundwater-like environment with low mineral and nutrient contents was shown to increase the electro-negativity of the sporangiospores in the pH range 6-10, if they were pre-germinated in groundwater ( Fig. 2A). Contrastingly, under a mineral-rich medium with high C-or N-contents 35 the zeta potential of sporangiospores' surfaces lost ca. 10% of the electro-negativity at pH 7, mainly due to the favored pre-occupation of some electronegative sites on the spore surfaces by the metal cations supplemented to the fungal growth medium. The surface potentials of the spores showed rapidly increasing electro-negativity (<−100 mV down to −480 mV) with increasing pH, indicating strong electrical attraction of aquatic M. hiemalis spore cell walls towards the positively charged metal cations.
Production of spherical metal nano-and micro-particles by activated and elicited sporangiospores. When a mixture of metal ions came into contact with cell surfaces of the activated sporangiospores of EH8, EH10 and EH11, metal spheres were readily produced. SEM showed the adsorption and precipitation of approx. 50-100 nm-sized spherical metal nanoparticles on the cell walls following 48 h incubation of mixed EH8, EH10 and EH11 activated sporangiospores in a metal mix solution (Fig. 2B1,B2,B2E). This was in accordance with the metal cation-attractive forces of the strongly electro-negative zeta-potential of spores' outer surfaces (Fig. 2C). EDX microanalysis localized the outer cellular surface of sporangiospores as the site of biosorption and precipitation of diverse metal nano-particles, e.g. Al, Pb, Cd and Cr, and showed the presence of P at the same spot, indicating selective spherical metal precipitations at the electro-negative reducing chitinised centers (Fig. S2) and phosphate-ligand sites of the cell surfaces (Fig. 2B2). This result is similar to the phenomenon of copper phosphate precipitation in Penicillium ochrochloron ATCC 36741 2 . The metal elimination power of the www.nature.com/scientificreports www.nature.com/scientificreports/ dead mixed insoluble cell walls, the spore mix and the mixed microbiomes of EH8, EH10 and EH11 is shown in Fig. 2D. Results showed a similar metal removal capacity of the microbiomes, the spore mix and the mixed cell walls.
In situ generation of diverse element particles by microbiomes. We observed precipitation of diverse nanometer-to micrometer-sized mineral particles, e.g. sphalerite (zinc sulfide), sulfur, iron colloids and iron sulfide, in microbiomes of sulfidic springs (Fig. 3). The precipitation of sphalerite in some microbiomes was also reported by Labrenz et al. 36 .
Kinetics of metal removal by elicitation of EH8, EH10 and EH11. The kinetics of metal removal depending on incubation time by fungal insoluble dead cell wall mix, spore mix and microbiome mix (EH8, EH10 and EH11) was compared with that of the control fungus (terrestrial M. hiemalis strain DSM2655, Fig. 4, Table 2). The hyper-accumulation of metals from contaminated water by fungal materials (see above) took place in two phases: an exponential phase (duration: ≤10 hours) and a slow phase (remaining duration: ≥38 hours) ( Fig. 4B-F). In the exponential phase, more than 50% of metals applied were removed; the remaining metals were removed in the slow phase. It was demonstrated that the mixed dead cellwalls from EH8, EH10 and EH11 when compared to their live spore mix and live microbiome mix were similarly effective with respect to (1) a wide range of metals removed (Al, Cd. Co, Cu, Cr, Ni, Pb, Zn, Ag, Au, U, Ti) and (2) rate of metal accumulation from water. We demonstrated the high efficiency of metal removal and enrichment by mixed fungal dead cell walls www.nature.com/scientificreports www.nature.com/scientificreports/ by applying a metal mix of Al, Cd, Co, Cu, Cr, Ni, Pb, Zn, Ag, Au, and U, each at a concentration of 1,000 µg/l. A universal three-parameter rational function with y = (a + b*x)/(1 + c*x), whereby y = concentration of metal (µg/l) and x = duration of incubation (h) was shown to fit the metal remediation kinetics of all the metals applied optimally depending on incubation duration (h) ( Table 2). Kinetics of residual mercury in water was inversely proportional to the amount of intracellular Hg fixed by EH8 independent of concentrations applied (1 mg.L −1 , 50 mg.L −1 ), apparently due to high mercury tolerance and detoxification activity (Fig. 4B,C) 5 .
Production of drinking water from metal-contaminated water by using successive fractionation and remediation biotechnology. The successive removal of various toxic metals by toxin-free insoluble cell walls of mixed grown EH8, EH10 and EH11 from water contaminated with 1,000 µg L −1 (Ag, Al, Cd, Co, Cr, Ni, Pb) is demonstrated for each metal in Fig. 5. Within 48 h the initial concentrations (each 1,000 µg/l) of aluminium and chromium decreased below the permitted threshold values for drinking water in the first step of fractionation e.g. following treatment of contaminated water with the insoluble mixed cell walls (Fig. 5A). After separation of cell wall materials in the first step, a re-treatment of water containing the residual metals Ni, Pb, Ag,  Table 3). The fitted curve parameters with statistical significance are given in Table 2. Standard deviations of measurements (n = 3) at each data point were less than 5%.
www.nature.com/scientificreports www.nature.com/scientificreports/ Cd and Co with the same amount of mixed cell walls removed all of them, keeping the residual concentrations below the permitted levels of metals in drinking water according to the German Water Ordinance 37 (Fig. 5B). The optimum metal removal by fungal cell walls occurred at neutral pH, although at pH below 7 or higher than 7 the percentage of metal removal decreased only slightly (Fig. 5C). Figure 6A-E show the concept of fractionation and biomining of metal ions based on experimental results. The first use of activated EH8 sporangiospores or microbiomes on expanded clay spheres will eliminate nearly all Hg(II) (99%), as well as some other metal ions as "fraction 1" from water by intracellular accumulation (Table S1) 5 . EH8 (Fig. 1A,B) of Marching spring (sporangiospores, microbiomes, cell walls) can retain Al (90%), Cr(III) (99%), Ni (86%), U (89%) as "fraction 2" 5 . EH10 (Fig. 1B) from Quarzitwerk spring biofilm will remove Cd (91%), Cu (85%), Pb (93%), and Zn (71%) as "fraction 3". In contrast, aerobically cultivated anaerobic EH11 with its unusual spring-like hyphal adaptation (Fig. 1C.3-C.5) from the microbiome of methane-emitting salty sulfidic Künzing spring can adsorb Al, Cd, Co, Cr, Cu, Pb, Ni and Zn as "fraction 4".

Fractionation and biomining of metal ions by successive use of EH8, EH10, EH11 and mixed grown EH8-EH10-EH11.
The characteristic ecological data and the glutathione S-transferase activities of the springs' key organisms (EH8, EH10 and EH11), microbiomes and the terrestrial strain Mhh (DSM 2655) are shown for the comparison of the biosorption properties of single strain, mixed grown aquatic strains and microbiomes (Table 3). In contrast to the biomass-mixture of separately grown (non-elicited) spores, the biomass-mixtures of activated and mixed-grown (elicited) sporangiospores, cellwalls and microbiomes of EH8, EH10 and EH11 can fractionate and even enrich precious metal ions, e. g. U, Ag, Au, Ti (Tables 2, 3 Table 2. Characteristic parameters of the generalized peak-fitting functions of insoluble dead cell walls (CW), spore mixtures (Sp) and cultivated fungal microbiomes (biofilms, Mb) describing highly efficient metal elimination by a combination of EH8, EH10 and EH11. Each row in the table represents an experimental subgroup, whereby a single ionic metal per subgroup was applied. The three parameters (a, b, c), the significance levels (p-values) by Student's t-tests and the correlation coefficients (r 2 -values) are given for each ionic metal applied in a mixture (1,000 µg/l per metal) after fitting the residual concentrations (y) of each metal ion in water by the generalized function y = (a + b*x)/(1 + c*x) depending on incubation duration (x, 0-48 h). Standard deviations of measurements at each data point were less than 5%. The asterisk mark (*) indicates the highest elimination of the respective metal ion from the aqueous phase.

Discussion
Harrison et al. 38 suggested a multifactorial model by which biofilm populations can endure metal toxicity by a process of biodiversification. In contrast to the overall microbial-community role for the metal precipitation 38 , it was shown by our previous studies that if a key microorganism for metal removal exists in microbiomes, it can remove toxic metals simultaneously with high efficiency which is comparable to the microbiome itself (Table S1) 5,29 .
Adaptation and evolution of some key M. hiemalis microorganisms in hostile environments of cold sulfidic springs are sufficient to confer special metal resistance, hyper-metal accumulation and detoxification power that can give the whole microbiome selective advantages for survival and facilitate their microbial community members to select favourable successions in extreme microbial habitats (Tables S1, 2).
Hyper-metal accumulating M. hiemalis strains EH8, EH10 and EH11 from sulfidic spring water were tested for metal accumulation after inhibition/toxicity assays. Using this assay, they were found to be physiologically tolerant to each other without showing any antagonistic inhibitory effects, and hence used in combination for the development of new metal remediation biotechnology. For fungal metal binding and remediation, the cell wall www.nature.com/scientificreports www.nature.com/scientificreports/ components 11 and their surfaces could play an important role. The cell wall of M. hiemalis consists of chitin (ca. 11%, Fig. S2), chitosan (32%), protein (5%) and phosphorus (1%) 11 , which are considered important for the determination of biotechnologically relevant features during growth and cell integrity 39 . In M. hiemalis MP/92/3/4 cell walls, the ratio of chitosan to chitin was ∼1: 0.84 (17.39% chitosan: 14.57% chitin, re-calculated), when the fungus was grown in adequate inorganic-N (sodium nitrate), whereas it decreased to ∼1:1.87 (8.7% chitosan: 16.3% chitin, re-calculated) when the fungus was grown in organic-N (peptone). Due to this ratio change with a concomitant decrease (about 21.8%, re-calculated) of Cr(III) biosorption, chitosan of M. hiemalis MP/92/3/4 cell walls is suggested to be involved in the biosorption of Cr(III) 11 . However, the increment of chitosan by ∼8.7% (re-calculated) is not in line with ~1.74% (re-calculated) decrease of chitin in the cell walls of inorganic-N grown M. hiemalis cells in comparison to the organic N-grown cells. Thus, chitosan cannot alone be responsible for the increment of biosorption by 27.91%, suggesting involvement of other functional groups, e.g. phosphate-group, in metal binding. In accordance with this interpretation, we found a strong P peak in aquatic M. hiemalis cell walls (Fig. 2B), where e. g. Al, Pb, Cd and Cr are adsorbed and precipitated as nanoparticles (Fig. 2B). The cell wall chemical composition of terrestrial M. hiemalis was variable, probably due to the influence of nutrition conditions, e.g. type of N-source 11 . Similarly, we also observed changes of cellular surface potential of aquatic M. hiemalis by varying nutritional conditions of growth medium (C-limitation; N-limitation; C-, and N-enriched medium; groundwater) (Fig. 2C). The zeta-potential of pure chitin showed enhancement of electro-negativity with increasing pH after pH ∼ 5.7 and reached the highest electronegativity ∼ −260 mV at pH ∼10.8 (data not shown). In contrast, the zeta potential of aquatic M. hiemalis spore surfaces at the lowest point is nearly double negative when compared to the zeta-potential of pure chitin, implying the involvement of other electronegative groups in the cell walls of aquatic M. hiemalis with twice the capacity for metal binding. The strong presence of a phosphorus peak at the location of precipitates of Al, Cd, Cr and Pb indicated high concentration of electronegative phosphate ligands for the complexation with positively charged metal cations at cell walls in aquatic M. hiemalis (Fig. 2B). Similarly, the deposition of Al and U cations into polyphosphate granules was also observed in the cell walls of Anabaena cylindrica 40 and in Citrobacter sp 41 . The metal precipitation at the cell walls of germinated spores could be catalyzed by acid phosphatase activity by increasing release of phosphate anion ligand for the complexation with metal cation and its precipitation as inorganic insoluble metal phosphate 2 . The concentration of the phosphate ligand may also exceed the solubility product of the metal phosphate in proximity to nucleation sites on the cell surface 41 . Similarly, it is possible in M. hiemalis by sequential biosorption and precipitation of Cr(III) 11 . Additionally, the functional groups of chitin, chitosan and other cell wall components as well as changes of molecular configurations and associated high electro-negativity of aquatic M. hiemalis cell surface might be involved in the cellular metal precipitation process (Fig. 2B-D), apparently via regulation of cell wall integrity and signalling pathways 39 . The binding of positively charged heavy metal cations to the fungal cell walls was previously suggested to be due to the electrostatic attractions 42 . Electro-negativity of the cell surface apparently results www.nature.com/scientificreports www.nature.com/scientificreports/ from the occurrence of negatively charged (dissociated) functional groups like carboxylic-, amino-, hydroxyl-and phosphate groups 42,43 . Hydroxyl-, amino-, acetyl-and carbonyl groups can be also found on chitin and chitosan of fungal cell walls, but additionally the presence of carboxyl groups on the fungal cell surface was suggested to be important for the biosorption of Hg 2+ (85.34%) and Co 2+ (73.05%) 44 .
EDX microanalysis demonstrated that the removed metals were deposited as nanoparticles (approx. 50-100 nm) in clusters on the surface of the aquatic M. hiemalis cell walls (Fig. 2B). Nanometer-to micrometer-sized particles were deposited in the natural microbiomes of the sulfidic springs, probably due to aggregations induced by microbial interactions with metal cations 45 (Fig. 3). Deposition of nanosized sphalerites (zinc sulfide) was shown to be mediated by the microbial community 36 . It was reported that microbes could change the orientation of metal ions in 3D space for the formation of nanospheres in the microbial community 45 . Here we demonstrate the use of a unique combination of EH8, EH10 and EH11 in producing diverse metal nanoparticles or nano-alloys, not only important for bioremediation, but perhaps also essential for e. g. biomedical (e.g. nano-contrasting substances in medical diagnosis), industrial (e.g. nano-catalysts) or biomining (e.g. selective enrichment) applications. Our discovery of this superior biotechnology for multiple nanometal particle and nano-alloy production is noteworthy as it supersedes biotechnologies focussed on single specialised fungi capable of producing only one particular metal nanoparticle 46 . For survival and additional resistance, the fungal microorganisms and microbiomes might have evolved the ability to interact their electronegative cell surfaces with toxic metal cations to produce non-toxic aggregates of nanospheres. SEM-EDX analysis showed intracellular accumulation of highly toxic ionic Hg and subsequent chemical reduction to metallic nanospheres by EH8's live (germinating elongating) sporangiospores. Fixing of intracellular Hg was inversely proportional to the diminishing residual Hg in water phase with time, following a generalized kinetics function of intracellular Hg accumulation independent of applied concentrations (Fig. 4A-C) 5 .
All of the three fungal materials (mixed microbiomes, spore mix, dead mixed insoluble cells/cell walls) of combinations of EH8, EH10 and EH11 share a mutual mechanism of metal biosorption. They mainly use cell surfaces for simultaneous hyper-accumulation and precipitation of twelve tested metal ions (Fig. 5D), apparently due to appearance of extra special chemical and physical properties in their cell walls by elicitation that could not be revealed by the toxicity/inhibitory tests (Fig. 1). As the cell wall surfaces play a great role, the dead cells/cell wall skeletons irrespective of preparation method (autoclaving/sterilization, solvent killing) mainly contribute to the removal of most of the metals. However, calculations using curve-fitting point to preferential binding of Cd, Co, Pb, Au, U and Ti by solvent killed and solvent washed dead cells (cell wall skeletons) in juxtaposition to other metal ions (Table 2), supporting the view that killing of fungus increases biosorption of some metals 47 . Our calculations also showed preferential fixing of Al, Cr, Ni and Ag by the live germination-activated spores. Contrastingly, the fixing of Cu, Pb and Zn beside Al was rather preferred by the live fungal microbiomes (Table 2). However, high metal removal values of all of the tested materials showed maximum 31% variations among them depending on metal ion and material used. Noteworthy, the three metals Au, Ag and Ti were fractionated and enriched only by the above mentioned growth combination of EH8, EH10 and EH11 by elicitation, i.e. they were not hyper-accumulated by any individual EH8 or EH10 or EH11 or by any combination not grown together. The underlying mechanism of gaining additional metal removal/enrichment functions by their combination of EH8, EH10 and EH12 can be mainly explained on the basis of elicitation/adaptation of cell wall composition and functions, whereby increment of surface display of pertinent functional (electronegative) groups and orientation via molecular, biochemical and physiological interactions can be assumed to take place during cell wall biosynthesis of germinating/elongating EH8, EH10 and EH11 sporangiospores in a mix. Thus, such changes apparently result in changes of ionic interactions of electronegative functional groups of cell surface with metal cations, complexation reactions, chemical reduction and acceleration of metal precipitation processes. Analogously, Kotrba et al. showed four times more binding of Cd 2+ , but Cu 2+ and Zn 2+ in less amounts, to cell surface by genetically modifying the surface display of short metal binding peptides on Escherichia coli 18 . On the contrary, our fungal technology using EH8, EH10 and EH11 covers a wide range of metals for binding. Inhibitory concentrations of copper or cobalt in the growth medium can also alter the cell wall's metal binding properties of Cunninghamella blakesleeana 48 . The diverse metal removal efficiency of mixed cell walls (EH8, EH10 and EH11) in the range of 67% to >99% of initial concentrations (Fig. 2C) was similar to that of the corresponding spore mix and microbiome, but much higher than that of the terrestrial control fungus M. hiemalis strain DSM2655, which removed only Cd (53%), Cr (90.5%), Pb (96%) and Zn (24.8%). The maximum biosorption capacity of Cr(III) by the cell wall of terrestrial M. hiemalis MP/92/3/4 was given as 132 mg/g dry-weight, whereas the whole cells of terrestrial M. hiemalis MP/92/3/4 reached a maximum of 21 mg/g dry-weight 11 . the future. Naturally occurring EH8, EH10 and EH11 simultaneously removed nine toxic metals (Al, Cd, Co, Cu, Cr, Ni, Pb, Zn) and thereby also enriched four precious metals (U, Ti, Ag, Au) (Fig. 6A-E), a result that was elusive to date. For the first time it was demonstrated that if the physiologically tolerant (compatible) fungi were grown together, their respective functions became biologically elicited (enhanced) and acquired additional properties of Ti, Ag and Au enrichment, not following the mathematical union of the three sets of metal accumulations by EH8, EH10 and EH11 individually. This mixed fungal technology may help in promoting cheap biotechnology for bioremediation, biofractionation and biomining of metals. A new bioremediation technology with its wide coverage of toxic metals at high concentrations without any genetic modifications or pre-treatment may have high competitive advantages in securing a significant part of the multi-billion dollar remediation and biomining market 8 . Landfill leachate, mine drainage and waste water with toxic metal mixes can also be decontaminated by using column fillings of mixed grown EH8, EH10 and EH11. A step-wise treatment of multimetal-contaminated water using insoluble EH8, EH10 and EH11 cell wall mix can even deliver drinking water (Fig. 5A,B), which is just one of the application areas in which the biotechnological potential of aquatic M.
www.nature.com/scientificreports www.nature.com/scientificreports/ hiemalis remains to be exploited. Even toxic Hg could be removed by live EH8 sporangiospores 5 . In contrast, most of the other microorganisms have so far failed to remediate a myriad of toxic metals simultaneously due to the extreme synergistic multimetal toxicity. As for example, simultaneous presence of Ni and Cd in solutions exerts toxicity to Clostridium thermoaceticum even at the low concentration of 1 mM 14 .
Some merits of this new biotechnology are worth mentioning: 1. Possibility of ex-situ and in-situ applications 2. Possible regeneration and recycling of biosorbents, e. g. by washing with 5 mM EDTA 48 , 3. Minimization of microbial waste by burning, 4. Possibility of multimetal nanospheres production for industrial and medical use, nearly independent of pH, e.g. only 10% efficiency loss was detected at pH 4 (Fig. 2B2, B2E), 5. Independence of growth from temperature variations in a wide range, enabling even an in situ treatment of a metal contaminated aquifer using fungal microbiomes on expanded clay down to near freezing temperature, 6. High safety assurance while using toxin-free fungal cell wall mixes for water purification, 7. Cheap biotechnology: even extracted fungal cell walls exhibited 2-3 times superior binding capacity for ionic metals compared to other materials used, thereby about 30% less cell wall biomass is required for the same efficiency of metal removal by using spore mixes or microbiomes. A calculation by Holan and Volesky 49 showed at least double the metal removal capacity of insoluble fungal cell walls compared to marine algae on the basis of dry-weight. 8. Biological elicitation by the fungal strains' mixed growth enhanced their ability to enrich precious metal ions more than the union of three sets of metal enrichments by the strain EH8, EH10 and EH11 individually, implying the biological set union has a system response greater than the corresponding mathematical set union. 9. In combination with Phanerochaete chrysosporium, aquatic M. hiemalis strains might also simultaneously remove many recalcitrant organic pollutants from their mixed contaminations with multiple metals in waste water [30][31][32] .
Last but not least, the presented results may inspire us to use these new strains EH8, EH10 and EH11 to concentrate and extract precious metals from oceans and seas at a low-cost, which was once tried in vain by the great German chemist F. Haber 50,51 . Later Necker also dreamt of extracting gold, silver and uranium from seas and oceans using emerging technology, but it was also found to be economically infeasible because of low concentrations (e.g. ~10 ng.L −1 gold and silver) 52 . A feasible low-cost method for biotechnological extraction of gold has awaited a break-through since the discovery that deposition of gold in ores was due to microbial activity 53 . Now we are on the verge of an economically feasible emerging biotechnology, simply showing nature's power in special microorganisms and microbiomes of selected sulfidic springs to remediate multimetal-contaminated water and even to extract and concentrate some precious ionic metals from extremely diluted solutions present in lakes and oceans (Fig. 6). The use of insoluble fungal dead cell-wall fractions avoids the eco-toxicological risk and conflict arising from current legislation restricting the use of live fungi, microbiomes and/or genetically modified/engineered microorganisms 8 for drinking water purification, bioremediation and/or biomining in the fields, as there is a risk of toxin release by live fungus. However, in presence of recalcitrant highly toxic organic substances mixed with toxic metals a combination of aquatic M. hiemalis with non-pathogenic P. chyrsosporium could be used 30 . The ability of mixed grown EH8, EH10 and EH11 to fractionate and enrich precious metal ions like U, Au, Ag and Ti from the aqueous phase opens a new possibility of simultaneous water purification, biofractionation and biomining of precious metals, even from highly diluted concentrations present in marine water. Thus, our results may be useful for the development of an efficient large-scale low-cost fungal biotechnology for the treatment of multimetal-contaminated industrial waste water, land-fill leachate and ground water 54,55 and/or for biomining 8 of valuable metals. toxicity/inhibition assay. The tests of strain toxicity/inhibition i.e. strain compatibility based on strain morphology were carried out on malt-extract agar plates 5 by inoculations of mycelium pieces 30,34 . Zeta-Potential of fungal spore surfaces. The surface charge (relative zeta-potential) of fungal spores at different pH was determined by measuring streaming current potential shift in the diffuse layer of fungal spore surfaces (electrostatic double layer model) after induction of a water stream by placing a current over the surface. For this purpose, the measuring electrode was fixed on a surface containing the fungal spores and the potential shift was detected using a particle charge detector Mütek PCD 03-pH (Mütek, Herrsching, Germany). Sporangiospores (2-5 ml spore suspensions) of aquatic M. hiemalis germinated in different nutrient liquid media (low-C; low-N; C-, and N-enriched media 35 ; water control) were diluted to give final mean concentrations of about 4.4 × 10 7 spores in 20 ml. The initial pH values of M. hiemalis spore dilutions were between 7.6 and 8.3. The www.nature.com/scientificreports www.nature.com/scientificreports/ titration of spore dilutions was started by using 0.1 M HClO 4 in the acid phase and then by using 0.1 M NaOH in the alkaline phase. All the measurements were duplicated.

Methods
Preparation of microbiomes, spore mix and dead insoluble mixed fungal cell walls. All the three strains, EH8, EH10 and EH11, were able to grow under aerobic conditions on malt extract-agar plates, although EH11 was also found to live under strictly anaerobic conditions in liquid cultures. The germinated spores of the strains were used as 1:1:1-mixture, as they were found to be physiologically compatible when challenged separately and against each other. Briefly, fungal sporangiospores (see results) from five malt extract-agar (2:1, wt/wt) plates of each strain were collected and enriched by filtration and centrifugation technique. The fungal spores were counted using a hemacytometer (Sigma) and image processing software (ProImage v.3.01, MicroMotion). Germination of sporangiospores (5×10 7 , EH8/EH10/EH11 = 1/1/1) was induced in 50 mL C-, N-enriched medium 35 or on expanded clay spheres at 30 °C under shaking at 120 rpm. Active germinated spores (1-3 cell stages) were purified by repeated washing with PBS (pH 7.4) and centrifugation cycles, and then absorptions (A 270 nm , A 650 nm ) of diluted suspensions were determined. Calibration curves for spore biomass (fresh-or dry-weight, mg) against absorption values (A 270 nm , A 650 nm ) were calculated by regression analysis, and used to determine spore biomass, as for example, y (mg, spore biomass dry-wt.) = 12.26 + 8.096 * A 270 nm , depending on absorbance. In contrast, microbiomes grown on expanded clay spheres in C-,N-enriched nutrient medium 35 using the same mixed spore count of EH8, EH10 and EH11 were repeatedly purified by washing with PBS (7.4) without centrifugation prior to their use for our study.
Insoluble dead mixed fungal cells with cell wall skeletons (cell wall materials, see below) were prepared following a protocol for preparation of plant cell walls (Bligh and Dyer) 56 with modifications. After aerobic germination of the three Mucor hiemalis strains, EH8, EH10 and EH11, (1: 1: 1, total 5×10 7 sporangiospores per 50 ml water) in associations (tri-cultures), the fungal sporangiospores (1−3 cell stage) were pelleted by centrifugation. The live pellets were chemically killed by incubation in Bligh-Dyer solution 56 for 4 h. After removal of supernatants, the insoluble dead cells with cell wall skeletons were successively washed with methanol-dichloromethane solution (3:1, v/v; 30 min), 1% SDS (12 h), 1 M NaCl (12 h), water (12 h) and acetone (30 min) to eliminate any remaining soluble cellular materials. The prepared dead insoluble fungal cell skeletons/cell walls were dried in a desiccator before use.
Metal Removal Assay. Spore mix, mixed microbiomes and insoluble dead cell walls of EH8, EH10 and EH11 (see below) were used for the metal removal assays. Pre-germinated EH8, EH10 and EH11 sporangiospores (5×10 7 spores, mixed in the 1:1:1 ratio) or mixed microbiomes or dead mixed cell walls (ca. 28 mg dry-weight) were suspended at least in triplicate in 50 ml ground water artificially contaminated either with a single heavy metal or with a heavy metal mix at a concentration of 1-50 mg/l, shaken continuously for 0-48 h at 120 rpm and 25 °C. After centrifugation at 4,000 × g (3 min) and after subsequent filtration (0.45 µm) and acidification (pH 1), with concentrated HCl or HNO 3 and/or in aqueous solutions, and finally digestion of pellets (spores) in concentrated HNO 3 and adjustment of pH to pH 1, the heavy metal contents in the supernatants were determined at element-specific wavelengths using an inductive coupled plasma detector (ICP, Model Liberty 200 A, Varian) as previously described 28 . The system was calibrated by using respectively 0, 1, 10, 100, 500 and 1000 µg/l heavy metal standards, whereby the calibration curves with 95% confidence intervals were calculated by linear regression analysis (r 2 = 0.99). The standard deviations of measurements (n = 3) at each data point were less than 5%. Analysis by energy-dispersive X-ray (EDX). After incubation with heavy metals, the fungal materials (spores) were separated by centrifugation, fixed with 1% glutaric aldehyde (15 min), then treated with 2% OsO 4 in phosphate-bufferd saline (PBS, pH 7.4), washed and dehydrated in an increasing ethanol gradient (50%, 80%, 100%) in water. The uncut samples were sprayed with nanogold particles prior to observations with scanning electron microscopy (SEM, model JSM 630 F, Jeol) at 5-15 kV in the secondary electron mode. The sites of metal adsorption and precipitations were observed by using the SEM's back scattering detector. A silicon peak detected is due to sample fixing on glass slides, traces of chloride and calcium peaks may also arise from the incubation medium. The elemental analysis of microbiomes, spores and cellwalls of EH8, EH10 and EH11 was carried out using eXL EDX-system (Oxford instruments). For this purpose, X-ray quanta were detected by a silicon detector, the data were analyzed using energy dispersive x-ray analysis software and the signals of element specific shells were resolved and annotated using Link-ISIS software 5 .
Assay of pH-stability of metal sorptions by fungal materials. The capacity of heavy metal adsorption and accumulation by microbiomes depending on pH was studied in the pH range 3-7. As some metal salts might precipitate under alkaline conditions due to the formation of insoluble hydroxides, the data taken at pH higher than ∼7.5 are not taken into account. The fungal spores or microbiomes were incubated in water of selected pH containing heavy metals of various concentrations for 48 h. The pH of incubation water was monitored before and after experiment.
Glutathione-S-transferase activity. The glutathione S-transferase activity of aquatic fungi was determined as described 28 . statistical analysis. The statistical analysis of data was performed using the Student's t-test and the Mann-Whitney U test 57 . The data for the curves were fitted by using the software Sigma Plot 8.0 for Windows (SPSS Inc., United States).
Identification. All the strains were isolated and identified by comparison of morphology, mating behavior and/or molecular biological sequencing 5,28 .