Energetics and kinetics of maltose transport in Saccharomyces cerevisiae: A continuous culture study

APPLIED ANDENVIRONMENTALMICROBIOLOGY, Sept. 1993,p.3102-3109 0099-2240/93/093102-08$02.00/0

Copyright ©1993, American Society for Microbiology

Energetics

and

Kinetics of Maltose

Transport

in

Saccharomyces

cerevisiae:

a

Continuous Culture

Study

RUUD A. WEUSTHUIS,* HENDRIK ADAMS, W. ALEXANDER SCHEFFERS,

ANDJOHANNES P. VANDIJKEN

Departmentof Microbiology andEnzymology, KluyverLaboratory of Biotechnology, Delft University of Technology, Julianalaan 67, 2628 BCDelft, TheNetherlands

Received1April1993/Accepted7July 1993

In Saccharomyces cerevisiae, maltose is transported by a proton symport mechanism, whereas glucose transportoccursvia facilitated diffusion. Theenergyrequirementfor maltosetransportwasevaluatedwitha

metabolic model basedon anexperimentalvalue of

YATP

forgrowthonglucoseandanATPrequirement for maltosetransportof 1 mol-mol-1.Thepredictionsofthe modelwereverifiedexperimentallywithanaerobic, sugar-limited chemostatculturesgrowingon arangeofmaltose-glucosemixtures atafixed dilution rate of0.1

h-1.

The biomass yield (grams of cells gram of

sugar-')

decreased linearly with increasing amounts of maltose in the mixture. Theyieldwas 25%lowerduring growth on maltose thanduringthaton glucose, in agreementwith the modelpredictions.During sugar-limited growth,theresidualconcentrations of maltose and glucose in the culture increasedin proportion to their relativeconcentrations inthe medium feed. From the residual maltoseconcentration, the in situ rates of maltose consumption bycultures,and the Km of the maltose carrier formaltose, itwas calculated thatthe amount ofthis carrierwas proportionalto the in situ maltose consumption rate. Thiswas also found for theamount of intracellular maltase. These two maltose-specific

enzymes therefore exert high control over the maltose flux in S. cerevisiae in anaerobic, sugar-limited, steady-state cultures.

Energy required for transport processes is derived from the gradient of the solute over the membrane (passive transport) or is delivered by metabolic processes at the

expense ofmetabolicenergy(active transport). Theenergy

costsof transportareespeciallyimportantincellular metab-olism when the actively transported solute serves as the source of carbon and energy. This affects growth in two ways: the ATPrequirement for biomass formation is higher,

and the energyyield ofdissimilation is lower. The energy

requirement fortransportprocessesmaytakealarge portion of the totalenergybudget of thecellwhen theenergyyield of the substrate is low. An example is the growth of Pseudomonas oxalaticus on oxalate. In this case, half the energy obtained in respiration of the growth substrate is required for transport of thedicarboxylic acid (4, 5).

How-ever, also during the growth ofyeastson sugarsinmineral media, the energy requirement for sugar transport can be

substantial. Verduyn (23) calculated that the theoretical

energy cost of sugar transport in the yeast Candida utilis growingonglucoseis8.2 mmolof ATP. gof

biomass-'

or

20% of the total ATP requirement.

Maltose is an important sugar in the production of beer and in theleavening ofcertaindoughs (1). So far, however, thespecificeffects ofmaltoseonyeast physiology, suchas

theenergetics of growth, have received littleattention. Most investigations have been aimedatglucosemetabolism.

How-ever, maltose is transportedbyproton symport in

Saccha-romycescerevisiae (20, 22), whereas glucose is takenupby

facilitated diffusion. In this study, an attempt is made to

quantify the ATP requirements ofmaltose transport in S. cerevisiae viaacomparison of growthonglucose and growth onmaltose.

* Correspondingauthor.

MATERIALSANDMETHODS

Organism and cultivation conditions. S. cerevisiae CBS 8066wasobtainedfromtheCentraalbureauvoor Schimmel-cultures (Delft, The Netherlands) and maintained on malt agar slopes at 4°C. Chemostat cultivation was performed with ADI 2-liter bioreactors (Applikon Dependable Instru-ments)atadilutionrateof0.10 h'- andaworking volume of 1.00 liter. Cultures were grown under carbon and energy

limitation on a mineral medium described below. The

re-moval of effluent by the standard procedure, continuous, upwardly directed suction from the surface of the culture,

gaverise to differences in cell density between the culture and the effluent ofup to 20%. Under these conditions, the continuous culturetheory isnotvalid(see alsoreference 13). Removing effluent from the middle of the culture when the culture surface made contact with an electrical sensor did notgive risetosuchadifference, and this methodwasused throughout this study.

Thetemperaturewas30°C, and the stirrer speedwas750 rpm. The pH was kept constant at 5.0 by an ADI 1020

biocontroller by the automatic addition of 2 M KOH. To

assure anaerobic conditions, the reactor and the reservoir vesselwere flushed with nitrogen gas at a flow rate of 0.5 liter. min-'. The flow ratewas keptconstantbya Brooks 5876 gas flow controller. The whole experimental setup (reactor, reservoir, and waste vessel) was equipped with

Norprene tubing (Cole-Palmer Corp.). Thedissolved-oxygen tension of theculture wascontinuously monitored with an oxygen electrode (Ingold, no. 34 100 3002) and wasbelow

0.1% air saturation.

The mineralmedium contained (each per liter): 5.0 g of

(NH4)2SO4,

3.0gofKH2PO4, 0.5gofMgSO4 7H20, 15.0

mg of EDTA, 4.5 mg of ZnSO4 7H20, 0.3 mg of CoCl2 6H20, 1.0 mg of MnCl2 4H20, 0.3 mg of CUSO4 5H20, 4.5 mg of CaCl2 2H20, 3.0 mg of FeSO4. 7H20, 0.4 mg of Na2MoO4 2H20, 1.0 mg of 3102 Vol.59,No. 9

MALTOSE TRANSPORT IN S. CEREVISIAE 3103

H3B04, 0.1 mg of KI, and 0.05 ml of silicone antifoam (BDH). After heat sterilization at 120°C and cooling, a filter-sterilized vitamin solution was added to final concen-trations per liter of 0.05 mg of biotin, 1.0 mg of calcium pantothenate, 1.0 mg of nicotinic acid, 25.0 mg of inositol, 1.0 mg of thiamine HCl, and 0.2 mg of para-aminobenzoic acid. Ergosterol and Tween 80 were dissolved in pure ethanol and steamedat 100°Cfor 10 min before being added to the medium to final concentrations of 10 and 420

mg.

liter-',

respectively. Because of this addition, the me-dium feed always contained 10 to 12 mM ethanol, which was taken into account in the calculation of the ethanol yield and

fluxes. Maltose monohydrate and glucose were sterilized separately (17) and added at the ratios indicated to a final sugar concentration of

approximately

25 g

liter-'.

Determination ofdry weight and elemental analysis. The dry weight of the cultures was determinedwith a microwave oven and 0.45-,um-pore-sizefilters accordingtothe method of Postma et al. (16). The carbon, hydrogen, and nitrogen composition of the biomasswasdetermined withan Elemen-talAnalyzer 240B (Perkin-Elmer) (24).

Sugar analysis. The sugar concentrations in the reservoir vessels were determinedwith a glucose analyzer (YSI 2000; YellowSprings Instruments). Maltosewasfirsthydrolyzed to glucose by a-glucosidase (Boehringer, no. 105 414) (19). Maltose was found to contain glucose and maltotriose as

impurities (both approximately 3% [wt/wt] after heat steril-ization). Maltotriose is also hydrolyzed toglucose by cx-glu-cosidase and is not metabolizedby S. cerevisiae CBS 8066

(seealso references 9 and26).The maltoseconcentrations in reservoir media and culture supernatants were therefore corrected for theamountsof maltotriose andglucosepresent. For the determination of residual substrate concentra-tions, aculturesamplewastaken from theculture,frozen in liquidnitrogen within2 s,and stored at -40°C. Priorto the determination of theconcentrations,thesamplewasthawed andcentrifuged at 0°C.The residual concentrations of mal-tose and glucosewere determined with Boehringer test kit 676543 before and after treatment of the supernatant with

a-glucosidase. Although

rapid sampling

was

performed

throughoutthisstudy, it canbe calculated that this isnot a

prerequisite for obtaining accurate data on residual sugar concentrations. Forexample, the

highest

maltose consump-tion rate was 3.5 mmol

g-1

h-1

at a residual maltose

concentration of 0.6 mM.

Thus,

with the amount of cells present(1.6g

liter-')

in 2smaximally0.003

[1.6

x

3.5/(60

x 30)] mmol of maltose or 0.5% (0.003/0.6) disappeared duringthe sampling.

Metabolite

analysis. Ethanol, glycerol, maltotriose,

and

organicacids(2-oxoglutaricacid,

pyruvic acid,

succinic

acid,

and fumaric acid)were determined

simultaneously by

high-pressure liquid chromatography

(HPLC) analysis

with an HPX-87H Aminex ion exclusion column

(300 by

7.8 mm;

Bio-Rad)at30°C.Thecolumnwaseluted with 5mMsulfuric acid at a flow rate of 0.6 ml.

min-'.

Organic

acids were

detected withaWaters 441UVmeter at214nm

coupled

with aWaters741data module.Ethanol,

glycerol,

andmaltotriose were detected by an ERMA ERC-7515A refractive index detector coupled with a Hewlett-Packard 3390A

integrator.

The amount of ethanol produced was corrected for the amount of ethanol in the reservoir

(approximately

10 mM

ethanol,originating fromthe addition of

ergosterol).

Acetic acid couldnotbedeterminedbythis HPLC

method,

since it had thesame retention time as one ofthe medium compo-nents. It was therefore determinedwith

Boehringer

test kit 148261.

Metabolite fluxeswerecalculatedasq =c

DIX,

in which q stands for flux

(mmoles

gram of

biomass-'

h-'),

c stands fortheamount ofsubstrateorproductconsumedor

produced

(mmoles.

liter-'),

D stands for the dilution rate

(per hour),

andXstands for the biomassconcentration in the culture (grams

[dry weight]

liter-').

Gasanalysis.The gas

flowing

outof thereactor wascooled in a condenser

(2°C)

and dried with a Perma Pure

Dryer

(PD-625-12P). CO2

was determined with a Beckman model 864 infrared detector. The effluent gas flow rate was

mea-sured withadevicewhichweconstructed. It consisted ofan

inverted

glass cylinder

filled withwater.Under this

cylinder

but without

touching

the

glass

cylinder,

a water reservoir was

placed

to prevent outflow of the water. The reservoir rested on an electronic balance. When a gas flow was

directed into the

cylinder,

waterflowed out into the

reser-voir. The amount of water, assessed

by

the electronic balance per unit of

time,

was a measurefor the gasflow after corrections for pressure

falls,

temperature, and water ten-sion. The accuracywas

3%,

and the

repeatability

was0.3%.

The

CO2

fluxwas calculated asq =

c.

VIX,

in which V

stands for the gas flow rate

(liters.

min-').

Other

designa-tionsare asdescribed above. Theamountof

CO2

leaving

the culture with the effluent mediumwas

negligible.

Enzyme analysis.

Enzyme

assays were

performed

with a

Hitachi model 100-60

spectrophotometer

at

30°C.

Reaction rates were

linearly

proportional

to the amount of enzyme added. The

preparation

of cellextracts and assays of pyru-vate

decarboxylase

(EC 4.1.1.1)

and

glucose-6-phosphate

dehydrogenase

(EC

1.1.1.49)

weredone

according

toPostma et al.

(18).

Citrate

synthase

(EC

4.1.3.7) activity

was deter-mined

according

to the method of Srere

(21),

and that of hexokinase

(EC 2.7.1.1)

was determined

according

to the method ofPostma et al.

(15).

Maltase

(EC 3.2.1.20)

activity

wasmeasured

by

a

discon-tinuous assay. The reaction mixture contained 100mM

ace-tate buffer

(pH 6.6)

and 60 mM maltose. The reactionwas

carriedoutin 1.0 ml of bufferat

30°C

andwasstarted

by

the addition ofcellextract.Thereactionwas

stopped

atdifferent time intervals

by

the addition of 10 ,ul of 75%

(wt/vol)

trichloroacetic acid. Before the amount of

glucose

in the reaction mixturewasdetermined with the

Boehringer

testkit described

above,

the

pH

of the

sample

wasneutralized

by

the addition of 4.5 ,ul of 10 M NaOH. All enzyme activities are

expressed

as micromoles of substrate converted per minute

milligram

of

protein-'.

Maltosetransport assay. Maltose

uptake

rateswere

deter-minedasdescribed

by

vanLeeuwenetal.

(22) by

measuring

the alkalinization of

weakly

buffered cell

suspensions

after the addition of maltose.

Proteindetermination. The

protein

content of wholecells was determined

by

a modified biuret method

(24).

The amount of

protein

in cell extracts was determined

by

the

Lowry

method.

Graphical

representation

ofdata.Allmetabolic

parameters

are

plotted

as a function of the

composition

of the sugar mixture thatis utilized in termsofhexoseunits on amolar

basis

according

to

(maltose

in feed - residual

maltose)

x

2/[(maltose

in feed - residual

maltose)

x 2+

glucose

infeed

- residual

glucose].

RESULTS

The disaccharide maltose is a

good

model substrate to

evaluatetheenergy

requirements

ofactive sugar

transport

in

S. cerevisiae. Since maltose

hydrolysis

does not

require

VOL.59, 1993

3104 WEUSTHUIS ET AL. A 5394glucose 5394gucose

.8

B out in 2697maltose H+ H+ 899 maltose

111I

E:'~

.1

A 2697maltose FT H- H ) 1102glycerol 8240 ethanol 1102glycerol 8240ethanol

585C02 8240CO2 585CO2 8240 C02 899 maltose 3596 ATP * 3596etanmol 3596 CO2 100gbiomass 100gbiomass

FIG. 1. Metabolicmodel forproductionof 100gof cells of S. cerevisiae CBS 8066 under anaerobic conditions withglucoseas acarbon source(24) (A)and transformation of the modeltothe utilization ofmaltose, assumingthat maltosetransportrequires1 mol of ATPpermol

of maltose(B).One quarter of the maltose that is utilized[899/(2,697+ 899) mmol]isrequiredtogenerate the ATPrequiredfortransportof

the disaccharide.

ATP,theonlydifference between theenergetics of maltose andglucosemetabolism resides in the transport step: trans-port of maltose is active, whereas transport ofglucose is passive.It is thereforeappropriatetouseestablished dataon

theenergeticsofgrowthonglucoseforpredictingthegrowth efficiencyonmaltose.

Metabolic model. A metabolic model for the anaerobic growthof S. cerevisiae CBS 8066 withglucose as acarbon andenergy source(24)is summarized inFig. 1A.The overall

equation for biomass formation is

5,394mmol ofglucose -- 100 g ofbiomass + 1,102mmol

ofglycerol + 8,240mmol of ethanol + 8,825mmol

ofCO2 (1)

where maltose is regarded as two glucose units (in fact, maltose consists of two glucose units minus one water molecule). Itwas therefore convenient to use maltose

mo-nohydrate as the carbon source. Assuming that the ATP

requirement for maltose transport is 1 mol ofATP, caused by a proton-maltose stoichiometry of 1 of the maltose transporter (20, 22)andaproton-ATP stoichiometryof 1 of

the plasma membrane ATPase (11, 12, 14), the net ATP productionfrom 1 mol of maltose is 3 mol of ATP(Fig. 1B). Therefore, compared with growth on glucose, a deficit of

2,697 (5,394/2)mmol of ATP willoccurinthe formation of 100gof biomass froman equivalent amountof maltose:

2,697mmol of maltose -- 100 g of biomass + 1,102mmol

ofglycerol + 8,240mmol of ethanol + 8,825 mmol ofCO2 - 2,697mmol of ATP (2)

Thiscanbereplenished bythe additional dissimilation of899 (2,697/3)mmol of maltose:

899 mmol of maltose --3,596mmol of ethanol + 3,596

mmol ofCO2 + 2,697mmol of ATP (3)

Therefore,the overallequation for the formationof100gof

biomass from maltoseunderanaerobic conditions is 3,596mmol of maltose -- 100g ofbiomass + 1,102 mmol

ofglycerol + 11,836 mmol of ethanol + 12,421 mmol

ofCO2 (4)

This model can be used to predict the production of dry weight, ethanol, CO2, and glycerol during growth of S. cerevisiae with glucose and maltose as carbon sources in

chemostat cultures at a dilution rate of 0.1 h-1. Thus, for example, whereas the biomassyield during growth on glu-cose is 0.103 [100/(5.394 x 180)] g* g-' (equation 1), the

theoretical biomass yield onmaltose is 0.077 [100/(3.596 x

360)] g*

g-1

(equation 4). The ethanol production during growthonglucoseis 82.4(8,240/100)mmol g of biomass-(equation 1).Withadilutionrateof0.1

h-1,

this isproduced

in 10 h or8.24 mmol g of

biomass-' h-1.

Similarly, from equation 4 it follows that the theoretical specific production

rate on maltose is 11.84 (11,836/100 x 0.1) mmol g of

biomass-'

*

h-1

(Table 1).

Methodology ofchemostat cultivation on sugar mixtures. Foracomparison of anaerobic glucose and maltose metab-olism it is important to keep the growth conditions, e.g., growth rate, temperature, and pH, etc., thesame.Theonly

suitable cultivation apparatus thatcan meet these

require-ments is the chemostat. Moreover, since carbon-limited chemostat cultivation gives rise to low residual substrate

concentrations, it is possible to circumvent repression ef-fects and to cultivate organisms on two carbon sources

simultaneously (8).Thisgivestheopportunitytocultivate S. cerevisiaeonvarious mixtures ofglucoseandmaltose in the feed(7).Themetabolic model can thus be tested not only for

growth onglucose andmaltose as sole carbon sources, but also forgrowthonvariousmixtures. Thetheoretical yields, production, and consumption rates with mixtures of glucose and maltose can be calculated as the summation of the parametersforgrowthonthe separate sugars.Thus,sincea

TABLE 1. Predictedyields of biomass and ethanol and specific fluxesof substratesandproducts inanaerobic, carbon-limited

chemostatcultures ofS. cerevisiae CBS 8066 withmaltose monohydrate as carbon sources

Yieldor mM carbonsourceb Maltose/glucose

flUXa Glucose Maltose (%)C

Ybiomass 0.103 0.077 75 Yethanol 0.39 0.42 108 qethanol 8.24 11.84 143 qCO2 8.83 12.42 140 qglycerol 1.10 1.10 100 qglucose 5.29 0 qmaltose 0 3.60 qhexose 5.29 7.20 133

aYields (1) are in grams -gram of

sugar-'

and fluxes (q) are in mmoles.gram-' hour-'.D =0.1h-1.

b Establisheddatafor growthonglucose(24)wereused to predict growth parameters on maltose with the model outlined inFig.1.

cValues of parameters for growth on maltose as percentages of the values forgrowthonglucose.

APPL.ENvIRON. MICROBIOL.

MALTOSE TRANSPORT IN S. CEREVISL4E 3105

FIG. 2. Phase-contrast micrograph of S. cerevisiae CBS 8066

cultivatedincarbon-limitedchemostat cultureataDof 0.1 h'- ona

mixture of75% maltose and 25% glucose. For anexplanation, see

thetext.

25% decrease in cell yield is predicted according to the model (Table 1), growth on a 1:1 mixture of maltose and glucose (in terms of hexose units) should result in a 12.5% decrease. The advantage of this approach is obvious: the calculation of the energetics of maltose transport is not

dependentanymoreon onemeasurement, but the modelcan

be checkedby manymeasurements.

Duringthe growthof S. cerevisiae CBS 8066 onmixtures of maltose andglucose,apeculiar problemwasencountered: cells tended to change their morphology upon changes in cultivation conditions. Culturesgrownonmixtures of these

sugars gaverisetopseudohyphaformation(Fig. 2).Thiswas

associated with adecreasein biomass densities in the culture aftersteady-stateconditions should have been reached(i.e., afterapproximatelyfive volumechanges).Thisphenomenon occurreddespitetheprecautions withrespecttothe effluent

removal system (see Materials and Methods). Regular checks on the biomass densities in the culture and the

culture effluent made it clear that the surface level sensing for effluent removalwasappropriate: the biomass densityin

the culture was always within 1% of that of the culture effluent, suggestingthatoccurrenceand selectionof pseudo-hyphaecouldnotbe ascribed toinadequateeffluent removal.

The formation of pseudohyphae occurred only during growth on glucose-maltose mixtures, not with glucose or

maltose as the sole carbon source. Itwas notreproducible but seemedtobetriggered by changesin theglucose-maltose composition of the medium. For example, when a culture

was switched from a steady-state situationwith maltose as

the sole carbon source to growth on a 75% glucose-25% maltose feedmixture, nopseudohyphaeweredetected after

1 month. When this culture was subsequently grown on

glucose as the sole carbon source until steady state was

establishedand then switchedtoa25%glucose-75%maltose

feed mixture, pseudohypha formation started after 2 days (Fig. 2). Platingof this cultureon maltagargave rise to the

developmentoftwocolonytypes, one consistingof normal

cells and oneconsisting ofelongatedcells. To exclude the

possibility that the elongated cells were an infection, both

colonytypesweretestedbythe Centraalbureauvoor

Schim-melculturesbystandard determination tests andwerefound tobegenuineS. cerevisiae cells.

Thevolume/surface ratio of cells isknown to influence the

cellular

energetics

(10).

In order to circumvent a

possible

effects of cell

morphology

onthe

bioenergetics

of

growth,

a

standard

procedure

was

adopted

for

steady-state

cultivation

on mixturesofmaltose and

glucose

thatavoided

changes

in

morphology

as follows: for each

mixture,

cultivation was

started in a sterilized fermentor

by

batch cultivation. The medium pump was switched on

immediately

after

inocula-tion,

and cellswereallowed togrow

aerobically

for 2 h and then switched to anaerobic conditions. To check for a

steady-state situation,

regular analyses

of biomass and

prod-uct concentrations were

performed. By

this

procedure,

steady

states without

pseudohyphae

were obtained within

six volume

changes.

Verification ofthe metabolic model.S. cerevisiae CBS 8066

was grown

anaerobically

at a D of 0.1

h-1

on

glucose-maltose mixtures

ranging

from 100%

glucose

to 100% mal-tose

according

to the

procedure

described above at a total

reservoirsugar concentration of

approximately

25g

liter-'.

With

increasing

maltose

concentrations,

the biomass de-creased inalinearfashion from 2.42 g

liter-'

on

glucose

to

1.62 g liter-1 on maltose. The

glycerol

concentration de-creased in

parallel

tothe

dry weight.

Theethanol

concentra-tion and the amountof

CO2

in the

off-gas

increased

slightly

(Table

2).

Notall

products

wereidentified

by

HPLC,

butthe

unidentified

products

were found under all

conditions,

and their concentrations did not

change

significantly.

Further-more, the carbon recovery of identified

products

in all

steady

stateswasbetween 98 and104%.

The

experimental

data ofTable 2were used to calculate the actual biomass

yields

and

specific

fluxes of

ethanol,

glycerol,

CO2, glucose,

and maltose. These

data,

plotted

in

Fig. 3,

fit well with the metabolic model. For

example,

during

growth

on amixture of 36.7mMmaltose and 52.2mM

glucose

in themedium feed

(Table

2)

itcanbecalculated that

thebiomass

yield

(defined

as grams of biomass per gram of

sugar) equals

0.086

{1.92/[(36.7

-

0.45)

x 0.360 +

(52.2

-0.36)

x

0.180]}

g gof

sugar-1

[grams

of

biomass/(maltose

in

feed - maltose in culture +

glucose

in feed -

glucose

in

culture)]. The metabolic model

(Table

1) predicts

avalue of

0.088

{[(36.7

-

0.45)

x 0.360x 0.077+

(52.2

-

0.36)

x0.180 x

0.103]/[(36.7

-

0.45)

x 0.360 +

(52.2

-

0.36)

x

0.180]}

g of biomass g of

sugar-1

{[(maltose

in feed - maltose in

culture)

x biomass

yield

on maltose +

(glucose

in feed

-glucose

in

culture)

x biomass

yield

on

glucose]/(maltose

in feed - maltose in culture +

glucose

in feed -

glucose

in

culture)}

and is within the accuracy of the biomass

weight

assay. Note that in this calculation the molecular mass of

maltoseis takenas360 rather than342 g

mol-1

tomake the biomass

yield

on maltose

comparable

to that on

glucose

when

expressed

onagram-per-grambasis. For the calcula-tion of the ethanol

production,

take into account that the reservoir medium contained

approximately

10 mM

ethanol,

whichwas used to dissolve

ergosterol

and Tween 80. The ethanol

yield

isdefined as

(ethanol

in culture - ethanol in

feed)/(maltose

infeed-maltose in culture +

glucose

infeed -

glucose

in

culture).

On

glucose,

the ethanol

yield equals

0.37{193 x

0.046/[(132.1

-

0.48)

x

0.180]}

gof

ethanol.

gof

biomass-1. On the sugar mixture with the

highest

amountof

maltose,

theethanol

yield

is 0.43{197 x

0.046/[(57.5

-

0.59)

x 0.360 +

(3.33

-

0.08)

x

0.180]}

g of ethanol g of

biomass-1. The

higher

ethanol

production during growth

on

maltoseis close to the

predictions

listedin Table 1. Themodel assumesthatthecell

composition

isconstant. Differencesincell

composition

may

change

theATP

require-mentsfor biomass

formation,

thus

leading

to

changes

in the metabolite fluxes related to assimilation and dissimilation.

3106 WEUSTHUIS ET AL. 0 0 .e LU 0 4) .0 CO C) 0 0 4) Cu 0 co OD 3 U, 4) Cu 4) cuo $t Q 4) 0 DU, C). 4)._ -4 .0_4 caC U, ._ Cu cn

m)

0 4)-Cu .0 .0 !9 4) 4-0 0 C) 0.0 011 U5 0 .. C) 00

Cuq

w m Cu la cn (A t) w 0 0 (A 0 0 ca 0 8

c:

0 _o 4) (A ce o 2 ojQ CA4 c = m.?5 0%1 C9'IC0% 00cn 1. 006o r;00 00 tl +1 +1 +1 +1 +1 ONCN C0 _ 0%000%e 00e C-+lcl+1 +1 +1 +1 +1C^ ti t 000o -N _ (N N(N (N (N (N (N +l +1 +1 +1 +1 +1 e0 CD -)CS -0--_ (N-o 000000 +l +1 +1 +1 +1 +1 (N o00n0r0 +1 +1+1+1+1 +1 Il0 0-40R4t +1 +1 +1+1+1 +1 C1400 (N 0* (N 'IO - o----+l +1 +1 +1 +1 +1 00 Lf')cn -0N 0 oN o o oN oN ol (N CD- M CD) +l +1 +1 +1 +1 +1 d 00 N 0N 0%~ 000 0%0%c 0% +1 +1+1+1+1 +1 \COOI'l Qt ON O n I" I,n (N_%oo N _0 +l +1 +1 +1 +1 +1 0No"cr*oo 0000t_ 0 - -- -- -+l +1 +1 +1 +1 +1

ooofi

os6

ofito

00 CD)4~10C)00 CN 0 (N (N(N +l+1 +1 +1 +1 +1 0IC"~' N (N e') 0 N ( oN _0NN en~-(NCq enU' (N 0 _ 4000 +l+1 +1 +1 +1 +1 -0t- 0t N 0 t0- a -00 N 00 & & - ." .' N .0% .5 2 E 0 ow 40 A T F T .0 U E E 0 ._ ED cr 6 4 2 0 15 10 5 0 0.10 ' 0.09 c o 0.08 0.0 0.07

APPL.ENVIRON. MICROBIOL.

0 20 40 60 80 100 maltose/(maltose+glucose) IX) 0 20 40 60 80 100 maltose/(maltose+glucose) (X) 0 20 40 60 80 100 maltose/(maltose+glucose) IX)

FIG. 3. Effect of amounts of maltose and glucose utilized on

specific rates of glucose and maltose consumption (A), specific production rates of ethanol, CO2, and glycerol (B), and biomass yieldsofanaerobic, carbon-limited chemostat cultures of S. cerevi-siae CBS8066 (C). The measured valuesarefitted withthe

meta-bolic model described in Results. The model graphs for sugar

consumption, ethanol, and CO2 productionarecurved because of

the linear decrease in cell yield with the increasing amount of maltose in the medium feed. The values on the x axis are the percentages of maltose from the total amount of sugar utilized. Standarddeviations areindicatedby bars.

Over theentire rangeofsugarmixtures, thecarbon, hydro-gen, andsulfurcontents of thecells didnotchange signifi-cantly (Fig. 4). The nitrogen content, however, increased from7.5to8.0%.Theproteincontentof the cells increased from45.3 + 0.5%on100%glucoseto48.7 + 0.9% (grams of

protein gramsofbiomass-')on 100% maltose.Intermsof thetheoretical ATPrequirementforbiomassformation, this increase in protein content would mean a 3% increase in energyexpenditure. Therefore,thissmall change in biomass

compositionwasneglectedin theevaluationof the energet-ics ofgrowthon maltose-glucose mixtures.

During anaerobic growth on glucose, S. cerevisiae

pro-duces organic acids that can be potent uncouplers. If the

B

co,,

III

I'm-I ethanol

MALTOSE TRANSPORT IN S. CEREV7AL4E 3107 50 40 10 S 0 0 90 0 p D

*

U - - a I ----0 20 40 60 80 100 maltose/(maltose+glucose) (X)

FIG. 4. Effect of amounts of maltose and glucose utilized on

carbon(0), hydrogen (L), nitrogen (0),sulfur(-),and protein (A)

contentsofbiomass in anaerobic, carbon-limitedchemostat cultures

of S. cerevisiaeCBS 8066. Thestandard deviation for the nitrogen,

carbon, hydrogen, and sulfurcontentsis 0.08, and those for protein are indicated. The values on thex axis are the percentages of

maltose from the totalamountofsugarutilized.

productionrate of these acidswereproportional totherate

of maltose metabolism, this would have severe

conse-quencesforthe validity of the model, asthis wouldcause a

progressive decrease of the cell yield with increasing amounts of maltose in the reservoir feed. However, the amounts of acetic acid and pyruvic acid, aswell as

2-oxo-glutaric acid and succinic acid,werelowand decreased with

increasing concentrations of maltose in the feed (Table 2). The specific production ratesof these acids were constant

(data not shown). The decrease of fumaric acid was

pro-nounced, but the absolute concentrations were very low.

Uncoupling of metabolism by such low concentrations of acetic acid andpyruvic acid canbeneglected (25).

Metabolic fluxes andenzymeactivities. During growth ofS. cerevisiae on glucose-maltose mixtures, the specificrateof

sugar consumption (expressed asmmoles of hexose gram

of cells-'

hour-')

slightly increased with increasing mal-toseconcentrations in themediumfeed(Table 1). This is due to the biomass yield on maltose being lower than that on

glucose. This small increase in fluxhadnosignificant effect on keyenzymesof theglycolytic pathway, suchas hexoki-naseandpyruvatedecarboxylase, which remained approxi-mately constant over the whole range of sugar concentra-tions. The same was true for glucose-6-phosphate dehydrogenase and citrate synthase (Fig. 5) (which under anaerobic growth conditions fulfill only assimilatory func-tions).Asexpected,adifferentpatternwasencountered for

themaltose-specific enzymes,maltase and the maltose

car-rier. Maltase was present in glucose-limited cultures at an

activityofapproximately1 U. mgof

protein-'.

Itsamount increased nearlysixfold withincreasing maltose concentra-tions in the feed (Fig. 6A). The amount of maltose carrier also increased withincreasingmaltoseconcentrationsin the medium feed. Thismaybeenvisagedasfollows: theamount of maltose carrier canbecalculated accordingto the

equa-tion V=

Vm.,

SI(Km + s), inwhich Vequalsthe specific

maltose consumption rate inthe culture

(qmaltose), Vma,

is equivalenttotheamountofcarrier,sis the residual maltose

concentration in the culture(Fig. 7),and Km is the Michaelis constantofthe carrier for maltose.By usingavalue of 4 mM for the last parameter (22) it can be calculated that the

T Z ._ 0 C. E -0 c) 1.2 0.9 0.6 0.3 0 5.5 6.0 6.5 7.0 7.5 qhexose

(mmol-g-'-h-')

c 40I._ 0 E 0 a. in 1.5 1.0 0.5 0 5.5 6.0 6.5 7.0 7.5 qhexose

(mmoI-g-1-h-')

FIG. 5. Effect ofglycolytic flux (in mmoles of hexose gramof

biomass-. h-1)onhexokinase (0) and glucose-6-phosphate dehy-drogenase (-) (A) and pyruvate decarboxylase (A) and citrate synthase (0) (B) activities of S. cerevisiae CBS 8066.

amountof maltose carrier(expressedas

Vm,,

inFig. 6B)also

increased withincreasingamountsof maltoseinthemedium feed.

DISCUSSION

Aquantitative analysisof anaerobic chemostatcultures of

S. cerevisiae CBS 8066growingonmixtures of maltose and

glucose showedthatgrowthandproductformation fitted in

amodelbasedon anATPrequirementfor maltose transport of 1 mol of ATP per mol of maltose (Fig. 1). The ATP requirementfor maltose transport is duetothesymportwith protonswhichsubsequentlymustbeexpelled bytheplasma membrane ATPase at the expense of ATP. The model as

presented in Fig. 1 predicts the various parameters for anaerobicgrowthaslisted in Table 1. Theexperimentaldata (Fig. 3) show an excellent fit with these predictions. For example, the cellyieldonmaltosewas25% lowerthan that on glucose. As expected, this decrease in cell yield is proportionaltothe amountof maltoseutilized.

Prerequisites for validation of the model. Relating cell yields to medium composition is oftenvery difficult, since

manyfactors,such asbiomasscomposition and the forma-tion ofby-products, mayaffect cellyields.Forthisreason,it

was decided to studycellyields notonlyon the individual

I

l

' I-

,I

t_s~~:

TF.E_

_

VOL. 59,1993 rr, x A 1.- -! 0 P=§=;w r-lw--l -I I I I I

3108 WEUSTHUIS ET AL. T . 4-0 En I 0 E E x E 6 4 2 0 30 20 10 0 FIG. 6. Relationsh of maltose gramofb

and maltosecarrier(E

maltose-glucose mixti

The Vmin valueform

maltosecarrier, wasc tration, therate ofma

deviationsare indicat

sugarsbutalsoonn not only tested for

sourcebutalsoveri

increase in protein

maltoseconcentrati cannot explain the

maltose-associated mentdue touncouj

Sincethe modeli difference between the transport step, changes due to a d avoided, especially

the cellular energc pseudohyphaeby S

tion on mixtures c

attempts toquantify

2). Usually, pseud

associatedwith nuti

tion did not occur

glucose and maltose

0 1 2 3 4 qmaltose

(mmolIg-l-h-1)

B

I 2

E

m (0 -o 0

0.6

0.4

0.2

o

f,

I I I I I

0

20

40

60

80

100

maltose/(maltose+glucose)

(O)

FIG. 7. Effect ofamounts ofmaltose andglucose utilizedbyS. cerevisiaeCBS8066,cultivated inanaerobic,carbon-limited

chemo-stats, onresidualglucose

(O)

andmaltose(0) concentrations. The

valueson thex axisare the percentages ofmaltose from the total

amountofutilizedsugar.Standarddeviations areindicatedbybars.

II

only

cause for the formation of

pseudohyphae

in this case.

Fortunately, pseudohypha formation by

adopting

a standard

procedure

that allowed

steady-state

0 1 2 3 4 growth with uniform yeast-like

morphology.

Kineticsof mixed substrateutilization. Ourresults confirm

qmaltose

(mmolg-1-h-1)

and extend the original

observations

by Egli et al.

obtained

for

Escherichia

coli (6), that during growth on mixed

sub-ipbetween maltose utilizationf(mmoles strates the residual concentrations

,iomass-

hgo)

and strates are lower than those during growth

ures inanaerobic,carbon-limitedchemostats. substrate

(Fig.

7).

The residual sugar

concentration

de-altoseuptake, with

Vmn

being the amount of pendedon the

composition

ofthemedium feed.Theresidual

calculated

from the residual substrateconcen- sugarconcentrations in carbon-limited

chemostat

culturesof

Lltoseutilization, andaKm of 4 mM. Standard S.cerevisiaewere2 orders ofmagnitude higherthanthose in

by bars. carbon-limited E. coli cultures. This is a reflection

affinity of the sugar uptake systems in S. cerevisiae being much lower than those in E. coli.

By the Michaelis-Menten equation, the residual substrate nixtures. Inthisway,themodelcouldbe concentrations, and the Km of the maltose

growth on maltose as the sole carbon substrate, it was calculated that the amount

ifiedunder additional conditions.Aslight carrier increasedwithincreasingmaltoseconsumption

content was observed with increasing In a separate experiment it was establishedthat

ionsin the medium feed. This, however, significant amount of maltose carrier was still

strong decrease in cell yield. Also, a glucose-limited cultures. The same was also

increase in maintenance energyrequire- amount ofmaltase. The most drastic adaptation

pling could beexcluded (Table 2). levelsoccurredatlow maltoseconcentrations

isbasedon the assumption thatthe only feed (i.e., at low maltose consumption rates).

growthonglucose and thatonmaltoseis maltose flux of 0.5 mmol g-' h- the

other factors such as morphological enzymes was linearly proportional to the consumption

lifferent medium composition should be (Fig. 6). The clear correlation of the

since the surface/volumeratiocan affect enzymes withthein siturateofmaltoseconsumption

etics

(10). The irregular formation of cultures confirms thattheseenzymes exert

cerevisiae growing under sugar limita- overmaltose fluxin the cells.

glucose and maltose interfered with Anaerobiccultivationfortheevaluation

(theenergeticsof maltosetransport (Fig. ofsugar transport. Forthe studyofthe

lohypha formation by S. cerevisiae is maltose-proton symport, anaerobic growth

rient limitation (2). Since thecell elonga- chosen, since particularly during anaerobic

r in sugar-limited cultures growing on ergyrequirements ofsugartransport become

alone, nutrient limitation cannotbe the is duetothe lowATPyieldincatabolism

It

It

MALTOSE TRANSPORT IN S. CEREVISUE 3109

under aerobic growth conditions. Under anaerobic condi-tions, a 25% difference between cell yield on glucose and maltose can be expected (Table 1; Fig. 3). During aerobic growth, the relative effect of the energy requirement for sugar transport is much smaller. For example, with a

P/O

ratio (number of molecules of ATP formed per atom of oxygenused)of 1, aerobic dissimilation of 1 mol of maltose yields 32 mol of ATP. If maltose transport requires 1 mol of ATP, this would be only 1/32 of the ATP produced in catabolism. As aresult, the cell yield (grams of cells gram ofmaltose-1)would be only 3% lower on maltose than on glucose, which is almost within the accuracy of yield deter-mination.

Practical implications of active maltose transport. The results described above may have practical implications for ethanol production with yeasts. When starch (cereals) is usedas afeedstock, it may be profitable to use hydrolysates with high maltose contents: in this way a higher yield of ethanol(grams of ethanol gram of

sugar-')

can be expected (Table 1). With respect to the effects of sugar transport on cellular energetics, it would also be of interest to study various Saccharomyces strains that differ with respect to their modes of hexose uptake. It has been previously re-ported, for example, that certain strains of brewer's yeast have the ability to take up fructose by a sugar-proton mechanism (3). Insuch a case the energetics of growth on sucrose and maltose would be comparable. In contrast to maltose, sucrose is not taken up by yeasts but is hydrolyzed extracellularly by invertase or inulinase to glucose and fructose. If fructose transport would require 1 mol ofATP, the ATPyield of anaerobic catabolism of sucrose would be 3 mol/mol, identical to that of maltose.

ACKNOWLEDGMENTS

We areindebted toC. Verduyn (BIRD Engineering, Schiedam, TheNetherlands) for valuable discussionsand to M. T. Smith of the Centraalbureau voor Schimmelcultures (CBS) for the taxonomic analysisof thepseudohypha-formingvariants.

REFERENCES

1. Beudeker, R. F., H. W.vanDam, J. B.vanderPlaat, and K. Vellenga. 1990. Developments inbaker'syeast production, p. 103-146. In H.Verachtert and R. De Mot(ed.),Yeast biotech-nology and biocatalysis. Marcel Dekker, Inc., NewYork. 2. Brown, C.M., and J. S. Hough. 1965. Elongation ofyeastcells

in continuousculture.Nature (London)206:676-678.

3. Cason, D. T., I. Spencer-Martins, and N. Van Uden. 1986. Transport of fructosebyaproton symport in abrewingyeast. FEMS Microbiol. Lett. 36:307-309.

4. Dikhuizen,L., L.Groen,W.Harder, and W. N. Konings. 1977. Activetransport of oxalate byPseudomonas oxalaticus OX1.-Arch. Microbiol. 115:223-227.

5. DUkhuizen, L., M. Wiersma, and W. Harder. 1977. Energy production and growth of Pseudomonas oxalaticus OX1 on

oxalateand formate.Arch.Microbiol. 115:229-236.

6. Egli, T., U. Lenden, and M. Snozzi. 1993. Microbialgrowth kinetics duringsimultaneous utilization of mixtures of carbon substrates, p. 118.SixthInternationalSymposiumonMicrobial Ecology, Barcelona, 1992.

7. Haggstrom, M. H., and C. L. Cooney. 1984. a-Glucosidase synthesis in batch and continuous culture ofSaccharomyces

cerevisiae. Appl. Biochem. Biotechnol. 9:475-481.

8. Harder,W., and L. Dikhuizen. 1976. Mixed substrate utiliza-tion, p. 297-314. In A. C. R. Dean, D. C. Ellwood, C. G. T. Evans,and J. Melling (ed.), Continuous culture 6: applications and newfields.Ellis Horwood, Chichester, England.

9. Harris, G., and C. C. Thompson. 1960. Uptake of nutrients by yeasts. II. Maltotriose permease and the utilization of maltotri-oseby yeasts. J. Inst. Brew. 66:293-297.

10. Kooijman, S. A. L. M., E. B. Muller, and A. H. Stouthamer. 1991. Microbial growth dynamics on the basis of individual budgets. Antonie van Leeuwenhoek Int. J. Gen. Mol. Biol. 60:159-174.

11. Malpartida, F., and R. Serrano. 1981. Proton translocation catalyzed by the purified yeast plasma membrane ATPase reconstituted in liposomes. FEBS Lett. 131:351-354.

12. Nelson, N., and L. Taiz. 1989. The evolution ofH'-ATPases. Trends Biochem. Sci. 14:113-116.

13. Noorman, H. J., J. Baksteen, J. J. Heijnen, and K. C. A. M. Luyben. 1991. The bioreactor overflow device: an undesired selectiveseparator in continuouscultures? J. Gen. Microbiol. 137:2171-2177.

14. Perlin, D.S., M. J. D. San Francisco, C. W. Slayman, and B. P. Rosen. 1986. H+/ATP stoichiometry of proton pumps from Neurospora crassa and Escherichiacoli.Arch. Biochem. Bio-phys. 248:53-61.

15. Postma, E., W. A. Scheffers, and J. P. van Diken. 1988. Adaptation of the kineticsofglucosetransporttoenvironmental conditions in the yeast Candida utilis CBS 621: a continuous culturestudy.J. Gen. Microbiol. 134:1109-1116.

16. Postma, E., W. A.Scheffers, and J. P.van

DQken.

1989.Kinetics ofgrowth and glucose transportin glucose-limited chemostat cultures ofSaccharomyces cerevisiaeCBS8066. Yeast 5:159-165.

17. Postma, E., and P. J. A. van den BroeL 1990. Continuous-culturestudyof theregulationofglucoseandfructose transport inKluyveromycesmarxianusCBS6556. J.Bacteriol. 172:2871-2876.

18. Postma,E., C. Verduyn, W. A.Scheffers,andJ.P.vanDiken. 1989.Enzymic analysisof the Crabtree effect inglucose-limited chemostat cultures ofSaccharomycescerevisiae. Appl. Envi-ron.Microbiol. 55:468-477.

19. Postma,E.,C.Verduyn,W. A.Scheffers,andJ.P.vanDiken. 1990.Substrate-accelerateddeath ofSaccharomycescerevisiae CBS 8066 under maltosestress.Yeast6:149-158.

20. Serrano,R. 1977.Energyrequirements formaltose transport in yeast. Eur. J. Biochem.80:97-102.

21. Srere, P. A. 1969.Citratesynthase. MethodsEnzymol.13:3-11. 22. vanLeeuwen, C. C.M.,R.A.Weusthuis,E.Postma,P.J.A.van

den Broek, and J. P. van Diken. 1992. Maltose/proton

co-transport inSaccharomycescerevisiae.Comparative studywith cellsandplasma membrane vesicles.Biochem. J. 284:441-445. 23. Verduyn, C. 1991. Physiologyof yeasts inrelationtobiomass yields. Antonie van Leeuwenhoek Int. J. Gen. Mol. Biol. 60:325-353.

24. Verduyn,C.,E.Postma,W. A.Scheffers,andJ.P.vanDiken. 1990. Physiology of Saccharomyces cerevisiae in anaerobic glucose-limitedchemostatcultures. J. Gen. Microbiol. 136:395-403.

25. Verduyn,C.,E.Postma,W. A.Scheffers,andJ.P.vanDiken. 1990. Energetics of Saccharomyces cerevisiae in anaerobic glucose-limited chemostatcultures. J.Gen. Microbiol. 136:405-414.

26. Yamamoto,Y., andT.Inoue. 1961. Studies of the poor

attenu-ative yeasts. I. Permeation of maltotriosethroughthe cell wall of poor attenuative yeast.Rep.Res.Lab. Kirin Brew. Co.Ltd. 1961:49-53.