In this study, we isolated and characterized the novel S. oneidensis MR-1 mutant strain EC-2, which exhibits an increased ability to generate current in MFC and EC reactors. In addition, strain EC-2 forms flat, rough, and extraordinarily large colonies that are clearly distinct from the morphology of WT colonies (Figure 1), implying that cell surface structure and physicochemical properties are altered in this mutant. As we previously provided evidence that cell surface hydrophobicity influences the adhesiveness of S. oneidensis cells to graphite felt electrodes and affects current generation in MFCs , we also examined the adhesiveness and hydrophobicity of EC-2 cells. Compared to WT, EC-2 cells attached with higher frequency to graphite felt anodes (Figure 3) and had a more hydrophobic surface (Figure 4). Taken together, these results suggest that mutations introduced into EC-2 modified the cell surface structure and hydrophobicity, resulting in the enhanced adhesiveness of cells to graphite felt electrodes and increased current generation. A similar trend has also been observed for strain ∆SO_3177, which contains a mutation in a gene necessary for CPS synthesis and shows altered cell-surface hydrophobicity and enhanced the adhesiveness to graphite felt anodes . We therefore hypothesized that genes involved in the synthesis of CPS or other cell surface structures were mutated in strain EC-2.
We determined that the SO_1860 (uvrY) gene, which encodes a DNA-binding response regulator of the BarA/UvrY two-component regulatory system, was disrupted in strain EC-2 by Tn insertion. In S. oneidensis MR-1, SO_1860 is involved in the transcriptional regulation of a large number of genes, including CPS biosynthesis genes , suggesting that disruption of SO_1860 in strain EC-2 modified cell surface structure and led to the enhanced current generation in MFCs. The involvement of SO_1860 in current generation was examined by constructing an in-frame deletion mutant of SO_1860 (∆SO_1860), which was found to generate higher current in an EC, indicating the involvement of SO_1860 in current generation. ∆SO_1860 cells also exhibited altered colony morphology and increased cell surface hydrophobicity (Figure 4), and qRT-PCR analyses demonstrated that the expression of several CPS biosynthesis genes (SO_3172, SO_3177, and SO_3179) was repressed (Figure 6). These results support the notion that disruption of the SO_1860 gene affects CPS synthesis and cell surface hydrophobicity, resulting in increased current generation. However, strain EC-2 formed larger colonies, exhibited higher hydrophobicity, and generated higher current than ∆SO_1860, indicating that the disruption of SO_1860 was not the only cause for the distinct phenotype of strain EC-2. As Southern-blotting analysis confirmed that EC-2 had a single Tn insertion in SO_1860 (data not shown), it is likely that unknown mutations, in addition to the Tn insertion, may have been spontaneously introduced during the long-term (approximately 40 days) electrochemical cultivation of strain EC-2. Another explanation is that gene(s) located downstream of the Tn insertion site in SO_1860 are differently expressed in EC-2. However, it is unlikely that the distinct phenotypic features of EC-2 are attributable to the decreased expression of SO_1861 (the excinuclease ABC subunit C gene; Table 1). Further investigation, such as genome sequencing of strain EC-2, is needed to examine these hypotheses.
We examined and compared the gene expression profiles of EC-2 and ∆SO_1860 to understand the physiological differences between these two strains. qRT-PCR (Figure 6) and microarray (Table 1) analyses demonstrated that expression of CPS biosynthesis genes did not largely differ between EC-2 and ∆SO_1860, suggesting that the distinct features of EC-2 are not attributable to differential expression of CPS synthesis genes. However, we identified a number of genes that were differentially expressed between EC-2 and ∆SO_1860 (Table 1). Notably, the methionine biosynthesis genes metR and metE were highly up-regulated (10- and 90-fold, respectively) in EC-2, although it remains unclear why these genes were overexpressed. It is also interesting that the luxS gene was up-regulated in EC-2, as this enzyme catalyzes the conversion of S-ribosyl homocysteine to homocysteine and 4,5-dihydroxy-2,3-pentanedione (DPD), and is widely conserved in both Gram-negative and Gram-positive bacteria [41–43]. LuxS functions as an autoinducer-2 (AI-2) synthase because DPD is spontaneously converted to AI-2 [42–44]. LuxS is also involved in the activated methyl cycle (AMC) [38, 43] which is responsible for the synthesis of homocysteine, methionine, and S-adenosylmethionine (SAM), a major methyl donor source that is utilized for various cellular processes including the methylation of DNA and methyl-accepting chemotaxis proteins [43, 45]. MetE (methionine synthase) and MetR (transcriptional activator for met genes) are also involved in the AMC. In S. oneidensis, disruption of luxS negatively affects biofilm development on solid surfaces by interrupting the AMC . It has been also reported that a luxS-complemented (overexpressing) mutant of E. coli was deficient in pili production and formed a thicker biofilm than the WT strain, phenotypes that were suggested to be due to the depletion of SAM resulting from elevated luxS expression . In addition, a metR mutant (∆metR) of Pseudomonas aeruginosa exhibited altered colony morphology due to a severe defect in swarming motility , suggesting that the AMC is associated with cell motility and colony morphology. It is therefore possible that the overexpression of AMC-associated genes in EC-2 influences cell surface structure or motility, resulting in the altered colony morphology and increased adhesiveness of cells to electrodes. Studies are underway to investigate this possibility.
The function and expression of the metR, metE, and luxS genes are considered to be closely related, because homocysteine, one of the products of LuxS, acts as a co-regulator for MetR and stimulates the transcription of metE
. We found a potential MetR-binding site in the upstream intergenic region of the MR-1 luxS gene (5’-TGAGATGATTTCA-3’) that closely matches the consensus MetR-binding sequence reported in E. coli and other bacteria (5’-TGAANNANNTTCA-3’) . A similar sequence was also identified in the intergenic region between metE and metR (5’-TGAGCGAAATTCA-3’). These findings suggest the possibility that MetR regulates the expression of luxS, as well as that of met genes, in S. oneidensis MR-1. A putative MetR-binding sequence was also found upstream of glyA (5’-TGAGGTGCATTCA-3’). Because MetR activates the transcription of glyA in E. coli
, it is likely that the overexpression of metR in EC-2 resulted in the increased expression of glyA. In addition to these AMC-related genes, the microarray analysis detected 20 genes other than SO_1860 and SO_1861 that were differentially regulated in EC-2, including 10 genes located within the LambdaSo prophage region (Table 1). Although it has been reported that prophage-mediated cell lysis enhances biofilm formation in S. oneidensis MR-1 , the regulation and involvement of these genes in the observed phenotype of EC-2 is unknown.