Vol. 52, No. 1APPLIED AND ENVIRONMENTAL MICROBIOLOGY, JUlY 1986, p. 185-1900099-2240/86/070185-06$02.00/0Copyright © 1986, American Society for Microbiology

Characterization, Biosynthesis, and Regulation of Granulose inClostridium acetobutylicum

A. L. REYSENBACH,' N. RAVENSCROFT,2 S. LONG,' D. T. JONES,' AND D. R. WOODS'*Department of Microbiology' and Department of Organic Chemistry,2 University of Cape Town,

Rondebosch 7700, South AfricaReceived 16 October 1985/Accepted 24 January 1986

Synthesis of granulose was investigated in 15 solvent-producing Clostridium strains. Only one of the strainsdid not produce granulose. The structure of granulose in Clostridium acetobutylicum P262 consisted of a

high-molecular-weight polyglucan containing only (1--4) linked D-glucopyranose units. Biosynthesis ofgranulose in C. acetobutylicum P262 was dependent on ADPglucose pyrophosphorylase, and granulose synthaseand mutants defective in granulose accumulation lacked either one or both enzyme activities. Granulose-positive revertants exhibited both enzyme activities. ADPglucose pyrophosphorylase and granulose synthasewere not subject to allosteric control by metabolites. Granulose accumulation and the biosynthetic enzyme

activities were initiated immediately before the pH breakpoint and were detected in cells only at the end of theexponential growth phase. Granulose accumulation did not occur under conditions of nitrogen limitation,excess carbon, or excess energy.

The accumulation of glycogen-like reserve polymers iswidespread among bacteria and usually occurs under condi-tions in which growth is limited in the presence of an excesscarbon and energy source (9, 28). Glycogen biosynthesisoccurs via the ADPglucose pathway in bacteria (28). The keyenzymes involved in glycogen synthesis are ADPglucosepyrophosphorylase (EC 2.7.7.27) and glycogen (granulose)synthase (EC 2.4.1.21) which catalyze the followingreactions:

ATP + glucose 1-phosphate = ADPglucose + PPiADPglucose + 1,4 glucan -- 1,4 glucosyl glucan + ADP

Among the clostridia, the accumulation of intracellulargranules of reserve material is widespread. In most Clostrid-ium species which have been investigated, the reservematerial has been shown to be a polyglucan which has beengiven the generic name granulose (10). The chemical struc-ture of granulose has been determined only in a few species,including Clostridium butyricum (4, 18) and Clostridiumbotulinum type E (34), in which granulose consisted of abranched glycogenlike polymer composed of linear chains ofa(1->4)-linked D-glucopyranose units showing occasionalbranching (1->6) at linked units. Brown et al. (6) reported thepresence of two types of reserve glucans in Clostridiumpasteurianum, one an amylopectin showing a(1--4) linkageand the other a dextran showing a(1-*6) linkage, both ofwhich exhibited only limited branching. However, Darvil etal. (8) reported that an amylopectinlike polysaccharide wasthe only reserve glucan which occurred in six wild-typestrains of C. pasteurianum. Methylation analysis demon-strated that (1->6)-linked D-glucose units accounted for lessthan 2% of the entire glucose content of these organisms.Granulose accumulation normally occurs at the end of

exponential growth before the onset of sporulation and isusually degraded during spore formation. This suggests thatthe reserve polyglucan can serve as a source of energy andcarbon for the formation of spores and their maturation (4,31, 33).

* Corresponding author.

A relationship between the accumulation of granulose,solvent production, and sporulation in an industrial strain ofClostridium acetobutylicum P262, was reported by Jones etal. (20). In this strain, the accumulation of granulose oc-curred concomitantly with the shift from an acid-producingmetabolism to a solvent-producing metabolism at the end ofthe exponential growth phase in batch culture. The accumu-lation of granulose was also linked to the production of aswollenphase-bright clostridial stage, capsule production,and the initiation of a forespore septum. Mutants of the P262strain were isolated which were unable to accumulategranulose, form a clostridial stage, produce capsules, un-dergo the shift to the solventogenic phase, or sporulate (20,22).The isolation of pleiotrophic cls mutants furnished evi-

dence that the various stationary-phase physiological andmorphological events may share common regulatory com-ponents. Further evidence for common regulatory compo-nents for these stationary-phase events was provided bynutrient studies. Under conditions of nitrogen limitation, thecells failed to undergo the shift to solvent production and didnot produce granulose, a clostridial stage, capsules, orendospores (22).A number of features indicate that the granulose

biosynthetic pathway could be a useful system for the studyof the regulation of stationary-phase events in C.acetobutylicum. Granulose is easy to assay, mutants arereadily isolated, and there seem to be common regulatorycomponents linking granulose and solvent production. Wehave investigated the nature, biosynthesis, and regulation ofgranulose in wild-type and mutant strains.

MATERIAL AND METHODSBacterial strains, media, and growth conditions. The C.

acetobutylicum, Clostridium beijerinckii, (Clostridiumbutylicum), and Clostridium saccharoperbutylacetonicumstrains used are listed in Table 1. The C. acetobutylicumP262 wild-type strain has been described previously (20-23,30).

Granulose-negative mutants of C. acetobutylicum P262were isolated by the method of Robson et al. (31) after

185

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from

186 REYSENBACH ET AL.

TABLE 1. Bacterial strains

CuLlture collectionno. or reference

C. ac(etobutylicuinn ............ ............ P262 (20)C. a(wetobutyvliuflll7 ............ ............ ATCC 824C. acetobutylihurn ........... ............. ATCC 3625C. ucetobutyli(wurM ............ ............ ATCC 4259C. (icetobutyliu/lon ............ ............ ATCC 8529C. aClCtobutyli(ln..........ur.. ............. ATCC 10132C. c(etobutyli(win ............ ............ NCIB 2951C. acetobittvlium.........u.n.. ............ NCIB 6441C. cwetobutyll(um ............ ............ NCIB 6442C. a( etobIItXicwn.........I.n.. ............. NCIB 6443C. acetobltylumi ..u ........................ NCIB 6444C. ac etobutvIkituin ............ ............ NCIB 6445C. wlcetobiutylilm.........11/.. ............. NCIB 8653C. beijerinckii (C. butykiwuin) ...................... NRRL B593C. sacha(Icaropehitvla( etonicitin.................... ATCC 27021

treatment of cells with ethyl methanesulfonate (2.5%[vol/vol]) as described previously (20, 23). The strains weremaintained as spore suspensions in sterile distilled water at4°C (22). The bacteria were propagated anaerobically at 37°Cin buffered clost/idial basal medium (CBM) (26). The pro-duction of granulose and endospores in the different strainsof Clostr-idiirn was determined after growth in reinforcedclostridial medium (RCM) (Biolab), CBM, tryptone yeastglucose medium (TYGM) (19), and the sporulation minimalmedium (CAMM) of Long et al. (21). Granulose productionand endospore formation were also determined after growthon CBM agar and CBM agar in which glucose was replacedby sucrose (4% [wt/vol]).The production of granulose, solvents, and endospores by

C. acetobittyliclirm P262 was determined in CAMM in whichglucose was replaced by fructose (60 g/liter), galactose (60g/liter), sucrose (60 g/liter), maltose (60 g/liter), raffinose (30g/liter), L-arabinose (30 g/liter), or D-xylose (30 g/liter). Theeffect of different ammonium sources on the production ofgranulose was determined in CAMM in which diammoniumhydrogen phosphate (DAP) was replaced by ammoniumacetate, ammonium sulfate, and ammonium nitrate added atequivalent ammonium concentrations.

Analytical and assay methods. Total bacterial counts,clostridial stage counts, and spore counts were determinedwith a Thoma counting chamber (Weber Scientific Interna-tional, Lancing, England) and Zeiss photomicroscope fittedwith phase- and interference-contrast optics. The presenceof granulose was determined by staining with iodine.

Granulose was assayed by the method of Robson et al.(31). Granulose was determined as glucose released afteracid hydrolysis by using a Beckman Glucose Analyzer 2(Beckman Instruments, Inc., Fullerton, Calif.). The gran-ulose was expressed as micrograms of glucose deposited per108 cells.Glucose consumption during the fermentation was deter-

mined by monitoring glucose oxidation of residual glucose inculture filtrates. Acetate, butyrate, acetone, butanol, andethanol were measured by gas chromatography as describedpreviously (22). Protein was measured by the method ofLowry et al. (24).

Extraction of granulose. Cells grown in CBM were har-vested during the early stationary growth phase, and thepellets were washed and suspended in 25 mM Tris hydro-chloride buffer (pH 8.0; 0.5 g [wet weight] per ml). The cellswere disrupted by passage through a Yeda Press (Yeda

Scientific Instruments, Rehovot, Israel). A preparation ofgranulose granules was obtained from the extract by differ-ential centrifugation as described by Robson et al. (31). Thepolysaccharide was extracted by the KOH and ethanolprocedure of Darvil et al. (8).

Gel filtration. Lyophilized cell samples (2 mg) in 1 ml of 1M NaCI solution were sonicated (10 min), applied to acolumn (60 by 0.9 cm) of Sepharose 4B, and eluted with 1 MNaCI (14.2 ml/h). The column was calibrated by usingdextran standards. Fractions (1.42 ml) were collected andanalyzed for carbohydrate by using the phenol-sulfuric acidcolorimetric method (7).

Acid hydrolysis. Polysaccharide samples were hydrolyzedwith 2 M trifluoroacetic acid at 100°C for 18 h. Afterhydrolysis, the acid was removed by codistillation withmethanol. Derived alditols were acetylated with sodiumacetate and acetic anhydride for 2 h at 100°C.

Methylation analysis. A lyophilized sample (10 mg) wasmethylated by a modification (15) of the standard proceduredescribed by Hakomori (14). Two such treatments werefollowed by a treatment described by Purdie and Irvine (29)to yield a permethylated polysaccharide showing no freehydroxyls in the infrared spectrum. The permethylatedpolysaccharide (4 mg) was hydrolyzed (2 M trifluoroaceticacid, 100°C for 16 h), and the derived partially methylatedalditols were acetylated as described above.

Thin-layer chromatography. Thin-layer chromatographywas done on Kiesel gel 60 F254 (Merck) by using the solventsystem CHC13-MeOH-H2O (10:10:3.5) for sugar analysis ofthe polysaccharide hydrolysate and by using the solventsystem 3%c NH3 in n-butanol for the partially methylatedhydrolysate.

Gas-liquid chromatography. Analytical gas-liquid chroma-tography separations were performed with a Carlo ErbaStrumentazione 4200 instrument (S.E.A.L. Instrumentation)fitted with dual flame-ionization detectors. A Supergrator-3Aelectronic integrator (Columbia Scientific Instruments) wasused to measure the peak areas. A 3% OV-225 bondedsilica-capillary column (J & W Scientific, Inc.) was used toseparate the acetylated alditols of methylated sugars.

Gas-liquid chromatography mass spectrometry. The gaschromatograph was interfaced with a V.G. Micromass 16Fmass spectrometer (V.G. Analytical). The mass spectrawere recorded at 70 eV, and the position of methoxylsubstitution was determined by using the data of Bjorndal etal. (5).

Proton magnetic resonance spectra. Samples (20 mg) wereprepared by dissolving in D2O after lyophilizing four timesfrom D2O solutions. The spectrum was recorded on a BrukerWH-90 instrument (Bruker) (90 MH2). A total of 983 scanswere accumulated with the probe at 80°C.

Iodine reaction. The adsorption spectrum of the iodinepolysaccharide complex was measured as described byArchibald et al. (2).

Preparations of cell extracts. Cells were harvested andwashed twice at 4°C in Tris hydrochloride buffer (25 mM, pH8.0) containing 1 mM dithiothreitol. Suspension was carriedout under an atmosphere of hydrogen. The bacteria (0.5 g[wet weight] per ml of buffer) were disrupted under nitrogenwith a Yeda Press, and the cell lysate was centrifuged at10,000 x g for 15 min at 4°C. The supernatant was used asthe source of ADPglucose pyrophosphorylase. The pelletwas suspended in Tris-dithiothreitol buffer and used as thesource of granulose synthase. Both crude extracts could bestored at -70°C without loss in enzyme activity.Enzyme assays. ADPglucose pyrophosphorylase was as-

Appt-. ENVIRON. MICROBIOL.

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from

REGULATION OF GRANULOSE IN C. ACETOBUTYLICUM

sayed by determining the rate of synthesis of ADP-[14CJglucose from [14C]glucose 1-phosphate and ATP asdescribed by Robson et al. (31). The assay mixture (0.2 ml)contained 0.15 Lmol of D-[14C]glucose 1-phosphate (3[LCi/[Lmol), 2 pLmol of ATP, 4 [imol of MgCl., 20 1Lmol of Trishydrochloride (pH 8.0), 1 U of inorganic pyrophosphatase,and crude enzyme. ADPglucose pyrophosphorylase activitywas expressed as nanomoles of ADPglucose produced perminute per milligram of protein. Granulose synthase activitywas assayed by measuring the rate of incorporation ofADP-[14C]glucose into ot 1,4-glucan as described by Robsonet al. (31). The assay mixture (0.2 ml) contained 0.2 nmol ofADP-[14C]glucose (0.1 RCi/imol), 0.05 Fg of amylopectin, 5p,mol of Tris hydrochloride (pH 8.0), 1 pmol of MgCl., andcrude enzyme. Granulose synthase activity was expressedas nanomoles of glucose incorporated per minute per milli-gram of protein.The effects of the following substances which act as

allosteric effectors of some bacterial ADPglucose pyrophos-phorylase and granulose synthase enzymes (28) were deter-mined at a final concentration of 3 mM: acetyl-coenzyme A,ADP, AMP, fructose 1-phosphate, fructose 6-phosphate,fructose 1,6 diphosphate, glucose 6-phosphate, PP.phosphoenolpyruvate, 3-phosphoglycerate, 6-phosphogluco-nate, pyridoxal 5'-phosphate, pyruvate, ribose 5-phosphate.NAD+, NADH, NADP', and NADPH.

RESULTS

Granulose in C. acetobutylicum and related strains. Thepresence of granulose in the 15 solvent-producing Clost/icl-iiun strains (Table 1) was determined by direct microscopicexamination of iodine-stained cells. In five of the strains(P262, ATCC 10132, ATCC 27021. NCIB 6445, and NCIB8653), granulose accumulation and phase-bright clostridialforms occurred readily in cells grown in complex liquidmedia and on CBM agar plates containing either glucose orsucrose. In RCM, TYGM, or CBM, at least 60% of the cellsof these strains exhibited substantial accumulation ofgranulose at the end of the exponential growth phase, and allof these strains sporulated. In TYGM, P262, ATCC 10132.NCIB 6445, and NCIB 8653 produced 5.8 to 8.4 g of solventper liter, whereas ATCC 2701 produced 2.4 g of solvent perliter. Four of these strains grew well and produced granulosein CAMM, but ATCC 10132 grew very poorly in this minimalmedium.Granulose was not observed in iodine-stained cells of nine

Clostridiium strains (ATCC 824, ATCC 3625, ATCC 4259.ATCC 8529, NCIB 2951, NCIB 6441, NCIB 6442, NCIB6443, and NCIB 6444) after growth in any of the liquidmedia. After growth for 2 to 4 days on CBM agar containing4% sucrose, seven of these strains accumulated smallamounts of granulose, but the ATCC 824 strain producedhigh levels of granulose. The NCIB 6444 strain did notproduce detectable granulose on the agar medium. C.beijerinckii NRRL B593 produced low levels of granulose onCBM 4% sucrose agar and in RC and TYG liquid media.Solvent production by these 10 strains in TYGM variedbetween 2.4 and 7.4 g of solvent per liter.The 10 strains which produced little or no granulose in

liquid media showed lower levels of sporulation in liquidmedia than did the 5 granulose-producing strains. Endosporeproduction by the 10 strains on CBM agar was very variableand ranged from >1% (ATCC 3625) to 70% (ATCC 824).

Characterization of granulose from C. acetobutylicum. Sam-ples of intracellular polysaccharide granules were prepared

from stationary-phase granulose-rich cells of C. acetobu-tvlicum P262. The crude polysaccharide was eluted in thevoid volume of the Sepharose 4B column, indicating that itconsisted of components with a molecular mass greater than6 x 106 daltons (5).The sample was hydrolyzed (18 h, 100C, 2 M trifluoro-

acetic acid) into the constituent sugars. Thin-layer chroma-tography indicated that glucose was the only sugar presentafter hydrolysis. This was confirmed by gas-liquid chroma-tography analysis of the derived alditol acetates.The nature of the linkage between the glucose residues

was investigated by methylation analysis. Full methylationwas achieved by using two treatments (described byHakomori [14]) followed by a subsequent (Ag2O/CH3I) treat-ment (described by Purdie and Irvine [29I) for 2 days.Methylation was assumed to be complete when the hydroxylpeak was no longer present in the infrared spectrum. Thehydrolysate of the permethylated polysaccharide yielded asingle spot after thin-layer chromatography and was shownto be 2,3,6-tri-O-methyl-glucose by gas-liquid chromatogra-phy and confirmed by gas-liquid chromatography mass spec-trometry. No evidence for branch points was found, indicat-ing that the polymer consisted of cx (1--4)-linked glucoseresidues. H-nuclear magnetic resonance of the polysaccha-ride showed a single peak in the anomeric region at 5.23 ppmwith a coupling constant of J1.2 = 3 Hz. These results aretypical of glucopyranosyl residues that are ot linked. Thepurified polysaccharide from C. acetoblitylicull P262 pro-duced a complex with iodine, with a maximum absorptionbetween 540 and 545 nm.

Biosynthesis of granulose in C. acetobutylicum P262. Theaccumulation of granulose and the activities of ADPglucosepyrophosphorylase and granulose synthase in crude cellextracts were determined in C. a(cetobutylicum P262 grownin CAMM. Long et al. (21, 22) showed that in CAMM, thetime sequence of the morphological changes associated withthe onset of solvent production and sporulation in C.acetobltylicumi P262 was sufficiently distinct and constant tofacilitate correlative physiological and biochemical studies.In this medium, granulose accumulation was first detectedafter approximately 25 h, when the cells reached the end ofthe exponential growth phase (Fig. 1). Initiation of granuloseaccumulation in a small proportion of the cells occurred afew hours before the pH breakpoint, and at the breakpoint,granulose accumulation was visible in the majority of cells.The pH breakpoint also coincided with the shift to solventproduction and an increase in glucose consumption. Maxi-mum granulose deposition (250 to 300 p.g of glucose depos-ited per 108 cells) occurred after approximately 45 h, whenover 80% of the available glucose had been consumed andforespore septation had been initiated. At this stage,granulose accumulation within the cells accounted for be-tween 40 and 55% of the dry weight of the cells (determinedin four separate experiments).ADPglucose pyrophosphorylase and granulose synthase

were not detected before 25 h, and the activity profiles ofboth enzymes coincided with the granulose profile (Fig. 1).The enzyme activities decreased during spore developmentand maturation when granulose was mobilized within thespore mother cells. Approximately 70 to 80% of thegranulose was mobilized during spore development, and nocell-free granulose was detected.

Effect of nutrients on granulose accumulation in C.acetobutylicum P262. The effect of the type and concentrationof carbohydrate and ammonia sources on the production ofgranulose in C. a(etobitylicumn P262 grown in CAMM was

VOL. 52, 1986 187

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from

188 REYSENBACH ET AL.

v 10004uJo 75

0- 500z2 25

ICU. 0

-zIui

a

Uo

0

400rz24-.-US 300

0200

n0

z 1004c

00

0 10 20 30 40 50TIME h

60 70 80

Q

U-

CL

!OUJ

FIG. 1. Accumulation of granulose and the activities ofADPglucose pyrophosphorylase and granulose synthase in C.acetobutylicum during a batch fermentation in CAMM. (a) 0, Totalcell count; A, percentage of total cells containing forespore septa;K, pH; O, total solvents. (b) 0, Granulose deposition; A,ADPglucose pyrophosphorylase activity; A, granulose synthase; 0,residual glucose in CAMM. Granulose deposition is expressed asmicrograms of glucose per 108 cells. Enzyme-specific activity isexpressed as nanomoles of glucose incorporated per minute permilligram of protein.

investigated. Cultures grown in CAMM containing glucose,fructose, galactose, sucrose, or maltose (60 g/liter) all exhib-ited a shift to solventogenesis and granulose accumulationduring the latter part of the fermentation, and 70 to 95% ofthe cells formed the clostridial stage. Granulose accumula-

0

I-

0a.

LU

LU

0-J

4

APPL. ENVIRON. MICROBIOL.

tion was initiated just before the pH breakpoint and reacheda concentration of between 190 and 260 pzg of granulose per108 cells approximately 15 h after the breakpoint. Thesecultures all produced between 11 and 15 g of solvent perliter.The growth of C. acetobutylicum P262 was inhibited by 60

g of raffinose per liter and the pentose sugars L-arabinose andD-xylose. Cultures of CAMM containing 30 g of these sugarsper liter produced solvents (5.2 to 6.5 g/liter) and granulose(60 to 80 [ig of glucose per 108 cells). Similar levels ofsolvents and granulose were produced by cells grown in 30 gof glucose per liter.The nature of the nitrogen source (DAP, ammonium

acetate, ammonium nitrate, ammonium sulfate), added atequivalent concentrations of ammonia, did not affect theproduction of solvents (11 to 15 g/liter) or granulose (200 to300 ji.g of glucose per 108 cells).Glucose (Fig. 2) and DAP (Fig. 3) concentrations affected

solvent and granulose production. Neither solvents norgranulose were detected in C. acetobutylicum P262 culturescontaining <10 g of glucose per liter or <0.5 g of DAP perliter. An increase in glucose or DAP concentrations resultedin a proportional increase in granulose and solvent produc-tion, and maximum amounts of granulose and solventsoccurred at 60 g of glucose per liter and 4 g of DAP per liter.

Characteristics of ADPglucose pyrophosphorylase.ADPglucose pyrophosphorylase was not detected in crudecell extracts of C. acetobutylicum P262 disrupted underaerobic conditions. When cell disruption was performedunder an atmosphere of hydrogen, the crude enzyme extractwas not affected by subsequent exposure to aerobicconditions and was stable at -70°C. Crude extracts ofADPglucose pyrophosphorylase utilized glucose 1-phosphate in the presence of ATP. Plots of enzyme activityagainst substrate concentration were hyperbolic for bothglucose 1-phosphate and ATP, and the Km values were 0.33and 3.23 mM, respectively. The optimum pH forADPglucose pyrophosphorylase was 8.0, and the enzymewas inactivated at pH values below 7.0 and above 11.0. Theenzyme showed a marked nucleotide specificity for ATP.Replacement ofATP with equimolar concentrations (10 mM)of UTP, GTP, or CTP resulted in a 68, 94, and 96% decrease,

3uu

250E

lox

co

z

-J

ul

20 O

en o

co cnc10 0

zO

100 cfl

Cl) 0

z

0 m0

2001-

1501

100

50

'0

GLUCOSE g /l

FIG. 2. Effect of glucose concentration on granulose and solventproduction in C. acetobutylicum P262. 0, Total cell count; 0,

granulose deposition; A, total solvents. Granulose deposition isexpressed as micrograms of glucose per 108 cells.

0 1 2 3 4 5 6DAP g/l

5

4 -- 20-

x en3 m 3co w

1O0

2. -lo-i -i

<

1

0 -0

FIG. 3. Effect of DAP concentration on granulose and solventproduction in C. acetobutylicum P262. 0, Total cell count; 0,

granulose deposition; A, total solvents. Granulose deposition isexpressed as micrograms of glucose per 108 cells.

- 6lu

04

- 40

-i

O.4

a

C"i1

b

50 -4C

10 IA\A - 2A

10 -2/0\ *

C' 0~~~~~~~~~~~~~~~~~~~~~~~~l

i - -

0/j '-0-------

'

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from

REGULATION OF GRANULOSE IN C. ACETOBUTYLICUM

respectively, in ADPglucose pyrophosphorylase activity.The enzyme was not activated or inhibited by the addition ofAMP or ADP at final concentrations of 3 to 50 mM. Inaddition, 13 other compounds, including those metabolitesreported to act as allosteric effectors of the ADPglucosepyrophosphorylase in other bacteria which accumulatepolyglucans, were tested as possible effectors of this enzyme(28). None of the following compounds (3 mM) enhanced orinhibited the activity of C. acetobutylicum P262 ADPglucosepyrophosphorylase: acetyl-coenzyme A, fructose 1-phosphate, fructose 6-phosphate, fructose 1,6-diphosphate,glucose 6-phosphate, PP1, phosphoenolpyruvate, 3-phosphoglycerate, 6-phosphogluconate, pyridoxal 5'-phosphate, pyruvate, ribose 5-phosphate, NAD+, NADP+,NADH, and NADPH.

Characteristics of granulose synthase. The preparation ofcrude enzyme extracts of granulose synthase from C.acetobutylicum P262 was not affected by exposure to aero-bic conditions, and 98% of the enzyme activity was associ-ated with the pellet obtained during extraction. There was alinear relationship between granulose synthase activity andADPglucose at substrate concentrations between 0 and 0.5mM, and the Km for ADPglucose was 0.4 mM. Granulosesynthase activity was not detected when UDPglucose (1mM) was substituted as a substrate. Granulose synthaseactivity was optimal between pH 5.5 and 9.4, was com-pletely inhibited at pH 4.0, and was 80% inhibited at pH11.0. Glucose 1-phosphate and ATP did not affect theenzyme, but the enzyme was sensitive to ADP and AMP (30mM), which caused a 50 and 20% decrease, respectively, inenzyme activity. None of the other compounds (3 mM)tested as possible effectors of ADPglucose pyrophos-phorylase enhanced or inhibited the activity of granulosesynthase.

Characterization of granulose-negative mutants. The isola-tion of 2 types of C. acetobutylicum P262 mutants whichwere defective in granulose accumulation and did not formswollen phase-bright clostridial stages was reported by Longet al. (23). One type of granulose-negative mutant producednormal levels of solvents and continued to produce matureendospores, although in most cases at reduced levels. Theother type of mutant (cis) did not undergo a shift to thesolventogenic phase, and these mutants failed to producesolvents, extracellular capsules, and endospores. Mutants ofboth phenotypes were assayed for granulose and granulosebiosynthetic enzyme activities in crude cell extracts ob-tained at the end of the exponential growth phase. Threegranulose-negative mutants which produced high levels ofsolvents (14 to 16 g/liter) and sporulated in CAMM wereassayed for enzyme activity. Two of these mutants showedno granulose synthase activity but produced normal levels ofADPglucose pyrophosphorylase. The other mutant lackedboth enzyme activities. Ten cis mutants were investigated,and they all failed to produce solvents or endospores andlacked both enzyme activities. Spontaneously occurringgranulose-positive revertants were isolated from two of thecls mutants. These produced solvents and endospores andcontained active granulose biosynthetic enzymes.

DISCUSSION

The structure of granulose in C. acetobutylicum P262 wassimilar to that in C. pasteurianum (6, 8) and consisted of ahigh-molecular-weight polyglucan containing only a(1->4)-linked D-glucopyranose units. This structure differed fromthe glycogen-like polymers with ot(1--6)-linked branched

chains which occur in C. butyricum and C. botulinum (4, 18,34).

Biosynthesis of granulose in C. acetobutylicum P262 wasdependent on ADPglucose pyrophosphorylase andgranulose synthase, and mutants defective in granuloseaccumulation were defective in either one or both enzymeactivities. Granulose-positive revertants exhibited both en-zyme activities.

In C. acetobutylicum P262, granulose accumulation andthe biosynthetic enzyme activities were detected in cellsonly at the end of the exponential growth phase. Theseevents were associated with the cessation of cell division andthe onset of solvent production and spore formation. Theabsence of granulose enzyme activities in vegetative cellsand the onset of enzyme activities associated with the shiftto the solventogenic phase suggest that these enzymes areinduced or derepressed in stationary phase. The derepres-sion of synthesis of ADPglucose pyrophosphorylase at theend of exponential growth was also reported to occur in C.pasteurianum (27). The absence of activity of both enzymesin all of the cls mutants and the reappearance of activity inthe revertants suggest that the enzymes responsible forgranulose biosynthesis may be associated in an operon ashas been demonstrated to occur in Salmonella typhimurium(32). The synthesis of enzymes involved in the final reactionsof solvent production has also been shown to be initiated justbefore the shift to the solventogenic phase in C.acetobutylicum (1, 16, 17). The isolation of cls mutantsdefective in both pathways and the resumption of granuloseand solvent production in revertants of these mutants sug-gests that both pathways may share some common regula-tory elements.The mechanism of enzyme regulation of both these path-

ways is not known, but the shift to solvent and granuloseproduction has been shown to be associated with the accu-mulation of threshold concentrations of acid end productswhich are produced during the initial phase of the fermenta-tion (3, 11-13, 22, 25, 35). Limitation of nutrients such asglucose or ammonia does not appear to be directly involvedin the initiation of the shift to the solventogenic phase andgranulose accumulation (22). In other bacteria which accu-mulate glycogen, the accumulation of this storage product isfavored under conditions of nitrogen limitation and excesscarbon and energy (9, 28). However, in C. acetobutylicumP262 cells grown in ammonium-limited media in the presenceof excess glucose, granulose accumulation did not occur andsolvent production and endospore formation were inhibited.The C. acetobutylicum P262 ADPglucose pyrophos-

phorylase was not affected by metabolites or allostericeffectors which control the enzyme activity in other groupsof bacteria. The absence of fine control of ADPglucosepyrophosphorylase activity has also been shown in C.pasteurianum (31), but in contrast to C. acetobutylicumP262, the C. pasteurianum enzyme was inhibited by ADP.The granulose synthase of C. acetobutylicum was similar toanalogous bacterial enzymes in that its activity was notsubjected to allosteric control by metabolites (28). In C.acetobutylicum P262 and C. pasteurianum (31), granulosesynthase was inhibited by ADP and AMP.

In C. acetobutylicum P262, the concentration of granulosewithin the cells was observed to decrease during sporematuration, and granulose mobilization was preceded by a

decline in both ADPglucose pyrophosphorylase andgranulose synthase activities. This suports the suggestion (4,31, 33) that granulose may be utilized as a source of carbonand energy during sporulation in Clostridium species. The

VOL. 52, 1986 189

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from

190 REYSENBACH ET AL.

isolation of granulose mutants which produced solvents andmature spores at a lower frequency indicated that granuloseis not essential for sporulation, but it may result in moreeffective sporulation under appropriate conditions.The ability to synthesize granulose appears to be wide-

spread among solvent-producing Clostl idiuiim strains, andonly 1 of the 15 strains investigated did not produce visiblegranulose under the conditions tested. However, only five ofthe strains tested showed a significant accumulation ofgranulose in all of the growth media tested. The remainderaccumulated a limited amount of granulose in a small pro-portion of the cells after prolonged incubation in one of theculture media.

LITERATURE CITED1. Andersch, W., H. Bahl, and G. Gottschalk. 1983. Level of

enzymes involved in acetate, butyrate, acetone and butanolformation by Clostriidiim (cetobutV1licum. Eur. J. AppI. Micro-biol. Biotechnol. 18:327-332.

2. Archibald, A. R., I. D. Fleming, A. M. Liddle, D. J. Manners, G.A. Mercer, and A. Wright. 1961. ox-1,4 Glucosans. XI. Theabsorption spectra of glycogen and amylopectiniodine com-plexes. J. Chem. Soc. 1961:1183-1190.

3. Bahl, H., W. Andersch, K. Braun, and G. Gottschalk. 1982.Effect of pH and butyrate concentration on the production ofacetone and butanol by Clostriiduiin a(cetobutl /liflu grown incontinuous culture. Eur. J. Appl. Microbiol. Biotechnol.14:17-20.

4. Bergere, J. L., M. Rousseau, and C. Mercier. 1975. Polyosideintracellulaire implique dans la sporulation de Clostiridiium71butYri umn. I. Cytologie, production et analyse enzymatiquepr6liminaire. Ann. Microbiol. (Inst. Pasteur) 126:295-314.

5. Bjorndal, H., B. Lindberg, and S. Svensson. 1967. Gas-liquidchromatography of partially methylated alditols as their ace-tates. Acta Chem. Scand. 21:1801-1804.

6. Brown, R. G., B. Lindberg, and E. J. Laishley. 1975. Charac-terization of two reserve glucans from Clostridiumn pasteur-ii11i11/1. Can. J. Microbiol. 21:1136-1138.

7. Churms, S. C. 1982. Section II. Detection techniques, p.171-20)7. In G. Zeweig and J. Sherma (ed.). CRC handbook ofchromatography. carbohydrates I. CRC Press, Inc., BocaRaton, Fla.

8. Darvil, A. G., M. A. Hall, J. P. Fish, and J. G. Morris. 1977. Theintracellular reserve polysaccharide of Clostiilidiur pasteur-iauuIrln. Can. J. Microbiol. 23:947-953.

9. Dawes, G. A., and P. J. Senior. 1973. The role and regulation ofenergy reserve polymers in microorganisms. Adv. Microb.Physiol. 10:135-266.

10. Gavard, R. J., and G. Milhaud. 1952. Sur un polysaccharideisole de Clostriidiurtn butvricitin. Ann. Inst. Pasteur (Paris)82:471-483.

11. Gottschal, J. C., and J. G. Morris. 1981. The induction ofacetone and butanol production in cultures of Clostr-idiuirniacetobItVIi(cuml by elevated concentrations of acetate and butyr-ate. FEMS Microbiol. Lett. 12:385-389.

12. Gottschal, J. C., and J. G. Morris. 1982. Continuous productionof icetone and butanol by Clostridiiul a(celobultyliclulm growingin turbidostat culture. Biotechnol. Lett. 4:477-482.

13. Haggstrom, L., and N. Molin. 1980. Calcium alginate immobi-lized cells of Clostridiumn a(cetobult/liculml for solvent produc-tion. Biotechnol. Lett. 2:241-246.

14. Hakomori, S. 1964. A rapid permethylation of glycolipid andpolysaccharide catalysed by methylsulfinyl carbanion indimethyl sulfoxide. J. Biochem. 55:205-208.

15. Harris, J. P., R. J. Henry, A. B. Blakeney, and B. A. Stone. 1984.

An improved procedure for the methylation analysis of oligosac-charides and polysaccharides. Carbohydr. Res. 127:59-73.

16. Hartmanis, M. G. N., and S. Gatenbeck. 1984. Intermediarymetabolism in Clostri-idiun acetobutvlicuin: levels of enzymesinvolved in the formation of acetate and butyrate. Appl.Environ. Microbiol. 47:1277-1283.

17. Hartmanis, M. G. N., T. Klason, and S. Gatenbeck. 1984.Uptake and activation of acetate and butyrate in Clostriidiuina(cetobutIlI/ulul. Appl. Microbiol. Biotechnol. 20:66-71.

18. Hobson, P. N., and H. Nasr. 1951. An amylopectin-type poly-saccharide synthesized from sucrose by C. butyricuit. J. Chem.Soc. 1951:1855-1857.

19. Hongo, M., A. Murata, K. Kono, and F. Kato. 1968. Lysogenyand bacteriocinogeny in strains of Clostridiulin species. Agric.Biol. Chem. 32:27-33.

20. Jones, D. T., A. van der Westhuizen, S. Long, E. R. Allcock, S.J. Reid, and D. R. Woods. 1982. Solvent production andmorphological changes in Clostridiurn acetoWbutyliculn. Appl.Environ. Microbiol. 43:1434-1439.

21. Long, S., D. T. Jones, and D. R. Woods. 1983. Sporulation ofClostridiurn acetobutv/icun P262 in a defined medium. Appl.Environ. Microbiol. 45:1389-1393.

22. Long, S., D. T. Jones, and D. R. Woods. 1984. Initiation ofsolvent production, clostridial stage and endospore formation inClostriidiliii ac(etobutyliclllm P262. Appl. Microbiol. Biotechnol.20:256-261.

23. Long, S., D. T. Jones, and D. R. Woods. 1984. The relationshipbetween sporulation and solvent production in Clostridiudna(cetohlbtWiciunl P262. Biotechnol. Lett. 6:529-534.

24. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

25. Martin, J. R., H. Petitdemange, J. Ballongue, and R. Gay. 1983.Effects of acetic and butyric acids on solvent production byClostr-idiunmi acetobut/icumi. Biotechnol. Lett. 5:89-94.

26. O'Brien, R. W., and J. G. Morris. 1971. Oxygen and the growthand metabolism of Clostridiiumn acetobutylic/un. J. Gen. Micro-biol. 68:307-318.

27. Otto, M. C., T. J. MacAlister, and E. J. Laishley. 1972. Thebiosynthesis of a polyglucose polymer in Clostridiu/npistellria/l1ii/f. Can. Fed. Biol. Soc. Proc. 15:764-767.

28. Preiss, J. 1984. Bacterial glycogen synthesis and its regulation.Annu. Rev. Microbiol. 38:419-458.

29. Purdie, T., and J. C. Irvine. 1903. The alkylation of sugars. J.Chem. Soc. 83:1021-1037.

30. Robson, P. M., and D. T. Jones. 1982. Production of acetone-butanol by industrial fermentation, p. 169-213. In D. Chaudeand G. Durand (ed.), Production d'intermediares industriels parculture anaerobie. Societe Francaise de Microbiologie, France.

31. Robson, R. L., R. M. Robson, and J. G. Morris. 1974. Thebiosynthesis of granulose by Clostridiumn pasteurianum.Biochem. J. 144:503-511.

32. Steiner, K. E., and J. Preiss. 1977. Biosynthesis of bacterialglycogen: genetic and allosteric regulation of glycogen biosyn-thesis in Sa/lmonella typ/hinuriulm LT-2. J Bacteriol.129:246-253.

33. Strasdine, G. A. 1972. The role of intracellular glucan inendogeneous fermentation and spore maturation in C/ostridiuinbotluli/iui/ type E. Can. J. Microbiol. 18:211-217.

34. Whyte, J. N. C., and G. A. Strasdine. 1972. An intracellularQX-D-glucan from Clostr-idiuln botulinuin, type E. Carbohydr.Res. 25:435-441.

35. Yu, E. K. C., and J. N. Saddler. 1983. Enhanced acetone-butanol fermentation by Clostridiuin acetobutylicuin grown onD-xylose in the presence of acetic or butyric acid. FEMSMicrobiol. Lett. 18:103-107.

APPL. ENVIRON. MICROBIOL.

on October 24, 2020 by guest

http://aem.asm

.org/D

ownloaded from