APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1984, p. 1277-12830099-2240/84/0601277-07$02.00/0Copyright ©D 1984, American Society for Microbiology

Vol. 47, No. 6

Intermediary Metabolism in Clostridium acetobutylicum: Levels ofEnzymes Involved in the Formation of Acetate and Butyrate

MARIS G. N. HARTMANIS* AND STEN GATENBECK

Department of Biochemistry and Biotechnology, Royal Institute of Technology, S-100 44 Stockholm, Sweden

Received 24 October 1983/Accepted 12 March 1984

The levels of seven intermediary enzymes involved in acetate and butyrate formation from acetylcoenzyme A in the saccharolytic anaerobe Clostridium acetobutylicum were investigated as a function oftime in solvent-producing batch fermentations. Phosphate acetyltransferase and acetate kinase, which are

known to form acetate from acetyl coenzyme A, both showed a decrease in specific activity when theorganism reached the solvent formation stage. The three consecutive enzymes thiolase, P-hydroxybutyryl-coenzyme A dehydrogenase, and crotonase exhibited a coordinate expression and a maximal activity aftergrowth had ceased. Only low levels of butyryl coenzyme A dehydrogenase activity were found. Phosphatebutyryltransferase activity rapidly decreased after 20 h from 5 to 11 U/mg of protein to below the detectionlimit (1 mU/mg). Butyrate no longer can be formed, and the metabolic flux may be diverted to butanol.Butyrate kinase showed a 2.5- to 10-fold increase in specific activity after phosphate butyryltransferaseactivity no longer could be detected. These results suggest that the uptake of acetate and butyrate duringsolvent formation can not proceed via a complete reversal of the phosphate transferase and kinasereactions. The activities of all enzymes investigated as a function of time in vitro are much higher than themetabolic fluxes through them in vivo. This indicates that none of the maximal activities of the enzymes

assayed is rate limiting in C. acetobutylicum.

The saccharolytic anaerobe Clostridium acetobutylicum isknown to utilize several different carbohydrates as the solesource of carbon and energy (22, 44). The main end productsof the fermentation are acetate, butyrate, acetone, butanol,ethanol, carbon dioxide, and hydrogen (15). The typicalacetone-butanol fermentation can be divided into two dis-tinctive phases. In the acidogenic phase acetate and butyrateare formed together with carbon dioxide and hydrogen.During this stage the organism grows rapidly and the pHdrops to 4 to 4.5 due to the accumulation of acids in themedium (15). The pH then remains at a fairly constant levelthroughout the remainder of the fermentation, In the secondphase the growth slows down, acids are taken up from themedium and metabolized, and acetone, butanol, ethanol,carbon dioxide, and hydrogen are formed (3, 8, 29). Al-though many investigators (7, 11, 23, 25, 28, 43, 45, 46) sinceSpeakman (32) have studied the biochemistry of the acetone-butanol fermentation, little is known about the intermediarymetabolism and its regulation and the factors responsible forthe shift from acid production to acid reutilization andneutral solvent formation. In this investigation we report onthe metabolic pathways of acetate and butyrate formationfrom acetyl coenzyme A (acetyl-CoA) in C. acetobutylicumgrown on glucose. The variations in specific activities of theenzymes responsible for the production of these acids fromacetyl-CoA are presented as a function of fermentation time.The studies were carried out on typical solvent-producing,butylic, batch fermentations.

MATERIALS AND METHODSMaterials. The lithium salt of butyryl phosphate was

prepared by the method of Avison (1). Potassium butyratewas prepared by neutralization of butyric acid with potassi-

* Corresponding author.

um hydroxide. Clostridium kluyiveri cells were a kind giftfrom T. C. Stadtman, National Institutes of Health. All co-factors, CoA thioesters, and electron transport dyes werepurchased from Sigma Chemical Co. All other materialswere obtained from commercial sources and were of thehighest available purity.

Bacterial strain and growth conditions. C. acetobutylicumATCC 824 was obtained from the American Type CultureCollection. The culture was maintained as spores on sterilesoil. The cells were grown in a medium containing thefollowing (grams per liter): (NH4)2SO4, 2.0; K2HPO4, 1.0;KH2PO4, 0.5; MgSO4 - 7H2O, 0.1; FeSO4 .7H20, 0.015;CaCI2, 0.01; MnSO4 - H20, 0.01; CoCl2, 0.002;ZnSO4 - 7H20, 0.002; Na,SeO3, 0.00025; tryptone, 2.0;yeast extract, 1.0; glucose, 50. The stock culture was heatshocked at 90°C for 5 min in 200 ml of growth medium andincubated in an anaerobic box (Forma Scientific; model1024) at 37°C for 24 h. The volume of the culture was scaledup to 20 liters in two successive transfers.

Fermentations. The butylic fermentations were carried outat 37°C in a jacketed 100-liter stainless steel substrate tank(AB Taby Rostfria, Taby, Sweden) with a 90-liter workingvolume with the medium described above. Glucose wasautoclaved separately from other components of the medi-um. Before inoculation anaerobic conditions were main-tained by sparging the medium with nitrogen. The initial pHbefore inoculation was 6.5. When temperature equilibriumhad been reached the medium was inoculated with a 10%(vol/vol) inoculum from the 20-liter culture. The pH was notcontrolled during the fermentation. Growth was followed bymeasuring the optical density at 540 nm. One optical densityunit at 540 nm corresponded to 0.14 g of cells (dry weight)per liter.

Samples. Samples (1 liter) were withdrawn at the timesindicated in Fig. 2 and centrifuged at 15,000 x g for 20 min.Samples of the supernatants were filtered and stored at-20°C for gas-liquid chromatographic analyses. The har-

1277

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1278 HARTMANIS AND GATENBECK

vested cells were immediately frozen and stored in liquidnitrogen until they were used for the preparation of cell-freeextracts.

Preparation of cell-free extracts. The cells harvested from1-liter samples were thawed in 4 ml of 50 mM 2-(N-morpho-line) propanesulfonic acid-KOH (pH 7.0) containing 1 mM1,4-dithiothreitol. The cell suspensions were sonicated fourtimes for 30 s at 45% of maximum power output at 0°C with aBranson B-12 sonifier. Cellular debris was removed bycentrifugation at 40,000 x g for 30 min at 4°C. All superna-tants were desalted by gel permeation chromatography onPharmacia PD-10 columns equilibrated with 50 mM 2-(N-morpholine)propanesulfonic acid-KOH (pH 7.0) containing1 mM 1,4-dithiothreitol. The desalted extracts were immedi-ately frozen and stored at -75°C until used for enzymeassays and protein determinations.

Protein determination. Protein concentrations in crudeextracts were estimated by the dye-binding method of Brad-ford (6). Determinations of the protein content of whole cellswere made by the biuret method as described by Herbert etal. (14). In both methods gamma globulin was used as thereference protein. For the calculations of the fluxes ofmetabolites in the cells a value of 0.7 mg of protein per mg ofcell dry weight was used.

Analysis of fermentation products. The concentrations ofacetone, butanol, ethanol, acetate, and butyrate were deter-mined by gas-liquid chromatography with a Perkin-Elmer3920B gas chromatograph equipped with a flame ionizationdetector. The instrument was fitted With a Teflon column (6ft [ca. 1.8 m] by 0.085 in. [ca. 0.22 cm], inner diameter)packed with Chromosorb 101 (80-100 mesh; E. Merck, A. G.Darmstadt, West Germany). The 0.5->.l samples, acidifiedwith metaphosphoric acid to ensure protonation of the acids,were injected at an injector temperature of 180°C. Theanalyses were run isothermally at a column temperature of180°C and a detector temperature of 215°C. The nitrogencarrier gas flow rate was 30 ml/min. External standardscontaining all five fermentation products to be separatedwere used.

Electrophoretic analysis. Disc gel electrophoresis of crudeextracts from different stages of the fermentations wasperformed with 7.5% polyacrylamide gels under nondenatur-ing conditions by a modification of the method of Laemmli(17).Enzyme activity staining of butyrate kinase after electro-

phoresis of crude extracts was carried out by the hydroxa-mate method of Rose (27) with the assay mixture describedbelow for the enzyme assays. The gel was incubated in 60 mlof reaction mixture for 20 min at 25°C. Then the reactionmixture was discarded, and the gel was immersed in 100 mlof FeCl3 reagent until a brown color developed.Enzyme assays. Unless otherwise indicated, all enzymes

were assayed in air at 30°C in masked 1-ml ctivettes in aGilford 2600 or a Shimadzu UV-120-02 spectrophotometer.At the dilutions of crude extracts used in the assays, NADHoxidase activity was negligible. One unit of enzyme activityis defined as the amount of enzyme catalyzing the conver-sion of 1 i,mol of reactant to products per min. All of theenzymes assayed as a function of time in crude extracts werestable to repeated feeziig and thawing and storage at -75°Cduring the course of this investigation.Phosphate acetyltransferase (acetyl-CoA:orthophosphate

acetyltransferase, EC 2.3.1.8) and phosphate butyryltrans-ferase (butyryl-CoA:orthophosphate butyryltransferase, EC2.3.1.19) were assayed by measuring the formation of acyl-CoA from acyl phosphate. The extinction coefficient used

was 4,440 M-1 cm-' at 233 nm. The 1.0-ml reaction mixturewas a modification of that described by Bergmeyer et al. (4)and contained 100 mM Tris-hydrochloride (pH 8.0), 100 mMKCI, 0.6 mM CoA, and 10 mM lithium acyl phosphate. Thereaction was initiated by the addition of extract. Phosphatebutyryltransferase was also assayed in the butyryl phos-phate-forming direction. The disappearance of the thioesterbond was measured spectrophotometrically at 233 nm. The1.0 ml reaction mixture contained 100 mM potassium phos-phate (pH 7.0) and extract. The addition of 0.1 mM butyryl-CoA initiated the reaction. The assay was linear with activi-ties up to 0.05 absorbancy units per min. To determinepossible thioesterase activity, controls containing 50 mM 2-(N-morpholine)propanesulfonic acid-KOH (pH 7.0), 0.1mM butyryl-CoA, and extract were run.

Acetate kinase (ATP:acetate phosphotransferase, EC2.7.2.1) and butyrate kinase (ATP:butyrate phosphotransfer-ase, EC 2.7.2.7) were assayed by the hydroxamate methodof Rose (27). The absorbances were measured at 540 nm byusing blanks where ATP had been omitted. Acetyl hydroxa-mate was used for preparation of all standard curves as theextinction coefficients of acetyl hydroxamate and butyrylhydroxamate are essentially identical (19). To investigate thereversibility of butyrate kinase the enzyme was also assayedin the butyrate-forming direction in a coupled assay withhexokinase and glucose 6-phosphate dehydrogenase (18).The 1.0-ml reaction mixture contained 75 mM Tris-hydro-chloride (pH 8.0), 6 mM ADP, 10 mM MgCl2, 2 mM glucose,1 U of hexokinase, 0.2 mM NADP+, 1 U of glucose 6-phosphate dehydrogenase, and rate-limiting amounts of en-zyme. The reaction was initiated by the addition of 7 mMbutyryl phosphate. The controls contained assay mixturewhere ADP had been omitted.

Determinations of thiolase (acetyl-CoA:acetyl-CoA C-acetyltransferase, EC 2.3.1.9) activity were made in thedirection of CoA-dependent acetoacetyl-CoA cleavage. Thedecrease in absorbance at 303 nm was measured spectropho-tometrically (36). The 1.0-ml reaction mixture contained 100mM Tris-hydrochloride (pH 8.0), 1 mM 1,4-dithiothreitol, 10mM MgCl2, 50 F.M acetoacetyl-CoA, and extract. Theaddition of 0.2 mM coenzyme A initiated the reaction. Themolar extinction coefficient used was 14,000 M-l cm-' (37).

,-Hydroxybutyryl-CoA dehydrogenase (L-3-hydroxyacyl-CoA:NAD+ oxidoreductase, EC 1.1.1.35) activity was as-sayed by a modification of the method of Madan et al. (21).The 1.0-ml assay mixture contained 50 mM 2-(N-morpho-line) propanesulfonic acid-KOH, pH 7.0, 1 mM dithiothrei-tol, 0.2 mM NADH, and extract. The addition of 75 ,uMacetoacetyl-CoA initiated the reaction.Crotonase (L-3-hydroxyacyl-CoA hydrolyase, EC

4.2.1.17) was assayed by determining the decrease in absor-bance at 263 nm due to hydration of the conjugated doublebond of crotonyl-CoA as described by Stern (35). Theextinction coefficient of crotonyl-CoA was 6,700 M-1 cm-'(20). The 1.0-ml reaction mixture contained 100 mM Tris-hydrochloride (pH 7.6) and extract. The addition of 0.15 mMcrotonyl-CoA initiated the reaction.For the assay of butyryl-CoA dehydrogenase (butyryl-

CoA:NAD+ oxidoreductase, EC 1.3.99.2) three differentmethods were tried. In the first method, where butyryl-CoAis oxidized to crotonyl-CoA, the reduction of iodonitrotetra-zolium [2-(4-iodophenyl)-3-(4-nitrophenyl)-5-phenyltetrazo-lium chloride] was followed spectrophotometrically at 492nm (9). The 1.0-ml reaction mixture contained 100 mMpotassium phosphate (pH 7.0), 0.15% Triton X-100, 0.2 mMMeldolablau (zinc chloride salt of 8-dimethylamino-2,3-ben-

APPL. ENVIRON. MICROBIOL.

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

LEVELS OF ENZYMES IN C. ACETOBUTYLICUM 1279

zophenoxazine), 0.3 mM iodonitrotetrazolium, and thecrude extract. The stoppered cuvettes were flushed withhydrogen gas for 2 min, and the reaction was initiated by theaddition of 0.1 mM butyryl-CoA. The extinction coefficientused was 19,400 M-1 cm-' (9). The necessary controlswithout butyryl-CoA and without enzyme were performed.In the second method the oxidation of prereduced Safranin0 was followed spectrophotometrically at 520 nm (30). Theassays were carried out in an anaerobic box in an atmo-sphere consisting of 85% N2, 10% H2, and 5% Co2. In thethird method the reduction of 2,6-dichlorophenolindophenolwas followed (10).

RESULTS AND DISCUSSIONIn the acetone-butanol fermentation of glucose by various

saccharolytic clostridia the final fatty acid and neutral sol-vent products are derived from acetyl-CoA by the series ofreactions outlined in Fig. 1. Although there is considerableinformation about the relative amounts of products formedand pH changes during the course of the fermentations, thefactors that control the shift from acid production to neutralsolvent production are not well understood. In the presentstudy the levels of enzymes that catalyze a number of theindividual steps of the overall process (Fig. 1) were mea-sured as a function of time in C. acetobutylicum cells from abutylic fermentation without pH control (Fig. 2). Crudeextracts prepared from cells harvested at the times indicatedin Fig. 2 were desalted and assayed for individual enzymeactivities as described above.A more butyric type of fermentation, where the pH was

maintained at 6 during the first 10 h by the addition ofpotassium hydroxide, was also investigated. The end prod-ucts of this fermentation were predominantly butyric andacetic acids. No major differences in the levels of theinvestigated enzymes were found between the butyric andthe butylic fermentations.Formation of acetate from acetyl-CoA. Two enzymes,

phosphate acetyltransferase and acetate kinase, which cata-lyze reactions 1 and 2, respectively, of Fig. 1, are involved inacetate formation from acetyl-CoA. These enzymes havebeen suggested to exist in all anaerobic bacteria that utilize

GLUCOSE

+1 2ACETYL-CoA ACETYL PHOSPHATE _ ACETATE

31ACETOACETYL-CoA - ACETONE

41,B-HYDROXYBUTYRYL-CoA

51CROTONYL-CoA

BUTYRYL-CoA - BUTYRYL PHOSPHATE - BUTYRATE

BUTANOL

FIG. 1. Metabolic pathways of acetate and butyrate formationfrom acetyl-CoA in C. acetobutylicum. The enzymes investigatedwere as follows: 1, phosphate acetyltransferase; 2, acetate kinase; 3,thiolase; 4, ,-hydroxybutyryl-CoA dehydrogenase; 5, crotonase; 6,butyryl-CoA dehydrogenase; 7, phosphate butyryltransferase; 8,butyrate kinase.

H~~~HURFI.2. Coreo uyi emnaini ach utr fC

z

C-)~~~~~~~~~-

0-

0 O

10- ~ 0203

acetobutylicum. Symbols: *, butanol; x, acetone; *, ethanol; 0,butyric acid; *, acetic acid; O, optical density at 540 nm;V, pH.

acetyl-CoA to synthesize ATP via substrate level phosphor-ylation (38).

Phosphate acetyltransferase. The variations in levels ofphosphate acetyltransferase during the course of the butylicfermentation are shown in Fig. 3. The enzyme was measuredin the direction of acetyl-CoA formation, which is thereverse of the in vivo reaction that leads eventually toacetate formation. The activity showed a marked decreaseafter 13 h, when the acetate formation rate slowed down andgrowth ceased. Phosphate acetyltransferase from C. aceto-butylicum shows similarities with the enzyme from C. kluy-yeri with respect to the influence of monovalent cations onactivity (33). Our investigations showed that the C. acetobu-tylicum enzyme was activated by K+ and NH4+ and wasinhibited by Na+. Half-maximal activation was obtained by60 mM K+ in the form of KCl.

Acetate kinase. The physiological role of acetate kinaseduring the first phase of the acetone-butanol fermentation isto catalyze the phosphorylation of ADP and the synthesis ofacetate from acetyl phosphate. In this study acetate kinasewas assayed in the direction of acetyl phosphate formation.The specific activities of the enzyme as a function of time areshown in Fig. 3. The specific activity exhibited a maximumafter only 4 h and dropped by 90% before growth ceased and

C 10 20 30HOURS

FIG. 3. Variation in levels of phosphate acetyltransferase (0)and acetate kinase (O) as a function offermentation time in the batchculture of C. acetobutylicum shown in Fig. 2. Enzyme activitieswere measured in desalted crude extracts as described in the text.

VOL. 47, 1984

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1280 HARTMANIS AND GATENBECK

before a corresponding decrease in phosphate acetyltrans-ferase activity was observed (Fig. 3). Under these growthconditions there was no coordinate expression of the twoenzymes responsible for acetate formation in C. acetobutyli-cum.

Butyryl-CoA formation from acetyl-CoA. The enzymesinvolved in this metabolic pathway, namely, thiolase, f-

hydroxybutyryl-CoA dehydrogenase, crotonase, and bu-tyryl-CoA dehydrogenase, were shown to be present in C.acetybutylicum. This is in accordance with the metabolismof C. kluyveri as shown by Barker (2) and also in agreementwith the postulations of Gottschalk and Bahl (12).

Thiolase. Thiolase activity was measured in the thiolyticdirection, whereas in vivo during butyrate formation thereaction is catalyzed in the biosynthetic direction. Thespecific activities found in desalted crude extracts, up to 11U/mg of protein, are comparable with those reported forother clostridia (5, 13). The variations in specific activities ofthe enzyme with time are shown in Fig. 4. Maximal specificactivity was observed after the organism ceased to grow.

,-Hydroxybutyryl-CoA dehydrogenase. Like other buty-rate-forming clostridia (42), C. acetobutylicum was found topossess an active NADH-dependent P-hydroxybutyryl-CoAdehydrogenase. The highest specific activity found was 8 U/mg of protein when the enzyme was assayed in the directionof acetoacetyl-CoA reduction. In contrast to C. kluyveri (21)no NADPH-dependent reduction of acetoacetyl-CoA couldbe demonstrated. The activity profile of the dehydrogenaseduring the fermentation is similar to those of thiolase andcrotonase (Fig. 4). In all cases the maximal specific activitieswere reached after growth had ceased.

Crotonase. A short-chain fatty acyl-CoA specific hydro-lyase with a high specific activity that catalyzes the revers-

ible hydration of crotonyl-CoA, has been purified to homo-geneity from C. acetobutylicum (43). In our investigationextremely high specific activities of this enzyme were found.In desalted crude extracts, an activity of 135 U/mg of proteinwas demonstrated when assayed in the hydration direction.The crotonase activity increased with growth up to 12 to 14h, at which time growth ceased (Fig. 4). The maximalspecific activity, however, was exhibited after 17 h offermentation.

Expression of thiolase, 0-hydroxybutyryl-CoA dehydroge-nase, and crotonase. There seems to be a coordinate expres-

sion of the metabolically consecutive enzymes thiolase, -

10 150

8

0 ° 1 0 20 3C 100

1-4

4z

2-

HOU RSFIG. 4. Variation in levels of thiolase (O), ,-hydroxybutyryl-

CoA dehydrogenase (0), and crotonase (A) as a function of fermen-tation time in the batch culture of C. acetobutylicum shown in Fig.2. Enzyme activities were measured in desalted crude extracts as

described in the text.

hydroxybutyryl-CoA dehydrogenase, and crotonase. Theyall showed maximum activity after 17 h of fermentation. Thereason for the high specific activities toward the end may bea reflection of the change in pH during the course of thefermentation. Berndt and Schlegel (5) have determined thepH optimum for thiolase from Clostridium pasteurianum tobe 8.1 in the condensation direction. To our knowledge nopH optima for clostridial NADH-dependent ,-hydroxybu-tyryl-CoA dehydrogenase and crotonase have been pub-lished. The pH of the interior of the cell follows the externalpH to a certain extent throughout the fermentation (26). Thiscreates an unfavorable milieu for thiolase and possibly alsofor the two other enzymes. Thereby more enzyme may beneeded to compensate for the decrease in enzyme activitiesdue to a low pH value. A drop in pH may be an importantfactor for the regulation of these three consecutive enzymes.Butyryl-CoA dehydrogenase. Butyryl-CoA dehydrogenase

has been shown to be present in several clostridia, such asClostridium butylicum and C. kluyveri (42). In desalted crudeextracts of C. acetobutylicum, however, we found onlyextremely low activities. Butyryl-CoA dehydrogenase couldnot be assayed by the 2,6-dichlorophenolindophenol (10) orthe Safranin 0 assays (30). No activity was exhibited withthe physiological electron donors NADH and NADPH.Activity could only be demonstrated with the Meldolablauassay (9) at pH 7.0 (0.1 M potassium phosphate). This assayworked only after the reaction mixture had been flushed withhydrogen gas, and then the enzyme showed a sluggish andnonlinear activity. In 0.1 M potassium phosphate (pH 6.2),only traces of activity could be found, and in 0.1 M Tris-hydrochloride (pH 8.0), no activity was detected. To test theassay procedures, extracts from C. kluyveri were preparedas described above, and the specific activity of butyryl-CoAdehydrogenase was determined by the Meldolablau andSafranin 0 assays. The activities were 0.8 and 0.7 U/mg,respectively. This was more than 100-fold higher than theactivities found in C. acetobutylicum extracts. Due to poorreproducibility, no determinations of the time dependence ofbutyryl-CoA dehydrogenase in C. acetobutylicum weremade. The reason for the low activity is not known. Onepossibility may be that the enzyme is oxygen sensitive and isinactivated during sonication and preparation of the desaltedcrude extracts.

Butyrate formation from butyryl-CoA. Saccharolytic clos-tridia form butyryl phosphate from butyryl-CoA via phos-phate butyryltransferase (41). Butyrate is formed from bu-tyryl phosphate with concomitant phosphorylation of ADP.The enzyme catalyzing this reaction is butyrate kinase (39,40).

Phosphate butyryltransferase. Phosphate butyryltransfer-ase was partially purified from C. acetobutylicum by Gavardet al. (11). They demonstrated that two different enzymes areresponsible for the formation of butyryl phosphate andacetyl phosphate from butyryl-CoA and acetyl-CoA, respec-tively. They reported a phosphate butyryltransferase activi-ty of 0.11 U/mg of protein in crude extracts. The presence ofthis enzyme in C. acetobutylicum was also shown by Valen-tine and Wolfe (41). In our investigation phosphate butyryl-transferase activity was assayed in both directions withidentical activity profiles as a result. Specific activities of 11U/mg were measured in the butyryl-CoA-forming direction(Fig. 5). The activity was not growth related and reached itsmaximum after only 5 h. It then decreased rapidly to a valuebelow the detection limit, 1 mU/mg, while the cells still wereactively growing. From the data in Fig. 5 it becomes evidentthat no more butyrate can be formed from butyryl-CoA via

APPL. ENVIRON. MICROBIOL.

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

LEVELS OF ENZYMES IN C. ACETOBUTYLICUM 1281

HOURS

FIG. 5. Variation in levels of phosphate butyryltransferase (0)and butyrate kinase (0) as a function of fermentation time in thebatch culture of C. acetobutylicum shown in Fig. 2. Enzymeactivities were measured in desalted crude extracts as described inthe text.

butyryl phosphate after 22 h. At this stage of the fermenta-tion a rise in the butanol formation rate occurred (Fig. 2).This is in accordance with a diversion of the metabolic fluxfrom butyrate to butanol at the branch point of butyryl-CoA.Butanol is formed from butyryl-CoA via butyraldehyde (24).The variations in the levels of these dehydrogenases in C.acetobutylicum as a function of time are not known. It isprobable that phosphate butyryltransferase, by adjusting theflux to either butyrate or butanol, plays a major role in theregulatory mechanism that causes the switch from acid tosolvent production.

Butyrate kinase. Butyrate kinase has been demonstrated indifferent clostridia, such as the saccharolytic species C.butyricum (39), Clostridium tetanomorphium, C. pastelr-ianum, Clostridium tyrobutyricum, Clostridium sporogenes,and a solvent former, Clostridium sp. McCoy A-14 (40). Wefound that desalted crude extracts of C. acetoblutylicumexhibited a reversible butyrate kinase activity during allstages of fermentation. The highest specific activities mea-sured were 5 U/mg in the butyrate-phosphorylating directionand 45 U/mg in the butyrate-forming direction. The activityprofile showed a marked increase in specific activity towardthe end, when cell growth had ceased (Fig. 5). This increasecoincided with the uptake and reutilization of butyrateproduced in the earlier stage (Fig. 2). The 2.5- to 10-foldincrease in butyrate kinase activity occurred after a sharpdecrease in phosphate butyryltransferase activity (Fig. 5).There was no concomitant change in acetate kinase activity(Fig. 3). This pattern was observed in all fermentations. Theexistence of two or more different enzymes that catalyze thesame reaction to meet different metabolic needs is notuncommon. For example, Clostridium cylindrosporum pos-sesses two different formate kinases (31). In that organismboth enzymes function in the formate-phosphorylating direc-tion, but only one is reversible and able to phosphorylateADP. In C. acetobutylicum we found some discrepancies inthe ratios of the specific activities of butyrate kinase of theforward and reverse reactions. This could indicate theexistence of multiple forms of butyrate kinase. Crude ex-tracts of C. acetobutylicum also phosphorylated propionatewith 75% of the activity exhibited with butyrate as a sub-strate.Uptake and reutilization of butyric acid. Already in 1945 it

was shown that when 13C-labeled butyrate was added to an

actively metabolizing culture of C. acetobutylicum, 84% ofthe label was found in butanol (45). The mechanism ofbutyric acid uptake and conversion to butanol is, however,still unknown. It is assumed that the transport of acetic andbutyric acids across the cytoplasmic membranes of clostridiais effected by passive diffusion of the uncharged forms ofthese acids (38). Acetic acid has been shown to cross themembrane of C. pasteurianum in the protonated form (16).During the solvent formation phase the activity of butyratekinase was still high, but the phosphate butyryltransferaseactivity was very low or not even detectable (Fig. 5).Therefore no more butyryl-CoA could be formed frombutyryl phosphate. Instead a direct reduction of butyrylphosphate to butyraldehyde and further to butanol may bepossible. Stadtman and Barker (34) demonstrated an enzy-matic reduction of butyryl phosphate to butanol by usingdried cells of C. kluyveri and hydrogen gas as the reductant.Although the direct reduction of butyryl phosphate to buta-nol was not demonstrated, this potential mechanism for thereutilization of butyric acid is metabolically more attractivethan a complete reversal of both the butyrate kinase andphosphate butyryltransferase steps. We are presently inves-tigating possible mechanisms for the uptake and activation ofacetic acid and butyric acid in C. acetobutylicum.

Metabolic fluxes. In an attempt to determine any possibleenzyme limitations for the metabolic fluxes through theseven enzymes investigated, the fluxes in vivo were com-pared with the enzyme capacities measured in vitro. Despitethe ambiguities of such a comparison when the in vivoactivities and regulating mechanisms are not known, it mayindicate possible rate-limiting steps in the metabolic path-ways leading to acetate and butyrate formation. The specificmetabolic fluxes that are brought about in vivo through theaction of the investigated enzymes were estimated from theproduct formation rates and protein content of the cells. Asan approximation, the fluxes through phosphate acetyltrans-ferase and acetate kinase reactions were estimated fromacetate formation rates. Butyrate, butanol, and acetoneformation rates were used for calculation of the fluxesthrough the thiolase reaction. The fluxes through the 1-hydroxybutyryl-CoA dehydrogenase and crotonase reac-tions were determined from the formation rates of butyrateand butanol. Butyrate formation rates were used for calcula-tion of the fluxes through the phosphate butyryltransferase-and butyrate kinase-catalyzed steps. It should be noted thatall enzyme activities were not measured under optimalconditions that gave maximal activities. Therefore the de-gree of utilization of the enzymes in vivo is, in most cases,even lower than is shown in Table 1. The activities of theenzymes in vitro are much higher than the fluxes, attribut-able to them in vivo, which means that in this fermentation

TABLE 1. Specific metabolic fluxes in C. acetobutvlicum in vivo

Enzyme % Utiliza-tion"

Phosphate acetyltransferase ......................... 0.3-10Acetate kinase .............. ............... 1.2-11.5Thiolase ............................. 0.3-2.83-Hydroxybutyryl-CoA dehydrogenase ....... ........ 0.3-4.3

Crotonase..................................... 0.01-1.1Phosphate butyryltransferase ........................ 0.08-1.3Butyrate kinase............................. 0.002-0.7

a Percentage of measured specific enzyme activities in vitro during thebutylic batch fermentation shown in Fig. 2.

VOL. 47, 1984

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1282 HARTMANIS AND GATENBECK

the enzymes are used to a low extent. Despite the approxi-mations made, the data in Table 1 indicate that there is norestriction of the fluxes due to enzyme shortages. Thisimplies that none of the maximal enzyme activities measuredseems to be rate limiting in the cells during an acetone-butanol fermentation.

ACKNOWLEDGMENTSWe thank Asalie Larsson at the Fermentation Centre, Department

of Biochemistry and Biotechnology, for kind help with the 100-literfermentations. We are grateful to Karl Hult for his advice andsuggestions throughout the course of this study. We also express ourappreciation to Thressa C. Stadtman for her critical reading of themanuscript and for helpful discussions.

This investigation was supported by a grant from the SwedishNatural Science Research Council.

LITERATURE CITED

1. Avison, A. W. D. 1955. The synthesis of acyl phosphates inaqueous solution. J. Am. Chem. Soc. 77:732-738.

2. Barker, H. A. 1956. Bacterial fermentations, p. 28-56. JohnWiley & Sons, Inc., New York.

3. Beesch, S. C. 1953. A microbiological process report: acetone-butanol fermentation of starches. Appl. Microbiol. 1:85-95.

4. Bergmeyer, H. U., G. Holz, H. Klotzsch, and G. Lang. 1963.Phosphotransacetylase aus Clostridium kluyi'eri. Biochem. Z.338:114-121.

5. Berndt, H., and H. G. Schlegel. 1975. Kinetics and properties ofP-ketothiolase from Clostridium pasteurianum. Arch. Micro-biol. 103:21-30.

6. Bradford, M. M. 1976. A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal. Biochem. 72:248-254.

7. Davies, R. 1942. Studies on the acetone-butyl alcohol fermenta-tion. 2. Intermediates in the fermentation of glucose by Cl.acetobitvlicum. 3. Potassium as an essential factor in thefermentation of maize meal by Cl. acetobutylicinn (BY). Bio-chem. J. 36:582-599.

8. Davies, R., and M. Stephenson. 1941. Studies on the acetone-butylalcohol fermentation. 1. Nutritional and other factors in-volved in the preparation of active suspensions of Clostridiumacetobutylicum (Weizmann). Biochem. J. 35:1320-1331.

9. Dommes, V., and W.-H. Kunau. 1976. A convenient assay foracyl-CoA-dehydrogenases. Anal. Biochem. 71:571-578.

10. Engel, P. C. 1981. Butyryl-CoA dehydrogenase from Mega-sphaera elsdenii. Methods Enzymol. 71:359-366.

11. Gavard, R., B. Hautecoeur, and H. Descourtieux. 1957. Phos-photransbutyrylase de Clostridium acetobutylicum. C. R. Acad.Sci. (Paris) 244:2323-2326.

12. Gottschalk, G., and H. Bahl. 1981. Feasible improvements of thebutanol production by Clostridium acetobutylicum, p. 463-471.In A. Hollaender (ed.), Basic life sciences, vol. 18. Trends in thebiology of fermentation for fuels and chemicals. Plenum Pub-lishing Corp., New York.

13. Hartmanis, M. G. N., and T. C. Stadtman. 1982. Identificationof a selenium-containing thiolase from Clostridium kluyveri:identification of the selenium moiety as selenomethionine. Proc.Natl. Acad. Sci. U.S.A. 79:4912-4916.

14. Herbert, D., P. J. Phipps, and R. E. Strange. 1971. The biuretmethod, p. 244-252. In J. R. Norris and D. W. Ribbons (ed.),Methods in microbiology, Sb. Academic Press, Inc., London.

15. Johnson, M. J., W. H. Peterson, and E. B. Fred. 1931. Oxidationand reduction relations between substrate and products in theacetone-butylalcohol fermentation. J. Biol. Chem. 91:569-591.

16. Kell, D. B., M. W. Peck, G. Rodger, and J. G. Morris. 1981. Onthe permeability to weak acids and bases of the cytoplasmicmembrane of Clostridium pasteurianum. Biochem. Biophys.Res. Commun. 99:81-88.

17. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

18. Lamprecht, W., and 1. Trautschold. 1970. Adenosin-5'-triphos-

phat. Bestimmung mit Hexokinase und Glucose-6-phosphatDehydrogenase, p. 2024-2033. In H. U. Bergmeyer (ed.), Meth-oden der enzymatischen Analyze, 2. Auflage. Verlag Chemie,Weinheim.

19. Lipmann, F., and L. C. Tuttle. 1945. A specific micromethod forthe determination of acyl phosphates. J. Biol. Chem. 159:21-28.

20. Lynen, F., and S. Ochoa. 1953. Enzymes of fatty acid metabo-lism. Biochim. Biophys. Acta 12:299-314.

21. Madan, V. K., P. Hillmer, and G. Gottschalk. 1973. Purificationand properties of NADP-dependent L(+)-3-hydroxyacyl-CoAdehydrogenase from Clostridium kluyveri. Eur. J. Biochem.32:51-56.

22. McCoy, E., E. B. Fred, W. H. Peterson, and E. G. Hastings.1926. A cultural study of the acetone butyl alcohol organisms. J.Infect. Dis. 39:457-483.

23. Petitdemange, H., C. Cherrier, J. M. Bengone, and R. Gay.1977. Etude des activites NADH et NADPH-ferredoxine oxy-doreductasiques chez Clostridium acetobutylicum. Can. J. Mi-crobiol. 23:152-160.

24. Petitdemange, H., J. Desbordes, J. Berthelin, and R. Gay. 1968.Conversion enzymatique du n-butyraldehyde en n-butanol chezClostridium acetobutylicum. C. R. Acad. Sci. (Paris) 266:1772-1774.

25. Petitdemange, H., J. Desbordes, M. Maugras, and R. Gay. 1969.Etude de la formation du n-butanol chez Clostridium acetobuty-licum. Bull. Soc. Chim. Biol. 51:157-165.

26. Riebeling, V., R. K. Thauer, and K. Jungermann. 1975. Theinternal-alkaline pH gradient, sensitive to uncoupler andATPase inhibitor, in growing Clostridium pasteurianum. Eur. J.Biochem. 55:445-453.

27. Rose, I. A. 1955. Acetate kinase of bacteria. Methods Enzymol.1:591-595.

28. Rosenfeld, B., and E. Simon. 1950. The mechanism of thebutanol-acetone fermentation. 1. The role of pyruvate as anintermediate. 2. Phosphoenolpyruvate as a new intermediate. J.Biol. Chem. 186:395-404, 405-410.

29. Ross, D. 1961. The acetone-butanol fermentation. Prog. Indust.Microbiol. 3:73-90.

30. Seubert, W., and F. Lynen. 1953. Enzymes of the fatty acidcycle. II. Ethylene reductase. J. Am. Chem. Soc. 75:2787-2788.

31. Sly, W. S., and E. R. Stadtman. 1963. Formate metabolism. II.Enzymatic synthesis of formyl phosphate and formyl coenzymeA in Clostridium cylindrosporum. J. Biol. Chem. 238:2639-2647.

32. Speakman, H. B. 1920. The biochemistry of the acetone andbutyl alcohol fermentation by Bacillus granulobacter pectino-vorum. J. Biol. Chem. 41:319-343.

33. Stadtman, E. R. 1952. The purification and properties of phos-photransacetylase. J. Biol. Chem. 196:527-534.

34. Stadtman, E. R., and H. A. Barker. 1950. Fatty acid synthesisby enzyme preparations of Clostridium kluyveri. VI. Reactionsof acyl phosphates. J. Biol. Chem. 184:769-793.

35. Stern, J. R. 1955. Crystalline crotonase from ox liver. MethodsEnzymol. 1:559-566.

36. Stern, J. R. 1955. Enzymes of acetoacetate formation andbreakdown. 3-Keto thiolase. Methods Enzymol. 1:581-585.

37. Stern, J. R. 1956. Optical properties of acetoacetyl-S-coenzymeA and its metal chelates. J. Biol. Chem. 221:33-44.

38. Thauer, R. K., K. Jungermann, and K. Decker. 1977. Energyconservation in chemotrophic anaerobic bacteria. Bacteriol.Rev. 41:100-180.

39. Twarog, R., and R. S. Wolfe. 1962. Enzymatic phosphorylationof butyrate. J. Biol. Chem. 237:2474-2477.

40. Twarog, R., and R. S. Wolfe. 1963. Role of butyryl phosphate inthe energy metabolism of Clostridium tetanomorphum. J. Bac-teriol. 86:112-117.

41. Valentine, R. C., and R. S. Wolfe. 1960. Purification and role ofphosphotransbutyrylase. J. Biol. Chem. 235:1948-1952.

42. v. Hugo, H., S. Schoberth, V. K. Madan, and G. Gottschalk.1972. Coenzyme specificity of dehydrogenases and fermenta-tion of pyruvate by clostridia. Arch. Microbiol. 87:189-202.

43. Waterson, R. M., F. J. Castellino, G. M. Hass, and R. L. Hill.1972. Purification and characterization of crotonase from Clos-

Appt-. ENVIRON. MICROBIOL.

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

LEVELS OF ENZYMES IN C. ACETOBUTYLICUM 1283

tridium acetobutylicum. J. Biol. Chem. 247:5266-5271.44. Weyer, E. R., and L. F. Rettger. 1927. A comparative study of

six different strains of the organism commonly concerned inlarge scale production of butyl alcohol and acetone by thebiological process. J. Bacteriol. 14:399-424.

45. Wood, H. G., R. W. Brown, and C. H. Werkman. 1945.

Mechanism of the butyl alcohol fermentation with heavy carbonacetic and butyric acids and acetone. Arch. Biochem. 6:243-260.

46. Zerner, B., S. M. Coutts, F. Lederer, H. H. Waters, and F. H.Westheimer. 1966. Acetoacetate decarboxylase. Preparation ofthe enzyme. Biochemistry 5:813-816.

VOL. 47, 1984

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from