Selection and subsequent physiological characterization of industrial Saccharomyces cerevisiae strains during continuous growth at sub- and- supra optimal temperatures

Delft University of Technology

Selection and subsequent physiological characterization of industrial Saccharomyces

cerevisiae strains during continuous growth at sub- and- supra optimal temperatures

Lip, Ka Ying Florence; García-Ríos, Estéfani; Costa, Carlos E.; Guillamón, José Manuel; Domingues,

Lucília; Teixeira, José; van Gulik, Walter M.

DOI

10.1016/j.btre.2020.e00462

Publication date

2020

Document Version

Final published version

Published in

Biotechnology Reports

Citation (APA)

Lip, K. Y. F., García-Ríos, E., Costa, C. E., Guillamón, J. M., Domingues, L., Teixeira, J., & van Gulik, W. M.

(2020). Selection and subsequent physiological characterization of industrial Saccharomyces cerevisiae

strains during continuous growth at sub- and- supra optimal temperatures. Biotechnology Reports, 26,

[e00462]. https://doi.org/10.1016/j.btre.2020.e00462

Important note

To cite this publication, please use the final published version (if applicable).

Please check the document version above.

Copyright

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy

Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.

This work is downloaded from Delft University of Technology.

Selection

and

subsequent

physiological

characterization

of

industrial

Saccharomyces

cerevisiae

strains

during

continuous

growth

at

sub-

and-supra

optimal

temperatures

Ka

Ying

Florence

Lip

a,

*

,

Estéfani

García-Ríos

b

,

Carlos

E.

Costa

c

,

José

Manuel

Guillamón

b

,

Lucília

Domingues

c

,

José

Teixeira

c

,

Walter

M.

van

Gulik

a,

*

a

DepartmentofBiotechnology,DelftUniversityofTechnology,Delft2629HZ,theNetherlands b

FoodBiotechnologyDepartment,InstitutodeAgroquímicayTecnologíadeAlimentos(IATA),ConsejoSuperiordeInvestigacionesCientíficas(CSIC),Valencia, Spain

c

CentreofBiologicalEngineering,UniversityofMinho,Braga4710-057,Portugal

ARTICLE INFO Articlehistory:

Received16January2020

Receivedinrevisedform22April2020 Accepted22April2020 Keywords: Chemostat Energeticefficiency Temperaturetolerance Saccharomyces SBR ABSTRACT

Aphenotypicscreeningof12industrialyeaststrainsandthewell-studiedlaboratorystrain CEN.PK113-7Datcultivationtemperaturesbetween12
Cand40
Crevealedsignificantdifferencesinmaximum growthratesandtemperaturetolerance.Fromthose12,twostrains,oneperformingbestat12
_Candthe

otherat40
C,plusthelaboratorystrain,wereselectedforfurtherphysiologicalcharacterizationin well-controlledbioreactors.Thestrainsweregrowninanaerobicchemostats,atafixedspecificgrowthrateof 0.03h1andsequentialbatchculturesat12
_{C,30
C,and39
C.Weobservedsignificantdifferencesin}

biomassandethanolyieldsonglucose,biomassproteinandstoragecarbohydratecontents,andbiomass yieldsonATPbetweenstrainsandcultivationtemperatures.Increasedtemperaturetolerancecoincided withhigherenergeticefficiencyofcellgrowth,indicatingthattemperatureintoleranceisaresultof energy wastingprocesses,suchasincreasedturnoverofcellularcomponents(e.g. proteins)dueto temperatureinduceddamage.

©2020PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http:// creativecommons.org/licenses/by-nc-nd/4.0/).

1.Introduction

Thealcoholicbeverageandbio-ethanolindustriesmainlyuse Saccharomyces yeasts as their workhorses, because of their robustness tolow pHand highethanol tolerance. The ethanol yieldandproductivityoffermentationprocesseshighlydependon theperformance of theyeast strains used at thetemperatures appliedintheseprocesses.Largedifferencesinperformanceand adaptation to working temperatures exist between individual yeaststrains[1].

Temperatureisoneofthepredominantfactorsdeterminingthe operationalcostsofindustrialfermentationprocesses.According toan energy studyof theEuropeanCommission, thealcoholic beverageandbio-energyindustriesspendaround30–60%oftheir total energy requirement of the whole production process to controlthecultivationtemperature[2].Ingeneral,theoptimum growthtemperatureofSaccharomycesyeastsliesbetween28
C–

33
C [3].However,this temperaturerangeis notapplicablefor boththealcoholicbeverageandbio-ethanolproductionprocesses in industry. In particular, cider, beer, white and rosé wine fermentation processes are commonly operated atsub optimal temperatures [4,5], range from 10
C to 25
C, to enhance and retaintheirflavorvolatiles[6].Theselowworkingtemperatures resultinprolongedfermentationprocessdurationandcausehigh risk of halted or sluggish fermentation [7]. Conversely, biofuel production processesare preferablyperformedat temperatures 40
_{Cespecially forfermentation processeswithsimultaneous}

saccharificationoflignocellulosicfeedstocks[8,9].Therefore,the adaptationofyeaststrainstotemperaturesoutsidetheoptimum rangeforgrowthprovidesanopportunitytomaketheproduction processmoreeconomicalandeco-efficient.

Temperaturetoleranceisapolygenictraitwhichisinfluenced byagroupofnon-epistaticgenes[10,11].Severalstudieshavebeen performed to increase the understanding of the impact of the cultivationtemperatureonthephysiologyofSaccharomycesyeasts andtoelucidatethemechanismswhichcontributetodifferences in temperature tolerance [5,12–23]. In the majority of these studies,temperatureshockswereappliedratherthanprolonged temperature stress,while the latter is much more relevant for

* Correspondingauthors.

E-mailaddresses:K.Y.F.Lip@tudelft.nl(K.Y.F.Lip),w.m.vangulik@tudelft.nl

(W.M. vanGulik).

https://doi.org/10.1016/j.btre.2020.e00462

2215-017X/©2020PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

ContentslistsavailableatScienceDirect

Biotechnology

Reports

industrial processes. To understand the cellular response and adaptationtotemperature, thechosenresearchmethodologyis crucialtoseparatetransientstressresponsesandadaptation.This work aims at addressing the long term impacts of different cultivation temperatures on Saccharomyces strains with better growthperformanceatsub-andsupra-optimaltemperatures.We firstcharacterizedacollectionofindustrialSaccharomycesstrains interms of theircapabilities togrowat sub-and-supraoptimal temperaturesrangingfrom12
_{Cto40
C.Thisallowedustoselect}

onestrainwhichperformedbestatsub-optimalandanotherstrain which performed best at supra-optimal temperatures. Subse-quently,thephysiologicalresponsesofthesestrains,togetherwith a well-characterized laboratory strain, CEN.PK113-7D, to sub-optimal,optimal and supra-optimal temperatures were investi-gatedinwell-definedchemostatculturesataconstantgrowthrate. 2.Methodsandmaterials

2.1.Yeaststrains,growthconditions,andstorage

Atotalof13Saccharomycesstrainsusedinthisstudyofwhich oneS.uvarum,oneS.cerevisiaeS.cerevisiaehybridandtheothers were S. cerevisiae species (Table 9). Inocula were prepared by introducinga singlecolonyofapurecultureofeach straininto 5mLsterilizedsyntheticmedium[24]with15gL1C6H12O6



H2O in a 30
C incubator shaker at 220rpm. Biomass stocks were preparedbytheadditionofsterilizedglyceroltotheexponentially growingculturesofall13strains,resultingthefinalconcentration of30%(v/v).Thebiomassstockswerestoredasepticallyat–80
_C.

Thesefrozenstockswereusedtoinoculatethedifferent experi-mentsdescribedasbelow.

Thegrowthprofilecanbeobtainedbygrowingtheyeaststrains onamicrotiterplateatdifferenttemperaturesrangingfrom12
C to40
Cunderaerobiccondition.Pre-culturewasgrownat30
C and 220rpmin sterilized syntheticmedium [24] with7.5gL1 C6H12O6



H2O. the pre-culturewas transferredto themicrotiter plate(24wells)withfreshsyntheticmediumresultedinaninitial opticaldensityat600nmofapproximately0.1andgrownatthe temperaturerangedfrom12
Cto40
Cwithcontinuousshaking (300rpm, 1-inch amplitude). Growth was monitored via the opticaldensityat600nm ina SynergyHTXMulti-Modereader (BioTek,USA),andmeasurementwastakenevery15minfor18h. Forthecultivationattemperaturebelow30
C,microtiterplates were cultivated in an incubator (New Brunswick Innova 44, Eppendorf)withcontinuousshaking(300rpm,1-inchamplitude) andmeasuredtheOD600bytheSynergyHTXMulti-Modereader every8hfor4days.Thegrowthprofileofeachstrainatdifferent temperatureswasobtainedbytriplicatemeasurementsandwasfit totwomodels.

2.2.Primarymodel

Themaximumspecificgrowthratesofeachstrainatdifferent temperatureswereobtainedby_fittingtheexperimentalOD600to thecorrectedmodifiedGompertzmodelequationmodifiedfrom theoriginalversion[25].

ln ODt OD0  

¼aexp  exp

m

max

expð1Þ a   ð

l

tÞþ1    

aexp  exp

m

max

expð1Þ a   ð

l

Þþ1    

WhereOD0istheinitialOD600andODtisthatattimet;aisthe asymptoticmaximumof lnðODt

OD0Þ;

m

maxis themaximumspecific

growthratewithaunitofh1,and

l

isthelagphaseperiod.Allthe parametersoftimehaveaunitofhour.

2.3.Secondarymodel

TheCTMImodelwasusedtofitwiththeobtained

m

maxofeach strain at different temperatures. The CTMI has the following expression;

m

¼0 ; if TTmin or TTmax  

m

¼

m

opt

D E  

; if Tmin <T<Tmax 

D¼ðTTmax ÞðTTmin Þ2

E¼ðTopt Tmin Þ½Topt Tmin 
T Topt 

 Topt Tmax 
Topt þTmin 2T

WhereTmaxisthetemperatureabovewhichnogrowthoccurs,Tmin isthetemperaturebelowwhichnogrowthisobserved,andToptis thetemperatureatwhich

m

maxisequalto

m

opt.Boththeprimary andsecondarymodelswere_fittedbyminimizingtheresidualsum ofsquares(RRS)withrespecttotheexperimentaldata.

2.4.Fermentationset-up

Allpre-culturesweregrownaerobicallyat220rpmandat30
C in the sterilized medium containing 5gL1 (NH4)2SO4, 3g



L1 KH2PO4,0.5gL1MgSO4



7H2O,22gL1C6H12O6



H2O,1.0mLL1of traceelementsolution,and1.0mLL1vitaminsolution[24].The sterilization of the medium was performed using a 0.2

m

m Sartopore2filterunit(SartoriusStedim,Goettingen,Germany).

Allbioreactors(describedindetailbelow)wereequippedwith norprene tubing, to minimizethe diffusion of oxygen into the vesselsandweresterilizedbyautoclavingat121
_C.Theexhaust

gasfromallfermentationswaspassedthroughacondenserkeptat 4.0
CandthenthroughaPermaPureDryer(InacomInstruments, Overberg, The Netherlands) to remove all water vapor and subsequently entered a Rosemount NGA 2000 gas analyzer (Minnesota, USA) for measurement of the CO2 concentration. Themediumofallfermentationswascontinuouslyspargedwith nitrogen gas prior to inoculation and contained 5.0gL1 (NH4)2SO4, 3.0gL1 KH2PO4, 0.5gL1 MgSO4



7H2O, 22.0gL1 C6H12O6



H2O,0.4gL1Tween80,10mgL1ergosterol,0.26gL1 antifoam C (Sigma-Aldrich, Missouri, USA), 1.0mLL1 trace element solution, and 1.0mLL1 vitamin solution [24]. The cultivations were carried out at temperatures of either 12.00.1
_{C,30.00.1
Cor39.00.1
C,bypumpingcooledor}

heatedwaterthroughthestainless-steeljacketsurrounding the bottompartofthereactorvesselusingacryothermostat(Lauda RE630,Lauda-Königshofen,Germany).Thewatertemperatureof thecryothermostatwascontrolledbyusingthesignalofaPt100 temperaturesensorinsidethereactor,foraccuratemeasurement andcontrolofthecultivationtemperature.Anaerobicconditions were maintained by continuously gassing of the reactor with nitrogengasataflowrateof1.000.01SLM(standardliterper minute)usingamassflowcontroller(Brooks,Hatfield,USA).Also, thefeedmediumwaskeptanaerobicbysparging withnitrogen gas.Thenitrogengaswassterilizedbypassingthrough hydropho-bic plate filters with a pore size of 0.2

m

m (Millex, Millipore, Billerica,USA).Theculturebrothinthereactorwasmixedusing one 6-bladed Rushton turbine(diameter 80mm) operatedat a

rotationspeedof450rpm.ThepHwascontrolledat5.000.05by automatictitrationwith2.0MKOH.

2.5.Sequentialbatchcultivations

Theprogressionofallbatchfermentationswasmonitoredby theCO2measurementintheexhaustgasfromtheoff-gasanalyzer andthebaseadditionintotheculturebroth.Thebasebottlewas placedonaloadcell(MettlerToledo,Tiel,TheNetherlands),and thustheamountofbaseadditionwas measuredbytheweight decreasedofthebasebottle.Whentherewasnobaseadditionfora de_fined amount of time, the culture broth was automatically drainedfromthebottomofthebioreactorbytheeffluentpump untiltherewas0.20kgleft.Freshmediumwassubsequentlyadded tofillthereactoruntilthetotalvolumewas4.0kg.Thevolumeof thebrothinthebioreactorwascontinuouslymonitoredviaaload cell (Mettler Toledo, Tiel, The Netherlands) which was placed underneaththebioreactor.Sixsequentialbatcheswerecarriedout intheabove-mentionedcycleateachtemperaturesetpoint(12
C, 30
C, and39
C). Thetemperaturesetpoint was changed after every6sequentialbatches.Themaximumspecificgrowthrateof each strainat each cultivation temperaturewas calculated and averagedfromtheCO2off-gasprofilesofthelastthreesequential batches.

2.6.Chemostatfermentation

Allchemostatcultivationswerecarriedoutatadilutionrateof 0.0300.002h1 in 7L bioreactors (Applikon, Delft, The Netherlands)equipped with a DCU3 control system and MFCS dataacquisitionandcontrolsoftware(SartoriusStedimBiotech, Goettingen,Germany).

Allchemostatfermentationswereinitiallyoperatedas anaero-bic batchcultivation with 4L culture broth in a 7L bioreactor (Applikon,Delft,TheNetherlands)toachieveenoughbiomassat thestartofthechemostatphase.FourhundredmLofpre-culture wasusedtoinoculateeachbatchcultivation.Whentheoff-gasCO2 levelfromthebatchcultivationdroppedclosetothelevelafterthe pre-cultureinoculation,thisindicatedtheendofthebatchphase. Thefermentationwasswitchedtochemostatphasebyswitching onthecontinuousfeedofsterilemediumtothebioreactor,which waspumpedintothereactorvesselataconstantflowrateusinga peristaltic pump (Masterflex, Barrington, USA), such that the outflowrateof theculturebrothwas1201gh1.Theeffluent vessel was placed on a load cell of which the signal was continuouslylogged for accurate determination of the dilution rateofthechemostatandmanualadjustmentofthemediumfeed rate if needed. The working volume was kept constant at 4.000.05kgusingtheloadcellofthebioreactor(MettlerToledo, Tiel, The Netherlands)which controlled theeffluent pump. All chemostatculturesreachedasteady-stateafter5volumechanges, whichwasapparentfromstableCO2levelsintheexhaustgasand biomassdry weightconcentrations.Afterreaching steady-state, triplicatesamplesatfoursamplingtimepointswerewithdrawn duringanotherperiodof4–5volumechanges,forquantificationof theconcentrationsofbiomass,residualglucoseandextracellular metabolites.

2.7.Analyticalmethods

Opticaldensity wasmonitored using a LibraSu spectropho-tometer (Biochrom Libra, Cambridge, UK) at a wavelength of 600nm.Biomass dry weightwasdeterminedusingfiltrationof sampledbrothoveradrynitrocellulosefilter(0.45

m

mporesize, Gelmanlaboratory,AnnArbor,USA)whichwasdried ina70
C oven overnight. After the filtration, two sample volumes of

demineralized water was used to wash thefilters which were subsequentlydriedinanovenat70
Cfortwodays.Priorandafter samplefiltrationthefilterswereweighedaftercoolingdownina desiccatorfor2h.Culturesupernatantwasobtainedusingthecold stainless-steelbeadsmethod[26].Theresultingsupernatantwas immediatelyfrozenbyliquidnitrogenandfollowedbythestorage at–80
_{C.Thesupernatantwasdefrostedandanalyzedinduplicate}

usinghigh-performanceliquidchromatography(HPLC)witha Bio-Rad Aminex column (Bio-Rad Laboratories, California, USA) at 60
C.Thecolumnwaselutedwith5.0mMphosphoricacidata flowrateof0.6mLmin1.Ethanolandglycerolweredetectedwith a Waters 2414 refractive index detector (Waters Corporation, Massachusetts,USA),whileaWaters1489UV–visdetector(Waters Corporation, Massachusetts, USA) was used to detect acetate, lactate,malate,andsuccinate.Residualglucosewasmeasuredby ion chromatography using Dionex-ICS 5000+ (Thermo Fisher Scientific,Massachusettts,USA).

2.8.Metabolicfluxanalysisanddatareconciliation

Themetabolicfluxdistributionsaswellasthebestestimatesof the biomass specific net conversion rates of the chemostat experiments wereobtained via metabolicflux analysis usinga stoichiometricmodel for anaerobicgrowth of S. cerevisiae[21]. With sufficient available conversionrates as inputvariables an overdetermined system was obtained, from which we could calculatethebestestimatesofthebiomassspecificnetconversion rates within theirerror margins as well as the metabolic flux distributions under the constraint that the elemental and compoundbalancesweresatisfied[27,28].

2.9.Totalorganiccarbonandtotalnitrogenmeasurement

Thetotalorganiccarbon(TOC)ofthechemostattotalbrothand supernatantwascalculatedfromthedifferencebetweenthetotal carbon(TC)andthetotalinorganiccarbon(TIC)whichwereboth measured by a total organic carbon analyzer (TOC-L CSH, Shimadzu,Kyoto,Japan).

The total nitrogen (TN) of the freeze-dried biomass (10mL culturebroth)fromthechemostatculturewasmeasuredbyatotal nitrogenunit(TNM-L,Shimadzu,Kyoto,Japan).TheTNcontentsof asamplewereintheformofammonium,nitrite,nitrate,aswellas organiccompounds.

TheinjectionofthesamplesforbothTOCandTNmeasurement was carried out by an auto-sampler (ASI-L, Shimadzu, Kyoto, Japan).

2.10.Cellularproteinmeasurement

Thirtymillilitreschemostatculturalbrothwaswithdrawnfrom thereactorandsubsequentlycentrifugedat4
_{Candat5000rpm}

for5min.Thesupernatantwasdiscarded,andthebiomasspellet was immediatelyfrozenin liquidnitrogenandstored in–80
_C

prior to freeze-drying. The cellular protein of the freeze-dried biomasswasdeterminedusingtheBiuretmethodasdescribedin [29]usingfreeze-driedBSAwasusedasstandard.

2.11.Cellularglycogenmeasurement

Approximately 2mg of biomass was quenched into 100% methanol, which was chilledto -40
C, using a rapid sampling setup[30].Thevolumeratioofsampletomethanolwas1–6.The quenchingsamplein100%methanolwascentrifugedat19
_Cand

at 5000rpmfor5min.Thesupernatant wasdiscarded, andthe biomass pellet was immediately frozen in liquid nitrogen and storedin–80
_C.

Thebiomasspelletwaswashedtwicewith1.5mLcoldMi-Q waterinanEppendorftubeandfollowedbycentrifugationat4
C andat8000rpmfor2min.Thesupernatant wasdiscarded.The pelletwas resuspendedin 250

m

L of 0.25M sodiumcarbonate solution and subsequently incubated at 95
C for 3h with continuousshaking.Aftertheincubation,600

m

Lof0.2Msodium acetatewasaddedtothemixture,thepHofwhichwasadjustedto 5.3 with 1M acetate acid afterwards. The hydrolysis of the glycogenwasperformedbyaddingα-amyloglucosidasedissolved in 0.2M sodium acetate to the mixture to have the final concentration of 1.2U/mL. The reaction of the hydrolysis was carried out at 57
_{C overnight with continuously shaking. The}

equivalent glucose released from the glycogen digestion was determinedintriplicateusingtheUVbioanalysiskit(R-Biopharm/ Roche,Darmstadt,Germany).Theresultingassaywasmeasuredat thewavelength340nmbyaspectrophotometer.

2.12.Cellulartrehalosemeasurement

Approximately 2mg of biomass was quenched into 100% methanol,which was chilled at -40
C, using a rapid sampling setup[30].Samplefiltratewaswashedtwicewith20mL80%(v/v) methanol,whichwas chilledat–40
_{C,and was extractedwith}

boiling ethanol as described in [30]. Cellular trehalose of the chemostatculturewasmeasuredbyGC–MSanalysisasdescribed in [31]. C13 _{labelled cell extract was added into the extracted} sampleasinternalstandard[32].

3.Results

3.1.GrowthphenotypiccomparisonofindustrialSaccharomyces strainsatdifferenttemperatures(12
C–40
_C)

The growth capacities of 13 Saccharomyces yeasts were determinedattemperaturesbetween12
C and40
Cinaerobic microtiterplatecultivations.Forallcultivationsthepurityofthe strainswasverifiedthroughanalysisofDNAdeltasequencesafter PCRamplificationasdescribedpreviously[33](datanotshown). Viathisgrowthphenotypicscreening,aninventorywasmadeof thetolerance of thesestrainsto thesub-optimal temperatures usedin thealcoholic beverage industry and the supra-optimal temperaturesusedinthebio-fuelproductionindustry.Duetothe Crabtree effect [34], all strains showed diauxic growth in the presenceofoxygen.Onlytheinitialpartsof thegrowthcurves,

representing growth onglucose with concurrentproduction of ethanol,wereused.Theobtainedgrowthprofilesofthestrainsat thedifferenttemperatureswerefittedtothecorrectedmodified GompertzmodelasproposedbyZwieteringetal.[25]whichwas modified from Salvado et al. [3]. The fit of this model to the experimentaldatayieldedtwoparameters,a maximumspecific growthrate(

m

max)andalagtime(

l

).Theobtainedlagtimesofthe strains at the different cultivation temperatures were not correlatedwiththe

m

max(datanotshown).Thefitofthemodel to the data was satisfactory for all strains at all cultivation temperatureswithR-squaredvalues(R2_{)rangingfrom0.92to0.99.} Ahierarchicalclusteranalysis(HCL) oftherelationbetween

m

maxandcultivationtemperaturewasperformedusingEuclidean distance to further analyze the growth performance between differentstrainswithinthetemperaturerangefrom12
Cto40
C (Fig.1).Thecolorsintheheatmaprepresentthevaluesof

m

max. Basedonthegrowthperformanceofallstrains,theHCLseperated theappliedtemperaturerangesintoarangewithsluggishgrowth (blue)andwithfacilitatedgrowth(blackoryellow).At12
C,15
C, and40
C,thegrowthratesofallstrainswerebelowthemedian (0.25h1).Theclusterfacilitatedgrowthoccurredcouldbedivided intotwosubgroups(optimalandsub-or-supraoptimalgrowth). Theoptimalgrowthtemperaturesofallstrainswereobservedat 28
Cand33
C,atwhichtheheatmapatthisareamostlyshowsa yellowcolor.Thesub-and-supraoptimalgrowthoccurredat25
_C

andbelowand 37
C andabove,respectively,atwhich theheat map mostly shows a black color. Regarding the sluggish and facilitatedgrowthconditions,theHCLdividedallstrainsintotwo major groups (1 and 2). Strains in group 1 had poor growth performanceatallgrowthconditions,whereasstrainsingroup2 hadacomparativelybettergrowthperformance.

Toobtainrelationsforthe

m

maxasafunctionofthecultivation temperatureforallstrainstested,thegrowthratesobtainedforthe different strains at the different cultivation temperatures were usedtofitthecardinaltemperaturemodelwithinflectionpoint (CTMI)[3].ThegoodnessoffitoftheCTMImodeltothe

m

maxdata wassatisfactoryinallcases,withp-valuesbetween0.97and0.99. The CTMI fitsof

m

max vs growthtemperaturefor thedifferent strainsareshowninFig.2.ADY5clearlyshowedthefastestgrowth in the temperaturerange from 12
C to27
C. At temperatures between 27
C and 33
C, ADY2 had the highest

m

max, while between33
C and40
C EthanolRedgrewfasterthanallother strains.EthanolRedandADY5showedthebestthermo-and cryo-tolerance,respectively,andwereselectedforfurtherinvestigation

Fig.1.HeatmaprepresentingtheHCLanalysisusingEuclideandistancewiththeareaunderthecurveofthe13Saccharomycesyeasts.Thedatawasobtainedfromthe phenotypicscreeningexperimentinmicrotiterplates.Growthraterangewithsluggishgrowthisrepresentedinblue,whilefacilitatedgrowthisrepresentedinblackor yellow.Strainscategorizedinred(group1)hadslowergrowthrateatallgrowthconditions,whilethoseingreen(group2)hadfasterrateatallgrowthconditions(For interpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle).

oftheunderlyingmolecularandmetabolicmechanisms.ADY5isa S. cerevisiae hybrid strain which is particularly used for the productionofaromaticwhiteandroséwinesintheindustryatlow fermentationtemperaturesandlownitrogenlevels.EthanolRedis an industrial yeast strain with high ethanol tolerance and is commonly used for the production of industrial ethanol at fermentationtemperaturesupto40
_C.

TheCTMIfitalsoprovidedthecardinalgrowthparametersof eachstrain(Tmax,Topt,Tmin,and

m

opt).Theaveragevaluesofthe estimated parameters for all strains were obtainedfrom three independentexperimentsandaresummarizedinTable1.TheTmin ofallstrainsrangedfrom1
Cto8
C,whereastheTmaxofallstrains rangedfrom39
Cto41
C.Forall13strains,theoptimumgrowth temperatures, Topt, were in therange between 29
C and 35
C wherein the corresponding specific growth rates ranged from 0.30h1to0.46h1.

3.2.Physiologicalcharacterizationoftheselectedstrainsat sub-optimal,optimalandsupra-optimaltemperatures

3.2.1.Maximumspecificgrowthratedeterminationoftheselected strainsinanaerobicsequentialbatchescultures

Tovalidatethephenotypicscreeningresultsandthemaximum growthrate estimationsfromtheCTMIatthedifferent temper-aturesfortheselectedstrains(ADY5 andEthanol Red)and the referencestrain(CEN.PK113-7D),weperformedsequentialbatch reactor(SBR)cultivationsofthethreestrains.SBRinsteadofsingle batchcultureswerechosenbecausethesewereshowntogivea betterreproducibilityandmoreconsistentresults[35]astheeffect ofcarryoverfromtheinoculum,whichmightplayaroleduringthe first batch, vanishes after a few repetitive batch cultivations. Within ten repetitive batches, we did not observe adaptation/

evolution ofcell culturesbecausenoincrease of themaximum specific growth rate occurred. The carbon dioxide production profilesduringtheexponentialphasesof theseSBRcultivations wereusedtocalculatethe

m

maxoftheselectedstrainsatthethree differentcultivationtemperatures(Table2).EthanolRedgrewthe fastest at 39
C, while ADY5grew the fastest at 12
C, thereby confirming the results from the screening experiments in microtiter plates. Compared to these strains, CEN.PK113-7D showed thelowestmaximum specificgrowth rateatthesupra optimaltemperature,whichwasroughly50%lowerthanthatof thecoldtolerantstrainADY5.

3.2.2.Furtherphysiologicalcharacterizationofthestrainsin anaerobicchemostatcultures

To identifypossibleunderlyingmechanismsfor thesuperior growthperformancesofADY5andEthanolRedatrespectively sub-and-supraoptimaltemperatures,wecomparedtheirphysiologyat 12
_{C,30
C,and39
Cwiththewell-studiedlaboratorystrainCEN.}

PK113-7D under well-defined conditions at a constant specific growthrate. Tothis end thestrainsweregrownat thesethree temperatures in anaerobicsteady-state chemostatcultures ata dilutionrateof0.03h1.Thisdilutionratewasslightlybelowthe

m

maxoftheselectedstrainsaswellasCEN.PK113-7Dat12
Cunder anaerobic conditions (Table 2). During all steady-states the measured residual glucose concentration was below 0.50mmolL1,confirmingglucoselimitedconditions.Chemostat instead of batch cultivation was chosen as this allowed us to separate thetemperatureeffects fromtheeffectsofthespecific growthrate.Itiswellknownthatdifferencesingrowthrateresult inphysiologicalchangesinyeast,suchastranscriptlevels[6].Fully anaerobicinsteadofmicro-aerobicconditionswerechosentorule out effectsof differencesindissolved oxygenlevelsatdifferent cultivationtemperatures.Besides,alcoholicbeverageandbio-fuel productionismainlyperformedintheabsenceofoxygen[36,37].

3.2.3.Largedifferencesbetweenstrainsintheeffectoftemperatureon netconversionrates

The biomass specific conversion rates (qi, mmolgDW1h1) werecalculatedforallsteady-statechemostatcultivationsatthe threedifferenttemperatures.Thefirstorderevaporationconstants

Table1

EstimatedparameterswithstandarddeviationfromtheCTMIfitforallthe13 strains. Standard deviations for each parameter were obtained from three independentnon-linearfits.

Strains mopt(h1) Tmax(
C) Tmin(
C) Topt(
C)

ADY1 0.3430.008 39.970.01 7.780.45 31.450.02 ADY2 0.4640.011 41.580.56 7.360.79 31.860.25 ADY3 0.3810.009 41.020.41 7.290.09 29.610.39 ADY5 0.4090.002 40.610.02 4.060.39 29.170.65 ADY6 0.2970.002 40.120.02 4.140.00 29.380.09 ADY7 0.3750.004 39.980.01 7.340.90 29.470.51 ADY8 0.3670.008 40.010.00 3.930.00 31.940.59 ADY20 0.3300.006 39.970.01 4.790.84 30.430.12 ADY21 0.3020.010 40.000.00 2.110.75 35.000.17 ADY22 0.3700.027 39.980.02 1.790.93 32.500.13 ADH30 0.2360.004 40.000.00 4.000.10 31.770.64 CEN.PK113-7D 0.3680.018 41.210.81 3.080.92 30.030.85 EthanolRed 0.4670.004 41.480.05 1.450.09 35.260.18 Table2

Maximumspecificgrowthrates(h1)inanaerobicSBRcultivationsat12
C,30
C, and39
_{C.Standarderrorsforeachstrainandforeachgrowthtemperaturewere} obtainedfromthreeindependentbatches.

Strains 12
_{C 30
C 39
C}

ADY5 0.0590.001 0.4160.004 0.2360.003

CEN.PK113-7D 0.0510.001 0.3610.001 0.1210.009

EthanolRed 0.0450.001 0.3580.001 0.3920.002

for culturebroth and ethanol during chemostat cultivations at differenttemperatureswereexperimentallydetermined(TableS3, supplementary)andusedforapropercalculationofthespecific ethanolproductionrate.Simultaneousmetabolicfluxanalysisand datareconciliationwasapplied(seematerialsandmethods)and yieldedthebestestimationsoftheconversionrateswithintheir errormargins(Tables3and4).Thereconcileddataforthenine chemostatconditionsfittedwellwiththeexperimentaldataasthe p-valuesofthereconciledconversionrateswereallgreaterthan thesignificancelevelof0.05(datanotshown).

As expectedfromthe stoichiometryof ethanol fermentation fromglucose,theratios(molmol1)of theCO2production rate (qco2)andethanolproductionrate(qeth)wereclosetooneforeach strainand each temperature condition (Table 3). Althoughthe dilutionrate,andthusthespecificgrowthrate,ofallcultivations wasthesame,significantdifferencesintheobtainedqivalueswere observedfordifferentstrainsandatdifferentcultivation temper-atures.Asanexamplethehighest(absolute)valueofthespecific glucoseuptakerate(qs),obtainedfortheCEN.PK113-7Dcultivated atthesupra-optimaltemperature(39
C),wasmorethanafactorof twohigherthanthelowest(absolute)valuewhichwasobtained fortheEthanolRedstraincultivatedatthesub-optimal tempera-ture(12
C).Comparabledifferenceswereobservedfortheethanol andcarbondioxideproductionrates.Forallthestrains,theglucose consumptionaswell asethanol andCO2 productionrates were highest at the highest cultivation temperature. However, the differencesbetweentheindividualstrainswerelargethereofCEN. PK113-7Dshowedthehighestvalues.

Withrespecttothesub-optimaltemperature,theresponsesof thethreestrainswerealldifferent.ForCEN.PK113-7Dtheqs,qeth, andqCO2valueswereallsignificantlyhigherat12
Ccomparedto thecontroltemperature;ADY5showednosignificantdifferences while for Ethanol Red these specificconversion rates were all

significantlylowerat12
Ccomparedtothecontroltemperature (30
C).Theseresultsclearlyindicatelargedifferencesincellular energetics between the distinct strains cultivated at different temperatures.

Therewerealsodifferencesintheproductionratesofglycerol (Table3)andacidicby-products(Table4)betweentheindividual strains, but there appeared tobe noclear correlationwiththe cultivationtemperature.Glycerolproductionratesweresimilarfor thedifferentstrainsandtemperatures,exceptforCEN.PK113-7D, whichhadasignificantlyincreasedglycerolproductionrateatthe sub-optimal temperature, which was accompanied with an increasedacetateproductionrate.Nevertheless,allthreestrains producedverysmallamountsofacids,withproductionratesupto 0.09mmolgDW1h1(TableS1,supplementary).

3.2.4.Largedifferencesbetweenstrainsintheeffectoftemperatureon yields

Afurtheranalysisofthephysiologicaldifferencesbetweenthe strains grown at different temperatures was performed by comparing the yields of biomass and (by)products on glucose forthedifferentcultivations(Tables5and6).Also,here,significant differences between strains and cultivation temperatures were observed.Forallthreestrainsthebiomassyieldsonglucosewere significantlylowerat39
_{Ccomparedto30
Cwherein}

CEN.PK113-7Dhadthelowestbiomassyield.Likewise,at12
CCEN.PK113-7D hadthelowestbiomassyield.ForADY5thebiomassyieldswere thesameat12
_{C and30
C,whilefor EthanolRedthebiomass}

yieldwashighestat12
C.Theethanolyieldsonglucoseforthe three strains at the different cultivation temperatures varied between1.52(EthanolRed,12
C)and1.70(CEN.PK113-7D,39
C) mol ethanol per mol glucose, whereby each individual strain showed a slightly different ethanolyield on glucosein general (Table 5). Although Ethanol Red and ADY5 were designated,

Table3

Reconciledspecificconversionratesofthethreestrainswiththeirstandarderrorsduringanaerobicchemostatcultivationat12
C,30
C,and39
Catadilutionrateof 0.03h1.Thenomenclatureofs,eth,andglyrepresentsassubstrate(glucose),ethanol,andglycerol,respectively.

qs(mmolgDW1h1) qCO2(mmolgDW1h1) qeth(mmolgDW1h1) qgly(mmolgDW1h1) ADY5 12
_{C 1.8040.092 3.0550.183 2.8990.183 0.2980.008} 30
_{C 1.7610.067 2.9550.134 2.7940.134 0.3060.008} 39
C 2.1580.049 3.5580.096 3.4640.097 0.2700.012 CEN.PK113-7D 12
_{C 2.7490.100 4.560.201 4.3390.201 0.4280.011} 30
_{C 2.0780.052 3.5680.104 3.3990.105 0.2990.011} 39
_{C 3.4110.007 4.8050.014 4.6580.014 0.3090.005} EthanolRed 12
C 1.5940.092 2.5990.184 2.4180.184 0.3340.006 30
C 1.8370.048 3.0300.096 2.8550.096 0.3260.013 39
C 2.0690.069 3.4040.138 3.2420.138 0.3650.009 Table4

Reconciledspecificconversionratesofthethreestrainswiththeirstandarderrorsduringanaerobicchemostatcultivationat12
_{C,30
C,and39
Catadilutionrateof} 0.03h1.Thenomenclatureofmal,suc,ace,andlacrepresentsasmalate,succinate,acetate,andlactate,respectively.

qmal(mmolgDW1h1) qsuc(mmolgDW1h1) qace(mmolgDW1h1) qlac(mmolgDW1h1) ADY5 12
_{C 0.0060.000 0.0130.000 0.0000.000 0.0000.000} 30
_{C 0.0060.006 0.0130.000 0.0000.000 0.0000.000} 39
_{C 0.0140.014 0.0940.000 0.0210.004 0.0210.004} CEN.PK113-7D 12
C 0.0080.000 0.0080.000 0.0670.004 0.0300.004 30
C 0.0000.000 0.0000.000 0.0000.000 0.0800.000 39
_{C 0.0070.001 0.0420.003 0.0160.002 0.0290.002} EthanolRed 12
_{C 0.0000.000 0.0130.000 0.0180.001 0.0180.001} 30
_{C 0.0000.000 0.0170.000 0.0120.004 0.0120.004} 39
_{C 0.0400.040 0.0010.000 0.0250.001 0.0250.001}

respectively, as hosts for the production of bioethanol and alcoholic beverages, they both had a lower ethanol yield on glucosethanthelaboratorystrainCEN.PK113-7D,regardlessofthe cultivation temperature. CEN.PK113-7D showed the highest ethanolyieldonglucoseat39
_{C,which wasaccompaniedwith}

acorrespondinglowbiomassyieldonglucose,asalargerpartof the consumed glucose was converted to ethanol. As the fermentationofglucosetoethanolisdirectlycoupledtocellular energygenerationintheformofATP,thisclearlyindicatedthat CEN.PK113-7D was negatively affected by the supra-optimal temperaturewhichincreasedthecellularenergydemand. 3.2.5.Temperatureeffectonmetabolicefficiencyofbiomassformation

Metabolic flux analysis was performed for each individual strain at each cultivation temperature using a stoichiometric modelforanaerobicgrowthofS.cerevisiaeonglucose.Herebythe metabolicflux distributions werecalculated using thebiomass speci_ficconversionratesobtainedfromthesteady-statechemostat cultivationsasinput.Thisallowed ustocalculatetheenergetic efficiencyofgrowthofeachindividualstrainasafunctionofthe cultivationtemperature.Foreachcondition,wecalculatedthenet biomassspecificrateofcatabolicATPproductionbysummingup the hexokinase, glycerol-3-phosphase, phosphofructokinase, phosphoglyceratekinase,and pyruvatekinasefluxes.Theratios ofthebiomassspecificgrowthratesandnetbiomassspecificATP productionratesprovidedthebiomassyieldswithrespecttothe

producedATP(Yx/ATP)atthedifferentcultivationtemperaturesfor eachstrain(Table7).Inspiteofthefixeddilutionrate,andthus identical specific growth rates, we observed large differences betweentheindividualstrainsatthedifferentcultivation temper-atures.EthanolRedcultivatedat12
Cproducedthreetimesmore biomasspermoleofATPthanCEN.PK113-7Dcultivatedat39
C. Foreachcultivationtemperature,theYx/ATPvaluesforEthanolRed and ADY5 were significantly higher than for CEN.PK113-7D. Ethanol Red grew most efficiently at both 12
C and 39
C. A possiblereasonforthedifferencesingrowthefficienciesmightbe differencesinbiochemicalcomposition,e.g.proteincontentsofthe cells among distinct strains. Therefore, for all chemostat culti-vationsthecellularcontentsofproteinandstoragecarbohydrates werequantifiedforeachstrainandcultivationtemperature. 3.3.Cellularprotein,glycogenandtrehalosecontents

Duringchemostatcultivationat12
C,thetotalcellularprotein contentwasverysimilarforthethreestrains(Table8), withan average value of about 0.32gproteingDW1. At a cultivation temperatureof30
C,theproteincontentsofADY5andEthanol Redwereslightlyhigher,whiletherewasasignificantincreasefor CEN.PK113-7Dcomparedtothe12
Ccultivations.Remarkably,the totalproteincontentsatthesupra-optimal cultivation tempera-tureof39
Cweresignificantlylowerforallstrains.Theseresults

Table5

Yieldsofbiomassandmainproductsonglucoseofthethreestrainswiththeirstandarderrors,calculatedfromthespecificnetconversionratesshowninTable3.

Ybiomass(molmol1) YCO2(molmol1) Yethanol(molmol1) Yglycerol(molmol1)

ADY5 12
_{C 0.5950.017 1.1610.046 1.6070.065 0.1650.005} 30
_{C 0.6250.014 1.4220.042 1.5870.049 0.1740.004} 39
_{C 0.5140.010 1.6490.096 1.6050.029 0.1250.003} CEN.PK113-7D 12
C 0.3960.009 1.7330.050 1.6490.049 0.1620.005 30
_{C 0.4980.011 1.7170.033 1.6360.032 0.1440.003} 39
_{C 0.4040.001 1.7580.003 1.7040.003 0.1130.001} EthanolRed 12
_{C 0.6750.020 1.6300.074 1.5170.072 0.2100.006} 30
C 0.6010.013 1.6490.034 1.5540.033 0.1780.004 39
C 0.5280.011 1.6450.043 1.5670.042 0.1760.004 Table6

Yieldsofby-productsonglucoseforthethreestrainswiththeirstandarderrors,calculatedfromthespecificnetconversionratesshowninTable4.

Ymalate(molmol1) Ysucinnate(molmol1) Yacetate(molmol1) Ylactate(molmol1) ADY5 12
C 0.0030.000 0.0070.000 0.0000.000 0.0000.000 30
_{C 0.0040.000 0.0070.000 0.0000.000 0.0000.000} 39
_{C 0.0070.000 0.0440.000 0.0100.001 0.0210.000} CEN.PK113-7D 12
_{C 0.0030.000 0.0030.000 0.0250.001 0.0110.000} 30
_{C 0.0000.000 0.0000.000 0.0000.000 0.0380.001} 39
C 0.0030.000 0.0160.000 0.0060.000 0.0110.000 EthanolRed 12
_{C 0.0000.000 0.0080.000 0.0110.000 0.0070.000} 30
_{C 0.0000.000 0.0090.000 0.0060.001 0.0330.003} 39
_{C 0.0190.004 0.0000.000 0.0120.000 0.0320.001} Table7

Yieldsof biomass on ATP (gDWmolATP1) with their standard errorsfor the anaerobicchemostatcultures.

12
C 30
C 39
C

ADY5 11.790.23 12.750.18 9.170.27

CEN.PK113-7D 7.640.17 9.450.12 5.350.22

EthanolRed 15.240.30 12.270.13 10.690.16

Table8

Totalproteincontentsofbiomass(gproteingDW1)forthethreestrainswiththeir standarddeviationsduringanaerobicchemostatcultivationatdifferent tempera-tures.

12
_{C 30
C 39
C}

ADY5 0.3160.004 0.3340.008 0.2740.030

CEN.PK113-7D 0.3320.010 0.3760.007 0.2460.002

wereconfirmed by quantification of the total cellular nitrogen contents (Table S2, supplementary) which indeed showed the sametrends.

We observed large differences for the cellular contents of glycogenandtrehalosebetweendifferentstrainsandcultivation temperatures(Figs.3and4).Forallcultivationtemperatures,both Ethanol Red and ADY5 had higher glycogen and trehalose accumulationsthanCEN.PK113-7D.Theaccumulationsofglycogen andtrehaloseforallthreestrainsshowedoppositepatternswith respecttothecultivationtemperature.Allthreestrainshadhigher glycogenbut lower trehalose accumulations at 12
C, and vice versaat39
_{C.At12
C,thetrehalosecontentsofallthreestrains}

wereextremelylow(below1%). Weobservedsignificant differ-encesin glycogenaccumulation betweenvarious strainsatthe sub-optimaltemperaturewhereinADY5hadthehighestvalueof about20%. At39
C, thetrehalose accumulations of ADY5 and

EthanolRedwereparticularlyhighwithvaluesaround10%.The glycogenaccumulationsatthistemperaturewerealsosignificantly differentbetweenstrains.ADY5hadthehighestvalue,morethan 5%,whereasCEN.PK113-7Dhadthelowestvalue,below1%.

4.Discussion

Thegrowthperformanceofthethreeselectedstrainsat12
C, 30
Cand39
CintheanaerobicSBRculturesalignedwellwiththe description of the CTMI model (Table 1) derived from the microtiterplatedata.The

m

maxvaluesobtainedfromtheanaerobic SBR cultivations (Table 2) showed that ADY5 and Ethanol Red clearlyperformedbetterat12
_{Cand39
C,respectively,compared}

toCEN.PK113-7D.The

m

maxvaluesofADY5andCEN.PK113-7Dat 30
Cwereverysimilartotheestimatedoptimal

m

maxobtained fromtheCTMI.TheestimatedoptimumtemperatureofEthanol RedfromtheCTMIwasfivedegreeshigherthanthatoftheother twostrains,highlightingitstemperaturetolerance,andwasclose tothevaluereportedintheliterature[38].Therefore,theCTMI modelwasusefulandreliableinourstudytodescribethegrowth profileoftheselectedstrainsoverthetemperaturerange.

Furtherphysiologicalcharacterization of thethreestrainsin anaerobicglucoselimitedchemostatculturesatafixeddilution rate revealed very large differences in the biomass yields on glucose between the different strains, but also between the differentcultivationtemperaturesfor thesamestrain(Table 5). Generally,lowerbiomassyieldscorrelatedwithhigherethanoland CO2 yields, suggesting differences in energy requirements for growth and maintenance for the different strains and temper-atures.Alsotheformationof increasedamountsofby-products (glycerol and acids) will result in decreased biomass yields. However,thetotalyieldofby-productsonglucosewasverysimilar for all chemostat cultivations and was on average 0.1000.005molofcarbonproducedpermolofcarbonconsumed asglucose.Notably,CEN.PK113-7D,ofwhichthebiomassyieldsat the different temperatures were the lowest, also showed the lowest average by-product yields, indicating that by-product formation was not the cause of the low biomass yields. The observedlargedifferencesinbiomassyieldsmustthereforehave been caused by large differences in cellular energy demands, whichwerequantifiedbycalculatingthebiomassyieldsonATP (YATP) (Table 7). The YATP of anaerobically grown S. cerevisiae (CBS8066)hasbeendeterminedpreviouslyfromglucoselimited chemostatexperimentsatacultivationtemperatureof30
C[39] inwhichthemaximumvaluewas16gDWmol1ATP.

From retentostat experiments it was found that the ATP dissipationrate formaintenance(mATP)ofCEN.PK113-7D under anaerobic conditions equals 1mmolATPgDW1h1at 30
C [40].

Table9

Yeaststrainsusedinthisstudy.

Strains Commercialname Species Origin Source

ADY1 Lalvin1_{QA23 S.cerevisiae Portugal LallemandInc.,France}

ADY2 Lalvin1_{ICVGRE S.cerevisiae France LallemandInc.,France}

ADY3 Lalvin1_{T73 S.cerevisiae Spain LallemandInc.,France}

ADY5 CROSSEVOLUTION1 _{S.cerevisiaexS.cerevisiae SouthAfrica LallemandInc.,France}

ADY6 VellutoBMV58TM

S.uvarum Spain LallemandInc.,France

ADY7 Lalvin1ICVOKAY S.cerevisiae France LallemandInc.,France

ADY8 Lalvin1_{Rhône2056 S.cerevisiae France LallemandInc.,France}

ADY20 Uvaferm1_{WAM S.cerevisiae Spain LallemandInc.,France}

ADY21 Lalvin1_{Rhône2226 S.cerevisiae France LallemandInc.,France}

ADY22 Uvaferm1_{CEG S.cerevisiae Germany LallemandInc.,France}

ADH30 – S.cerevisiae Spain LallemandInc.,France

CEN.PK113-7D – S.cerevisiae Unknown FungalBiodiversityCentre,

Utrecht,TheNetherlands

EthanolRed EthanolRed S.cerevisiae Unknown Fermentis,S.I.Lesaffre

Note:(–)Labstrainornon-marketedstrain.

Fig.3.Glycogenaccumulationofthethreestrainsinanaerobicchemostatat12
_C, 30
_{C,and39
C.Errorbarsrepresentstandarddeviationsofaveragevaluesof} measurementsin biomasssamples fromidentical chemostatcultures atfour differenttimepointsinsteady-state.

Fig.4.Trehaloseaccumulationofthethreestrainsinanaerobicchemostatat12
C, 30
_{C,and39
C.Errorbarsrepresentstandarddeviationsofaveragevaluesof} measurementsinbiomasssamplesobtainedfromidenticalchemostatcultures collectedatfourdifferenttimepointsinsteady-state.

FromthesefiguresitcanbecalculatedthatYATPshouldbearound 14gDWmol1ATPata growthrateof0.1h1,whichwasindeed observedexperimentally[39]andduetoanincreasedcontribution ofmaintenanceenergyrequirements,around10.5gDWmol1ATP atthegrowthrateof0.03h1usedinourchemostatcultivations. TheYATPvaluesweobservedforthethreestrainsat30
C(between 9.5 and 12.8gDWmol1 ATP) are close to this value, whereby differencesinbiomasscomposition,especiallyproteincontentof whichthebiosynthesisisthemostenergydemanding,couldbe responsibleforthedifferencesinYATPvaluesbetweenthestrainsat the same cultivation temperature. Quantification of the total protein contents revealed, however, that there were minor differencesinproteincontents betweenthethreestrainsatthe sametemperature,thusrulingoutthattheobserveddifferencesin YATP betweenthestrains werecaused bydifferencesin protein content. The cultivation temperature itself had more effect, especiallyat39
Ctheproteincontentsofallthreestrainswere significantlylowerthanat30
Cand12
C.Thiscouldbeastrategy ofthecellstodecreasetheirATPexpensestocopewiththestress duringgrowthatsupra-optimaltemperatures.

It is well known that maintenance energy requirements of microorganismsincreasewithincreasingcultivationtemperature [41].Usingthecorrelationproposedbytheseauthors,anincrease ofthecultivationtemperaturefrom30
Cto39
Cwouldresultina 2.2 fold increase of the maintenance coefficient, which would resultina decreaseof YATP with30% to7.4gDWmol1ATPata growthrateof0.03h1.ThedecreaseofYATPoftheADY5strainat 39
Cisindeedverycloseto30%valuewhileforCEN.PK113-7Dthe decreaseismorethan40%.EthanolRed appearedclearlybetter adaptedtohighercultivationtemperaturesastheYATPdecreased withonly13%.Conversely,adecreaseofthecultivation tempera-turefrom30
C to 12
C would, accordingtothe correlationof Tijhuisetal.[41],resultinadecreaseofthemaintenanceenergy requirementswithmorethanafactorof5and,consequently,an increaseofYATPwithalmost40%atagrowthrateof0.03h1.Such anincreasewasnotobservedinourchemostatcultures,onthe contrary,fortwostrains(CEN.PK113-7DandADY5)YATPwaslower at12
_{Cthanat30
C.NeverthelessforEthanolRedY}

ATPwas24% higherat12
Ccomparedto30
C.

AnotherfactorwhichcanleadtodifferencesinYATPbetween differentstrainsand/orcultivationtemperaturesisdifferencesin theconcentrations of weakacids [38]. Passive diffusionof the undissociatedform intothecells and subsequent activeexport resultsinanATPdissipatingfutilecycleleadingtoincreased non-growthassociatedenergyrequirements.Oftheacidicbyproducts excreted,aceticacid(pKa=4.76)wouldhavethemostsignificant influenceonthemaintenanceenergyrequirementsbecauseatthe cultivationpHof5,37%oftheacidispresentintheundissociated form. The maximum residual acetic acid concentration of the chemostatcultureswas2.84mmolL1forCEN.PK113-7D cultivat-edat12
_{C(TableS1).Atthisresidualaceticacidconcentration,the}

maintenance energy requirements would be approximately 2.3mmolATPgDW1h1atpH5,and 30
C[39]and wouldresult inanYATPof7.2gDWmol1ATP,whichisclosetotheobserved valueof7.64(Table7).Foralltheotherstrainsandtemperatures theresidualaceticacidconcentrationswerearound1mMorlower andthustheeffectonthemaintenanceenergyrequirementswere assumedtobesmall.Interestingly,ADY5didnotproduceacetic acidat12
Cand30
C.Possibleuncouplingofaceticacidseemsnot toattributesignificantlytotheYATPofEthanolRedat12
Cwhere theresidualaceticacidconcentrationwastwotimeshigherthan thatat30
C.

Anothercauseof theeffectoftemperatureonYATP couldbe proteinmisfoldingathightemperaturesandaggregationatlow temperatures.Remarkablytheproteomeanalysisoftheidentical chemostatculturedstrainsrevealedthatforbothCEN.PK113-7D

andADY5theproteinsrelatedtoproteinfoldinganddegradation processeswereupregulatedat 12
C, incontrasttoEthanol Red [42].Thiscouldindicatethatproteinaggregationand/or misfold-ing and subsequent degradation and re-synthesis might have occurredinthesestrainsat12
C,resultinginanincreasedenergy demandandthusadecreasedYATP.

In Ethanol Red Erg13, one of the first and rate controlling enzymesintheergosterolbiosynthesispathwaywasupregulated atboth12
_{Cand39
C[42].Althoughergosterolwasoneofthe}

anaerobicgrowthfactorssupplementedtothechemostatmedium asitssynthesisrequiresoxygen,thisupregulationcouldindicate increased incorporation of ergosterol in the cell membrane of EthanolRed.Severalstudieshavereportedthattheactivationof theergosterolpathwaysmakesyeastcellsmoreresistant/tolerant to a variety of stresses, including low temperature, low-sugar conditions,oxidativestressandethanol[43–46].

Allthreestrainsshowedupregulationofproteinsinvolvedin transportandmetabolismofcarbohydratesaswellasenergyand aminoacidmetabolismat12
Ccomparedto30
C[42].Thisshows thatmaintainingthesamespecificgrowthrateof0.03h1inthe chemostat at 12
C, where maximum enzyme capacities have decreased, requiresupregulationof proteinsin central metabo-lism.

It is well-known that the accumulation of the storage carbohydrates glycogen and trehalose in S. cerevisiae strongly dependsonthegrowthrate[47]andthatinparticulartrehalose wasshowntoprotectcellsduringstressconditions[48–50].Asin thisworkallcultivationswerecarriedoutatafixeddilutionrate, differences in storage carbohydrate accumulation can only be attributed to the particular strain used and/or the cultivation temperature. During glucose limited chemostat cultivation all threestrainsaccumulatedbothtrehaloseandglycogen,whereby the differencesin total accumulations (glycogen and trehalose) betweenstrainsweremoresignificantthanbetweencultivation temperaturesforthesamestrain.ItiswellknownthatS.cerevisiae accumulatesthese carbohydratesat growth ratesbelow 0.1h1 wherebythecontentsarerelatedtothedurationoftheG1phase [51].Undercarbonlimitedconditionstrehaloseandglycogenserve as carbon and energy reserves to enable the survival during starvationbutarealsomobilizedtofacilitateatransientincreasein theATPfluxforprogressionthroughthecellcycle[52].Forallthree strainstheaccumulationsoftrehaloseandglycogenwerestrongly dependentonthecultivationtemperature,withhighestglycogen accumulationat12
Candhighesttrehaloseaccumulationat39
C. Increasedtrehaloseaccumulationathighcultivationtemperatures havebeenobservedbeforeandwerecaused byastimulationof trehalosesynthaseandinhibitionoftrehalose[53].Becauseofthe α-1,1-glycosidicbondlinkagewithinthestructure,trehalosehas stronger resistance to heat and acid and was shown to be a preferableenergyreserveoverglycogenduringstressconditions [48,54],althoughthesynthesisoftrehaloserequiresmoreATPper glucose than that of glycogen[47]. Besides, thedegradation of trehalosereleasestwoglucoses,whereasoneglucoseisreleased afterthedegradationofthe_{α-1,4-glycosidic}bondwithinglycogen. Apartfromitsroleasareservecarbohydrate,trehalosealsohasa protective functionduring stressconditions e.g. thermal stress, wherebyitactsastheprotectorofmembranesandproteins[55– 57].Therefore,thesignificantlyhighertrehaloseaccumulationof ADY5andEthanolRedmighthavecontributedtothebettergrowth performanceat39
CcomparedwithCEN.PK113-7D.

The superior growth performance of ADY5 during SBR cultivationat12
Ccoincidedwithahighcapacitytoaccumulate glycogen. Increased carbohydrate accumulation, in particular glycogen,as a responsetoprolonged exposure ofyeast tocold (10
C) has been observed before [58]. Furthermore, a positive correlationbetweencellwallboundglycogenandviabilityunder

glucose deprived conditions was reported [59]. Although the precisefunctionof thisglycogenpoolremainsunclear, itmight play a role in membrane stabilizationwhich may improvethe resistancetocold[60].

5.Conclusions

Fromagrowthphenotypicscreeningof12industrial Saccharo-mycesstrainsfortheirtemperaturetolerance,weselectedADY5, EthanolRed,andCEN.PK113-7Dtofurtherelucidatethepossible underlyingmechanismsfortemperaturetolerance.Thechemostat resultsrevealedsigni_ficantdifferencesinthemetabolicresponse and cellular energetics between strains and among different growthtemperatures.Despiteafixedgrowthrate,differentgrowth temperatures resulted in large differences between the three strains in terms of net conversion rates, substrate yields and energeticefficiency of biomassformation. Allstrains showeda decreaseofproteincontentatsupra-optimaltemperatureswhich was nevertheless accompanied with a decrease of YATP, thus implyinganincreaseofnon-growthassociatedenergydemands. Increasedtemperaturetolerancecoincidedwithhigherenergetic efficiencyofcellgrowth,indicatingthattemperatureintoleranceis aresultofenergywastingprocesses,suchasincreasedturnoverof cellular components due to temperature induced damage, e.g. proteinmisfolding. Further research is required todeepen our comprehensionontheunderlyingmechanisms.Withthis knowl-edge,wecandevelopandapplystrategiestoobtaintailored cryo-and-thermotolerantyeastsforindustrialapplications.

Authorcontributions

Allauthorshavebeeninvolvedintheconceptualizationandthe methodologyoftheexperiments.JMGprovidedinformationabout theindustrialstrains. CEC,EGR,andKYFLperformedthegrowth phenotypic screenings in microtiter plates. EGR and KYFL performed the computational simulations. EGR performed the HCLanalysis.KYFLperformedthemeasurementsoftotalproteins, total nitrogen, and storage carbohydrates of the chemostat cultures. KYFL and WVG performed the fermentations in bioreactors,analysisoftheresults, andwrotetheoriginaldraft ofthepaper.CECandLDmadethegraphicabstract.Allauthors reviewed,edited,andapprovedthefinalmanuscript.

DeclarationofCompetingInterest

The authors declare that they have no known competing financial interests or personal relationships that could have appearedtoaffecttheworkreportedinthispaper.

Acknowledgements

WewouldliketothankJudithCohenandKristenH.Davidfor technicalassistancewiththechemostatfermentationsandJoséMa

Heras(LallemandIbéria, SA)for kindlyproviding theindustrial strains.Thisresearchwascarried outwithintheERA-IB project “YeastTempTation” (ERA-IB-2-6/0001/2014) and partially sup-portedbythePortugueseFoundationforScienceandTechnology (FCT) through strategic funding UID/BIO/04469/2020 and Bio-TecNorte(NORTE-01-0145-FEDER-000004).

AppendixA.Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,in the online version, at doi:https://doi.org/10.1016/j.btre.2020. e00462.

References

[1]AmparoQuerol,G.H.Fleet,YeastsinFoodandBeverages,Springer,Berlin, 2006.

[2]Y.Chan,R. Kantamaneni, Study onEnergyEfficiency andEnergy Saving PotentialinIndustryandonPossiblePolicyMechanisms(ICFConsulting Limited)Availableat:,(2016). https://ec.europa.eu/energy/en/studies/study- energy-efficiency-and-energy-saving-potential-industry-and-possible-policy-mechanisms.

[3]Z.Salvadó,F.N.Arroyo-López,J.M.Guillamón,G.Salazar,A.Querol,E.Barrio, Temperatureadaptationmarkedlydeterminesevolutionwithinthegenus Saccharomyces,Appl.Environ.Microbiol.77(2011)2292–2302.

[4]R.E.Kunkee,Selectionandmodificationofyeastsandlacticacidbacteriafor winefermentation,FoodMicrobiol.1(1984)315–332.

[5]M.J.Torija,G.Beltran,M.Novo,M.Poblet,J.M.Guillamón,A.Mas,N.Rozès, EffectsoffermentationtemperatureandSaccharomycesspeciesonthecell fattyacidcompositionandpresenceofvolatilecompoundsinwine,Int.J.Food Microbiol.85(2003)127–136.

[6]G.Beltran,M.Novo,V.Leberre,S.Sokol,D.Labourdette,J.-M.Guillamon,A. Mas,J.François,N.Rozes,Integrationoftranscriptomicandmetabolicanalyses forunderstandingtheglobalresponsesoflow-temperaturewinemaking fermentations,FEMSYeastRes.6(2006)1167–1183.

[7]L.F.Bisson,Stuckandsluggishfermentations,Am.J.Enol.Vitic.50(1999)107– 119.

[8]M.Kelbert,A. Romaní, E.Coelho, L.Pereira, J.A. Teixeira,L. Domingues, Simultaneoussaccharificationandfermentationofhydrothermalpretreated lignocellulosicbiomass:evaluationofprocessperformanceundermultiple stressconditions,Bioenerg.Res.9(2016)750–762.

[9]U. Breuer, Yeast biotechnology: diversity and applications, in: T. Satyanarayana,GotthardKunze(Eds.),Biotechnol.J.,52010,pp.427–428. [10]Y. Yang, M.R. Foulquié-Moreno, L. Clement, É. Erdei, A. Tanghe, K.

Schaerlaekens,F.Dumortier,J.M.Thevelein,QTLanalysisofhigh

thermotolerancewithsuperioranddowngradedparentalyeaststrainsreveals newminorQTLsandconvergesonnovelcausativeallelesinvolvedinRNA processing,PLoSGenet.9(2013).

[11]Z.Wang,Q.Qi,Y.Lin,Y.Guo,Y.Liu,Q.Wang,QTLanalysisrevealsgenomic variantslinkedtohigh-temperaturefermentationperformanceinthe industrialyeast,Biotechnol.Biofuels12(2019)59.

[12]J.Verghese,J. Abrams,Y.Wang,K.A.Morano,Biologyof theheatshock responseandproteinchaperones:buddingyeast(Saccharomycescerevisiae) asamodelsystem,Microbiol.Mol.Biol.Rev.76(2012)115–158.

[13]V.A.Kalyuzhin,HeatresistanceinSaccharomycescerevisiaeyeast,Biol.Bull. Rev.1(2011)207–213.

[14]L.Castells-Roca,J.García-Martínez,J.Moreno,E.Herrero,G.Bellí,J.E. Pérez-Ortín,Heatshockresponseinyeastinvolveschangesinbothtranscription ratesandmRNAstabilities,PLoSOne6(2011)e17272.

[15]E.G.Rikhvanov,N.N.Varakina,T.M.Rusaleva,E.I.Rachenko,V.A.Kiseleva,V.K. Voinikov,Heatshock-inducedchangesintherespirationoftheyeast Saccharomycescerevisiae,Microbiology70(2001)462–465.

[16]D.A.Costa,C.J.A.deSouza,P.S.Costa,M.Q.R.B.Rodrigues,A.F.dosSantos,M.R. Lopes,H.L.A.Genier,W.B.Silveira,L.G.Fietto,Physiologicalcharacterizationof thermotolerantyeastforcellulosicethanolproduction,Appl.Microbiol. Biotechnol.98(2014)3829–3840.

[17]J.Postmus,A.B.Canelas,J.Bouwman,B.M.Bakker,W.Gulik,M.J.T.vanMattos, BrulS.de,G.J.Smits,Quantitativeanalysisofthehightemperature-induced glycolyticfluxincreaseinSaccharomycescerevisiaerevealsdominant metabolicregulation,J.Biol.Chem.283(2008)23524–23532.

[18]K.A.Morano,C.M.Grant,W.S.Moye-Rowley,Theresponsetoheatshockand oxidativestressinSaccharomycescerevisiae,Genetics190(2012)1157–1195. [19]L.Caspeta,J.Nielsen,Thermotolerantyeaststrainsadaptedbylaboratory evolutionshowtrade-offatancestraltemperaturesandpreadaptationtoother stresses,mBio6(2015)e00431-15.

[20]L.Caspeta,Y.Chen,J.Nielsen,Thermotolerantyeastsselectedbyadaptive evolutionexpressheatstressresponseat30
C,Sci.Rep.6(2016)27003. [21]S.L.Tai,P.Daran-Lapujade,M.A.H.Luttik,M.C.Walsh,J.A.Diderich,G.C.Krijger,

W.M.Gulik,J.T.vanPronk,J.-M.Daran,Controloftheglycolyticfluxin saccharomycescerevisiaegrownatlowtemperatureamulti-levelanalysisin anaerobicchemostatcultures,J.Biol.Chem.282(2007)10243–10251. [22]S.L.Tai,P.Daran-Lapujade,M.C.Walsh,J.T.Pronk,J.-M.Daran,Acclimationof

Saccharomycescerevisiaetolowtemperature:achemostat-based transcriptomeanalysis,Mol.Biol.Cell18(2007)5100–5112.

[23]E. García-Ríos,M. Morard, L. Parts, G. Liti, J.M. Guillamón, Thegenetic architectureoflow-temperatureadaptationinthewineyeastSaccharomyces cerevisiae,BMCGenomics18(2017)159.

[24]C. Verduyn, E. Postma, W.A. Scheffers, J.P. van Dijken, Physiology of Saccharomycescerevisiaeinanaerobicglucose-limitedchemostatculturesx, Microbiology136(1990)395–403.

[25]M.H.Zwietering,I.Jongenburger,F.M.Rombouts,K.van’tRiet,Modelingofthe bacterialgrowthcurve,Appl.Environ.Microbiol.56(1990)1875–1881. [26]M.R.Mashego,W.M.vanGulik,J.L.Vinke,J.J.Heijnen,Criticalevaluationof

samplingtechniquesforresidualglucosedeterminationincarbon-limited chemostatcultureofSaccharomycescerevisiae,Biotechnol.Bioeng.83(2003) 395–399.

[27]R.T.J.M.Heijden,J.J.vanderHeijnen,C.Hellinga,B.Romein,K.C.A.M.Luyben, Linearconstraintrelationsinbiochemicalreactionsystems:I.Classificationof

thecalculabilityandthebalanceabilityofconversionrates,Biotechnol.Bioeng. 43(1994)3–10.

[28]R.T.J.M.Heijden,B.vander,Romein,J.J.Heijnen,C.Hellinga,K.C.A.M.Luyben, Linearconstraintrelationsinbiochemicalreactionsystems:II.Diagnosisand estimationofgrosserrors,Biotechnol.Bioeng.43(1994)11–20.

[29]H.C. Lange, J.J. Heijnen, Statistical reconciliation of the elemental and molecularbiomasscompositionofSaccharomycescerevisiae,Biotechnol. Bioeng.75(2001)334–344.

[30]A.B.Canelas,A.tenPierick,C.Ras,R.M.Seifar,J.C.vanDam,W.M.vanGulik,J.J. Heijnen,Quantitativeevaluationofintracellularmetaboliteextraction techniquesforyeastmetabolomics,Anal.Chem.81(2009)7379–7389. [31] D.Visser,G.A.vanZuylen,J.C.vanDam,A.Oudshoorn,M.R.Eman,C.Ras,W.M.

vanGulik,J.Frank,G.W.K.vanDedem,J.J.Heijnen,Rapidsamplingforanalysis ofinvivokineticsusingtheBioScope:asystemforcontinuous-pulse experiments,Biotechnol.Bioeng.79(2002)674–681.

[32]L.Wu,M.R.Mashego,J.C.vanDam,A.M.Proell,J.L.Vinke,C.Ras,W.A.van Winden,W.M.vanGulik,J.J.Heijnen,Quantitativeanalysisofthemicrobial metabolomebyisotopedilutionmassspectrometryusinguniformly 13C-labeledcellextractsasinternalstandards,Anal.Biochem.336(2005)164–171. [33]A. Xufre, H. Albergaria, F. Gírio, I. Spencer-Martins, Use of interdelta polymorphismsofSaccharomycescerevisiaestrainstomonitorpopulation evolutionduringwinefermentation,J.Ind.Microbiol.Biotechnol.38(2010) 127–132.

[34]R.H.DeDeken, TheCrabtreeeffect:aregulatorysysteminyeast,J. Gen. Microbiol.44(1966)149–156.

[35]A.l.b.Cruz,A.j.Verbon,L.j.Geurink,P.j.t.Verheijen,J.j.Heijnen,W.M.vanGulik, Useofsequential-batchfermentationstocharacterizetheimpactofmild hypothermictemperaturesontheanaerobicstoichiometryandkineticsof Saccharomycescerevisiae,Biotechnol.Bioeng.109(2012)1735–1744. [36]P.Saranraj,P.Sivasakthivelan,M.Naveen,Fermentationoffruitwineandits

qualityanalysis:areview,Aust.J.Sci.Technol.1(2017)85–97.

[37] S.H.MohdAzhar,R.Abdulla,S.A.Jambo,H.Marbawi,J.A.Gansau,A.A.Mohd Faik,K.F.Rodrigues,Yeastsinsustainablebioethanolproduction:areview, Biochem.Biophys.Rep.10(2017)52–61.

[38]A.Ullah,R.Orij,S.Brul,G.J.Smits,Quantitativeanalysisofthemodesofgrowth inhibitionbyweakorganicacidsinSaccharomycescerevisiae,Appl.Environ. Microbiol.78(2012)8377–8387.

[39]C. Verduyn, E. Postma, W.A. Scheffers, J.P. Van Dijken, Energetics of Saccharomycescerevisiaeinanaerobicglucose-limitedchemostatcultures, Microbiology136(1990)405–412.

[40]L.G.M.Boender,E.A.F.Hulster,A.J.A.deMaris,P.A.S.van,Daran-Lapujade,J.T. Pronk,Quantitativephysiologyofsaccharomycescerevisiaeatnear-zero specificgrowthrates,Appl.Environ.Microbiol.75(2009)7578.

[41]L.Tijhuis, M.C. VanLoosdrecht,J.J. Heijnen,Athermodynamically based correlationformaintenancegibbsenergyrequirementsinaerobicand anaerobicchemotrophicgrowth,Biotechnol.Bioeng.42(1993)509–519. [42]T.Pinheiro,K.Y.F.Lip,E.Garcia-Rios,A.Querol,J.A.Teixeira,W.Gulik,J.M.van

Guillamon,L.Domingues,DifferentialproteomicanalysisbySWATH-MS unravelsthemostdominantmechanismsunderlyingyeastadaptationto non-optimaltemperaturesunderanaerobicconditions,bioRxivorg(2020) 2020.01.06.895581.

[43]F.Aguilera,R.A.Peinado,C.Millán,J.M.Ortega,J.C.Mauricio,Relationship betweenethanoltolerance,H+-ATPaseactivityandthelipidcompositionof

theplasmamembraneindifferentwineyeaststrains,Int.J.FoodMicrobiol. 110 (2006)34–42.

[44]J.Ding,X.Huang,L.Zhang,N.Zhao,D.Yang,K.Zhang,Toleranceandstress responsetoethanolintheyeastSaccharomycescerevisiae,Appl.Microbiol. Biotechnol.85(2009)253.

[45]T.M.Swan,K.Watson,Stresstoleranceinayeaststerolauxotroph:roleof ergosterol,heatshockproteinsandtrehalose,FEMSMicrobiol.Lett.169(1998) 191–197.

[46]G.Liu,Y.Chen,N.J.Færgeman,J.Nielsen,Eliminationofthelastreactionsin ergosterolbiosynthesisalterstheresistanceofSaccharomycescerevisiaeto multiplestresses,FEMSYeastRes.17(2017).

[47]C.A.Suarez-Mendez,M.Hanemaaijer,A.tenPierick,J.C.Wolters,J.J.Heijnen,S. A.Wahl,Interactionofstoragecarbohydratesandothercyclicfluxeswith centralmetabolism:aquantitativeapproachbynon-stationary13Cmetabolic fluxanalysis,Metab.Eng.Commun.3(2016)52–63.

[48]T.Higashiyama, Novelfunctionsandapplicationsoftrehalose,PureAppl. Chem.74(2002)1263–1269.

[49]M.A.Singer,S.Lindquist,Multipleeffectsoftrehaloseonproteinfoldingin vitroandinvivo,Mol.Cell1(1998)639–648.

[50]M.B.França,A.D.Panek,E.C.A.Eleutherio,Oxidativestressanditseffects duringdehydration,Comp.Biochem.Physiol.PartAMol.Integr.Physiol.146 (2007)621–631.

[51]J.W.G.Paalman,R.Verwaal,S.H.Slofstra,A.J.Verkleij,J.Boonstra,C.T.Verrips, TrehaloseandglycogenaccumulationisrelatedtothedurationoftheG1phase ofSaccharomycescerevisiae,FEMSYeastRes.3(2003)261–268.

[52]H.H.W.Silljé,J.W.G.Paalman,E.G.terSchure,S.Q.B.Olsthoorn,A.J.Verkleij,J. Boonstra,C.T.Verrips,Functionoftrehaloseandglycogenincellcycle progressionandcellviabilityinSaccharomycescerevisiae,J.Bacteriol.181 (1999)396–400.

[53]J.L.Parrou,M.-A.Teste,J.François,Effectsofvarioustypesofstressonthe metabolismofreservecarbohydratesinSaccharomycescerevisiae:genetic evidenceforastress-inducedrecyclingofglycogenandtrehalose, Microbiology143(1997)1891–1900.

[54]B.M.A. Abdel-Banat, H. Hoshida, A. Ano, S. Nonklang, R. Akada, High-temperaturefermentation:howcanprocessesforethanolproductionat hightemperaturesbecomesuperiortothetraditionalprocessusing mesophilicyeast?Appl.Microbiol.Biotechnol.85(2010)861–867. [55]E.Eleutherio,A.Panek,J.F.D.Mesquita,E.Trevisol,R.Magalhães,Revisiting

yeasttrehalosemetabolism,Curr.Genet.61(2015)263–274.

[56]P.A.Gibney,A.Schieler,J.C.Chen,J.D.Rabinowitz,D.Botstein,Characterizing theinvivoroleoftrehaloseinSaccharomycescerevisiaeusingtheAGT1 transporter,ProcNatlAcadSciUSA112(2015)6116–6121.

[57]C.Coutinho,E.Bernardes,D.Félix,A.D.Panek,Trehaloseascryoprotectantfor preservationofyeaststrains,J.Biotechnol.7(1988)23–32.

[58]B.Schade,G.Jansen,M.Whiteway,K.D.Entian,D.Y.Thomas,Coldadaptationin buddingyeast,Mol.Biol.Cell15(2004)5492–5502.

[59]R. Pérez-Torrado, J.V. Gimeno-Alcañiz, E. Matallana, Wine yeast strains engineeredforglycogenoverproductiondisplayenhancedviabilityunder glucosedeprivationconditions,Appl.Environ.Microbiol.68(2002)3339–3344. [60]S.Rodríguez-Vargas,A.Sánchez-García,J.M.Martínez-Rivas,J.A.Prieto, F.

Randez-Gil,Fluidizationofmembranelipidsenhancesthetoleranceof Saccharomycescerevisiaetofreezingandsaltstress,Appl.Environ.Microbiol. 73(2007)110–116.