View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 580 (2006) 5084–5088

Respiratory behaviour of a adhB::kanr mutant supports the hypothesis of two isoenzymes catalysing opposite reactions

U. Kalnenieksa,*, N. Galininaa, M.M. Tomaa, J.L. Pickfordb, R. Rutkisa, R.K. Pooleb a Institute of Microbiology and Biotechnology, University of Latvia, Kronvalda Boulv. 4, LV-1586, Riga, Latvia b Department of Molecular Biology and Biotechnology, The University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, UK

Received 10 July 2006; revised 9 August 2006; accepted 11 August 2006

Available online 28 August 2006

Edited by Barry Halliwell

However, under aerobic growth conditions, synthe- Abstract Perturbation of the aerobic steady-state in a chemo- stat culture of the ethanol-producing bacterium Zymomonas sis competes for reducing equivalents with respiration. Re- mobilis with a small pulse of ethanol causes a burst of ethanol cently, we found that competition between the two ADH oxidation, although the reactant ratio of the alcohol dehydroge- isoenzymes and the respiratory chain may lead to aerobic stea- nase (ADH) reaction ([NADH][][H+])/([etha- dy-states with unexpected properties. Perturbation of the aer- + nol][NAD ]) remains above the Keq value. Simultaneous obic steady-state in a chemostat culture with a small pulse of catalysis of ethanol synthesis and oxidation by the two ADH iso- ethanol caused a rapid, transient burst of ethanol oxidation, enzymes, residing in different redox microenvironments, has been seen as a rise of acetaldehyde concentration and intracellular proposed previously. In the present study, this hypothesis is ver- NADH [7], and a fall of pO2 in the culture. Notably, a slow ified by construction of an ADH-deficient strain and by demon- net synthesis of ethanol was taking place during the aerobic stration that it lacks the oxidative burst in response to steady-state, in full accordance with the measured bulk reac- perturbation of its aerobic steady-state with ethanol. 2006 Federation of European Biochemical Societies. Pub- tant ratio, which was well above Keq even after addition of lished by Elsevier B.V. All rights reserved. the ethanol pulse. In order to explain our finding, we proposed operation of an ‘‘ethanol cycle’’ – a novel redox-driven futile Keywords: Alcohol dehydrogenase; Continuous culture; cycle, involving both ADH isoenzymes, simultaneously cata- Respiration; Metabolic channelling; Zymomonas mobilis lysing opposite reactions (Fig. 1A). Both reactions were sup- posed to proceed several times faster than the resulting slow net synthesis of ethanol, thus explaining the rapid burst of oxi- dation upon ethanol addition by perturbation of the cycle [7]. It must be noted that a simultaneous catalysis of opposite reac- 1. Introduction tions by two isoenzymes in a homogeneous medium as pic- tured in Fig. 1A would not be thermodynamically feasible Reversible reduction of acetaldehyde to ethanol, which ter- without an external energy supply to one of the reactions. As minates the fermentative pathways of various ethanol-produc- the possibility of energization of the ADH reaction by ATP ing microorganisms, is catalysed by NAD(P)H-dependent or by other metabolically available energy source seems highly unlikely, we proposed that the reaction medium should be het- alcohol dehydrogenases (ADH) [1]. Keq for the ethanol oxida- tion reaction ([NADH][acetaldehyde][H+])/([ethanol][NAD+]) erogeneous. In other words, for the operation of ethanol cycle, is as low as 6.92 · 1012 M, hence the reaction equilibrium is a different redox microenvironment for each reaction must be shifted far towards ethanol synthesis [2]. For the Gram-nega- expected, which in the single compartment of the bacterial cell tive, aerotolerant ethanol producer Zymomonas mobilis, the would imply some kind of enzyme–enzyme interaction and reaction is catalysed by two ADH isoenzymes: the zinc-con- substrate channelling [8,9]. taining ADH I, encoded by adhA, and the iron-containing Although the ethanol cycle hypothesis largely helps to ADH II, encoded by adhB [3–6]. Both of them are NAD+- understand the response of aerobic steady-state culture to per- dependent, and under anaerobic conditions they carry out turbations, it still lacks direct experimental verification. First the rapid and efficient synthesis of ethanol seen in this bacte- of all, it would be necessary to demonstrate that the burst of rium. ethanol oxidation, following the ethanol pulse, does not take place with only one of the ADH isoenzymes present. In partic- ular, the hypothesis proposes that ADH II is the enzyme pri- marily responsible for ethanol oxidation. Studies of this type *Corresponding author. Fax: +371 703 4885. have been hampered by the limited availability of Z. mobilis E-mail address: [email protected] (U. Kalnenieks). knockout mutants of respiratory and fermentative catabolism, obtained by molecular cloning methods. Here, we describe Abbreviations: ADH, alcohol dehydrogenase; E-D, Entner-Doudoroff construction of an ADH II-deficient Z. mobilis strain, with a glycolytic pathway; YETOH, ethanol yield (grams of synthesised etha- kanamycin-resistance gene inserted in the coding region of nol per gram of consumed ); Qglucose, specific rate of glucose consumption (millimoles of consumed glucose per gram of dry weight adhB. We show that the response of the recombinant strain per hour) to ethanol pulse indeed dramatically differs from that of the

0014-5793/$32.00 2006 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2006.08.034 U. Kalnenieks et al. / FEBS Letters 580 (2006) 5084–5088 5085

was used. The Z. mobilis adhB gene (GenBank accession number A Electron transport chain NADH M15398) was amplified by PCR, using primers Zm1 (CGGAATTCAAAATCGGTAACCACATCTCAATTA) and Zm2 (CGCTGCAGTGGCAAATTATTTATGACGGTAGGC) supplied ADH II by Sigma Genosys; the engineered restriction sites for EcoRI and PstI, respectively, are underlined. PCR reactions were carried out in a Ther- moHybaid gradient thermocycler, using Accuzyme DNA polymerase +NAD+ (Bioline). Purification of the obtained PCR products was done with the QIAquick PCR purification kit, while for the recovery of plasmids Ethanol E-D Acetaldehyde and PCR products from agarose gels the QIAquick gel extraction kit +NADH was used (Qiagen). T4 DNA ligase (Fermentas) was used for ligation. ADH I Restriction, ligation and cloning assays essentially followed standard procedures [12]. All constructs were confirmed by DNA sequencing done by Lark Technologies Inc.

2.3. Continuous cultivation B Electron transport chain NADH Continuous cultivation was carried out in a Labfors fermenter (Inf- ors), of 1 L working volume, at 30 C, pH 5.5 and 0.25 h1 dilution rate. The growth medium contained 20 g L1 glucose, 5 g L1 ex- ADH II tract and mineral salts, as described previously [10]. Growth was initi- ated under anaerobic conditions, gassing the culture with nitrogen at 0.4 L min1 flow rate. After anaerobic steady-state was reached, aera- tion of the culture was started, and gradually increased, until aerobic +NAD+ steady-state was established, with air flow 2 L min1 and stirring rate ADH II Ethanol E-D Acetaldehyde 550 r.p.m. The pO2 level was not feedback-regulated by either stirring +NADH speed or air flow; however it was maintained remarkably stable by the ADH I steady-state culture itself (see Section 3).

2.4. Analytical methods Alcohol dehydrogenase activity was estimated in the direction of Fig. 1. The ethanol cycle. (A) ADH I catalysing ethanol synthesis, and ethanol oxidation, as described by Neale et al. [4]. The total ADH ADH II catalysing ethanol oxidation; (B) ADH I and part of ADH II activity and the activity of ADH I were measured spectrophotometri- catalysing ethanol synthesis, with part of ADH II catalysing ethanol cally at 340 nm after transfer of an aliquot (10–20 ll) of the cell suspen- oxidation. sion into a cuvette with 1.5 ml of 30 mM Tris/HCl buffer, pH 8.5, containing 1 mM NAD+. Alcohol dehydrogenase isoenzymes were dif- ferentiated according to Kinoshita et al. [3]. For measurement of the parent strain under similar culture conditions, and that in gen- total activity of both isoenzymes, 1 M ethanol was added to the buffer. eral, ADH II-deficiency causes marked alterations of the respi- Discrimination between the two isoenzyme activities was based on the ratory catabolism of Z. mobilis. fact that only ADH I, but not ADH II, could oxidise butanol. For measurement of the ADH I butanol-oxidising activity, which was ta- ken to be half of the ADH I ethanol-oxidising activity, 200 mM buta- nol was added in place of ethanol. ADH II activity was found by 2. Materials and methods subtraction of the estimated ADH I ethanol-oxidising activity from the total ethanol-oxidising activity. 2.1. Bacterial strains, plasmids and transformation Intracellular NADH was determined luminometrically, as described Escherichia coli JM109 and plasmid pGEM-3Zf(+) were purchased previously [11]. Ethanol concentration was determined by gas chroma- from Promega, and the strain JM109 was used as the host strain for tography (Varian). Acetaldehyde was assayed using the alcohol dehy- cloning of recombinant plasmids. Z. mobilis ATCC 29191 (ZM6) drogenase reaction, and glucose was assayed with the glucose oxidase was grown and maintained as described previously [10,11]. Plasmids, method, as described previously [15]. Cell concentration was deter- constructed and used in the present study, are shown in Table 1. mined as optical density at 550 nm, and dry cell mass of the suspen- sions was calculated by reference to a calibration curve. All results E. coli was transformed by the CaCl2 procedure as described by Sam- brook et al. [12]. Z. mobilis was transformed by electroporation [13]. are means of at least three replicates. Standard errors are given in brackets in Table 2. 2.2. PCR and DNA techniques Amplification and cloning of adhB was done following the approach of Delgado et al. [14], who previously used the same plasmid vector for 3. Results insertion of gfpuv, a reporter gene encoding the green fluorescent pro- tein, into the adhB locus by means of homologous recombination. Genomic DNA from Z. mobilis was isolated using a Promega Wizard 3.1. Construction of an ADH II-deficient strain Genomic DNA purification kit, following the manufacturer’s instruc- The amplified product of 1.7 kb, in contrast to the pub- tions. For plasmid isolation, the QIAprep Spin Miniprep kit (Qiagen) lished sequence of adhB, appeared to contain an additional

Table 1 Plasmids used in the study Plasmid Essential characteristics of genotype Source pGEM-3Zf(+) ampr Promega pUC4K kanr determinant from Tn903 Amersham Biosciences pZADH2 pGEM-3Zf(+) derivative, carrying a 0.6 kb fragment Present work of PCR-amplified adhB gene, cloned between EcoRI and PstI sites of the MCS region pZADH2Tn903 pZADH2 derivative, carrying the kanr determinant Present work from pUC4K, cloned in the EcoRV site of the 0.6 kb fragment of adhB 5086 U. Kalnenieks et al. / FEBS Letters 580 (2006) 5084–5088

Table 2 was able to grow on agar plates in the presence of 40 mg L1 Parameters of aerobic continuous cultivations at 0.25 h1 dilution rate, 1 kanamycin, while in liquid culture medium it tolerated up to 550 r.p.m. stirring speed, and 2 L min air flow 400 mg L1 kanamycin concentration. However, when culti- Parameter Zm6 adhB::kanr vated without kanamycin, the adhB::kanr strain gradually re- 1 X (gdry wt L ) 0.71 (±0.01) 0.58 (±0.01) gained its ADH II activity. A statistically significant rise of pO2 (% of saturation) 43.6 (±0.7) 42.9 (±1.5) ADH II activity was noted after 10 generations in kanamy- 1 Ethanol (g L ) 5.9 (±0.8) 2.5 (±0.2) cin-free growth medium (not shown), indicating a certain selec- Acetaldehyde (g L1) 0.42 (±0.18) 1.33 (±0.42) 1 tive advantage of the wild type phenotype. NADHin ðnmol mgdry wtÞ 1.82 (±0.09) 1.20 (±0.25) 1 YETOH ðggglucoseÞ 0.32 0.22 1 Qglucose (mmol (g dry wt h) ) 36.1 26.8 3.2. Effects of ADH II-deficiency on the properties of the aerobic steady-state restriction site for EcoRI. A 0.6 kb fragment of the obtained Aerobic catabolism of the ADH II-deficient strain and its re- PCR product (localized between the additional EcoRI site sponse to perturbation of steady state with ethanol pulse was and the 30-terminal PstI site) was directionally cloned into studied in aerobic continuous culture and compared to the pGEM-3Z(f+), using the restriction sites for EcoRI and PstI behaviour of the parent strain, ZM6. To obtain physiologically of its multiple cloning site, resulting in plasmid pZADH2. comparable results, both strains were cultivated in a chemostat Plasmid pUC4K, carrying the kanamycin resistance determi- under identical conditions; therefore no kanamycin was added nant from Tn903 [16], was digested with EcoRI, and a 1.3 kb to the ADH II-deficient strain during the experiment. Due to fragment, containing the kanr determinant, was isolated. The the limited stability of the adhB::kanr mutant, all measure- fragment was blunt-ended, using T4 DNA polymerase, and ments and perturbations of the steady state were done within was cloned into pZADH2 in the middle of the 0.6 kb adhB a time span of less than 10 generations. In addition, absence fragment at its EcoRV site, following the protocol of blunt- of ADH II activity in the recombinant strain was verified at end cloning of PCR products by Sambrook and Russell [17], the end of the chemostat experiment. yielding plasmid pZADH2Tn903. The ADH II-deficient strain showed lower steady-state bio- Plasmid pZADH2Tn903 was used to transform Z. mobilis mass concentration, lower specific rate of glucose utilisation by electroporation, and selection for homologous recombi- and lower ethanol yield (Table 2). Although no direct measure- nants was carried out on kanamycin-containing plates. One ments of the respiratory rate were done during the continuous colony was selected, in which the kanamycin-resistance gene culture experiment, the data presented in Table 2 clearly indi- appeared to be inserted in the coding region of adhB. Insertion cate that the specific respiration rate was higher in the recom- was verified by PCR reaction on the genomic DNA template binant strain. This can be concluded from the fact that, under with primers Zm1 and Zm2. The adhB::kanr strain was found identical aeration conditions, the ADH II-deficient culture to be severely deficient in ADH II activity (see inset of Fig. 2), with a lower biomass concentration maintained the same which had reached almost zero level. The recombinant strain pO2 value as that of the parent strain (note that pO2 was not regulated – see Section 2). Higher acetaldehyde concentration in the medium (Table 2) also points to an in- 60 creased respiratory activity of the ADH II-deficient strain. Not only continuous culture, but also exponentially growing batch culture of the ADH II-deficient strain typically had high- 50 2 er respiration rates than ZM6, as measured by an oxygen elec- trode in samples taken from shaken flasks (not shown). Taken together, these data indicate that under aerobic growth condi- 40 1 tions ADH II activity serves to increase the ethanol yield and facilitates fermentative metabolism. 1.2

) On the other hand, perturbation of the steady state with eth- (%)

30 -1 2 1 anol in both cultures showed that the presence of ADH II

pO 0.8 activity was essential for rapid ethanol oxidation. Respiratory 20 0.6 responses to the addition of 10 ml of ethanol (providing an in- crease of ethanol concentration in the fermenter by a step of 0.4 approx. 0.8%) were monitored as changes of the steady-state

ADH (U mg dry wt 0.2 10 pO2 value (Fig. 2). For ZM6, a steep decrease of pO2 followed 0 the ethanol addition, while the ADH II-deficient strain showed 12 an opposite effect – a gradual rise of pO2 a few minutes after 0 ethanol addition, indicating a fall of the respiration rate. The 0 5 10 15 20 25 reason for the fall is not self-evident, but it may be related to Time (min) a slight, transient slowing down of the Entner-Doudoroff (E- D) pathway in response to the added ethanol, causing, accord- Fig. 2. Response of the steady-state pO2value in aerobic continuous ingly, a decrease in the NADH flux from the E-D pathway to r cultures of ZM6 (1) and the recombinant strain adhB::kan (2). The the respiratory chain. In any case, against such a background arrows denote addition of 10 ml absolute ethanol to 1 L of culture. Inset: specific activities of ADH II (hatched bars) and ADH I (black of respiratory slowdown the total contribution of ADH II in bars) in ZM6 (1) and adhB::kanr (2) at the end of continuous the respiratory burst appears to be even greater than the net cultivation. effect seen in ZM6. U. Kalnenieks et al. / FEBS Letters 580 (2006) 5084–5088 5087

4. Discussion unless we assume that some part of the measured ADH II activ- ity does not participate in catalysis of ethanol synthesis in vivo. The presented data clearly do not support a picture of two A simple calculation shows that about 30% of the measured ADH isoenzymes, localized in a homogeneous intracellular ADH II activity in tandem with ADH I would be sufficient to medium, and both operating at the same bulk intracellular produce the observed YETOH in the parent strain. Hence, at concentrations of reactants. It must be emphasized that the least 70% of the ADH II pool (which makes 0.7–0.8 U mg measured bulk reactant ratios for the ADH reaction in both dry wt1 in our experiment) might constitute the ‘‘oxidative’’ strains were closely similar, generally exceeding Keq by 3 orders fraction. Notably, the activity of the ‘‘oxidative’’ ADH II is of magnitude, and hence could not explain the extremely differ- comparable to, or even exceeds the activity of the NADH dehy- ent responses of the strains to the ethanol pulse. Taking the drogenase in Z. mobilis electron transport chain [7]. Therefore, measured values for ethanol, acetaldehyde and intracellular it may be regarded as quantitatively relevant for explanation of NADH concentrations (Table 2), 6.5 for intracellular pH the observed pO2 response to the ethanol pulse. [18], 3.3 ll per mg of dry weight for the intracellular volume The putative role of both fractions of ADH II in maintain- [19], and (for the lowest estimate of the reactant ratios) assum- ing intracellular redox homeostasis and in distribution of ing that NAD+ makes up most of the 5 mM intracellular NADH fluxes between respiration and ethanologenesis in Z. NAD(P)(H) pool, detected in Z. mobilis by NMR [20],we mobilis deserves a detailed study. Special functions for partic- come to reactant ratios of 2.6 · 109 M for ZM6 and ular ADH isoenzymes might be common among ethanologenic 12.6 · 109 M for the ADH II-deficient strain. These are in- microorganisms. An interesting scheme of two alcohol dehy- deed the lowest estimates of the reactant ratios, because: (i) drogenases catalysing opposite reactions has been postulated by assuming the intracellular acetaldehyde/ethanol ratio equal by Burdette and Zeikus [22,23] for anaerobically growing to that of the outer medium, we are neglecting the fact that Thermoanaerobacter ethanolicus. They suggested that one of acetaldehyde is a more volatile compound than ethanol, and the two ADH isoenzymes in this anaerobe was participating hence, the true intracellular acetaldehyde/ethanol ratio might both in ethanol synthesis from acetaldehyde and NADH, be higher; (ii) a lower value for the intracellular volume of and in ethanol oxidation coupled to NADP+ reduction, thus Z. mobilis has been reported by Osman et al. [21], which would maintaining the redox balance between the NAD(P)(H) cofac- result in a higher measured intracellular NADH concentra- tor pools. tions. Furthermore, the ethanol pulse per se could not signifi- Most probably, in aerobically growing Z. mobilis, the oxida- cantly affect the proximity of ADH reaction to equilibrium in tive microenvironment for ADH II could be maintained near either of the strains, as it decreased the reactant ratios only by (or in direct contact with) the respiratory chain. An intriguing a factor of 2 or 3. The estimated bulk reactant ratios thus do problem for future research might be to find out whether the not help to explain the culture responses to perturbation of putative ‘‘respiratory’’ part of the ADH II pool directly inter- their steady states. Therefore, interpretation of our data neces- acts with the respiratory chain, by channelling NADH to a sarily involves concepts of enzyme–enzyme interactions and particular membrane NADH dehydrogenase, and what might substrate channelling. be the regulatory impact of such a mechanism. The interaction In order to account for the observed effects of ADH II-defi- of a soluble protein with a specific component of an aerobic ciency, we suggest a dual role for ADH II in the aerobic catab- respiratory chain is not unprecedented: haemoglobin Vgb of olism of Z. mobilis. Apparently, a limited fraction of ADH II Vitreoscilla binds specifically to subunit I of the cytochrome under aerobic conditions must be operating in a specific, ‘‘oxi- bo-type terminal oxidase [24], perhaps facilitating oxygen dative’’ microenvironment, which should differ very much delivery. from the measured bulk, and should have the local reactant ra- Acknowledgements: This work was supported by an ISIS International tio below Keq, thus enabling oxidation of the added ethanol pulse. At the same time, the rest of the ADH II pool in tandem Fellowship (Ref. 1436) from BBSRC (for UK) and by Grant No. 04.1101 from the Latvian Council of Science. with ADH I contributes to ethanol synthesis (Fig. 1B). Comparison of the ethanol yields (YETOH) in both strains (Table 2) can provide a rough idea of the putative ‘‘oxidative’’ References ADH II fraction. It is well established that the maximum value of YETOH in Z. mobilis catabolism under anaerobic conditions [1] Reid, M.F. and Fewson, C.A. (1994) Molecular characterization is close to 0.51 g of ethanol per gram of consumed glucose of microbial alcohol dehydrogenases. Crit. Rev. Microbiol. 20, [18,21], and that under aerobic growth conditions YETOH is 13–56. lower due to competition for NADH between the respiratory [2] Burton, K. (1974) The enthalpy change for the reduction of nicotinamide–adenine dinucleotide. Biochem. J. 143, 365–368. chain and ethanol synthesis [15]. In the aerobic chemostat cul- 1 [3] Kinoshita, S., Kakizono, T., Kadota, K., Das, K. and Taguchi, ture of the mutant strain, we found YETOH to be 0.22 g g , H. (1985) Purification of two alcohol dehydrogenases from which constitutes 43% of the maximum value. That corre- Zymomonas mobilis and their properties. Appl. Microbiol. Bio- sponds to a nearly equal distribution of the reducing equiva- technol. 22, 249–254. lents between equally active consumers of NADH: ethanol [4] Neale, A.D., Scopes, R.K., Kelly, J.M. and Wettenhall, R.E.H. (1986) The two alcohol dehydrogenases of Zymomonas mobilis. synthesis and respiration. At the same time, in the parent strain Purification by differential dye ligand chromatography, molecular the total activity of ADH (due to the presence of ADH II) is at characterisation and physiological roles. Eur. J. Biochem. 154, least 5-fold higher (Fig. 2), while the specific rate of oxygen con- 119–124. [5] Conway, T., Sewell, G.W., Osman, Y.A. and Ingram, L.O. (1987) sumption is slightly lower. However, the observed YETOH is 1 Cloning and sequencing of the alcohol dehydrogenase II gene only 0.32 g g , or 63% of the maximum value. Apparently, from Zymomonas mobilis. J. Bacteriol. 169, 2591–2597. more than 5-fold excess of the ADH activity over respiration [6] Keshav, K.F., Yomano, L.P., An, H.J. and Ingram, L.O. (1990) should result in YETOH around 80–85% of the maximum value, Cloning of the Zymomonas mobilis structural gene encoding 5088 U. Kalnenieks et al. / FEBS Letters 580 (2006) 5084–5088

alcohol dehydrogenase I (adhA): sequence comparison and [17] Sambrook, J. and Russell, D.W. (2001) Molecular Cloning: A expression in Escherichia coli. J. Bacteriol. 172, 2491–2497. Laboratory Manual, 3rd edn, Cold Spring Harbor Laboratory, [7] Kalnenieks, U., Galinina, N., Toma, M.M. and Marjutina, U. NY. (2002) Ethanol cycle in an ethanologenic bacterium. FEBS Lett. [18] Barrow, K.D., Collins, J.G., Norton, R.S., Rogers, P.L. and 522, 6–8. Smith, G.M. (1984) 31P nuclear magnetic resonance studies of the [8] Ova´di, J. and Saks, V. (2004) On the origin of intracellular fermentation of glucose to ethanol by Zymomonas mobilis. J. Biol. compartmentation and organized metabolic systems. Mol. Cell. Chem. 259, 5711–5716. Biochem. (256/257), 5–12. [19] DiMarco, A.A. and Romano, A.H. (1985) D-glucose transport [9] Huang, X., Holden, H.M. and Raushel, F.M. (2001) Channeling system of Zymomonas mobilis. Appl. Environ. Microbiol. 49, 151– of substrates and intermediates in enzyme-catalyzed reactions. 157. Annu. Rev. Biochem. 70, 149–180. [20] de Graaf, A.A., Striegel, K., Wittig, R.M., Laufer, B., Schmitz, [10] Kalnenieks, U., de Graaf, A.A., Bringer-Meyer, S. and Sahm, H. G., Wiechert, W., Sprenger, G.A. and Sahm, H. (1999) Metabolic (1993) Oxidative phosphorylation in Zymomonas mobilis. Arch. state of Zymomonas mobilis in glucose-, -, and xylose-fed Microbiol. 160, 74–79. continuous cultures as analysed by 13C- and 31P-NMR spectros- [11] Kalnenieks, U., Toma, M.M., Galinina, N. and Poole, R.K. copy. Arch. Microbiol. 171, 371–385. (2003) The paradoxical cyanide-stimulated respiration of Zymo- [21] Osman, Y.A., Conway, T., Bonetti, S.J. and Ingram, L.O. (1987) monas mobilis: cyanide sensitivity of alcohol dehydrogenase Glycolytic flux in Zymomonas mobilis: enzyme and meta- (ADH II). Microbiology 149, 1739–1744. bolite levels during batch fermentation. J. Bacteriol. 169, 3726– [12] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular 3736. Cloning: A Laboratory Manual, 2nd edn, Cold Spring Harbor [22] Burdette, D. and Zeikus, J.G. (1994) Purification of acetaldehyde Laboratory, NY. dehydrogenase and alcohol dehydrogenases from Thermoanae- [13] Liang, C.-C. and Lee, W.-C. (1998) Characteristics and transfor- robacter ethanolicus 39E and characterization of the secondary- mation of Zymomonas mobilis with plasmid pKT230 by electro- alcohol dehydrogenase (2 ADH) as a bifunctional alcohol poration. Bioproc. Eng. 19, 81–85. dehydrogenase-acetyl-CoA reductive thioesterase. Biochem. J. [14] Delgado, O.D., Martinez, M.A., Abate, C.M. and Sin˜eriz, F. 302, 163–170. (2002) Chromosomal integration and expression of green fluores- [23] Burdette, D.S., Jung, S.-H., Shen, G.-J., Hollingworth, R.I. and cent protein in Zymomonas mobilis. Biotechnol. Lett. 24, 1285– Zeikus, J.G. (2002) Physiological function of alcohol dehydro- 1290. genases and long-chain (C30) fatty acids in alcohol tolerance of [15] Kalnenieks, U., Galinina, N., Toma, M.M. and Poole, R.K. Thermoanaerobacter ethanolicus. Appl. Environ. Microbiol. 68, (2000) Cyanide inhibits respiration yet stimulates aerobic growth 1914–1918. of Zymomonas mobilis. Microbiology 146, 1259–1266. [24] Park, K.-W., Kim, K.-J., Howard, A.J., Stark, B.C. and Webster, [16] Taylor, L.A. and Rose, R.E. (1988) A correction in the nucleotide D.A. (2002) Vitreoscilla hemoglobin binds to subunit I of sequence of the Tn903 kanamycin resistance determinant in cytochrome bo ubiquinol oxidases. J. Biol. Chem. 277, 33334– pUC4K. Nucleic Acid. Res. 16, 358. 33337.