Supplementation with Amino Acid Sources Facilitates Fermentative Growth of Shewanella oneidensis MR-1 in Defined Media

ABSTRACT

Shewanella oneidensis MR-1 is a facultative anaerobe that grows by respiration using a variety of electron acceptors. This organism serves as a model to study how bacteria thrive in redox-stratified environments. A glucose-utilizing engineered derivative of MR-1 has been reported to be unable to grow in glucose minimal medium (GMM) in the absence of electron acceptors, despite this strain having a complete set of genes for reconstructing glucose to lactate fermentative pathways. To gain insights into why MR-1 is incapable of fermentative growth, this study examined a hypothesis that this strain is programmed to repress the expression of some carbon metabolic genes in the absence of electron acceptors. Comparative transcriptomic analyses of the MR-1 derivative were conducted in the presence and absence of fumarate as an electron acceptor, and these found that the expression of many genes involved in carbon metabolism required for cell growth, including several tricarboxylic acid (TCA) cycle genes, was significantly downregulated in the absence of fumarate. This finding suggests a possibility that MR-1 is unable to grow fermentatively on glucose in minimal media owing to the shortage of nutrients essential for cell growth, such as amino acids. This idea was demonstrated in subsequent experiments that showed that the MR-1 derivative fermentatively grows in GMM containing tryptone or a defined mixture of amino acids. We suggest that gene regulatory circuits in MR-1 are tuned to minimize energy consumption under electron acceptor-depleted conditions, and that this results in defective fermentative growth in minimal media.

IMPORTANCE It is an enigma why S. oneidensis MR-1 is incapable of fermentative growth despite having complete sets of genes for reconstructing fermentative pathways. Understanding the molecular mechanisms behind this defect will facilitate the development of novel fermentation technologies for the production of value-added chemicals from biomass feedstocks, such as electro-fermentation. The information provided in this study will also improve our understanding of the ecological strategies of bacteria living in redox-stratified environments.

INTRODUCTION

Fossil resources are not only the major energy sources of human society but also the primary chemical industry feedstocks for producing commodity chemicals (1). Despite its importance to human society, the amount of usable fossil resources deposited globally is steadily decreasing (2). It has been argued that industrial systems based on alternative resources must be developed to sustain the development of our society. A possible solution for this concern would be to use biomass as feedstocks in chemical industries (3). Sugars, such as glucose, are primarily used as microbial fermentation feedstocks for producing commodity chemicals in biomass-based industries (4), while those produced from biomass currently make up only a small fraction of industrially produced chemicals.

To use biomass-derived feedstocks more widely for producing commodity chemicals, microbial fermentation processes must produce more diverse chemical species profitably. An obstacle limiting the utility of biomass feedstocks in the chemical industries is that microbial fermentation can produce sufficient chemicals only when redox equivalents between substrates and products are balanced (5). Therefore, conventional fermentation technologies have not yet been exploited to produce many petroleum-derived chemicals that are more reduced than biomass feedstocks.

Recent research has proposed a novel technology, termed electro-fermentation, that facilitates redox-unbalanced fermentation by using electrochemical interactions between electrochemically active bacteria (EAB) and electrodes in bioelectrochemical systems (6). Electro-fermentation is expected to be utilized to produce reduced products using electricity as an inexpensive reducing power. However, EAB currently known for their high electrochemical activities, including Shewanella oneidensis and Geobacter sulfurreducens, are incapable of fermentation (7), and it is currently difficult to use these bacteria for efficient electro-fermentation. For instance, a previous study has reported that S. oneidensis MR-1 harboring a plasmid containing the hexose kinase (glk) and sugar permease (galP) genes from Escherichia coli, MR-1(pBBR-glk-galP), can grow on glucose in the presence of electron acceptors, including oxygen, fumarate, and electrodes (8). However, in the absence of electron acceptors (here also referred to as fermentative conditions), MR-1(pBBR-glk-galP) cannot grow on glucose. This is despite this strain having a complete set of genes for lactic acid fermentation, including the NADH-dependent l-lactate dehydrogenase gene ldhA (GenBank accession no. AAN54042.1) (8, 9). A similar defect in fermentative growth was also observed when wild-type (WT) MR-1 was cultured on N-acetylglucosamine (NAG), a sugar that WT MR-1 can metabolize (9). Given their high electrochemical activities and potential utility in electro-fermentation processes, it is important to understand why they are incapable of fermentation.

The present study was conducted to gain insights into why S. oneidensis MR-1 is incapable of fermentative growth. The expression of some metabolic genes essential for cell growth was hypothesized to be downregulated under fermentative conditions. To analyze this hypothesis, comparative transcriptomic profiling of the MR-1 derivative was performed in the presence and absence of an electron acceptor. Results suggest that MR-1 and its derivatives are incapable of fermentative growth due to the repression of carbon metabolic pathways involved in amino acid synthesis.

RESULTS

Transcriptomic profiling of MR-1 under fermentative conditions.

S. oneidensis MR-1 harboring a plasmid containing glk and galP genes from Escherichia coli, MR-1(pBBR-glk-galP), is capable of growth in minimal medium (MM) supplemented with glucose (GMM, containing 10 mM glucose) when in the presence of electron acceptors such as oxygen, fumarate, and electrodes (8). The present study also confirmed that this strain can grow anaerobically in GMM supplemented with 40 mM fumarate (Fig. 1A). Under the fumarate-supplemented conditions, fumarate was stoichiometrically converted to succinate and acetate was produced as the major metabolite of glucose (Fig. 1B), whereas other fermentative metabolites, including dl-lactate, formate, and ethanol, were not detected. These results, together with the observation that MR-1(pBBR-glk-galP) did not grow anaerobically in GMM supplemented with 40 mM succinate (Fig. 1A), indicate that this strain utilizes fumarate as a terminal electron acceptor for anaerobic glucose catabolism (glucose oxidation to acetate and carbon dioxide). However, in the absence of fumarate, MR-1(pBBR-glk-galP) cannot grow in GMM (Fig. 1A), despite having a complete set of genes for lactic acid fermentation from glucose (Fig. 2). To address the idea that the expression of some genes required for cell growth in MR-1 is downregulated in the absence of electron acceptors, resulting in defective fermentative growth, the present study conducted comparative transcriptomic profiling of MR-1(pBBR-glk-galP) cells pregrown in GMM supplemented with fumarate (40 mM) and subsequently incubated for 3 h in the presence or absence of fumarate.

FIG 1.

Effect of fumarate on the growth of MR-1(pBBR-glk-galP) in GMM. (A) Growth of MR-1(pBBR-glk-galP) on 10 mM glucose in the presence and absence of 40 mM fumarate or succinate. (B) Concentrations of glucose, fumarate, succinate, and acetate in the fumarate- or succinate-supplemented GMM culture supernatants after 5 days of incubation. No significant glucose or succinate consumption was observed in the succinate-supplemented GMM cultures. Error bars represent standard deviations calculated from the results of three independent experiments. OD₆₀₀, optical density at 600 nm. ND, not detected.

FIG 2.

Central catabolic pathways in MR-1(pBBR-glk-galP). Red arrows and letters indicate enzymatic steps whose genes are upregulated in the absence of fumarate, blue indicates downregulation, and black indicates those whose expression was not significantly changed by the presence or absence of fumarate. Amino acids synthesized from metabolites in the pathways shown in this figure are indicated in boxes with broken arrows.

It was observed that among 4,214 genes in the MR-1 genome, 993 genes were differentially expressed between these two conditions (statistically significant at P < 0.05 and a log₂ fold change [FC] of ≥ 1 or ≤–1) (see Data Set S1 in the supplemental material). The reliability of these transcriptome analyses was validated by quantitative reverse transcriptase PCR (qRT-PCR) for selected genes (Fig. S1). Among the 993 differentially expressed genes, 500 genes were upregulated, while 493 genes were downregulated in the absence of fumarate. Given that the expression of approximately one quarter of all MR-1 genes was affected by the absence of fumarate, it is suggested that the absence of fumarate and the consequent inhibition of cell growth in GMM triggered substantial changes in the intracellular redox and nutrient states of MR-1, resulting in the drastic shift in the transcriptome profiles of this strain. Many of the upregulated and downregulated genes are found in category of “energy production and conservation [C]” in the Clusters of Orthologous Groups of proteins (COG) database (10) (Fig. S2), suggesting that MR-1 responds to the absence of electron acceptors mainly by regulating catabolic pathways. Subsequent analyses therefore focused on genes for catabolic pathways, including glycolysis, the TCA cycle, and lactate and alcohol dehydrogenases (Table 1 and Fig. 2). It was found that the expression of genes for glycolysis was mostly unchanged. In contrast, many genes for the TCA cycle were downregulated in the absence of fumarate, while genes for enzymes catalyzing the consumption of the reducing equivalent (NADH), including ldhA (encoding fermentative lactate dehydrogenase) and adhE (alcohol dehydrogenase), were upregulated. These transcriptional responses appear reasonable, since under fermentative conditions (in the absence of electron acceptors), NADH produced by glycolysis must be oxidized through reducing intermediate metabolites produced during glycolysis (e.g., pyruvate) to produce fermentation products (e.g., lactate and ethanol), for which the TCA cycle is unnecessary.

TABLE 1.

Differentially expressed genes discussed in this study

Locus tag	Gene	Annotation	Log2 FCa
Upregulated in the absence of fumarate
SO_0396	frdC	Quinol:fumarate reductase menaquinol-oxidizing subunit	2.71
SO_0397	frdC	Quinol:fumarate reductase menaquinol-oxidizing subunit	1.54
SO_0968	ldhA	Fermentative lactate dehydrogenase NADH dependent	1.38
SO_0970	fccA	Periplasmic fumarate reductase	2.77
SO_1490	adhB	Alcohol dehydrogenase II	1.72
SO_1520	lldE	l-Lactate dehydrogenase complex protein	1.11
SO_2136	adhE	Aldehyde-alcohol dehydrogenase	2.55
SO_2913	pflA	Pyruvate formate-lyase 1 activating enzyme	2.07
SO_2912	pflB	Pyruvate formate-lyase	2.36
Downregulated in the absence of fumarate
SO_0432	acnB	Aconitate hydratase	−1.51
SO_0770	mdh	NAD-dependent malate dehydrogenase	−1.69
SO_1484	aceA	Isocitrate lyase	−1.44
SO_1521	dld	Respiratory FAD-dependent d-lactate dehydrogenase	−1.03
SO_1926	gltA	Citrate synthase	−2.06
SO_1928	sdhA	Succinate dehydrogenase flavoprotein subunit	−1.54
SO_1929	sdhB	Succinate dehydrogenase iron-sulfur protein	−1.59
SO_1930	sucA	2-Oxoglutarate dehydrogenase complex E1 component	−2.02
SO_1931	sucB	2-Oxoglutarate dehydrogenase complex E2 component	−2.08
SO_1932	sucC	Succinyl-CoA synthase beta subunit	−2.66
SO_1933	sucD	Succinyl-CoA synthase alpha subunit	−2.06
SO_2629	icd	Isocitrate dehydrogenase NADP-dependent	−1.62
SO_3505	nagP	N-acetyl glucosamine transporter NagP	−1.07

^aLog₂-transformed fold change (in the absence/presence of fumarate).

Supplementation with tryptone facilitates fermentative growth.

Based on the above-described transcriptomic profiles of the MR-1 derivative, we hypothesized that the downregulation of some carbon metabolic pathways, including the TCA cycle, would precipitate a shortage of metabolites required for the cell growth of MR-1 and the resulting growth defect in the absence of electron acceptors. In particular, the downregulation of the TCA cycle was considered likely to result in defects in amino acid synthesis, since many amino acids are produced from TCA cycle components (Fig. 2) (11). To examine this hypothesis, we attempted to fermentatively grow MR-1(pBBR-glk-galP) in GMM supplemented with amino acids (Fig. 3). In this trial, 0.1% (wt/vol) tryptone, 0.1% (wt/vol) Casamino Acids (CA), or 0.5% (wt/vol) amino acid mixture (AAM, a mixture of 16 amino acids described in Materials and Methods) were used as an amino acid source. We found that in the presence of tryptone or AAM, the strain grew on glucose fermentatively (Fig. 3A and B), while CA did not promote growth (Fig. 3C). MR-1(pBBR-glk-galP) was also found not to grow solely on tryptone or AAM at the concentrations used (without glucose) (Fig. 3A and B).

FIG 3.

Effects of amino acid sources on fermentative growth of MR-1(pBBR-glk-galP) on glucose. (A to C) Medium was supplemented with 0.1% (wt/vol) tryptone (A), 0.5% (wt/vol) AAM (B), or 0.1% (wt/vol) CA (C) as an amino acid source. (D) Effects of tryptone on fermentative growth of MR-1 on NAG. Error bars represent standard deviations calculated from the results from three independent experiments.

Although CA is widely used as an amino acid source for the cultivation of microbes, tryptophan (Trp) is not included in CA since it is degraded during the acid hydrolysis of casein (12). In contrast, tryptone is produced by the trypsin treatment of casein and contains all 20 free amino acids, including Trp, as well as oligopeptides produced by incomplete enzymatic hydrolysis of casein (13). Therefore, we hypothesized that the supplementation of GMM with CA and Trp would facilitate the fermentative growth of the MR-1 derivative on glucose. To examine this hypothesis, we cultivated MR-1(pBBR-glk-galP) in GMM supplemented with 0.1% (wt/vol) CA and 0.006% (wt/vol) Trp in the absence of electron acceptors and found that supplementation with CA and Trp resulted in a small but significant increase in the final optical density at 600 nm (OD₆₀₀) of the cultures (Fig. 3C). However, the final OD₆₀₀ observed in this medium (Fig. 3C) was much lower than that obtained in GMM supplemented with AAM (Fig. 3B), suggesting that CA plus Trp still lacks some amino acid(s) that promotes the growth of MR-1 or that this strain cannot sufficiently take up Trp. Further investigation is needed to elucidate why CA cannot promote the fermentative growth of MR-1(pBBR-glk-galP). It should also be noted that the final OD₆₀₀ in the AAM-supplemented cultures (Fig. 3B) was lower than that in the tryptone-supplemented cultures (Fig. 3A). Therefore, it is suggested that oligopeptides or another unidentified compound(s) specifically contained in tryptone affected the growth of MR-1(pBBR-glk-galP).

When WT MR-1 was grown in MM supplemented with 10 mM NAG (NAG minimal medium; NMM), the same phenomenon as seen in MR-1(pBBR-glk-galP) grown in GMM was observed (Fig. 3D); namely, WT MR-1 grows fermentatively if the NMM is supplemented with tryptone. However, the culture growth rate and final OD₆₀₀ were much lower than those of MR-1(pBBR-glk-galP) grown on glucose. This is likely due to relatively low capacities of MR-1 to take up and catabolize NAG compared with those of MR-1(pBBR-glk-galP) which constitutively expresses glk and galP to take up and catabolize glucose. This concept was supported by an experiment in which the WT and MR-1(pBBR-glk-galP) were grown on NAG and glucose, respectively, in the presence of fumarate (Fig. S3). Additionally, the very slow fermentative growth of the WT on NAG (Fig. 3D) would be partially attributable to the downregulation of the NAG permease gene nagP in the absence of electron acceptors (Table 1 and Fig. 2), although the data were obtained for MR-1(pBBR-glk-galP) incubated in GMM. Supplementation with tryptone was also found to enable the fermentative growth of WT MR-1 on pyruvate (Fig. S4), although previous studies have reported that fermentative pyruvate degradation (pyruvate disproportionation) can provide energy for cell survival but cannot support cell growth (14). Collectively, these results indicate that MR-1 can grow fermentatively when external amino acid sources are available.

Characterization of fermentation by MR-1.

To further characterize the fermentative growth of MR-1(pBBR-glk-galP) on glucose in the presence of tryptone, glucose consumption and metabolite production were analyzed. As shown in Fig. 4A and B, glucose (10 mM) was mostly consumed by day 7, when dl-lactate, formate, acetate, and ethanol were detected in culture media, and d-lactate was the most abundant. Amounts of the substrate consumed and the metabolites produced during the fermentative growth are summarized in Fig. 4C and show that carbon and electron balances were obtained between the substrate and metabolites. These results indicate that approximately 70% of the glucose in the culture media was consumed for lactic acid fermentation, while the remaining 30% was consumed for mixed-acid fermentation producing formate, acetate, and ethanol as end products. In mixed-acid fermentation, a redox balance should be obtained when the ratio of formate:acetate:ethanol is 2:1:1. A similar ratio was observed in the MR-1(pBBR-glk-galP) culture (Fig. 4C). Formate, acetate, ethanol, and dl-lactate were also detected when WT MR-1 was grown fermentatively on NAG (Fig. S5), although more acetate was detected in the NAG-grown WT cultures. This result is reasonable given that the NAG-specific catabolic pathway (15) yields an additional molecule of acetate (Fig. 2).

FIG 4.

(A and B) Glucose consumption (A) and metabolite production (B) during fermentative growth of MR-1(pBBR-glk-galP) on glucose in the presence of tryptone. Error bars represent standard deviations calculated from the results from three independent experiments. (C) Typical results and proposed fermentative pathways.

MR-1 has three lactate dehydrogenases (Lld, Dld, and LdhA), among which Lld is suggested to be an NADH-independent enzyme that catalyzes the oxidation of l-lactate to pyruvate, while Dld and LdhA catalyze the conversion of d-lactate (16). It has also been suggested that LdhA is an NADH-dependent fermentative lactate dehydrogenase that preferentially converts pyruvate into d-lactate (9), while Dld is an NADH-independent respiratory lactate dehydrogenase that preferentially converts d-lactate into pyruvate (17). To identify d-lactate dehydrogenases that are involved in fermentative d-lactate production by MR-1(pBBR-glk-galP), single- and double-deletion mutants of genes for these d-lactate dehydrogenases (Δdld, ΔldhA, and ΔdldΔldhA) were transformed with pBBR-glk-galP, and their growth on glucose in the absence of electron acceptors (tryptone was added), glucose consumption, and metabolite production were analyzed (Fig. 5 and Table 2). All deletion mutants grew fermentatively on glucose, while their growth was inferior to that of MR-1(pBBR-glk-galP) (Fig. 5), and glucose consumption was suppressed in these deletion mutants (Table 2). Analyses of metabolites indicate that both Dld and LdhA were involved in the production of d-lactate, with LdhA contributing much more than Dld (Table 2). It is also shown in this table that the ldhA-deletion mutant did not produce l-lactate. This result was unexpected, but a possibility remains that LdhA in MR-1 would be a low-specificity enzyme. The dld/ldhA double-deletion mutant produced formate, acetate, and ethanol as major products (Table 2), indicating that this mutant catabolized glucose via the formate-acetate-ethanol fermentation pathway (Fig. 4C).

FIG 5.

Growth of Δdld(pBBR-glk-galP), ΔldhA(pBBR-glk-galP), and ΔdldΔldhA(pBBR-glk-galP) on glucose in the absence of electron acceptors. Medium was supplemented with 0.1% (wt/vol) tryptone. For comparison, the growth curve of MR-1(pBBR-glk-galP) shown in Fig. 3A is also plotted. Error bars represent standard deviations calculated from the results of three independent experiments.

TABLE 2.

Glucose consumption and metabolite production of Δdld(pBBR-glk-galP), ΔldhA(pBBR-glk-galP), and ΔdldΔldhA(pBBR-glk-galP) after 7 days of incubation in GMM supplemented with 0.1% (w/v) tryptone^a

Strain	Glucose consumption (mM)a	Metabolite production (mM)b
		d-Lactate	l-Lactate	Formate	Acetate	Ethanol
MR-1(pBBR-glk-galP)	9.2 ± 0.3	12.5 ± 1.5	2.3 ± 1.0	7.1 ± 0.4	2.9 ± 0.5	2.6 ± 0.2
Δdld(pBBR-glk-galP)	6.2 ± 0.9	6.0 ± 1.3	1.7 ± 0.7	3.8 ± 0.7	3.5 ± 0.3	1.2 ± 0.6
ΔldhA(pBBR-glk-galP)	5.5 ± 1.0	0.7 ± 0.3	ND	6.0 ± 0.5	4.0 ± 0.8	5.2 ± 2.2
ΔdldΔldhA(pBBR-glk-galP)	6.1 ± 0.8	ND	ND	7.0 ± 0.3	4.7 ± 0.3	6.8 ± 0.6

^aValues are means ± standard deviations (n = 3).

^bND, not detected.

To examine the contribution of acetate synthesis to the fermentative growth of MR-1, we constructed a deletion mutant of genes for acetate synthesis, ΔptaΔacs(pBBR-glk-galP) (Table 3). Under aerobic conditions, ΔptaΔacs(pBBR-glk-galP) grew but did not produce acetate, whereas the control strain, MR-1(pBBR-glk-galP), produced acetate, confirming the inability of ΔptaΔacs(pBBR-glk-galP) to synthesize acetate (Fig. S6). We found that this mutant did not grow fermentatively under anaerobic conditions in GMM even in the presence of 0.1% (wt/vol) tryptone (Fig. S7). This result suggests that ATP generated via acetate synthesis from acetyl-CoA (Fig. 4C) (18) is required for the fermentative growth of MR-1.

TABLE 3.

List of Shewanella oneidensis strains and a plasmid used in this study

Strain or plasmid	Description	Source or reference
MR-1	Wild type	ATCC
MR-1(pBBR-glk-galP)	MR-1 harboring pBBR-glk-galP	8
Δdld(pBBR-glk-galP)	ldhA (AAN54042.1) disrupted, harboring pBBR-glk-galP	8
ΔldhA(pBBR-glk-galP)	dld (AAN54582.2) disrupted, harboring pBBR-glk-galP	8
ΔdldΔldhA(pBBR-glk-galP)	ldhA and dld disrupted, harboring pBBR-glk-galP	8
ΔptaΔacs(pBBR-glk-galP)	pta (AAN55930.1) and acs (AAN55771.1) disrupted, harboring pBBR-glk-galP	This study
pBBR-glk-galP	Broad host-range plasmid containing glk and galP, Gmr	8

DISCUSSION

The data presented here show that the defect in fermentative growth of MR-1 can be rescued by supplementing culture media with tryptone or AAM (Fig. 3A and B), indicating that supplementation with appropriate amino acid sources allows MR-1 to grow in the absence of electron acceptors. We also found that supplementation with tryptone allowed MR-1(pBBR-glk-galP) to convert glucose to dl-lactate, formate, acetate, and ethanol in the absence of electron acceptors and that the redox balance between the substrate and these fermentation products was maintained (Fig. 4C). Taken together, our findings suggest that MR-1 can ferment sugars and other fermentable substrates when external amino acid sources are available. However, supplementation with AAM was less effective than supplementation with tryptone in promoting the fermentative growth of MR-1(pBBR-glk-galP) (Fig. 3A and B). Since tryptone contains oligopeptides as well as free amino acids (13), it is possible to speculate that MR-1 may not be able to sufficiently take up some free amino acids, whereas it can take up oligopeptides contained in tryptone via peptide transporters, such as Sap and DtpA (13). However, the identification of additional growth-promoting factors contained in tryptone requires further investigation.

There exist bacteria whose growth is solely dependent on respiration, including obligate aerobes such as Pseudomonas putida (19) and Azotobacter vinelandii (20). Since genome analyses indicate that these bacteria also possess the genes necessary for completing fermentative pathways from glucose (19, 20), we speculate that these obligate aerobes may also be able to grow fermentatively in the presence of appropriate nutrients for assimilation. In support of this idea, comparative genomic analyses of obligate aerobic and facultative anaerobic pseudomonads have suggested genes that are possibly necessary for growth under anaerobic conditions, including those for vitamin and amino acid biosynthesis (21).

While the MR-1 genome is predicted to encode a pathway for ethanol synthesis from acetyl-CoA (AdhE and AdhB; Fig. 2), to our knowledge, no previous study has reported ethanol production by this strain. This is likely because many other studies have analyzed catabolic products when MR-1 was cultivated in the presence of electron acceptors (8, 18, 22–27). An exception was the study by Pinchuk et al. (14) that identified catabolic products from pyruvate in the absence of electron acceptors. Under pyruvate disproportionation conditions, however, MR-1 produced lactate, acetate, and formate as the major metabolites but did not produce ethanol (14). This metabolic profile is reasonable considering that reducing equivalents required for ethanol production are not generated by pyruvate disproportionation. In contrast, the present study analyzed metabolites produced from sugars under fermentative conditions and found that MR-1 ferments sugars through a combination of lactic acid fermentation and formate-acetate-ethanol mixed-acid fermentation (Fig. 4C and Fig. S5). Similar fermentation profiles have been observed in Escherichia coli (28) and other heterofermentative bacteria (29), although E. coli produces small amounts of succinate during the fermentation of glucose (30). In E. coli, formate-acetate-ethanol fermentation is the major route for fermenting glucose, while the flux of carbon into lactic acid production is relatively minor (30, 31). In MR-1 however, the contribution of lactic acid fermentation was larger than that of formate-acetate-ethanol fermentation (Fig. 4 and Fig. S5). This metabolic characteristic seems not to be energetically beneficial given that acetate synthesis in mixed-acid fermentation yields an additional ATP per sugar molecule (Fig. 4C). It is therefore of interest to consider why MR-1 ferments sugars primarily to lactate. One possible explanation is that MR-1 produces lactate as a temporal electron sink when electron acceptors are unavailable (9). Lactic acid fermentation may also play a role in preventing the excessive accumulation of formate that inhibits cell growth (32). The latter idea is supported by the observation that the inhibition of lactate production by the deletion of dld and ldhA resulted in impaired growth in the absence of electron acceptors (Fig. 5). The ΔptaΔacs (pBBR-glk-galP) mutant was also found to be unable to grow on glucose in the absence of electron acceptors (Fig. S7). This finding suggests that ATP generation associated with acetate synthesis is essential for fermentative growth of MR-1 and that lactic acid fermentation from glucose alone does not provide sufficient ATP for growth. Given that a pta-deletion mutant of E. coli which metabolizes glucose via the Embden-Meyerhof-Parnas pathway grew in the absence of electron acceptors (33), this observation may be due to the low yield of ATP in the Entner-Doudoroff pathway that MR-1 uses for glycolysis (22, 34).

In nature, fermentation most often occurs in organics-rich reduced habitats, such as digestive tracts and lake sediments where electron acceptors are depleted (35). In such habitats, there may be sufficient nutrients, such as amino acids, available for fermentative microbes to assimilate. In contrast, in electron acceptor-rich oxidized environments, such as aerobic environments, organics, especially easily assimilable amino acid sources, are generally depleted, and organisms must synthesize amino acids to grow (36). Therefore, we consider that the repression of carbon metabolic pathways involved in amino acid synthesis in the absence of electron acceptors (or its upregulation in the presence of electron acceptors) is an ecologically feasible strategy for microbes to thrive in redox-stratified habitats. It is also likely that the gene regulatory circuits in MR-1 are tuned to efficiently utilize external resources under energy-limited fermentative conditions, resulting in the defect in fermentative growth in laboratory-defined media.

Respiratory bacteria, such as P. putida, have been examined for their applicability to the bioproduction of value-added chemicals. Studies have suggested that these bacteria would be suitable for the production of chemicals that are difficult to produce through fermentative bacteria in terms of their unique catabolic potentials, such as the biodegradation of aromatic hydrocarbons (37, 38). To diminish the loss of carbon atoms in substrates by evolving carbon dioxide and to produce target chemicals at high yields, the supply of electron acceptors must be minimized. On the other hand, EAB, such as S. oneidensis MR-1, are expected to be applied to electro-fermentation for producing reductive chemicals using electrons supplied from low-potential electrodes (39, 40). In such processes, EAB are subjected to highly reduced conditions and utilized to convert organic substrates to target compounds in the absence of electron acceptors. Therefore, we suggest that the supplementation of culture media with appropriate amino acid sources would be the key to the efficient utilization of these bacteria for the bioproduction of value-added chemicals. Our findings also suggest the possibility that these bacteria can be engineered to grow fermentatively without amino acid supplementation if gene regulatory circuits are appropriately modified.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains used in this study are listed in Table 3. These strains were cultivated at 30°C in MM containing an organic substrate. MM (pH 7.4) contained 9.0 mM (NH₄)₂SO₄, 5.7 mM K₂HPO₄, 3.3 mM KH₂PO₄, 2.0 mM NaHCO₃, 30 mM 4-(2-hydroxyethyl)-1-piperazine ethanesulfonate, 10 mL L⁻¹ each of amino acid solution (containing 2 g L⁻¹ each of l-glutamic acid, l-arginine, and dl-serine) and trace mineral solutions (8). MM was supplemented with glucose (10 mM), NAG (10 mM), or pyruvate (20 mM) that served as the carbon and energy source. In some cultures, MM was supplemented with fumarate (40 mM) that served as an electron acceptor. Tryptone (Becton, Dickinson and Company, Sparks, MD, USA) or CA (Becton, Dickinson and Company) was added to MM at a final concentration of 0.1% (wt/vol). Trp was added to GMM supplemented with CA at a final concentration of 0.006% (wt/vol). AAM contained an equal amount of the following amino acids at a total final concentration of 0.5% (wt/vol): Ala, Val, Leu, Ile, Ser, Thr, Cys, Met, Asp, Glu, Arg, Asn, Gln, Phe, Trp, and Pro. For aerobic cultivation, 5 mL of GMM or lysogeny broth (LB) in a test tube (30 mL capacity) was inoculated with an S. oneidensis strain at an OD₆₀₀ of 0.05 and was shaken at 180 rpm. For anaerobic cultivation, a strain was pregrown aerobically in LB medium, and cells were harvested by centrifugation at 15,000 × g for 5 min, washed twice with MM, and transferred to a screw-top test tube (10-mL capacity; for growth curve analysis) or vial (100-mL capacity; for metabolite analysis) containing 5 mL or 80 mL medium, respectively, at an initial OD₆₀₀ of 0.01. A test tube or vial containing inoculated medium was sealed with a butyl rubber septum, purged with high-purity nitrogen (99.99%), and incubated without shaking. The OD₆₀₀ was measured using a mini-photo 518R photometer (Taitec, Tokyo, Japan). When necessary, 15 μg mL⁻¹ gentamicin (Gm) was added to a culture medium. Agar plates contained 1.6% Bacto agar (Becton, Dickinson and Company).

Mutant construction.

In-frame disruption of the pta and acs genes in MR-1 was performed using a two-step homologous recombination method with suicide plasmid pSMV10 as described previously (41, 42). Briefly, a 1.6-kb fusion product, consisting of upstream and downstream sequences of the pta or acs gene joined by an 18-bp linker sequence, was constructed by PCR and in vitro extension using the primers listed in Table 4. The amplified fusion product was ligated into the SpeI site of pSMV10, generating pSMV-pta or pSMV-acs. To construct a pta/acs double-deletion mutant (ΔptaΔacs), pSMV-pta was first introduced into MR-1 by filter mating with E. coli WM6026. Transconjugants (single-crossover clones) were selected on LB plates containing 50 μg/mL kanamycin (Km) and were further cultivated for 20 h in LB medium lacking antibiotics. The cultures were then spread onto LB plates containing 10% (wt/vol) sucrose to isolate Km-sensitive double-crossover mutants. Disruption of the target gene in the obtained strains was confirmed by PCR and DNA sequencing. One representative mutant strain in which the pta gene was disrupted was selected and subjected to filter mating with E. coli WM6026 harboring pSMV-acs. The double-crossover mutants in which both pta and acs were disrupted were screened as described to obtain ΔptaΔacs.

TABLE 4.

Primers used in this study

Primer	Sequence (5′–3′)	Used for:
pta_5-O-SpeI	CTACACTAGTTGATCTGTGCGCATTTAGGC	Construction of pSMV-pta
pta_5-I	CTGAGCCACGATTGAGTTGCCGATGGGGATTAACAT	Construction of pSMV-pta
pta_3-I	AACTCAATCGTGGCTCAGGCGTCACAAAACGATGCT	Construction of pSMV-pta
pta_3-O-SpeI	TACCACTAGTGTCTTTGGTGAACTTGCGG	Construction of pSMV-pta
acs_5-O-SpeI	GTATACTAGTGGCTAAACACCATCGCCTTA	Construction of pSMV-acs
acs_5-I	AGCCGTGTTGGTGCTTGACCAGAAACCCTCTGGATTGA	Construction of pSMV-acs
acs_3-I	TCAAGCACCAACACGGCTCCGCTAACGAAGTGACCAAT	Construction of pSMV-acs
acs_3-O-SpeI	TACCACTAGTCCGATAGGGAGTGGCAGTAA	Construction of pSMV-acs
qRT_16S_F	ACCGCAACCCCTATCCTTAT	Quantitative RT-PCR for 16S rRNA
qRT_16S_R	CGTAAGGGCCATGATGACTT	Quantitative RT-PCR for 16S rRNA
qRT_frdC_F	GCTGCTGTACGCCTTATTGC	Quantitative RT-PCR for frdC
qRT_frdC_R	ACTAAGCTGGCGCTACCAAG	Quantitative RT-PCR for frdC
qRT_fccA_F	GGTATGAACGCCGCAGAAAC	Quantitative RT-PCR for fccA
qRT_fccA_R	CAGTCATGTCGGCACCCATA	Quantitative RT-PCR for fccA
qRT_acnB_F	GTGGATTGCACCACCGACTA	Quantitative RT-PCR for acnB
qRT_acnB_R	CGATTCGGGAAGTTACGGGT	Quantitative RT-PCR for acnB
qRT_gltA_F	GTGCGTTTAGCGGGTTCTTC	Quantitative RT-PCR for gltA
qRT_gltA_R	CGCACGGGCGATAAATTCAG	Quantitative RT-PCR for gltA
qRT_sucC_F	CATCGTTAATCTGCACGGCG	Quantitative RT-PCR for sucC
qRT_sucC_R	CGCCCACTTCTTTAACAGCG	Quantitative RT-PCR for sucC

RNA extraction.

MR-1(pBBR-glk-galP) was grown in GMM supplemented with fumarate (40 mM) as an electron acceptor, and cells were harvested at the logarithmic growth phase (OD₆₀₀, 0.2 to 0.3). These cells were incubated for 3 h in the presence or absence of fumarate, and then RNA was extracted using a TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. The extracted RNA was purified using an RNeasy minikit and RNase-free DNase set (Qiagen, Valencia, CA, USA). The quality of purified RNA was evaluated using an Agilent 2100 Bioanalyzer, RNA 6000 Pico reagents, and RNA Pico chips (Agilent Technologies, Inc., Santa Clara, CA, USA) according to the manufacturer’s instructions.

Transcriptomic analyses.

Transcriptome analysis was performed using a custom DNA microarray for MR-1 (8 × 15K; Agilent Technologies) which was designed in a previous study (43) and was validated in previous studies (44–47). Cyanine 3 (Cy3)-labeled cRNA was synthesized from 50 ng total RNA using a low-input Quick Amp WT labeling kit (Agilent Technologies) and purified using an RNeasy minikit (Qiagen). For each array, 19 μL of purified Cy3-labeled cRNA (600 ng) was mixed with 5 μL of 10× blocking agent, 1 μL of 25× fragmentation buffer, and 25 μL of 2× GE hybridization buffer Hi-RPM. The resultant mixtures were hybridized to the arrays at 65°C for 17 h. After hybridization, each microarray slide was washed with gene expression wash buffer 1 (Agilent Technologies) at room temperature for 1 min, followed by gene expression wash buffer 2 (Agilent Technologies) at 37°C for 1 min. Slides were air dried for 1 min and scanned using an Agilent DNA microarray scanner at 5-μm resolution. Gene expression data (n = 4 biological replicates) were normalized and statistically analyzed using the limma software package (version 3.36.2) (48) for R (https://www.r-project.org). A paired Student’s t test and the Benjamini-Hochberg false-discovery rate correction were used for statistical analyses. Differential expression for each probe was considered statistically significant when an absolute value of the log₂ fold change (log₂ FC) was >1.0 at P < 0.05.

qRT-PCR.

qRT-PCR was performed using a LightCycler 1.5 instrument (Roche, Indianapolis, IN, USA) according to a previously described method (49). Briefly, a PCR mixture contained 15 ng of total RNA, 1.3 μL of 50 mM Mn(OAc)₂ solution, 7.5 μL of LightCycler RNA master SYBR green I (Roche), and 0.15 μM primers listed in Table 4. To generate standard curves, DNA fragments of target genes were amplified by PCR using Ex Taq DNA polymerase (TaKaRa, Tokyo, Japan) and the primer sets listed in Table 4. These were purified by gel electrophoresis using a QIAEX II gel extraction kit (Qiagen) according to the manufacturer’s instructions. Standard curves were generated by amplifying a dilution series of the purified DNA fragments of each gene. The specificity of quantitative PCR was verified by a dissociation-curve analysis. Expression levels of target genes were normalized to the expression level of the 16S rRNA gene.

Analyses of metabolites.

After cells were removed by filtration through a membrane filter (0.20-μm pore size, DISMIC-13JP; Advantec Toyo, Tokyo, Japan), the amounts of acetate, fumarate, succinate, and formate were measured by high-performance liquid chromatography (HPLC) using an Agilent 1260 Infinity II LC system (Agilent Technologies) equipped with an Aminex HPX-87H column (300 × 7.8 mm; Bio-Rad Laboratories, Hercules, CA, USA). Sample solutions were eluted at 28°C with 10 mM H₂SO₄ at an elution rate of 0.6 mL min⁻¹. A UV detector (Agilent Technologies) set at a wavelength of 210 nm was used for signal detection. Glucose, d/l-lactate, and ethanol in the filtrate were measured using F-kit enzymatic assays (J. K. International, Tokyo, Japan) according to the manufacturer’s instructions.

Data availability.

The microarray data obtained in this study have been deposited in the NCBI Gene Expression Omnibus under accession number GSE220284.