Background

Because of global modernization and industrialization, pollution caused by the release of heavy metals into the environment has become a critical issue, and thus concerns surrounding the recovery of such heavy metals have been rising (Dodson et al. 2015). Among all metals commonly seen in wastes, gold is a noble metal and can be used in luxury jewelry, electronics, and medical applications because of its unique physical and chemical properties such as high biocompatibility and long-term stability (Ramesh et al. 2008; Spitzer and Bertazzoli 2004). Gold ions fabricated in nanoscale (gold nanoparticles or Au@NPs) with shape-dependent and optoelectronic properties have broadly applicable physiochemical characteristics and biological functions, and are thus of great interest to scientists (Klaus et al. 2001; Shedbalkar et al. 2014; Suresh et al. 2011).

Many traditional methods of gold recovery, such as cyanide leaching, precipitation and filtration, and electrochemical treatments, have been reported (Mata et al. 2009). However, these methods are challenged by restricted selectivity and large amounts of toxic chemicals involved in the process resulting in secondary pollution (He et al. 2015). Compared with traditional methods, biological methods have great appeal because of their simplicity, elimination of toxic chemicals (Mishra et al. 2014), and capability of controlling nanoparticle size (Bai et al. 2009; Sathishkumar et al. 2010a). Over the last decade, adsorption and reduction of heavy metals by microorganisms including bacteria, yeast, and fungi have been reported (Farooq et al. 2010; Lo et al. 2014; Tan et al. 2017). The biofabrication of nanoparticles with optimal conditions, such as pH, incubation time, and metal ion concentration (Narayanan and Sakthivel 2010) also showed high potential in commercial applications and large-scale production.

Shewanella species, which are dissimilatory metal-reducing bacteria widely distributed in sediment or seawater, grow optimally between 25 and 40 °C and can reduce metal ions via a special electron pathway called the Mtr pathway (Fredrickson et al. 2008; Shi et al. 2012; Wang et al. 2017). Shewanella oneidensis MR-1, isolated from Oneida Lake, New York (Myers and Nealson 1988), has been reported to reduce Au3+ into discrete spherical Au@NPs that are well dispersed with homogeneous sizes. The nanoparticle sizes of gold produced by MR-1 were in the range of 2–50 nm and had high potential for use in biomaterials (Suresh et al. 2011). In the MR-1 strain, the outer membrane c-type cytochromes, MtrC and OmcA, were considered as important proteins for metal reduction (Wu et al. 2013a). However, the MR-1 wild type and its mutants omcA and mtrC were still capable of reducing Au3+ into Au@NPs (Wu et al. 2013b), while the particle size of the extracellular nanoparticles were decreased mtrC and omcA mutants (Ng et al. 2013). On the other hand, Shewanella xiamenensis, which is isolated from coastal sediment collected off Xiamen, China, is a close-related strain to S. oneidensis based on phylogenetic tree analysis by 16S rRNA and gyrase gene (Huang et al. 2010). It has also been reported to reduce mediators from the medium to nanoparticles (Ng et al. 2015a) and also showed resistance to different kinds of metal ions (Ng et al. 2015b).

The mechanism and conditional optimization of Au@NP production by Shewanella have never been reported. The aim of this study is to explore the mechanism and to accomplish optimization by determining the effects of pH, biomass, and gold concentration on Au@NP production by S. oneidensis MR-1 and S. xiamenensis BC01 (SXM). Finally, the resting cell activity on the biofabrication of Au@NPs and silver nanoparticles (Ag@NPs) is also examined.

Methods

Chemicals

For the reduction of gold and silver ions, chloroauric acid (HAuCl4·3H2O) was purchased from Alfa Aesar, silver nitrate (AgNO3) was purchased from Sigma (209139), and sodium lactate (50% w/w) was purchased from Showa (G1510E). For scanning electron microscopy, formvar solution was purchased from Sigma (09823), and tert-butanol was purchased from Shimakyu Chemical Co. Ltd.

Bacteria culture

Both S. xiamenensis BC01 (SXM) and S. oneidensis MR-1 were grown in Luria–Bertani broth containing yeast extract (5 g/L), sodium chloride (10 g/L), and tryptone (10 g/L). The cells were maintained at 4 °C on LB plates, and a single colony was inoculated into 2 mL of LB medium and cultured at 30 °C and 150 rpm for 12 h for preculture. Then, 1% (v/v) of the precultured cells was transferred into a 50-mL flask containing 10 mL of LB medium and grown aerobically at 30 °C and 150 rpm for another 12 h.

Biofabrication and characterization of nanoparticles

Cells were collected by centrifugation at 8000×g for 5 min and washed twice with 0.5 mL of distilled water to obtain a final biomass concentration of 2.4 g/L. The precipitate was resuspend in 50 mM sodium lactate as the electron donor with 0.5 mL of 300 ppm Au3+ or 100 ppm Ag+ solution. The samples were incubated at room temperature (approximately 30 °C) with a light intensity of 100 µmol/m2/s for 24 h.

UV–vis spectroscopy analysis of nanoparticle formation

The nanoparticles were analyzed using UV–vis spectroscopy (Molecular Devices, SpectraMax 340PC384, USA) to measure the surface plasmon resonance at 530 nm for Au@NPs and 410 nm for Ag@NPs. The color of the sample changed from pale yellow to purple to indicate the formation of Au@NPs (Kumar et al. 2008). For Ag@NPs, the color changed from pale yellow to orange.

Scanning electron microscopy (SEM)

The samples were fixed in 2.5% (w/v) glutaraldehyde for 2 h and washed three times with phosphate buffer (0.1 M, pH 7.4). The sample (100 μL) was carefully dropped onto a formvar-coated silicon chip for 1 h and washed three times with phosphate buffer. The cells were dehydrated in a series of ethanol washes with increasing ethanol concentration (30, 50, 70, and 100%). After three final washes in 100% ethanol, the samples were immersed in tert-butanol and dried by lyophilization (KINGMECH, FD3-12P, Taiwan) for 0.5 h. Dehydrated samples were analyzed using SEM (JEOL JSM-6700F, Japan).

Inductively coupled plasma-optical emission spectrometry (ICP-OES)

Cells were centrifuged at 10,000 rpm for 10 min, and the supernatant was filtered using a 0.22-µm filter (Millipore, USA). Au3+ concentration was measured using ICP-OES (ULTIMA 2000, Japan). A standard solution containing 1000 ppm Au3+ (High-Purity Standards, USA) or 200 ppm Ag+ was used as the starting solution and diluted in the range from 0 to 100 ppm for ICP-OES analysis.

Zeta potential

Samples for zeta potential measurement were prepared as follows. Cells were added into 1 mL of distilled water in Falcon tubes with a final biomass concentration of 1.2 g/L. Different samples with a volume of 750 µL were introduced into cuvettes, and the zeta potential was measured (Malvern, Zetasizer Nano ZS, UK).

Preparation of resting cells

Strains were grown aerobically at 30 °C and 150 rpm in 250-mL flasks containing 50 mL of Luria–Bertani medium for 12 h. Cells were collected by centrifugation at 8000×g for 10 min and washed twice by distilled water. The precipitate was resuspended by 1 mL of distilled water, and the sample was quickly frozen in liquid nitrogen. The resting cells were generated by overnight lyophilization into a powder and stored at − 20 °C for long-term storage.

Results and discussion

Effect of pH

There have been reports on the biosynthesis of metal nanoparticles using Shewanella species (Ng et al. 2013; Suresh et al. 2011), but the physical factors for optimization have rarely been reported. Therefore, it is critical and meaningful to explore the optimal conditions of Au@NP biofabrication by SXM and MR-1. By adjusting the pH to 3, 4, 5, and 6, Shewanella produced different quantities of Au@NPs. It was obvious that SXM produced more Au@NPs than MR-1 did, as shown in Table 1. SXM produced 116 and 108 ppm Au@NPs at pH 4 and 5, respectively, while MR-1 only produced 62 ppm Au@NPs at pH 5. The production of Au@NPs by both strains decreased dramatically at pH 6. The value of pH was considered an important factor for gold reduction and nanoparticle formation (Mishra et al. 2012), as the different pH values affect the zeta potential, thus influenced the electric properties on the cell surface to form nanoparticles. The effect of pH was consistent with the results of S. haliotis, which has an optimal pH of 5 (Zhu et al. 2016). In addition, neutral conditions were not appropriate for Shewanella for the production of Au@NPs.

Table 1 The effect of pH on Shewanella at biomass of a 0.6 g/L to produce Au@NP with 300 ppm Au3+ after 24 h
Full size table

Effect of biomass and concentration of gold ions

As shown in Fig. 1a, the dark purple colors indicated Au@NPs formation by both Shewanella. As shown in Fig. 1b, the gold ions were adsorbed onto the cell surface when the biomass increased, and the concentration of gold dropped from 300 to 3 ppm with 6 g/L biomass. As a result, the residual concentration of Au decreased significantly when 3.6 g/L biomass was used, and residual Au3+ was relatively low (i.e., < 5 ppm) in both strains. However, the quantity of Au@NPs did not increase when the biomass increased from 2.4 to 6.0 g/L (Fig. 1b, bar graph). Maximal amounts of Au@NPs were generated at 2.4 g/L biomass (108 and 58.9 ppm for SXM and MR-1, respectively). The morphology of Shewanella in different biomass concentrations and the amounts of Au@NPs formed could be further analyzed by SEM as shown in Fig. 2. It was evident that Au@NPs were fabricated on the outer membrane of cells at 2.4 and 3.6 g/L biomass. When the biomass increased to 6.0 g/L, the cell membrane became thicker and sticky (shown in red rectangle in Fig. 2e). Although gold ions would adsorb on the cell surface, it was not beneficial to generate Au@NPs through nucleation layer by layer when biomass increased up to 3.6 g/L.

Fig. 1
figure 1

a The image of Au@NP production by SXM and MR-1. b The effect of Shewanella biomass on Au@NP formation when immersed in 300 ppm Au3+ after 24 h. Left axis, bar: Au@NPs. Right axis, dot: residual Au3+ concentration

Full size image
Fig. 2
figure 2

SEM analysis of SXM at biomass of a 0.6 g/L, c 2.4 g/L, and e 6.0 g/L, and of MR-1 at b 0.6 g/L, d 2.4 g/L, and f 6.0 g/L when immersed in 300 ppm Au3+ for 24 h

Full size image

The effect of gold concentration on Au@NP formation was tested within the range of 10–300 ppm Au3+. As shown in Table 2, no Au@NPs were produced at 0.6 and 2.4 g/L biomass when the concentration of gold ions was 50 ppm. Even when the concentration of gold ions increased to 100 ppm, only approximately 5 ppm Au@NPs were generated in (i.e., 5% conversion). When the gold concentration was greater than 100 ppm, Shewanella produced Au@NPs that proportionally increased when the Au3+ increased from 100 to 300 ppm. As shown in Table 2, it was obvious that SXM produced more Au@NPs than MR-1 with the increasing amount of Au3+. Moreover, the threshold of gold concentration for Shewanella to produce Au@NPs is reported here for the first time.

Table 2 The effect of gold ion concentration on Shewanella in Au@NP formation after 24 h
Full size table

A comparison of Au@NP formation using other microbes is shown in Table 3. The optimal biomasses of S. haliotis, Aspergillus oryzae var. viridis, and Sargassum sp. were 5.3, 10, and 5 g/L, respectively (Binupriya et al. 2010; Sathishkumar et al. 2010b; Zhu et al. 2016). Herein, the optimal biomass of SXM and MR-1 was 2.4 g/L, which was much lower than those of the other bacteria. Moreover, the formation of Au@NPs occurred within 4 h, which was much faster than those shown in previous reports. The absorbance intensities of SXM and MR-1, corresponding to the amounts of Au@NPs, were significantly higher than those of the other species. Because of the lower biomass used, higher reaction rates, and higher productivity, MR-1 and SXM have great advantages in gold reduction to nanoparticles.

Table 3 Comparison of different strains in biofabrication of Au@NPs
Full size table

Photo effect on Au@NP formation

As shown in Fig. 3a, the color of the solution changed rapidly within 4 h from pale yellow to purple when exposed to light at an intensity of 100 µmol photons/m2/s, indicating that the reaction rate of Shewanella in the presence of light was far greater than that of the solution in darkness. We also compared the wild type Shewanella (i.e., MR-1 and SXM) and genetically modified strains (i.e., cells harboring mtrC and mtrCAB genes). As shown in Fig. 3a, cells placed in light produced more Au@NPs in the time series from 0 to 24 h. Temperature was not a critical factor for Au@NP formation as shown in Fig. 3b. Very low levels of Au@NPs were produced at 25 and 37 °C without light induction. In addition, the MR-1 strains harboring mtrC or mtrCAB genes showed the same levels of nanoparticle production under light induction, which implies that the Mtr pathway proteins are not stimulated by light. It is known that cyanobacteria possess light-sensitive phytochromes to control photosynthesis, phototaxis, and production of pigments (Schmitz et al. 2000; Yeh et al. 1997). A phytochrome is a two-component system, with a membrane-bound sensor protein and an intracellular response regulator protein, which function in sequence in response to an extracellular signal. An E. coli strain has been artificially engineered to respond to light by replacing the osmolarity sensing domain EnvZ to a photosensing domain from the cyanobacteria in the native EnvZ–OmpR two-component system (Levskaya et al. 2005). The engineered E. coli strain was photosensitive and could turn on or shut down the expression of the reporter gene according to illumination. Therefore, we hypothesize that photoinduction in Shewanella which caused the formation of Au@NPs could be regulated by a two-component protein system. In our experiments, the formation of Au@NPs was significantly increased by photoinduction. From the genomic database of MR-1 (Accession Number NC004347), a putative two-component system with a photoreactivation-associated protein (PhrA) and photolyase (PhrB) could be involved in sensing of light. These light-sensitive proteins may be the key factor in stimulating the formation of Au@NPs. The search for key proteins by genetic method, i.e., knock-out phrA and phrB, is an inevitable further research motive for this study. In the past, the S. algae strain BRY was found to reduce Au3+ to Au using hydrogen as the electron donor (Kashefi et al. 2001). Alternatively, Shewanella species have been considered for the bioremediation of different kinds of metals ions or for use in microbial fuel cells (Chen et al. 2015; Liu and Logan 2004; Xu et al. 2006). This is the first attempt to discover the effects of temperature and light on the formation of Au@NPs.

Fig. 3
figure 3

Effect of light on Shewanella in Au@NP formation when immersed in 300 ppm Au3+. a Photos of Shewanella placed under light (top) and in the dark (bottom) within 24 h. b Amount of Au@NPs produced by Shewanella at 25 °C, light effect, and 37 °C for 24 h

Full size image

The proposed mechanism of Au@NP formation

The zeta potential analysis is shown in Fig. 4a. Significant differences in zeta potential were observed at biomasses of 2.4 and 1.2 g/L for SXM and MR-1, respectively. The difference in zeta potential gradually decreased when the biomass was greater than 3.6 g/L. At 6.0 g/L biomass, no difference in zeta potential was detected. This shows the disadvantage of high biomass (6.0 g/L) for Au@NP formation. Recent research shows that the changes in zeta potential of MR-1 not only reflect on the production of Au@NPs, but also the formation of biofilm (Ishiki et al. 2017). Therefore, with higher cell density, increased biofilm formation could affect the nucleation of gold atoms to NPs. On the other hand, the ICP-OES analysis of residual ion concentration after immersion of SXM or MR-1 in 300 ppm gold solution or 100 ppm silver solution with biomass concentrations ranging from 2.4 to 6 g/L, we found that only 20.8 to 3.29 ppm of residual gold ions or 2.26 to 0.81 ppm of residual silver ions were present for both SXM and MR-1 at lower or higher biomass (Table 4). Our hypothesis for the mechanism of Au@NPs formation by Shewanella is shown in Fig. 4b. Sodium lactate acted as the electron donor at a final concentration of 50 mM. Then, Shewanella adsorbed the gold ions, and nucleation was triggered on the cell membrane layer by layer, while higher biomass blocked the nucleation. Because of the distinct electron pathway of Shewanella, gold ions were reduced on the outer membrane of the cells, further accomplishing the process of nanoparticle fabrication and resulted in different zeta potentials on cell.

Fig. 4
figure 4

a Zeta potential analysis of S. oneidensis MR-1 and S. xiamenensis BC01 before and after the formation of Au@NPs. b Proposed mechanism of Au@NP production using Shewanella

Full size image
Table 4 ICP-OES analysis of residual ion concentration after immersion in Au3+ or Ag+ by SXM or MR-1 at different biomasses for 24 h
Full size table

In the formation of Au@NPs, the first step is the reduction of gold ions to metallic gold atoms. This process is supposed to be the common reduction, where the electron donor lactate goes through the Mtr pathway (i.e., mtrA, mtrB, mtrC, and omcA) and excess electrons enter the cytoplasmic membrane-anchored tetrahaem c-type cytochrome CymA to accomplish the reduction. The importance of proteins in Mtr pathway and cytochrome c-type has also been demonstrated in DMSO or dinitrotoluene reduction (Coursolle and Gralnick 2010; Liu et al. 2017). Moreover, as Au@NPs formation by Shewanella was photoinduced, there is the supposed involvement of a prospective two-component system [photoreactivation-associated protein (PhrA) and photolyase (PhrB)] which was activated by light energy to drive electron transfer and accelerate the reduction (Ng et al. 2000; Sancar 2003). Second, the nucleation of gold atom as a nanoparticle on cell surface should be a “layer by layer” processing, obeying thermodynamics and kinetic behavior. Thus, with higher cell density, the thicker biofilm formed would be a drawback for nanoparticles formation.

The activity of resting cells

Apart from the fresh cells, resting cells were also used for Au@NP formation. Resting SXM and MR-1 cells were prepared based on the method described in “Preparation of resting cells”. They were stored at − 20 °C for long-term conservation as shown in Fig. 5a. The ability of resting cells to produce Au@NPs was analyzed after 1, 15, 20, and 25 days. The resting cells had nearly 60% capability for Au@NPs production after 25 days (Fig. 5b), the variations in the activity could be attributed to the nonuniform cell powder. The resting cells were preferable as they could be used at any time without several repeated steps. Although there was a report showing that resting S. algae cells generated reduced amounts of platinum nanoparticles (Konishi et al. 2007), this is also the first attempt in showing that resting Shewanella cells possess long-term stability for the fabrication of Au@NPs.

Fig. 5
figure 5

a Resting Shewanella cells. b Biofabrication of Au@NPs by resting cells after 1, 15, 20, and 25 days

Full size image

The selectivity of gold and silver ions by Shewanella

The ability for Ag@NPs formation by SXM and MR-1 was shown in the blue line shown in Fig. 6. The Ag@NPs exhibited surface plasmon resonance at 410 nm. In order to confirm the selectivity of Au3+ and Ag+ ions by Shewanella, both strains were examined in a solution containing 100 ppm Au3+ and 100 ppm Ag+ ions (red line in Fig. 6) or 250 ppm Au3+ and 100 ppm Ag+ ions (green line in Fig. 6) for 24 h. The results showed that both SXM (Fig. 6a) and MR-1 (Fig. 6b) reduced silver ions at the same concentration, and reduced Au3+ when immersed in 250 ppm Au3+ and 100 ppm Ag+ ions. The selectivity can also be controlled by the concentration of ions inside the system.

Fig. 6
figure 6

Selectivity of a SXM and b MR-1 immersed in only 300 ppm Au3+ (pink), only 100 ppm Ag+ (blue), 100 ppm Au3+ mixed with 100 ppm Ag+ (red), or 250 ppm Au3+ mixed with 100 ppm Ag+ (green)

Full size image

Conclusions

Shewanella xiamenensis BC01 and MR-1 reduced Au3+ to Au@NPs, which were localized on the surface. By measuring the optimal condition, the highest amounts of Au@NPs were 108 ppm for SXM and 62 ppm for MR-1, respectively. It was found that an increase in biomass resulted in a decrease in Au@NPs. The presence of light dramatically accelerated nanoparticle formation. The mechanism of Au@NP formation and light-induced effect in Shewanella have been reported for the first time. The recovery of Au ions from industrial waste via Shewanella is a potential bioremediation option.