OUP user menu

Acetaldehyde dehydrogenase activity of the AdhE protein of Escherichia coli is inhibited by intermediates in ubiquinone synthesis

Shashi Gupta, Fairoz Mat-Jan, Maryam Latifi, David P. Clark
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb08872.x 51-55 First published online: 1 January 2000


Defects in the acd gene (which may be allelic to ubiH) result in the inactivation of the coenzyme A-linked acetaldehyde dehydrogenase activity of the multifunctional AdhE protein of Escherichia coli. This activity is restored by addition of ubiquinone-0 to cell extracts. However, the alcohol dehydrogenase activity of the AdhE protein is not decreased by an acd mutation. Abolition of ubiquinone biosynthesis by mutation of ubiA or ubiF does not affect either the acetaldehyde dehydrogenase or the alcohol dehydrogenase activity of AdhE. Guaiacol (2-methoxyphenol), which resembles the intermediate that builds up in ubiH mutants, except in lacking the octaprenyl side-chain, was found to inhibit ethanol metabolism in vivo, presumably via inhibition of acetaldehyde dehydrogenase. In vitro assays confirmed that guaiacol inhibited acetaldehyde dehydrogenase. This suggests that the acetaldehyde dehydrogenase activity of AdhE is specifically inhibited by intermediates of ubiquinone synthesis that accumulate in acd mutants and that this inhibition may be relieved by ubiquinone.

  • Ethanol
  • Anaerobic metabolism
  • Alcohol dehydrogenase
  • Guaiacol
  • Fermentation

1 Introduction

The AdhE protein of Escherichia coli is responsible for three distinct enzymatic activities. Two of these, alcohol dehydrogenase (ADH) and coenzyme A-linked acetaldehyde dehydrogenase (ACDH), are involved in the conversion of acetyl-CoA to ethanol during fermentation [1]. The third activity, pyruvate formate lyase (PFL) deactivase, is regulatory and serves to inactivate PFL when aerobic conditions prevail [2]. Some time ago we isolated a mutation, designated acd, in which the ACDH activity was lacking but the ADH activity was intact [3]. At that time we did not know that the same protein, AdhE, was responsible for both enzyme activities. Later work showed that most mutants lacking either ACDH or ADH activities of the AdhE protein mapped in the adhE gene at 27.9 min [1,4]. The acd mutation, located at 65–66 min, remained an anomaly. Investigation of the growth properties of the acd mutant, strain DC349, showed that it grew well on glucose and other sugars but was unable to grow on succinate, lactate and acetate. In addition, this strain was resistant to kanamycin and neomycin [4]. These properties are characteristic of mutants lacking ubiquinone or ATP synthase [5] and suggested that acd might be allelic with ubiH, which is located in the same region of the genetic map [6].

2 Materials and methods

2.1 Bacterial strains and culture media

All bacteria used were strains of E. coli K-12, and are described in Table 1. Minimal medium was M9 [7]. Rich broth contained (per liter), tryptone (10 g), NaCl (5 g) and yeast extract (1 g). Minimal medium was supplemented with carbon sources at 0.4% (w/v). Solid media contained 1.5% (w/v) Difco Bacto-agar. Alcohol indicator agar [8] contained M9 minimal medium plus Difco Proteose Peptone #3 (2.0 g l−1), triphenyltetrazolium chloride (0.025 g l−1), and ethanol (5.0 g l−1). Anaerobic minimal media were supplemented with trace metals (50 μM Fe; 5 μM Mo; 5 μM Mn; 5 μM Zn; 5 μM Se). Aerobic cultures (200 ml) were grown in 500-ml conical flasks in a New Brunswick gyratory shaking water bath. Growth was followed turbidimetrically using a Klett-Summerson colorimeter equipped with a green (540 nm) filter. Anaerobic growth was performed in anaerobic jars (Oxoid Ltd., London, UK) under an atmosphere of H2/CO2 generated by Oxoid Gas Generating Kits. Resazurin indicators (Oxoid Ltd., London, UK) were used to ensure anaerobic conditions. Anaerobic liquid cultures were grown without agitation in 160-ml milk dilution bottles filled to the top before sealing [9]. Transductions using P1vir were performed essentially as described by Miller [7]. Transductants were selected on medium E [10] containing glucose (0.4% w/v) plus 0.1% casein hydrolysate. Tetracycline was used at 10 mg l−1 to select for the presence of Tn10. Kanamycin was used at 30 mg l−1 in plates containing succinate (0.4% w/v) plus 0.1% casein hydrolysate.

View this table:
Table 1

Strains of E. coli used

StrainRelevant genotypeSource/reference
CL556aroD::Tn10R. Meganathan
DC80fadR mel adhC80 aceF10[8]
DC271fadR mel[3]
DC272fadR mel adhC81[3]
DC349fadR mel adhC81 acdA1[3]
DC489fadR mel adhC81ΔaceEFLaboratory collection
DC1192zch::Tn10 adhC81 of JF496P1 (DC300)×JF496
DC1335ubiA::Cam of DC489P1 (MU1227)×DC489
DC1365aroD::Tn10 of DC271P1 (CL556)×DC271
DC1366aroD::Tn10 of DC272P1 (CL556)×DC272
DC1391aroD::Tn10 ldhA::Kan of DC271P1 (NZN111)×DC1365
DC1392aroD::Tn10 ldhA::Kan of DC272P1 (NZN111)×DC1366
DC1402ubiA::Cam of DC272P1 (MU1227)×DC272
GD1ubiG::Kan zei::Tn10dTetC.F. Clarke [18]
JC7623Δ4-1ubiE::Kan thr leu proA2 his thi argE rpsLC.F. Clarke [17]
JF496ubiF411 metB asnA asnB nagB bglRCGSC6057
LCB320thr-1 leu-6 thi-1 lacY tonA22 rpsL[15]
MC4100Δ(argF-lac)U169 rpsL thiA relA araDM. Casabadan
MU1227ubiA::CamM. Kawamukai
NZN111ΔpflAB::Cam ldhA::Kan[15]
SHH642acd-2, Kanr ACDH of DC272see text
SHH643acdA3, Kanr ACDH of DC272see text
SHH644acd-4, Kanr ACDH of DC272see text
W1485wild-typeLaboratory collection

2.2 Enzyme assays

Soluble cell extracts for enzyme assays were made by growing 200-ml batch cultures to about 109 cells ml−1. Cells were harvested by centrifugation at 8000 rpm for 10 min at 4°C. The supernatant was decanted. After washing the cell pellet in 50 mM KH2PO4 buffer, pH 7.4, the cells were resuspended in 3.0 ml phosphate buffer and were ruptured by passage through a French pressure cell (American Instrument Co., Silver Springs, MD). The high speed supernatant fraction was obtained by centrifugation at 145 000×g for 1 h at 4°C. The cell extract was stored at −20°C until used. Protein content was assayed by the method of Bradford [11] using the Bio-Rad reagent. Alcohol dehydrogenase was assayed spectrophotometrically by following the reduction of NAD+ to NADH at 340 nm with a Perkin Elmer 552A UV/Vis spectrophotometer. In a 1-ml cuvette, 900 μl of 12 mM sodium pyrophosphate (pH 8.5), 50 μl 1.5 mM NAD, and 30 μl of soluble cell-free extract were mixed. The assay was started by adding 20 μl of absolute EtOH. Acetaldehyde dehydrogenase was also assayed by observing the reduction of NAD+ to NADH coupled to the conversion of acetaldehyde to acetyl-CoA. To 800 μl of 50 mM CHES buffer (pH 9.5) were added 50 μl of 4 mM dithiothreitol, 50 μl of 1.5 mM NAD, 50 μl of 2 mM CoASH and 40 μl of enzyme preparation. To start the reaction 10 μl of 1 M acetaldehyde was added [3]. For both ADH and ACDH, one unit of enzyme activity is defined as 1 nmol of NADH formed per minute [8].

3 Results and discussion

3.1 Isolation of more acd mutations and effect of ubiquinone

The lack of ACDH but presence of ADH activity in the acd mutant, DC349, implied that the AdhE protein was being produced but that its ACDH activity was somehow impaired. Strain DC349 was unable to grow on respiratory carbon sources and was resistant to kanamycin [4]. This suggested that the defect in DC349 was in the respiratory chain. As ubiH is located very close to where acd mapped, we thought it possible that acd is an allele of ubiH. Although lacking ACDH activity in air, DC349 grew well on sugars anaerobically. This fits with the observation that the hydroxylation step catalyzed by the UbiH enzyme uses molecular oxygen and is performed by an alternative enzyme anaerobically [13]. Later, we found that adding micromolar amounts of ubiquinone-0 (2,3-dimethoxy-5-methyl-1,4-benzoquinone, i.e. ubiquinone without its hydrophobic tail) to cell extracts resulted in recovery of the ACDH activity in DC349 (data not shown). Unfortunately the acd mutation of DC349 was unstable and reverted at a high frequency, making consistent analyses difficult. We therefore isolated more acd-type mutants in order to confirm the effect of ubiquinone.

The adhC strain DC272 expresses high levels of AdhE allowing it to use ethanol as carbon source in air and it therefore forms red colonies on alcohol indicator agar. Strain DC272 was mutagenized with ethyl methane sulfonate (EMS) and plated for single colonies on alcohol indicator agar containing kanamycin (30 μg ml−1). Colonies that were both white, implying defective ethanol metabolism, and kanamycin-resistant were screened for growth properties resembling the original acd-1 mutant, DC349 [4]. Three mutants, SHH642, SHH643, and SHH644, were isolated which grew aerobically on glucose but not on succinate alone. Like DC349 they grew on succinate supplemented with casein hydrolysate or proteose peptone. Anaerobically all three, SHH642, SHH643, and SHH644, were able to grow on sugars or sugar alcohols as sole carbon source, as does DC349. (Note that mutations in the adhE gene that inactivate only the ACDH activity of AdhE are known [4]. These differ from acd mutants in being unable to grow anaerobically on sugars or sugar alcohols as sole carbon source.)

The ADH and ACDH activities of SHH642, SHH643, and SHH644 were assayed and all three were found to have 5% or less of the ACDH activity of their parent, DC272. When 75 μM ubiquinone-0 was added, the ACDH activity recovered to about 50% of the parental level (Table 2). Addition of the same concentration of menadione had no effect (data not shown) indicating that the effect is specific for ubiquinone. Cotransduction with Tn10 insertions located close to ubiH, ubiAC, ubiBDE, ubiF, and ubiX indicated that the acd-3 mutation in SHH643 was very close to the ubiH locus like that of DC349 (data not shown). However, the mutations in SHH642 and SHH644 were not at any of the tested locations. If acdA really is allelic to ubiH, it is conceivable that these other acd-type mutants carry defects in the still genetically uncharacterized step preceding ubiH[5]. We have provisionally designated the acd mutations of DC349 and SHH643 that map close to ubiH as acdA.

View this table:
Table 2

Effect of ubiquinone-0 on ACDH activities of acd mutants

StrainAlcohol dehydrogenaseAcetaldehyde dehydrogenase
All cultures were grown aerobically in M9 plus glucose (0.4% w/v) and casein hydrolysate (0.1% w/v). Cells were broken and enzymes assayed as described in Section 2. Enzyme activities are expressed in nmol of NADH produced min−1 mg−1 protein. Ubiquinone-0 (UQ0) was added at 75 μM where indicated.

3.2 Effect of mutations blocking ubiquinone biosynthesis

The recovery of ACDH activity in the acd mutants, upon provision of ubiquinone-0, suggested that ubiquinone is necessary for the action of ACDH. If so, other ubiquinone-deficient mutants should also lack ACDH activity. We therefore transduced the adhC mutation that allows aerobic expression of high levels of AdhE into the ubiF mutant, JF496. Several adhC transductants and their parent JF496 were grown aerobically in rich broth and assayed for ADH and ACDH. As expected, the parent, JF496, which lacks the adhC mutation, showed no detectable ADH or ACDH activity when grown in air. The adhC derivative, DC1192, expressed 50% of the ADH and 39% of the ACDH activity seen in the donor of the adhC mutation, strain DC272 (see Table 2 for DC272). Several other adhC derivatives of JF496 gave similar values (data not shown). There was no significant deficit of ACDH activity relative to ADH as seen in the acd mutants. Furthermore, when ubiquinone-0 was added to cell extracts of DC1192 it had no effect on the ACDH activity (data not shown).

Since JF496 carries an ubiF point mutation it remained possible that small but sufficient amounts of ubiquinone were made to allow ACDH activity. We therefore constructed strain DC1402 by moving the ubiA::Cam insertion mutation [12] into DC272 and assayed for ADH and ACDH activity as before. The ACDH level seen in DC1402 (2.4 nmol NADH min−1 mg−1 protein) was 39.3% of that in its parent DC272 (6.1 nmol NADH min−1 mg−1 protein); however, the level of ADH activity was similarly reduced. Thus, although ubiA knockout strains have less overall AdhE protein, they have no specific defect in their ACDH activity.

3.3 Effect upon growth of mutations blocking ubiquinone biosynthesis

If ubiquinone is necessary for the action of ACDH then mutants lacking ubiquinone should fail to grow anaerobically on hexoses or sugar alcohols, such as sorbitol [1]. Although none of DC349, SHH642, SHH643, or SHH642 showed any anaerobic growth defects, the hydroxylation step catalyzed by the UbiH enzyme aerobically uses molecular oxygen and is performed by an alternative enzyme anaerobically [13]. We therefore tested mutations that prevent ubiquinone synthesis irrespective of aeration.

The AroD protein makes shikimic acid, the common precursor to aromatic amino acids and cofactors including ubiquinone [14]. An aroD mutant will grow on sugars aerobically if supplied with tyrosine, phenylalanine, tryptophan and p-aminobenzoate (for folate synthesis). It cannot respire aerobically unless provided with p-hydroxybenzoate, the precursor for ubiquinone synthesis [14]. We transduced aroD::Tn10 into both DC271 and its adhC derivative, DC272. The resulting pair of strains, DC1365 and DC1366, grew well anaerobically on glucose, mannose, sorbitol, mannitol and several other sugar derivatives irrespective of the presence or absence of p-hydroxybenzoate. To eliminate the possibility of growth by lactate fermentation we introduced the ldhA::Kan mutation [15] into DC1365 and DC1366, so giving DC1391 and DC1392 which no longer make lactate dehydrogenase. This pair of strains also grew well anaerobically on sugars and sugar alcohols and the presence or absence of p-hydroxybenzoate had no effect. Aerobically, deprivation of p-hydroxybenzoate prevented the aroD mutants DC1365, DC1366, DC1391 and DC1392 from growing on the respiratory substrate succinate, but allowed growth on glucose. This confirmed that p-hydroxybenzoate starvation does indeed prevent synthesis of ubiquinone in these strains. (Incidentally, the aerobic growth of DC1391 and DC1392 on glucose casts doubt on the supposed requirement of lactate fermentation for respiration defective strains to grow in air [16].)

We also tested knockout mutants in the ubiquinone biosynthetic pathway. Strains MU1227 (ubiA::Cam) [12], JC7623Δ4-1 (ubiE::Kan) [17], and GD1 (ubiG::Kan) [18] all grew well anaerobically on both sugars and sugar alcohols as sole carbon sources. As expected, they were unable to grow aerobically on respiratory substrates such as succinate. Since ubiA comes several steps before ubiH in the ubiquinone pathway [5], we transduced ubiA::Cam from MU1227 into several other strains. Derivatives of W1485, LCB320, DC271 and DC272 carrying ubiA::Cam grew as well anaerobically on glucose, mannose, sorbitol and mannitol as their parents. Thus ubiquinone is not needed for ethanol fermentation, in particular it is not required for the ACDH reaction which is an essential part of this pathway.

Although unlikely, it was conceivable that ubiquinone was required for in vivo ACDH function only under aerobic conditions. This was tested as follows. Strain DC489 carries an ΔaceEF deletion and so lacks pyruvate dehydrogenase and cannot grow aerobically unless supplied with a source of acetate. Due to its adhC mutation, DC489 expresses a high level of AdhE aerobically which allows it to convert ethanol to acetyl CoA and thus supply its acetate requirement. We moved the ubiA::Cam insertion into DC489, to give DC1335. Both DC1335 and DC489 grew on glucose plus acetate or on glucose plus ethanol but failed to grow on glucose alone. Thus DC1335 was able to use ethanol as an acetate source like its parent DC489, indicating that both the ADH and ACDH reactions of the AdhE protein were still functional in vivo aerobically in an ubiA::Cam knockout strain.

3.4 Inhibition of alcohol metabolism by guaiacol

Apart from lacking the octaprenyl side chain, 2-methoxyphenol, commonly known as guaiacol, is identical to the intermediate just preceding the UbiH step of the ubiquinone pathway. We therefore tested the effect of guaiacol on ethanol metabolism. Strains of E. coli, with adhC mutations, such as DC272, can grow on ethanol as sole carbon source in air, due to high levels of AdhE [8]. We found that 7.3 mM guaiacol prevented growth of DC272 on ethanol but that 21.8 mM guaiacol was needed to stop growth on other carbon sources, such as acetate or succinate (Table 3). We also tested the effect of guaiacol on ethanol supplementation of strains DC489 and DC80 which require an exogenous source of acetate due to lack of pyruvate dehydrogenase. Both are adhC and so can use ethanol as an acetate source. Both strains were more sensitive to guaiacol when supplied with ethanol than when given acetate itself (Table 3). Strain DC80 was much more sensitive than DC489, presumably because the adhC80 mutation only results in the expression of about 20% as much AdhE as the adhC81 mutation of DC489 and DC272 [8].

View this table:
Table 3

Effect of guaiacol on alcohol metabolism

StrainCarbon sourceaGuaiacol (mM)
All carbon sources were tested in M9 minimal agar. Carbon sources were present at 0.4% (w/v). Acetate or ethanol used as supplements were present at 0.1% (w/v). The growth tests with DC272 were done at 30°C as ethanol becomes inhibitory at higher temperatures. The tests with DC80 and DC489 were performed at 37°C. ++, good growth; +, poor but visible growth after 2 days; −, no growth after 3 days.

Using cell extract from strain DC272, we tested the effect of guaiacol upon ACDH activity in vitro. Guaiacol at 14.5 mM, the concentration necessary to prevent growth of DC489 with ethanol as acetate source (Table 3), caused 78% inhibition of ACDH activity and at 36 mM it caused 91% inhibition. Thus concentrations of guaiacol causing in vivo growth inhibition were also severely inhibitory to ACDH enzyme activity in vitro.

4 Conclusion

The visB mutant of E. coli was isolated for sensitivity to visible light and turned out to be in ubiH[19]. When ubi mutations earlier in the pathway than ubiH were introduced, the light sensitivity was abolished, suggesting that build-up of a biosynthetic intermediate was responsible. However, light sensitivity of ubiH mutants is seen only in some genetic backgrounds, and probably requires additional mutations [19]. The acd mutations may have a similar explanation. It is possible that the acd mutations are in ubiH or a nearby uncharacterized gene affecting ubiquinone metabolism. It is clear from the use of knockout mutations that ubiquinone is not required for ACDH activity. Furthermore, guaiacol, which mimics the precursor to the UbiH step, specifically inhibits growth that depends on alcohol metabolism and also inhibits ACDH activity in vitro. It seems likely that ACDH is inhibited by intermediates of ubiquinone synthesis that accumulate in the acd mutants, and that addition of ubiquinone-0 reverses this.


This work was supported by a grant to D.C. from the Department of Energy, Office of Basic Energy Sciences (Contract DE-FG02-88ER13941).


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
View Abstract