Benefits from Tween during enzymic

Benefits from Tween During Enzymic Hydrolysis of Corn StoverWilliam E.Kaar,*Mark T.HoltzappleDepartment of Chemical Engineering,Texas A&M University,College Station,Texas77843-3122;telephone:409-845-9708;fax:409-845-6446; e-mail:mth4500@Received7February1997;accepted27December1997Abstract:Corn stover is a potential substrate for fermen-tation processes.Previous work with corn stover demon-strated that lime pretreatment rendered it digestible by cellulase;however,high sugar yields required very high enzyme loadings.Because cellulase is a significant cost in biomass conversion processes,the present study fo-cused on improving the enzyme efficiency using Tween 20and Tween80;Tween20is slightly more effective than Tween80.The recommended pretreatment condi-tions for the biomass remained unchanged regardless of whether Tween was added during the hydrolysis.The recommended Tween loading was0.15g Tween/g dry biomass.(The critical relationship was the Tween load-ing on the biomass,not the Tween concentration in so-lution.)The72-h enzymic conversion of pretreated corn stover using5FPU cellulase/g dry biomass at50°C with Tween20as part of the medium was0.85g/g for cellu-lose,0.66g/g for xylan,and0.75for total polysaccharide; addition of Tween improved the cellulose,xylan,and total polysaccharide conversions by42,40,and42%, respectively.Kinetic analyses showed that Tween im-proved the enzymic absorption constants,which in-creased the effective hydrolysis rate compared to hy-drolysis without Tween.Furthermore,Tween prevented thermal deactivation of the enzymes,which allows for the kinetic advantage of higher temperature hydrolysis. Ultimate digestion studies showed higher conversions for samples containing Tween,indicating a substrate ef-fect.It appears that Tween improves corn stover hydro-lysis through three effects:enzyme stabilizer,lignocellu-lose disrupter,and enzyme effector.©1998John Wiley& Sons,Inc.Biotechnol Bioeng59:419–427,1998.Keywords:corn stover;Tween;hydrolysis;enzyme;sac-charificationINTRODUCTIONA previous article(Kaar and Holtzapple,in press),showed that lime is an effective pretreatment for the enzymic sac-charification of corn stover;however,high enzyme loadings were required for high conversions.For example,the7-d glucan conversion at50°C using5FPU cellulase/g dry bio-mass was about60%compared to88%when25FPU/g dry biomass was used(Kaar and Holtzapple,in press).An im-portant and expensive input into any potential biomass con-version system is the enzyme loading,which can amount to as much as60%of the process cost(Helle et al.,1993; Nguyen and Saddler,1991;Wilke et al.,1981);hence,en-zyme use should be as low as possible.Clearly,a method to increase enzyme effectiveness would be beneficial.Toward this aim,it has been reported that addition of non-ionic surfactants,particularly Tween species,effectively im-proves the cellulase activity as well as preserves them for recycle(Castanon and Wilke,1981;Helle et al.,1993;Kaya et al.,1995;Park et al.,1992).This article focuses on the benefits derived from adding Tween20and Tween80to the corn stover saccharification hydrolytic medium.Tween20(polyoxyethylene sorbitan monolaurate,CAS #9005-64-5)and Tween80(polyoxyethylene sorbitan mo-nooleate,CAS#9005-65-6)are surfactant polysorbates. Previous reports using Tween additives for enhancing en-zymic saccharification have focused on non-native cellulose and cellulose analogs such as newsprint,microcrystalline cellulose,and carboxymethylcellulose(CMC).Castanon and Wilke(1981)reported a14%increase in glucose yield and more than twice as much recovered enzyme from news-paper saccharification(5%suspension,13FPU/g newsprint, 45°C,48h)when0.02g Tween80/g newsprint was added. These researchers postulated that Tween80assists the de-sorption of enzyme from substrate.Park et al.(1991),also working with newspaper,examined several surfactants and found Tweens to be among the best performers.The rec-ommended Tween80loading(4%suspension,0.75FPU/g newsprint,50°C,22h)was0.0125g Tween80/g newsprint, with little enhancement with loadings greater than0.025g Tween80/g newsprint and no enhancement when loadings exceeded0.25g Tween80/g newsprint.After adding0.16g Tween80/g cellulose,Helle et al.(1993)reported a seven-fold increase in the saccharification of Sigmacell100cel-lulose(2.5%suspension,40FPU/g cellulose,50°C,100h); they concluded that the Tween80disrupted the cellulose structure as well as facilitated enzyme desorption from the substrate.Finally,Kaya et al.(1995)conducted an extensive evaluation of surfactants used during the saccharification of CMC and xylan.Tween80at a loading of0.015g Tween 80/g substrate increased the conversion of CMC(0.5%so-*Present address:The Foxboro Company,10707Haddington Drive,Houston,Texas77043Correspondence to:Mark HoltzappleContract grant sponsors:National Renewable Energy LaboratoryContract grant numbers:DE-AC02-83CH10093;XAW-3-11181-03©1998John Wiley&Sons,C0006-3592/98/040419-09lution,0.77FPU/g CMC,50°C,1h)by50%and improved the xylan(1%suspension,0.59IU/g xylan,50°C,1h)hy-drolysis by about9%.Concerning the mechanism,Kaya et al.(1995)referred to the work of Kim et al.(1991)who concluded that surfactants help prevent cellulase denatur-ation at the air interface by competing with the cellulase for the free surface area.Thus,while the experimental sub-strates and Tween loadings have differed,the unanimous conclusion in the literature has been that,by whatever mechanism,Tween assists the enzymic saccharification of cellulose and cellulose analogs.Based on these previous results,it was suspected that Tween might improve the enzymic saccharification of a native lignocellulose like corn stover.Experiments were conducted to identify the recommended Tween loading and to verify the recommended pretreatment conditions when using Tween as an additive for the enzymic saccharification. Additional tests were performed to show that Tween acts as enzyme stabilizer,lignocellulose disrupter,and enzyme ef-fector.MATERIALS AND METHODSGeneralHydrolyses were performed in parallel with and without Tween.Except for the lime loading study,pretreatment of corn stover samples(−20+80mesh;39.0%glucan,20.1% xylan,2.0%arabinan,21.5%lignin,6.8%ash,and3.9% protein,by weight)was conducted according to previously reported methods(Kaar and Holtzapple,in press)using standard conditions(0.1g Ca(OH)2/g dry biomass,10g H2O/g dry biomass,120°C,5h).Unless otherwise speci-fied,the enzymic saccharification was performed at a solids content of50g biomass/L in500-mL culture flasks at a pH of4.8using procedures described previously(Kaar and Holtzapple,in press).Tween20and Tween80were obtained from Fisher Sci-entific(Pittsburg,PA).The cellulase mixture(Spezyme-CP, from Trichoderma reesei with reported activity91.9FPU/ mL)was provided by the National Renewable Energy Labo-ratory(NREL)and had an assumed specific activity of700 FPU/g enzyme(Holtzapple et al.,1991).The␤-glucosidase (Novozym188)was obtained from Novo Nordisk(Fran-klinton,NC),had a vendor-reported activity of250IU/mL, and was used at a loading of28.4IU/g dry biomass.The typical procedure for conducting the hydrolysis was:trans-fer of pretreated solids to culture flask;addition of citrate buffer and sodium azide solutions;neutralization to pH4.8 with glacial acetic acid;preheating sample for10min in the incubator shaker;and,addition of Tween or an equal vol-ume of water,followed by immediate addition of the en-zymes.(Note:For this enzyme system,there was no inhi-bition by the calcium acetate resulting from the lime neu-tralization.)Post-hydrolysis samples were heated in a boiling water bath for10min to denature the enzymes,centrifuged,and the clear supernatants analyzed by DNS assay(Miller,1959)or by High Performance Liquid Chro-matography(HPLC)(Kaar et al.,1991).(Note:Tween does not affect,nor contribute to,the absorbance measurement in the DNS assay.)Tween Loading StudiesThis experiment was conducted using Tween80.The hy-drolysis was conducted at50°C.Aliquots(4mL)were re-moved after hydrolysis times of0.5,1,1.5,2,3,4,5,6,7, and8d;the supernatants were analyzed by HPLC. Pretreatment StudiesExperiments were repeated to verify that the previously de-termined recommended pretreatment conditions(Kaar and Holtzapple,in press)were also best for samples saccharified in the presence of Tween80.For all cases the samples were pretreated at120°C.In the pretreatment time experiment, the zero-time samples were mixed with the standard loading of0.1g Ca(OH)2/g dry biomass at room temperature and immediately neutralized to pH4.8with acetic acid.For the calcium hydroxide loading evaluation,pretreatments were conducted for5h using lime loadings of0.0,0.025,0.05 0.075,0.1,and0.125g Ca(OH)2/g dry biomass;the inter-mediate loadings of0.055,0.060,0.065,and0.070g Ca(OH)2/g dry biomass were also evaluated with Tween only.The enzymic hydrolysis time was100h at50°C.At the conclusion of the hydrolysis,the samples were analyzed by DNS assay.Enzyme Loading DependenceThe hydrolysis was conducted at50°C with Tween80and 40°C with Tween80or Tween20as additives.For both experiments,4-mL samples were removed at time intervals of1,3,6,10,16,24,36,48,72,and100h and analyzed by DNS assay.Hydrolysis Temperature DependenceThe hydrolyses were conducted for72h at temperatures of 30,40,50,and60°C.Before adding the enzymes,the samples were placed in the shaker for5min to ensure thor-ough mixing and to warm the samples partially to the hy-drolysis temperature.The samples were not allowed to com-pletely reach the designated temperature to avoid heat shocking the enzymes upon their addition.After the5-min warm-up,Tween20,Tween80,or water was added to a pair of samples.The samples were hydro-lyzed in sets of six and the hydrolyzates were analyzed by HPLC.Tween Dependence:Loading or Concentration? Corn stover samples were pretreated in triplicate using3.75, 7.5,11.25,and15.0g of dry biomass per reactor.The420BIOTECHNOLOGY AND BIOENGINEERING,VOL.59,NO.4,AUGUST20,1998samples were prepared for hydrolysis by the previously de-scribed standard methods;the resulting solids concentra-tions in the hydrolytic solutions were25,50,75,and100g dry biomass/L.The pretreated samples were separated into three sets such that each set contained one of each of the four different substrate concentrations.To one set of samples,Tween80 (0.15g Tween/g dry biomass)was added;another set re-ceived aliquots of water equal in volume to the Tween80 added to the first set;and the third set of samples received a combination of water and Tween80such that the25g/L sample received a double loading(0.3g Tween/g biomass) of Tween80;the75g/L sample received a two-thirds load-ing(0.1g Tween/g biomass);and the100g/L sample re-ceived a one-half loading(0.075g Tween/g biomass).The samples were hydrolyzed in the shaker at50°C for100h. Ultimate DigestionThe experiment was conducted in duplicate according to NREL Chemical Analysis and Testing Standard Procedure No.009(Torget,1993)with Tween20and Tween80as additives and distilled water as the control.The following loadings were used(all per gram dry biomass):Tween20or Tween80,0.15g/g;␤-glucosidase,28.4IU/g;and cellulase, 25FPU/g.The samples were hydrolyzed for7d at50°C.At the conclusion of the hydrolytic period,the supernatants were analyzed by HPLC.Initial Reaction Rate MeasurementInitial reaction rates were measured for two sets of samples; for one set the enzyme concentration was constant(variable substrate concentration)and for the other the substrate con-centration was constant(variable enzyme concentration). For the sample set with variable substrate concentration, pretreated stover was weighed into500-mL screw-cap cul-ture flasks and distilled water was added as appropriate, such that the volume for enzymic hydrolysis after the ad-dition of all reagents would be150mL and the resulting biomass concentrations would be25.0,33.3,41.7,and50.0 g/L.Six50g/L samples and three samples for each of the other concentrations were prepared.After neutralization,the samples were placed in the incubator/shaker to preheat to the desired hydrolytic temperature(40and50°C).Once the samples had reached the incubator temperature,Tween20 (0.15g/g dry biomass),Tween80(0.15g/g dry biomass),or water,was added.Then213IU␤-glucosidase was pipetted into each flask,the shaker was started,and37.5FPU cel-lulase was added to each at30-s intervals.As each flask received its aliquot of cellulase,the cap was secured on the flask.The first sample to receive cellulase was removed from the shaker after exactly1h and a4-mL aliquot was immediately removed and heated in a water bath for15min to denature the enzymes.At30-s intervals,the remaining samples were similarly removed from the shaker.When the last sample had been removed from the water bath,the tubes were centrifuged and the hydrolyzates were analyzed by DNS assay.The50g/L samples from each temperature were also analyzed by HPLC to determine the relative amounts of glucose and xylose.The variable-enzyme samples were prepared by the pro-cedure described above using a substrate concentration of 50g/L and enzyme concentrations of0.36,0.50,0.71,and 1.07g/L.The samples were hydrolyzed at40and50°C and were analyzed by DNS assay.RESULTS AND DISCUSSIONTween Loading StudiesA portion of the data from the Tween loading experiment is plotted in Figure1.The Tween80clearly aided the enzymic hydrolysis of the pretreated corn stover;the average rate and extent of conversion in the Tween80samples was higher than for Tween-free samples.At the apparent optimum loading level of0.15g Tween/g dry fiber,a30%increase in the total sugar yield in the192-h saccharification was ob-tained.Interestingly,for unknown reasons,the Tween80 appeared to be slightly inhibitory at loadings of around 0.005g Tween/g dry fiber.For comparison,the individual glucan and xylan conver-sions for the0.15g Tween/g dry fiber loading are plotted in Figure2.In the Tween-free samples,the192-h conversion for glucan and xylan were approximately the same,about 0.60g anhydromonomer/g polysaccharide.In the samples which were loaded with0.15g Tween80/g dry fiber,the xylan conversion was about0.71g anhydromonomer/g xylan,an improvement of about18%,whereas the glucan conversion increased by36%to0.83g anhydromonomer/g glucan.This would indicate that Tween80is more benefi-cial for the cellulose hydrolysis,as was also reported by Kaya et al.(1995).Pretreatment StudiesFigure3shows that,whereas the yields obtained in the Tween samples were universally higher than the Tween-free counterparts,the recommended pretreatment conditions were identical.Therefore,using Tween during the hydroly-sis does not eliminate,or even reduce,the requirements for effective pretreatment,but does lead to much higher yields. Comparisons should also be made between the pretreated and untreated samples.The lime pretreatment,when con-ducted at the recommended conditions(0.1g Ca(OH)2/g dry biomass,10g H2O/g dry biomass,120°C,5h),increased the saccharification of the stover from about50mg equiv glucose/g dry stover to450mg equiv glucose/g dry fiber. Adding Tween80during the hydrolysis further improved the sugar yield to665mg equiv glucose/g dry stover.Only marginal hydrolysis improvement,from50to80mg equiv glucose/g dry stover,was achieved by adding Tween to the untreated biomass samples.Enzyme LoadingThe results from the50°C experiment are plotted in Figure 4as a function of hydrolysis time;this figure contains sev-eral interesting and important features.First and foremost,the enzymes performed better in the samples containing Tween80;a loading of1FPU/g dry biomass with Tween80achieved the same yield after100h of hydrolysis as a7FPU/g dry biomass loading without Tween80.Furthermore,the slopes of the yield lines for the 1and3FPU/g dry biomass samples that contained Tween 80are positive after100h at50°C indicating that the en-zymes were still active.These slopes can be compared to the flat yield lines for all of the Tween-free samples;without Tween80the enzymes were completely deactivated after about72h.Finally,a loading of3FPU/g dry biomass with Tween80resulted in a yield that was higher than that ob-tained with15FPU/g dry biomass without Tween80.In fact,when the data are plotted as a function of enzyme loading(Fig.5),the recommended enzyme loading,identi-fied as the shoulder of the yield curve,with Tween80is about3FPU/g dry biomass,vs.10FPU/g dry biomass without Tween80.When the hydrolysis was conducted at40°C(Fig.6), samples containing Tween20performed better than Tween 80samples,which in turn outperformed the Tween-free samples.However,the advantage of the Tween samples over the Tween-free samples was less at40°C than at50°C, especially with the higher enzyme loadings.For example,at 50°C an enzyme loading of15FPU/g resulted in a sugar yield(100h reaction)of697mg equiv glucose/g dry bio-mass in the Tween80sample,vs.only534mg equiv glu-cose/g dry biomass in the Tween-free sample,an improve-ment from the Tween80addition of30.5%.At40°C the corresponding samples had656mg equiv glucose/g dryFigure2.Conversion of glucan(a)and xylan(b)from saccharification ofpretreated corn stover.Pretreatment conditions:0.1g Ca(OH)2/g dry bio-mass and10g H2O/g dry biomass for5h at120°C.Hydrolysis conditions: 5FPU cellulase/g dry biomass at50°C.Legend:ࡗ0.15g Tween80/g dry biomass;᭡no Tween.Figure3.Sugar yields from pretreatment variations:(a)time at120°C,and(b)lime loading.Pretreatment conditions:(a)0.1g Ca(OH)2/g drybiomass and10g H2O/g dry biomass;(b)10g H2O/g dry biomass for5hat120°C.Hydrolysis conditions:5FPU cellulase/g dry biomass at50°C. Legend:ࡗ0.15g Tween80/g dry biomass;᭡no Tween;ииииии0.15g Tween80/g dry biomass,no pretreatment;------no Tween,no pretreat-ment.biomass and599mg equiv glucose/g dry biomass,respec-tively,an increase of only9.5%.A key to the increased yield in the Tween-free samples can be seen in the slopes of the yield lines for the lower enzyme loadings.Unlike the curves at50°C(Fig.4),the slopes for the Tween-free samples at40°C(Fig.6)remained positive,indicating that the enzymes continued to be active throughout the reaction period.However,plotting the data as a function of enzyme loading(Fig.7)reveals that,despite the increased perfor-mance in the Tween-free samples at40°C,the recom-mended enzyme loadings for hydrolysis at40°C are com-parable with those identified for hydrolysis at50°C;when using Tween the loading should be3FPU/g dry biomass, whereas if Tween is not used,the loading should be about 10FPU/g dry biomass.Hydrolysis Temperature Dependence:Tween as an Enzyme StabilizerThe results from the experiment are plotted in Figure8;the conversions are plotted as g equiv anhydromonomer/gram of original polysaccharide as a function of hydrolysis tem-perature.As expected,the samples containing Tween had higher conversions,with Tween20performing slightly bet-ter than Tween80.Furthermore,the Tween samples had higher conversions at50°C than at40°C,whereas the con-versions for the Tween-free samples were higher at40°C.In all cases,the extent of hydrolysis was much lower at30°C and60°C than at the middle temperatures.Considering that the optimum temperature in the Tween samples was10°C higher than that for the Tween-free samples,it was con-cluded that adding Tween protects the enzymes from ther-mal deactivation.Substrate Dependence of Tween:Tween as a Lignocellulose DisrupterThe results from both the Tween Dependence and Ultimate Digestion experiments indicate that Tween acts as a ligno-cellulose disrupter.The sugar yields from the Tween Dependence experiment are plotted in Figure9.It was expected that the sugar yieldFigure5.Sugar yields for100h of hydrolysis(50°C)as a function ofenzyme loading.Pretreatment conditions:0.1g Ca(OH)2/g dry biomassand10g H2O/g dry biomass for5h at120°C.Legend:ࡗ0.15g Tween80/g dry biomass;᭡no Tween.KAAR AND HOLTZAPPLE:BENEFITS FROM TWEEN DURING ENZYMIC HYDROLYSIS OF CORN STOVER423would increase with increasing substrate concentration be-cause of the correspondingly higher enzyme concentration. Sugar yield increased with increasing concentration for the Tween-free samples,although the increase beyond the50 g/L sample was slight.In the samples for which the Tween loading was constant,a large increase in yield was observed up to the75g/L sample,after which the yield decreased slightly for the100g/L sample;the decrease was probably due to product inhibition.However,in samples for which the Tween concentration was constant,the sugar yield de-pended on the Tween loading.Doubling the Tween80load-ing in the25g/L samples did not improve the yield.This is consistent with the previous finding(Fig.1)in which adding Tween80beyond a loading of0.15g Tween/g original dry fiber did not provide additional benefit.On the other hand, the75g/L and100g/L samples both had lower yields than when the Tween80loading was maintained constant.Thus, it was concluded that the benefit of Tween is loading de-pendent,not concentration dependent,which implies a re-lationship between Tween and the substrate.This result was corroborated by the Ultimate Digestion experiment.The average HPLC analysis results and corresponding conversions for each pair of samples from the ultimate di-gestion are listed in Table I.As can be seen from Table I, while the conversions were high for all samples,the con-versions for total sugars averaged about10%higher(0.97g/g)for Tween20samples,and6%higher(0.93g/g)for Tween80samples,compared to the Tween-free samples (0.88g/g).The ultimate digestion utilizes high enzyme loadings and long hydrolysis times which obfuscate the sta-bilizer and effector roles of Tween.Therefore,the higher ultimate conversions in the Tween samples must be due to more substrate being available to the enzymes;this requires that Tween disrupt the lignocellulose matrix.Furthermore, the results from this experiment indicate that Tween20is better at disrupting the matrix than Tween80,which was also the conclusion derived by Kurakake et al.(1994). The Effector Role of TweenAn effector in enzymic processes aids the interaction of the substrate and the enzyme by either making the substrate or enzyme conformation better for enzyme attachment to the substrate or facilitating the release of the enzyme from the substrate once reaction has occurred.In either case,the benefits appear in the kinetic constants associated with the reaction.Before kinetic parameters can be evaluated,an adequate model is needed.Holtzapple et al.(1984a)have developed a model for cellulose hydrolysis shown to correlate experimental data better than models derived from the traditional Michaelis-Menten approach.The HCH-1model of Holtzapple et al. accounts for enzymic limitations encountered with the in-soluble cellulose substrate.In its complete form,HCH-1 predicts the hydrolysis velocity,v,to be:v=d͓G s͔dt=␬͓G x͔͓E͔ͩ11+͚␤i͓G i͔ͪ␣+␾͓G x͔+␧͓E͔(1) where[G s]is the product concentration(cellobiose and glu-cose),␬is the reaction rate constant,[G x]is the total cel-lulose concentration,[E]is the total enzyme concentration,␤i is the enzyme binding constant for inhibitor G i(either glucose,G1,or cellobiose,G2),␣is the affinity constant,␾is the fraction of free cellulose sites,and␧is the number of cellulose sites covered by an adsorbed cellulase. Although the HCH-1model was developed for a specific enzyme(cellulase)reacting with a homogeneous substrate (cellulose),it can legitimately be used to model the enzyme mixture and heterogeneous substrate that characterize the present experiment.For example,it is likely that the mecha-nism of the reaction between the enzyme(s)and the hemi-cellulose is analogous to the cellulase-cellulose interaction because both cellulose and xylan are constructed of similar,␤-O-4linked,monomers.Furthermore,the HPLC analyses of the50g/L samples revealed that the major monosaccha-ride present in the1-h hydrolyzates was glucose(Table II); this is important for two reasons.First,the1-h hydrolyses, and therefore,the experimental data represented by the DNS analyses,are essentially indicative of the action of cellulase on cellulose.And second,the1-h xylan conversions implied in Table II are low relative to the corresponding1-h cellu-Figure9.Saccharification results with constant Tween80loading(0.15 g Tween/g biomass)or concentration(10g Tween/L solution).Pretreat-ment conditions:0.1g Ca(OH)2/g dry biomass and6g H2O/g dry biomassfor5h at120°C.Hydrolysis conditions:5FPU cellulase/g dry biomass for 100h at50°C.Legend:᭹constant Tween loading;᭿constant Tween concentration;᭡no Tween.Table I.Conversions from ultimate digestion of pretreated corn stover.Sample type Glucan(g/g)aXylan(g/g)aArabinan(g/g)aTotal(g/g)aNo Tween0.880.880.920.88Tween20 1.000.920.970.97Tween800.950.900.960.93a g anhydromonomer/g polysaccharide.424BIOTECHNOLOGY AND BIOENGINEERING,VOL.59,NO.4,AUGUST20,1998lose conversions,despite the fact that xylan conversions achieved after extended time under similar experimental conditions are approximately the same as the cellulose con-versions(Figure2).This behavior suggests the action of a single enzyme that reacts with,but has different affinities for,two substrates;as the preferred substrate(cellulose)is consumed,the enzyme is increasingly available for reaction with the second substrate(hemicellulose).Thus,it is prob-able that cellulase,rather than a different enzyme,performs the hydrolysis of the hemicellulose.If a different enzyme were responsible for the hemicellulose hydrolysis,the xy-lose concentration in the1-h samples would be close to that of glucose.Therefore,the hydrolytic system of the present experiment can best be described as a single enzyme(cel-lulase)acting on two,structurally similar,substrates(cellu-lose and hemicellulose),with the initial hydrolysis of one of the substrates(cellulose)proceeding at an approximately 10-fold faster rate than the hydrolysis of the other substrate (hemicellulose).Because of the mechanistic similarity between the present experiment and the cellulose hydrolysis model proposed by Holtzapple et al.(1984a),the HCH-1model was selected to evaluate the kinetic data.However,to provide a clearer nomenclature and differentiate the model equation,as ap-plied to heterogeneous substrates,from the original HCH-1 model equation[eq.(1)],the following changes are made: The variable representing the cellulose concentration in eq.(1),G x,is replaced with S to designate total substrate con-centration;the variable representing the concentrations of inhibitors(products),G i,is replaced with S i to indicate all possible soluble inhibitors;and the concentration of soluble products,G s,is replaced with S s to account for hydrolysis products derived from all polysaccharides.Hence,the model equation for heterogeneous substrates becomes:v=d͓S s͔dt=␬͓S͔͓E͔ͩ11+͚␤i͓S i͔ͪ␣+␾͓S͔+␧͓E͔(2)where S is the total substrate concentration(for convenience assumed to be g dry biomass/L)and the other parameters are as described for eq.(1).The variable␾in eq.(2)represents a complicated ex-pression involving[S],[E],␣,and␧.However,for the pur-poses of data correlation,Holtzapple et al.(1984b)have determined that␾can be set equal to1.The conditions of the experiments used in this study allow eq.(2)to be sim-plified even further.For T.reesei cellulase Holtzapple et al. (1990)reported the glucose enzyme-binding constant,␤1,to be equal to0.0031L/g,whereas the enzyme-binding con-stant for cellobiose,␤2,was reported as0.0184L/g.In the present experiment,the maximum glucose concentration at-tained was about13g/L,and the high level of cellobiase added to the system eliminated cellobiose entirely(as ob-served in HPLC chromatograms).If the enzyme binding constants for the other hydrolysis products,such as xylose, are assumed to be of the same order of magnitude(or smaller)as the binding constant for glucose,the product inhibition term,⌺␤i[S i],is small and can be ignored.Thus, eq.(2)reduces to:v=d͓S s͔dt=␬͓S͔͓E͔␣+͓S͔+␧͓E͔(3) Equation(3)is similar in form to expressions derived from the Michaelis-Menten model.In the absence of product in-hibition,Michaelis-Menten kinetics would predict the hy-drolysis velocity to be:v=k͓S͔͓E o͔K m+͓S͔(4)where k is a constant,E o is the total enzyme concentration, and K m is the Michaelis-Menten constant.The product of the constant,k,and the total enzyme concentration,[E o],is also referred to as the maximum reaction velocity,␯max. Hence,the only difference between the simplified HCH-1 and Michaelis-Menten models is the appearance of the en-zyme coverage parameter,␧[E],in the denominator of eq.(3);the HCH-1model completely reduces to the Michaelis-Menten model for␧ס0.Consequently,the HCH-1expres-sion,eq.(3)can be manipulated and the constants obtained using techniques typically used with Michaelis-Menten ki-netics.For example,both sides of eq.(3)can be inverted and linearized at either constant enzyme or constant sub-strate concentration.This allows the preparation of Line-weaver-Burk-type plots for parameter evaluation.At con-stant-enzyme concentration,1v=␣+␧͓E͔1+1(5) and at constant-substrate concentration,1v=␣+͓S͔␬͓S͔1͓E͔+␧␬͓S͔(6) Double-reciprocal plots were prepared for the initial reac-tion rates obtained with variable enzyme concentration and variable substrate concentration;the slopes and intercepts of the fitted lines are listed in Table III.The HCH-1constants were calculated from eqs.(5)and(6)for the various reac-tion systems using the respective slopes and intercepts listed in Table III;the resulting values for␬,␣,and␧are also reported in Table III.Table II.Sugar concentrations in50g biomass/L samples,1-h hydroly-sis.Sample type Temperature(°C)Glucose(g/L)Xylose(g/L)Arabinose(g/L)No Tween40 2.980.320.40Tween2040 3.310.390.54Tween8040 2.960.340.43No Tween50 4.260.680.53Tween2050 4.540.590.64Tween8050 4.680.590.44KAAR AND HOLTZAPPLE:BENEFITS FROM TWEEN DURING ENZYMIC HYDROLYSIS OF CORN STOVER425。