Bacterial strains and culture conditions

The bacterial strains, plasmids, and oligonucleotides used in the present study are summarized in S1 Table. The G. sulfurreducens strains were routinely grown in NBAF anoxic medium (acetate-fumarate) [15] at 30°C and Escherichia coli strains were grown in LB medium with ampicillin (200 μg/ml) at 37°C.

Total RNA was isolated from cells derived from biofilms grown on glass and graphite. For the first condition, the cells were cultured in NBAF medium at 25°C for 48 h and the biofilms generated at the bottom of the culture flask (glass) were washed and separated from the planktonic cells. For the second condition, biofilms grown on a graphite anode assembled in an H-type MCF with FWAF medium were washed with PBS buffer and collected for RNA extraction. The biofilms were then isolated and resuspended in 1 ml of fresh NBAF medium and 100 μl of RNA later (Invitrogen). The mixture was incubated on ice for 30 min and the cells were collected by centrifugation at 14,000 rpm for 2 min. The cell pellets were then stored at –70°C until required.

DNA extraction and manipulations

Genomic DNA, plasmids, and PCR products were purified using the DNeasy blood and tissue kit (Roche), the High Pure Plasmid Isolation kit (Roche), and the GeneJET PCR purification kit (Thermo Scientific), respectively. E. coli transformations and other routine DNA manipulations were conducted following standard procedures [16].

Analysis of biofilm production and structure by CLSM

The biofilm structure and the ratio of live cells to dead cells were determined by confocal laser scanning microscopy (CLSM). Glass and graphite plates were used as supports for biofilm formation inside hermetically sealed test tubes in anaerobic conditions with NBAF medium. Incubation was performed without shaking at 25°C for 48 h. All of the solutions used in the following procedures were sterile and anaerobic. After removing the electrodes from the culture medium, the planktonic cells were removed from the biofilm with a mixture of 0.002 M cysteine and 0.9% saline isotonic solution. Afterward, a mixture of dyes from the LIVE/DEAD® BacLight Bacterial Viability kit (0.00334 mM SYTO9 and 0.02 M propidium iodide) dissolved in 0.9% saline isotonic solution and 0.1 M cysteine was added to the samples. The samples were dyed for 10 minutes, during which they were protected from any extraneous light sources. The dye was then washed with 0.002 M cysteine and 0.9% saline solution. Finally, images were captured with an Olympus FV1000 microscope at excitation wavelengths of 488 nm (green channel) and 559 nm (red channel). Imaging was performed using an immersion objective (LUMFLN 60 X 1.1 W). Fluorescence was obtained with a spectral detector at a 500–545 nm range (SDM560) for the green channel and a 570–670 nm range (Mirror) for the red channel. Images were acquired through the Z-axis of the biofilm at regular thickness intervals. Image analysis was performed using the Comstat2 (version 2.1) and Fiji (version 2.9.0) software [17,18].

RNA extraction

RNA was extracted from cells recovered from the biofilms after 48 h of incubation on glass and two weeks of growth on the graphite electrode using the RNeasy mini kit (Qiagen) according to the manufacturer’s instructions. Residual genomic DNA was digested via DNase I treatment (Thermo Scientific). The concentration and purity of the RNA samples were quantified using a NanoDrop 200c spectrophotometer (Thermo Scientific), after which sample integrity and quality were assessed with an Agilent 2100 Bioanalyzer. The extracted RNA was used for both RNA-seq and RT-qPCR analyses. The experiments were performed in duplicate, independent experiments.

RNA-seq and data analysis

RNA-seq analyses were conducted using RNA samples extracted from G. sulfurreducens biofilms [strains DL1 (wild-type) and Δgsu1771]. Illumina sequencing was performed at the Unidad Universitaria de Secuenciación Masiva y Bioinformática (UUSMB; National Autonomous University of Mexico, Mexico). All RNA samples were processed as previously described [19,20]. Briefly, ribosomal RNA was removed using the Ribominus kit (Thermo Scientific) and cDNA libraries were constructed using the TruSeq Stranded mRNA kit (Illumina), after which they were purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research). Finally, the libraries were sequenced on an Illumina NextSeq 500 sequencer and differential expression analysis was performed on the IDEAMEX web server (http://www.uusmb.unam.mx/ideamex/) [21] using the ‘edgeR’, ‘DESeq2’, ‘limma–voom’, and ‘NOISeq’ packages. Differentially expressed (DE) genes were defined as those having a p-value <0.01 and a Log2 fold change >1.5, and candidate genes were exclusively selected among the genes that were differentially expressed according to all four of the aforementioned analysis methods. Functional annotation of the DE genes was conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [22] using a custom R script. RNA-seq transcriptome data were deposited in the NCBI Gene Expression Omnibus database under accession number GSE223184.

GSU1771 protein purification

The pBAD/His-GSU1771 plasmid [14] was transformed into the E. coli MC1061 (S1 Table) strain and the expression of the 6his-GSU1771 protein was induced with 0.2% arabinose at 37°C for 4 h. The protein was purified under non-denaturing conditions using Ni-NTA resin (Qiagen) at 4°C according to the manufacturer’s instructions. The obtained 6his-GSU1771 purified protein was concentrated using Ultra 0.5 mL centrifugal filters (Amicon) and the elution buffer was replaced with storage buffer (HEPES 40 nM, KCl 50 mM, MgCl 8 mM). Protein concentrations were determined via the Bradford assay (Bio-Rad). The integrity and molecular mass of 6his-GSU1771 (approximately 28.46 kDa) were confirmed via SDS-PAGE.

DNA gel mobility shift assay

Fragments of the regulatory regions of the pgcA (391-bp), pulF (191-bp), gsu1771 (195-bp), gsu3356 (200-bp), omcM (369-bp), and omcB (210-bp) genes were amplified through PCR using G. sulfurreducens genomic DNA and the corresponding oligonucleotide pairs listed in S1 Table. A 146-bp fragment corresponding to the intergenic region between gsu1704 and gsu1705 was used as a negative control. The PCR products were purified using the GeneJet PCR purification kit (Thermo Scientific). Binding assays were performed by mixing 100 ng of each PCR fragment and 100 ng of the control fragment at increasing concentrations (0.1, 0.25, 0.5, and 1.0 μM) of purified GSU1771 protein. The DNA-protein mix was incubated in 20 μl of binding buffer (40 mM HEPES, 8 mM MgCl₂, 50 mM KCl, 1 mM DTT, 0.05% NP-40, and 0.1 mg/ml BSA) at 30°C for 30 min [20]. Afterward, the reactions were separated on a 6% polyacrylamide gel under native conditions in 0.5X TBE buffer. The gels were stained with ethidium bromide and visualized in a Gel Doc DZ imaging system (Bio-Rad).

RT-qPCR

The expression of a selection of genes that were differentially expressed in the RNA-seq analysis was quantified by RT-qPCR. Total RNA from G. sulfurreducens biofilms was obtained as described above. cDNA was synthesized using the Revert Aid First Strand DNA Synthesis kit (Thermo Scientific) and the specific reverse oligonucleotides listed in S1 Table. RT-qPCR was then performed using the Maxima SYBR Green/ROX qPCR Master Mix (Thermo Scientific) on a Rotor-Gene® Q MDx instrument (Qiagen). The relative expression of the target genes was calculated with the Rotor-Gene Q Series Software using the 2^-ΔΔCT method. The expression of the gsu2822 gene was used as an internal control. All reactions were performed in triplicate and their average values were calculated.

Cytochrome c content and immunoblot analysis

Cell-free protein extracts from G. sulfurreducens biofilms were prepared as previously described [14]. Biofilm cells were resuspended in 200 μl of B-PER II Bacterial Protein Extraction Reagent (Thermo Scientific) and incubated for 15 minutes at room temperature. Cell debris and non-lysed cells were removed by centrifugation at 14,000 rpm for 5 min. Total protein content was quantified via the Bradford assay (Bio-Rad). Afterward, 30 μg of proteins were separated via 15% SDS-PAGE and heme groups were stained with 3,3′,5,5′-tetramethylbenzidine following standard procedures [23,24]. The same concentration of protein used in heme-staining was separated as a control-loading protein and observed by coomassie staining (S1 Fig). The SDS-PAGE gels were visualized using a Gel Doc DZ imager (Bio-Rad).

Next, immuno-detection of the PilA, OmcS, and OmcZ proteins was conducted using 1, 10, and 100 μg of protein extract from G. sulfurreducens biofilms, respectively. The proteins were separated via 15% SDS-PAGE and transferred to nitrocellulose membranes (Merck-Millipore). Western blot analyses were then conducted using PilA-, OmcS-, and OmcZ-specific rabbit polyclonal antibodies [25–27]. The membranes were blocked with 10% low-fat dry milk overnight at 4°C, after which they were thoroughly washed with PBST (PBS, 0.3% Tween) and PBS. The primary anti-PilA 1:1000, anti-OmcS 1:1000, and anti-OmcZ 1:500 antibodies in PBS-BSA 0.3% were then added and incubated overnight with gentle agitation at 4°C. The membranes were then washed once more and subsequently treated with alkaline phosphatase-coupled anti-rabbit IgG secondary antibodies (Invitrogen) at a 1:5000 dilution in PBS-BSA 0.3%, after which they were allowed to incubate overnight with gentle shaking at 4°C. Finally, the membranes were revealed with 1 ml BCIP/NBT solution (Sigma-Aldrich). The protein amount used for each western blot was observed by coomassie staining as a loading control (S1 Fig).

Current production

The current production of each strain was compared in a two-chambered H-cell system with a continuous flow of acetate-containing medium (10 mM) as the electron donor and graphite stick anodes (65 cm²) poised at 60 mV versus Ag/AgCl as the electron acceptor. Once current production was initiated, the anode chamber received a steady input of fresh medium as described previously [7].

Biofilm CLSM analysis on two different supports

Previous studies have demonstrated that the Δgsu1771 mutant strain forms a thicker and more structured biofilm on FTO electrodes compared with the wild-type biofilm strain [14]. To investigate whether the Δgsu1771 strain also produces thick and structured biofilms on different support materials, biofilms were grown on both non-conductive (glass) and conductive (graphite electrode) support materials, after which their structures and key parameters were characterized through CLSM coupled with the Comstat2 and Fiji software [17,18]. Fig 1 shows the CLSM images of the three-dimensional structures of the biofilms produced by the DL1 wild-type and Δgsu1771 strains at 48 h on the glass and graphite surfaces, respectively. The wild-type strain formed a thinner and more continuous biofilm on the glass support than on the graphite surface, where it formed smooth aggregates. In contrast, the Δgsu1771 strain formed a thicker and more structured biofilm on both support materials compared to the wild-type strain. Particularly, the biofilm produced by the mutant strain exhibited column-like structures with channels, as seen in the top-down view in Fig 1, which were not present in the wild-type strain. This morphology was consistent with that of a biofilm grown in FTO in a recent study conducted by our work group [14]. On the graphite support, the Δgsu1771 mutant strain exhibited similar growth patterns to those observed on glass. Column-like structures and channels were also observed, and the biofilm formed by the mutant strain grown on graphite was thinner than the biofilm formed on the glass surface (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CLSM images of wild-type (DL1) and Δgsu1771 biofilms formed on glass and graphite supports in fumarate-containing medium.

The top and bottom panels respectively illustrate the top and side view projections generated at 48 h of growth. Live and dead cells are indicated in green and red, respectively.

https://doi.org/10.1371/journal.pone.0293359.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Biofilm parameters quantified from CLSM image analysis.

https://doi.org/10.1371/journal.pone.0293359.t001

Table 1 shows the biofilm parameters measured through image analysis using the Comstat2 and Fiji software. The thickness of the wild-type biofilm on the graphite support was almost three times higher than that of the wild-type biofilm grown on the glass support. On graphite, the roughness coefficient was also higher than on glass, which was consistent with the observed morphology on the side view images. The roughness coefficient is a measure of the variability in the height of the biofilm [28]. High roughness values indicate higher heterogeneity in the biofilm surface. Therefore, given that graphite is not a completely smooth material, higher roughness values were expected in the biofilm samples grown on the graphite electrodes. Finally, the viability (i.e., live/death cell ratio) of the wild-type cells that formed biofilms on the glass surface was slightly higher than that of the cells grown on the graphite surface. As seen in the CLSM images, the Δgsu1771 biofilm was thicker than the wild-type biofilm, particularly on the glass supports, which is consistent with previous characterizations of this mutant [14]. The following sections provide evidence of the role of GSU1771 as a regulator of important genes involved in the formation of electroactive biofilms, which provides insights into the molecular mechanisms underlying the phenotypes observed through confocal microscopy.

Differentially expressed genes in biofilm grown on glass

To characterize the GSU1771 regulon in G. sulfurreducens during biofilm formation on a glass surface, the transcriptional profile of this bacterium was examined through RNA-seq analysis. DE genes were defined as those having a p-value <0.01 and a Log2FC >1.5 according to all four of the different analysis methods selected in the IDEAMEX platform (edgeR, DESeq2, limma–voom, and NOISeq). Based on these criteria, a total of 467 DE genes were identified (Fig 2A). After classifying the genes according to their LogFC values (positive or negative), 167 genes were upregulated and 300 were downregulated (Fig 2B) (S2 Table). The DE genes were then classified into the following functional categories according to KEGG enrichment analysis: “energy metabolism and electron transport,” “carbohydrate metabolism,” “transport,” “regulatory functions and transcription,” “signal transduction,” “cell envelope,” “lipids metabolism,” “nucleotide metabolism,” “DNA/RNA metabolism,” “protein synthesis,” “proteolysis,” “metabolism of proteins and cofactors,” “unknown function,” and “others” (Fig 2B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Differential gene expression in Δgsu1771 biofilm versus wild-type biofilm (DL1) grown in glass support.

A) Venn diagram of DE genes identified using four different analysis methods. B) Functional overview of genes that were differentially expressed in the Δgsu1771 biofilm.

https://doi.org/10.1371/journal.pone.0293359.g002

“Energy metabolism and electron transport” was the category with the most DE genes (11.9%), which was followed closely by “regulatory functions and transcription” (10.4%), “transport” (9.6%), and “carbohydrate metabolism” (6%). Importantly, some of the DE genes in these functional categories are known to play important roles in biofilm formation and EET [7,9,10,29,30].

Gene expression confirmation of selected genes with RT-qPCR

To validate the results obtained by the RNA-seq approach from biofilms grown in glass, RT-qPCR analyses were conducted on a selection of target genes. The selected genes encoded proteins involved in carbohydrate metabolism (gsu1979 and acnA), energy metabolism and electron transport (hybA, pgcA, omcM, ppcD, and pilA), regulatory functions (gnfK, gsu2507, and csrA), transport (dcuB and gsu0972), lipid metabolism (gsu0490), amino acid metabolism (gsu3142), metabolism of cofactors and vitamins (gsu1706), cell envelope (gsu0810), and signal transduction (gsu3356) (Table 2). As expected, our RT-qPCR analyses confirmed the upregulation of gsu1979, gsu0972, hybA, pgcA, omcM, gsu3142, gsu1706, gsu0941, and gsu2507, as well as the downregulation of dcuB, pilA, gsu0490, gsu0810, acnA, ppcD, csrA, and gsu3356 observed in Δgsu1771 during biofilm formation according to our RNA-seq analyses. Moreover, although the RNA-seq and RT-qPCR data exhibited differences in expression levels, both datasets showed the same regulation trends.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. RT-qPCR validation of differentially expressed genes elucidated by RNA-seq.

https://doi.org/10.1371/journal.pone.0293359.t002

Expression of c-type cytochromes and PilA

Given their critical role in direct EET, c-type cytochromes had been extensively studied in G. sulfurreducens [3]. According to our transcriptome analyses, 14 genes encoding c-type cytochromes involved in energy metabolism and electron transport were differentially expressed (6 upregulated and 8 downregulated) (S2 Table).

The c-type cytochromes upregulated in the Δgsu1771 strain included omcM, gsu2808, pgcA, gsu2495, gsu3615, and gsu2937. Among these, omcM and gsu2808 are also expressed during Fe(III) and Pd(II) reduction [3], whereas gsu2937 also is expressed during Pd(II) reduction [19]. Additionally, pgcA codifies an extracellular cytochrome necessary for the reduction of Fe(III) and Mn(IV) oxides [31]. A recent study reported the potential involvement of PgcA in periplasmic electron transfer [32]. Furthermore, gsu2495 encodes a periplasmic cytochrome that is upregulated in an omcB-deficient strain in the presence of Fe(III) oxide, whereas gsu3615 is upregulated in the presence of acetate as an electron donor [33]. Finally, gsu2937 encodes a periplasmic c-type cytochrome, which is upregulated during Pd(II) reduction and is possibly involved in selenite and tellurite reduction [19,34].

The c-type cytochromes downregulated in the Δgsu1771 strain were macA, ppcD, gsu0068, gsu2811, gsu2743, gsu1740, gsu3259, and gsu2724. MacA is an inner membrane cytochrome that participates in the reduction of Fe(III) and U(VI) oxides and forms a complex with PpcA [35–37]. PpcD is a tri-heme cytochrome involved in Fe(III) reduction and its expression increases during fumarate reduction using a graphite electrode as an electron donor [3,35,38]. Additionally, genetic studies have shown that gsu0068 is important for insoluble Fe(III) reduction [3]. On the other hand, gsu2811 is abundant in co-cultures of Syntrophobacter fumaroxidans and G. sulfurreducens compared to pure G. sulfurreducens cultures, as well as when the bacteria are grown in the presence of hydrogen as an electron donor [33,39]. The GSU2743 and GSU1740 cytochromes are abundantly expressed when formate is used as an electron donor and their expression is upregulated during fumarate reduction [33,40]. Moreover, the expression of gsu2743 decreases in outer biofilms versus inner biofilms grown on a graphite anode that produces current [41]. GSU3259 (imcH) is an inner membrane cytochrome required for the reduction of electron acceptors with reduction potentials above −100 mV. Mutations in imcH have been reported to inhibit the reduction of Fe(III) citrate and Mn(IV) oxides [42]. GSU2724 is upregulated in the presence of formate and hydrogen as donor electrons [33]. Moreover, the transcript abundance of gsu2724 increases during growth on Fe(III) and Mn(IV) oxide compared with growth on Fe(III) citrate, and the deletion of this gene impairs growth on Fe(III) oxide [3].

In a recent study, we reported increases in the total content of c-type cytochromes in planktonic cells of the Δgsu1771 strain grown in NBAF medium [14]. To determine whether this increase in c-type cytochromes was conserved in biofilms of the Δgsu1771 strain, the total content of c-type cytochromes was determined by SDS-PAGE and heme staining. As illustrated in Fig 3A, the whole-cell protein extracted from the Δgsu1771 biofilm exhibited an increase in the abundance of c-type cytochromes compared to the DL1 biofilm. Although the omcS and omcZ genes were not differentially expressed in the Δgsu1771 strain according to our RNA-seq analyses, western blot analyses were conducted to assess whether there were differences in their expression at the protein level. Our findings indicated that the OmcS and OmcZ cytochromes were upregulated in the Δgsu1771 biofilm compared to DL1. This included the active form of OmcZ (OmcZ_S), which is obtained through the post-translational processing of OmcZ_L by OzpA, an enzyme that was also found to be upregulated in our RNA-seq analyses (Fig 3B and 3C, S2 Table) [43,44].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Expression of c-type cytochromes and PilA content in Δgsu1771 and wild-type biofilms.

A) SDS-heme staining from whole biofilm extracts. The asterisks indicate the major c-type cytochromes in the Δgsu1771 biofilm. Western blot analysis of OmcS (B), OmcZ (C), and PilA (D) in biofilms. Heme-staining and western blot analysis are representatives of several replicates.

https://doi.org/10.1371/journal.pone.0293359.g003

Surprisingly, the pilA gene was downregulated in the Δgsu1771 biofilm according to our RNA-seq analyses, which was contrary to its expression pattern during planktonic growth and Fe(III) reduction [12,14]. Therefore, immunodetection of the PilA protein was conducted to confirm whether its expression was correlated with that of the pilA transcript. As illustrated in Fig 3D, the content of PilA was similar in both the DLI and Δgsu1771 biofilms, suggesting the involvement of an unknown post-transcriptional regulatory mechanism that promotes the translation of PilA and other proteins such as OmcS and OmcZ in the Δgsu1771 biofilm.

DE genes involved in exopolysaccharide production

Exopolysaccharides are essential components of the extracellular matrix of bacterial biofilms [45]. Previous studies in G. sulfurreducens have characterized the expression of the xapD (gsu1501) gene, which belongs to an ABC transporter-dependent exopolysaccharide production pathway, to investigate the role of exopolysaccharides in electroconductive biofilms. Exopolysaccharides promote the adherence between cells and abiotic surfaces, in addition to enabling the anchoring of c-type cytochromes in the extracellular matrix [9–11]. However, additional studies are needed to gain more insights into the role of the exopolysaccharides in the formation of electroconductive biofilms by G. sulfurreducens. RNA-seq analysis revealed a cluster of upregulated genes related to exopolysaccharide synthesis in the Δgsu1771 strain (S2 Table). Among this cluster of genes, gsu1963 encodes a putative flippase, which is an essential protein in Wzx/Wzy-dependent exopolysaccharide synthesis pathways. The aforementioned gene cluster also included the gsu1959, gsu1961-62, gsu1976-77, gsu1952, and gsu0991 genes, which encode putative glycosyltransferases that are essential for exopolysaccharide synthesis in bacteria [46]. The gsu1985 and epsH genes were also identified within this cluster. Among them, gsu1985 is a homolog to epsE of Methylobacillus sp. strain 12S, which encodes a polysaccharide co-polymerase enzyme (PCP) that participates in the synthesis of the polysaccharide methanolan in a Wzx/Wzy-dependent pathway. epsH, a member of the exosortase family of proteins, might also participate in methanolan synthesis. However, its specific role in this process remains unknown [47]. In addition to polysaccharide synthesis, EpsH might also be involved in protein export sorting [48].

Other genes related to exopolysaccharide synthesis that were upregulated in the Δgsu1771 strain included neuB, gsu1972, and gsu1973, which could be involved in the synthesis of sialic acid. These sugars have been mainly studied due to their role in concealing or masking pathogenic bacteria to avoid detection by the host’s immune system [49]. Although the role of sialic acids in nonpathogenic bacteria is poorly understood, a recent study demonstrated that neuB mutation negatively affects biofilm formation, suggesting that sialic acid production could play an important role in the generation of biofilms by G. sulfurreducens [50].

The gsu1958 and gsu1980 genes, which encode putative polysaccharide deacetylase proteins, were upregulated in the Δgsu1771 strain. In E. coli, Pseudomonas aeruginosa, Klebsiella pneumoniae, and Yersinia pestis, deacetylation is a crucial step for polysaccharide maturation and correct positioning in the cell wall. However, in G. sulfurreducens, the role of acetylation in exopolysaccharide maturation remains unknown [51]. All of the above-described genes could thus belong to an uncharacterized pathway of exopolysaccharide synthesis in G. sulfurreducens, where exopolysaccharide modifications such as the addition of sialic acids and deacetylation are important contributors to electroactive biofilm formation [51].

DE genes involved in transmembrane transport

Another functional category with a large number of differentially expressed genes in Δgsu1771 biofilm was “transport” (S2 Table). Among the upregulated genes in this category, the pulGPQF genes encode the proteins necessary to form a type II secretion system, which in turn is required for the secretion of proteins critical for Fe(III) and Mn(IV) reduction such as OmpB [52].

Additionally, we detected DE genes that encode subunits of the RND superfamily of transporters. RND transporters function as major drug efflux pumps in many gram-negative bacteria [53]. In the Δgsu1771 biofilm, the gsu2135 and gsu2136 genes were upregulated and encoded two of the three subunits of the RND-superfamily CzcABC transporter. In G. sulfurreducens, CzcABZ transporters are implicated in the export of metals from the periplasm across the outer membrane and play a relevant role in Co(II) detoxification [54]. In contrast, the gsu1330-32 genes, which also encode an RND transporter, were downregulated in the Δgsu1771 biofilm [54].

ABC transporters are involved in the export and import of a wide range of molecules, such as the export of exopolysaccharides across the periplasm and through the outer membrane [55]. In the Δgsu1771 biofilm, we identified upregulation of the lptFG (gsu1922-23) genes, which are homologs to subunits of the ABC-transporter Lpt complex of E. coli [56]. Previous studies have also reported the upregulation of ABC-transporters in G. sulfurreducens MFCs, suggesting that the transport of extracellular matrix components plays an important role in electroactive biofilm formation [7,30]. Furthermore, the upregulation of genes that encode transport systems putatively involved in the export of exopolysaccharides in the Δgsu1771 biofilm was consistent with the previously reported increase in the content of exopolysaccharides in this strain [14].

Genes for regulation and signal transduction

Another functional category that contained a large number of DE genes in Δgsu1771 biofilm was “regulatory function and transcription” (S2 Table). Some of the genes in this functional category encoded two-component systems (TCSs). These TCSs consist of a histidine kinase (HK) that activates a response regulator (RR) by phosphotransfer, with the RR typically being a transcriptional regulator [57]. A total of 90 HKs and 93 RRs are encoded in the genome of G. sulfurreducens. Some of these RRs have been identified as enhancer-binding proteins (EBP). However, most of the TCSs in G. sulfurreducens have not been characterized [58]. In the present study, among the DE genes identified in Δgsu1771 biofilm via RNA-seq analysis, 12 were HKs (7 upregulated and 6 downregulated) and 9 were RRs (3 upregulated and 6 downregulated).

Among the upregulated genes identified in our study, gsu0470 encodes a sigma-54 dependent RR that shares 45.41% of identity with Nla6 from Myxococcus xanthus DK1622. In M. xanthus, Nla6 is a key regulator of sporulation and the development of fruiting bodies [59]. Additionally, the gsu0470-71 genes were upregulated in the Δgsu1771 biofilm. These genes encode a putative RR and HK respectively, which are homologs to the TCS ZraS/ZraR genes of E. coli and whose function is to activate the expression of genes in response to stress [60]. Among the downregulated TCS genes, kdpD/kdpE (gsu2483 and gsu2484) are homologs to the KdpD/KpdE system of E. coli that regulates the kdpFACB operon in response to potassium limitation or salt stress [61]. The decrease in the transcription of the kdpD/kdpE system was correlated with the decrease in the transcription of the kdpABC genes that code for the putative potassium-transporting ATPase complex, suggesting that this regulatory mechanism was conserved in G. sulfurreducens (S2 Table).

Cyclic diguanylate (c-di-GMP) is an important second messenger in bacteria, which modulates many physiological processes including biofilm formation. This molecule is synthesized by enzymes containing the GGDEF domain known as diguanylate cyclases (DGCs) and is degraded by phosphodiesterase enzymes [62]. The genome of G. sulfurreducens harbors 29 genes that encode diguanylate cyclases. However, only a few have been characterized [5]. Among the DE genes detected in Δgsu1771 via RNA-seq analysis, 6 genes encode putative diguanylate cyclases (1 upregulated and 5 downregulated), whereas 2 were identified as phosphodiesterase-encoding genes (1 upregulated and 1 downregulated) (S2 Table). Furthermore, gsu0895 was upregulated in the Δgsu1771 biofilm, whereas gsu1037, gsu1937, gsu3356, gsu1400, and gsu1149 were downregulated. Additionally, the putative phosphodiesterase genes gsu2622 and gsu1007 were upregulated and downregulated, respectively. Nevertheless, additional studies are needed to characterize the role of these diguanylate cyclases and phosphodiesterases in Δgsu1771 biofilm formation.

In addition to transcriptional regulation, post-transcriptional regulation also contributes greatly to controlling global expression patterns. One of the most studied regulators that control post-transcriptional expression levels in bacteria is CsrA. In bacteria, CsrA binds to specific sequences in mRNA near or overlapping ribosome binding sites and the establishment of this mRNA-protein complex inhibits mRNA translation [63]. In our RNA-seq analyses, we identified a CsrA homolog (gsu3041) that was downregulated in the Δgsu1771 biofilm. In E. coli and related bacteria, CsrA plays a key role in controlling physiological processes such as carbon metabolism, virulence, motility, and biofilm development [63]. Although the role of CsrA as a post-transcriptional regulator has not been studied in G. sulfurreducens and in phylogenetically related bacteria, its downregulation in the Δgsu1771 strain coupled with the increase in biofilm thickness in this strain suggests that this gene is involved in the regulation of biofilm development [14]. Furthermore, LepA is a highly conserved protein that participates in 30S ribosomal subunit biogenesis and translation initiation [64]. In this study, the gsu1266 (lepA) gene was upregulated in the Δgsu1771 strain. Therefore, the upregulation of lepA and downregulation of csrA in Δgsu1771 could promote the increased translation of some genes involved in electroactive biofilm development.

Additionally, the gsu2236 gene was upregulated in Δgsu1771. This gene encodes a homolog to RelA, which is classified as a stringent factor. The RelA enzyme synthesizes hormone-like molecules such as (p)ppGpp (guanosine pentaphosphate) in response to nutrient starvation, particularly amino acid shortages [65]. The deletion of the gsu2236 (rel_Gsu) gene in G. sulfurreducens causes a deficiency in Fe(III) reduction, suggesting that Rel_Gsu regulates the expression of genes involved in Fe(III) reduction, in addition to participating in the response to various environmental stresses [65]. In strain Δgsu1771, Rel_Gsu upregulation could favor the expression of genes relevant to biofilm formation and EET.

DE genes in biofilm grown on the surface of a graphite electrode

To compare the transcriptional response of the Δgsu1771 biofilm grown on glass versus graphite in current production mode, the global transcriptional response of Δgsu1771 biofilm grown on a graphite electrode in an MFC was analyzed through RNA-Seq analysis. The performance of the MFC is shown in Fig 4. The Δgsu1771 strain was more efficient in charge transfer, producing approximately 20% more charge than the wild-type strain, which was consistent with previously reported data [14]. Under these conditions, a total of 119 DE genes were identified after statistical analysis (79 upregulated and 40 downregulated). The DE genes were classified into the following metabolic categories: “unknown function,” “energy metabolism and electron transport,” “regulatory functions and transcription,” “transport,” “amino acid metabolism,” “cell envelope,” “proteolysis,” “signal transduction,” “carbohydrate metabolism,” “lipids metabolism,” “DNA/RNA metabolism,” and “others” (Fig 5). The categories containing the most DE genes were “unknown function” (41), “energy and electron transport” (20), “regulatory functions and transcription” (17), and “transport” (13) (S3 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Current production of Geobacter sulfurreducens strains.

The black and red lines represent the current production of the Δgsu1771 and wild-type strains, respectively, as a function of time. The data presented in the figure are representative time courses for multiple replicates for each treatment.

https://doi.org/10.1371/journal.pone.0293359.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Differential gene expression of the Δgsu1771 biofilm versus the wild-type biofilm (DL1) grown on the graphite electrode.

Functional overview of the DE genes detected in the Δgsu1771 biofilm.

https://doi.org/10.1371/journal.pone.0293359.g005

Our RNA-seq analyses also revealed 67 DE genes that were shared between both Δgsu1771 biofilms grown in graphite and glass. Among them, 42 were upregulated, 25 were downregulated, and 56 genes shared the same type of regulation (Table 3). In contrast, 11 DE genes in the Δgsu1771 biofilm grown on the graphite electrode were counter-regulated with respect to the biofilm grown on the glass support (Table 3). Among these 11 genes that presented counter-regulation in both RNA-seq datasets (glass and graphite), 7 encode hypothetical proteins, 3 encode proteins related to transcriptional regulation, and dcuB encodes a fumarate transporter protein [66].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Differentially expressed genes in Δgsu1771 biofilm in both transcriptome data sets.

https://doi.org/10.1371/journal.pone.0293359.t003

Among the DE genes upregulated in the Δgsu1771 strain in both conditions, the omcM, gsu2808, and gsu3615 genes, which encode c-type cytochromes, were also identified. The upregulation of these cytochromes in the Δgsu1771 strain suggests that they play a relevant role in biofilm-related EET. Additionally, a gene cluster composed of gsu0972-73, gsu0975-77, gsu0979, gsu0982-83, gsu0987-89, and gsu0991 encoding hypothetical proteins was upregulated in both conditions. Previous studies have proposed that the aforementioned genes can be acquired by lateral gene transfer; however, their function remains unknown [67]. The upregulated genes encoding proteins related to transcriptional regulation included gsu0470, gsu0471, gsu1265, gsu1268, and gsu2670.

DE genes detected exclusively in the graphite electrode biofilm

Our transcriptome analyses indicated that 52 genes (37 upregulated and 15 downregulated) were uniquely expressed in the Δgsu1771 biofilm formed on the graphite electrodes (Table 4). Importantly, the majority of these unique DE genes included c-type cytochromes (11), putative proteins of secretion systems (7), transcriptional regulators (7), and hypothetical proteins (10) (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Differentially expressed genes in Δgsu1771 biofilm only in graphite electrode.

https://doi.org/10.1371/journal.pone.0293359.t004

In the Δgsu1771 biofilm grown on the electrode, a group of genes encoding proteins homologous to the type VI secretion system (T6SS) were upregulated, including gsu3167 (vasJ), gsu0433 (vasG), gsu0428 (tssJ), gsu3166 (ImpK), gsu3166 (impL), and gsu3174 (hcp). In bacteria, T6SS has been linked to virulence, antibacterial activity, metal ion uptake, transport, and biofilm formation [68–71]. In P. aeruginosa, P. fluorescens, and Acidovorax citrulli, T6SS has been related to the formation of mature biofilms [70–72]. In P. aeruginosa and A. citrulli, mutations in the hcp gene had negative effects on the development of mature biofilms and exopolysaccharide production [71,72]. However, although G. sulfurreducens harbors the genes that comprise the T6SS, few efforts have been made to characterize their expression and function in this species [73]. To the best of our knowledge, our study is the first to demonstrate the potential participation of the T6SS in the formation of electroconductive mature biofilms in G. sulfurreducens.

The c-type cytochromes upregulated in the MFC included omcQ, gsu1538, gsu2513, gsu2748, gsu2801, gsu2887, gsu0702, and gsu2642 (omcW) (Table 4). OmcQ is a 12-heme cytochrome that is also associated with Pd(II) reduction [19]. The gsu1538 and gsu2513 genes, which encode cytochromes, were upregulated in response to Co(II) accumulation in the periplasm [54]. Additionally, GSU1538 has a putative domain of cytochrome c peroxidase and is involved in the reduction of hydrogen peroxide to avoid oxidative stress damage. In bacteria, the activity of cytochrome c peroxidase is dependent on the availability of electrons from small mono-heme cytochromes, suggesting that GSU2513 is a redox partner of GSU1538 [54,74]. Therefore, the upregulation of the gsu1538 and gsu2513 genes in Δgsu1771 biofilm suggests their protective role against the oxidative stress generated from metabolism in MFCs. Moreover, previous studies reported that in G. sulfurreducens the gsu2748 cytochrome was downregulated in a rel_Gsu mutant and during growth on Fe(III) [65].

The gsu2801 gene encoding a cytochrome related to the U(VI) response was highly expressed. However, its function remains unknown [75]. In another work, the gsu2887 gene was associated with Fe(III) citrate reduction because its encoded protein was upregulated in the omcF-mutant [76]. Additionally, the GSU2887 lipoprotein cytochrome c was highly expressed in a co-culture of Desulfotomaculum reducens MI-1 and G. sulfurreducens in Fe(III)-reducing conditions [77]. A recent study reported that the gsu0702 gene was upregulated in biofilms grown in graphite electrodes poised to a –0.17 V potential in the presence of acetate with respect to biofilms poised with the same potential in the presence of formate [78]. Although this large GSU0702 cytochrome was predicted to be extracellular, its function outside the cell has not been established. Finally, omcW (gsu2642), which is predicted as an outer surface cytochrome, was upregulated in cells grown in Fe(III) oxides but not in citrate Fe(III). In contrast, a mutant strain lacking omcW exhibited no Fe(III) reduction phenotype [3,79].

Unlike in planktonic cells and glass-generated biofilms, omcS was downregulated in electrode-grown biofilms. omcS is transcribed in a transcriptional unit and in an operon together with omcT [80]. omcS and omcT are expressed in the presence of Fe(III) oxides and fumarate as electron acceptors. However, only omcS is necessary for the reduction of Fe(III) oxides. In the Δgsu1771 biofilm grown on the graphite electrodes, omcT was also downregulated. The change in the regulation of omcS expression in biofilms grown on MFCs suggests that other regulatory proteins could be involved under these conditions, resulting in different transcriptional control mechanisms than those previously reported.

Four putative HKs (gsu1264, gsu3419, gsu1939, and gsu2815) and one RR (gsu3261) exhibited transcriptional changes in the Δgsu1771 electrode biofilm. Among these genes, gsu1264, gsu3419, and gsu3261 were upregulated, whereas gsu1939 and gsu2815 were downregulated (Table 4). The function of the products of these genes is unknown. However, the gsu3162 and gsu2815 genes have been associated with the reduction of Fe(III) oxides and Pd(II) [12,24]. The gsu3370 gene encodes a member of the GntR family regulators and was upregulated. In G. sulfurreducens, gsu3370 is a transcriptional regulator that binds to the promoter region of gltA, an enzyme that plays a key role in the tricarboxylic acid cycle. Nevertheless, little is known regarding its regulatory mechanisms and its role in controlling central metabolism [81]. In bacteria, c-di-GMP is a major signaling molecule that modulates several bacterial functions such as virulence, motility, and biofilm formation [62]. The gsu0537 gene encodes a putative diguanylate cyclase and was upregulated in the Δgsu1771 electrode biofilm. Previous studies have also reported that gsu0537 was upregulated in an adapted omcB-mutant associated with Fe(III) and Pd(II) reduction [3,19]. Transcriptional changes were also identified in five genes containing riboswitch regulation elements of the GEMM-I (genes for environment, membranes, and motility) family, which respond to cyclic dinucleotides (Table 4). Moreover, gsu1556, gsu1948, gsu1018, and gsu1945 were upregulated and omcS was downregulated. Nevertheless, the function of these positively regulated genes and the mechanism of the GSU1771 regulator remain uncharacterized.

GSU1771 binding to the promoter regions of pgcA, pulF, gsu1771, gsu3356, and relA

To determine whether GSU1771 directly regulates the expression of the pgcA, pulF, gsu1771, gsu3356, and relA genes, we analyzed the binding of GSU1771 to their promoter regions through electrophoretic mobility shift assays (EMSA). The gsu1704-gsu1705 intergenic region (control) and the omcB promoter region were used as negative controls [14]. As shown in Fig 6, GSU1771 was bound directly to all of the examined sequences except the negative control, meaning that this protein specifically interacted with all of the promoter regions. Taken together, our findings suggest that this regulator acts as a repressor of pgcA, pulF, and relA and as a transcriptional activator of gsu1771 and gsu3356.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. GSU1771 protein binding to the DNA of promoter regions.

DNA-protein binding was assessed via competitive non-radioactive EMSA. DNA of the promoter region of pgcA, gsu1771, pulF, relA, gsu3356, and omcB was incubated with increasing concentrations of purified GSU1771 (0, 0.1, 0.25, 0.5, and 1 μM). A fragment containing the gsu1704-gsu1705 intergenic region was included in each reaction as a negative control. The asterisks indicate the DNA-protein complexes.

https://doi.org/10.1371/journal.pone.0293359.g006