纤维素酶经典综述_免费下载

REVIEW

TowardanAggregatedUnderstanding

ofEnzymaticHydrolysisofCellulose:

NoncomplexedCellulaseSystems

Yi-HengPercivalZhang,

1

1,2

ThayerSchoolofEngineering,DartmouthCollege,Hanover,

NewHampshire03755;e-mail:

@

,

@

2

DepartmentofBiologicalSciences,DartmouthCollege,Hanover,

NewHampshire03755

Received2June2004;accepted29July2004

Publishedonline10November2004inWileyInterScience().DOI:10.1002/bit.20282

1

Abstract:Informationpertainingtoenzymatichydrolysisof

cellulosebynoncomplexedcellulaseenzymesystemsis

reviewedwithaparticularemphasisondevelopmentof

aggregatedunderstandingincorporatingsubstratefeatures

inadditiontoconcentrationandmultiplecellulasecompo-

consideredincludepropertiesofcellulose,

adsorption,cellulosehydrolysis,andquantitativemodels.

Aclassificationschemeisproposedforquantitativemodels

forenzymatichydrolysisofcellulosebasedonthenumber

ofsolubilizingactivitiesandsubstratestatevariablesin-

estthatitistimelytorevisitandreinvig-

oratefunctionalmodelingofcellulosehydrolysis,andthat

thiswouldbehighlybeneficialifnotnecessaryinorder

tobringtobearthelargevolumeofinformationavailable

oncellulasecomponentsontheprimaryapplicationsthat

motivateinterestinthesubject.

B2004WileyPeriodicals,Inc.

Keywords:adsorption;cellulose;cellulase;hydrolysis;ki-

neticmodel

INTRODUCTION

Thepotentialimportanceofcellulosehydrolysisinthecon-

textofconversionofplantbiomasstofuelsandchemicals

iswidelyrecognized(Lyndetal.,1991,1999;Himmeletal.,

1999),andcellulosehydrolysisalsorepresentsoneofthe

largestmaterialflowsintheglobalcarboncycle(Falkowski

etal.,2000).Thequantityofscientificinformationon

componentsofcellulose-hydrolyzingenzymesystemhas

e12-year

periodfrom1991to2003,forexample,thenumberof

knownglycosylhydrolasesgenesequenceshasincreased

from

f

300to>10,000,andthenumberofcellulase

crystalstructureshasincreasedfromseveralto

f

230(H.

Correspondenceto:Y.-

Contractgrantsponsors:DepartmentofEnergyandNationalInstituteof

StandardsandTechnology

Contractgrantnumbers:DE-FG02-02ER15350and60NANB1D0064

Henrissat,.).Alsoduringthisperiod,exten-

sivestructurallybasedclassificationschemeshavebeen

introducedforbothcatalyticandcellulose-bindingmod-

ules,andhaveledtonewinsightsandhypotheseswith

respecttotheevolutionofcellulasesystems(Henrissat,

1991;HenrissatandBairoch,1993,1996),updatedfre-

quentlyathttp//:/CAZY.

Inorderforthelargevolumeofavailableinformationon

cellulasecomponentstobebroughttobearontheprimary

applicationsthatmotivateinterestincellulosehydrolysis,

e.g.,conversionofrenewablyproducedbiomasstofuels

andcommoditychemicals,itisnecessarytoincorporate

thisinformationintoanunderstandingofcellulasesystems

comprisedofmultiplecomponentswithdistinctmodesof

uationisfurthercomplicatedbecausethe

actionofcellulaseenzymesystemsisimpactedbysubstrate

propertiesinadditiontoconcentration—suchasdegreeof

polymerization,crystallinity,accessiblearea,thepresence

oflignin—whichdependontheparticularsubstratebeing

courseofseekingan‘‘aggregated’’understandingofenzy-

matichydrolysisofcellulosethatincorporatesinforma-

tionaboutcellulasecomponentsandsubstratefeaturesin

additiontoconcentration,quantitativemodelsaretremen-

icularimportance,measuredpa-

rametersforcellulasecomponentsandsubstratescould

inprinciplebeincorporatedintomodelsusedtopredict

thebehaviorofmulticomponentcellulaseenzymesystems.

Comparisonofsuchpredictionstoexperimentalmeasure-

mentsisthemostsystematicandrigorousmeansavailable

bywhichtotestwhetherunderstandingofcellulasecompo-

nentsandtheirinteractionsissufficienttoexplainagiven

tion,onceaquantitativemodelisvali-

dated,itcanbeusedtorapidlyformulatenewhypothesesof

significanceinbothfundamentalandappliedcontexts.

B2004WileyPeriodicals,Inc.

Thisarticlereviewsavailableinformationonenzymatic

hydrolysisbynoncomplexedcellulasesystems;thatis,sys-

temsbasedoncomponentsthatactdiscretelyratherthan

asstablecomplexes(Lamedetal.,1983;Tommeetal.,

1995a).Aconsiderableportionofthisreviewisspent

onthepropertiesofcelluloseinlightofthecentralrole

suchpropertiesplayinmechanisticallybasedquantitative

icular,thefollowing

sectionconsiderscrystallinity,degreeofpolymerization,

accessibility,preparationandpropertiesofmodelsub-

strates,tion

CellulaseAdsorptionisdevotedtoadsorptionleadingtothe

formationofcellulose–cellulasecomplexes,includingad-

sorptionmodels,reversibility,andenzymemobility,aswell

asinferredaccessibilityofcellulosefromcellulaseadsorp-

fter,mechanisticunderstandingofcellulose

hydrolysisbynoncomplexedsystemsisaddressedin

CelluloseHydrolysis,withattentiongiventoconcep-

tualunderstandingofcellulosehydrolysis,featuresofthe

widelystudiedTrichodermareeseicellulasesystem,docu-

mentationandunderstandingofsynergismamongcellu-

lasecomponents,andasummaryofcurrentmechanistic

tionQuantitativeModelspresents

aclassificationschemeandsummarizesfeaturesofmod-

alsectionofferscon-

cludingperspectivesandoutlinesoutstandingchallenges

associatedwithunderstandingandmodelingnoncomplexed

urprimaryfocusisonthefunction

ofcellulasesratherthantheirstructure,weusetheolder,

functionallydefinednomenclatureratherthanthenewer

nomenclaturebasedonamino-acidsequenceandmolecu-

larstructure.

CELLULOSE

ellulose

productionbyphotosynthetichigherplantsandalgaeis

thoughttobebyfarthemostimportantintermsofglobal

carbonflows,celluloseproductionbynonphotosynthetic

organisms(certainbacteria,marineinvertebrates,fungi,

slimemoldsandamoebae)hasalsobeendocumented

(Coughlan,1985;Jarvis,2003;Lyndetal.,2002;Tomme

etal.,1995a).Celluloseisalinearcondensationpolymer

consistingofD-anhydroglucopyranosejoinedtogetherby

h-1,ocellobioseistherepeating

unitofcellulose,sinceadjacentanhydroglucosemole-

culesarerotated180jwithrespecttotheirneighbors

(Fig.1a).Thisrotationcausescellulosetobehighlysym-

metrical,sinceeachsideofthechainhasanequalnumber

ngofadjacentcellulosemol-

eculesbyhydrogenbondsandvanderWaal’sforcesre-

sultsinaparallelalignmentandacrystallinestructure.

Theextensivehydrogenbondsofinterchain(2peranhy-

droglucopyranose)andintrachain(2

f

3peranhydrogluco-

pyranose)producesstraight,stablesupramolecularfibers

ofgreattensilestrength(GardnerandBlackwell,1974a,b;

Krassig,1993;NevellandZeronian,1985).Incontrast,

starchcontainsamyloseandamylopectinconnectedby

a-1,4andtosomeextenta-1,6glucosidicbonds,forming

atightlycoiledhelicalstructuremaintainedbyinterchain

hydrogenbonds(Buleonetal.,1998;Calvert,1997).Na-

tivecellulose,referredtoascelluloseI,hastwodistinct

crystalliteforms,I

a

,whichisdominantinbacterialand

algalcellulose,andI

h

,whichisdominantinhigherplants

(AtallaandVanderhart,1984).Nativecellulose(celluloseI)

canbeconvertedtoothercrystallineforms(II–IV)byvar-

ioustreatments(KleinandSnodgrass,1993;Krassig,1993;

O’Sullivan,1997).

Celluloseexistassheetsofglucopyranoseringslyingin

aplanewithsuccessivesheetsstackedontopofeachother

eofthisar-

rangement,thesurfaceofacelluloseparticlehasdistinct

‘‘faces’’thatinteractwiththeaqueousenvironmentand

carbonsintheglucopyranose

ringandinternalh-glucosidicbondslieintheabplaneor

¯

0faceconsistsof‘‘110’’face,whereastheacplaneor11

theedgesofrings(seeFig.1b).Additionalfacespresent

reducingandnonreducingends,eating

unitofthe110faceisthecellobioselattice,whichmea-

sures1.04nmalongtheaxisofthecellulosemoleculeand

100cellu-

loseglucansareaggregatedintoelementaryfibrilswith

acrystallinewidthof4–5nm(O’Sullivan,1997),and

bunchesofelementaryfibrilsareembeddedinamatrixof

hemicellulosewithathicknessof7–nifica-

tionprocessoccurslateintheprocessofsynthesizingnat-

uralfibers,soligninislocatedprimarilyontheexterior

ofmicrofibrilswhereitcovalentlybondstohemicellulose

(Fig.1c;KleinandSnodgrass,1993).

Therelationshipbetweenstructuralfeaturesofcellu-

loseandratesofenzymatichydrolysishasbeenthesubject

ofextensivestudyandseveralreviews(Converse,1993;

CowlingandKirk,1976;Lyndetal.,2002;Mansfieldetal.,

1999;McMillian,1994),butisstillincompletelyunder-

uralfeaturesofcellulosecommonlyconsidered

asrate-impactingfactorsincludecrystallinityindex,degree

ofpolymerization,andaccessiblearea.

CrystallinityIndex(CrI)

Crystallinityhasoftenbeenthoughtofasprovidingan

indicationofsubstratereactivity,andisprominentlyfea-

turedinthemodelofWood(1975)aswellasothermodels.

Thecrystallinityofdriedcellulosesamplescanbequan-

titativelymeasuredfromthewide-rangeX-raydiffraction

pattern(Krassig,1993).Inthecaseofcellulose-I,thecrys-

tallinityindex(CrI)iscalculatedusingtheformula:

CrI¼1Àh

am

=h

cr

¼1Àh

am

=ðh

tot

Àh

am

Þð1Þ

basedontheratiooftheheightofcrystallinecellulosein

the002reflectionat2u=22.5j(h

cr

)totheheightof

amorphouscellulose(h

am

),andh

tot

=h

cr

+h

am

.Cotton

798BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

Figure1.a:Structureofcellulosefeaturingrepeatingh1,4-linkedanhydrocellobioseunits.b:softherepeatingunit

(cellobiose)are:a=0.817nm,b=1.04nm,andc=esoftheglucopyranoseringsareparalleltotheabplane(110face)ofthecrystal

(Mosieretal.,1999).c:Organizationoflignocelluloseoriganizationintoelementaryfibrilsandmicrofibrils(KleinandSnodgrass,1993).

(Hoshinoetal.,1997;Leeetal.,1982;Sinitsynetal.,1991),

bacterialcellulosefromAacetobacterxylinum(Boisset

etal.,1999;Gilkesetal.,1992;Valjamaeetal.,1999),and

cellulosefromthealgaValoniaventricosa(Boissetetal.,

1999;Fierobeetal.,2002)provideexamplesofhighly

crystallinecellulose,whilephosphoricacidswollencellu-

loseandball-milledcelluloseareregardedasamorphous

cellulose(Hoshinoetal.,1997;Leeetal.,1982;Ooshima

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS799

etal.,1983).Commonmodelsubstratesderivedfrom

bleachedcommercialwoodpulps,suchasAvicel(Wood

andBhat,1988;Wood,1988),filterpaper(Henrissatetal.,

1985),andSolkaFloc(BertrainandDale,1985;Fanetal.,

1980;Leeetal.,1982;Sinitsynetal.,1991)areregardedas

ablendofamorphousandcrystallineforms(Gilkesetal.,

1991).TypicalvaluesofCrIforvariousmodelcellulosic

valueofcel-

luloseincreasesafteraperiodofwaterswellingduetore-

crystallization(Fanetal.,1980;Leeetal.,1983;Fengeland

Wegener,1984),andthevariationsindryingconditionprior

tomeasurementofCrImaycausedifferencesbetween

substratesarisingfromthemethodofsubstratepreparation

ratherthanpropertiesofthesubstrateperse(Lenzeetal.,

1990;Weimeretal.,1995).Thepresenceofresidualcells

andproteinscanalsoresultinartifactsintheCrIassay

(Converse,1993).

Cellulosehydrolysisratesmediatedbyfungalcellulases

aretypically3–30timesfasterforamorphouscelluloseas

comparedtohighcrystallinecellulose(Lyndetal.,2002;

TableIII).Thisobservationledinvestigatorsinthe1980s

topostulateamodelforcellulosestructureconsistingof

amorphousandcrystallinefractions(Fanetal.,1980,1981;

Leeetal.,1983).Ifthishypothesiswerecorrect,itwould

beexpectedthatcrystallinityshouldincreaseoverthe

courseofcellulosehydrolysisasaresultofpreferential

reactionofamorphouscellulose(BetrabetandParalikar,

1977;Ooshimaetal.,1983).However,severalstudieshave

foundthatcrystallinitydoesnotincreaseduringenzymatic

hydrolysis(Lenzeetal.,1990;Ohmineetal.,1983;Pulsand

Wood,1991;Schurzetal.,1985;Sinitsynetal.,1989).Con-

sideringboththeuncertaintyofmethodologiesformea-

suringCrIaswellasconflictingresultsonthechangeof

CrIduringhydrolysis,itisdifficulttoconcludeatthistime

thatCrIisakeydeterminantoftherateofenzymatichy-

drolysis(Lyndetal.,2002;Mansfieldetal.,1999).

Futurestudiesaimedatdevelopingandapplyingim-

provedmethodswouldbeusefultomoredefinitivelyre-

rpreting

crystallinitydata,andindeeddataforallcellulosephysical

properties,caremustbetakentodistinguishcorrelation

mple,severaltreatmentsthat

decreasecrystallinityalsoincreasesurfacearea,andithas

beensuggestedthattheincreasedhydrolysisratesobserved

withsubstratesarisingfromsuchtreatmentsmaybedueto

increasingadsorptivecapacityratherthansubstratereac-

tivity(CaulfieldandMoore,1974;HowellandStuck,1975;

LeeandFan,1982).Comparingthehydrolysisrateson

varioussourcesofmodelcellulosicsubstrates,Fierobeetal.

(2002)concludedthataccessibilityofcelluloseisamore

importantfactorthancrystallinityindexindeterminingthe

hydrolysisrate.

DegreeofPolymerization

Thedegreeofpolymerization(DP)ofcellulosicsubstrates

determinestherelativeabundanceofterminalandinterior

h-glucosidicbonds,andofsubstratesforexo-actingand

endo-actingenzymes,edefinedin

termsofthenumberaverageDP(DP

N

),weightaverageDP

(DP

W

),orDPinferredfromviscosity(DP

V

):

P

M

n

N

i

M

i

DP

N

¼¼

P

=MW

glu

MW

glu

N

i

ð2Þ

P

N

i

M

i

2

M

W

P

DP

W

¼¼=MW

glu

N

i

MW

glu

P

M

V

N

i

D

¼

P

=MW

glu

DP

V

¼

MW

glu

N

i

ð3Þ

ð4Þ

yofsomephysicalpropertiesofmodelcellulosic

substrates.

Substrate

1

Avicel

BC

PASC

Cotton

FilterPaper

Woodpulp

1

CrI

2

0.5–0.6

0.76–0.95

0–0.04

0.81–0.95

–0.45

0.5–0.7

SSA

2

(m

2

/g)

20

200

240

na.

na.

61–55

DP

N

2

300

2000

100

1000–3000

750

500–1500

F

RE

(%)

0.33

0.05

1.0

0.1–0.033

0.13

0.06–0.2

BC,bacterialcellulose;PASC,phosphoricacidswollencellulose;CrI

denotescrystallinityindex;SSAdenotesspecificsurfaceareabyBET;

DP

N

denotesthenumber-averagedegreeofpolymerization;F

RE

denotes

thefractionofreducingends.

2

Referencesintext.

whereN

i

isthenumberofmolesofagivenfractionihaving

molarmassM

i

,M

N

isthenumber-averagemolecular

weight,M

w

istheweight-averagemolecularweight,M

V

is

theviscosity-averagemolecularweight,MW

glu

isthe

molecularweightofanhydroglucose(162g/mol),andDis

ementofDPbeginswithdissolutionof

celluloseusingatechniquethatdoesnotalterchainlength.

Severalsuchmethodsappearsatisfactory,including:1)

metalcomplexsolutionssuchasCuamsolution(Klemm

etal.,1998)andcupriethylenediamine(Klemen-Leyeretal.,

1992,1994,1996);2)formingcellulosederivativesby

reactingwithorganicsolvents(NgandZeikus,1980)or

inorganicacidssuchasnitricacid(Whitaker,1957);and3)

ionicsolutionssuchasN,N-dimethylacetamide(DMAc)/

LiCl(Striegel,1997).Afterdissolution,DP

N

canbemea-

suredbymembraneorvaporpressureosmometry,cry-

oscopy,ebullioscopy,determinationofreducingendcon-

centration,orelectronmicroscopy(Krassig,1993).DP

W

canbemeasuredbasedonlightscattering,sedimentation

equilibrium,andX-raysmallanglescattering,andDP

V

is

cosityofdissolved

celluloseorcellulosederivativeshasbeenfoundtoequal:

D¼K

m

M

i

aþ1

ð5Þ

800BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

inwhichK

m

=constant,withthevalueofaforcellulose

andcellulosederivativesinmostcasesrangingfrom0.75to

1(Krassig,1993).Therefore,DP

V

canbewrittenas:

P

N

i

M

i

1:75À2

P

=MW

glu

ð6Þ

DP

V

¼

N

i

Sincecelluloseispolydisperse,DP

W

zDP

V

>DP

N

.The

DP

N

valuesareadequateindealingwithcellulosehydrol-

ysis,andDP

W

andDP

V

frequentlyshowagoodcorrelation

topolymerproperties(Klemmetal.,1998;Krassig,1993).

ThedistributionofDPsamongapopulationofcellulose

moleculescanbemeasuredbysizeexclusionchromatog-

raphy(Yauetal.,1979).ThereciprocalofDPcorresponds

tothefractionofreducingendsrelativetoallglucanunits

present(F

NR

,unitless).

Cellulosesolubilitydecreasesdrasticallywithincreasing

extrins

withDPfrom2–6aresolubleinwater(Klemmetal.,1998;

Miller,1963;Pereiraetal.,1988),whilecellodextrinsfrom

7–13orlongeraresomewhatsolubleinhotwater(Zhang

andLynd,2003;Schmidetal.,1988).AglucanofDP=30

alreadyrepresentsthepolymer‘‘cellulose’’initsstructure

andproperties(Klemmetal.,1998).

TheDPofcellulosicsubstratesvariesgreatly,from<100

to>15,000,dependingonsubstrateoriginandpreparation,

fwoodafterpulpingis

reducedto500–1,500(BertrainandDale,1985;Kleinand

Snodgrass,1993;Leeetal.,1982;Swatloskietal.,2002).

Afterpartialacidhydrolysis,theDPofAvicelisfurther

decreasedto130–800(Hoshinoetal.,1997;NgandZeikus,

1980;Ross-Murphy,1985;Steineretal.,1988;Wood,1985),

dependingonhydrolysisconditions(Dongetal.,1998)and

theDPoftheoriginalsubstrate(Wood,1988).Similarly,the

DPofnaturalcottoncanbeashighas15,000,butisreduced

to1,000–3,000orlessinthepreparationofcottonlinters

involvingtreatmenttoaccomplishdewaxingandwhitening

(Kleman-Leyeretal.,1992,1996;OkazakiandMoo-

Young,1978;RyuandLee,1982),andfilterpapermade

fromcottonpulphasaDPof500–1,000orhigher

(Nisizawa,1973;Kongruangetal.,2004).Bacterialcel-

lulose(BC)hasanaverageDPof2,000–3,000(Hestrin,

1963;Fierobeetal.,2002;Valjamaeetal.,1999),while

bacterialmicrocrystallinecellulose(BMCC)preparedby

treatmentofBCwithacidsrangesfrom130–1,300,de-

pendingonhydrolysisconditions(Valjamaeetal.,1999).

TheDPofphosphoric-acidswollencellulose(PASC)ranges

from30tomorethan1,000(Fanetal.,1980;Krassig,1985;

Petreetal.,1981;WoodandMcCrae,1972),depending

ontheDPofthestartingsubstrate(Wood,1988;Hoshino

etal.,1997),aswellasthephosphoricacidincubationtime

andtemperature(Krassig,1993).

ThechangeinDPoverthecourseofhydrolysisfor

cellulosicsubstratesisdeterminedbytherelativepropor-

tionofexo-andendo-actingactivitiesandcelluloseproper-

canasesactonchainends,andthusdecrease

DPonlyincrementally(Kleman-Leyeretal.,1992,1996;

Srisodsuketal.,1998).Endoglucanasesactoninterior

portionsofthechainandthusrapidlydecreaseDP(Kleman-

Leyeretal.,1992,1994;Selby,1961;Srisodsuketal.,1998;

Whitaker,1957;WoodandMcCrae,1978).Exoglucanase

hasbeenfoundtohaveamarkedpreferenceforsubstrates

withlowerDP(Wood,1975),aswouldbeexpectedgiven

thegreateravailabilityofchainendswithdecreasingDP.

Itiswellknownthatendoglucanaseactivityleadstoan

increaseinchainendswithoutresultinginappreciable

solubilization(Irwinetal.,1993;Kruusetal.,1995;Re-

verbel-Leroyetal.,1997).Weknowofnoindicationinthe

literaturethattherateofchainendcreationbyendogluca-

naseisimpactedbysubstrateDP.

Accessibility

Cellulaseenzymesmustbindtothesurfaceofsubstrate

particlesbeforehydrolysisofinsolublecellulosecantake

3Dstructureofsuchparticles(includingmicro-

structure)incombinationwiththesizeandshapeofthe

cellulaseenzyme(s)underconsiderationdeterminewhether

h-glucosidicbondsareorarenotaccessibletoenzymatic

osicparticleshavebothexternalandinternal

ral,theinternalsurfaceareaofcelluloseis

1–2ordershigherthantheexternalsurfacearea(Chang

etal.,1981),butthisisnotalwaysthecase,forexample,in

ernalsurfaceareacan

bemeasuredbysmallangleX-rayscattering(SAXS),mer-

curyporosimetry,watervaporsorption,andsizeexclusion

(Grethlein,1985;NeumanandWalker,1992;Stoneetal.,

,

naturalcotton;NW,naturalwood;P,pulp;CT,cottonlinter;FP,filterpaper.

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS801

1969).Theinternalsurfaceareaofporouscelluloseparticles

dependsonthecapillarystructureandincludesintrapar-

ticulatepores(1–10nm)aswellasinterparticulatevoids

(>5Am)(MarshallandSixsmith,1974).Grethlein(1985)

foundlinearcorrelationsbetweentheinitialhydrolysisrate

ofpretreatedbiomassandtheporesizeaccessibletoa

˚

,similartothesizeofmoleculewithadiameterof51A

surfaceexposedto

dextrancannotdistinguishthespecificactivecellulosesur-

faceareaatwhichenzymatichydrolysisoccursfromthe

surfaceareawhichisnotasiteforenzymaticattack(Chanzy

etal.,1984;Gilkesetal.,1992;Lehtioetal.,2003),re-

sultinginpotentialoverestimationofeffectivecellulase-

quesformeasuringinternalsurface

generallydonotestimateexternalarea(Converse,1993).

Externalsurfaceareaiscloselyrelatedtoshapeandpar-

ticlesize,andcanbeestimatedbymicroscopicobservation

(Gilkesetal.,1992;Henrissatetal.,1988;Reinikainenetal.,

1995b;Weimeretal.,1990;WhiteandBrown,1981).For

example,theexternalsurfaceareaofBMCCis

f

115m

2

/g

(Gilkesetal.,1992)whereasthatofAvicelis

f

0.3m

2

/g

(Weimeretal.,1990).Increasingcellulaseadsorptionand

cellulosereactivitywithdecreasingparticlesizehasbeen

reported(Kimetal.,1992;Mandelsetal.,1971).However,

thismaybeduetocausesotherthanincreasedexternalarea,

perhapsdecreasingmasstransferresistance,sinceexternal

surfaceisthoughttobeasmallfractionofoverallsurface

areaformostsubstrates.

Thegrosscelluloseaccessibilityisgenerallymeasuredby

thesorptionofnitrogen,argonorwatervapor,dimensional

changeorweightgainbyswellinginwaterororganic

liquids,andexchangeofHtoDatomswithD

2

t

widelyusedprocedureforspecificsurfacearea(SSA)is

theBrunauer-Emmett-Teller(BET)methodusingnitrogen

ariationsintheexperimentalcondi-

tionssuchasadsorptiontime,vacuumtimeandvacuum

pressure(MarshallandSixsmith,1974),sampleprepara-

tion(Grethlein,1985;Leeetal.,1983),andsampleorigin

andfeatures(MarshallandSixsmith,1974;Weimeretal.,

1990),awiderangeofgrossareavalueshavebeenreported

cific

areaofAvicelPH102increasesfrom5.4m

2

/gsurfacearea

to18m

2

/gafteralongtimeofwaterswelling,becausethe

capillarystructureofair-driedcellulosefromthewater-

swollenstatecollapses,resultingindrasticchangesinphys-

icalparameters(Grethlein,1985;Leeetal.,1983).Tokeep

substratecapillarystructureasitexistsinthehydrated

state,itisrecommendedthatSSAbemeasuredusing

solvent-driedsamples(Grethlein,1985;Leeetal.,1983).

ThetypicalSSAofBMCC,Avicel,andwetpulpare

f

200m

2

/gBMCC(Bothwelletal.,1997),1.8–22m

2

/g

Avicel(Fanetal.,1980;Leeetal.,1983;Marshalland

Sixsmith,1974),and55–61m

2

/gpulp(Fanetal.,1980;

Kyriacouetal.,1988).ThespecificsurfaceareaofPASC

fromSolkaFlocincreasesfrom19.5to239m

2

/gwhen

phosphoricacidconcentrationincreasesfrom75%to85%

(Leeetal.,1982).Becauseanitrogenmoleculeismuch

smallerthancellulase,ithasaccesstoporesandcavities

ore,

thereislimitedbasistoinferthatSSAmeasuredusingthe

BETmethodisakeydeterminantofenzymatichydrolysis

rate(Mansfieldetal.,1999).

PreparationandPropertiesofModelSubstrates

Woodpulpismadefromwoodusingseveralsteps,in-

cludingshredding,delignification,bleaching,andwashing

(Klemmetal.,1998).Forexample,SolkaFlocismade

fromSO

2

-bleachedsprucepulpbyballmilling(Ghose,

1969).Avicel,alsocalledhydrocelluloseandmicrocrystal-

linecellulose,ispreparedfromcellulosicfibers(woodpulp)

bypartialacidhydrolysisandthenspraydryingofthe

washedpulpslurry,butmicrocystallinecellulose(Avicel)

stillcontainsasubstantialamount(

f

30–50%)ofamor-

phouscellulose(Krassig,1993).Bacterialcellulose(BC)

ispreparedfromthepellicleproducedbyAcetobacter

xylinum(ATCC23769)(Hestrin,1963)orfromNatade

Coco(DaiwaFineProduces,Singapore;Boissetetal.,

2000).Bacterialmicrocrystallinecellulose(BMCC)ispre-

paredfromBCbypartialacidhydrolysistoremoveamor-

phouscellulose(Valjamaeetal.,1999).Cottoncellulose

ismadefromnaturalcottonafterremovingimpurities

suchaswax,pectin,andcoloredmatter(Corbett,1963).

WhatmanNo.1filterpaperismadefromcottonpulp

(Dongetal.,1998).Homogenousamorphouscellulosecan

,Avicel,

cottonlinters,byswellingtreatmentssuchasphosphoric

acid,alkali,DMSO,DMAc/oricacidswol-

lencellulose(PASC)ismostcommonlymadebyswelling

cellulosepowderusingconcentratedphosphoricacid,re-

sultingindecreasedcrystallinity(Wood,1988).Typical

valuesforCrI,DP,grosssurfaceareavalues(SSAbyBET),

andfractionofreducingends(F

NR

,reciprocalofDP)for

modelcellulosicsubstratesarepresentedinTableI.

CharacteristicsofPretreatedLignocellulose

Naturalcellulosemoleculesoccurinelementaryfibrils

closelyassociatedwithhemicelluloseandotherstructural

polysaccharidesaswellaslignin(Fig.1c).Suchligno-

cellulosetypicallycontainscellulose(35–50wt.%),hemi-

cellulose(20–35wt.%),andlignin(5–30wt.%)(Chang

etal.,1981;KleinandSnodgrass,1993;Lyndetal.,2002;

Mansfieldetal.,1999).Adetailedconsiderationofen-

zymatichydrolysisofnativelignocellulosemaybefound

elsewhere(Hatfieldetal.,1999).Sinceenzymatichydro-

lysisofnativelignocelluloseusuallyresultsinsolubiliza-

tionofV20%oftheoriginallypresentglucan,someform

ofpretreatmenttoincreaseamenabilitytoenzymatichy-

drolysisisincludedinmostprocessconceptsforbiological

atment,underappro-

priateconditions,retainsnearlyallofthecellulosepres-

entintheoriginalmaterialandallowsclosetotheoretical

edpretreatment

802BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

processesincludediluteacid,steamexplosionathighsolid

concentration,‘‘hydrothermal’’process,‘‘organosolv’’pro-

cessesinvolvingorganicacidsolventsinanaqueousphase,

ammoniafiberexplosion(AFEX),strongalkaliprocess

(Lyndetal.,2002),aswellasmechanicaltreatmentssuch

ashammerandballmilling(Millettetal.,1976;Sunand

Cheng,2002).Comparativefeaturesoftheseprocessesas

wellasconsiderationofsubstratefactorsimpactingthehy-

drolysisratearereviewedelsewhere(Changetal.,1981;

Converse,1993;CowlingandKirk,1976;Dale,1985;Hsu,

1996;Ladischetal.,1983;Mansfieldetal.,1999;McMillian

1994;Lynd,1996;SunandCheng,2002;Weiletal.,1994;

WoodandSaddler,1988).

Hydrolysisoflignocellulosicbiomassismorecompli-

catedthanthatofpurecelluloseduetothepresenceof

nonglucancomponentssuchasligninandhemicellulose.

Ligninremovaland/orredistributionarethoughttohavea

significanteffectonobservedratesofenzymatichydrolysis

(Chernoglazovetal.,1988;Converse,1993;Lyndetal.,

2002).Ligninhasbeenimplicatedasacompetitivecel-

lulaseadsorbentwhichreducestheamountofcellulase

availabletocatalyzecellulosehydrolysis(Bernardezetal.,

1993;Ooshimaetal.,1990;SutcliffeandSaddler,1986).

Inaddition,ithasbeensuggestedthatresidualligninblocks

theprogressofcellulasedownthecellulosechain(Eriksson

etal.,2002;Mansfieldetal.,1999).

Themeasuredcrystallinityindexoflignocelluloseis

,

caremustbetakenincomparingCrIvaluesforlignocellu-

losicsubstratestovaluesforcellulosicsubstrates,andalso

incomparingtheCrIoflignocellulosicsubstratesbefore

edCrIvaluesforpretreated

materialsaregenerallyintherangeof0.4–0.7(Changand

Holtzapple,2000;Gharpurayetal.,1983;Koullasetal.,

1992;Sinitsynetal.,1989,1991).Pretreatmentbyeither

dilute-acidorsteamexplosionunderconditionsthatare

quiteeffectiveinenhancinghydrolysishasbeenfoundto

increasethecompositeCrIoflignocellulose(Deschamps

etal.,1996;Kimetal.,2003;Knappertetal.,1980;

Meunier-Goddiketal.,1999).Consistentwiththis,a

negativecorrelationbetweenhydrolysisrateandCrIhas

beenshowninexperimentsthatinvolvedchemicalpre-

treatmentsfollowedbyballmilling(ChangandHoltzapple,

2000;Gharpurayetal.,1983;Knappertetal.,1980;Koullas

etal.,1992;Sinitsynetal.,1989,1991),andalsoex-

perimentsthatexaminedvariouspretreatmentconditions

(ChangandHoltzapple,2000).Incontrasttothetrendob-

servedforotherpretreatmentprocesses,AFEXpretreat-

menthasbeenreportedtoresultinadecreaseinCrI

(Gollapallietal.,2002).Severalinvestigatorshaveimpli-

catedaccessiblesurfaceareaasanimportantfactorin

determiningtheeffectivenessofpretreatment(Gharpuray

etal.,1983;Grethlein,1985;GrethleinandConverse,1991;

Sinitsynetal.,1991).Asignificantdifficultyininterpreting

theeffectsofpretreatmentatamechanisticlevelisthat

exposureofsubstratestoconditionsthatcauseonepoten-

tialdeterminantofreactivitytochangeusuallybringabout

mple,

Sinitsynetal.(1991)foundastrongnegativecorrelation

betweenCrIandaccessiblesurfaceareaaccompanying

ectthattheimpact

ofincreasedsurfaceareaaccompanyingpretreatmentmay

inmanycasesbemoreimportantthanchangesinCrI,al-

thoughfurtherworkwillbeneededtoestablishthispoint

andtherelativesignificanceoftheseandotherfactorsmay

wellbedifferentfordifferentprocesses.

DPvaluesoflignocellulosicsubstratessuchasba-

gasse,wheatstraw,andEucalyptusregnanspretreated

usingsteamexplosion,supercriticalCO

2

,alkali,andozone

mostlyfallintherangeof600–1,100,althoughvaluesas

highas3,000havebeenrecordedforPinusradiatachips

(Puri,1984;Sinitsynetal.,1991).Duringdiluteacid-

catalyzedcellulosehydrolysis,theDPofcellulosicma-

terialsdecreasesrapidlyinitiallyandachievesanearly

constantvaluethereaftercalledthelevel-offDP(LODP)

(Klemmetal.,1998;Krassig,1993;Wood,1988).LODP

valuesintherangeof100–300havebeenmeasured,de-

pendingonthesubstrateandconditionssuchastemperature

andacidconcentration(Krassig,1993;Wood,1988).This

LODPvaluemaylimittheratesofhydrolysisthatcanoc-

curwithdiluteacidpretreatedlignocellulose,althoughthis

entcon-

clusionsabouttheimportanceofDPindetermining

hydrolysisratesofpretreatedcellulosicbiomasshavebeen

drawn,withSinitsynetal.(1991)concludingthatDPis

relativelyunimportant,butPuri(1984)concludingthatitis

quiteimportant.

CELLULASEADSORPTION

Adsorption

Cellulaseadsorptionisrapidcomparedtothetimere-

quiredforhydrolysis,withmanystudiesfindingthatad-

sorptionreachessteady-statewithinhalfanhour(Lynd

etal.,2002).Themostcommondescriptionofcellulase

adsorptionistheLangmuirisotherm(Eq.[7]),derivedas-

sumingthatadsorptioncanbedescribedbyasinglead-

sorptionequilibriumconstantandaspecifiedadsorption

gmuirisothermmayberepresentedas:

E

a

¼

W

max

K

P

E

f

1þK

P

E

f

ð7Þ

inwhichE

a

isadsorbedcellulase(mgorAmolcellulase/L),

W

max

isthemaximumcellulaseadsorption=A

max

*

S(mgor

Amolcellulase/L),A

max

isthemaximumcellulaseadsorp-

tionpergcellulose(mgorAmolcellulase/gcellulose),Sis

celluloseconcentration(gcellulose/L),E

f

isfreecellulase

(mgorAmolcellulase/L),andK

P

isthedissociationconstant

E

a

(K

P

¼

E

)intermsofL/tributioncoef-

f

S

ficientorpartitioncoefficient,R,isdefinedas:

R¼K

P

W

max

ð8Þ

803ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS

RhasdimensionsofL/gcelluloseandcorrespondstothe

ratioofE

a

/E

f

whensubstrateisexcess,andhenceE

f

=0

(Beldmanetal.,1987;Klyosov,1988,1990;Kyriacouetal.,

1988;Medveetal.,1997).Inadditiontoequilibriumad-

sorptionmodels,adynamicadsorptionmodelhasbeenused

bysomeinvestigators(Converseetal.,1988;Converse

andOptekar,1993;NidetzkyandSteiner,1993;Nidetzky

etal.,1994c).

TheLangmuirequationiswidelyusedbecauseit

providesagood(andoftenverygood)fittothedatain

mostcases,anditrepresentsasimplemechanisticmodel

thatcanbeusedtocomparekineticpropertiesofvarious

cellulase–sevidentthatcellulase

bindingdoesnotcomplywithassumptionsimplicitinthe

Langmuirmodelduetooneormoreofthefollowing:1)

partiallyirreversiblecellulaseadsorption(Palonenetal.,

1999);2)interactionamongadsorbingcellulasecompo-

nents,especiallyathighconcentrations(Jeohetal.,2002);

3)multipletypesofadsorptionsites,evenforonecellulase

molecule(LinderandTeeri,1997;CarrardandLinder,

1999);4)cellulaseentrapmentbyporesofcellulose(Lee

etal.,1983);and5)multicomponentcellulaseadsorptions

inwhicheachcomponenthasdifferentconstants(Beld-

manetal.,1987).Inlightoftheseconsiderations,several

equilibriummodelsrepresentingalternativestosimple

Langmuiradsorptionhavebeenproposed,includingtwo-

siteadsorptionmodels(Linderetal.,1996;Medveetal.,

1997;Stalhbergetal.,1991;Woodwardetal.,1988a),

Freundlichisotherms(Medveetal.,1997),andcombined

LangmuirFreundlichisotherms(Medveetal.,1997).

Langmuirparametersforcellulaseadsorptionarepres-

entedinTableII,withanemphasisonnoncomplexed

ghwidevariationsareobservedin

thevaluesofparametersfordifferentcombinationsofen-

zyme,substrate,andtemperature,reproducibilityamong

measurementsfromdifferentlabstakenforthesameen-

-

sider,forexample,

cellulaseslistedinTableII,includingCBH1onBMCCat

4jC(Reinikainenetal.,1995b;Srisodsuketal.,1993)

and50jC(Bothwelletal.,1997;Tommeetal.,1995b),

CBH1onAvicelat20–25jC(KimandHong,2000;

Stahlbergetal.,1991;Tommeetal.,1990),andunfraction-

atedcellulaseadsorbingtoAvicelat4jC(Leeetal.,1982;

Luetal.,2002;Ooshimaetal.,1983).Thisreproducibility

suggeststhatexperimentalmethodsformeasurementof

adsorptionparametersmaybesufficientlystandardizedsuch

thatvaluesfromdifferentlabscanbemeaningfullycom-

estthatitmaybeusefultocalibratetech-

niqueswithmeasurementsmadeunderwell-characterized

tion

toexperimentalvariables,differentregressionmethods

canleadtodifferentvaluesforparameters(Bothwelland

Walker,1995).

GhoseandBisaria(1979)foundthatendoglucanasesad-

etal.(1984)foundthatcellulasecontainedtightlyad-

sorbedcellobiohydrolases,somelooselyboundEG1,and

a

etal.(1983)

endoglucanasesandcellobiohydrolaseswastemperature-

dependent,withendoglucanasespreferentiallyadsorbedat

5jC,andcellobiohydrolasespreferentiallyboundat50jC.

Bycontrast,Kyriacouetal.(1989)foundthatadsorptionof

CBH1wasstrongerthanadsorptionofEG1-3on

SolkaFlocat5jC,butthatpreferentialadsorptionofCBH1

wasdiminishedat50jC,andsuchpreferentialadsorption

wasalsoobservedtobelesspronouncedwithdecreasing

al.(2002)reportedthatthecombined

ellulasesCel5A,Cel6B,andCel9A

waslowerthanthesumofindividualadsorptionatlow

temperaturebuthigherat50jConBMCC.

Mostearlypublishedstudieshavedealtwiththerevers-

ibilityofcellulaseadsorptionbymeasuringtheamountof

enzymereleasedintosolutionascellulosehydrolysisprog-

ressed(Huang,1975;LeeandFan,1982;Mandelsetal.,

1971;MoloneyandCoughlan,1983).ButBeltrameetal.

(1982)determinedthattheadsorptionofproteinconsisted

ofirreversiblesteps,whichwerethoughttoarisefromcon-

al.

(1984)contradictedBeltrame’sfindingbyreportingthat

adsorbedcellulasecanberemovedbywashingwithbuf-

ractionatedcellulase,Kyriacouetal.(1989)

foundcellulaseadsorptionwasirreversible,whileBeldman

etal.(1987)foundcellulaseadsorptiontobepartially

,

Palonenetal.(1999)foundthatdesorptionofCBH2in

responsetosampledilutionshowedhysteresis(60–70%

reversible),whiledesorptionofCBH1wasmorethan90%

zkyetal.(1994b)

CBH1adsorptionispartiallyreversibleduetoitsbifunc-

CBH1CBMon

microcrystallinecellulosewasreportedtobereversible

(LinderandTeeri,1996),CBH2CBMcould

notbedissociatedfromcellulose(CarrardandLinder,

1999).el5A,Cel6B,

Cel48AontoBMCCwasreversibleatlowconcentration

butirreversibilitywasobservedathighcellulasecon-

centrations,apparentlyduetointerstitialentrapment(Jung

etal.,2002).

Inanagitatedbatchreactor,theintensityofagitationhas

littleeffectoncellulosehydrolysisaslongascellulosepar-

ticlesarecompletelysuspended(Huang,1975).Jervisetal.

(1997)studiedsurfacediffusionofCellulomonasfimicel-

lulasesCexandCenAonthesurfaceofValoniaventricosa

microcrystallinecelluloseusingfluorescencerecoveryaf-

terphotobleaching(FRAF).Basedoncomparisonofthe

valueofdiffusioncoefficientandspecificcellulaseactivity,

theseinvestigatorsinferredthatexternaldiffusionofcel-

lulaseisnotarate-limitingfactorforthewholereaction.

Ingeneral,experimentsexaminingstirringratealsosug-

gestthatexternaldiffusionofcellulaseonthesurfaceis

notrate-limiting(Fanetal.,1981;FanandLee,1983).But

wheninternalareaisfarlargerthanexternalsurface,which

804BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

yofLangmuircellulaseadsoprtopmparameterfornoncomplexcellulasesandtheirsolecellulose-bindingdomains.

a

Strain

cellum

cellum

ovorans

ovorans

ovorans

cellum

cellum

a

Cellulase

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBH3(CBH1)

CBH1

CBH2

CBH2

CBH2

CBH2

CBH2

CBH2

EG1

EG1

EG2

EG3

EG4

EG5

EG6

EG3

total

total

total

total

total

CBM

CipA

CBM

CipA

CBM

CbpA

CBM

CbpA

CBM

CipA

CBM

Ce1K

CBM

Ce1K

CBM

Cex

CBM

Cex

CBM

Cex

CBM

E3

CBM

E3

CBM

CBH1

CBM

CBH2

SubstrateTemp.(jC)

50

4

4

50

50

20

25

4

20

40

30

50

25

4

20

20

30

50

50

30

30

30

30

30

30

50

5

4

4

4

2–8

25

25

37

37

37

A

max

mg/g(Amol/g)

(4.6)

(6.0)

(4.2)

(2.63)

(0.48)

69(1.1)

70(1.07)

48(0.74)

51.8

40

63

(0.17)

64(1.10)

28(0.52)

54.3

48.9

6.6

(0.258)

(0.166)

126

90

26

2.8

105

4.1

(0.308)

55.6

64

95.2

1224

78–89

10(0.54)

200(1.08)

(2.1)

(6.4)

(0.2)

(17.1)

(3.95)

40

3

13.3

(1.65)

(1.77)

Kp

L/g(L/Amol)

(0.28)

(8.33)

(7.14)

(4.03)

(0.09)

(0.278)

(0.01)

(0.93)

0.0192

0.0123

6.92

(1.41)

(0.01)

(1.92)

0.0071

0.0066

4.96

(0.95)

(0.56)

0.88

0.28

11.67

2.5

0.89

3.44

(0.91)

3.21

1.23

0.3

0.06

1.3–1.48

(2.5)

—

(1)

(1.25)

(1.4)

—

—

—

R

L/gCellulose

1.29

50

30

10.6

0.043

0.30

0.011

0.69

0.99

0.53

0.436

0.24

0.011

1.0

0.039

0.033

0.037

0.246

0.093

0.111

0.025

0.303

0.007

0.094

0.014

0.28

0.178

0.079

0.029

0.073

0.2

1.35

—

2.1

8

0.28

(2.33)

(9.87)

Reference

Bothwelletal.,1997

Reinikainenetal.,1995b

Srisodsuketal.,1993

Tommeetal.,1995b

Bothwelletal.,1997

Stahlbergetal.,1991

Tommeetal.,1990

Medveetal.,1997

KimandHong,2000

KimandHong,2000

Beldmanetal.,1987

Nidetzkyetal.,1994c

Tommeetal.,1990

Medveetal.,1997

KimandHong,2000

KimandHong,2000

Beldmanetal.,1987

Nidetzkyetal.,1994

Nidetzkyetal.,1994

Beldmanetal.,1987

Beldmanetal.,1987

Beldmanetal.,1987

Beldmanetal.,1987

Beldmanetal.,1987

Beldmanetal.,1987

Nidetzkyetal.,1994

Ooshimaetal.,1983

Leeetal.,1982

Luetal.,2002

Leeetal.,1982

Beltrameetal.,1982

Moragetal.,1995

Moragetal.,1995

Goldsteinetal.,1993

Goldsteinetal.,1993

Goldsteinetal.,1993

Ketaevaetal.,2001

Ketaevaetal.,2001

Ongetal.,1993

Ongetal.,1993

Ongetal.,1993

Bothwelletal.,1997

Bothwelletal.,1997

Palonenetal.,1999

Palonenetal.,1999

BMCC

BMCC

BMCC

BMCC

Avicel

Avicel

Avicel

Avicel

Avicel

Avicel

Avicel

FilterPaper

Avicel

Avicel

Avicel

Avicel

Avicel

FilterPaper

FilterPaper

Avicel

Avicel

Avicel

Avicel

Avicel

Avicel

FilterPaper

Avicel

Avicel

Avicel

PSAC

Cotton

Avicel

PSAC

Avicel

.

ose

PASC

BMCC

PASC

Avicel

BMCC

BMCC

Avicel

BMCC

BMCC

22

22

22

50

50

22

22

(0.124)

(0.182)

2.05

0.322

1.5

1.0

.,absorbentcotton;Fbcellulose,fibrouscotton.

isthecaseformostcellulosicsubstrates,itislikelythat

somecellulaseisentrappedinpores,resultinginlowerhy-

drolysisrates.

SpatialAnalysisofAdsorptionandInferred

AccessibilityofCellulose

Analysisofadsorptioninspatialtermsisaprerequisitefor

understandingcellulosehydrolysisatamechanisticlevel,

andalsoprovidesapotentiallypowerfulapproachtoeval-

aoc-

cupiedbyanadsorbedcellulasemoleculeismuchlarger

thattheareaoftherepeatingcellobioselattice(shownin

Fig.1b)forallcellulasesforwhichinformationisavailable.

Asaresult,thenumberofcellulasemoleculesthatcanbind

toacellulosesurfaceisingeneralsubstantiallysmallerthan

thenumberofaccessiblecellobioselatticesonthatsurface.

Adsorptionofcellulaseexhibitsapreferenceforthe110

face(Fig.1b)CBH1(Chanzyetal.,1984;

Lehtioetal.,2003)llulases(Gilkesetal.,

1992).Itseemsreasonabletohypothesizethatthisisgen-

erallytruesincethisisthefaceonwhichh-glucosidicbonds

areaccessiblebycellulase.

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS805

icactivitiesofTrichodermacellulasecomponentsoninsolublecellulosesubstrates.

Strain

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

viride

viride

reesei

reesei

reesei

reesei

reesei

reesei

reesei

viride

reesei

reesei

reesei

reesei

reesei

reesei

reesei

reesei

viride

viride

reesei

reesei

Enzyme

CBH

CBH

CBH

CBH1

CBH1

CBH1

CBH1

CBH1

CBH1

CBHIII(CBH1)

CBH2

CBH2

CBH2

CBH2

EG

EG

EG1

EG1

EG3(EG1)

EG3(likeEG1)

EG1

EG1

Temp.(jC)

40

40

50

50

50

40

50

45

40

30

50

40

50

50

50

50

45

40

40

30

50

50

Specificactivity(substrate)(AmolGE/mg/min)

0.42(Av)

0.53–1.0(AC)

0.08(FP)

0.014(Av),0.039(AC)

0.22(FP)

0.0175(Av)

0.065(Av)*

0.04(Av),0.6(AC)

0.012(Av)*,0.0046(FP)*

0.019(Av),0.03(AC)

0.36(FP)

0.0391(Av)

0.027(Av),0.052(AC)

0.065(Av)*

0.18(FP)

3.6(AC)

0.17(Av),26(AC)

0.0046(Av)*,0.0023(FP)*

0.13(Av),9.9(AC)

0.196(Av),0.45(AC)

0.045(Av)*

1.20(FP)

Reference

BerghemandPettersson,1973

GumandBrown,1977;GritzaliandBrown,1978

Ryuetal.,1984

Tommeetal.,1988

Nidetzkyetal.,1994c

vanTilbeurghetal.,1984

Bakeretal.,1998

Shoemaker,1983

Henrissatetal.,1985

Beldmanetal.,1985

Nidetzkyetal.,1994c

vanTilbeurghetal.,1984

Tommeetal.,1988

Bakeretal.,1998

Ryuetal.,1984

Niku-Paavolaetal.,1985

Shoemaker,1983

Henrissatetal.,1985

GritzaliandBrown,1978;Shoemaker,1978

Beldmanetal.,1985

Bakeretal.,1998

Nidetzkyetal.,1994c

*Long

incubationtime.

Gilkesetal.(1992)definedparametersconsistentwitha

spatialinterpretationofadsorptionandincorporatedthese

parametersintoamodifiedLangmuirequation:

E

a

¼

N

0

K

P

0

E

f

1þaK

P

0

E

f

ð9Þ

whereN

0

=Amolaccessiblecellobioselattices/gcellulose,

a=cellobioselatticesoccupied/boundcellulasemolecule,

K

P

V=K

P

/a.

Itmaybenotedthatthecellobioselatticesoccupied/

boundcellulasemolecule,a,maybecalculatedfrom:

a¼N

0

=A

max

ð10Þ

Foracellulasewithagivenvalueofa,thesurfacearea

accessibletothatcellulase(AS,m

2

/g)maybecalculated

fromthemaximumadsorptioncapacityasfollows:

AS¼A

max

N

A

aA

G2

ð11Þ

whereN

A

=Avogadro’sconstant(6.023

Â

10

23

molecules/

mol),A

G2

=areaofthecellobioselattice(0.53

Â

1.04nm=

5.512

Â

10

À19

,m

2

;GardnerandBlackwell,1974a).

ThevalueofASisdependentonthevalueofa,which

willvarydependingonwhichenzymeisunderconsid-

ecellulosesubstrates,thefractionof

h-glucosidicbondsaccessibletocellulaserelativetothe

totalnumberofglucosidicbonds(F

a

)isdefinedas:

F

a

¼2aA

max

MW

anhydroglucose

whereMW

anhydroglucose

=162g/molanhydroglucose.

ð12Þ

BMCChasbeenusedinmoststudiesaimedatdeter-

miningparametervaluesforspatialanalysisofadsorption.

ThisislikelybecausethegeometryofBMCCiswelles-

tablished,incontrasttomostothercellulosicsubstrates.

Inparticular,BMCCexistsasamicrofiberribbonwitha

crosssectionof15

Â

40nm,inwhichthenarrowerofthe

nacelluloseden-

sityof1.5–1.63g/cm

3

,Gilkesetal.(1992)andReinikainen

etal.(1995b)estimatedN

0

forBMCCat93–100Amolcel-

lobioselattice/g.

Atthistime,thelargestbodyofinformationrelevantto

.

,Gilkesetal.(1992)estimatevaluesof32.9,

39.2,and27.9forCenA,thecellulosebindingdomain

ofCenA,andCex,alyticdomainof

CBH1isbelievedtooccupyabout48cellobioselatticeson

atotallyanisotropicsurface(Sildetal.,1996),basedon

structuralinformationinferredfromX-raycrystallography

(Divneetal.,1994).is

thoughttooccupyabout10cellobioselatticesbasedon

nuclearmagneticresonancedata(Kraulisetal.,1989;

Reinikainenetal.,1995b).Reinikainenetal.(1995b)re-

portedarangeofvaluesforA

max

forCBH1bindingto

BMCC,fromwhichvaluesofafrom15–40canbecal-

culatedusingEq.[10].Theseauthorsestimateavalueof

about40fora,whichisveryclosetothevalueof38.7

estimatedbyTommeetal.(1995b)andisintermediate

betweenthesizeofthecatalyticdomainandtheCBM.

SincebindingofCBHIoccursprimarilytothereactive

faceofBMCC(Chanzyetal.,1984;Gilkesetal.,1992;

Lehtioetal.,2003),thevalueofamayalsobeestimated

fromtheratioofthereactivesurfaceareatototalsurface

806BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

area,15/(15+40)=nthisvalue,aforBMCC

canbecalculatedasfollows:

a¼N

0

=A

max

¼0:27ÃS=ðA

max

ÃA

G2

ÃN

A

Þð13Þ

whereSisthetotalexternalsurfaceareaofBMCCfrom

itsgeometricshape=1kgBMCC/(1.5–1.63

Â

10

3

kg/m

3

)/

(15

Â

10

À9

m

*

40

Â

10

À9

m)

*

2

*

(15+40)

Â

10

À9

m=

122–112m

2

max

=6AmolCBH1/g

BMCC(Reinikainenetal.,1995b),a=15.3–16.7cello-

portanttonote

thattheinferredvalueofaisinfluencedbyexperimen-

talconditionssuchastemperatureandionicstrength

(Reinikainenetal.,1995b).

BasedonarepresentativeA

max

valueof4.6Amol/gfor

CBH1adsorptiontoBMCCat50jC(TableII)andain

therangeof15–40,ASvaluesforBMCCof23–61m

2

/g

maybecalculatedusingEq.[11].Thisvaluecorresponds

to18–50%ofthetotalexternalsurfaceareaoftheMBCC

ribbon(15m

2

/g).Regardlessoftheavalue,itappearsthat

cellulasedoesnotadsorbtoasignificantfractionofthe

externalsurfaceofBMCC.

ForAvicel(FMCPH105),0.48Amol/gisarepresentative

A

max

valueforCBH1adsorptionat50jC(TableII),from

whichtheAS

CBH1

ofAvicelisfoundtobe6.4m

2

/gusing

Eq.[10]

CBH1

val-

ueofAvicelPH105ismuchlargerthantheexternalsur-

facearea(0.3m

2

/g;Weimeretal.,1990),indicatingthat

>

f

95%r,AS

CBH1

is

muchsmallerthanthetotalsurfaceareaaccessibletoni-

trogen,

f

20m

2

forAvicel(MarshallandSixsmith,1975),

indicativeofthepresenceofextensiveinternalsurfacearea

inporestoosmalltobeaccessedbycellulasemolecules.

ConsiderationofAvicelandBMCCclearlyshowsthatthe

magnitudeofexternal,internal,andgrosssurfacearea,as

wellastherelativeimportanceofthese,isquitedifferent

q.[12]witha=40,F

a

is

foundtobe6.0%forBMCCand0.62%forAvicel.

Availabledatasuggestthattheareaaccessibletocel-

lulaseenzymes,asindicated,forexample,byAS

CBH1

,

tiontothe

10-folddifferenceforAS

CBH1

notedaboveforBMCCas

comparedtoAvicel,

cellulasehavereporteda3-foldhighercellulaseadsorption

capacityforSolkaFlocSW40comparedtoAvicel(Steiner

etal.,1988),anda20-folderhighercapacityforPASC

comparedtoAvicel(Leeetal.,1982;Moragetal.,1995).

AccessibleareaintheorderAvicel

alsosupportedbydatafromtheCBMsisolatedfromC.

fimi(Ongetal.,1993)cellum

(Kataevaetal.,2001).

Forpretreatedlignocellulosicmaterials,adsorptionto

lignintypicallyoccursatthesametimeasadsorptiontocel-

aetal.(1990)estimatedthemaximumad-

cellulasewith

respecttobothcelluloseandligninpresentindilute-acid-

undtheadsorptioncapacity

forcellulose(asdistinctfromlignin)increasedfrom14.1

to80.6mgcellulasepergramcelluloseasthepretreatment

temperatureincreasedfrom180–220jC,whilethecapacity

forlignindecreasedfrom100to12.3mgcellulase/glignin

.(2002),also

cellulase,reported

cellulaseadsorptioncapacitiesof180mg/gcelluloserel-

ativetothecellulosefractionofDouglasfirpreparedby

SO

2

-catalyzedsteamexplosionfollowedbyperoxidetreat-

ment,and95.2mgcellulase/esultssuggest

thattheaccessibilityofcellulosepresentinpretreatedbio-

masscanvarysignificantlyasafunctionofconditions,but

isoftenofamagnitudecomparabletoAvicel.

CELLULOSEHYDROLYSIS

OntheMechanismofCelluloseHydrolysis

(NoncomplexedSystems)

BeginningwithReese’soriginalhypothesisfortheaction

ofC1(Reeseetal.,1950,1968;Reese,1976),therehave

beensuggestionsthatthemechanismofcellulosehydroly-

sisinvolvesphysicaldisruptionofinsolublecellulosein

ortance

ofsuchdisruption,aswellasthecellulasecomponents

responsibleforit,an(1985)

usedtheterm‘‘amorphogenesis’’todescribephysical

,swelling,segmentation,ordestratificationof

cellulose)thatenhanceenzymatichydrolysisandrender

sed

celluloseaccessibilityduringenzymatichydrolysishasbeen

ncludeH

2

O

2

production

inthepresenceofFeion(Koenigs,1975),ortheshort-

ii(Halliwell

andRiaz,1970),CBH1(Chanzyetal.,1983;

Leeetal.,2000)oritscatalyticdomain(Leeetal.,1996)

ortheCBH2catalyticdomain(Woodwardetal.,1992),

endoglucanase–exoglucanasecomplex(Spreyand

Bochem,1993),HumicolainsolensCBH2(Boissetetal.,

2000),ThermomonosporafuscacellulasesE3andE5

(Walkeretal.,1990,1992),somenoncatalyticdomains

doglucanaseA

(Dinetal.,1991,1994),ashortfiber-generatingpolypep-

koningii(Wangetal.,2003),

fibril-formingprotein(MW=11.4kD)(Bankaetal.,1998),

proteincalledswollenin(MW=

49kD)(Saloheimoetal.,2002).

Itiswidelyobservedthattheheterogeneousstructureof

cellulosegivesrisetoarapiddecreaseinrateashydrolysis

proceeds,evenwhentheeffectsofcellulasedeactivation

andproductinhibitionaretakenintoaccount(Zhangetal.,

1999;Valjamaeetal.,1999).Explainingthisobservationat

amechanisticlevelisanoutstandingissue,withimportant

ghverylittle

workhasbeendoneinvolvingdetailedcharacterization,it

wouldseemlogicaltoexpectthatthedecliningreactivity

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS807

ofresidualcelluloseduringenzymatichydrolysisisare-

sultoffactorssuchaslesssurfaceareaandfeweraccessi-

blechainendsand/oradsorptionofinactivecellulaseon

thesurfaceofcellulose(orlignocellulose)particleswhich

roscopiclevel,boththe

accessibleareaofcellulose(basedontheBETassay;Fan

etal.,1980)andcellulaseadsorptivecapacity(Ooshima

etal.,1983)pergramcellulosehavebeenreportedto

ulatethat

theavailabilityofglucanandchainendspergrammay

roscopiclevel,the

CBH1disruptsfibers,resultinginmoresurface

area(Leeetal.,1996),whileEGIIappearstosmoothfiber

surface,resultinginlesssurfacearea(Leeetal.,2000).

Freshadditionofsubstratescanstimulatemoresoluble

sugarrelease(Carrardetal.,2000),alsoindicatingtheloss

ofcellulosereactivityattheendofhydrolysisand/orin-

creasedreactivityfor‘‘new’’cellulase/celluloseencounters

ascomparedto‘‘old’’encounters.

Whencellulaseenzymesystemsactinvitrooninsoluble

cellulosicsubstrates,threeprocessesoccursimultaneously:

1)chemicalandphysicalchangesintheresidual(notyet

solubilized)solid-phasecellulose;2)primaryhydrolysis,

involvingthereleaseofsolubleintermediatesfromthe

surfaceofreactingcellulosemolecules;and3)secondary

hydrolysis,involvinghydrolysisofsolubleintermediatesto

lowermolecularweightintermediates,andultimatelyto

glucose,alchangesinresidual

cellulosearemanifestedaschangesintheDPandchainend

ucanaseincreasestheconcentration

ofchainendsandsignificantlydecreasesDPbyattacking

canases

shortenDPincrementallyandonlyoccasionallydecrease

,endoglucanaseac-

tivityisthoughttobeprimarilyresponsibleforchemical

changesinsolid-phasecellulosethatoccuroverthecourse

ofhydrolysis,butplaysaminorroleinsolubilizationrel-

ativetoexoglucanase,whileexoglucanaseactivityis

thoughttobeprimarilyresponsibleforsolubilizationbut

playsaminorroleinchangingthechemicalpropertiesof

alchangesinresidualcelluloseare

manifestedaschangesinaccessiblesurfaceareaduetogeo-

metricalchangesresultingfromtheconsumptionorenlarge-

mentofaccessiblesurfaceofcelluloseduetoprogressive

eculativelyatpresent,componentsof

cellulaseenzymesystemsmaymakeadditionalsurfacearea

availablebymechanismsotherthanhydrolysisperse.

Sincetherateofsecondaryhydrolysisismuchfasterthan

therateofprimaryhydrolysis,itispossible—althoughat

thispointspeculative—thatsolublecellodextrinscouldac-

countforasignificantfractionoftheimmediateproducts

enzy-

matichydrolysis,cellodextrinswithDP>4arepresent

inthesolidphaseassociatedwithcrystallinecellulose

(Kleman-Leyeretal.,1994,1996;Srisodsuketal.,1998;Stal-

brandetal.,1998),andithasbeensuggestedthatthis

associationimpedesreleaseofsuchcellodextrinstoso-

r,cellodextrinswithDP>4arenotfound

associatedwithamorphouscellulose(Stalbrandetal.,

1998).Thus,enzymatichydrolysisofcellodextrinsof

length4–6associatedwiththesolidphasemaybean

importantpartoftheoverallsolubilizationprocessfor

crystallinesubstrates,butnotforamorphoussubstrates.

Mostoftheavailabledataoncellulosehydrolysiscon-

cernstherateofsolubilization(process2)above,oftenbased

onreleaseofreducingsugarsorsolubleglucoseequiva-

pinion,bettercharacterizationofchemical

andphysicalchangesassociatedwithresidualcelluloseas

wellassecondaryhydrolysisarepromisingareasofinquiry

inordertoimprovefundamentalunderstandingofcellu-

losehydrolysis.

TrichodermareeseiCellulaseSystem

CellulasesofthegenusTrichodermahavereceivedin-

tensiveattentiondueinsignificantparttothehighlevels

dermavirideisavalidspecies

aggregate,whichisusedforallunknownTrichodermaspe-

cies;aredevelopedfromasingleisolate

(QM6a),namedinrecognitionofthepioneeringcontribu-

mmercialcellulasesare

producedfromTrichodermaspp.,withafewalsoproduced

byAspergillusniger(Esterbaueretal.,1991;Nievesetal.,

1998).Thereaderisreferredtorecentcomprehensivere-

viewsthataddressfeaturesofnoncomplexedcellulase/

hemicellulasesystemsproducedbyorganismsotherthan

(BhatandBhat,1997;Brodaetal.,1996;Ito,1997;

ShallomandShoham,2003;Singhetal.,2003;Subrama-

niyanandPrema,2000;Tommeetal.,1995a;Warren,1996;

Wilson,2004).

cellulasemixtureconsistsofmany

ttwocellobiohydro-

lases(CBH1-2),fiveendoglucanases(EG1–5),h-glucosi-

dases,andhemicellulaseshavebeenidentifiedby2D

electrophoresis(Vinzantetal.,2001).CBH1,CBH2,and

cellu-

istichypothesisofenzymatichydrolysisforcellulose

cellulase.

808BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

lasesystem,representing60F5%,20F6%,and12F3%

oftotalcellulaseprotein,respectively(Goyaletal.,1991;

GritzaliandBrown,1978;Knowlesetal.,1987;Kyriacou

etal.,1987;NidetzkyandClaeyssens,1994).Reconstituted

cellulasepreparationsbasedonpurifiedcomponentsin

theseproportionsexhibitspecificactivityequivalentto

unfractionatedpreparations(Bakeretal.,1998).Thestruc-

tureofCBH1,CBH2,andEG1featuresacatalyticdomain

andacellulose-bindingdomainconnectedbyaglycolysated

peptidelinker(Gilkesetal.,1991;LeeandBrown,1997;

LinderandTeeri,1997).

ThecatalyticdomainstructuresofCBH1andCBH2are

entirelydifferentbutbothfeaturetunnel-shapedstructures

2,twowell-ordered

˚

longtunneladjacenttoana/h-barrelloopsforma20A

structure(Rouvinenetal.,1990).InCBH1,foursurface

loopsformatunnelof50Aadjacenttoah-sandwich

structure(Divneetal.,1993,1994).Thetunnel-shaped

topologyofCBH1andCBH2allowsforastructuralinter-

-

alyticsitesofbothcellobiohydrolasesarewithinthetunnel

neartheoutlet,sothath-glucosidicbondsarecleaved

byretaining(CBH1)orinverting(CBH2)mechanisms.

Structuralanalyses,asopposedtomeasurementofhy-

drolysisproducts,providesdirectevidencethatcellobiose

istheprimaryproductofhydrolysismediatedbyCBH1and

CBH2(Divneetal.,1993,1994;Daviesetal.,1997).The

CBH1andCBH2cancleaveseveralbondsfol-

lowingasingleadsorptioneventbeforethedissociationof

theenzymesubstratecomplex(Imaietal.,1998;Teerietal.,

1998a,b;Valjamaeetal.,1998).Therefore,theactionof

CBH1andCBH2resultinagradualdecreaseinthedegree

ofpolymerization(DP)ofcellulose(Kleman-Leyeretal.,

1992,1996;Srisodsuketal.,1998).Cellobiohydrolaseac-

tivityisoftenmeasuredbyreducingsugarreleasefrom

Avicel,oftencalled‘‘Avicelase’’isagood

substrateformeasuringexoglucanaseactivity,althoughnot

exclusively,becauseithasthehighestratioofchainends

toaccessibleinternalh-glucosidicbondsamongmodelcel-

lulosicsubstrates(seeTableIandAdsorption,above).

EG1andCBH1havesignificanthomology(45%identity,

Penttilaetal.,1986),belongtothesamefamily(Cel7),and

ivesiteofEG1isa

grooveratherthanatunnel(Henrikssonetal.,1996),

allowingglucanchainstobecleavedrandomlytotwo

shorterchainsresultinginarapiddecreaseinDP(Kleman-

Leyeretal.,1992,1994;Srisodsuketal.,1998;Whitaker,

1957;Selby,1961;WoodandMcCrae,1978).Endogluca-

naseactivityismostoftenmeasuredbasedontherateof

changeoftheviscosityofasolublecellulosederivativesuch

ascarboxymethylcellulose(CMC)(Milleretal.,1960;

WoodandMcCrae,1972).ItmaybenotedthatCMCase

activityhasbeenshowntocorrelatepoorlywiththeability

tohydrolyzeinsolublecelluloseevenforpurifiedendo-

glucanases(Himmeletal.,1993;Klyosov,1988;Klyosov,

1990).endoglucanasesobtained

byShoemakerandBrown(1978),theoneexhibitingthe

highestratesofAvicelhydrolysishadthelowestCMCase

v(1990)clearlypointedoutthatthespe-

cificendoglucanaseactivitiesfrommanymicroorganisms

measuredonCMCdonotcorrelatewithactivitiesagainst

insolublecellulose.

Itisapparentthatthedivisionintoendo-andexogluca-

nasesisinmanycasesnotabsolute(Barretal.,1996;Irwin

etal.,1993;HenrissatandDavies,1997;Teeri,1997;Teeri

etal.,1998a,b).Irwinetal.(1993)documentedaprocessive

doglucanaseactivity

CBH2(EnariandNiku-

Paavolar,1987;Kyriacouetal.,1987)andCBH1(Schmid

andWandrey,1990),nsCBH2

(Boissetetal.,2000).Stahlbergetal.(1993)concludedthat

beensuggestedthatexoglucanasecouldexhibitsome

endoglucanaseactivityduetotemporaryconformational

changesofloopsonthetunnelstructurethatexposetheir

activesites(Warren,1996;ZhangandWilson,1997).This

hypothesisissupportedbytheobservationthatdisruption

oftheloopscomprisingthetunnelofexoglucanaseresults

inincreasedendoglucanaseactivityaswellashigherk

cat

(Kleywegtetal.,1997;Meinkeetal.,1995).Inaddition,it

maybeobservedthatCBH2containsfewerloopsalongthe

catalytictunnelandexhibitsgreaterendoglucanaseactivity

relativetoCBH1.

RemovaloftheCBMofTrichodermacellulasesresults

inaseveral-foldreductionintherateofhydrolysisof

insolublecellulosebuthaslittleeffectonhydrolysisof

solublesubstrates(Glikesetal.,1988;Irwinetal.,1994;

Reinikainenetal.,1992;Srisodsuketal.,1997;Stahlbeg

etal.,1993;Tommeetal.,1988).CBMsbe-

longtofamily1(CBM1),characterizedbyasmallwedge-

shapedfoldfeaturingacellulosebindingsurfacewith

threeexposedaromaticresidues(Hoffrenetal.,1995;

Lehtioetal.,2003;Kraulisetal.,1989).Thesearomatic

residuesarethoughttobecriticalforthebindingofa

cingofthethree

aromaticresiduescoincideswiththespacingofeverysec-

ondglucoseringonaglucanchain,andithasbeenpos-

tulatedthatthearomaticaminoacidsoftheCBMsformvan

derWaalsinteractionsandaromaticringpolarizationin-

teractionswiththepyranoseringsonthesurfaceofcel-

lulose(Lehtioetal.,2003).

EG1,CBH1,andCBH2on

variousinsolublecellulosicsubstratesarepresentedin

aexhibitsubstantialvariabilityevenfor

apparentlysimilarenzymepreparationsandsubstrates.

Notwithstandingthisvariation,thedatasupportthe

followingobservations:1)someearlyvaluesforexogluca-

naseandendoglucanaseactivitywerehigherthanvalues

reportedmorerecently,possiblyduetouseoflowerpurity

enzymepreparationsinearlierstudies;2)ratesmeasuredat

longerreactiontimesaremuchslowerthanthoseatshorter

times,whichappearsdueatleastinparttocellulose

heterogeneity(Klyosov,1990;Valjamaeetal.,1998;Zhang

etal.,1999);and3)therateofgenerationofsoluble

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS809

reducingsugarsbyEG1relativetoCBH1isJ1for

amorphouscellulose,V1forAvicel,andV1forBMCCand

ativelylowrateofreducingsugarrelease

exhibitedbyEG1oncrystallinecelluloseisconsistentwith

mostofthereducingendsgeneratedbyendoglucanase

activityremaininginthesolidphase,anddoesnotnec-

essarilyimplyalowerrateofh-glucosidicbondcleavage.

ThespecificactivityofCBH2hasbeenfoundtobenearly

twicethatofCBH1inmost(Henrissatetal.,1985;Medve

etal.,1994;Nidetzkyetal.,1994c;Tommeetal.,1988)but

notall(Bakeretal.,1998)studies.

EG1,CBH1,CBH2,andh-glucosidaseonsolubleglucans.

Whilevariabilityisagainevident,thefollowingtrends

maybenoted:1)therateofreactioncatalyzedbyexoglu-

canaseandendoglucanaseincreaseswithincreasingsolu-

blesubstratechainlength,whereasdecreasingactivityof

h-glucosidasewithincreasingchainlengthisobservedin

thesinglestudyforwhichcomparativedataareavailable;

2)significantlyhigherratesareobservedforEG1ascom-

ingdatainTablesIIIand

IV,itmaybeseenthatthespecificactivitiesofexoglu-

canasesandendoglucanasesactingonsolublesubstrates

arehigherbyatleastanorderofmagnitudethanactivities

,therateofprimaryhydroly-

sis(fromcellulosetosolubleglucans)ismuchslowerthan

secondaryhydrolysis(fromsolubleglucanstocellobiose

andglucose).

Synergism

Synergismissaidtooccurwhentheactivityexhibitedby

mixturesofcomponentsisgreaterthanthesumoftheac-

tivityofthesecomponentsevaluatedseparately(Walker

andWilson,1991;WoodandMcCrae,1979;Woodand

Garcia-Campayo,1990;Woodward,1991).Quantitative

representationoftheextentofsynergismisusuallyex-

pressedintermsofa‘‘degreeofsynergism’’(DS)—equal

totheratiooftheactivityexhibitedbymixturesofcom-

ponentsdividedbythesumoftheactivitiesofseparate

fsynergismproposedinthecellulose

hydrolysisliteratureinclude:1)endoglucanaseandexoglu-

canase;2)exoglucanaseandexoglucanase(Fagerstamand

Pettersson,1980;Tommeetal.,1988,1990;Woodand

McCrae,1986;WoodandGarcia-Campayo,1990);3)en-

doglucanaseandendoglucanase(Mansfieldetal.,1998;

Tukaetal.,1992;Walkeretal.,1992);4)exoglucanaseor

endoglucanaseandh-glucosidase,whichreducesinhibition

bycellobiose(Lamedetal.,1991;Woodward,1991);5)

intramolecularsynergybetweencatalyticdomainandCBM

(Dinetal.,1994)ortwocatalyticdomains(Riedeland

Bronnenmeier,1998;Te’oetal.,1995;Warrenetal.,1987;

Zverlovetal.,1998);6)cellulose-enzyme-microbe(CEM)

synergism(Lyndetal.,2002);and7)aproximitysynergism

duetoformationofcellulasecomplexes(Fierobeetal.,

2001,2002;Mandels,1985;Schwarz,2001).Notall

synergiesarenecessarilyoperativeinanygivensituation.

Forexample,synergismbetweenthecatalyticdomainand

cottonfibersbut

notonBMCC(Dinetal.,1994).Cell-enzyme-microbe

synergismhasbeenpostulatedforsystemsinwhicha

metabolicallyactivecelltogetherwithadheredcellulase

bindstocellulose(Lyndetal.,2002),buthasnotbeen

quantitativelyevaluated.

Synergismbetweenendoglucanasesandexoglucanases

isthemostwidelystudiedtypeofsynergyandisamong

themostquantitativelyimportantforhydrolysisofcrystal-

ninTableV,thehighestreported

DSvaluesareforBC(5–10)andcotton(3.9–7.6).Less

pronouncedbutstillsignificantsynergismisexhibited

forAvicel(DS1.4–4.9),whilethesmallestsynergistic

effects(DS0.7–1.8)havebeenreportedforphosphoric

acid-swollenandotheracid-treatedamorphouscelluloses.

DPappearstoplayanimportantandquitepossiblydom-

icactivityofTrichodermacellulasecomponentsonsolublesubstrates.

Specificactivity(Amolbond-breaking/mg/min)SubstrateDP

Strain

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

T.

viride

reesei

reesei

reesei

reesei

reesei

reesei

reesei

reesei

viride

viride

viride

reesie

reesie

Enzyme

CBH

CBH

CBH1

CBH1

CBH1

CBH2

CBH2

EG

EG1

EGIII(EG1)

BG

BG

BG1

BG1

BG2

Temp.(jC)

39

50

25

25

50

27

27

50

25

40

40

50

45

50

50

G2G3

0.013

0.1

0.23

0.013

0.056

0.074

11

24.4

19

G4

2.7

G5G6G7Reference

Lietal.,1965

Hsuetal.,1980

vanTilbeurghetal.,1982

Claeyssensetal.,1989

Nidetzkyetal.,1994a

Koivulaetal.,1998&2002

Harjunpaaetal.,1996

Niku-Paavolaetal.,1985

Claeyssensetal.,1989

ShoemakerandBrown,1978

BerghemandPettersson,1974

Gongetal.,1977

Shoemakeretal.,1983

Chenetal.,1992

0.41

3.78

2.86

0.49

1.01

0.74

0.98

12.9

11.0

17.5

0.81

66.7

33

58

31.4

43.5

9.8

810BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

mreporteddegreeofexo/endosynergismforvariousmodelsubstrates.*

Maximumdegreeofsynergism

Strain

rarium

cellum

Humicolainsolens

i

ulentum

ii

Enzymecombination

Exo/Endo

Exo/Endo

(CBH1+CBH2)/EG1

CBH/EG

Exo/Endo

CBH/EG

CBH1/EG1

(CBH1+CBH2)/EG1

CBH1,CBH2/EG1

CBH1/EG1

CBH1/EG1

CBH1/EG2

CBH1/EG2

CBH1/EG1

CBH1/EG2

Exo/Endo

Exo/Endo

<2

0.7(a)

2to5

1.4–2.1(Av)

2.5(Av)

4.9(Av)

2.9(FP)

6.8(BC)

1.8(AC)

1(AC)

3.9(ct)

7.6(ct)

1.3–1.4(Av)

1.5–2

2

(Av)

f

2(Av)

f

1.5–2(b)

4.1(c)

2.1(Av)

2.2(Av),2.5(d)

1.2(e)

1.8(AC)1.7–3.5(Av)

2.1(Av)

>5Reference

Riedeletal.,1997

Tukaetal.,1992

Boissetetal.,2001

Sadana,1985

Streameretal.,1975

WoodandMcCrae,1978

Medveetal.,1998

Woodwardetal.,1988a

Bakeretal.,1998

Srisodsuketal.,1998

Valjamaeetal.,1999

Hoshinoetal.,1997

Valjamaeetal.,1999

Henrissatetal.,1985

Samejimaetal.,1998

Beldmanetal.,1988

Kimetal.,1992

1.7(c)

1.5(AC)

7.8(BC)

3.2(ct)

3(FP)

5(f)

f

6(BC)

5(BC)

10(BC)

*Av,Avicel;FP,filterpaper;ct,cotton;BC,bacterialcellulose;AC,amorphouscellulose;a,acid-treatedAvicel;b,acid-treatedcotton;c,acid-treated

BC;d,homogenizedAvicel;e,acid-treatedBC;f,SO

2

-treatedBC.

inantroleindeterminingwhethertheDSislargeorsmall.

Insupportofthisinterpretation,wenotethattheabove-

listedorderingofcellulosicsubstrateswithrespecttoDS

isthesameastheorderingwithrespecttodegreeof

polymerization(seeCellulose,above)andisalsoconsist-

entwithmodelingresults(OkazakiandMoo-Young,1978).

Higherendo-exosynergyhasbeenreportedforsubstrates

thathavebeentreatedtoreduceCrI,forexample,homo-

(Henrissatetal.,1985)and

loc(Fanetal.,1981).

However,Hoshinoetal.(1997)observedhigherDSas

dpreviouslyinourdiscussionof

CrI,itisdifficulttoattributeobservedchangestoCrIbased

onworkinvolvingtreatmentsthatalsochangeaccessible

surfacearea.

Inadditiontosubstrateproperties,experimentalcon-

beenreportedthatendo-exosynergyincreaseswithan

increaseinenzymeloadingbelowsaturationbutdecreases

withoversaturatedenzymeloading(Tukaetal.,1992;

Watsonetal.,2002;Woodwardetal.,1988a,b;Woodward,

1991).Inaddition,suchsynergyisreportedtobegreater

underconditionschosentominimizeinhibitionbysoluble

hydrolysisproductsinsome(Fierobeetal.,2001,2002;

Srisodsuketal.,1998)butnotall(Erikssonetal.,2002;

Medveetal.,1998)studies.

ComparisonofCelluloseandStarchHydrolysisRates

Forthepurposeofunderstandingfactorslimitinghydrolysis

ofcellulosebycellulases,itisinformativetoconsider

dbyseveralauthors

(Mandels,1985;Klyosov,1988),ratesofstarchhydroly-

siscanbeabout100-foldfasterthanhydrolysisratesfor

celluloseunderconditionsanticipatedforindustrialpro-

cessesand/orusingcrystallinemodelsubstrates.

Inadditiontoanydifferenceintheintrinsicreactivityof

h-linkedglucansascomparedtoa-linkedglucans,three

propertiesofcelluloseandstarchinfluencetheirhydrolysis

rates:1)thefractionofbondsaccessibleforinsoluble

substrates,2)theavailabilityofchainendsforinsoluble

substrates,and3)thesolubilitiesofhydrolysisproducts.

Thefractionofaccessibleglucose-glucosebonds,F

a

,ranges

fromlessthan0.002to0.12forcellulose(basedonEq.[11]

witha=40).Thisis8–500-foldlowerthanforsoluble

starch(F

a

=1),solublemalto-oligosaccharides,orsoluble

cellulosederivativeslikeCMC(F

a

=1),andis5–200-fold

lowerthaninsolublestarch(F

a

=

f

0.2;Fujiietal.,1981).

Thelowfractionofaccessiblebondsisthoughttolimitrates

,per

unitmass)islowerforcellulosethanforstarchbecauseof

thehighDPofcelluloseaswellastheincidenceof

lulose,theratioofglucosylunits

perchainendisequaltotheDPandrangesfrom300–2,000

(seeDegreeofPolymerization,above).Forstarch,which

exhibitsbranchesevery17to26glucoseunits(Bertoldoand

Antranikian,2002;Bueleonetal.,1998),eachbranchgives

risetoanewchainendandtheratioofglucosylunitsto

chainendsisthus

f

osehydrolysisratesare

thoughttolimitedbytheavailabilityofchainendsfor

cellobiohydrolase(Schulein,2000;Valjamaeetal.,2001;

ZhangandWilson,1997),andchain-endlimitationhasalso

beenproposedfortheactionofglucoamylaseonmalto-

saccharides(MazurandNakatani,1993).Whereascello-

dextrinsareessentiallyinsolubleatDP>6–10(Miller,

1963;Pereiraetal.,1988;ZhangandLynd,2003),malto-

oligosaccharidesaresolubleatDPupto60(Johnetal.,

ZHANGANDLYND:NONCOMPLEXEDCELLULASESYSTEMS811

1982).Thisdifferencecanbeattributedtotheplanarlin-

earstructureofcellodextrinsascomparedtothehelical

ultofthesedifferences

inthesolubilityofhydrolysisproduct,manyfewerbond

cleavagesneedoccurbeforesolublehydrolysisproducts

aregeneratedfromstarchascomparedtocellulose,anda

correspondinglylargerfractionofbondscanbecleavedby

enzymesactingintheliquidratherthansolidphasefor

ary,mostcrys-

tallinecellulosicsubstratesexhibitaz10-foldsmaller

fractionofaccessiblebonds,az10-foldsmallerfrequency

ofchainends,andamuchsmallerfractionofbondscleaved

inthesolublephaseduringenzymatichydrolysisascom-

paredtostarch.

Incontrasttothemarkedlydifferentpropertiesof

celluloseandstarchassubstratesforenzymatichydrolysis,

availabledatasuggeststhatthespecificrateofsolubiliza-

tionexhibitedbyexo-actingsaccharolyticenzymesappears

,thespecific

activityofCBH2oncellohexaoseat27jC(k

cat

=14s

À1

;

Harjunpaaetal.,1996;Koivulaetal.,1998,2002)isquite

comparabletothatforAspergillusawamoriglucoamylase

onmaltohexaose(G

6

)at45jC(49s

À1

;Fierobeetal.,1998),

particularlywhenthedifferentmeasurementtemperatures

3.5-foldhighervalueofk

cat

observed

forglucoamylaseat45jCrelativetoCBHat27jCisvery

closetowhatwouldbeexpectedbasedonthewidely

observedtrendofdoubledactivityforevery10jCincrease

intemperature(GodfreyandWest,1996).

Inlightoftheseconsiderations,itappearstousthatthe

largedifferenceintherelativehydrolysisratesofcellulose

andstarchisdueprimarilytodifferencesinsubstratechar-

acteristicsratherthantoh-linkedglucosidicbondsbeing

intrinsicallymoredifficulttohydrolyzethana-linked

tentwiththishypothesis,were-

centlyfoundthattheinitialhydrolysisrateofPASCismore

than100-foldhigherthanthatofAvicel.

QUANTITATIVEMODELS

Aclassificationschemeforquantitativemodelsofenzy-

theterm‘‘nonmechanisticmodels’’formodelsbasedondata

correlationwithoutanexplicitcalculationofadsorbed

uchmodelsmaybeuseful

forcorrelatingdata,theyareunlikelytobereliableunder

conditionsdifferentfromthoseforwhichthecorrelationwas

developed,andtheyhavelimitedutilityfortestingand

featuringadefensible

adsorptionmodelbutwhicharebasedonconcentrationas

theonlyvariabledescribingthestateofthesubstrateand/or

arebasedonasinglecellulosehydrolyzingactivityare

termed‘‘semimechanistic.’’Inparticular,modelsfeaturing

concentrationastheonlysubstratestatevariablearereferred

toas‘‘semimechanisticwithrespecttosubstrate,’’whereas

modelswithasinglecellulosehydrolyzingactivityare

referredtoas‘‘semimechanisticwithrespecttoenzyme.’’

Mostofthehydrolysismodelsproposedtodateforde-

signofindustrialsystemsfallintothecategoryofsemi-

chanisticmodelscanbeuseful

inthecontextofexercisesmotivatedbyincludingthe

minimalinformationnecessaryfordescriptivepurposes.

However,semimechanisticmodelswithrespecttosubstrate

cannotdescribeorlendinsightintobehaviorsdetermined

rly,

semimechanisticmodelswithrespecttoenzymecannotde-

scribeorlendinsightintobehaviorsdeterminedbymultiple

featuringanadsorption

ficationschemeformodelsofenzymaticcellulosehydrolysis.

Modelcategory

Nonmechanistic

Definingfeature&basis

Notbasedonadefensible

adsorptionmodel

Utility

.

Datacorrelation

Limitations

.

Reliabilityunderconditions

differentfromthoseusedto

developthecorrelation

.

Doesnotenhanceunderstanding

.

Datacorrelation

.

Reactordesign

.

Identificationofessentialfeatures

.

UnderstandingatthelevelofSemimechanistic

spect

tosubstrate

spect

toenzyme

Functionallybased

Basedonadefensibleadsorptionmodel

Concentrationtheonly

substratestatevarible

Onesolubilizingactivity

substratefeaturesandmultiple

enzymeactivities

Includesanadsorptionmodel,

substratestatevariablesin

additiontoconcentration,

multiplesolubilizingactivities

.

Testinganddeveloping

understandingatthelevel

ofsubstratefeaturesand

multipleenzymeactivities

.

Identifyingrate-limitingfactors

.

Reactordesign(potentially)

.

Moleculardesign

.

Testinganddevelopingunderstanding

.

Moleculardesign

.

Stateofmodeldevelopment

anddataavailabilitycurrently

limitapplicationtodesign

StructurallybasedBasedonstructuralinformation

pertainingtocellulasecomponents

.

Challengingtodevelop

ofstructure/functionrelationships

meaningfulkineticmodels

basedonstructuralinformation

812BIOTECHNOLOGYANDBIOENGINEERING,VOL.88,NO.7,DECEMBER30,2004

model,substratestatevariablesinadditiontoconcentration,

andmultipleenzymeactivitiesaredenoted‘‘functionally

basedmodels.’’Functionallybasedmodelsareparticularly

usefulfordevelopingandtestingunderstandingatthelevel

ofsubstratefeaturesandmultipleenzymeactivities,in-

cludingidentificationofrate-limitingfactorsandstrategies

unctionallybasedmodels

couldconceivablybeusedforbioreactordesign,application

erlimitationof

functionallybasedmodelsisthattheyprovidelittlebyway

ofguidancerelativetodesignofcellulasesatthemolecular

y,modelsbasedonstructuralfeaturesofcellu-

lasecomponentsandtheirinteractionwiththeirsubstrates

aretermed‘‘structurallybasedmodels.’’Toamuchgreater

extentthanmodelsinothercategories,structurallybased

modelsareusefulformoleculardesignaswellastesting

anddevelopingunderstandingoftherelationshipbetween

tionofmeaningful

kineticmodelsbasedonstructuralmodelscannotbedone

atthistime,andawaitsmajoradvancesinthegeneralfieldof

thatthe

vastmajorityofavailablekineticmodelsdonottakeinto

considerationchangesinhydrolysisrateoverthecourseof

hydrolysis,andthosethatdorepresentsuchchangesusing

empiricallyfittedparametersratherthanmechanistically

basedparameters.

andlignincontentenhancehydrolysis,withspecificsurface

areathemostinfluentialofthestructuralfeatures,followed

ndHoltzapple(2000)reporta

modeltocorrelatemaximumconversioninrelationtore-

siduallignin,crystallinityindex,

authorsfoundthatlignincontentandCrIhavethegreatest

impactonfinalconversion,whereasacetylcontenthada

setal.(1992)alsoattemptedtore-

latemaximumconversionwithCrIanddegreeofdiligni-

fication,andobtainedasimilarconclusionaboutCrIand

lignineffects.

Sattleretal.(1989)developedthefollowingequationto

describefinalfractionalconversionafterenzymatichydrol-

ysisofpretreatedpoplarinrelationtocellulaseloading:

YY

max

½E

本文标签： 参数 作者