Alanine, a Novel Growth Substrate for the Acetogenic Bacterium Acetobacterium woodii

Peptides and amino acids are widespread in nature, but there are only a few reports that demonstrated use of amino acids as carbon and energy sources by acetogenic bacteria, a central and important group in the anaerobic food web. Our finding that A. woodii can perform alanine oxidation coupled to reduction of carbon dioxide not only increases the number of substrates that can be used by this model acetogen but also raises the possibility that other acetogens may also be able to use alanine. Indeed, the alanine genes are also present in at least two more acetogens, for which growth on alanine has not been reported so far. Alanine may be a promising substrate for industrial fermentations, since acid formation goes along with the production of a base (NH₃) and pH regulation is a minor issue.

ABSTRACT

Acetogenic bacteria are an ecophysiologically important group of strictly anaerobic bacteria that grow lithotrophically on H₂ plus CO₂ or on CO or heterotrophically on different substrates such as sugars, alcohols, aldehydes, or acids. Amino acids are rarely used. Here, we describe that the model acetogen Acetobacterium woodii can use alanine as the sole carbon and energy source, which is in contrast to the description of the type strain. The alanine degradation genes have been identified and characterized. A key to alanine degradation is an alanine dehydrogenase which has been characterized biochemically. The resulting pyruvate is further degraded to acetate by the known pathways involving the Wood-Ljungdahl pathway. Our studies culminate in a metabolic and bioenergetic scheme for alanine-dependent acetogenesis in A. woodii.

IMPORTANCE Peptides and amino acids are widespread in nature, but there are only a few reports that demonstrated use of amino acids as carbon and energy sources by acetogenic bacteria, a central and important group in the anaerobic food web. Our finding that A. woodii can perform alanine oxidation coupled to reduction of carbon dioxide not only increases the number of substrates that can be used by this model acetogen but also raises the possibility that other acetogens may also be able to use alanine. Indeed, the alanine genes are also present in at least two more acetogens, for which growth on alanine has not been reported so far. Alanine may be a promising substrate for industrial fermentations, since acid formation goes along with the production of a base (NH₃) and pH regulation is a minor issue.

INTRODUCTION

Acetogenic bacteria are a specialized group of strictly anaerobic bacteria characterized by their ability to produce acetate from carbon dioxide and to make a living from this reaction (1–4). CO₂ is reduced to acetate in a special pathway, the Wood-Ljungdahl pathway (WLP) (5, 6). Electrons for CO₂ reduction to acetate may stem from molecular hydrogen, thus enabling lithoautotrophic growth of acetogens (7). However, this group of strictly anaerobic CO₂ reducers is metabolically very versatile, and the reductant for CO₂ reduction may also be derived from the oxidation of organic substrates, such as sugars (8, 9), alcohols (10–14), aldehydes (15), carboxylic acids (11, 16), and methyl groups (17, 18). Thus, the WLP acts as an electron sink for the oxidation of many different organic substrates. Among the organic substrates used as nutrients by acetogens, amino acids are rare (19). This is surprising, since proteins derived from the breakdown of organic material are readily available in oxic ecosystems and also in anoxic ecosystems. There are only a few reports on the oxidation of amino acids in acetogenic bacteria. Clostridium mayombei can degrade alanine, glutamate, serine, and valine (20); Clostridium methoxybenzovorans is able to utilize Casamino Acids for growth (21); isoleucine and valine could serve as the substrates for Eubacterium limosum (22); Natronincola histidinovorans uses histidine, glutamate, and Casamino Acids (23); and Sporomusa aerivorans is able to use alanine as the substrate (24). Balch et al. (25), in their original description of Acetobacterium woodii, noted that this acetogen can oxidize substrates other than hydrogen but that the substrate range is rather narrow, being limited to fructose, glucose, lactate, glycerate, and formate. The amino acids aspartate, glutamate, glycine, serine, and alanine were reported to not support growth (25). However, inspection of the genome sequence of A. woodii revealed a gene cluster possibly encoding the capacity to take up alanine and to oxidize it to pyruvate, a central intermediate of the Embden-Meyerhoff-Parnas pathway. This prompted us to reinvestigate growth of A. woodii cells on alanine.

RESULTS

Growth of A. woodii cells on alanine.

Cells transferred from a fructose-grown preculture into the same complex medium containing 50 mM alanine as the sole carbon and energy source showed a lag phase of 70 h before growth started (Fig. 1), indicating the induction of genes required for alanine metabolism. Growth occurred with a rate of 0.031 h⁻¹ to a final optical density at 600 nm (OD₆₀₀) of 0.75, whereas the growth rate and final OD of cells adapted to 50 mM alanine were 0.074 h⁻¹ and 1.0, respectively. Final OD was dependent on the alanine concentration, with a maximum at 50 mM alanine (Fig. 2).

FIG 1.

Adaptation and growth of A. woodii cells on l-alanine. A. woodii was cultivated as described in the Materials and Methods in carbonate-buffered complex medium with 10 mM fructose (■) as the carbon and energy source and transferred to medium with 50 mM alanine as the carbon and energy source (▲). Additionally, growth of a culture adapted to alanine for 5 transfers (▼) is shown. Each data point is a mean ± standard error of the mean (SEM); n = 3 independent experiments.

FIG 2.

Alanine dependence of growth of A. woodii. Carbonate-buffered complex medium was supplemented with different final concentrations of l-alanine, namely, 0, 20, 50, 75, 100, and 200 mM. The medium itself contains small amounts of l-alanine (1.26 mM) due to the yeast extract. The precultures were grown in the presence of 50 mM l-alanine. The final OD₆₀₀ (■) and the growth rate (▲) were determined. Each data point is a mean ± SEM; n = 2 independent experiments.

Alanine degradation was accompanied by acetate production. In addition, ammonium was produced and excreted. The alanine:acetate:NH₄⁺ ratio was 1:1.65:1.19. The pH fell from 7.3 to 6.8 (Fig. 3). Degradation of alanine led to the production of NH₄⁺, which is in equilibrium with NH₃, a chemical known to uncouple ATP synthesis from respiration by its action as a weak protonophore (26). To test the effect of NH₄⁺ on growth of A. woodii, cells were grown on pyruvate, the degradation product of alanine, in the presence of increasing amounts of NH₄Cl. Concentrations up to 250 mM did not affect growth. At 500 mM NH₄Cl, growth was drastically reduced, and at concentrations of 750 mM and higher, growth was no longer observed (Fig. 4). When cells were grown on alanine, inhibition by NH₄Cl was more pronounced (Fig. 5A). In the presence of 100 mM NH₄Cl, the maximum OD₆₀₀ already decreased from 0.95 to 0.45. The growth rate dropped from 0.07 h⁻¹ in the presence of 50 mM NH₄Cl to 0.044 h⁻¹ in the presence of 100 mM NH₄Cl. At NH₄Cl concentrations of ≥500 mM, no growth was observed (Fig. 5B).

FIG 3.

Conversion of l-alanine to ammonia and acetate during growth of A. woodii. Carbonate-buffered complex medium was supplemented with 30 mM l-alanine, and the OD₆₀₀ was followed for 98 h (■). The contents of l-alanine (○), acetate (□), and ammonia (⬥), as well as the pH (▲), were determined during the growth of A. woodii cells. The precultures were grown in the presence of 50 mM l-alanine. Each data point is a mean ± SEM; n = 3 independent experiments.

FIG 4.

Inhibition of growth of A. woodii on pyruvate by NH₄Cl. (A) The medium was supplemented with 50 mM pyruvate, and NH₄Cl was added to final concentrations of 0 (△), 50 (▼), 100 (◆), 250 (●), 500 (□), 750 (▲), and 1,000 mM (▽). As a negative control, no substrate was added (■). (B) The maximum OD₆₀₀ (△) and the growth rate (■) were determined and plotted against the ammonium chloride concentration. The precultures were grown in the presence of 50 mM pyruvate. Each data point is a mean ± SEM; n = 2 independent experiments.

FIG 5.

Inhibition of growth of A. woodii on l-alanine by NH₄Cl. (A) The medium was supplemented with 50 mM l-alanine, and NH₄Cl was added to final concentrations of 0 (▲), 50 (▼), 100 (⬥), 250 (●), 500 (□), 750 (△), and 1,000 mM (▽). As a negative control, no substrate was added (■). (B) The maximum OD₆₀₀ (△) and the growth rate (■) were determined and plotted against the ammonium chloride concentration. All precultures were grown in the presence of 50 mM l-alanine. Each data point is a mean ± SEM; n = 2 independent experiments.

Identification and transcriptional organization of genes involved in alanine metabolism.

To identify genes that are involved in alanine metabolism, the genome sequence was inspected. The genome harbors one gene encoding a potential alanine dehydrogenase (Awo_c24350). The gene is flanked by three other genes that are potentially involved in alanine metabolism, a gene encoding a potential sodium:alanine symporter (Awo_c24340), a gene encoding a potential pyruvate:ferredoxin oxidoreductase (Awo_c24330), and a gene encoding a potential transcriptional regulator (Awo_c24360) (Fig. 6), all transcribed in the same direction. To analyze whether the genes are expressed during growth on alanine, cells were grown in minimal medium with 50 mM alanine, and samples were taken at an OD of 0.14. RNA was isolated, cDNA was synthesized and transcript analyses were performed, as described in the Materials and Methods. All three genes, Awo_c24340, Awo_c24350, and Awo_c24360, were expressed during growth on alanine, but not during growth on fructose, indicating that they are involved in alanine metabolism (Fig. 6). In contrast, nifJ, encoding a putative pyruvate:ferredoxin oxidoreductase, was expressed while cells were grown in the presence of alanine or fructose (Fig. 6).

FIG 6.

Genetic organization and expression of the ald genes in A. woodii. (A) The alanine gene cluster in A. woodii. (B) Total RNA was isolated from alanine- and fructose-grown cells in exponential growth phase. The contaminating DNA was digested, and cDNA was synthesized. PCR analysis was performed using the oligonucleotides listed in Table 1 and indicated in panel A. As a template, cDNA from alanine-grown cells (Lane 1), cDNA from fructose-grown cells (Lane 2), genomic DNA (Lane 3), or RNA (Lane 4) was used. (C) To analyze the transcriptional organization, bridging PCR with cDNA as the template was used to amplify the intergenic regions of the ald genes in A. woodii (Lane 1). As positive and negative controls, genomic DNA and RNA were used (Lanes 2 and 3), respectively. The oligonucleotides used are listed in Table 1, and their position is indicated in panel A. Hyp. protein, hypothetical protein.

To test for the genetic organization of the alanine genes, cDNA was transcribed using mRNA isolated from alanine-grown cells. The cDNA, genomic DNA, and RNA were used in PCRs with primers that bridge the intergenic regions between Awo_c24320 and Awo_c24330, Awo_c24330 and Awo_c24340, Awo_c24340 and Awo_c24350, Awo_c24350 and Awo_c24360, and Awo_c24360 and Awo_c24370, as indicated in Fig. 6C. PCR products were obtained for Awo_c24330 and Awo_c24340 (871 bp), Awo_c24340 and Awo_c24350 (600 bp), and Awo_c24350 and Awo_c24360 (458 bp), indicating that Awo_c24330, Awo_c24340, Awo_c24350, and Awo_c24360 form a transcriptional unit. Awo_c24320 and Awo_c24370 are apparently not part of the transcript. No PCR products were obtained with isolated RNA as the template (Fig. 6C).

Properties of alanine utilization genes and deduced proteins.

Awo_c24360 is 462 bp long and encodes a protein of 17 kDa. It harbors an N-terminal asnC-type helix-turn-helix (HTH) domain and a C-terminal ligand-binding domain known from other transcription regulators of the AsnC/Lrp family. The deduced protein is 29% similar to BkdR of Pseudomonas putida, which controls the transcription of the bkd operon that encodes a branched-chain ketoacid dehydrogenase (27). Fifty-eight base pairs downstream of Awo_c24360 is Awo_c24350, which encodes a potential alanine dehydrogenase with a molecular mass of 39 kDa. It shares a sequence identity of 64% with the alanine dehydrogenase of Methanococcus maripaludis S2 (28). Awo_c24340, the next gene in the cluster, is located 175 bp downstream of Awo_c24350. It is 1,425 bp long and encodes a putative sodium:alanine symporter family protein with a molecular mass of 50 kDa. The protein is 55% identical to the sodium/alanine symporter AgcS of M. maripaludis S2 (28). Three hundred forty-three base pairs downstream of Awo_c24340 is Awo_c24330, which encodes a potential pyruvate:ferredoxin oxidoreductase (nifJ) with a molecular mass of 128 kDa. It shares a sequence identity of 59% with the pyruvate:ferredoxin oxidoreductase of Desulfovibrio africanus (29).

Enzymes involved in alanine metabolism.

The genetic data suggest that alanine is oxidatively deaminated to pyruvate by an alanine dehydrogenase. This enzyme activity could indeed be measured in cell extracts of alanine-grown cells, but not in those of fructose-grown cells. Upon addition of alanine, NAD⁺ was reduced with an activity of 3.4 U/mg protein. The activity was clearly dependent on the addition of cell extract (Fig. 7A). NAD⁺ could not be replaced by NADP⁺. The pH optimum was between pH 9.5 and 10.5 (Fig. 7B). As described for the alanine dehydrogenase from Bacillus sphaericus (30), a weak stimulation of the enzyme activity of about 16% could be observed when increasing amounts of KCl were added to the assay; the activity was maximal at 500 mM KCl. The K_m values for NAD⁺ and alanine were 0.23 mM (Fig. 8A) and 7 mM (Fig. 8B), respectively. The enzyme was specific for l-alanine, whereas d-alanine was not oxidized. The presence of a gene encoding NifJ, a pyruvate:ferredoxin-oxidoreductase, implies that pyruvate produced from alanine is further decarboxylated to acetyl-coenzyme A (CoA) by NifJ. Indeed, this activity was present in alanine-grown cells with 5.7 U/mg protein. In fructose-grown cells, this activity was only 1.7 U/mg protein.

FIG 7.

Dependence of alanine-dependent NAD⁺ reduction on protein concentration (A) and pH (B). Alanine-dependent reduction of NAD⁺ was measured as described in the Materials and Methods. In panel A, 1 mM NAD⁺, 25 mM l-alanine, and different amounts of cell extract (0 to 40 μg) were added. In panel B, the pH optimum for the alanine:NAD⁺ oxidoreductase activity was determined in buffer that contained 50 mM Tris, 50 mM CHES, 500 mM glycine, 500 mM KCl, 4 mM DTE, and 4 μM resazurin, which was adjusted to different pH values (7.5, 8.5, 9.5, 10.5, and 11.5). The reduction of NAD⁺ was followed by measuring the absorption at 340 nm. Each data point is a mean ± SEM; n = 3 independent experiments.

FIG 8.

Dependence of alanine:NAD⁺ oxidoreductase activity on NAD⁺ (A) and l-alanine (B). In panel A, the NAD⁺-dependent reduction of l-alanine to pyruvate was measured in 500 mM glycine buffer (pH 10.5), 500 mM KCl, 4 mM DTE, and 4 μM resazurin. l-Alanine (25 mM) and cell extract (10 μg) were added, as well as increasing concentrations of NAD⁺. In panel B, the dependence of alanine:NAD⁺ oxidoreductase activity on l-alanine was measured in 500 mM glycine buffer (pH 10.5), 500 mM KCl, 4 mM DTE, 4 μM resazurin, and 1 mM NAD⁺, and 10 μg cell extract was added. Different concentrations of l-alanine were added. The NAD⁺ reduction was followed at 340 nm. The determination of the K_m and V_max values and the curve fitting were performed using the GraphPad Prism program (version 4.03) and the Michaelis-Menten equation [Y = (V_max × X)/(K_m + X)]. Each data point is a mean ± SEM; n = 3 independent experiments.

The reducing equivalents derived from alanine oxidation (NADH) and pyruvate oxidation (reduced ferredoxin) are then suggested to be funneled into the Wood-Ljungdahl pathway and reoxidized by reduction of CO₂ to acetate. Indeed, enzymes of the Wood-Ljungdahl pathway, such as hydrogenase (44.9 U/mg), hydrogen-dependent CO₂-reductase (2.4 U/mg), methylene-tetrahydrofolate (THF) reductase (2.16 U/mg), CO dehydrogenase (4.9 U/mg), and ferredoxin:NAD⁺ oxidoreductase (Rnf; 0.5 U/mg) were present in cell extract of alanine-grown cells. The activities in fructose-grown cells were 37 U/mg for the hydrogenase, 2 U/mg for the hydrogen-dependent CO₂ reductase (HDCR)/formate dehydrogenase (13), 3.9 U/mg for the methylene-THF reductase (31), 3.4 U/mg for the carbon monoxide dehydrogenase (CODH) (13), and 0.15 U/mg for the Rnf complex (13).

DISCUSSION

Here, we have provided evidence that the model acetogen A. woodii uses alanine as a sole carbon and energy source. Awo_c24350 encodes the potential alanine dehydrogenase (AldD) that converts alanine to pyruvate and ammonium. The gene preceding aldD in the same transcriptional unit encodes a protein that is similar to regulators of the AsnC/Lrp family, and the gene downstream of aldD encodes a potential transporter. We propose that Awo_c24360 is the transcriptional regulator of the ald genes and Awo_c24340 is the alanine transporter and suggest naming the genes aldR and aldT, respectively. Downstream of aldT is nifJ, encoding a pyruvate:ferredoxin oxidoreductase (Pfo). This is the only Pfo-encoding gene in the genome of A. woodii, and since Pfo is also required for the metabolism of fructose (and other substrates), it is plausible that it is also expressed during growth on fructose. The mode of transcriptional regulation of nifJ is an interesting question for the future.

From the genetic and biochemical data presented we propose that alanine is oxidized according to the following equation:

$$2{\text{CH}}_3{\text{CHN}{\mathrm{H}}_3}^{+}\text{CO}{\mathrm{O}}^{-}+2{\mathrm{H}}_2\mathrm{O}\to 3\mathrm{C}{\mathrm{H}}_3\text{CO}{\mathrm{O}}^{-}+2{\mathrm{N}{\mathrm{H}}_4}^{+}+{\mathrm{H}}^{+}$$ (1)

Alanine is first oxidized to pyruvate, which is further decarboxylated by Pfo to acetyl-CoA, CO₂, and reduced ferredoxin. Alanine oxidation yields 2 mol NADH and 2 mol reduced ferredoxin. The WLP in A. woodii requires 1 mol H₂, 1 mol reduced ferredoxin, and 2 mol NADH for the reduction of 2 mol CO₂ to acetate, and redox carrier conversion is achieved by two redox-balancing modules, the electron-bifurcating hydrogenase and the ferredoxin:NAD⁺ oxidoreductase (Rnf complex) (7, 32, 33). Reduced ferredoxin (0.5 mol) is oxidized by the Rnf complex to provide 0.5 mol NADH, and this, together with the remaining 0.5 mol reduced ferredoxin, is converted to 1 mol hydrogen by the electron-bifurcating hydrogenase. Altogether, the pathway yields 1.15 mol ATP/mol alanine (Fig. 9).

FIG 9.

Pathway of acetogenesis from alanine in A. woodii. Two moles of alanine are imported via the alanine transporter (AldT), indicated in gray, and converted into 2 mol of pyruvate. This step is catalyzed by the alanine dehydrogenase (AldD). The reducing equivalents that derived from the oxidation of alanine to pyruvate and from pyruvate to acetyl-CoA, via NifJ, are funneled into the WLP. The correct stoichiometry of all reducing equivalents is achieved by action of the bifurcating hydrogenase (HydABC) and the Rnf complex.

Alanine dehydrogenase catalyzes the NAD⁺-dependent reversible amination of pyruvate to alanine and has various cellular functions. It was shown previously that in many bacteria, alanine dehydrogenase is involved in biosynthesis of alanine using pyruvate and ammonia as the substrates (34, 35). This enzyme also plays an important role in alanine degradation, as exemplified here and also in Bacillus species, where alanine is a major carbon and energy source for spore germination (36–38). For some organisms, it is also suggested that alanine dehydrogenase plays an important role in the assimilation of NH₄⁺ (39, 40). This may also be true for A. woodii and other acetogens and needs to be investigated in the future.

Inspection of the genome sequences of acetogenic bacteria revealed that all of them have an alanine dehydrogenase homolog. When we refined our search to genomes that encode an alanine dehydrogenase along with a transporter and a regulator, only Treponema primitia turned out to have all three genes, but aldR is in the opposite orientation. Blautia producta and Sporomusa ovata have the transporter and dehydrogenase but are missing a regulator in the same genetic context. Interestingly, the genome of S. ovata encodes two alanine dehydrogenases, both in context with a potential transport protein. One of these alanine dehydrogenases is in a genetic context with a predicted pyruvate:flavodoxin oxidoreductase (Fig. 10). Thus, alanine degradation is not restricted to A. woodii, but it is still a special and rather unique feature in acetogens.

FIG 10.

Organization of potential ald operons in acetogens.

MATERIALS AND METHODS

Cultivation of A. woodii.

The cultivation of A. woodii cells was carried out at 30°C in the medium previously described by Heise et al. (41, 42). Alanine (50 mM) or fructose (20 mM) were used as the growth substrate, and growth was monitored by measuring the optical density at 600 nm (OD₆₀₀).

Analysis of transcriptional units.

A. woodii was routinely cultivated on alanine or fructose and harvested in the mid-exponential growth phase (OD₆₀₀, 0.14). The pellet was resuspended in buffer containing 10 mM Tris (pH 8)-1 mM EDTA, and 600 μl of RLT buffer, provided with the RNeasy minikit (Qiagen), containing 1% of β-mercaptoethanol, was added. Cells were disrupted in a MM 301 mixer mill (Retsch, Germany) with 0.1 mm zirconia/silica beads for 5 min at 30 Hz. After 1 min of incubation, the cells were shaken again for 2 min at 30 Hz. Total RNA was isolated using the RNeasy minikit (Qiagen) according to the manufacturer's protocol. To remove remaining DNA contaminations, the samples were digested by adding Invitrogen Turbo DNase (Thermo Fisher), according to the manufacturer's protocol. The isolated RNA was stored at −80°C. cDNA synthesis was performed using the RNase H-minus point mutant Moloney murine leukemia virus (MMLV) reverse transcriptase (Promega) according to the manufacturer's protocol.

The cDNA was used as the template for analyzing the transcriptional organization and the transcript levels of the three putative ald genes via PCR. As a positive control, chromosomal DNA of A. woodii was used, and as a negative control, the isolated RNA was used. Primers used are listed in Table 1.

TABLE 1.

Oligonucleotides used in this study

Oligonucleotide	Sequence	Usage
DnaG_NifJ_fwd	ATGACATTACCGCCTACCCC	Operon analysis: intergenic region dnaG-nifJ
DnaG_NifJ _rev	CGGATTGGATCATGTGCTGG
NifJ_AldT_fwd	CGTCCTTGTGAAGCCCATAC	Operon analysis: intergenic region nifJ-aldT
NifJ_AldT_rev	GGGGCGAATCTTACCATTGC
AldT_AldD_fwd	AAGGTGAAATGTCCCCGCT	Operon analysis: intergenic region aldT-aldD
AldT_AldD_rev	AGTTGCTAATATGCCCGGTG
AldD_AldR_fwd	TTCAATACCGGCCCCAGTTT	Operon analysis: intergenic region aldD-alrR
AldD_AldR_rev	GGTTGCTTATATGTTCGTTTCGC
AldR_hyp_fwd	TTTGCGATGCTGTCACTCTG	Operon analysis: intergenic region aldR-hypothetical protein
AldR_hyp_rev	ATTTGAAGGTTGGGATGCGG
NifJ_fwd	AAAGAACCATGTACCGCACC	Transcription analysis: nifJ
NifJ_rev	GTATGGGCTTCACAAGGACG
AldT_fwd	CCGGACGAAGCAATAACGAG	Transcription analysis: aldT
AldT_rev	AAACCGACCATCCTTCACGA
AldD_fwd	GGTAACGCCGCCTAATAACA	Transcription analysis: aldD
AldD_rev	TGAGTGAAGTTGCGGGAAGA
AldR_fwd	TGAATAATCCCCGTAGAATCCAA	Transcription analysis: aldR
AldR_rev	TGCCAGAGTGACAGCATCG

Determination of acetate, ammonia, and alanine contents.

During growth of A. woodii, samples were taken and the content of acetate was determined. Therefore, the samples were centrifuged for 5 min at 13,000 rpm, and 400 μl of the cell-free supernatant were used for determination of acetate, ammonia, and alanine contents. For acetate determination, 50 μl 2 M phosphoric acid, 500 μl of 13.6 M acetone, and 50 μl of 200 mM 1-propanol, as an internal standard, were added to 400 μl of the supernatant. Acetate was determined by gas chromatography, as described previously (43). The ammonia content was determined using the ammonia assay kit (Sigma-Aldrich) according to the manufacturer's protocol, and l-alanine was determined using an alanine dehydrogenase that is recombinantly expressed in Escherichia coli (Sigma-Aldrich) that was dissolved in 2.4 M ammonia sulfate solution. For this assay, a buffer containing 40 mM Tris-HCl (pH 9), 1.4 mM EDTA, and 1 M hydrazine was used, and the NAD⁺-dependent conversion of alanine in the samples to pyruvate was measured by following the change in the absorption at a 365-nm wavelength. Every sample contained 750 U/liter enzyme.

Preparation of cell extracts.

A. woodii was cultivated in the presence of either 50 mM l-alanine or 20 mM fructose, and cells were harvested in the exponential growth phase (OD₆₀₀ = 0.3 for alanine-grown cells, and OD₆₀₀ = 1 for fructose-grown cells). The pellet was washed and resuspended in 20 ml washing buffer (25 mM Tris-HCl [pH 7], 420 mM sucrose, 2 mM dithioerythritol [DTE], and 4 μM resazurin), 5 mg/ml lysozyme was added, and the cells were incubated for 1 h at 37°C. After centrifugation for 10 min at 8,000 × g, the cell pellet was resuspended in 3 ml lysis buffer (25 mM Tris-HCl [pH 7.5], 20 mM MgSO₄, 2 mM DTE, and 4 μM resazurin) and 100 mM PMSF, and a spatula tip of DNase I was added. Cells were disrupted by two passages through a French press (900 lb/in²), and the cell extract was separated from cell debris and whole cells by centrifugation. To remove the remaining H₂, the gas phase of the samples was changed to 100% N₂, and the cell extracts were stored at 4°C. The protein content was determined according to Bradford (44).

Measurement of enzymatic activities.

All enzymatic activities were determined in 1.8-ml anoxic cuvettes (Glasgerätebau Ochs, Bovenden, Germany) under a N₂ atmosphere (alanine:NAD⁺ oxidoreductase activity, pyruvate:ferredoxin oxidoreductase activity, HDCR activity, methylene-THF reductase activity), 100% CO atmosphere (CODH activity and Rnf activity), or 110 kPa H₂ atmosphere (hydrogenase activity) at 30°C. For the alanine:NAD⁺ oxidoreductase activity, the reduction of NAD⁺ was followed at 340 nm (ε = 6.2 mM⁻¹ · cm⁻¹) in a buffer containing 500 mM glycine (pH 10.5), 500 mM KCl, 4 mM DTE, and 4 μM resazurin with 1 mM NAD⁺, 25 mM l-alanine, and cell extract added (30). For measurements of the pH optimum, a combined buffer was used, which contained 50 mM Tris, 50 mM CHES (N-cyclohexyl-2-aminoethanesulfonic acid), 500 mM glycine, 500 mM KCl, 4 mM DTE, and 4 μM resazurin. The influence of potassium ions on the enzyme activity was tested by using a buffer containing 500 mM glycine (pH 10.5) without KCl. KCl was then added to final concentration of 50, 100, 200, 300, 400, or 500 mM. The pyruvate:ferredoxin oxidoreductase activity was measured using a 50 mM Tris buffer (pH 7.5) containing 10 mM NaCl, 4 mM DTE, and 4 μM resazurin by following the decrease of absorption at 430 nm. Pyruvate (10 mM), CoA (200 μM), ferredoxin (30 μM), and cell extract were added. CODH activity was determined in buffer containing 100 mM HEPES (pH 7), 2 mM DTE, and 2 μM resazurin by following the reduction of methyl viologen (10 mM) at 604 nm (32). The measurement of hydrogenase activities was performed using a 50 mM Tris-HCl buffer (pH 7) containing 10 mM NaCl, 4 mM DTE, and 4 μM resazurin. The reduction of methyl viologen was followed at 604 nm (7). HDCR measurements were performed in buffer containing 100 mM HEPES, 20 mM MgSO₄, 2 mM DTE, and 4 μM resazurin (45). Formate (10 mM) and benzyl viologen (1 mM) were added, and the reduction was followed at 578 nm. Methylene-THF reductase activity was measured using a buffer containing 50 mM morpholinepropanesulfonic acid (MOPS), 10 mM NaCl, 20 mM MgSO₄, 2 mM DTE, and 4 μM resazurin. NADH (0.25 mM), formaldehyde (1.5 mM), THF (0.5 mM), and flavin mononucleotide (FMN) (10 μM) were added (31). The Rnf activity was determined using buffer containing 20 mM Tris (pH 7.7), 20 mM NaCl, 4 mM DTE, and 4 μM resazurin. NAD⁺ (3 mM), ferredoxin (30 μM), and CODH were added, and the reduction of NAD⁺ was followed at 340 nm (32). Ferredoxin was purified from Clostridium pasteurianum as described previously (46).