Introduction

In the last two decades, the alternative of using bioreductive deposition of precious metals, such as platinum-group metals (PGMs; e.g., Pt, Pd, Rh), for their recovery, has been widely explored. Special interest has been focused on palladium (Pd) due to its high value and extensive use as catalyst. Fe(III) reducing bacteria (IRB), such as Shewanella oneidensis (De Windt et al. 2006) and Geobacter sulfurreducens (Yates et al. 2013; Pat-Espadas et al. 2013), as well as sulfate reducing bacteria (SRB), such as Desulfovibrio desulfuricans (Lloyd et al. 1998), have extensively been explored for this purpose. Despite the great demand for developing efficient microbial processes to recover this valuable element, the mechanisms involved in Pd(II) reduction and subsequent deposition of Pd(0) are poorly understood.

Extracellular formation of Pd(0) nanoparticles (NPs) has been reported in G. sulfurreducens (Pat-Espadas et al. 2013). Moreover, differences regarding location of deposited NPs, depending on the strain, have been documented (Lloyd et al. 1998). These findings could be related to specific mechanisms used by each strain, as well as to experimental conditions prevailing. However, further studies are required to fully elucidate the mechanisms involved.

Geobacter species are abundant in nature and have the ability to perform extracellular electron transfer (EET) to reduce a broad array of heavy metals, such as Fe(III), Mn(IV), U(VI), Co(III), and Ag(I), among others (Caccavo et al. 1994; Sanford et al. 2007; Law et al. 2008; Lovley et al. 2011). It has been proposed that the electron transfer mechanisms reported for iron reduction could also be involved in Pd(II) reduction (Pat-Espadas et al. 2014). Nevertheless, the genome of G. sulfurreducens has 111 predicted c-type cytochromes (Mehta et al. 2005; Ding et al. 2008); hence, it is likely that other cytochromes, as well as conductive pili, could play a role in Pd(II) reduction (Childers et al. 2002; Reguera et al. 2005). This is supported by the observation that some proteins or cytochromes are specifically required to achieve the reduction of certain metals, such as soluble Fe(III), Fe(III) oxides, and U(VI) (Shelobolina et al. 2007; Shi et al. 2007; Ding et al. 2008).

The purpose of this study was to elucidate the mechanisms involved in Pd(II) reduction by G. sulfurreducens, based on global transcriptome analysis using RNA sequencing analysis. Quantitative reverse transcription PCR, as well as genetic and physiological tests were also performed to identify the genes involved in Pd(II) reduction.

Materials and methods

Culture procedures

Bacterial strains and oligonucleotides used in this study are listed in Table 1. G. sulfurreducens PCA (DSM 12127; ATCC51573) was grown anaerobically at 30 °C in NBAF medium, supplemented with acetate and fumarate; these culture conditions were referred to as “non-Pd(II)-reducing conditions” (Coppi et al. 2001). For experiments conducted under “Pd(II)-reducing” conditions, late logarithmic phase cultures of G. sulfurreducens were used. The protocol comprised harvesting cells by centrifugation at 9000g for 20 min and washing with sterilized, osmotically balanced buffer. The buffer composition was as follows (in grams per liter): NaHCO3, 2.5; NH4Cl, 0.25; NaH2PO4·H2O, 0.006; and KCl, 0.1. Reduction experiments were performed in 120-ml glass serum bottles, including 100 ml of anaerobic basal medium. Anaerobic conditions were established as follows: medium was dispensed into anaerobic pressure bottles, which were sealed with butyl rubber stoppers and flushed with N2/CO2 (80:20, v/v) gas mixture to remove dissolved oxygen. The headspace was saturated with the same gas mixture in all bottles, which were subsequently sterilized at 121 °C for 20 min. Cell suspensions, as well as acetate and Na2PdCl4 (Sigma-Aldrich) stock solutions were added to yield a concentration of 800 mg l−1 cell dry weight (CDW), 5 mM of acetate, and 25 mg Pd(II) l−1, respectively (Pat-Espadas et al. 2013; Pat-Espadas et al. 2014). RNALater stabilization solution (Ambion) was added to cultures for harvesting cells. Bacterial pellets were flash-frozen and stored at − 70 °C.

Table 1 Bacterial strain and primer sequences used in this work and for RT-qPCR validations
Full size table

RNA extraction

G. sulfurreducens cells from both “Pd(II) reduction” and “non-Pd(II) reduction” experimental conditions were used for RNA-Seq and quantitative real-time PCR (RT-qPCR) analyses. All experiments were performed in duplicate by using independent samples. For each biological sample, total RNA samples were extracted using the RNeasy mini kit (Qiagen), then they were examined with an Agilent 2100 Bioanalyzer and quantified using NanoDrop 200c (Thermo Scientific).

RNA-Seq and data analysis

RNA-Seq was performed using RNA samples extracted from the two experimental conditions tested. Illumina sequencing was performed at USMI (Unidad de Secuenciación Masiva, UNAM, Mexico). Briefly, after removing residual DNA using DNase I (ThermoScientific) and ribosomal RNA with Terminator 5´-Phosphate-dependent exonuclease (Epicentre), the mRNA-enriched RNA was chemically fragmented to 150–200 bp. Based on these cleaved RNA fragments, cDNA was synthesized using a random hexamer primer and reverse transcriptase. After final reparation and ligation of adaptors, obtained products were amplified by PCR, further purified, and used to create the final cDNA library. Libraries were sequenced on an Illumina Genome Analyzer IIx. Differential expression analyses were performed through IDEAmex website (http://zazil.ibt.unam.mx/ideamex/) using three methods: edgeR (Robinson and Oshlack 2010), DESeq (Anders and Huber 2010), and NOISeq (Tarazona et al. 2011). edgeR and NOISeq were performed by applying TMM (Robinson and Oshlack 2010) as the normalization method. To identify differentially expressed genes, we selected those whose p value were < 0.05 and fold change > 2, for each method. Finally, we considered as the best candidates, only genes that appeared differentially expressed in the three methods. The functional annotation of differentially expressed genes, regarding the affected pathways, was obtained from Kyoto Encyclopedia of Genes and Genomes (KEGG) (Kanehisa and Goto 2000), using our own R’s scripts. RNA-Seq transcriptome data were deposited in the NCBI Gene Expression Omnibus database under accession number GSE113152.

RT-qPCR

To validate the quality of sequencing data, some differentially expressed genes were selected for RT-qPCR analysis. mRNA was extracted as described in section “RNA extraction” and residual DNA was removed using DNase I (Thermo Scientific). cDNA synthesis was performed using RevertAid H Minus First Strand cDNA Synthesis kit (Thermo Scientific). Subsequently, RT-qPCR was performed using a Maxima SYBR Green/ROXq PCR Master Mix (Thermo Scientific) in a 96-well plate with the Light-Cycler II (Roche). Gene-specific primers used for RT-qPCR are shown in Table 1. recC was used as internal gene standard for PCR amplification. Normalized fold changes of the relative expression ratio were quantified by the 2−ΔΔCT method (Livak and Schmittgen 2001). All experiments were performed in triplicate, using independent samples, and their average values were calculated.

Cytochrome c content

Membrane fractions of G. sulfurreducens were isolated as previously described (Kim et al. 2005; Juárez et al. 2009). Outer membrane-enriched fractions were prepared by treating crude membranes with a sarkosyl (sodium N-lauroyl sarcosinate) solution at 1% (wt/vol) to extract inner membrane proteins. Outer membrane proteins were analyzed by Tris-Glycine denaturing polyacrylamide gel electrophoresis, and c-type cytochromes were detected by staining with N,N,N,N-tetramethylbenzidine, as previously described (Thomas et al. 1976; Francis and Becker 1984). PageRuler pre-stained protein standards were purchased from Thermo Scientific. The Tris-Glycine gel image was digitized using a Gel-doc (Bio Rad).

Immunoblot analysis

Protein extraction from cultures performed under “Pd(II)-reducing” and “non-Pd(II)-reducing” conditions was conducted by western blot as follows: cells pellets were re-suspended in 150 μl of B-PER II Bacterial Protein extraction reagent (Pierce) and incubated for 15 min. Afterwards, 1 mg of total protein per sample was incubated with PAGE-Buffer and boiled for 5 min before separation on a 15% SDS-PAGE. After separation, proteins were transferred to nitrocellulose membranes (Merck-Millipore) for immunoblot analysis using rabbit polyclonal antibodies raised against G. sulfurreducens (Yi et al. 2009). Blots were blocked with 3% BSA in PBS overnight at 4 °C and then incubated with a 1/1000 dilution of primary antibody for 4 h at room temperature, washed with PBS, and incubated with a 1/5000 dilution of goat anti-rabbit alkaline phosphatase-conjugated secondary antibody for 3 h at room temperature. After being washed, blots were developed with 5-bromo-4-chloro-3-indolylphosphatase (BCIP)-Nitro Blue Tetrazolium (Pierce) following manufacturer’s instructions.

Viability assay

Cell viability assays after exposure to Pd(II) were performed by recovering resting cells in NBAF medium and incubated to measure microbial growth. Prior to inoculation, cell suspensions under Pd(II)-reducing conditions were incubated for 3 h and further transferred to NBAF medium to yield a cellular density of 0.05 (OD 600nm). Cultures were incubated at 30 °C and growth was periodically monitored as OD 600nm.

Analytical techniques

Reduction of Pd(II) was quantified as follows: 5 ml of samples were filtered using 0.22 μm membrane filters (Millipore, Bedford, USA). Filtered samples were then analyzed by inductively coupled plasma-optic emission spectroscopy (ICP-OES, Varian 730-ES). Cell counts were performed using a fluorescent microscopy and acridine orange to stain cells. Briefly, a 100-μl sample was added to a 900-μl, 2.5% glutaraldehyde solution and mixed thoroughly. Bacterial cell suspensions were stained with an acridine orange solution (final concentration, 0.01%) and incubated at room temperature for 2 min. Samples were then vacuum filtered through a black Isopore membrane filter (pore diameter, 0.2 μm; Millipore).

X-ray diffraction analysis

Analysis of Pd(0) NPs deposited on the different strains of G. sulfurreducens was conducted in an X-Ray diffractometer Bruker D8 Advance. Samples were treated as previously described (Pat-Espadas et al. 2013). X-ray diffraction (XRD) patterns were recorded from 20°–90° 2θ with a step time of 2 s and step size of 0.01° 2Ө.

Results

The main results obtained from the analysis of c-type cytochromes, the genes up- and downregulated under Pd(II)-reducing conditions, as well as the contribution of the pili to palladium reduction are described in the following sections.

Differentially expressed genes during Pd(II) reduction

We used high-throughput RNA-Seq approach to profile transcriptional responses during Pd(II) reduction in G. sulfurreducens. p values and fold changes (FC) were calculated and only genes that showed differential expression by all methods were selected, resulting in 393 differentially expressed genes (Fig. 1a). A cutoff p value < 0.05 and FC > 2 were used.

Fig. 1
figure 1

Transcriptome analysis results from Geobacter sulfurreducens under Pd(II)-reducing conditions. a Venn diagram representing differential gene expression analysis from Pd(II)-reducing conditions by three statistical methods. b Functional overview of the genes that were differentially expressed during Pd(II) reduction

Full size image

Out of the 393 genes, 252 displayed statistically significant upregulation (FC > 2), while 141 downregulated under Pd(II)-reducing conditions. Genes showing significant differences in transcript levels were classified into the following functional categories: regulatory functions and transcription, energy metabolism and electron transport, DNA metabolism, transport, carbohydrate metabolism, proteolysis, protein synthesis, amino acids metabolism, mobile and extrachromosomal elements, metabolism of cofactors and vitamins, cell envelope, lipid metabolism, unknown function, and others (Fig. 1b).

The main differentially expressed genes were those involved in protein synthesis, where 50 genes were upregulated, while 11 were downregulated. The second group corresponds to genes involved in regulatory functions and transcription with 55 genes (42 upregulated and 13 downregulated). The third group of genes is involved in energy metabolism and electron transport (12 upregulated and 29 downregulated), from which 20 code for c-type cytochromes.

Expression of c-type cytochromes genes during Pd(II) reduction

Approximately half of the differentially expressed genes involved in energy metabolism and electron transport were related to c-type cytochromes (9 upregulated and 11 downregulated); 8 are located in the outer membrane, 8 in the periplasmic, 1 in the cytoplasm, and 1 attached to the inner membrane, while the location of the remaining 2 is unknown (Table 2).

Table 2 c-Type cytochrome and putative cytochrome genes differentially expressed during Pd(II) reduction
Full size table

Among the most highly upregulated c-type cytochromes under Pd(II)-reducing conditions were GSU1062, GSU2808 and PpcA. Additionally, the outer membrane cytochromes OmcH and OmcM were also overexpressed under Pd(II)-reducing conditions. Other c-type cytochromes upregulated during Pd(II) reduction were PpcD, GSU2513, GSU2934, and GSU3107. PpcD was also overexpressed in the reduction of Mn(IV) oxides, while the gsu2934 gene was overexpressed during the reduction of Fe(III) oxides (Aklujkar et al. 2013). The putative cytochrome GSU2513 has not been reported previously and its function needs to be elucidated. Additional genes expressed under Pd(II)-reducing conditions by G. sulfurreducens are presented and discussed in supplementary material (SM, Tables S1 and S2).

Cytochrome c and PilA proteins content during palladium reduction

In order to assess if mRNA expression correlates with protein content of some c-type cytochromes differentially expressed under Pd(II)-reducing conditions, we evaluated their content by heme-staining of SDS-PAGE gels. As shown in Fig. 2a, inner membrane and outer membrane proteins extracted from cultures incubated under Pd(II)-reducing and non-P(II)-reducing conditions revealed differences in abundance of c-type cytochromes. This was particularly evident for OmcB and OmcC outer membrane multiheme c-type cytochromes, which are required for Fe(III) reduction in G. sulfurreducens (Leang et al. 2005; Liu et al. 2015). OmcS, which is also required for Fe(III) oxide reduction (Qian et al. 2011), as well as the outer membrane multiheme c-type cytochrome, OmcZ, which is essential for optimal current production in microbial fuel cells (Inoue et al. 2011), were more abundant in non-Pd(II)-reducing conditions (acetate/fumarate) as compared to the level observed under Pd(II)-reducing conditions.

Fig. 2
figure 2

c-Type cytochrome and PilA protein content found under Pd(II)-reducing conditions. a SDS-PAGE heme stained. Outer membrane (OM), inner membrane (IM) and soluble fraction (SF) were prepared from PCA strain with acetate-fumarate (Ac-F) or acetate-Pd(II) (Ac-Pd(II)). The localization of OmcC, OmcB, OmcS, OmcZ, and PpcA were labeled based on expected molecular weight (78.96, 74.89, 42.94, 47.09, and 7.72 kDa, respectively). b Immunoblot analysis for PilA. The PageRuler Pre-stained Protein Ladder standard (ThermoScientific) was used as a molecular weight

Full size image

We also examined the expression of pilA gene (GSU1496), which was upregulated under Pd(II)-reducing conditions. To verify the PilA protein content under these conditions, Immunobloting analysis was performed using anti-PilA antibodies. As shown in Fig. 2b, PilA was overproduced under these conditions.

Contribution of pili on Pd(II) reduction

In order to assess the contribution of pili to the reduction of Pd(II), experiments were performed with ΔpilA (pilin-deficient mutant) and ΔpilR mutant strains. Strain ΔpilR does not produce PilR, which is the main transcriptional activator of pilA gene, encoding for pilin, the structural protein of pili (Reguera et al. 2005). Therefore, in this mutant, PilA is severely decreased (Juárez et al. 2009). Results showed differences in Pd(II)-reducing capacity for the mutant strains as compared to the wild type (WT) strain during the same incubation period (Fig. 3a). It was quantified 98%, 81%, and 61% of Pd(II) reduction for WT, ΔpilA and ΔpilR mutant strains, respectively. The nature of produced NPs was analyzed by XRD (Fig. 3b), which confirmed the formation of Pd(0). The pattern of XRD in all samples showed five strong Bragg reflections at 2 < theta > values around 40.11, 46.66, 68.13, and 82.11, which correspond to planes (111), (200), (220), and (311) of a face-centered cubic lattice (fcc) (XRD pattern was indexed to ICDD card 89–4897 (fcc palladium syn)). XRD pattern showed that Pd NP’s were crystalline in nature.

Fig. 3
figure 3

Palladium reduction by different strains of G. sulfurreducens. a Kinetics of Pd(II) reduction. b Comparison of XRD patters corresponding to cells and black precipitates obtained from cultures of wild type (WT), ΔpilA and ΔpilR strains under Pd(II)-reducing conditions. WT strain, blue line; ΔpilA strain, black line; ΔpilR strain, orange line

Full size image

Viability of G. sulfurreducens after exposure to Pd(II)

In order to verify if G. sulfurreducens is able to grow after exposure to Pd(II), cells were harvested from Pd(II)-reducing incubations and subsequently cultured in NBAF medium. Results showed that G. sulfurreducens could recover its viability after exposure to Pd(II), as shown in SM (Fig. S1). However, ΔpilA and ΔpilR mutant strains spent slightly more time than the WT strain to recover viability after exposure to Pd(II) (SM, Fig. S1). While our results confirmed that Pd(II) can be used as electron acceptor by G. sulfurreducens, no evidence demonstrating microbial growth was obtained under P(II)-reducing conditions. To corroborate if the lack of growth was not related to electron acceptor limitation, subsequent additions of Pd(II) were done and cells were able to reduce Pd(II) in several consecutive cycles, but growth was not observed (SM, Fig. S2).

Validation of selected differentially expressed genes using RT-qPCR

To verify the results obtained from RNA-Seq experiments and to get quantitative data to compare the transcript abundances under Pd(II)-reducing conditions, RT-qPCR analyses from 24 selected genes encoding proteins involved in electron transfer, transcriptional regulators and central metabolism (Table 3) were performed. These genes include the c-type cytochromes genes (omcH, omcM, omcB, omcC, omcS, omcZ, ppcD, gsu1062, ppcB, omcE, ccpA, gsu0615, gsu2937, gsu2513, gsu2808, omcQ, and gsu2495), as well as the cold shock DNA/RNA-binding protein, gsu0207, the transcriptional regulators, hgtR and gsu0837, the menaquinol oxidoreductase complex Cbc3, gsu1650, the pilin protein, pilA, the NADH dehydrogenase I, nuoH-1, and the citrate synthase I, gltA.

Table 3 Expression of genes with relevant phenotype observed in RNA-Seq analysis validated by RT-qPCR
Full size table

Upregulation of omcH, omcM, gsu1062, gsu2513, gsu2808, and ppcD genes that encoded for c-type cytochromes under Pd(II)-reducing conditions was confirmed by RT-qPCR. Similarly, the expression of pilA, gsu0207, hgtR, and gsu1650 was high under these conditions according to RT-qPCR results. On the other hand, the low transcription of omcB, omcC, omcS, omcZ, gsu0837, gsu0345, gsu2813, gsu0615, gsu2937, omcQ, gsu2495, and gsu1106 observed in RNA-Seq analyses was confirmed by RT-qPCR. Furthermore, the low transcription of nuoH-1 and gltA genes was also observed, in agreement with the high expression of hgtR, which is a negative regulator of these genes (Ueki and Lovley 2010).

Discussion

The results obtained from the analysis of the genes differentially expressed under Pd(II)-reducing conditions accounted for approximately 11% of the genes in G. sulfurreducens genome, indicating that Pd(II) reduction triggered significant global gene expression changes. The group of genes with the main notable change under these conditions was that involved in tRNA synthesis and ribosomal proteins, such as rpsU-1, rpsB, rpsB, rpsT, rpmB, rplU, and rpsL. The results also revealed a high number of upregulated transcriptional regulators, which points out the response of this bacterium to use Pd(II) as electron acceptor. Significant number of differentially expressed genes was associated to c-type cytochromes under Pd(II)-reducing conditions as it is shown in Table 2. For instance, the highly upregulated c-type cytochrome GSU1062 is a putative c-type cytochrome, which is abundant under ferric-citrate-reducing conditions (Ding et al. 2006). Similarly, gsu2808 encodes for an outer membrane cytochrome, and also reported overexpressed under Fe(III)-reducing conditions, while its expression decreases in OmcB-deficient mutant (Leang et al. 2005; Methé et al. 2005). On the other hand, cytochrome PpcA, upregulated during Pd(II) reduction, participates in electron transfer in the periplasm. A ppcA mutant showed a decrease in Fe(III) and U(VI) reduction capacities (Lloyd 2003; Mehta et al. 2005).

Overexpression of cytochromes omcH and omcM under Pd(II)-reducing conditions suggests that they could be involved in the extracellular reduction of this electron acceptor since previous work has shown that omcH was overexpressed under growing conditions with insoluble Fe(III) oxides, while mutations in the omcH and omcM genes affect the reduction of Fe(III) oxides (Aklujkar et al. 2013).

Surprisingly, during Pd(II) reduction, cytochromes OmcB, OmcC, OmcS, and OmcZ, which are involved in the reduction of Fe(III), Mn(IV), and U(VI) (Aklujkar et al. 2013), were downregulated. A previous model of microbial reduction of Pd(II) by G. sulfurreducens suggested that cytochromes OmcB and OmcS, which are important in the reduction of Fe(III), Mn(IV) or U(VI), could be involved in the reduction process (Pat-Espadas et al. 2014). However, our data suggest that Pd(II) reduction does not involve those common cytochromes (Fig. 2a), but others with different biochemical characteristics. Moreover, the overproduction of PilA was surprising since the pili is required for extracellular electron transfer to insoluble electron acceptors, such as metal oxides and electrodes (Reguera et al. 2005), but not for reducing soluble metals. In G. sulfurreducens, the pili is an important structure participating in long-distance extracellular electron transfer towards Fe(III) oxides, syntrophic partners, as well as electrodes to generate bioelectricity (Reguera et al. 2005; Summers et al. 2010; Smith et al. 2013). It has been observed that the production of some c-type cytochromes decreased under U(VI)-reducing conditions, when a ΔpilA mutant strain was used, which resulted in a slight decrease in the reduction of U(VI) to U(IV) (Cologgi et al. 2011). It has been suggested that these effects on the reduction of U(VI) in the pili-deficient strain is due to the decrease in outer membrane c-type cytochromes and not to pili deficiency (Orellana et al. 2013). Therefore, the negative effect obtained on the reduction of Pd(II) by the ΔpilA mutant strain, in our experiments, could be due to a decrease in the production of c-type cytochromes of the outer membrane instead of a negative effect related to the pili. Since PilR is a transcriptional regulator that controls the expression of at least 44 genes, among which are several c-type cytochromes, we suggest that the decreased in Pd(II) reduction observed in the ΔpilR mutant strain could be related to c-type cytochrome content rather than to the absence of pili.

The study revealed an important aspect related to G. sulfurreducens and the use of Pd(II) since no microbial growth could be verified, though this metal could be used as electron acceptor. Palladium is a toxic element for many microorganisms, which may inhibit the activity of creatine kinase, aldolase, succinate dehydrogenase, carbonic anhydrase, alkaline phosphatase, and prolyl hydroxylase (Liu et al. 1979). Moreover, it is important to mention that it has been reported that chemically synthesized Pd NP’s inhibit growth in Staphylococcus aureus and E. coli (Wilkins et al. 2013). Thus, deposition of Pd(0) NP’s on cells surface, which has been observed in G. sulfurreducens under these conditions (Pat-Espadas et al. 2013), may be a limiting factor affecting growth during the reduction of Pd(II).

Biological reduction of Pd(II) has been poorly studied. To date, several bacteria have been reported with the ability to reduce Pd(II) to Pd(0) NPs (De Corte et al. 2012). In Desulfovibrio fructosivorans and Escherichia coli, the biological reduction of Pd(II) to Pd(0) is linked to the activity of a hydrogenase (Mikheenko et al. 2008; Deplanche et al. 2010). Shewanella oneidensis and D. desulfuricans present a similar Pd(II) reduction mechanism, suggesting that hydrogenase and cytochrome c3 are involved in the reduction process (Lloyd et al. 1998; De Corte et al. 2012). Moreover, Enterobacter cloacae SgZ-5 T reduces Pd(II) to Pd(0) nanorods in the extracellular space with hydroquinone and riboflavin as redox mediators (Gardy et al. 2003). The present study contributes to elucidate the mechanisms involved in Pd(II) reduction by G. sulfurreducens. However, further studies are required to clarify if Pd(II) reduction proceeds in the periplasm.