Strains S. thermosulfidooxidans DSM 9293 and L. ferriphilum DSM 14647 were purchased from Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Germany. The two strains were cultivated in Mackintosh (MAC) medium (Mackintosh, 1978). 0.02% yeast extract was supplemented in the medium for S. thermosulfidooxidans. 2% (m/v) pyrite grains or 1% sulfur prills or sulfur slices were added as energy source for S. thermosulfidooxidans. 4 g/L ferrous ion was added as energy source for L. ferriphilum. The initial pH of the medium was adjusted to 2.5 when the energy source was sulfur, while the pH was adjusted to 1.8 when the energy source was pyrite or ferrous ion. Both strains were cultivated on a rotary shaker at 140 rpm. Culturing temperature for S. thermosulfidooxidans and L. ferriphilum was 45 and 37°C, respectively.

All pyrite was museum grade in form of naturally crystalized cubes from Freiberg, Germany. Smooth and shiny pyrite cubes were used for the experiments. The cubes were cut by a diamond saw and slices with one shiny side and one rough side were obtained. Only the shiny side was used for biofilm formation. The remaining pyrite was ground and pyrite grains with a size of 50–100 μm were used for cultivation or bioleaching experiments. Pyrite grains and slices were cleaned and sterilized according to Schippers et al. (1996). Briefly, they were washed with 6 M boiling HCl for 30 min, then rinsed with deionized water till the pH was neutral. After washing with acetone for three times, they were dried at 80°C for 12 h and sterilized for 24 h at 120°C under a nitrogen atmosphere.

To produce sulfur prills, elemental sulfur powder was firstly molten in a glass beaker at 130°C and then poured into cold deionized water with an agitation of 250 rpm. Due to rapid cooling sulfur prills with a diameter of 2 mm were formed. To produce sulfur slices, molten sulfur was poured onto glass plates to obtain a sulfur layer after its solidification. Then sulfur slices with a size of 3 cm × 3 cm × 0.2 cm were obtained by manually cutting the sulfur layer. Both sulfur prills and slices were autoclaved at 110°C for 90 min.

Pyrite grains were first incubated at a pulp density of 2% (m/v) with 10⁸ cells/mL of iron-grown L. ferriphilum in 50 mL MAC medium with a pH of 1.8 for 1 day on a rotary shaker at 37°C. Afterward, the culture supernatants were discarded and the pyrite residues were washed three times with fresh MAC medium. After washing the pyrite residues were processed as following: Resuspended the pyrite residues in 50 mL fresh medium; Resuspended the pyrite residues in 50 mL fresh medium which contained 0.02% yeast extracts and 10⁸ cells/mL of sulfur-grown cells of S. thermosulfidooxidans; Incubated the pyrite residues at 100°C for 2 h to heat-inactivate the biofilms of L. ferriphilum. And after washing, the pyrite residues were resuspended in 50 mL fresh medium which contained 0.02% yeast extracts and 10⁸ cells/mL of sulfur-grown cells of S. thermosulfidooxidans. Two more sets of experiment were set: (1) 5 × 10⁷ cells/mL of sulfur-grown S. thermosulfidooxidans cells and 5 × 10⁷ cells/mL of iron-grown L. ferriphilum cells were inoculated simultaneously to 2% (m/v) freshly ground sterile pyrite; (2) 10⁸ cells/mL of S. thermosulfidooxidans were inoculated to 2% (m/v) sterile freshly ground pyrite. All the experiments were performed at 37°C on a rotary shaker at 140 rpm with an initial pH of 1.8. Each experiment was done in triplicate.

The pyrite grains with 1 day old living or inactivated biofilms of L. ferriphilum were prepared as described above. Then cell attachment of S. thermosulfidooxidans to the pyrite with living or heat-inactivated biofilms followed. Briefly, 10⁸ cells/mL of S. thermosulfidooxidans were inoculated into flasks filled with 50 mL MAC medium, pH 1.8, and with freshly ground sterile pyrite or with living or inactivated biofilms of L. ferriphilum. Then they were cultivated at 37°C with shaking at 140 rpm. The number of planktonic cells was determined by direct counting within 3 h. The amount of attached cells was calculated by subtracting the planktonic cells from the initial cell number. All the experiments were performed in triplicate.

The culture supernatants of L. ferriphilum were collected after 3 or 14 days of incubation with pyrite. Then they were filtered through a filter with a pore size of 0.22 μm. Afterward, pyrite-grown cells of S. thermosulfidooxidans were harvested by centrifugation and after washing with fresh MAC medium they were inoculated into the filtered leachate with a final concentration of 10⁸ cells/mL. 2% (m/v) freshly ground sterile pyrite were then added to all the flasks. Considering the importance of ferric ion for leaching, two batches of experiment were set as control, in which the same amount of iron as detected in the leachate was added to the fresh medium. One control got an inoculum of S. thermosulfidooxidans, named as biological control, and an assay without inoculum of S. thermosulfidooxidans functioned as chemical control. The added iron was ferric chloride. The experiments were performed on a rotary shaker at 140 rpm, 37°C. Data were collected from three parallel experiments.

The planktonic cell density was determined by direct counting with a Thoma counting chamber (Assistant, Germany) under a light microscope (Leica DMLS, Wetzlar GmbH) in phase contrast mode with a magnification of 400 times.

To determine the attached cells on the pyrite grains, samples were stained with SYTO9 which combines with the attached cells and gives fluorescent signals. Then an epifluorescence microscope (Axio Imager A1m, Zeiss, Germany) was used to take images of the pyrite grains. Afterward, the images were analyzed by platform Fiji according to Schindelin et al. (2012) to obtain the number of attached cells. Fiji identifies and enumerates the objects in the stack with a user-defined threshold for voxel intensity value. For counting first the value of the voxel intensity threshold was set manually by visual inspection of the images (Fiji: Image > Adjust > Threshold). Some incomplete cells, caused by the adjustment of the threshold, were then filled by function of fill holes (Fiji: Process > Binary > Fill Holes). To distinguish conjunct cells the watershed function was applied (Fiji: Process > Binary > Watershed). Afterward, the counting process was performed (Fiji: Analyze > Analyze Particles). Each cell number was collected from 25 images.

The pH was determined with a digital pH meter (Model pH 537, WTW, inLab 422 Combination Semi-micro pH Electrode, Mettler Toledo).

Ferrous and total iron ions were quantified by the procedures described by Tamura (Tamura et al., 1974). The principle is that 1,10-Phenanthroline complexes ferrous ions. The complex gives a red color and can be measured spectrophotometrically at 492 nm. Subsequently, the total iron can be quantified by adding hydroxylamine, which reduces ferric to ferrous ions. The samples were measured in triplicate within microtiter plates with a UV-Vis spectrophotometer (TECAN, Infinite pro^® 200).

A “BioMaterials Workstation” (JPK Instruments, Germany) composed of a NanoWizard II atomic force microscope (JPK Instruments) and an upright epifluorescence microscope (Axio Imager A1m, Zeiss, Germany) was applied to visualize the cells and the biofilms. A sample can be visualized at the same site between the two microscopes with 2 μm deviation (Mangold et al., 2008). In this study all samples were first stained by 6 μM SYTO9. The location was first chosen by EFM, then visualized by AFM. For AFM imaging, the contact mode was used. The applied setpoint was below 1 nN and the scan rate ranged from 0.1 to 0.5 Hz. A CSC38/NO AL (Mikromasch, Tallinn, Estonia) probe was used for imaging and cantilever B with the following parameters were chosen for scanning: length, 350 μm; width, 32.5 μm; thickness, 1.0 μm; resonance frequency, 10 kHz; shape, cone with a full cone angle of 40^o; and force constant, 0.03 N/m. Before the experiments the cantilevers were immersed in Piranha solution for 10 min and then washed with sterile MilliQ water. All AFM operations were performed in MAC medium at pH 2.5 at room temperature. The sensitivity of the optical lever system and the cantilever spring constant were calibrated in liquid prior to each experiment. The spring constant for each cantilever was compatible with the manufacturer’s specifications.

Development of biofilms on pyrite slices by a pure culture of S. thermosulfidooxidans or of L. ferriphilum within 14 days is shown in Figure 1 or Figure 2, respectively. Figure 1A clearly indicates that after 1 day incubation cells of S. thermosulfidooxidans had attached on the pyrite slices either in the form of single cells or flocs. However, the attached cells detached gradually and almost no cells could be observed after 16 days of incubation (Figure 1C). It has been reported that cells of S. thermosulfidooxidans are unable to form stable biofilms on pyrite, if a routine culturing procedure is used. The cells are non-motile, but they prefer to colonize the defect sites of mineral surfaces (Sand et al., 2001). Also the culturing environment deteriorates as the experiment proceeds (e.g., decrease of organic compounds and increase of metabolic end products), which might stimulate formation of endospore (Li et al., 2016). In the assays with L. ferriphilum after 1 day of incubation most cells had attached on the surface of the pyrite slices in the form of single cells (Figure 2A). The cells were observed to grow in colonies after 7 days of incubation (Figure 2B). More robust biofilms of L. ferriphilum were observed after 16 days of incubation (Figure 2C).

Figure 3 demonstrates that the number of attached cells of S. thermosulfidooxidans on pyrite changes as a function of time. The pyrite was pre-colonized with 1 day old living or inactivated biofilms of L. ferriphilum. In the sterile control after 3 h about 27% of the inoculated cells of S. thermosulfidooxidans had attached on the pyrite. Compared with the sterile control, the percentage of attached cells of S. thermosulfidooxidans increased by around 5%, if the pyrite had been pre-colonized with living cells of L. ferriphilum. However, it decreased by approximately 5%, if the pyrite had been pre-colonized with inactivated cells of L. ferriphilum. It has been reported that bacterial adhesion is influenced by many factors, including surface roughness, type of mineral, hydrophobicity and medium (Dexter et al., 1975; Absolom et al., 1983; Helke et al., 1993; Quirynen et al., 1993; Blackman and Frank, 1996). It is likely that a conditioning film formed on the pyrite surface by excretions of the cells of L. ferriphilum like EPS compounds, which changed surface properties of the pyrite surfaces. This influenced the adhesion of cells of S. thermosulfidooxidans. However, in the assay with the inactivated biofilm of L. ferriphilum the conditioning film and the inactivated cells of L. ferriphilum might have been destroyed or deteriorated and blocked the attaching sites. Consequently, the attached number of cells of S. thermosulfidooxidans decreased.

Figure 4 illustrates biofilm formation on sterile pyrite by a binary culture of S. thermosulfidooxidans with L. ferriphilum within 16 days. The two species were simultaneously inoculated. Figures 5, 6 show the biofilm formation by the culture of S. thermosulfidooxidans on pyrite pre-colonized with living or inactivated biofilms of L. ferriphilum, respectively. It is visible that the cells of S. thermosulfidooxidans are long rod-shaped, while the cells of L. ferriphilum are vibrioid and/or round-shaped and small. The two cell forms can be distinguished well with the microscope. In all three assays after 16 days incubation the biofilms of S. thermosulfidooxidans remained on the pyrite. It indicates that S. thermosulfidooxidans can form biofilms on pyrite in the presence of L. ferriphilum. However, the architecture of the biofilm is different. Biofilms of S. thermosulfidooxidans in the form of single cells are observed on the pyrite pre-colonized with living biofilms of L. ferriphilum (Figure 5). In the other two assays biofilms of S. thermosulfidooxidans grow in the form of colonies with biofilms of L. ferriphilum (Figures 4, 6). In all assays a physical contact between the cells of S. thermosulfidooxidans and L. ferriphilum can be observed, as is visible in the images of EFM and AFM in Figure 7.

A physical contact has been reported also for other microorganisms and it happens during microbial coaggregation. Coaggregation, the process of adhesion between genetically distinct bacterial partners, was first found between bacteria isolated from human dental plaque in 1970 (Gibbons and Nygaard, 1970). It is an integral process in the development of multispecies biofilms and is caused by specific ligand receptor interactions (Rickard et al., 2003). In a dual-species coaggregation one of the species will benefit from being within a coaggregation, while the other may also gain or not gain advantages as compared to planktonic cells. Katharios-Lanwermeyer et al. (2014) reviewed coaggregation mediated interactions and concluded five potential outcomes for biofilm communities: (1) The interspecies can rapidly form aggregates because of the enhanced cell-cell adhesion; (2) Species within the coaggregation can exchange of metabolites; (3) The coaggregation can aid one species to kill/inactivate another or to protect both species from predation; (4) The coaggregation reduces the distance between cells, thereby facilitating cell-cell signaling; (5) The coaggregation can protect cells from adverse events, enhance their antimicrobial resistance and genetic exchange.

In this study, obviously, coaggregation enhanced the adhesion of S. thermosulfidooxidans. However, how the two species developed the coaggregation or whether it was also caused by specific ligand receptor interactions, and whether this coaggregation could facilitate metabolite exchange or cell-cell signaling remains unclear. More delicate experiments need to be done to answer these questions.

For an improved understanding the assays are numbered as follows: (1) S. thermosulfidooxidans incubated with pyrite pre-colonized with 1 day old living biofilms of L. ferriphilum; (2) S. thermosulfidooxidans incubated with pyrite pre-colonized with 1 day old inactivated biofilms of L. ferriphilum; (3) S. thermosulfidooxidans and L. ferriphilum were simultaneously inoculated into assays with sterile pyrite; (4) pyrite colonized with 1 day old biofilms of L. ferriphilum (without planktonic cells); (5) S. thermosulfidooxidans incubated with sterile pyrite. Data for an assay with sterile pyrite are not shown, since no dissolution was detected.

The leaching of pyrite by S. thermosulfidooxidans in the presence of L. ferriphilum was calculated and the results are shown in Figure 8. In all assays the pH decreased, which indicates that bioleaching occurred. Since L. ferriphilum can oxidize only ferrous ion for energy but S. thermosulfidooxidans can use both, ferrous ion and RISCs, as energy source, the pH in assay 4 decreased the least. The pH decreased the most in assays 1 and 3, but in the first 14 days the pH in assay 3 was lower than it in assay 1. After 16 days of leaching the highest iron concentration was detected in assay 4 with a final value of 1.5 g/L, followed by the one in assay 3 with 1 g/L. There were 0.5 g/L and 0.3 g/L iron detected in assay 1 and assay 2, respectively. The data indicate that with biofilms of L. ferriphilum at the beginning of leaching, no matter whether living or inactivated, these biofilms exerted a negative effect on leaching by S. thermosulfidooxidans.

It is known that leaching bacteria prefer to colonize the crystal defect sites on a mineral surface and only attached cells can initiate bioleaching (Bellenberg et al., 2015). Obviously the inactivated biofilms of L. ferriphilum in assay 2 (Figure 8) could not perform leaching and to some extent they protected the pyrite surfaces from being attacked by ferric ions. Also the sites for attachment of S. thermosulfidooxidans were blocked. Consequently, the leaching needed longer time to start and the amount of leached iron was the least among all assays (Figure 8). For living biofilms of L. ferriphilum the introduction of S. thermosulfidooxidans might mean introducing a competitor for iron, which can be indicated by the higher iron concentration in assays 4 and 5 (single culture) than in assay 1. Interestingly, also with living L. ferriphilum the pH in assay 3 was much lower and the concentration of iron ion was much higher than in assay 1 (Figure 8). An increased concentration of iron ions mean dissolution of pyrite and a decreased pH means oxidation of RISCs. S. thermosulfidooxidans is the only sulfur-oxidizer in the system. Thus, most likely in assay 1 S. thermosulfidooxidans grew mainly on ferrous ion and the competition led to the bad leaching performance. In assay 3 the decreased pH illustrated the oxidation of RISCs by S. thermosulfidooxidans, but their worse leaching performance than that of the pure culture of L. ferriphilum indicated a negative effect on growth of L. ferriphilum. A possible explanation is that in assay 3 S. thermosulfidooxidans grew on both ferrous ion and RISCs.

Based on the results above two points can be highlighted:

The effects of spent leach liquor from L. ferriphilum on pyrite leaching by S. thermosulfidooxidans was also investigated. In order to reproduce a natural leaching system yeast extract was not added in this section. Firstly, bioleaching of pyrite by S. thermosulfidooxidans without addition of yeast extract was tested and the results are shown in Figure 9. From Figure 9A it can be seen that in the first 12 days the iron concentration increased from 0 to 260 mg/L. Afterward it almost leveled off at 300 mg/L. The iron ions were detected mostly in the form of Fe(III) in the first 9 days. However, from 12 days on Fe(II) became gradually the dominant form. At the end of the experiment the iron ions were almost all in the form of Fe(II). Data in Figure 9B indicate that the pH of the leaching system decreased rapidly from the 6th day on till the 15th day when it did not change any more. The numbers of planktonic cells decreased, from the initial cell density of 1.1 × 10⁸ cells/mL to the final cell density of 4 × 10⁶ cells/mL. The number of attached cells was also monitored. The bar chart (Figure 9C) indicates that the number of attached cells increased during the leaching and, finally, the pyrite surfaces were covered with 5 × 10⁶ cells/cm² (Figures 9D,E). Our data clearly indicate that after around 14 days of incubation the bioleaching stopped and the cells of S. thermosulfidooxidans started to adhere to pyrite surfaces. It remains, however, unclear, why these events occurred. A similar phenomenon has been reported, where adhesion of Acidithiobacillus ferrooxidans to pyrite was enhanced if no phosphate was added to the growth medium (Bellenberg et al., 2012). In our study lack of nutrition, such as yeast extract, might account for the observed events.

Bioleaching of pyrite by S. thermosulfidooxidans with pyrite leachate from L. ferriphilum was then investigated. Figure 10 shows, how each parameter changes during bioleaching, if the bioleaching was performed in 3 days or 14 days old pyrite leachate from L. ferriphilum. Although the leachates were of different age, the results of the three assays were similar: (1) The bioleaching ran better in the biological control than with the leachate; (2) A large amount of Fe(II) was detected in all leaching assays at the end of the experiment (Figures 10A,C); (3) Except for the chemical control the pH in all leaching assays decreased and the planktonic cell densities of S. thermosulfidooxidans also decreased (Figures 10B,D); (4) The number of attached cells increased in all leaching assays (Figure 10E).

There are also differences: (1) The age of the leachate influenced the time until the iron concentration started to differ. Specifically, if S. thermosulfidooxidans was incubated with 3 days old leachate, only after 12 days the total iron concentration became lower than in the biological control (Figure 10A). If the cells were incubated in 14 days old leachate, the concentration of total iron differed between the test assays and the biological control from the 3rd day on (Figure 10C). Also due to the age of the leachate the percentage of Fe(II) was different. With 3 days old leachate after 21 days of bioleaching the Fe(II) concentration amounted to 76%. In the biological control the percentage of Fe(II) was 52% (Figure 10A). If the leachate was 14 days old, the assays after 21 days contained 53% of Fe(II), whereas 41% of Fe(II) was measured in the biological control (Figure 10C). (2) The time, until a decrease of the pH occurred, differed also. In the assays with 3 days old leachate the pH started to decrease from the 3rd day on in both assays (Figure 10B). In the assays with 14 days old leachate the pH in both groups decreased right from the start (Figure 10D). (3) For adhesion differences were noted. In the assays with 3 days old leachate in all assays from the 14th day on the cells were detected mostly attached to pyrite surfaces, while in the assays with 14 days old leachate from the very beginning the cells of both assays were detected largely attached on pyrite surfaces. However, in the assays with 14 days old leachate the numbers of attached cells were always lower than it in the biological control, which should be related to the organic products produced by L. ferriphilum (Figure 10E).

From these results three points can be concluded: