Archives of Microbiology

, Volume 187, Issue 6, pp 499–510

Purification and characterization of an iron-containing alcohol dehydrogenase in extremely thermophilic bacterium Thermotoga hypogea


  • Xiangxian Ying
    • Department of BiologyUniversity of Waterloo
  • Ying Wang
    • Department of BiologyUniversity of Waterloo
    • Xinjiang Agricultural University
  • Hamid R. Badiei
    • Department of ChemistryUniversity of Waterloo
  • Vassili Karanassios
    • Department of ChemistryUniversity of Waterloo
    • Department of BiologyUniversity of Waterloo
Original Paper

DOI: 10.1007/s00203-007-0217-x

Cite this article as:
Ying, X., Wang, Y., Badiei, H.R. et al. Arch Microbiol (2007) 187: 499. doi:10.1007/s00203-007-0217-x


Thermotoga hypogea is an extremely thermophilic anaerobic bacterium capable of growing at 90°C. It uses carbohydrates and peptides as carbon and energy sources to produce acetate, CO2, H2, l-alanine and ethanol as end products. Alcohol dehydrogenase activity was found to be present in the soluble fraction of T. hypogea. The alcohol dehydrogenase was purified to homogeneity, which appeared to be a homodimer with a subunit molecular mass of 40 ± 1 kDa revealed by SDS-PAGE analyses. A fully active enzyme contained iron of 1.02 ± 0.06 g-atoms/subunit. It was oxygen sensitive; however, loss of enzyme activity by exposure to oxygen could be recovered by incubation with dithiothreitol and Fe2+. The enzyme was thermostable with a half-life of about 10 h at 70°C, and its catalytic activity increased along with the rise of temperature up to 95°C. Optimal pH values for production and oxidation of alcohol were 8.0 and 11.0, respectively. The enzyme had a broad specificity to use primary alcohols and aldehydes as substrates. Apparent Km values for ethanol and 1-butanol were much higher than that of acetaldehyde and butyraldehyde. It was concluded that the physiological role of this enzyme is likely to catalyze the reduction of aldehydes to alcohols.


Thermotoga hypogeaAlcohol dehydrogenaseIronExtreme thermophileAnaerobeThermostability



Alcohol dehydrogenase


3-(Cyclohexylamino)-1-propanesulfonic acid


Ethylenediaminetetraacetic acid


4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid


In-torch vaporization-inductively coupled plasma-atomic emission spectrometry


1,4-Piperazine-bis-(ethanesulfonic acid)


Hyperthermophiles are a group of microorganisms that grow optimally at temperatures of ≥80°C (Stetter 1989) or are capable of growing at 90°C or above (Kelly and Adams 1994; Robb and Maeder 1998). They are classified as members of the domains archaea and bacteria (Stetter 1996). The majority of them are anaerobic organisms and many of them have a fermentative metabolism via modified Embden–Meyerhof–Parnas and Entner–Doudoroff pathways (Schönheit and Schäfer 1995; Selig et al. 1997; de Vos et al. 1998; Verhees et al. 2003; Sakuraba et al. 2004; Siebers and Schönheit 2005), by which the important intermediate metabolite pyruvate is produced. Pyruvate serves as a precursor for the biosynthesis of amino acids and production of acetate and ethanol (Ingram et al. 1999). A relatively small number of hyperthermophiles including species of Pyrococcus, Thermococcus and Thermotoga can produce ethanol, one of the most desirable renewable energy sources, as the end product (Kengen et al. 1994; Sheehan 1994; Ma et al. 1995; Fardeau et al. 1997; Jochimsen et al. 1997). However, the metabolic pathway for its production in this group of organisms is still not fully elucidated. It has been proposed that enzymes responsible for alcohol production are pyruvate decarboxylase and alcohol dehydrogenase (ADH) (Ma et al. 1997).

ADHs (EC are a family of oxidoreductases that catalyze the interconversion between alcohols and the corresponding aldehydes or ketones, and they are widely distributed in all three domains of life (Reid and Fewson 1994). They can be classified into the following categories: short-chain ADHs (lack of metal ions), zinc-containing ADHs and Fe-dependent ADHs (Reid and Fewson 1994). Zinc ions have catalytic or structural functions in several enzymes including hyperthermophilic zinc-containing ADHs (Brinen et al. 2002; Esposito et al. 2002; Guy et al. 2003; Littlechild et al. 2004). However, only a few iron-dependent ADHs are known, of which four are purified from hyperthermophiles (Radianingtyas and Wright 2003). So far, the iron-dependent ADHs that still contain iron after purification are found only in hyperthermophilic archaea. The importance and catalytic mechanism of iron in such enzymes are not well understood. Furthermore, there has been great interest in studying ADHs in hyper/thermophilic microorganisms (Radianingtyas and Wright 2003) because of their thermostability, broad range of substrate specificity and high tolerance to solvents, which are competitive features for use in the synthesis of various alcohols (Ma et al. 1994, 1995; Ma and Adams 1999; Hirakawa et al. 2004).

Thermotoga hypogea is an anaerobic, extremely thermophilic bacterium that can grow at 90°C (Fardeau et al. 1997). It can utilize carbohydrates including xylan as carbon and energy sources and can produce acetate, CO2 and hydrogen as the major end products. Ethanol is also produced as an end product of glucose/xylose fermentation (Fardeau et al. 1997). Therefore, T. hypogea may have potential application in biomass conversion to ethanol and hydrogen, which are alternative sources of fuel. It was intriguing to study properties of all essential enzymes such as ADH involved in alcohol fermentation in T. hypogea. Particularly, the type and metabolic function of the ADH in this extremely thermophilic bacterium warrant further investigation. In this work, we report the properties of the ADH in T. hypogea, the first purified iron-containing ADH from the hyperthermophilic bacteria. Its proposed physiological role is to catalyze the reduction of aldehydes to alcohols.

Materials and methods

Materials and chemicals

Chromatography columns and protein molecular mass calibration kits were purchased from Amersham Biotech (QC, Canada). Ultrafiltration membranes and Amicon microcentrifuge tubes were from Millipore (MA, USA). All chemicals used with high purities were commercially available products, unless specified.

Organism and growth conditions

T. hypogea (DSM 11164) was obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen, Germany. T. hypogea was grown in a medium as described previously (Fardeau et al. 1997; Yang and Ma 2005) and was routinely cultured in a 20-l glass carboy at 70°C. The cells were harvested by centrifugation at 13,000×g. The resulting cell pellet was frozen in liquid nitrogen immediately and stored at −80°C until use.

Preparation of cell-free extracts

All procedures for the preparation of cell-free extracts were carried out anaerobically. The frozen cells (50 g) of T. hypogea were re-suspended in 200 ml of 50 mM Tris–HCl buffer (pH 7.8) containing 2 mM dithiothreitol, 2 mM sodium dithionite and 5% (v/v) glycerol. Subsequently, lysozyme (0.1 mg ml−1) and DNase I (0.01 mg ml−1) were added to the cell suspension. The suspension was incubated at 37°C for 2 h. After centrifugation at 10,000×g for 30 min, the supernatant was collected as cell-free extract for further use.

Localization of alcohol dehydrogenase in T. hypogea

Fractionation of cell-free extract of T. hypogea was achieved by centrifugation at various g values. After each centrifugation, the supernatant and the pellet were separated in an anaerobic chamber (MBraun, USA), and the pellet was re-suspended to its original volume (6 ml). The crude extract of T. hypogea (6 ml) was centrifuged at 20,000×g for 30 min at 8°C. The supernatant was the cell-free extract that was then used for further centrifugation at 30,000×g for 1 h at 10°C. Finally, the supernatant obtained was centrifuged at 115, 000×g for 1 h at 10°C. ADH activity was measured using the method described below and the glutamate dehydrogenase activity was measured as described previously (Kort et al. 1997).

Enzyme assay and protein determination

The catalytic activity of T. hypogea ADH was measured anaerobically at 80°C by monitoring the substrate-dependent absorbance change of NADP(H) at 340 nm (ε340 = 6.3 mM−1 cm−1, Ziegenhorn et al. 1976). Unless otherwise specified, the enzyme assay was done in duplicate using the assay mixture (2 ml) for the oxidation of alcohol that contained 20 mM 1-butanol and 0.2 mM NADP in 100 mM 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS) buffer (pH 11.0). The assay mixture (2 ml) used for the reduction of aldehyde contained 22 mM butyraldehyde and 0.1 mM NADPH in 100 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer (pH 8.0). One unit of the activity is defined as 1 μmol NADPH formation or oxidation per minute. For determining enzyme kinetic parameters, various substrate concentrations (0 to ≥10 × Km unless specified) for NADPH (0, 0.007, 0.014, 0.026, 0.037, 0.050 and 0.092 mM), butyraldehyde (0, 0.05, 0.23, 0.45, 1.10, 2.30, 5.60 and 22.00 mM), acetaldehyde (0, 1.0, 2.2, 4.3, 11.0 and 42.0 mM), NADP (0, 0.009, 0.024, 0.047, 0.095 and 0.380 mM), butanol (0, 1.0, 2.1, 5.2, 10.0 and 20.0 mM) and ethanol (0, 4, 6, 10, 20, 40, 60 and 120 mM) were used for the determination of corresponding activities at 80°C, while concentrations of the corresponding co-substrates were kept not only constant, but also higher than 10 × Km. Apparent values of Km and Vmax were calculated from their Lineweaver-Burk plots. For determining pH optimum of the enzyme activity, 100 mM buffer of 1, 4-piperazine-bis-(ethanesulfonic acid) (PIPES, pH 6.0, 6.5, 7.0), HEPES (7.0, 7.5, 8.0), Tris/HCl (8.0, 8.5, 9.0), glycine (9.0, 9.5, 10,0), CAPS (10.0, 10.5, 11.0, 11.5) and phosphate (12) were used and the assays were performed at 80°C. The protein concentrations of all samples were determined using the Bradford method and bovine serum albumin served as the standard protein (Bradford 1976).

Enzyme purification

The cell-free extract of T. hypogea was applied to a DEAE-sepharose column (5 × 10 cm) that was equilibrated with buffer A [50 mM Tris–HCl buffer containing 5% (v/v) glycerol, 2 mM dithiothreitol, 2 mM sodium dithionite, pH 7.8]. After eluting the column using 150 ml buffer A, a gradient (0–0.5 M NaCl) was applied at a flow rate of 3 ml min−1. ADH was eluted while buffer A was applied. Fractions containing enzyme activity were then pooled and loaded onto a Hydroxyapatite column (2.6 × 15 cm) at a flow rate of 2 ml min−1. The column was applied with a gradient (0–0.5 M potassium phosphate in buffer A) and ADH started to elute from the column at the concentration of 0.05 M potassium phosphate. Fractions containing enzyme activity were pooled and applied to a phenyl-sepharose column (5 × 10 cm) equilibrated with 0.8 M (NH4)2SO4 in buffer A at a flow rate of 2 ml min−1. A linear gradient [0.82–0 M (NH4)2SO4 in buffer A] was applied and the ADH started to elute at a concentration of 0.64 M (NH4)2SO4. Fractions containing ADH activity were pooled and concentrated by ultrafiltration using PM-30 membrane. The concentrated sample was applied to a gel filtration column (Superdex 200, 2.6 × 60 cm) equilibrated with buffer A containing 100 mM KCl at a flow rate of 2 ml min−1. The purity of the fractions containing ADH activity was verified using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described by Laemmli (1970). The purified enzyme sample was sent to the Molecular Biology Core Facilities at Dana-Farber Cancer Institute (Boston, MA, USA) for N-terminal sequence determination (Deutscher 1990).

Activation of oxygen-inactivated ADH

Dithiothreitol and sodium dithionite present in the purified enzyme sample (0.08 mg ml−1) were first removed using a Microcon YM-30 (Millipore, MA, USA) with aerobic 50 mM Tris/HCl buffer (pH 7.8) containing 5% (w/v) glycerol at the room temperature, which took approximately 30 min. The enzyme sample without dithiothreitol and sodium dithionite was adjusted to its original volume, and an aliquot of the oxygen-inactivated sample was added into different stoppered vials. Each vial was then immediately degassed for approximately 5 min before the addition of the following compounds (final concentration): 1 mM dithiothreitol, 1 mM sodium dithionite, 0.1 mM Fe2+, or 1 mM dithiothreitol plus 0.1 mM Fe2+. Each vial was then incubated at room temperature and the ADH activities were assayed at different intervals as specified. To determine if the length of the enzyme exposure to air would affect the activity recoverability by incubating dithiothreitol and Fe2+ under anaerobic conditions, the sample solutions were vortexed well in the air. At each specified length of exposure time (0.5, 5 and 24 h), the oxygen-inactivated sample was transferred to a stoppered vial that was then degassed and filled with oxygen-free nitrogen gas. Both dithiothreitol and Fe2+ were added to a final concentration of 1 mM and 0.1 mM, respectively. At different intervals of the incubation at room temperature, ADH activities were determined using an aliquot (0.5 μg) of each sample at 80°C, as described above.

Determination of metal contents in the purified ADH

The iron (Fe) and zinc (Zn) contents of the purified enzyme were determined using in-torch vaporization inductively coupled plasma atomic emission spectrometry (ITV-ICP-AES), as described previously (Badiei et al. 2002). The remarkable aspect of the ITV-ICP-AES system used for elemental determinations is that it can provide accurate, precise and simultaneous multi-element determinations from microliter volumes (e.g., 1–3 μl) of metal-containing enzymes (with an absolute weight of the metal content in the picogram range). In general, ITV can be used with micro- to nano-amounts of liquids or solids by employing interchangeable rhenium (Re) coiled filaments or Re cups (Badiei et al. 2002). Prior to the determination, the enzyme samples were washed ten times with ultra pure buffer (10 mM Tris–HCl prepared with 18.2 MΩ cm de-ionized water) in the anaerobic chamber to remove Na+ and unbound ions by using YM-30 Amicon microcentrifuge tubes. Due to the difficulty in obtaining fully active enzyme after such filtration procedures, 2 mM dithiothreitol was added to the ultra pure buffer, which resulted in fully active ADH samples after completing the washing procedures. Aerobically prepared enzyme samples were obtained by exposure to the air after removing dithiothreitol. To construct calibration curves, standard solutions containing Fe and Zn with concentrations of 10, 30, 100 and 300 ng ml−1 (for each element) were prepared via serial dilution using 18.2 MΩ cm distilled, de-ionized water and 1,000 μg ml−1 stock solutions (SCP Science, Quebec, Canada). The stock solutions of Fe and Zn were in 4% (v/v) HNO3 and 2% (v/v) HCl, respectively. Throughout this work, 3 μl aliquots were used for ITV-ICP-AES experiments.

Determination of molecular mass and effects of metal ions

The molecular mass of the enzyme subunit was determined using SDS-PAGE (Laemmli 1970) and a calibration curve was obtained using the low molecular mass standard from Bio-Rad (14–97 kDa, Bio-Rad Laboratories, ON, Canada). The native molecular mass of the purified enzyme was estimated by gel filtration on the Superdex 200 column (2.6 × 60 cm). The column was equilibrated with buffer A containing 100 mM KCl at a flow rate of 2 ml min−1 before applying standard samples of protein kits (Pharmacia, NJ, USA) that contained blue dextran (molecular mass, Da, 2,000,000), thyroglobulin (669,000), ferritin (440,000), catalase (232,000), aldolase (158,000), bovine serum albumin (67,000), ovalbumin (43,000), chymotrysinogen A (25,000) and ribonuclease A (13,700). In order to determine the effect of the metal ions, the chelator and the reducing agents on the enzyme activity, the enzyme activity was assayed using an NADPH-dependent reduction of butyraldehyde in 100 mM PIPES at pH 7.0 to obtain improved solubility of cations over those at alkaline pHs. Each or combination of the following compounds (1 mM unless specified)—FeCl2, ZnCl2, CoCl2, NiCl2, MgCl2, CuCl2, MnCl2, CaCl2, CdCl2, ethylenediaminetetraacetic acid (EDTA), dithiothreitol (2 mM) and mercaptoethanol (2 mM) was added into the anaerobic ADH assay mixture containing 100 mM PIPES buffer (pH 7.0), 22 mM butyraldehyde and 0.1 mM NADPH. The enzymatic reaction was started by the addition of the purified enzyme (0.9 μg). The assay control did not contain any of the compounds listed above. For the pre-incubated enzyme samples (0.045 mg protein per ml) with these compounds (1 mM) and HgCl2 (1 mM) for either 1 h at 4°C or 10 min at 70°C, the activities were measured using the assay mixture that did not contain any of the compounds, and the enzyme sample without any treatment was used as an assay control.


Growth and ADH activities

T. hypogea had grown well and approximately 2 g cells (wet weight) per liter were obtained. Only an NADP-specific ADH activity (0.20 ± 0.05 U mg−1) was found to be present in the cell-free extract of T. hypogea. The ADH was located in the soluble fractions of the cell of T. hypogea because more than 95% of the ADH activity was present in the supernatant fraction after ultracentrifugation (115,000×g, 1 h) of the cell-free extract. This supernatant fraction also contained more than 90% of the glutamate dehydrogenase activity, a known cytoplasmic enzyme (Ma et al. 1994; Robb et al. 2001).

Purification and physical properties of T. hypogea ADH

ADH was partially eluted during the loading of the sample on to the DEAE-sepharose column and completely eluted once buffer A was applied. This implied that ADH did not bind to the column well and its separation from the majority of other proteins in the cell-free extract was achieved. The ADH was eluted out as a predominant single peak in all subsequent purification steps. The final purification step using gel-filtration chromatography resulted in a preparation of ADH with a specific activity of 57 U mg−1 (Table 1). Its purity was determined by SDS-PAGE, which showed a single type of subunit with a molecular mass of 40 ± 1 kDa (Fig. 1). A single protein peak corresponding to the ADH activity was eluted from the gel-filtration column Superdex 200, which had a molecular mass of 70 ± 5 kDa, indicating that the purified ADH was a homodimer enzyme.
Table 1

Purification of ADH from T. hypogea

Purification steps

Total protein (mg)

Total activitya (U)

Specific activity (U mg−1)


Yield (%)

Cell-free extract
























Gel filtration






aADH activities were determined using the method described in “Materials and methods”, except the CAPS buffer at pH 10.5
Fig. 1

SDS-PAGE analysis of ADH purified from T. hypogea. Lane 1, protein molecular mass marker (Bio-Rad Laboratories, ON, Canada); lane 2, 1.5 μg of ADH

Amino-terminal sequence analysis of the purified ADH gave rise to a single sequence (MENFVFHNPTKLIFG) that shows no similarity to other Fe-containing ADHs in hyperthermophilic archaea (Ma et al. 1995), but shows significant sequence similarity to the enzymes in hyper/thermophilic bacteria, such as Symbiobacterium thermophilum (MENFVFHNPTRLIFG) (Ueda et al. 2004), Thermotoga maritima (MENFVFHNPTKIVFG) (Nelson et al. 1999), Fervidobacterium nodosum Rt17-B1 (MNNFVFHNPTKLVFG) (Copeland et al. 2006a), Thermotoga petrophila (MKNFVFHNPTKIVFG) (Copeland et al. 2006b) and Thermoanaerobacter ethanolicus X514 (MENFVFSSPTKIIFG) (Copeland et al. 2006c). Genes encoding these enzymes are annotated as either iron-containing ADH or NADH-dependent butanol dehydrogenase; however, none of them has been purified and characterized, indicating that T. hypogea ADH is the first one purified from this uncharacterized type of hyper/thermophilic bacterial ADHs.

Both Fe and Zn were detected from the T. hypogea ADH using ITV-ICP-AES analyses (Fig. 2a, b). Prior to both Fe and Zn signals, a decrease in plasma background levels was observed, which was attributed to a pressure pulse. This pulse is due to a momentary increase in the carrier gas flow rate, which, in turn, is due to a gas expansion in the chamber by the rapidly heated Re coil. For plasma background levels in UV range, the pressure pulse typically becomes more pronounced with the wavelength of a spectral line (e.g., Fe II 259.940 nm vs. Zn I 213.856 nm). Fe and Zn contents of the fully active enzyme were found to be 1.02 ± 0.06 g-atoms per subunit and 0.08 ± 0.01 g-atoms per subunit, respectively, indicating that T. hypogea ADH is an iron-containing enzyme and that the trace amount of Zn detected may come from non-specific binding.
Fig. 2

Determination of Fe and Zn in the ADH purified in T. hypogea using ITV-ICP-AES. a Representative Fe signal detected at 259.940 nm; b representative Zn signal detected at 213.856 nm; see “Materials and methods” for details

Thermostability of T. hypogea ADH was determined by monitoring its temperature-dependent change of enzyme activity. The activity of the purified ADH increased along with the rise of assay temperature up to 95°C (data not shown). However, the enzyme activities were not measured at temperatures above 95°C because of the instability of NADPH at such high temperatures (Robb et al. 1992). An Arrhenius plot showed no obvious transition point between 30 and 95°C. The time required for a 50% loss of activity (t1/2) in the presence of dithiothreitol was approximately 10 and 2 h at 70 and 90°C, respectively (data not shown). There was a decrease in t1/2 values in the absence of dithiothreitol (t1/2 = 6 h at 70°C and t1/2 = 1 h at 90°C).

Catalytical properties of the purified T. hypogea ADH

The optimal pH for the alcohol oxidation was determined to be 11, while that of the aldehyde reduction was about 8 (Fig. 3), which is similar to the common feature of pH preference of many ADHs (Rella et al. 1987; Ma et al. 1994; Guagliardi et al. 1996; Antoine et al. 1999; van der Oost et al. 2001; Hirakawa et al. 2004). It showed that the aldehyde reduction activity was about 17 times higher than that of the alcohol oxidation at pH 8.0, indicating that this enzyme is more efficient for catalyzing the aldehyde reduction in the cell where approximately neutral pH is maintained.
Fig. 3

pH dependence of alcohol oxidation and aldehyde reduction activities catalyzed by the ADH in T. hypogea. Filled symbols represent the reduction of butyraldehyde; open symbols represent the oxidation of 1-butanol. Buffers used (100 mM) were PIPES (pH 6.0, 6.5, 7.0 and 7.5; filled triangle), HEPES (pH 7.0, 7.5 and 8.0; filled circle), Tris/HCl (pH 8.0, 8.5 and 9.0; both filled and open inverted triangle), glycine (pH 9.0, 9.5 and 10.0; open square), CAPS (pH 10.0, 10.5, 11.0 and 11.5; open circle), and phosphate (pH 12.0; open diamond). The relative activity of 100% equals to 69 and 71 U mg−1 for the oxidation of 1-butanol and the reduction of butyraldehyde, respectively

T. hypogea ADH was able to transform a broad range of primary alcohols and diols, but secondary alcohols such as 2-propanol, 2-butanol and polyols such as glycerol were not oxidized, and acetone was not reduced (Table 2). Therefore, this enzyme can be classified as a primary ADH. The purified T. hypogea ADH was NADP(H) specific, a characteristic of Fe-containing ADHs in other hyperthermophiles (Radianingtyas and Wright 2003). The apparent Km value for NADPH was more than three times lower than that of the NADP (Table 3). Further more, the apparent Km values for ethanol and 1-butanol were approximately three to four times higher than those for acetaldehyde and butyraldehyde (Table 3). Specificity constant kcat/Km for aldehyde reduction using NADPH as electron donor (8.5 × 106 s−1 M−1) was also more than three times higher than that of alcohol oxidation using NADP as electron acceptor (2.5 × 106 s−1 M−1). These catalytic properties suggest that the enzyme can play an important role in reduction of aldehydes, rather than the in vivo oxidation of alcohols.
Table 2

Substrate specificity of the purified ADH from T. hypogea


Relative activity (%)






100 ± 3


298 ± 6


400 ± 47


287 ± 3

 Hexyl alcohol

245 ± 3


210 ± 19


190 ± 3






132 ± 10


63 ± 5


91 ± 7


147 ± 4

Aldehydes or ketones



100 ± 6


307 ± 7




362 ± 12

aThe concentration of substrates used was 60 mM. The relative activity of 100% corresponds to 17.3 ± 0.5 U mg−1 for the alcohol oxidation at pH 11.0 and 19.7 ± 1.1 U mg−1 for the aldehyde reduction at pH 8.0, respectively

Table 3

Enzyme kinetics parameters of ADH from T. hypogea

Substrate (mM)

Co-substrate (mM)

Apparent Km (mM)

Apparent Vmax (U mg−1)

kcat (s−1)

kcat/Km (s−1 M−1)

NADPH (0.007–0.092)a

Butyraldehyde (22)

0.006 ± 0.001

75.8 ± 2.2



Butyraldehyde (0.05–22)a

NADPH (0.1)

0.45 ± 0.06

73.1 ± 2.1



Acetylaldehyde (1–42)a

NADPH (0.1)

3.10 ± 0.04

21.8 ± 1.5



NADP (0.009–0.380)b

Butanol (20)

0.020 ± 0.002

75.8 ± 3.1



Butanol (1–20)b

NADP (0.2)

1.90 ± 0.05

73.5 ± 2.6



Ethanol (4–120)b

NADP (0.2)

9.7 ± 0.1

21.5 ± 1.7



aAssays were performed at pH 8.0

bAssays were performed at pH 11.0

Effects of metal ions and thiol reagents on the T. hypogea ADH activity

It was observed that less iron and more zinc present in the purified sample correlated to lower specific activity, and the lowest iron content and the highest zinc content resulted from the aerobically prepared samples (Table 4), indicating that iron was lost more readily in the presence of oxygen and so was the activity. The enzyme activities increased when the iron content was higher in the samples up to 1 atom/subunit. In contrast, an increase in zinc content in the enzyme (up to 0.5 g-atoms/subunit or Zn/Fe ratio >5) was at the expense of a decrease in its iron content and, hence, its activity. This suggests that the extraneous zinc can substitute iron that may be dissociated after reaction with oxygen. The decrease of iron content in the enzyme sample prepared anaerobically in the presence of 2 mM dithiothreitol and 1 mM EDTA (Table 4) suggests that iron-enzyme binding affinity is only strong enough to survive the purification procedures.
Table 4

Metal contents and activities of the purified ADH from T. hypogea

Treatment of samples

Protein concentration (mg ml−1)

Specific activity (U mg−1)a

Metals (g-atoms/subunit)




0.100 ± 0.005b

70.4 ± 2.1

1.02 ± 0.06

0.08 ± 0.02

0.060 ± 0.002c

59.1 ± 1.9

0.60 ± 0.02

0.02 ± 0.01

0.040 ± 0.002

53.7 ± 1.5

0.76 ± 0.03

0.11 ± 0.02

0.045 ± 0.001

31.6 ± 0.2

0.28 ± 0.04

0.31 ± 0.11

0.080 ± 0.003

20.3 ± 0.6

0.21 ± 0.03

0.34 ± 0.16

0.040 ± 0.001

18.1 ± 0.5

0.12 ± 0.01

0.46 ± 0.06


0.057 ± 0.002

3.6 ± 0.2

0.18 ± 0.01

0.47 ± 0.09

0.210 ± 0.005

3.1 ± 0.1

0.09 ± 0.01

0.49 ± 0.06

aThe purified ADH had a specific activity of 68.0 ± 2.2 U mg−1, which was kept in 50 mM Tris/HCl, pH 7.8, containing 2 mM dithiothreitol, 2 mM sodium dithionite and 5% (w/v) glycerol; however, the values in the table were determined from samples that were washed ten times using ultra-pure buffer without any reducing reagent (10 mM Tris/HCl, pH 7.8 unless otherwise specified) under anaerobic (in the anaerobic chamber) and aerobic (in the air) conditions, respectively

bThe washing buffer contained 10 mM Tris/HCl (pH 7.8) and 2 mM dithiothreitol

cThe enzyme sample was incubated with buffer containing 50 mM Tris/HCl (pH 7.8), 5% (w/v) glycerol, 2 mM sodium dithionite, 2 mM dithiothreitol and 1 mM EDTA at room temperature for 1 h before washing using anaerobic buffer containing 10 mM Tris/HCl (pH 7.8) and 2 mM dithiothreitol

When pre-incubated with the enzyme, EDTA, ZnCl2, CuCl2, CdCl2 and HgCl2 inhibited the activity by 80–100%, but dithiothreitol, mercaptoethanol and FeCl2 stimulated the activity by 10–15%, while CaCl2 did not have any effect on the activity. When added to the enzyme assay mixture, ZnCl2 and Zn(SO4)2 inhibited the activity completely; however, such inhibition was reduced to about 20% when dithiothreitol was also present in the assay mixture, which was similar to an inhibitory effect achieved by ZnCl2 alone at a much lower concentration (0.02 mM); EDTA, CuCl2 (0.02 mM) and CdCl2 inhibited enzyme activity by 20–60%, and FeCl2, CoCl2, NiCl2, MnCl2 stimulated the enzyme activity by 20–40%, while dithiothreitol, mercaptoethanol, MgCl2 and CaCl2 showed almost no effect on the activity (data not showed). These results indicate that Fe2+ is a required metal ion for the enzyme activity, while Zn2+ inhibits the catalytic activity. The effects of both thiol reagents, dithiothreitol and mercaptoethanol, could be seen only when they were incubated with the enzyme for a period of time (10–60 min) before the assay, and the Zn2+ inhibition could be reduced by dithiothreitol, indicating that thiol reagents may have a role in decreasing interaction of toxic metal ions such as Zn2+ with the enzyme.

Oxygen sensitivity and recoverability of oxygen-inactivated T. hypogea ADH

The purified enzyme lost about 80% of its activity within the first 0.5 and 3 h of exposure to air in the absence and presence of dithiothreitol and sodium dithionite, respectively (Fig. 4). This suggests that the enzyme is very oxygen sensitive and that the reducing agents can slow down such oxidation processes.
Fig. 4

Oxygen sensitivity of the purified ADH in T. hypogea. Filled circles in the presence of 2 mM dithiothreitol and 2 mM sodium dithionite, open circles in the absence of dithiothreitol and sodium dithionite. The relative activity of 100% equals to the ADH activity prior to exposure to air (58 U mg−1)

The lost activity of oxygen-inactivated T. hypogea ADH could not be recovered merely by removing oxygen from the environment. But, the lost activity was partially recovered by adding either dithiothreitol or Fe2+ and fully recovered by adding both dithiothreitol and Fe2+ under anaerobic conditions for about 1 h. Addition of sodium dithionite did not result in any recovery of the lost enzyme activity. This may further indicate that Fe2+ is the required metal ion for the catalytic activity and dithiothreitol may help complete a full binding of Fe2+ by not only providing the reducing equivalents that are required, but also decreasing the possible inhibitory effect of other metal ions such as Zn2+.

The recoverability and rate of recovery of the oxygen-inactivated enzyme were dependent on the duration of enzyme exposure to air. The lost activity of the enzyme exposed to air for 0.5 and 5 h could be recovered within 20 and 120 min, respectively. It appeared that the shorter the time of exposure to air, the faster the rate of recovery. However, only 50% of the lost activity of the enzyme exposed to air for 24 h could be recovered, even with an extended incubation time to 5 h. It was also observed that the recovery rate for the 66-h exposure to air at 4°C was faster than that of the 24-h exposure to air at room temperature, which might be caused by the slow oxidation of ferrous iron at low temperatures. Obviously, the precise mechanism of oxygen inactivation and reactivation are not known, which warrants further investigation.


ADHs are present in all three domains of life, and they can be classified into three groups based on their molecular properties and metal content (Reid and Fewson 1994). Group I contains long-chain ADHs whose size varies from approximately 350 to 900 amino acid residues and zinc is at their catalytic site and, sometimes, it also has a structural function (Littlechild et al. 2004). Group II contains short-chain ADHs with approximately 250 amino acid residues and lacks metals (Reid and Fewson 1994). Group III consists only of a small number of iron-dependent ADHs including examples of mesophilic iron-activated ADH2 from Zymomonas mobilis (Scopes 1983) and hyperthermophilic iron-containing ADH from Thermococcus strain ES-1 (Ma et al. 1995). All groups of ADHs are found to be present in hyperthermophiles (Radianingtyas and Wright 2003). Apparently, group III has a dominant number of iron-containing ADHs in hyperthermophilic archaea (Table 5). T. hypogea enzyme represents the first hyperthermophilic bacterial ADH that contains iron with full activity after purification, whose catalytic properties show similarities to the enzyme in archaea (Ma et al. 1995). The bacterial T. hypogea ADH is one of the most thermostable iron-containing ADHs known (Table 5). The N-terminal sequence of T. hypogea ADH has no similarity to archaeal ADHs, an indication of the divergence of iron-containing ADHs from hyper/thermophiles. However, more sequence information is needed for understanding the evolutionary relationship between the bacterial iron-containing ADH and the archaeal ADHs.
Table 5

Comparison of iron-containing ADHs from hyperthermophilic microorganisms


T. hypogea

T. hydrothermalis

Thermococcus strain ES−1

T. litoralis

P. furiosus

Subunit molecular mass (structure, kDa)

40 (α2)

45 (α4 or α2)

46 (α4)

48 (α4)

48 (α6)

Metal (g-atoms/subunit)

Fe (1.0)

Fe (0.5)

Fe (0.95)

Fe (0.45)

Fe (0.9), Zn (0.8)







App. Km for alcohol (mM)

1.9 (1-Butanol)

9.7 (Ethanol)

2.0 (Benzylalcohol)

8.0 (Ethanol)

11.0 (Ethanol)

29.4 (Ethanol)

App. Km for aldehyde (mM)

0.45 (Butyraldehyde)

3.1 (Acetyraldehyde)

0.01 (Benzaldehyde)

0.25 (Acetaldehyde)

0.4 (Acetaldehyde)

0.17 (Acetaldehyde)

App. kcat (s−1)

48 (Butanol)

14 (Ethanol)

23 (Benzyl alcohol)

48 (Ethanol)

26 (Ethanol)

19 (Ethanol)

48 (Butyraldehyde)

15 (Acetaldehyde)

2.1 (Benzaldehyde)

19 (Acetaldehyde)


5.5 (Acetaldehyde)

App. kcat/App. Km (s−1 M−1)

2.7 × 104 (1-Butanol)

1.1 × 105 (Butyraldehyde)

1.2 × 104 (Benzyl alcohol)

6.0 × 103 (Ethanol) 1.1 × 105

2.3 × 103 (Ethanol)

6.5 × 102 (Ethanol)

1.5 × 103 (Ethanol)

5.0 × 103 (Acetaldehyde)

2.1 × 105 (Benzaldehyde)

7.6 × 104 (Acetaldehyde)


3.2 × 104 (Acetaldehyde)

Optimal pH for alcohol oxidation






Optimal pH for aldehyde reduction






Optimal temperature (°C)






Stability (t1/2, h)

10 (70°C), 2 (90°C)

0.25 (80°C)

35 (85°C), 4 (95°C)

5 (85°C), 0.3 (96°C)

160 (85°C), 7 (95°C)


This work

Antoine et al. (1999)

Ma et al. (1995)

Ma et al. (1994)

Ma and Adams (1999)

A recombinant 1,3-propanediol dehydrogenase in T. maritima was reported to contain iron by X-ray crystallography analysis, but no further catalytical properties are available (Schwarzenbacher et al. 2004)

ND not determined

The inhibition of T. hypogea ADH activity by zinc is similar to a sulfur-regulated iron-containing ADH in Thermococcus strain ES-1 (Ma et al. 1995). But, the mechanism of zinc inhibition in hyperthermophilic enzymes is not understood yet. It seems that iron in the bacterial T. hypogea enzyme is more easily lost compared to the archaeal-type enzyme in the absence of reducing reagent dithiothreitol, and iron substitution by zinc is also higher under aerobic conditions (Table 4). It is plausible to speculate that the loss of activity could be partially attributed to oxidation of Fe2+ to Fe3+ in the absence of the reducing agents, and Fe3+ may have less affinity to the iron-binding site of the enzyme, so that zinc may substitute the Fe3+ at a greater rate under aerobic conditions.

Iron-containing ADHs prefer to use NADPH as a coenzyme, which is similar to other ADHs characterized from Pyrococcus furiosus and Thermococcus spp. (Table 5), albeit it is contrary to the usual use of NAD(H) in catabolic pathways. These ADHs have a very broad substrate specificity to utilize a series of aliphatic primary alcohols (C2–C8), diols (C3–C5) and aromatic primary alcohol (2-phenylethanol), except for secondary alcohols. The ADHs have much lower Km values for aldehydes. All organisms that possess such ADH produce ethanol as an end product (Kengen et al. 1994; Ma et al. 1995). This is despite the fact that T. hypogea ADH has a relatively high value of kcat/Km for butanol, while archaeal ones show preference for benzyl alcohol or ethanol (Table 5). It has been proposed that these ADHs are responsible for the alcohol formation using the excess of reducing equivalents generated during their growth (Ma et al. 1995), which can also be supported by the optimal pH for aldehyde reduction of 8.0 that is closer to the pH values in the cells. This is true for T. hypogea ADH that catalyzes the aldehyde reduction much faster than the alcohol oxidation at pH 8.0. Among the organisms that have ADHs with very high N-terminal sequence similarity to T. hypogea ADH, F. nodosum (Patel et al. 1985), T. ethanolicus (Wiegel and Ljungdahl 1981) and T. maritima (Ying and Ma, University of Waterloo, unpublished results) produce ethanol. It is plausible to predict that S. thermophilum and T. petrophila can produce ethanol because they may have the same type of ADH, and Thermotoga lettingae may have a similar type of ADH because it produces ethanol and is a member of Thermotoga (Balk et al. 2002). Such speculation needs to be proved experimentally in the future.

It is not uncommon that the iron-containing enzyme can be reactivated by incubation with ferrous iron, for example, fumarase in Bacteroides thetaiotaomicron was restored to its full activity by ferrous ion and aconitase in the same organism was recovered by the treatment of iron and dithiothreitol (Pan and Imlay 2001). Both enzymes contain iron-sulfur clusters, which are vulnerable to oxygen stress. It is unlikely that the purified T. hypogea enzyme contains any iron-sulfur cluster because of the low content of iron present in the enzyme (1.02 ± 0.06 g-atoms/subunit). Further structural and spectroscopic analyses and comparison of protein sequences can be useful for understanding mechanisms of catalysis and activation of T. hypogea ADH.

In addition to its principal role in alcohol production, the activity of iron-containing ADHs is found to be associated with other physiological processes. In hyperthermophiles, the production of the ADH in Thermococcus strain ES-1 is regulated by the concentration of elemental sulfur, and the ADH activity is much higher when S0 concentration is low (Ma et al. 1995). The regulation of the hyperthermophilic bacterial T. hypogea ADH production is not known and needs to be further investigated.


This work was supported by research grants from Ontario Ministry of Agriculture and Food, Rural Affairs, Natural Sciences and Engineering Research Council (Canada) and Canada Foundation for Innovation, and funds from the University of Waterloo to KM. We thank Feng Zhang for helping grow T. hypogea and Xianqin Yang for measuring glutamate dehydrogenase activities.

Copyright information

© Springer-Verlag 2007