Tomato juices and tomato juice concentrates: a study of factors contributing to their gross viscosity

CENTRALE LANDBOUWCATALOGUS

0000 0174 1715 m mi BISLI0T3BEK DER LANDPO'CVHOOKSCHOOL

Promotor: dr. W. Pilnik, hoogleraar in delevensmiddelenlee r Co-promotor: dr. ir. A.G.J. Voragen, wetenschappelijk hoofdmedewerker ^JfVJO ??oi,'^'

R. Heutink

Tomato juices and tomato juice concentrates: a study of factors contributing to their gross viscosity

Proefschrift ter verkrijging van de graad van doctor in de landbouwwetenschappen, op gezag van de rector magnificus, dr. C.C. Oosterlee, in het openbaar te verdedigen op woensdag 8 oktober 1986 des namiddags te vier uur in de aula van de Landbouwuniversiteit te Wageningen

H i / 0' J ^u9p?o, , fo9

STELLINGEN

1. Dedoo r Shomer et al. (1984) gesussereerde rol van endogeen cellulase bij het uitzakken van deeltjes in tomatensap wordt niet ondersteund door de in de literatuurvermeld ecellulas ehoeveelhede ni nd etomaat . Shomer,I. ,P . Lindnere nR . Vasiliver,198 4 Journalo f FoodScience ,49:628-63 3

2. Geziend egroott eva nd edeeltje s ishe ttwijfelachti g of Raoe t al.(1986 )d ejuist e meetgeometriegebruike nvoo r hunrheologisch e metingenaa n appelmoes. Rao, M.A., H.J.Cooley ,J.N .Nogueir ae nM.R .McLellan ,198 6 Journalo f FoodScience , 51:176—179

3. Bij het gebruik van polysaccharide-afbrekende enzymente r opheldering van de structuurva npolysaccharide n moetmee raandach tworde n besteed aan: 1)d ezuiverhei dva n hetenzym , 2)d ezuiverhei dva nd esubstrate ndi e worden gebruiktvoo r het karakteriserenva nhe tenzy me n3 )ee nmogelijk emultisubstraa t specificiteitva nhe tenzym .

4. De vorming van hydroxymethylfurfural is geen bruikbare indicator voor het optredenva nniet-enzymatisch ebruinkleurin g invruchtesappen .

5. Alleen door het geven van uitgebreide voorlichting kanworde n bereikt dat de consument bestraling van voedingsmiddelen alsee nalternatiev e conserverings- methode aanvaardt. Het effect van een voorlichtingscampagne zal worden tenietgedaan wanneer dit proces in verhullende benamingen op het pirodukt wordtgedeclareerd .

6. Het zich toeeigenen van het woord significant door statistici kanto t verwarring leiden wanneer dit woord in het Engels wordt gebruikt in de betekenis van veelbetekenend.

7. De viscositeit van tomatensap en concentraten daarvan wordt grotendeels bepaalddoo r hetgehalt eaa n wateronoplosbare bestanddelen. Dit proefschrift

9. Een enzymatische hydrolyse met een mengsel van pectolytische en cellulo- lytische enzymen, voorafgaand aan een chemische hydrolyse met 2N TFA is noodzakelijk om de neutrale suikersamenstelling in tomaten WIS te kunnen bepalen. Dit proefschrift BIBLIOTHEEK DER LAHDBGUWHOGESCHOOI WAGENINGEN 10. Het is mogelijk om tomatensap te concentreren zonder dat naverdunnin g van het concentraat een verlies in viscositeit ten opzichte van het ongeconcen- treerdesa pza l optreden. Ditproefschrif t

11. Verschillen inviscositei t tussenho t breake ncol d breaktomatensa pworde n niet alleen veroorzaakt door de snelheid van enzym-inactivatie maar ook door de temperatuurtijden s hetpasseren . Dit proefschrift

12. Het gebruik van microscopische technieken om veranderingen in voedings- middelen vast te stellen kan gemakkelijk leiden tot "wishful seeing" als variant van"wishfu lthinking" . Abstract

R. Heutink, 1985.Tomat o juices and tomato juice concentrates:a study of factors contributing to their grossviscosity .Doctora lThesis , AgriculturalUniversity ,Wageningen ,Th eNetherlands .

The grossviscosit y of tomato juice and tomato juice concentrateswa s found tob e determined primarily by thewate r insoluble solids (WIS)content . The serumviscosit y did not contribute to grossviscosity . TheWI S consisted ofwhol e tomato cells,vascula r bundles and skin fragments. In general theWI S could be fractionated into 40-45% pectin, 25-30% hemicellulosean d 30-35% cellulose.Highl ybranche d aswel l asmor e linear pectin fragmentswer e found tob e present in tomatoWIS .Xylans , arabinans and (arabino)galactans were involved in the attachment of pectic substances to the cellwal lmatrix .Th e presence ofxylans ,xyloglucans , (gluco)mannansan d a limited amount of galactan in thehemicellulos e was indicated. Pectin fragments,characterize d by a low content of arabinose and galactose containing side chains,an d esterified in such awa y that theywer e degraded bybot h PG and PL,mad e an important contribution to grossviscosity . Cellulosewa s found tomak e a relatively smaller contribution togros sviscosit y than these pectin fragments.Th e rigidity of the cellwalls ,cause d by the cellulosic structure,seeme d to influence grossviscosit y athighe rWI S concentrations.Th e differenceswhic h exist naturally inWI S compositionwer e found tob e too small to cause large differences ingros sviscosity . Higher hotbrea k temperatureswer e found to result in amor e effective removal ofWI S from skinmaterial .Th e influences ofho tbrea k temperature,finishe r temperature and size of finisher screen opening on the finisher operationwer e closely related to the efficiency of the finisher.Th e concentration of a tomato juice by evaporation resulted in a loss in grossviscosity ,afte r dilution of the paste to the original strength.Th e causes for this "dilution-loss"wer e studied and could be related to the simultaneous concentration of tomato cells and,presumably , ions.Th e centrifugation-serum-concentrationmetho d resulted in a significantly decreased "dilution-loss". Curriculum vitae

RobHeutin kwa sbor n inEnschede ,th eNetherlands ,o n the 29th ofMa y 1956. From 1974 to 1980h e studied food science and technology at the Agricultural University ofWageningen ,th eNetherlands . In 1980h e obtained the degree ofMaste r of Science,cu m laude,wit h majors in food chemistry and foodmicrobiolog y andwit h amino r in toxicology. From October 1980 to May 1985 heundertook , in theDepartmen t of Food Science of the AgriculturalUniversity ,Wageningen ,th eH.J . Heinz Co. funded research work presented in thisthesis . Acknowledgemen ts

Thewor kpresente d inthi sthesi swa scarrie dou tunde rth eguidanc e ofProf .Dr .W .Pilnik ,Hea do fth eLaborator yo fFoo dChemistr yo fth e Departmento fFoo dScienc eo fth eAgricultura lUniversit yo fWageninge ni n theNetherlands .I remai nver ygratefu lt ohi mfo rpersuadin gm et oaccep t thechalleng eo fthi sPh.D .study .Hi sinspiratio nan dnumerou sidea swer e ofgrea tvalu et oth ecompletio no fthi sthesis . Ia mals over ygratefu lt om yco-promoto rDr .Ir .A.G.J .Vorage nfo r manyhour so ffruitfu ldiscussio nan dfo rhi scritica lrevie wo fth e manuscript. Ia mindebte dt oDr .Ir .F.M .Rombout sfo rth estimulatin gdiscussion s andfo rhi scarefu lreadin go fth ereport sprepare d inth ecours eo fthi s study. Iwoul dlik et othan kth eH.J .Hein zCo. ,Pittsburgh ,US Afo r initiatingan dfundin gthi sproject .I woul despeciall y liket othan kMr . A.Nip ,Dr .Ir .P .Folstar ,Dr .K .Venka tan dMr .S .Martine zfo rthei r contributiont othi sresearch . Iwoul dlik et oexpres sm ygratitud eto :Mis sG .Lokhors tfo rhe r skilfulassistenc ei nanalysin gth esamples ;Ir .G .va nd eHoe kan dIr .B . vanValkengoe d forthei rstudie so nth esubject sdiscusse d inchapter s4 and5 ;Mr .H .Schol sfo rstudyin gth estructura lfeature so fth e hemicellulosei ntomatoe san dfo ral lhi sassistenc ei ngeneral ;Dr .H.A.I . Siliha,m yfello wdoctora lscholar ,fo rhi sfriendshi pan dfo rth e interestingdiscussion so nbot hth eclou dstabilit yo faprico tnecta ran d thegros sviscosit yo ftomat oproducts ;Mrs .M.J.F .Searle-va nLeeuwe nfo r purifyingth eenzyme suse di nth eexperimenta lwork ;th eentir estaf fo f theDepartmen to fFoo dScienc efo rbein gfantasti ccolleague san dfo ral l theirassistence ;Mr .M .Schimme lfo rpreparin gth edrawings ;Mr .W .va n Hoffo rpreparin gth ephotographs ;Mis sJ .Mulde rfo rtypin gth emanuscrip t andMis sJ .Truijen sfo rcorrectin gth eEnglis htext . FinallyI wan tt othan kJolanda .Withou the rpatience ,understandin gan d encouragementthi sthesi swoul dno thav ebee nwritten . SUMMARY

The subject and objectives of thepresen t study are introduced in chapter 1. The relevant data from the literature are discussed in chapter 2. This literature review covers topics such as:th eproduction ,genera l composition and processing of tomatoes (2.1.); definitions and quality parameters for tomato products (2.2.); factors contributing to the gross viscosity of tomato products (2.3.); factors affecting serum separation in tomato products (2.4.)an d the composition of tomato polysaccharides (2.5.). For convenience the lengthy and extensive review on the factors contributing togros sviscosit y of tomato products is summarized in2.3.1 . The general analyticalmethod s used throughout this study are presented in chapter 3. Special attention hasbee n paid to a comparison of twomethod s ofmeasurin g grossviscosit y and to thedeterminatio n of the neutral sugarspresen t in tomatoWIS . Itwa s found that the Bostwick consistometer isno t sensitive enough formeasurin g high-viscosity tomato products. This instrument,however , isver y suitable formeasurin g the grossviscosit y of tomato productswit h a Bostwick flowo f 5 cmo rmore . Thehydrolysi s of thepolysaccharide s in tomatoWI Swit h 0.8 N H SO, or 2N H SO,,bot h after prehydrolysiswit h 72%H SO,,di d not result ina n adequate determination of theneutra l sugars.Prehydrolysi s of the polysaccharides with an enzymemixtur e containing pectolytic and cellulolytic enzymes,prio r tohydrolysi swit h 2NTFA , increased in particular the amounts of glucosewhic h were determined. The influence of soluble solids (serumviscosity ) and insoluble solids on thegros sviscosit y of tomato juice and concentrates isstudie d and discussed in chapter 4.Th e samples analysed differed invariet y and methods of processing (break temperature and concentration).Th e gross viscosity of tomato juices and concentrateswa s found tob e determined primarily by theirWI S content.Variet y andmetho d ofprocessin g influenced grossviscosit y at a specific °Bxb y affecting the amount ofWIS .Gros s viscosity at a specificWI S amountwa sno t influenced by variety, serum viscosity ormetho d ofprocessin g unless themetho d of processing resulted indegradatio n of theWIS . In chapter 5 the results of studies on the composition and structural features of tomatoWI S arepresente d and discussed.WI S fragments, solubilized fromWI Swit h solvents like oxalate,HC 1an d NaOH orb y using purified enzymes,wer e characterized by gelpermeatio n and ion exchange chromatography and by analysis of neutral sugar and glycosidic linkage composition. Ingenera l tomatoWI S could be fractionated into 40-45% of pectin, 25-30% ofhemicellulos e and 30-35% of cellulose. Thepecti n fraction contained large amounts ofAG A and protein and moreover approximately 25-35% of thearabinos e and galactose present in theWIS . Most of this arabinose and galactosewa s found in theHC 1 soluble pectin and formed part of ahighl y branched pectin segment solely linked to the cellwal lmatri x throughhighl y esterified AGA chains. A largeproportio n of thepecti n occurred as fragmentswit h a low content of arabinose and galactose side chains.Thes e fragmentswer e esterified in such awa y that theywer e susceptible tobot h PG and PL action. Xylans,arabinan s and (arabino)galactans were found tob e involved in the attachment of pectic substances to the cellwal lmatrix . The tomatovarietie swhic hwer e studied,differe d mainly in theWI S amount and in the arabinose and galactose content of thisWIS .Thi sdifferenc e was most clearly expressed in theHC 1 soluble pectin fraction. The composition of this fractionwa s found to changewhe n tomatoeswer e processed into juice and paste. The tomatohemicellulos e contained approximately 80%o f theWIS-xylose .60 % of this amountwa s present in g-l,4-xylans which are probably bound to pectin. 30%o f thehemicellulos e xylosewa s found tob e present in fragmentswit h a composition typical forxyloglucans .Th e hemicellulose also contained ±50 %o f the arabinose and galactose present in theWI S and 15-20% of theWIS-glucose .Al l the arabinose and themajo r part of the galactose found inhemicellulos e were present as terminal residues.Th e presence of (gluco)mannans and a limited amount of galactan in the hemicellulose fractionwa s indicated. The cellulose fraction consisted mainly of glucosebu t also contained ±15 % of the arabinose present inWIS . The results of studies on the relative contribution of WIS constituents to the grossviscosit y of tomato juices and tomato juice concentrates arepresente d and discussed in chapter 6. Technical aswel l as purified,wel l characterized enzymeswer euse d todecreas e the gross viscosity of a diluted tomato paste and of a 10°Bx tomato juice concentrate.Al l results,irrespectiv e of the tomatoproduc t used,pointe d to the important contribution of themor e highly esterified parts of the WIS pectin to grossviscosity . Itwa s considered likely that thesepart s of the pectinwer e identical to thepart s susceptible todegradatio nb y both PG and PL,a s described in chapter 5.Th e importance of the low esterified pectin and thehemicellulos e as contributors to grossviscosit y remained unclear.Th e role of the proteinpresen t inWI S as a contributor to gross viscosity was found tob e negligible. Cellulosewa s found tomak e a relatively smaller contribution to grossviscosit y than esterified pectin. The rigidity of the cellwall s caused by thepresenc e of cellulosic structures,seeme d to influence grossviscosit y athighe r WIS concentration. Itwa s concluded that the small differences inWI S composition,whic h occurred naturally in the tomatovarietie swhic hwer e studied, cannot cause large differences in grossviscosity . The grossviscosit y of tomato juices and tomato juice concentrates therefore seemed tob e determined primarily by theamoun t ofWIS ,whic h isprobabl y proportional to theamoun t of tomatocells . Inchapte r 7 the influence of processing factors such asho t break temperature,finishin g operation,concentratio n and pasteurization on the WIS characteristics and grossviscosit y of tomato juice and concentrates is discussed. The increase in the grossviscosit y of tomato juices and tomato juice concentrates as a result of an increase inho tbrea k temperaturewa s attributed to amor e effective removal ofWI S from the skinmaterial .WI S "quality" and themicroscopi c appearence of the tomato cellswer e not affected by higher hotbrea k temperatures.Th e influences ofho t break temperature,finishe r temperature and size of finisher screen opening on the finisher operationwer e found tob e closely related to the effectiveness of the finisher. Neither themetho d of concentration nor the type of evaporator used for concentration had an influence on the grossviscosit y of concentrates.I n general theWIS/T S ratiowa s found to remain fairly constant during concentration. A "dilution-loss"wa s observed when certain serum components (probably ions)wer e concentrated by evaporation in thepresenc e of the tomato cells.Th emagnitud e of the "dilution-loss"wa s found to be influenced bymechanica l shear,vacuu m and heat processing of the concentrated juice.Ho t break temperature and temperature of evaporation had no influence on themagnitud e of the "dilution-loss".Th e quality of thejuic ebefor e concentration also influenced the "dilution-loss". The centrifugation-serum-concentration techniquewa s found to prevent "dilution-loss"whe n abenc h top centrifugewa s used and to reduce "dilution-loss"whe n a decanterwa s used. Bothhea t processing and freezing were found to decrease the grossviscosit y of tomato juice and concentrates. LIST OF ABBREVIATIONS

AGA = anhydrogalacturonic acid AIS = alcohol Insoluble solids Ara. = arabinose DE = degree of esterificatlon ofAG A (%) DEAE = diethylamino-ethyl EDTA = ethylenediamlnetetra-acetate g = acceleration due togravit y (m.s ) Gal. = galactose glc = gas liquid chromatography glc-ms • combined gas liquid chromatography andmas s spectrometry Glue. = glucose HPLC =hig h pressure liquid chromatography Man. =mannos e MW =molecula r weight NS =neutra l sugar n.s. =no t significant PE = pectinesterase PG » polygalacturonase PL = pectinlyase pnp = para-nitrophenyl Rham. = rhamnose TFA = trifluoracetic acid WIS =wate r insoluble solids WIS-ara.= arabinose present inWIS .I n the samewa yWIS-AG Aetc . Xyl. = xylose n = grossviscosit y measured with theHaak e rotovisco and expressed app asapparen t viscosity (mPa.s). n <•seru m viscosity •loot-"™ "* CONTENTS

Summary

Listo fabbreviation s

1.Introductio n 1

2.Revie wo fliteratur e 3 2.1.Th etomato :production ,genera lcomposition ,processin g 3 2.2.Definition san dqualit yparameter sfo rtomat oproduct s 9 2.3.Factor scontributin gt oth egros sviscosit yo ftomat oproduct s 12 2.3.1.Summar y 12 2.3.2.Contributio no fvarieta lan dhorticultura lfactor st o grossviscosit y 15 2.3.3.Polysaccharid edegradin genzyme si ntomatoe s 23 2.3.4.Contributio no fcompositiona lfactor st ogros sviscosit y 31 2.3.5.Contributio no fprocessin gfactor st ogros sviscosit y 45 2.4.Factor saffectin gseru mseparatio ni ntomat oproduct s 64 2.5.Compositio no ftomat opolysaccharide s 66

3.Genera lanalytica lmethod s 73 3.1.Viscosit ymeasurement s 73 3.2.Determinatio no fsolid sconten t 74 3.3.Chemica lanalysi so fWI S 75 3.4.Propertie so ftomat oparticle s 80 3.5.Conductivit ymeasuremen t 81 3.6.Determinatio no fseru msugar s 81 3.7.Enzymes :purificatio nan dactivit ydeterminatio n 81

4.WI Sa sa facto rdeterminin ggros sviscosit y 85 4.1.Introductio n 85 4.2.Method s 85 4.2.1.Preparatio no fho tbrea kan dcol dbrea ktomat ojuic e 85 4.2.2.Concentratio no ftomat ojuic e 87 4.2.3 Reductiono fseru mviscosit yan dexchang eo fseru m 88 4.2.4.Dialysi so ftomat ojuic e 88 4.3.Result san dDiscussio n 89 4.4.Conclusion s 102

Compositionan dstructura lfeature so ftomat oWI S 105 5.1.Introductio n 105 5.2.Method s 105 5.2.1.Fractionatio no ftomat oWI S 105 5.2.2.Characterizatio no fpecti nfraction s 106 5.2.3.Characterizatio no fth ehemicellulos efractio n 107 5.2.4.Enzymati cfragmentatio no fWI San dcharacterizatio n ofth efragment ssolubilize d 109 5.3.Result s 111 5.3.1.Quantit yan dcompositio no fWI Sfro mfres htomatoe s 111 5.3.2.Microscopi cappearenc eo fWI Si ntomat ojuic e 112 5.3.3.Genera lcompositio no fWIS ,pectin ,hemicellulos ean d celluloseprepare d fromtomat ojuic e 112 5.3.4.Compositio nan dstructura lfeature so fWI Specti n 114 5.3.5.Compositio nan dstructura lfeature so ftomat o hemicellulose 119 5.3.6.Compositio nan dstructura lfeature so fth efragment s solubilizedfro mWI Sb yenzymes . 124 5.4.Discussio n 135 5.5.Conclusion s 139

Contributiono fWI Sconstituent st ogros sviscosit y 142 6.1.Introductio n 142 6.2.Method s 142 6.2.1.Enzym etreatmen to fdilute dho tbrea kpast e 142 6.2.2.Enzym etreatmen to fa 10°B xtomat ojuic econcentrat e 146 6.3.Result san dDiscussio n 147 6.3.1.Effec to fenzym etreatmen to ngros sviscosit yan d compositiono fdilute dho tbrea kpast e 147 6.3.2.Effec to fenzym etreatmen to ngros sviscosit yan d compositiono fa 10°B xtomat ojuic econcentrat e 157 6.4.Conclusion s 159 Influenceo fprocessin gfactor so nWI Scontent ,WI Scompositio n andgros sviscosit yo ftomat ojuic ean dconcentrate s 161 7.1.Introductio n 161 7.2.Method s 161 7.2.1.Productio nscal eequipmen tuse dt oproduc esample sfo r studyingth eeffect so fprocessin gcondition so ngros s viscosity 161 7.2.2.Productio no ftomat ojuic ean dpast ea tvariou sho t breaktemperature s 163 7.2.3.Productio no ftomat ojuic ean dpast eusin gdifferen t finishersan dfinishe rcondition s 163 7.2.4.Concentratio no ftomat ojuic e 164 7.3.Result san dDiscussio n 166 7.3.1.Effec to fho tbrea ktemperatur eo ngros sviscosit y 166 7.3.2.Effec to ffinishin go ngros sviscosit y 171 7.3.3.Effec to fconcentratio no ngros sviscosit y 175 7.3.4.Effec to fhea tprocessin gan dfreezin go ngros s viscosity 189 7.4.Conclusion s 191

References 192

Samenvatting INTRODUCTION

The tomato,a membe r of the Solanaceae family and therefore related to thenightshades ,wa s not generally accepted as a food Itemunti l themiddl e of the 19th century. Since then,th eworl d production of tomatoes for consumption has increased and in 1979 it ranked third in theworl d production of fruits (grapes and citrus fruits ranking first and second). Thenutrien t value of the tomato isno t particularly highbu tbecaus e of the large consumption the tomato stands first in the total contribution of a group of 10vitamin s and minerals in theU S diet (Rick 1978). Approximately 70-80% of the total annual tomato crop isuse d in the tomato processing industry. In Europe tomato paste is themai n tomatoproduc t (± 70%o f the total processed tomatoes). Other products are:peele d tomatoes (21%); passata,a slightly concentrated tomato juice (6%);an d tomato juice(2%) . Secondary tomato products such asketchu p and sauces are produced either from fresh tomato juice or from tomatopaste . An important characteristic of tomato products isgros s viscosity. Grossviscosit y describes the result of aviscosit y measurement of the whole tomatoproduct ,i n contrast to serumviscosit y which describes the viscosity of the tomatoproduc t freed of the suspended particles. It iswel lknow n that tomato juices and pastes ofhig h grossviscosit y are obtained when tomatoes areheate d to a temperature exceding 90°C (hot break)befor e the removal of the skins and seeds.Juice s and pastes of low grossviscosit y are obtained when tomatoes areheate d to 40-60°C (cold break)befor e the removal of skins and seeds.Thi sdifferenc e has always been ascribed to the degradation of pectinb y pectolytic enzymes.Th e literature,however , showsn o agreement on the contribution of serum viscosity andwate r insoluble solids constituents such aspectin , cellulose and protein,t o grossviscosity . No exact data are available on the composition and structure of the polysaccharides present in tomato products and on the relationbetwee n changes in these polysaccharides and changes in grossviscosity . Thepresen t study isa n attempt to establish the role ofwate r insoluble solids (WIS)an d soluble solids as contributors to the gross viscosity of tomato juices and tomato juice concentrates.Th e composition and structural features of tomatoWI S aswel l as the relative contribution ofWI S constituents to grossviscosit y is described. The influence of processing factors on the amount and composition ofWI S and on gross viscosity isals o studied. REVIEW OF LITERATURE

2.1. The tomato: production, general composition, processing

The tomato forms a part of the genusLycopersicon ,a membe r of the Solanaceae family.Th eLycopersico n genus isnormall y subdivided into two groups; namely the Eulycopersicon and the Eriopersicon. Fruits of the Eriopersicon,whic h isnativ e to theAndea n region (Chile,Peru ,Ecuador) , remain green throughout development. The red coloured Eulycopersicon contains two species:Lycopersico n pimpinellifolium,know n as the currant tomato,an d the cultivated tomato Lycopersicon esculentum (Rick 1978,Sim s 1980,Davie s &Hobso n 1981,Goodenoug h 1981). The origin and early events of the tomatoe'sdomesticatio n remain rather obscure,althoug h it is generally believed thatMexic o is the region of originaldomestication . In theNahu a language ofMexic o theplan t isknow n as tomatlwhic h is undoubtedly the origin of themoder nname .Th e cultivated form of the tomato probably originated from thewil d formL . esculentumvar . cerasiforme,know n as the cherry tomato.Shortl y after the discovery of the NewWorl d the tomatowa s introduced to the European region.Th e first descriptions on the occurrence of the tomatoplan t in southernEurop e were published by Pier AndreaMattiolu swho ,i n 1544,describe d aplan t known as pomid'oro ,mal a aurea (golden apple) or pornoamori s (love apple). Because of its known relationship to thenightshades ,whic h also belong to the Solanaceae family, it took until themiddl e of the 19th century for the tomato tob e generally accepted as a food item. Since then, world production of tomatoes for consumption has increased and in 1979 it ranked third in theworl d production of fruits after theproductio n of grapes and citrus fruits (Davies &Hobso n 1981). In general,70-80 % of the total annual tomato crop isuse d in the tomato processing industry (Rick 1978,Mores i &Liverott i 1982). Table 1show sproductio n data for theU.S .an d three E.E.C. countries for theyear s 1982-1985. Table 1:Productio n of fresh tomatoes (tonnes)fo r processing industry.

1982 1983 1984 1985

U.S. total 7.298.990 7.029.840 7.679.570 7.021.320 California 6.148.000 5.972.930 6.591.750 6.020.000 Italy 3.037.900 4.348.296 5.597.327 3.669.973 Greece 1.008.696 1.076.211 1.600.000 1.300.010 France 375.367 304.500 355.000 391.233

Sources: AnnualVegetables ,Cro p Reporting Board, SRS,USDA ,Jun e 1983. California Vegetable Review,vo l 6 (10),1985 . Data for Europewer e kindly supplied by mr. R. Gandolfi, Parma, Italy.

Although theU.S . Supreme Court in 1893 adjudged the tomato botanically tob e a fruit,i n common language it is still often considered tob e avegetable . Botanically the tomato isa fleshyberr y (Fig. 1) covered by an epidermis (skin)consistin g of four to five layers of small sized cellsunde r a thincuticle .

epidermis outer wall of pericarp seeds radial wall of locular cavity pericarp with jelly-like parenchyma columella inner wall of pericarp

vascular bundles placental tissue

Fig.1 :Transvers esectio no fmatur etomat ofrui t (from:Davie s& Hobso n1981) .

Thebod y of the fruit isknow n as thepericar p and consists of outer, radial and innerwall s (columella). The locular cavities appear as gaps in thepericar p and contain the seeds embedded ina jelly-like parenchymateous tissue originating from theplacenta .Th enumbe r of locules is two ormore , dependingo nth evariety .Vascula rbundle sradiat efro mth este men do fth e fruit,bot hroun dth epericar pan ddow nth ecolumella ,t oth ebotto mend . Thenutrien tvalu eo fth etomat oi sno tparticularl yhigh .A si sshow nb y Rick (1978)th etomat orank s16t hwhe ncompare dwit hothe rfruit san d vegetableso nit srelativ econcentratio no fa grou po f1 0vitamin san d minerals.However ,becaus eo fth elarg econsumptio ni tstand sfirs ti n totalcontributio no fthes enutrient si nth eU.S .die t (Table 2).

Table 2:Consumptio n vs nutritional content of tomatoes in theU.S .

nutrient Contribution of concentration nutrients to diet Crop Rank Crop Rank Broccoli 1 ,Tomatoe s 1 Spinach 2 Oranges 2 Brussels sprouts 3 Potatoes 3 Lima beans 4 Lettuce 4 Peas 5 Sweet corn 5 Asparagus 6 Bananas 6 Artichokes 7 Carrots 7 Cauliflower 8 Cabbage 8 Sweet potatoes 9 Onions 9 Carrots 10 Sweet potatoes 10 Sweet corn 12 Peas 15 Potatoes 14 Spinach 18 Cabbage 15, Broccoli 21 Tomatoes 16/ Lima beans 23 Bananas 18 Asparagus 25 Lettuce 26 Cauliflower 30 Onions 31 Brussels sprouts 34 Oranges 33 Artichokes 36

Table3 show sth emea ncompositio no ftomat ofruits .Th edat aar ecompile d fromrevie warticle sb yHerman n (1979)an dDavie s& Hobso n (1981).Solubl e sugarsrepresen tapproximatel y50 %o fth etota ltomat osolids .Anothe r25 % ofth etomat osolid si so fhig hmolecula rweigh tan dinsolubl ei nalcoho l (AIS).Othe rimportan tconstituent so fth etomat osolid sar eorgani cacid s andminerals .A mor edetaile dcompositio no fthi sAI Swil lb ediscusse di n section2.5 . Table 3: Mean composition of tomato fruit compiled from Hermann (1979) and Davies & Hobson (1981).

fructose 52% .- soluble sugars 50% glucose 45% sucrose 3%

r-cellulos e 30%

pectlc substances25 % .AI S 25%_ •-water 94% - protein 20% I-hemlcellulos e 10%

citric acid 70% dry matter 6% organic acids 12% malic acid 30%

potassium 60% -mineral s 8% phosphorus 4% { calcium 3%

- lipids

- amino acids

_ vitamins

•—other s 5% _- polyphenols ascorbic acid

- pigments -volatile s

As already mentioned, the largest portion of the annual tomato crop is processed. In the EEC countries themai n tomato product is tomato paste (Table 4).Ital y is,base d on quantities, the only producer of passata,a slightly concentrated tomato juice of approximately 8°Bx. Table4 :Productio no ftomat oproduct si n% o ftota lprocesse dtomatoe s (1985).

peeled tomato tomato Juic e passata paste

Italy 28.6 3.1 8.2 59.1 Greece 1.4 - - 97.6 France 12.9 1.0 4.8 78.8 totalEE C 20.8 2.2 5.9 69.9

Datakindl ysupplie db ymr .R .Gandolfi ,Parma ,Italy .

Table5 give sa nindicatio no fth eamount so ftomat opast eproduce di nsom e Europeancountrie si n198 4an d 1985.I tshow stha tItal yi sb y farth e largestproduce ro ftomat opaste .

Table5 :Productio no ftomat opast ei nEurop e (tonneso ftomatoes ) in198 4an d1985 .

1984 1985

Italy 3.187.894 2.169.395 Greece 1.558.400 1.268.890 Turkey* 930.000 900.000 Portugal* 580.000 630.000 Spain* 480.000 350.000 France 279.000 308.267

Datakindl ysupplie db ymr .R .Gandolfi ,Parma ,Ital y *approximat efigures .

Asimplifie d schemefo rth eproductio no ftomat opast ei soutline di n Fig. 2.Mor edetaile dinformatio nca nb eobtaine dfro mNelso n& Tressle r (1980),Goos e (1981),L eMair e (1981)an dGoul d (1983). tomatoes

harvesting

transporting

pooling/fluming

washing

sorting

crushing

crushed tomatoes

preheating

pulping /finishing

•filling

Fig.2 :Flo wdiagra mfo rth eproductio no ftomat opaste . Synonymsfoun di nliterature :crushin g- choppin g breaking, preheating- scalding ,juic e=pulp .

After hand picking ormechanica l harvesting, the tomatoes are,i ngeneral , transported to theprocessin g plant by means ofbul k trucks.Th e trucks are unloaded with large capacitywate rhose s and flumed to stock poolswhic h provide a continuous supply of tomatoes to theplant .Befor e being processed the tomatoes are thoroughlywashe d (5-10pp m chlorine)an d sorted out for defects.Th ewashe d tomatoes are chopped into smallpiece s and heated toa specific temperature.A distinction hasbee n drawnbetwee n a hotbrea k process and a coldbrea k process. In theho tbrea k process tomatoes areheate d as quickly as possible to a temperature higher than 90°C inorde r to inactivate pectolytic enzymes,whil e in the cold break process the tomatoes areheate d to any temperature between 40°C and 60°C. Often,th e chopping and heating procedure isperforme d simultaneously ina recirculating chop/scald system consisting of a chopperbo x and tubular heater.Th eRoss i& Catelli "Eldorado" chopper/scalder,equippe dwit h a pressurized section,offer s theopportunit y tohea t the tomatoes almost instantaneously toan y temperature up to 110-115°C.Th ehot ,crushe d tomatoes pass a two or three stagepulper/finishe runit . Seeds and skins are removed;moreove r thejuic e is refined.Typica lvalue s for the pulping screen are 1.0-1.5mm ,whil e the finisher screensma y range from 0.4-1.0m m depending on tomato quality and product specifications.Man y factories increase theiryiel d by pressing thewast e materialwhic h remains after the pulping/finishing step (LeMair e 1981). The juice obtained after this pressing action iscentrifuge d to remove impurities,an d then the centrifuged juice isadde d to the refined juice. In the early days of the tomato paste industry thejuic ewa s concentrated in open pans; due to the high temperatures theproduc t wasver y susceptible to caramelisationan d scorched flavours.A majo r improvement was the introduction ofvacuu m pans orboules ;man y factories stillus e these type of evaporators.Moder n tomato paste plants,however ,us e large continuous double or triple effect evaporators,wit h capacities ranging from 60 to over 250 tonnes of 28-30°Bx tomato paste/24 hours (Goose 1981,Anonymou s 1981). The rise in energy costs focussed attention onmor e energy efficient concentration methods such as themechanica lvapou r recompression evaporator and reversed osmosis (Dale et al. 1982). Reversed osmosis isalread y used on a commercial scale for thepreconcentratio n of tomato juice to approximately 8.5°Bx.Th e final step in theproductio n of tomato paste ispackaging .Traditionall y tomato pasteha sbee n packed in 5k g cans.Althoug h largevolume s are still packed inthi sway ,mor e and more tomato paste ispacke d aseptically inmetallize d plasticbag s (Orbell 1980).

2.2. Definitions and quality parameters for tomato products

In their standards of identity (Lamb 1977)th eU.S .Foo d and Drug Administration (F.D.A.)define s tomato juiceas :

"theunconcentrate d liquid extracted frommatur e tomatoes of red or reddish varietieswit h orwithou t scalding followed by draining. Inth e extraction of such liquid,hea tma y be applied by anymetho d which doesno t add water 10

thereto. Such liquid is strained free of skins,seed s and other coarse or hard substances,bu t carries finely divided insoluble solids from the flesh of the tomato. Such liquid may be homogenized andma y be seasoned with salt."

Thenam e concentrated tomato juicema y be used when tomato juice is concentrated to a levelno t less than 20%natura l tomato soluble solids (N.T.S.S.,measure d on the sugar scale of a refTactometer). Tomato puree and tomato paste areprepare d from one or any combination of the following optional tomato ingredients: 1)th e liquid obtained frommatur e tomatoes of red or reddishvarieties . 2)th e liquid obtained from peelings and cores remaining after preparation of tomatoes for canning,wit h orwithou twhol e tomatoes orpiece s of tomatoes. 3)th e liquid obtained from the residue from partial extraction ofjuic e from tomatoes. Themetho d ofpreparin g tomato puree/paste is the same as for tomatojuice . As a result of concentration,tomat o puree contains not less than8.0 % but less than 24.0%N.T.S.S. ;tomat o paste contains not less than24.0 % N.T.S.S. Tomato ketchup isprepare d from the same ingredients as tomato puree/paste but in addition salt,vinegar ,spice s and/or flavourings,onio n and/or garlic and sweetening agents (sucrose,dextrose ,glucos e sirup)ar e added. Codex Alimentariusdefinition s for tomato juice and tomato paste correspond with theF.D.A .definition s for tomato juice and concentrated tomatojuice . Table 6 shows theU.S .Departmen t ofAgricultur e (U.S.D.A.) standards for tomato juice,puree ,past e and ketchup (Lamb 1977). 11

Table6 :U.S.D.A . standardsfo rtomat ojuice ,puree ,past ean d ketchup.

tomato juice tomato puree/paste tomatoketchu p factor maximum gradeA gradeC maximum gradeA gradeC maximum gradeA gradeC points points points

colour 30 26-30 23-25 50 45-50 40-44 25 21-25 17-20 consistency 15 13-15 10-12 - - - 25 22-25 18-21 defects 15 13-15 10-12 50 45-50 40-44 25 21-25 18-20 flavour 40 33-40 27-32 1 - - 25 21-25 17-20 totalscor e 100 85 70 100 90 80 100 85 70

1:qualit yfactor ,howeve rno trated . 2:grad eA >33 %tota lsolubl esolids ,grad eB ^29% ,grad eC J25% .

Consistency has reference to theviscosit y of theproduct .Th e tendency of the insoluble solids to separate,leavin g practically clear liquid at the top, isals o tob e noted in this connection."Goo d consistency" in ketchup (score 22-25), for example,mean s that theproduc t showsno tmor e thana slight separation of free liquidwhe n poured ona flat grading tray; isno t excessively stiff and flowsno tmor e than 9 cm in 30 seconds at 20°C in the Bostwick consistometer (for description of Bostwick consistometer see: Davis et al. 1954 and section 3.1.). In general,however ,th e terms consistency,viscosit y and gross viscosity areuse d interchangeably todescrib e theviscosit y of thewhol e tomato product; this in contrast to the term serumviscosit y which describes theviscosit y of the tomato product freed of suspended particles. In this thesis the terms grossviscosit y and serumviscosit y willb e used todescrib e theviscosit y of "whole tomato product" and "serum" respectively. 12

2.3. Factors contributing to the gross viscosity of tomato products

2.3.1. Summary

Invie w of the extent of section 2.3. the data on the factors contributing to thegros sviscosit y of tomato products as found in the literature are summarized below. Grossviscosit y isa n important quality attribute of tomato products,suc h as juice,concentrates ,ketchu p and other sauces.I n the literature review the factorswhic h contribute to grossviscosit y were divided intovarieta l and horticultural factors, compositional factors and processing factors.I t is realized that these factors are interrelated,bu t for the sake of clarity theywer e dealt with separately. The influence of temperature and degree of concentration on the grossviscosit y of tomato productswa s not discussed. The grossviscosit y of tomato products depends primarily on the viscosity potential of the tomatoes used for processing. Tomato variety (firmness),climate ,degre e of ripeness at the time ofharvest ,damag e as result of harvest and transportation,an d length of storage before processing, influence theviscosit y potential of tomatoes. In general itha sbee n shown thatvarietie s high in insoluble solids (firm tomatovarieties )giv eproduct s ofhighe r grossviscosity . Tomato firmness depends not only on thevariet y but also on the stateo f ripeness of the tomato fruit.Juice s prepared from firmunrip e tomatoes are higher in grossviscosit y thanjuice s from soft ripe tomatoes.Softenin g as a result of tomato ripening has alwaysbee n associated with enzymatic modifications of cellwal l polysaccharides. Severalauthor shav e observed a decrease inAI S content and cellwal lweigh t and a solubilization of pectin as a result of tomato ripening.Th e enzymewhic h hasbee n reported as playing ake y role in tomato fruit softening is polygalacturonase. Ripening has alsobee n found toresul t ina ne t losso f galactose and, toa lesser extent,o f arabinose from the cellwalls ;moreove r marked changes in themolecula r weight of tomato cellwal lhemicellulose shav ebee n reported. These changes occurred independently of thepecti n solubilization. However the role of enzymes other thanP G (B-1,3 glucanase, g-galactosidase, cellulase,pectinesterase ) in the softening of tomato fruit hasbee n found 13

tob e doubtful. The extent towhic h the fruits are damaged during harvesting,an d the time betweenharves t and processing havebee n found tob ever y important for product characteristics.Damag e to the tomatoes hasbee n found to result in large serum losses and in reduced serumviscosity . An increased concentration of insoluble solids in thejuice ,le d to increased gross viscosity but negatively influenced other juice characteristics (yield, serum separation). The effects of other factorswhic h might influence theviscosit y potential of tomatoes,lik e culturalpractice s (timeo f planting,metho d of planting, use of fertilizers) and climate (temperature,amoun t of sun)ar eno t well documented. The abundant presence ofpectolyti c enzyme activity intomatoe sha s resulted in the fact that for years the emphasisha sbee n put on the importance ofpecti n andpecti n degradation as factors determining the viscosity of tomatoproducts .Theoretically ,however ,al l the polysaccharide degrading enzymes in tomatoes are able to degrade their substrates and influence product viscosity from themomen t the tomatoes are crushed. The tomatoha sbee n found to containvariou s typeso f polysaccharide degrading enzymes ofwhic h themos t important are: polygalacturonase,pectinesterase ,cellulas e degrading enzymes, g-l,3-glucanasean d B-galactosidase. The grossviscosit y of tomato products isals o influenced by the composition of theproducts . A comparison of the data found in the literature has showndisagreemen t on the importance of thewate r soluble solids or serumviscosit y in contributing to grossviscosity .Man y authorshav e concluded that serum viscosity influences grossviscosit y to a great extent.Othe r authors,o n the contrary,hav e shown that grossviscosit y isno t affected by large differences in serumviscosity . The important role ofwate r insoluble solids (WIS) indeterminin g gross viscosity ismor e generally accepted. Proportion (WIS/TS ratio), size and shape of thewate r insoluble solidshav ebee n reported as influencing gross viscosity. The alcohol insoluble solids (AIS)hav e alsobee n found tob e highly correlated togros sviscosity . In theavailabl e literature,however , there 14 isn o consensus on the relative contribution of thevariou sAI S constituents to grossviscosity . Conflicting resultshav ebee n found regarding the influence of totalpectin ,pecti c acids,pectini c acids, cellulose and protein on grossviscosity . Electrolytes such as soluble pectin,citri c acid, sodium chloride and calcium chloride,whic h are present in tomato juice,hav e sometimesbee n reported ashavin g anegativ e influence on gross viscosity. Apart frombein g influenced by varietal,horticultura l and compositional factors,th e grossviscosit y of tomato products is also influenced by processing conditions. The preheating of crushed tomatoes results in the inactivation ofenzymes , including the pectolytic enzymes.Hea t treatments of 15 sec.a t82° C and 15 sec.a t 93°C havebee n found to inactivate pectinesterase; the inactivation of polygalacturonase was accomplished after 100 sec.a t 88°C, 15 sec.a t 93°C and 15 sec.a t 104°C. Thebeneficia l effect of higher preheat temperatures on the gross viscosity has oftenbee n explained in termso f quicker enzyme inactivation and a better pectin retention. Some authors have concluded that softening of the tissue isals o important in determining the effect of preheating on gross viscosity. A delay between crushing and preheating has usually been found to result in decreased grossviscosit y of the finished juice.Thi snegativ e effect on grossviscosit y isdetermine d by the coarseness of theparticle s in the crushed tomatoes (smallparticle s favour enzyme action),b y the temperature and duration ofholding ,an d probably by thevariet y (amount of pectolytic enzymes).Th eholdin g of crushed tomatoes at temperatures inexces s of 90°C often resulted in increased grossviscosit y of thejuice . The finisher operation,wher e skins and seeds are separated from thejuice , hasbee n reported as influencingbot h theyiel d and the grossviscosit y of tomato juice.Th eus e of apaddl e type finisher resulted injuice s higher ingros sviscosit y than thoseprepare d with a tapered screw finisher.A n increase inpaddl e speed or an enlargement of the finisher screen openings hasbee n found to increase grossviscosity . Almost no datawer e found on the effect of the finisher temperature onjuic e gross viscosity. Preheating was,however ,foun d to influence the finisher operation, probably due to the softening of the tomato tissue. 15

The effects of concentration on thegros sviscosit y ofho tbrea k tomato juice hasbee n studied by only a few authors.Thei r datahav e shown that concentration results in a loss in grossviscosit y when the concentrate is diluted to the strength of thejuic ebefor e concentration. The extent of desiccation of thewate r insoluble solids and their inability to resorb maximally havebee n proposed as the cause of thisphenomenon . Increases in the grossviscosit y of tomato juices and concentrates have been found to occur after homogenization. These increaseshav e been attributed toa decrease in cell size,a n increase in total surface area and entanglement of the cell fragments. Thehea t processing of tomato juice and paste,befor e or after packing,ha s been found to influence grossviscosit y negatively. Itha s been hypothesized that the decrease in grossviscosit y was due to a decrease in themolecula r weight ofpectins . Itha s been found that storage doesno t affect the grossviscosit y of tomato juice,bu t does cause a decrease in the grossviscosit y of dilutions ofhig h solidspaste s (30°Bx).

2.3.2. Contribution of varietal and horticultural factors to gross viscosity

Theviscosit y (grossan d serum)o f tomato products depends primarily on theviscosit y potential of the tomatoesuse d for processing. Tomato variety, climate,degre e of ripeness at the time of harvest,damag e as result ofharves t and transportation,an d length of storage before processing,hav ebee n found to influence theviscosit y potential of tomatoes. Pear shaped tomatoeshav e oftenbee n reported to containmor e AIS,WI S and total pectin,an d toyiel d juices and pasteshighe r inviscosit y than the round varieties (McColloch et al. 1950,Lu h et al. 1954,Smi t &Nortj e 1958,Twig g 1959,Lu h et al. 1960,Coole r 1962,Ro y & Choudhury 1972). Bel-Haj (1981)showe d that five cultivarswhic h didno t differ significantly in chemical composition also failed to show differences in juice grossviscosity . These examples show that varietal differences in composition affect theviscosit y of theproducts .Fo r this reason tomatovarietie s are often classified according to their viscosity 16 potential inhigh-viscosity ,medium-viscosit y and low-viscosity varieties. In general itha sbee n shown thatvarietie shig h in insoluble solids (measured as alcoholo rwate r insoluble solids)wil lproduc e products of higherviscosity . In thepas t 10-15 years numerous newly developed tomato cultivarshav ebee n studied with respect toyiel d and juice characteristics.Base d on the results of their studies,Mars h et al. (1982, 1983)stat e that themos t important attribute of a processing tomato cultivar after its soluble solids (NTSS)content ,i s the% of insoluble solids in the total solids (WIS/TS). These twovalue s provide critical information forplannin g purposes.NTS Swil l determine theyiel d for a product concentrated to some given level of solids,bu t bothNTS S and WIS/TS ratio are necessary topredic t theyiel d for aproduc twhic h has both grossviscosit y and solids specified in its standard of identity. Johanessen (1981), therefore,pointe d to thenee d for developing tomato varietieswit h high soluble solids andWIS/T S ratios tob e used for gross viscosity oriented tomato products.Thi sWIS/T S ratiowa s found tovar y widely inne w tomato cultivars (Table 7); however, the ratio increased as a result of the search for tomatovarietie s suitable for mechanical harvesting (firmvarieties) .

Table7 :Trend si nWIS/T Srati o(fro mMars he tal .1980 ,1982 , 1983).

year 1976 1977 1978 1979 1980 1981 1982 no.var . 23 12 12 12 12 12 12 range 0.101-0.1820.109-0.17 30.125-0.18 50.137-0.19 50.114-0.17 10.125-0.18 60.130-0.16 4 mean 0.135 0.139 0.152 0.167 0.145 0.166 0.152

The success indevelopin g firmer tomatovarieties ,whic h also yield juices high ingros sviscosit y is illustrated with the following example. In 1971 theV F 145variet y accounted for about 50%o f themechanicall y harvested tomatoes in theU.S .Thi svariet y was said tohav eman y desirable characteristics highly acceptable togrower s aswel l asprocessor s (Luhe t al. 1971).Mars h et al. (1979)considere d thisvariet y tob e only medium firm and susceptible to damage during mechanical harvest and 17 transportation.Finally ,Lu he tal . (1985)classifie d thisvariet y asa low-viscosit yvariety . Thedevelopmen to fthes efir mvarietie ssuitabl efo rmechanica l harvestingcause d someunexpecte dproblem swit hth eproductio no fcol d breaktomat opaste .Du et oth ehig hinsolubl esolid sconten to fth e tomatoes,col dbrea kyield swer einsufficien tan dproblem swer eexperience d inpumpin gth epast e (Bartholin 1981).T oovercom ethes eproblem sBartholi n (1981)studie dth eusefulnes so faddin gpectolyti cenzyme sa swel la s removingo finsolubl esolid sb ycentrifugin ga sa mean so fdecreasin gth e grossviscosit yo fcol dbrea ktomat opaste . Tomatofirmnes sdepend sno tonl yo nvariet ybu tals oo nth estat e ofripenes so fth etomat ofruit .Hal l (1963)reporte da decreas ei nAI S contento ftomat ofrui tdurin gripening .Whittenberge r& Nuttin g (1957) showedtha tjuice sprepare dfro mgree ntomatoe swer ehighe ri ngros san d serumviscosit ya swel la scel lwal lan dpecti ncontent ,tha njuice s prepared fromrip etomatoes .Thi sinfluenc eo fdegre eo fripenes so n viscosity isconfirme db ysevera lothe rauthor s (Twigg 1959,Lu he tal . 1960,Coole r1962 ,Bartolom e1972) . Thebiochemica lchange si nfruit saccompanyin gripenin ghav ebee n extensivelyreviewe db yPresse y (1977),Hobso n (1981)an dKne e& Bartle y (1981).Tomatoe shav eofte nbee nuse da sa syste mfo rstudyin gfrui t ripeningbecaus eo fth eavailabilit yo fgeneti cmutant stha td ono tripe n normally;neve rrip e (Nr),ripenin g inhibited (rin),an dnon-ripenin g (nor).Softenin ga sa resul to ftomat oripenin galway sha sbee nassociate d withenzymati cmodification so fcel lwal lpolysaccharides . Wallner& Bloo m (1977)reporte da 33 %decreas ei ncel lwal lweigh tdurin g ripeningfro mth ematur egree nt oth ere drip estage .Th eenzym ewhic hha s beenreporte da splayin ga ke yrol ei ntomat ofrui tsoftenin gi s polygalacturonase (PG). PGi sabsen ti nmatur egree nfruit sbu tincrease s dramaticallydurin gripenin g (Foda1957 ,Hobso n1964 ,Gros s& Wallne r1979 , Tuckere tal .1980 ,Tucke r& Grierso n 1982).Solubilizatio nan ddegradatio n ofpecti nha sbee nreporte dt ocoincid ewit hth eincreas ei nP Gactivit y (Gross6 .Wallne r1979 ,Hube r1983 ,Rushin g& Hube r 1984).Hobso n(1963 ) 18 reported a decrease in total pectin of 37-41% and a loss in insoluble pectin of 87-89%.Th e ratio insoluble pectin/total pectin decreased. The ripening mutants show decreased PG activity and rate of softening (Nr-mutant), or noPG-activit y and softening at all (rin-mutant) (Buescher & Tigchelaar 1975,Tucke r et al. 1980,Tucke r & Grierson 1982). Hobson (1980)propose d that thever y low activities ofPG ,phosphofructokinas e and NADP -malic enzyme in the fully developed mutant lines offered an explanation for thewea k climacteric respiration rise and slow ripening of Nr-fruit and thenon-climacteri c behaviour of rin-fruit. Firm tomato varietieswhic h ripen slowly also contained lessP G activity and showed less pectin degradation during ripening than the softervarietie s (Hobson 1965, Tucker et al. 1980,Malis-Ara d et al. 1983). Someyear s ago itwa s suggested that PG activity is an essential catalyst for thewhol e tomato ripening process (Tigchelaar et al.1978 , Poovaiah &Nukay a 1979). However,othe r authors (Sawamura et al.1978 , Brady et al. 1982,Grierso n &Tucke r 1983,Crooke s &Grierso n 1983) clearly showed that PG activity increased not before 2-3 days after the onset of the respiratory climacteric as detected by an increase in ethylene production. Itwa s concluded, therefore,tha t ripening appears as a coordinated series of eventswhic h are set in trainb y ethylene.Brad y et al. (1982)hypothesize d that although PG does not have an initiating effect it canpla y a catalytic role.I t ispossibl e that the initial intracellular events, indicated by a rise in ethylene production and respiration, induce PG production.Activit y ofP G in turnma y release catalysts (seeStran d et al. 1976)previousl y inactivated on thewalls ,whic h lead to further increases in ethylene and to other intracellular responses of ripening, including e.g. the chloroplast to chromoplast transition. Ethylene production, in turn,ma y be triggered by theplan t hormone auxin (Goodenough 1981). Recently Rushing &Hube r (1985)presente d results illustrating the role of copper in thebiosynthesi s and activity of ethylene in ripening fruit. In thematurin g tomato fruit the cells are connected by themiddl e lamellawhic h is thought to contain large amounts of pectin. The non-esterified carboxylgroups are important inmaintainin g cohesionb y cross-linkingwit h divalent metal ions (calcium). In the cells of mature 19 green tomatoes a certain degree of cell separation has already occurred as a result of cell elongation during growth (Pressey 1977). Changes in the structure of cellwall s asa result of ripening startwit h the dissolution of themiddl e lamella; later amor e severe disruption of theprimar y cell wallma y occur (Crookes& Grierson 1983). Many authors showed the existence of at least two isoenzymes (Fig.3 )whic h act sequentially during ripening (Pressey &Avant s 1973,Tucke r et al. 1980,1981 ,Zaino n & Brady 1982, Crookes& Grierson 1983). PGactivit y (nMmin" 1g fruit"1)

800

2 4 6 8 10 days from onset of ethy - lene production

Fig. 3:Change si nP Gisoenzyme s duringripenin g(fro m Crookes& Grierso n1983) .

Crookes & Grierson (1983)propose d thatP G was responsible for themiddl e lamella dissolution and that the later stage inwhic h the primary cellwal l was attacked was due toP G .On e to two days after the onset of ethylene production an increase inendoplasmati c reticulumwit h ribosomes attached to itwa s detected. It ispossibl e that PG is synthesized on this rough endoplasmatic reticulum justbefor e theappearenc e of its activity (Crookes & Grierson 1983). Other authors also reported "denovo " synthesis of PG during ripening (Tucker etal . 1980,Tucke r & Grierson 1982,Brad y etal . 1982). 20

Chemicalanalysi so ftomat ocel lwall sa tsevera lstage so fripening , andth eincubatio no fcel lwall sfro mmatur egree ntomatoe swit htomat o enzymeextract so rpurifie dPG ,confirme d therol eo fP Gi nsoftening ,bu t alsoshowe dtha tbeside sth edissolutio no fpecti nothe rchange stak eplac e inth ecel lwall . (Wallner& Bloo m1979 ,Gros s& Wallne r1979 ,Presse y& Avants1982 ,Themme ne tal . 1982,Crooke s& Grierso n1983 ,Hube r1983 , Gross 1984).A ne tlos so fgalactos ean dt oa lesse rexten to farabinos e fromth ecel lwall swa sfoun ddurin gripening .Matur egree ntomat ocel l wallscontai n 12-15%galactos ewhil eth ecel lwall so frip efruit scontai n only4-7 %galactose .(Wallne r& Bloo m1977 ,Gros s& Wallne r1979 ,Gros s 1984).Dat apresente db yWallne r& Bloo m (1977)sho wtha tth etota lne t losso fgalactos ei ntomat ofrui tma yamoun tt o60 %durin gripening .Gros s & Sams(1984 )showe dtha tothe rfruit sals olos tgalactos eand/o rarabinos e duringripening .Th esolubilizatio no fgalactos eseem st ooccu r independentlyo fpecti nsolubilization .I nvitr oadditio no ftomat oenzym e extractt ocel lwall so fgree ntomatoe sfaile dt osolubiliz ea larg epar t ofth egalactos ei nadditio nt oth epecti n (Wallner& Bloo m 1977).Gros s& Wallner (1979)showe dtha tdetache drin-mutan tfrui tlos tneutra lsugar si n absenceo fP Gan dpecti nsolubilization .Bot hWallne r& Bloo m (1977)an d Gross& Wallne r (1979)indicate dtha tpar to fth egalactos elos salread y occurredbefor eth eonse to fripening .Lacke ye tal .(1980 )suggeste dtha t thedeclin ei nwal lgalacta ni spartl ydu et oreduce dsynthesi si n senescingnorma lfruit san di ndetache drin-tomatoes .I nattache d rin-tomatoesth edeclin ei ngalactos ewa smuc hsmaller .Indee dPresse y (1983)isolate da g-1,4-galactosidas eI Iwhic hwa sabl et ohydrolys ea galactanric hpolysaccharid e (Pressey& Himmelsbac h1984 )int oit s monomericconstituents .Th eincreas ei nfre egalactos etha toccur si n tomatoesdurin gripenin g (Gross& Saltvei t1982 ,Gros s1983 )ma yb ea producto fthi senzyme .Gros s(1984 )foun dtha tdecrease si ngalactos e contentoccurre d inbot hionicall yassociate dpecti n (chelatorsoluble )an d covalentlyboun dpecti n (coldalkal isoluble )bu tals oi na non-cellulosi c polysaccharidefractio nwhic hwa spar to fth ecellulosi cfractio n (resistantt oextractio nwit h4 M KOH) .Dat aindicate d thata tleas ttw o galactosecontainin gpolymer sar einvolve di nth ene tlos so fwal l galactoseresidue sdurin gripening . 21

Inadditio nt othes ecompositiona lchange si nth ecel lwall ,Hube r (1983)observe da marke dchang ei nth emolecula rweigh to ftomat ocel lwal l hemicelluloses.Hemicellulose s(ric hi nxylos ean dglucose )fro mimmatur e greenfrui twer esimila rt othos efro mmatur egree nfruit ,wherea s hemicellulosesfro mfrui tharveste ddurin gripenin gshowe dprogressivel y lowerquantitie so fhig hmolecula rweigh tpolymer san dhighe rquantitie so f lowmolecula rweigh tpolymer s (MW< 40.000).Th echange si nhemicellulose s occurredindependentl yo fpecti ndegradation . Attemptshav eofte nbee nmad et odetec ta rol efo renzyme sothe rtha n PGi ntomat ofrui tripening ;thi sbecaus ea certai namoun to fsoftenin g alreadyoccur sbefor eth eonse to fripenin gan dth eincreas ei nP Gactivit y (Hobson1965 ,Besfor d& Hobso n1972) .I nadditio nt othis ,Sawamur ae tal . (1978)showe dtha twate rsolubl epecti n (WSP)increase dbefor eP Gwa s active.The yfoun dtha ta nearl ydetachmen to ftomatoe sadvance dripening . Theethylen etreshol dfo ra nincreas ei nWS Pdecrease dwhil eth etreshol d forth eincreas ei nP Gactivit ydi dno tchange .Th eincreas ei nWS Pwa s attributed toeithe rth eactivit yo fa nenzym eothe rtha nPG ,o rt o"d e novo"synthesi so fWS P (compareKne e 1982). Thepresenc eo fsevera lcarbohydrases ,o fwhic hB-l,3-glucanas ean d 3-galactosidasear eth emos timportant ,ha sbee nreporte di ntomatoe s (Wallner& Walke r 1975,Phar re tal .1976 ,Hinto n& Presse y 1980,Presse y 1983).Th erol eo fthes eenzyme si nripenin gi sstil ldoubtful :o nth eon e handbecaus ethes eenzyme swer eno tabl et oinitiat ewal ldegradatio n (Wallner& Bloo m1977 )o rt odegrad ea tomat ocel lwal lpolysaccharid e (Gross& Wallne r 1979),o nth eothe rhan dbecaus ei twa sshow ntha tthes e enzymeswer epresen tthroughou tgrowt han dripenin g (Wallner& Walke r1975 , Pharre tal .1976) . Thesam ei smor eo rles stru efo rth einfluenc eo fcellulase so nsoftening . AlthoughHal l (1964),Dickinso n& McCollu m (1964)an dSobotk a& Watad a (1971)reporte da relationshi pbetwee nfirmnes san dcellulas eactivity , otherdat ad ono tconfir msuc ha relationshi p (Hobson1968) ;moreover ,i t hasbee nshow ntha tcellulas ei salread ypresen ti nth egree nfrui t (Buescher& Tigchelaa r 1975,Wallne r& Walke r 1979)an dwil lincreas eals o inth erin-mutan t (Poovaiah& Nukay a 1979). Pectinesterase (PE)i sno tthough tt ob edirectl yinvolve di ntomat ofrui t ripening.Hal l& Denniso n (1960)an dHobso n (1963)foun dn orelationshi p 22 betweenP Eactivit yan dfirmnes so ftomatoes ;othe rauthor sshowe dth e presenceo fP Ei ngree ntomat ofruit san di nfruit so fth eNr -an drin - mutants(Buesche r& Tigchelaa r1975 ,Sawamur ae tal .1978 ,Tucke re tal . 1982).However ,i ti spropose dtha talthoug hP Eactivit ydoe sno tresul t directlyi nsoftenin gi tma yinfluenc ecel lwal ldegradatio nb yothe r enzymes,especiall yb yP G (Tuckere tal .1982) .Indeed ,Presse y& Avant s (1982)showe dtha tpurifie dP Gwa smuc hmor eactiv eo ndeesterifie dcel l wallstha no nnativ eesterifie dcel lwall s (Degreeo festerificatio n (DE) ±47%).P Estimulate d theactio no fP Ga tp H5 ,th eoptimu mp Hfo r degradationo fdeesterifie d tomatocel lwalls ,an di nsmal lamount sa tp H 3.5,th eoptimu mp Hfo rdegradatio no fnativ eesterifie d cellwalls .Hig h PEactivit ya tp H3. 5 inhibitedcel lwal lsolubilizatio nb yP Gaction .N o datawer epresente do nth einfluenc eo fP Eo nP Gactivit ya tth enatura lp H ofth etomat ofruit .However ,i twa sconclude dtha tpecti nsolubilizatio n duringripenin gi sdu et oth eactio no fP Go nrelativel yhighl yesterifie d substrates.Thi sconclusio ni si nagreemen twit hth eresult so fThemme ne t al. (1982)wh oshowe dtha tcel lwal ldegradatio nwit hpurifie dPG T didno t differfro mcel lwal ldegradatio nwit hth etota lcel lwal lboun dprotei n extractcontainin gPG ,P Ea swel la s2-galactosidase . Hobson (1981)propose da ninaccessibilit yo fth emethoxygroup si n protopectinan dpecti nmolecules ,preventin gP Efro mquickl yan dcompletel y deesterifyingth esubstrate ,a sa nexplanatio nfo rth estil lhig hdegre eo f esterificationi nripene dfruits .I fth eenzym ecoul dac tunrestrictedl yo n allth epecti ni na tomat ofruit ,th esubstrat ewoul db ecompletel y demethoxylatedi nfou rminutes . Ionsals oma yb eimportan t intomat ofrui tripening .Will s& Rigne y (1979)foun dtha tth eactivit yo fP Gan dPE ,isolate dfrom-tomat ofruits , wasinhibite db yth eadditio no fhig hconcentration so fcalciu mions . Accordingt othem ,thi smigh texplai nth ecomplet einhibitio no fripenin g whenwhol etomat ofruit swer einfiltrate dwit hsubstantia lamount so f calcium.Rigne y& Will s (1981)hav epropose dtha tth esolubilizatio no f cellwal lboun dcalciu mi sa prerequisit efo rth einitiatio no fripenin g andtissu edegradatio nb yPG .Buesche r& Hobso n (1982)conclude dtha t solubilizationo fcalciu mregulate sth erat ean dexten to fcel lwal l degradationb yP Gdurin gnorma ltomat ofrui tripening .Mizrah i(1980 ) reportedtha tnor-mutan tfruit so nplant sgrow ni nnutrien tsolutio n 23

containing 3g/ 1NaC lexhibite d ripening responseswhic h included increases in respiration rate and ethylene production and thedevelopmen t of pectolytic activity and red colour.Rushin g &Hube r (1984)reporte d similar effects of calcium and sodium on the solubilization of galacturonic acid from cellwall s under the influence of cellboun d PG. The extent towhic h the fruits are damaged during harvesting (especially withmechanica l harvesting) and the timebetwee n harvest and processing isver y important forbot h yield and product characteristics (York et al. 1967,Mier s et al. 1971,Mars h et al. 1979 b). Damage to the tomatoes during harvest or transportation,an d storage of thedamage d tomatoeswil l result in large serum losses and in reduced serumviscosities .A s a result of this, the concentration of insoluble solids in thejuic ema y increase giving rise to increased grossviscosity . However,th eyiel d willb ever y low and serum separation may be negatively influenced by theactio n of enzymes on serum and insoluble solids. The effects of other factorswhic h might influence theviscosit y potential of tomatoes,lik e cultural practices (timeo f planting,metho d of planting,us e of fertilizers) and climate (temperature,amoun t of sun),ar e notwel l documented andwil l not be discussed here.

2.3.3. Polysaccharide degrading enzymes in tomatoes

Ripe tomatoes contain several types ofpolysaccharid e degrading enzymes ofwhic h the pectin degrading enzymes PE and PG (Fig.4 ) are present inver y large amountswhe n compared to other fruits (Hobson 1962, Polacsek-Racz &Pozsar-Hajna l 1976). Because of this abundant presence of pectolytic enzyme activity, tomatoes have oftenbee nuse d to characterize thenatur e of plant pectolytic enzymes.O n the other hand itha s resulted in the fact that foryear s the emphasiswa s put on the importance of pectin and pectin degradation as factors determining theviscosit y of tomato products. Theoretically,however ,al l thepolysaccharid e degrading enzymes in tomatoes are able todegrad e their substrates and influence product viscosity from themomen t the tomatoes are crushed. 24

PE concentration (U g"1) to-f PG concentration (arbitrary units)

30- •110

PE D P Jl 1 PG

1 Jl * JL I 6 7 9 10 Tl 12 13 14 =•- -fc ?" o a

Fig. 4: PG and PE activity in fruits and vegetables 0 ~ PE activity on low-ester pectin, P " PE activity on high-ester pectin (from: Polacsek-Racz & Pozsar-Hajnal 1976).

The different polysaccharide degrading enzymes will be discussed in the following part.

Polygalacturonase (PC)

Ingeneral ,polygalacturonase sar edivide dint otw ogroups :exo-P G (EC. 3.2.1.67)remove ssingl egalacturoni caci dunit sfro mth enon-reducin g endo fpolygalacturoni cacid ;an dendo-P G (EC.3.2.1.15)hydrolyze sth e substratea trando mconvertin gi tint oa mixtur eo fdime ran dmonomer . PGactivit yi ntomatoe sarise sfro ma tleas ttw oisoenzyme s (Pressey& Avants1973 ,Tucke re tal .1980 ,Zaino n& Brad y 1982,Moshref i& Lu h1984 ) whichac tsequentiall ydurin gripenin g (Tuckere tal .1980) .P G ,whic h actsa tth ebeginnin go fth eripenin gprocess ,i sthough tt ob ea dime ro f PG (Tuckere tal .1981) .Polyacrylamid ege lelectrophoresi sshowe dtha t

PGT ,th edominan tP Gisoenzym ei nrip etomatoes ,consist so ftw oseparat e polypeptides (Zainon& Brad y 1982,Crooke s& Grierso n1982 ,Moshref i& Lu h 1984).Th eexistenc eo fmor etha non eP Gisoenzym ewa salread yindicate db y McColloch& Kertes z (1948)wh ofoun da residua lP Gactivit y (20%o finitia l 25 activity)afte r5 minute sheatin gi nboilin gwater .Indeed ,severa lauthor s showeddifferen thea tstabilitie sfo rP G andP G .50 %inactivatio no fP G isobtaine dafte r5 min .heatin ga t70-78° Cwhil eP G isalread y inactivatedfo r50 %afte r5 min .52-57° C (Pressey& Avant s1973 ,Tucke re t al.1980 ,Zaino n& Brad y 1982,Moshref i& Lu h 1984).Bot hisoenzyme sar e basicglycoprotein s (Takehanae tal .1977 ,Zaino n& Brad y 1982,Moshref i& Luh1984 )wit ha reporte dp io f8. 6 forP G and9. 4 forP G (Zainon& Brady 1982).Th eoptimu mp Hfo rhydrolysi so fpolygalacturoni caci di s approximately4. 5bu tthi svalu edecrease swit ha decreas ei nsubstrat e molecularweigh t (Patel& Pfaf f196 0a,b ,Presse y& Avant s1972 ,1973) . Studieswit himpur etomat oP Gindicate d thatdurin gth efirs t25-30 % ofhydrolysis ,polygalacturoni caci dchain sar ecleave da trandom ; oligogalacturonic acidswer eth eproduct so fhydrolysis ,n ofre e galacturonicaci dwa sfound .A t80 %hydrolysi smono -an ddigalacuroni caci d wereth eonl yreactio nproduct sfound ;th edigalacturoni caci dwa scleave d very slowly.I fth einitia lreactio nvelocit ywit hpolygalacturoni caci di s seta t 100,thos ewit htetra- ,tri- ,an ddigalacturoni cacid sar e approximately7.0 ,1.5 8an d1.05 ,respectivel y (Luhe tal .1956) . Non-purified tomatoP Gattacke d tetra-an dtrigalacturoni caci dfro mth e reducingendgrou p (Patel& Pfaf f196 0b) .McCread y etal .(1955 )showe d alsotha tmonogalacturoni caci dwa sth efina lreactio nproduc twhe n polygalacturonicaci dwa sincubate dwit hcrud etomat oPG ,indicatin gsom e exo-PGactivity .However ,al lth epurifie d isoenzymeso ftomat oP Gar e apparentlyo fth eendo-typ e (Pressey& Avant s1973 ,Markovi c& Slezari k 1977, 1981,Takehan ae tal .1977 ,Dongowsk ie tal .1980 ,Zaino n6 .Brad y 1982).Presse y& Avant s (1973)reporte dtha tP G accomplished in5 min . thesam eviscosit ydecreas eo fa pecti caci dsolution ,wit hth esam e increasei nreducin gendgroups ,a sP G after3 0min .Thi sindicate sa differentmod eo factio nfo rth etw oisoenzymes .Th eoptima lsubstrat efo r PG wasfoun dt ob ea full ydemethylate dcitru specti cacid .Th eenzym e stillshowe d± 20 %o fth eoptimu mactivit yo na substrat ewit ha D Eo f60 % (statisticallydistributed) ;n oactivit ywa sfoun do na substrat ewit ha D E of80-90% .P G showedhighe ractivitie so npartia lesterifie d polygalacturonicacid s (DE<32 )wit ha block-wis edistributio no fth e methylestergroup swhe ncompare dt osimila rpolygalacturoni cacid swit ha 26 statisticaldistributio no fth eeste rgroup s (Dongowskie tal . 1980).A positivecorrelatio nwa sfoun dbetwee nP G activityan dsubstrat e molecularweigh t (Dongowskie tal .1980 ,Markovi c& Slezari k1977 ,1981) . Themod eo factio no noligogalacturoni cacid swa ssimila rt otha treporte d forimpur eP G (Markovic& Slezari k1981 )wit hth eexceptio ntha tth efina l reactionproduct swer emono -an ddigalacturoni cacid ;digalacturoni caci d wasno tcleaved .

Bi , B2 I (XXX) (XXX)

*

IWfVI

CXXX># (XXX>4 IWTYI rwrvn Fig. 5: Enzyme-substrate complex of tomato PG with oligogalacturonic acids and their reduced derivatives A- trigalacturonic acid, B andB - tetragalacturonic acid, C» redu­ ced tetragalacturcnic acid, D and D « reduced pentagalacturonic acid, 1-3 » binding subsites, arrow - catalytic site (from: Markovic &Slezari k 1981). Markovic & Slezarik (1981) concluded that the primary binding site of tomato PG is composed of three subunits and the catalytic site is situated close to the first bond from the reducing end of the substrate segment bound into the complex (Fig. 5). Pectinesterase (PE) Pectinesterase (E.C. 3.1.1.11) catalyzes the hydrolysis of methylester groups present in pectinic acids and pectin. The tomato pectinesterase complex is composed of five to eight multiple forms as is shown by Delincee (1970) and by Rexova-Benkova et al. (1977). Only one type is dominant in ripe tomatoes (Pressey & Avants 1972, Tucker et al. 1982). The enzyme attacks the pectin chain from the reducing 27 endan da tothe rsite so nth esubstrat emolecule ,mos tprobabl ynex tt oa freecarboxy lgroup ;th eenzym eproceed slinearl yalon gth echai n (Lee& MacMillan1968 ,1970) .Deesterificatio nresult si nth eformatio no f segmentsi nth epecti nmolecul ewit ha block-wis earrangemen to ffre e carboxylgroup salternate dwit hsegment scontainin gfull yesterifie d galacturonicaci dunit s (Kohne tal . 1983).Th epartiall ymethylate d (60%) pentagalacturonatewa sfoun dt ob eth eshortes tsubstrat edeesterifie db y tomatopectinesteras e (Markovice tal .1983) .Th einitia lreactio nrat eo n thepentamer ,however ,wa sonl y4.58 %o fth einitia lrat eo na hig h molecularweigh tpecti nwit ha D Eo f65% .I ngenera lth ereactio nrat e increaseswit ha nincreas ei nsubstrat emolecula rweight .(Markovi ce tal . 1983).Polygalacturoni caci dan dreactio nproduct so foligomeri csubstrate s competitively inhibitth eactivit yo ftomat opectinesteras e (Lee& MacMillan 1968,Markovi ce tal .1983) .Markovi ce tal .(1981) ,however , foundtha tth epurifie ddominan tfor mo fP Ewa sabl et odeesterif ya purified citruspecti n (DE64.3% )withi nthre ehour st oa D Eo f1. 8± 0.5 % withoutdegradatio no fth egalacturoni caci dchains .

activity (% 100 1 -10 seconds 2-20 seconds 3-40 seconds 4-1 minute 5-2 minutes 6-5 minutes 7-10 minutes 8-20 minutes 9-40 minutes 10-1 hour

70 90 temp. (°C)

Fig.6 :Therma l inactivationo ftomat oP E (from:Draett ae tal .1979) . 28

TomatoP Ei sa basi cprotei n (Delincee& Radol a1970 )wit ha p H optimumi nth erang e7- 8 (Lee& MacMilla n 1968,Nakagaw ae tal .1970 , Pozsar-Hajnal& Polacsek-RSc z1975 ,Draett ae tal .1979 ,Tucke re tal . 1982).Theenzym ei sstil lactiv ea tth enatura lp Hi ntomat ofruit ; however,th eactivit yi sdependen tupo nth eD Eo fth esubstrate . Pozsar-Hajnal& Polacsek-RSc z (1975)reporte dlac ko fP Eactivit yo na 70 % DEsubstrat ebelo wp H5.0 ;however ,o na substrat ewit ha D Eo f30-36 % tomatoP Ewa sstil lactiv ea tp H4. 5 (±30%o factivit ya toptima lpH) . 50%o ftomat oP Eactivit yi slos tafte r5 min .heatin ga t60-67° C (Pressey& Avant s1972 ,Nakagaw ae tal .1970 ,Tucke re tal .1982) ;dat ao n timean dtemperatur ecombination sfo rcomplet eP Einactivatio nar e presentedb yDraett ae tal .(1979 )(Fig . 6).

Celluloses Thecellulas egrou po fenzyme si sclassified ,accordin gt oth emod eo f action,i nendo-glucanase ,exo-glucanas ean dcellobiase . Endo-glucanase(E.C .3.2.1.4. ), 1,4-g-D-gluca nglucanohydrolase ,hydrolyse s cellulosepolymer sa trandom ,th eminimu msubstrat esiz ebein ga trimer . Theenzym ei smainl yactiv eo namorphou scellulos ean do nsolubl e derivativesa scarboxymethylcellulos e (CMC);i trapidl ydecrease sth e viscosityo fCM Csolutions . Exo-glucanase (E.C.3.2.1.91) ,1,4-B-D-gluca ncellobiohydrolase ,attack s fromth enon-reducin gen do fth epolyme ran drelease smolecule so f cellobiose.I ti sactiv eo ncrystallin ecellulos e (Avicel,Whatma n cellulosepowder )bu tshow smino ractivit yo nCMC ;activit ydecrease swit h increasingchai nlength . Cellobiase (E.C.3.2.1.21) ,a 6-glucosidase ,hydrolyse scellobios eint otw o glucosemolecules . Selby& Maitlan d (1967)showe da larg esynergisti ceffec twhe nth e cellulasegrou po fenzyme sproduce db yTrichoderm avirid eacte dtogethe ro n cottonfibre .I ti sprobabl etha tendo-glucanas eprovide sth enon-reducin g endgroup snecessar yfo rexo-glucanas eactivity .Cellobiose ,whic hinhibit s bothendo -an dexo-glucanas eactivity ,i sdegrade db ycellobiase . Thepresenc eo fth eendo-typ eo fglucanas ei ntomatoe sha softe nbee n reported (Dickinson& McCollu m1964 ,Hal l& Mullin s1965 ,Hobso n1968 , Sobotka& Watad a1971 ,Buesche r& Tigchelaa r 1975,Wallne r& Walke r1975 , 29

Poovaiah &Nukay a 1979). Information on the existence of exo-glucanase and cellobiase,however ,i srathe r limited.Phar r &Dickinso n (1973)reporte d activities of endo-glucanasean d cellobiase in tomatoes. Cellobiase weakly inhibited endo-glucanase activity. Cellobiasewa s found tob e lesshea t stable thanendo-glucanase .A 50%cellobias e inactivation at 50°Cwa s obtained after 3min . compared to 40min . for endo-glucanase.Muc h of the information on theoccurrenc e of a tomato cellulase-complexi s obtained from a studyb y Sobotka & Stelzig (1974). They detected two fractions with endo-glucanase activity (CMC-liquifying),on ewit h ap H optimum of 7.5 the otherwit h ap H optimum of 5.2.Moreover ,thes e fractionswer e able to hydrolyse Avicel,Whatma n fibrous cellulose powder and beanhypocoty l cell wallmaterial .Th e fractionwit h a pH optimum of 7.5 showed properties pointing to thepresenc e of some exo-glucanase activity. This enzymewa s active on insoluble celluloses; it degraded short chain cellulodextrins (except cellobiose)an d showed CMC-saccharifying activity which was indicated by the release of largeamount s of reducinggroups . Inadditio n to the endo-glucanase,a no n specific B-glucosidase,activ e on a number of short chain cellulodextrins and p-nitrophenyl-B-D-glucoside, was isolated.Th e activity of this enzyme decreased with an increase in chain length.Moreover ,a n exo-B-1,4-glucanase,highl y active on cellobiose,cellotriose ,cellotetraos e but not onp-nitrophenyl- B -D-glucoside,wa s found.Th e total complex of tomato cellulasewa s able to degrade completely the insoluble cellulose from different sources.

Other glycanases Only a few authorshav e studied thepresenc e in tomatoes of glycanases other than thosebelongin g toth e cellulase complex.On e of the glycanases found in tomatoes is B-l,3-glucanase (laminarinase,E.C . 3.2.1.6), an enzymewhic h hydrolyses the glycosidic bond in a 3-1,3 linked glucose polymer. This enzyme,wit h aM W of 12.000, ispresen t in the same quantities asP G (Wallner &Walke r 1975). B-l,3-glucanaseha s ap H optimum in the range 5-6 and israthe r heat stable,a s it is inactivated for only 15% after 5min .heatin g at 60°C (Hinton &Presse y 1980).Wallne r &Walke r (1975)foun d only little or no activities on substrates such as glucomannan,galactomanna n (guar gum),g-l,6-gluca n (postulan), arabinogalactan (stractan)an d xylan. 30

Glycosidases Usuallyth epresenc eo fglycosidase si steste do np-nitropheny l glycosides.Usin gthes esubstrates ,activitie so f a- andg-galactosidas e (Wallner& Walke r1975 ,Phar re tal .1976 ,Presse y 1983),a-mannosidas e (Pharre tal .1976 )an d g-glucosidase(Phar re tal .1976) ,wer efoun di n tomatoes.Th eg-glucosidas e showedonl ywea kactivitie so ncellobios ean d gentobiose (3-1,6linkage) ,whil eth eactivit yo nlaminari nwa ssomewha t higherbu tstil lno tmor etha n10 %o fth ep-nitrophenyl-g-D-glucosidas e activity.On eo fth eg-galactosidase sisolate db yPresse y (1983)wa s capableo fhydrolysin ga polysaccharid etha twa sisolate dfro mtomatoe san d thatconsiste dprimaril yo fg-1, 4linke dgalactose .N oo ronl ywea k activitieso fa-D-glucosidase , a- and g-L-fucosidase,a -an dg-D-xylosidas e and g-D-mannosidasewer edetecte d intomatoe s (Wallner& Walke r1975 ,Phar r etal .1976) .

Toobtai na tomat oproduc to fhig hviscosit yal lthes eenzyme shav et o beinactivated ,eithe rbefore ,durin go rimmediatel yafte rcrushing ,i n ordert opreven tthei rdegradativ eactio no ntomat opolysaccharides .Tomat o enzymesca nb einactivate db yhea t (Kertesz& Locont i1944 )o rb yadditio n ofenzym einhibitor s (Wagnere tal .1967) ;th eforme rmethod ,however ,i s commonlyused . Duringhea tinactivatio no fth eenzyme sth etemperatur ealway spasse s througha rang ewher eenzym eactivit yi soptimal .Therefore ,th eheatin g rateshoul db ehig henoug ht opreven ta lon gresidenc etim eo fth ecrushe d tomatoesa tth eoptima ltemperatur efo renzym eaction .Moreover ,th efina l temperatureshoul db ehig henoug hfo rcomplet eenzym einactivation .Th e temperaturesrecommende d forth ehea tinactivatio no fenzyme sha schange d overth elas t4 0year sfro m82° C (Kertesz& Locont i1944 )t o85° C (McColloche tal .1950) ,93° C(Twig g1959 )an dhighe rtha n93° C(Lu h& Daoud 1971).Tomat oenzyme sca nals ob einactivate d"i nsitu "b ya hea t treatmento fth ewhol etomatoe sbefor ecrushing .Thi smetho dwa sfoun dt o bepreferabl et oth eheatin go fcrushe dtomatoes ,unles sth efina l temperatureo fth eheate dtomatoe swa sinsufficien tfo rcomplet eenzym e inactivation (Kertesz& Locont i1944 ,McColloc he tal .1952 ,Coole r1962 , Crandall& Nelso n 1975).Becaus eo fth erapi dactio no fpectolyti cenzymes , Wagneran dMier s (1967)searche dfo renzym einhibitor st opreven tth e 31

enzyme catalyzed loss Inviscosit y during preparation of tomatojuices . Strong acidswer e themos t effective inhibitors among the compounds tested.

2.3.4. Contribution of compositional factors to gross viscosity

Contribution of water soluble solids (serum viscosity) to gross viscosity

Kertesz& Locont i (1944)conclude d that aviscou s serum is an indispensable prerequisite for a tomato juice of desirable grossviscosity . Results of serum exchange experiments and experimentswit h enzymatically degraded serumviscosit y or artificially increased serumviscosity , led to the conclusion that serumviscosit y strongly influences grossviscosity . In the serum thepresenc e or absence of pectic constituentswa s thought tob e of great importance.Robinso n et al. (1956)suggeste d that grossviscosit y isth e result of "particle"viscosit y and serumviscosity . Their assumption wasbase d on analogywit h colloidal solutionswher eviscosit y isrelate d to theviscosit y of the solvent and thevolum e of the colloid insolution . Cooler (1962)develope d a theory of awea k serum gelwhic hwa s thought to be necessary forhig h grossviscosity . The formation of the serum gelwa s dependent upon thepreservatio n of solublepectins .Othe r authors also reported high correlation coefficients between the serumviscosit y and grossviscosit y of tomatopul p or paste samples (Stevens &Paulso n 1976, Marsh et al. 1980,Bel-Ha j 1981,Lu h etal . 1984). Stevens &Paulso n (1976) found indications that,i nmos t situations,eithe r serumviscosit y or gross viscosity could be used with considerable confidence to evaluate genotypic variation inviscosity .Mars h et al. (1980) showed thatbot h serum viscosity and theWIS/T S ratiomad e an important contribution to the Bostwickvalu e of a juice.O n thebasi s of a theoreticalmode l they predict the effect of changes in serumviscosit y on the grossviscosit y of juices and concentrates.Bel-Ha j (1981)considere d serumviscosit y tob e an important factor inalterin g grossviscosity ,Lu h et al. (1984 a) found a high correlationbetwee n gross and serumviscosit y and concluded thatbot h pectinic and pectic acid fractions contribute to the grossviscosit y of canned juices andpastes . 32

In contrast to theabov e cited authors,Han d et al. (1955)foun d no relationship between serumviscosit y and grossviscosity . Injuice s of uniform grossviscosit y and partical size,a 100%differenc e in serum viscosity had no perceptible effect on grossviscosit y as determined by a panel.The y showed that serumviscosit y was approximately 1%o f thegros s viscosity of ajuice .Simila r resultswer e reported by Robinson etal . (1956), Twigg (1959)an dMars h et al. (1979 b).Robinso n et al. (1956) found that serumviscosit y isproportiona l to the soluble pectin content and they stated that pectin is the only constituent contributing to serum viscosity. In general it appeared thatwhe n the pectin content of the serum washig h theamoun t of pectinassociate d with the suspended particles was alsohigh .Thi s isprobabl y the reason for a correlationbetwee n serum viscosity and grossviscosity . However,Robinso n et al. (1956)prov e that by adjusting preheating and finishing conditions,juic e samples canb e prepared withver y different grossviscosity/seru mviscosit y ratio's, indicating that serumviscosit y doesno t determine grossviscosity . This finding issupporte d by Twigg (1959),wh o concluded that serum viscosity was completely unsatisfactory as a single predictor of gross viscosity. Very low coefficients of correlation (r=0.29)wer e reported between ketchup grossviscosit y (sensory panel) on the onehan d and serumviscosit y (Ostwald-Cannon-Fenske pipette) on the other hand.Mars h et al. (1979 b) presented resultswhic h indicate that low serumviscositie s asa result of enzyme action during crushing didno t result in an increased ketchup yield factor. This factorwa s defined as the percentage of tomato solids required tomak e a standard batch ofketchu p having a Bostwickvalu e of 6 cma t33 % totalsolids .

Whittenberger &Nuttin g (1958)eve n pointed to thenegativ e effect of the serum on grossviscosity . According to them the serum contains substances (electrolytes)whic h inhibit thedevelopmen t ofmaximu m gross viscosity inwhol ejuice . Becker et al. (1972)reporte d that individually,neithe r the serum soluble polymers nor the cellwal l fibers can cause gross viscosity changes. In thepresenc e of intact cells substantial increases in the serum viscosity resulted inmino r grossviscosit y increases. Increased amounts of dispersed (=homogenized) cellwal l fibers also resulted inmino r viscosity increases,unles s ahig hviscosit y serumwa s present.Whe n a 33 juice contained both large amounts of dispersed cellwal l fibers and a high viscosity serum,th e grossviscosit y was greatly elevated. These results may indicate that the influence of serumviscosit y on gross viscosity depends upon thephysica l state of the cellwalls .

Contribution of water insoluble solids (WIS) or suspended particles to gross viscosity

Many authors report on the influence ofwate r insoluble solids or suspended particles on the grossviscosit y of tomato products.Th e majority of them agreewit h the conclusion ofKertes z &Locont i (1944)tha t the proportion of insoluble solids,a swel l as their shape,siz e and other properties,ar e of importance in determining the grossviscosit y of tomato juice.Whittenberge r &Nuttin g (1957, 1958)dissecte d tomatoes into four fractions: shell tissue 48%,placent a or center tissue 26%,fre e juice occurring in thevicinit y of the seeds 18%,an d gelatinous envelopes surrounding the seeds8 % (Fig.7) .

shell tissue free juice center tissue seed envelope

Fig.7 :Cross-sectio no ftomat ofrui t (from:Whittenberge r& Nuttin g 1957).

Juices from the outer shell and from the center tissueswer e moderately thick and contained moderate quantities of cellwalls .Juic e from the gelatinous seed envelopeswa s thick and contained cellwall s heavily impregnated with pectin. Free juice from the seed cavitieswa s exceedingly thinan d was devoid of cellwalls .The y concluded that the cell wall, comprising only 6% of the total solids and less thanhal f of the insoluble solids,wa s the structuremos t closely related to the grossviscosit y of tomatojuice . 34

York et al. (1967)foun d ahig h correlationbetwee n the amount of insoluble solids and theBostwic k consistency of tomato products.Bartolom e (1972) also concluded that the amount of insoluble solidsha d the greatest influence onjuic e grossviscosity ,bu t that the effects of serum viscosity, titratable acidity and thenatur e of the suspended particles were also important.Fod a &McCollu m (1970) reported a great reduction in grossviscosit y whenwashe d solidswer e resuspended in their original serum. They concluded that highmolecula r weight polymerswit h ahig h ratio saccharide/uronide and associated with the insoluble solidsma y play an important role indeterminin g grossviscosity . Marsh et al. (1980) studied the influence of composition on the rate of change in grossviscosit y during concentration. They found that the change inBostwic k consistency of tomato juice as result of a change in solids content during concentrationwa s related to theWIS/T S ratio and to theviscosit y of the serum of clarified juice.Thes e characteristics were found tob ewidel y variable amongne w tomato cultivars.Th e slope of the concentration curve (logBostwic k = A % solids+ B,Lu h 1954)wa s primarily dependent upon theWIS/T S ratio; the locus of the concentration curve,o n the other hand,wa s found tob e dependent onbot hWIS/T S ratio and serum viscosity and could be located by theBostwic k consistency of thejuic e and initial solids content.WIS/T S ratio isa function of cultivar,hea t treatment and finisher screensize,whil e the serumviscosit y depends upon theeffectivenes so f thehea t treatment in destroying the enzymesystems . Marsh (1977)reporte d that for 67 samples the relationship between WIS/TS ratio and Bostwick value at 12°B xha d a correlation coefficient of -0.95. Luh etal . (1984b,c )als o considered theWIS/T S ratio indicative for pulp and paste viscosity. Marsh et al. (1979 a,b) tried to study and determine the factors responsible for the quality and grossviscosit y of ketchup.Fo r this purpose they developed ametho d of determining the percentage of tomato solids required tomak e a standard batch of ketchup having aBostwic k value of 6 cma t 33%tota l solids.Th e amount of tomato solids expressed as% of the total solidswa s defined as theyiel d factor.Gros sviscosit y and serum separationwer e found tob e unrelated quality attributes.Gros s viscosity was found tob e directly dependent upon the fractiono fwate r insoluble solids of the total solids of the tomato juiceused . Ifketchu p was prepared from low solids pasteswhic h require almost no dilution,a good 35 correlationwa s found betweenWIS/T S ratio of thejuic e before concentration and theyiel d factor (r« -0.97). Multiplying ratio and yield factor yielded a constant factorK of 5.38 ±0.1 2 (for 24 different varieties). In fact,thi s factor indicates that allketchup swit h33 % solids and aBostwic ko f 6 cm contained the sameWI S content.Thi s constant also shows that theyiel d factor canb e predicted when theWIS/T S ratio of thejuic e isknown .Mars h et al. (1982, 1983,1984 )als o studied the usefulness of theWIS/T S ratio of juice as a reliable predictor for the grossviscosit y potential of ahig h solidspaste . Itwa s found that the WIS/TS ratio of juice doesno t predict the grossviscosit y potential of high solidspaste s as accurately as the grossviscosit y potential of a low solids paste.I t isprobabl e that concentration,storag e and subsequent dilution altered thephysica l properties of the carbohydrate polymers. Marsh et al. (1983)presente d resultswhic h show that a standard batch of ketchup prepared from ahig h solids paste (24-33%TS)neede d approximately 9%mor e tomato solids than abatc h prepared from a low solids paste (13-17%TS).Mars h et al. (1977)reporte d that themor ewate rwa s required todilut e a concentrated sample to 12° Bx thehighe rwa s theBostwic k value reported for theproduc t and the further itdeviate d from thevalu ebefor e concentration.Th e extent of dessication of thewate r insoluble solids and their inability to resorb maximally ispropose d as the cause of the phenomenon. Smit &Nortj e (1958)foun d that theholdin g of crushed tomatoes at 21-27°Cbefor e preheating resulted inhighe r grossviscositie s ofjuice , puree and paste.A s theWI S contentwa s rather similar for sampleswit h and without holding, itwa s suggested that the difference in grossviscosit y was due toa qualitative (structural)differenc e in insoluble solids. Hand et al. (1955), studying the effects of process conditions on grossviscosity , concluded that the quantity of cellularmateria l alonewa s not the only factor governing grossviscosity ; the configuration or shape of the cell fragmentswa s also important.Durin g juicemanufactur e the cellsma y become separated,distorte d and broken,althoug hman y of the structural parts are still recognizable in the finished juice.The y agree withWhittenberge r &Nuttin g (1957)that ,i n general,sheetlik e or rodlike walls orwal l fragments offermor e resistence to flow and givea more stable juice than spherical cells.Th e irregularity of cellwal l form 36 depends largely on themechanica l treatment thewall s receive during juice manufacture.Forcin g cellwall s through passages of small clearance,an d other types of shearing,stirrin g orbeatin g actions increase linearity (Whittenberger& Nuttin g 1957). Kertesz &Locont i (1944)reporte d that the particles suspended in tomato juiceswer e of irregular shape and to a certain extent resembled cogwheels.Thes e irregularitieswer e explained by a number of factors such as tearing of the tissue duringmanufactur e and swelling oruneve n corrosion during heating. Becker et al. (1972) studied the effect of pH during breaking on the grossviscosit y of tomatojuice . Juices prepared at thenatura lp Hha d almost nobroke n cells;juice s prepared at lowp Hha d very high grossviscositie s but alsohig h amounts of brokencells . In contrast to all these authorsMcColloc h et al. (1950)foun d no relationship between insoluble solids content and paste gross or serum viscosity asmeasure d with thepenetromete r and the rate of filtration test. Twigg (1959)calculate d that 49%o f thevariation s inketchu p gross viscosity as determined by apane l could be attributed to the compositional factors,structura l solids,no n structural solids,tota l pectin and pH.Th e relative influence of the structural solids (=tota lAlS-tota lpectin )wa s only14% .

Contribution of AIS to gross viscosity

The importance ofAI S as a contributor to the grossviscosit y of tomato products iswel l recognized. However,AI S ismainl y used either for the calculation of other components or for the extraction and determination of totalpecti n and pectin fractions.Twig g (1959)use d theAI S contents of ketchups to calculate the structural solids (=AIS-totalpectin )an d non-structural solids (=tota l solids-AIS). Cooler (1962) calculated the particulate-AIS fractionb y substracting the serum-AIS from the total-AIS. He observed increases in allAI S fractions and increases in the gross viscosity of tomato juice as a result of changes inprocessin g conditions. McColloch etal . (1950)an d Luh et al. (1954 a,b) found good correlations betweenAI S content on the one hand and thegros sviscosit y of tomato paste 37 prepared from SanMarzan o and Pearson tomatoes on theothe rhand . Both authors pointed to the fact that constituents in theAI S other than pectin were important contributors to grossviscosity . Luhe t al. (1954 a,b) showed that pastewit h a pectin content of 1.53% (drybasis )prepare d from SanMarzan o tomatoesb y a coldbrea k process had ahighe r gross viscosity (Bostwick at 12°Bx)tha npast ewit h 3.48% pectinprepare d from Pearson tomatoesb y theho tbrea k process.Base d on anAI S composition of 10% pectic substances,30 %crud e fiber,5 % ash and 20-50%nitroge n asprotein , McColloch et al. (1950)propose d that proteinmigh t be an important component contributing togros sviscosity . Wagner etal . (1969) reported increased grossviscosit y and AIS content of juices as a result of the addition of acid during theho tbreak . They ascribed these effects to quicker enzyme inactivation andbette r extraction and dispersion of constituents. Alcohol insoluble solids content of tomato fruits and the gross viscosity of juicewer e found tob e highly correlated (Janoria & Rhodes 1974). The correlation coefficient values,base d on 12cultivar swere : r= 0.97 forAI S ofwhol e fruit extracted with 50%ethano l and r= 0.94 for extractionwit h 75%ethanol .Th e correlation ofAI Swit h gross viscosity was high for outer pericarp,r = 0.93 and inner pericarp,r = 0.78 but low for locular contents,r = 0.18. Janoria &Rhode s (1974)conclud e that a cause-and-effeet relationship existsbetwee n a certain fraction ofAI S and gross viscosity. Stevens and Paulson (1976)use d genotypic variation in composition and viscosity to study the effects of thevariou s components of theAI S on viscosity. They state that any attempt to isolate a component of tomato juice and todetermin e the relative effect of this component onjuic e gross viscosity will destroy interrelationships between components and in this way undoubtedly affect the grossviscosit y expression of these components. After preparing AIS from the different tomato puree samples they fractionated theAI S according to the reproduced scheme (Fig.8) . Amounts ofpolysaccharide s (PS)an d polygalacturonides (PG)wer e determined colourimetricallywit h anthron and carbazole respectively. TotalAI S was found to correlatever y wellwit h grossviscosit y (r= 0.94). Stepwise regression analysis of the data showed thatwate r soluble and water insoluble polygalacturonides accounted formos t of thevariatio n in gross 38

viscosity (WSPG,WISPG) .Th edat aindicate dtha twate rinsoluble ,pectino l solubilizedpolysaccharide s (WISPS)hav eth epotentia lfo rmakin ga larg e contributiont ogros sviscosit ya thighe rconcentrations .Th ewate rsolubl e polysaccharides (WSPS)an daci dhydrolyse dpolysaccharide s (AHPS) contributed littlet ogros sviscosity .

AIS

boiling H20

WSS WIS

NQOH/EDTA NQOH/EDTA pectinol pectinol

| acid WSPS WSPG AHPS

WISPS WISPG

Fig.8 :Schem efo rfractionatio no fAI S (from:Steven s& Paulso n 1976).

Contribution of pectin to gross viscosity

Mosto fth einformatio no nth econtributio no fpecti nt oth egros s viscosityo ftomat oproduct si sobtaine d fromstudie srelate dt oth e inactivationo fpecti cenzyme sb yhea t (Kertesz& Locont i1944 ,Robinso ne t al. 1945,McColloc he tal .1950 ,Lu he tal .1954 ,Lu h& Daou d 1971,Garce s & Luh1972 ,Sherka t& Lu h1976 ,Sherka t& Lu h1977 ,Bel-Ha j1981 ,Lu he t al. 1981,1982 ,198 4a ,198 4b ,198 4c )o rb yaci d (Wagnere tal .1967 , Mierse tal .1967 ,Becke re tal .1968 ,Wagne re tal .1968 ,Wagne re tal . 1969). Th eus eo fhig hho tbrea ktemperature sresult si nquic kinactivatio n ofth epectolyti cenzymes ,a hig hretentio no fbot htota la swel la sseru m pectinsan da hig hgros sviscosit yo fjuice ,pure eo rpaste .Accordin gt o Kertesz& Locont i (1944)th esolubl epecti nwil ldetermin eth eseru m viscositywhil eth einsolubl epecti csubstance swil ldetermin eth e properties (e.g.separation )an dth econtributio nt ogros sviscosit yo fth e suspendedparticles .Failur et oinactivat epectolyti cenzyme searl yi nth e processingprocedur eresult si na rapi ddestructio no fth epecti c substancesi nth ecrushe dtomatoes ,whic hi ntur nresult si npaste so flo w 39 pectin content and low grossviscosit y (McColloch et al. 1950). The addition of acid during theho tbrea k influences juice gross viscosity by at least 2mechanisms :a )juic e grossviscosit y and pectin content increase as a result of retarded or prevented enzymeactio nan db ) juice grossviscosit y and pectin content increase as a result of increased extraction or improved dispersion of tomato components (Wagner etal . 1969). Twigg (1959) studied the factors contributing to the gross viscosity ofketchup .Multipl e regression analysiswit h the fourvariables : structural solids,no n structural solids,tota l pectin and pH,yielde d a correlation coefficient of 0.70. Thismean t that approximately half of the variation inketchu p grossviscosit y as detected by a sensory panel could be attributed to these four factors.Th e relative influence of the factor totalpecti n on the explainable grossviscosit y variationswa s found tob e 44%. Luh etal . (1954, 1981,1982 ,198 4a , 1984b , 1984 c)reporte d that tomato juices and diluted tomato pasteswit h thehighes t total pectin content also had thehighes t grossviscositie s asmeasure d with the Bostwick consistometer,Storme rviscosimete r orLibb y pipette.Lu h etal . (1954)foun d that Stormerviscosimete r readings of commercial diluted paste samples at 10°Bx soluble solids correlated wellwit h thepecti n content on a dryweigh t basis (r= 0.88). On semi-logaritmicbasi s the coefficient of correlation increased to 0.97. The pectic substances represent a complexmixtur e of polysaccharides which canb e partially extracted from the cellwall swit hwate r (pectinic acidswit h a high degree of esterification) orwit h aqueous solutions of chelating agents such as EDTA, sodium oxalate or sodiumhexametaphosphat e (pectic acidswit h a lowdegre e of esterification).Approximatel y 60-80% of the totalpecti c substances present in the cellwall s of parenchymatous tissues are extractable.Th e insoluble pecticmateria l isalway s found in close associationwit h other cellwal l constituents,particularl y the -cellulose fraction (Luh et al. 1984 a).Bot h pectinic and pectic acid fractions of tomatoes contribute to the grossviscosit y of canned juice and paste.Highe r grossviscosit y isno t only explained by the larger quantity ofpecti n retained after processing but alsob y the weight-average molecular weight of thepecti c fractions.Large r carbohydrate polymers are 40 capableo fholdin gmor ewater ,resultin gi nhighe rgros sviscosit y (Luhe t al.198 4a , c). Whittenberger& Nuttin g (1957)an dSmi t& Nortj e (1958)detecte d significantdecrease si ngros san dseru mviscositie swhe ntomat ojuic eo r dilutedpast ewa streate dwit ha commercia lpectinase .Thi sdecreas ewa s notaccomplishe db ya clea rchang ei nth emicroscopi cappearenc eo fcel l wallsalthoug hth ewall sappeare dt oincreas eslightl yi nbirefringenc e (Whittenberger& Nuttin g 1957).Th echaracte ro fcel lwall swa sfoun dt o varywit hthei rpecti ncontent .Wall spermeate dwit hpectin sar etacky , resilient,an dcapabl eo fbindin gappreciabl equantitie so fwater ,wherea s wallsdevoi do fpectin sar ebrittle ,friabl ean dles shydrophilic .Unde r identicalcondition sth epecti ncontainin gwall syielde dth ethicke rjuice s (whittenberger& Nuttin g 1957).whittenberge r& Nuttin g (1958)presente d resultswhic hdeemphasiz eth erol eo fsolubl epectin si ndeterminin ggros s viscosity.A sa nelectrolyt ei teve nha da negativ eeffec to nth egros s viscosityo ftomat ojuice .O nth eothe rhan dCoole r (1962)postulate dtha t juicesfro meffectivel ystea mblanche d tomatoesexiste da sa wea kge ldu e toth epresenc eo fa continou sstructur eo fhydrophylic ,macromolecula r polyelectrolytes (mostlypectins) .Cel lwall san dothe rsuspende dparticle s werethough tt othicke nth esyste man dt oincreas eth eshearin gforc e necessaryt oinduc ean dmaintai nit sflow .Steven s& Paulso n(1976 ) emphasizedth eimportanc eo fbot hinsolubl ean dsolubl epolygalacturonide s indeterminin ggros sviscosity .± 90 %o fth edifference si nth egros s viscosityo fjuice sprepare dfro mlo wan dhig hviscosit yvarietie scoul db e attributed tothes etw ofractions .I twa sfoun dtha tbot hfraction scoul d beequall yimportant .

Luhe tal .(1960 )studie dth erelationshi pbetwee ngros sviscosit yan d insolublepectin .Washe dho tbrea ktomat ojuic ewa sextracte dwit h0.5 % EDTAa tneutra lpH .Afte rremova lo fth eEDT Ath eseru mwa sadde dbac kt o theEDT Atreate djuic efraction .I twa sshow ntha tth etomat ojuic eo f unripetomatoe swhic hcontaine d largeamount so finsolubl epecti nha da highinitia lgros sviscosit ybu tlos t38 %o fthi sgros sviscosit yupo n extractionwit hEDTA .I nsoft ,rip etomatoe sa larg epar to fth einsolubl e pectinha sbee nconverte d tosolubl epectin .Th ejuice so fsoft ,rip e tomatoeswer elowe ri ngros sviscosity ,bu tretaine dmos to fthi sviscosit y 41

(89%)afte r extractionwit h EDTA. These results seem to indicate that the EDTA soluble fraction of thejuic e is an important factor determining gross viscosity. Foda &McCollu m (1970)repeate d these experiments and found a viscosity drop of 80%afte r resuspension of the EDTA treated cells in the original serum.Relativel y lowamount s ofuronide s and reducing sugarswer e detected in theEDT A extract; therefore the effect ofwate r extraction on grossviscosit y was studied. Readings for apparent viscosity (measured with theBrookfield )wer e 32 for the originaljuice ,4 3 forwashe d solids and 8 forwashe d solids in the original serum.Hig hmolecula rweigh t polymers with a neutral sugar/uronide ratio of approximately 11,closel y associated with the insoluble solids,wer e extracted during thewate rwashing s and were responsible for the grossviscosit y drop.Fod a &McCollu m also showed that degradation ofwashe d tomato cellswit h pectinase after degradation with cellulase (8M urea added to inhibit pectolytic enzyme activity) did not result in a further grossviscosit y loss. Itwa s concluded that the data did not indicate a significant contribution of theuronide s to the grossviscosit y of tomatojuice . Stephens et al. (1970), studying thepecti n content of tomatoes and juices of fourvarieties ,conclude d that the content of pectic substances alone did not determine thegros sviscosit y of canned juice.The y refer to thewor k ofLu h et al. (1954b )wh o concluded that not only the quantity but also thequalit y of thepecti c materialswa s important in forecasting the grossviscosit y of tomatoproducts .

Contribution of cellulose to gross viscosity

There areno tman y reports on the contribution of cellulose to the grossviscosit y of tomatoproducts .Whlttenberge r and Nutting (1957) observed that theadditio n of a technical cellulase preparation (contaminated with pectinase) toho tbrea k juice resulted in the digestion ofpecti n (serumviscosit y drop)an d thebul k of the cellulose as indicated bymicroscopi c examination. The grossviscosit y of thejuic e decreased dramatically, to a level lower than after the addition ofpectinas e alone. Their results indicate the importance of cellulosic structures in maintaining grossviscosity . Boiling cellwall s successively in 2% sulphuric acid and 2% sodiumhydroxid e and then extracting themwit h 42 ethylalcoholan dether ,remove dal lnon-cellulosi csubstance sfro mth ecel l walls.Th etreate dcel lwall swer ebrough tbac kt oth eorigina lvolum ewit h distilledwater .Th egros sviscosit yo fjuic edecrease dwit hth e treatments,indicatin gtha tnon-cellulosi ccomponent so fth ewall smad ea n importantcontributio nt ogros sviscosity .However ,th epur e "cellulose-walls"yielde d suspensionso fappreciabl egros sviscosit y (Whittenberger& Nuttin g 1958). I ncontras tt oth eopinio no fWhittenberge r andNutting ,Twig g (1959)reporte dtha tth estructura lsolids ,whic hh e considered tob eidentica lt oth ecellulosi ccel lwalls ,exerte dles s influenceo nth egros sviscosit yo fketchu ptha ntota lpectin .Fod a& McCollum (1970)o nth eothe rhan dattribut ea nimportan trol et ocellulos e indeterminin ggros sviscosity .A technica lcellulas epreparatio ni nwhic h pectolyticenzym eactivitie swer einactivate db ya treatmen twit h8 M urea , greatlydecrease dth eviscosit yo fwashe d solids.Experiment swit ha cellulaseextracte dfro mtomatoe sals odecrease d thegros sviscosit yfo r tomatojuice .Thei rdat adi dno tindicat ea nimportan trol eo fpecti ni n determininggros sviscosity . Luhe tal . (1960)foun dtha tth eamoun to fcellulos edi dno tchang e substantiallydurin gripenin go rwithi nseason .Therefor ethe yconclude d thata relationshi pbetwee ngros sviscosit yan dcellulos edi dno thold . Luhe tal .(1981 ,1982 ,198 4 a), studyingcarbohydrat epolymer si n processingtomatoes ,showe dtha ttomat opast esample shig hi ntota lpecti n contentan dwit ha hig haverag emolecula rweigh to fth epecti nwer ehig hi n grossviscosities .Thei rresult sals oprov etha tth esam epast esample s werehig hi naci ddetergen tfibr e(ADF )an dcellulos econtent .Roughly ,th e ADFwa scompose do f80 %cellulose ,an d20 %ligni nan dcutin .Lu he tal . (1984b,c )repor ttha ttota lpectin ,ADF ,cellulos ean dWIS/T Srati oar e indicativeo fth egros sviscosit yvalu eo fcanne dtomat ojuic ean dpaste .

Contribution of protein to gross viscosity

Inspit eo fth efac ttha tprotein sar epresen t inlarg eamount si nth e AISo ftomatoe san dtomat oproduct s (Table8) , therol eo fprotei ni n determiningviscosit y isno tver ywel ldocumented . A3

Table8 :% protei ni nAI So ftomatoe san dtomat oproducts . reference %protein *

McColloche tal .195 0 20-50 Williams& Bevenu e195 4 17 Brown& Stei n197 7 23-29 Davies& Hobso n198 1 ±30 Takada& Nelso n198 3 20-23

* totalN expresse da sprotei n

McColloche tal .(1950 )foun dlosse si nAI Sa sa resul to fprocessin gwhic h couldno tb eexplaine db ylosse si npecti nalone .Base do nth eamount s presentthe ysuggeste dtha tprotei nmigh tpla ya rol ei ndeterminin ggros s viscosityo ftomat opaste .O nth eothe rhan dFod aan dMcCollu m (1970) showedtha tth edegradatio no fprotei ni ntomat ojuic ewit hpronase ,whic h containedn opectinas eo rcellulas eactivity ,cause donl ya relativel y smalllos si ngros sviscosity .Recentl yTakad a& Nelso n (1983)reporte d thatth eviscosit yo fa tomat oproduc tmode lsyste mconsistin go fcitru s pectinan dbovin eseru malbumin ,change da sa resul to fa p Hchange .The y attributedthi schang et oth ereversibl eelectrostati c complexformatio no f pectinan dprotei nbu tno tt owea khydroge nbond sbetwee nthem .A simila r patterno frespons eo fgros sviscosit y top Hchang ewa sfoun di ntomat o puree (10°Bx)an di ntomat ojuic e (5.5°Bx)prepare db ydilutio nfro mth e puree.Th emaximu mgros sviscosit ywa sobtaine da tp H4.4 ,an di twa sfoun d thatth epH-effec twa sreversible .Tomat opure edilute dfro mpast ewit ha highersolid sconten t (20°Bx)di dno tsho wa chang ei ngros sviscosit ywit h achang ei npH .I twa sconclude dtha tsever eo rprolonge dheatin gma y denatureth eprotei nan dstabiliz eit scomple xwit hpectin ,resultin gi na n irreversiblecomple xformation .Takad a& Nelso nspeculate d thatth epectin s associatedwit hth ecellulos efibril si nth ecel lwall smak etigh t complexeso rfascicle swit hproteins ,resultin gi na stron ginteractio no r networkaroun dth ecel lsurface .I nthi swa yprotein sma ycontribut et oth e grossviscosit yo ftomat oproducts . 44

Contribution of electrolytes to gross viscosity

Thegros sviscosit y oftomat ojuic ei sgreatl yelevate db ya complet e replacemento fth eseru mwit hdistille dwate r (Whittenberger& Nuttin g 1958,Fod a& McCollu m (1970).Th eadditio no fnon-electrolyte ssuc ha s sucrose,glyceri nan dethylalcoho lt osuc ha washe djuic eha dn oeffec to n grossviscosity .However ,additio no fsmal lamount so felectrolyte s includingsolubl epectin ,citri cacid ,sodiu mchlorid ean dcalciu mchlorid e resultedi ndecrease dgros sviscosity . (Whittenberger& Nuttin g1958) . Accordingt oWhittenberge ran dNutting ,th esurface so fth ecel lwall sbea r electriccharge s (e.g.insolubl epectin )whic hhel pmaintai nth ewall si n suspension.I nth eabsenc eo fsolubl eelectrolytes ,th echarge sexhibi t theirmaximu meffect .Th ewall sswell ,bin dwate ran dgiv eris et o increasedgros sviscosity ;i nfact ,sometime sth ejuic ethicken st oa sem i gel. Inth epresenc eo felectrolytes ,however ,th echarge so nth ewall sar e neutralized,th ewall sshrink ,an da dro pi ngros sviscosit yensues .Th e quantityo felectrolyte soccurrin gnaturall y infres htomatoe si s sufficientt okee pgros sviscosit ya ta lo wlevel .(Whittenberge r& Nuttin g 1958).Fod a& McCollu m (1970)conclude dtha tth edrasti creductio nfoun di n grossviscosit yafte radditio no fEDT A (0.5%,neutra lpH )t owashe dtomat o juice,an dth esubsequen tincreas ewhe ni twa sremoved ,indicate da nioni c role,an dno ta depectinizatio na ssuggeste db yLu he tal . (1960).Th elo w contento fpecti ni nth eEDT Aextract soffere dadditiona lsuppor tfo rthi s view.

pH and gross viscosity

Twigg (1959)reporte da relationshi pbetwee nth ep Han dth egros s viscosityo fketchu pa smeasure dwit ha sensor ypanel .H epostulate dtha t theincreas ei nketchu pgros sviscosit ywit ha decreas ei np Hresulte dfro m increasesi nth eviscosit yo fth elo weste rpecti csubstance sdissolve di n theketchup .Dat apresente d inhi sstud yindicat etha tapproximatel yhal f ofth ewate rsolubl epecti nwa ssufficientl y lowi nmethoxy lconten tt o formcalciu mpectates . 45

Callose and gross viscosity

Callose,a namorphou spolyme ro fglucos eresidue si nP-1, 3linkages , iswidesprea dthroughou tth eplan tkingdo man di sforme di ncertai ntype s ofcell si nrespons et oinjury ,ultrasoun dan dhig htemperature .Callos ei s insolublei nwater ,ethanol ,ho tacid ,o rho talkali .I thas ,however ,a strongaffinit yfo rwate r (Dekazos1972) . Dekazos (1972)suggeste dtha tperhap scallos ei sth eplan tpolysaccharid e thati sth emajo rcontributo rt oth egros sviscosit yo ftomat oproduct san d totextur ei ntomat ofruit .Callos ewa sdetecte d inheate dunbruise d tomatoes (50°C),artificiall ybruise d tomatoes (35-45°C),mechanicall y harvested tomatoesan di ntomat ojuic efro mcommerciall yharveste d tomatoes.Callos econtainin gtomatoe swer enoticeabl y firmert oth etouch . Thepropertie so fcallos ewer ecompare dt oth epropertie so fth ecellulosi c cellwall sdescribe db yWhittenberge ran dNuttin g (1958).

2.3.5. Contribution of processing factors to gross viscosity

Effect of preheating on gross viscosity

Inth epas t4 0year sman yauthor shav ereporte do nth eeffect so f preheatingtemperatur eo nth egros sviscosit yo fjuice ,past ean dsauce s (Kertesz& Locont i1944 ,Robinso ne tal .1945 ,McColloc he tal .1950 ,Lu h etal .1954 ,Han de tal .1955 ,Whittenberge r& Nuttin g 1957,Twig g1959 , Cooler1962 ,Lu h& Daou d 1971,Garce s& Lu h1972 ,Sherka t& Lu h1976 ,1977 , Bel-Haj1981 ,Lu he tal .1981 ,1982 ,198 4a,b,c ,X ue tal .1986) .Thi s preheatingtemperatur ei sdefine da sth etemperatur et owhic htomatoe sar e heateddurin gth ecrushin gprocess .Kertes z& Locont i(1944 )wer eamon gth e firstt ostud yth ecause sfo rth eincreas ei njuic egros sviscosit y asa resulto fhighe rprehea ttemperatures .The ystate dtha tpreheatin gtomatoe s to82° Cwa snecessar y inorde rt oinactivat eP E (theydi dno tdetec tP G activityi ntomatoes )an dt oobtai na hig hseru mviscosity ,an da sa resul t oftha ta juic eo fpleasin ggros sviscosity .Lik eKertes z& Loconti ,man y otherauthor sconclud etha tth ebeneficia leffec to fpreheatin gi sdu et o pectolyticenzym einactivatio nan dtherefor et obette rpecti nretention . 46

However,th eliteratur eshow ssom edisagreemen tabou tth edegre eo f theprehea ttemperatur enecessar y foreffectiv einactivatio no fpectolyti c enzymes.McColloc he tal . (1950)showe dtha ti ntomat opast eproductio n 82°Cmigh tno tb eenoug ht oinactivat eenzyme san dthe yrecommen da prehea t temperatureo f85°C .Lu h& Daou d (1971)an dGarce s& Lu h (1972)reporte d thatP Ewa sinactivate dafte r1 5sec .82° Can dP Gafte r1 5sec .104°C . Bel-Haj (1981)indicate d that1 5sec .93° Cwoul db eenoug hfo rP Ean dP G inactivation.Lu he tal .(1981 ,1982 ,198 4a )state dtha ta ho tbrea k treatmenta t88° Cfo r10 0sec .appeare dt ob eenoug hfo rP Ginactivation . Informationo nth eeffect so fspecifi cprehea ttemperature so ngros s viscosity isabundan tbu tno tver yconsistent .Robinso ne tal .(1945 ) detected increasedgros san dseru mviscositie swhe nth epreheatin g temperaturewa selevate dfro m69° Ct o77° Can dthe nt o93°C .I nanothe r serieso fexperiment sn odifferenc ei nviscosit y (seruman dgross )wa sfoun d betweenjuice swithou ta prehea ttreatmen t (temp.± 27°C )an dthos epreheate d to51°C .Twig g (1959)wh ostudie dth eeffect so fpreheatin ga t66 ,79 ,9 3an d 100°Co nketchu pviscosity ,reporte d thata nincreas ei nprehea ttemperatur e above93 CCdi dno tincreas egros sviscosity .I ngeneral ,highe rtemperature s resultedi nles sseru mseparation ,highe rseru mviscosity ,mor etota lan d watersolubl epectin ,an dmor eAI San dstructura lsolids .However ,th e intrinsicviscosit yo fwate rsolubl ean doxalat esolubl epecti ndecrease d stronglydurin gpreheatin ga t100° Ci ncompariso nt o93°C .Coole r(1962 ) reportedtha tth ehighes tjuic eviscositie swer eobtaine dafte rstea m blanchingwhol etomatoe s (insit uinactivatio no fenzymes )an dheatin gth e coarsechoppe dtomatoe st o90 CCo r120°C .Choppin ga troo mtemperatur e( ± 30°C)yielde djuice swit hmuc hlowe rviscosities ,independen to fth e finishingtemperature .Juice swit hhighe rviscositie sha dhighe rAIS , serum-AISan dparticulate-AI Scontent .Lu h& Daou d (1971)an dGarce s& Lu h (1972)studie dth eeffect so fprehea t temperaturean dholdin gtim eo n enzymeinactivatio nan dgros sviscosity .Seru mviscosit y increasedwhe n preheattemperature sincrease dfro m71° Ct o115°C .Onl ysligh tgros s viscosity increaseswer edetecte da sresul to fa nincreas ei nprehea t temperaturesfro m60° Ct o82°C .However ,whe nth etemperatur ewa sincrease d from93° Ct o115°C ,gros sviscosity ,AI San dtota lpecti nincrease d substantially.Sherka t& Lu h (1976)reporte d thatjuice spreheate da t100° C or104° Cwer esignificantl yhighe ri ngros sviscosit ytha nthos epreheate d 47 at64° Co r77 CC.Bel-Ha j (1981)detecte dn odifference sbetwee npreheatin g at66° Co r79°C ;however ,a prehea ttemperatur eo f93° Cresulte di n significantlyhighe rgros sviscosity .Seru mviscosit y increasedwit hal l thebrea ktemperatures .Lu he tal . (1981,1982 ,198 4a,b,c )studie dth e effectso fprehea ttemperatur eo nth echaracteristic so fjuice sfro m severalvarieties .Increase si nprehea ttemperatur e (66,77 ,88 ,9 6an d 107°C)alway sresulte di nhighe rgros san dseru mviscosities .No tonl ydi d totalpecti nincreas ea sa resul to fhighe rtemperatures ,bu ts odi daci d detergentfibre ,cellulos ean daci ddetergen tlignin .Cellulos econten t (wetbasis )double dwhil etota lpecti nconten ttreble da sa resul to fth e increasei ntemperatur efro m66°C-107°C .(Lu he tal .198 4b) . Allthes e resultswer eexplaine d interm so fquicke renzym einactivatio n(als o cellulases:Lu h198 4b )an dbette rretentio no fcomponent sa sa resul to f thehighe rprehea ttemperature .However ,usin gthi stheor y iti sdifficul t toexplai nth ebeneficia leffec to fprehea ttemperature shighe rtha nthos e neededfo ra nefficien tenzym einactivation . Xue tal .(1986 )studie d therheologica lpropertie san dmicrostructur e ofcanne djuice san dconcentrate smad efro mfou rtomat ocultivar sa tthre e preheatingtemperature s (85°C,96° Can d 107°C).Th ehighe rgros sviscosit y ofjuic ean dpaste sa sresul to fa highe rpreheatin gtemperatur ewa sno t onlyattribute d toa highe rdegre eo finactivatio no fP Gan dP Ebu tals ot o ahighl ydisrupte dcel lstructur ei nth esamples . Hande tal .(1955 )observe dthat ,i ngeneral ,a nelevatio no fth e preheattemperatur eincrease dth egros sviscosit yo ftomat ojuice ;however , theeffect swer edependen to nth efinishin gprocedure .Th eeffec to f preheatingtemperatur eo nseru mviscosit ywa smuc hmor eclea rcut .Seru m viscositywa sminima la ta prehea ttemperatur eo f60-66°C .A thighe r temperaturesth epecti cenzyme swer einactivated ,a tlowe rtemperature sth e enzymeswer einhibited .I nbot hcase sthi sresulte di nhighe rseru m viscosity.Han de tal .(1955 )conclude d thatth eeffec to fpreheatin go n grossviscosit ywa sdu et oa softenin go fth ecrushe dtomat otissue sprio r tofinishin gi nadditio nt oth epreservatio no fth epecti ni nth eserum . Whittenberger& Nuttin g (1957)presente dresult swhic hsuppor tth etheor y that,beside senzym einactivation ,softenin go fth etissu ei simportan ti n determiningth eeffec to fpreheating .Unheate dgree ntomatoe sgav ea juic e lowi ngros sviscosit yan dcel lwal lcontent .Whol egree ntomatoe s 48

preheated to 88°Cyielde d a juice high in grossviscosity ,cel lwal l and pectin content.Whe n ripe fruitswer e used,difference s in preheat treatment caused smaller differences injuic e structure (cellswer e easily separable)an d pectin content;however ,difference s ingros sviscosit y were significant.Twig g (1959)conclude d that preheat temperatures above 93°C appear toallo w for the extraction of more structural solids,probabl y due toa softening effect on the tomato tissue,bu tma y also cause a decrease inaverag emolecula rweigh t of the pectic constituents and in thiswa y negatively affect grossviscosity . Luh &Daou d (1971)an d Luh et al. (1984 c) suggest that the increase in serumviscosit y athighe r preheat temperatures may be explained by the solubilization of insoluble pectin to the soluble form. Inadditio n to its effects onviscosity ,preheatin g may also affect the retention of ascorbic acid (Luh &Daou d 1971,Sherka t &Lu h 1976,1977 , Bel-Haj 1981), colour (Cooler 1962,Sherka t & Luh 1976,1977 ,Lu h & Daoud 1971)an d flavour (Sherkat &Lu h 1976, 1977).

Effect of holding before finishing on gross viscosity

In industrial practice there isofte n some delay between the crushing orbreakin g of the tomatoes and heating the crushed tomatoes to the desired preheat temperature.A s early as 1944,Kertes z & Loconti stated that a juice of thehighes t possibleviscosit y cannot be obtained when the inactivation of the enzymes isperforme d after breaking the tomatoes.The y showed that theholdin g of finished juice at room temperature foronl y 30 sec. caused a drop inbot h gross and serumviscosity . Even in samples heated immediately aftermakin g the juice (within 2-5 sec.) therewa s some viscosity losswhe n compared to theviscosit y ofjuic emad e from canned tomatoes.McColloc h et al. (1952)therefor e concluded that the action of enzymes is so rapid that a 100%retentio n ofpecti c substances isno t obtained evenunde r thebes t practical industrial conditions of rapidly preheating freshly crushed tomatoes.Result spresente d by several other authors seem to confirm this statement.Lu h et al. (1954)prepare d juices according to an activated coldbrea k process.Tomatoe swer e dipped in boilingwate r for 2min . and finished through a 0.8 mm screen.Th e temperature of thejuic ewa s approximately 45°C.Afte r 2hour s the enzymes 49 were heat-inactivated at 100°C.Thes e juiceswer emuc h lower in gross viscosity, totalpecti n and AISwhe n compared tonorma lho tbrea kjuices . Cooler (1962)reporte d that crushing tomatoes at room temperature (± 30°C) followed by holding at this temperature and finishing at 90°C yielded juicesmuc h lower in grossviscosit y thanjuice s obtained after crushing, quick heating to 90°C and finishing at 90°C.Mier s et al. (1967) studied thepossibilit y of inactivating pectolytic enzymes by theadditio n of acid. They found that a delay of 5 seconds in the addition of acid tomacerate d tomatoes already resulted in a sharp grossviscosit y drop. In contrast to these authors,Han d etal . (1955) showed that holding chopped tomatoes for severalhour s at temperatures of 21°C and 54°C before heating caused no significant reduction in either serum or gross viscosity. However,holdin g finished,unheate d pulp for onehou r at 38°C resulted in grossviscosit y and serumviscosit y losses.The y conclude that instead of using the terms "hot and coldbreak " it issufficien t to statewhethe r the chopped tomatoeswer e preheated prior to finishing and to report the preheating temperature.Accordin g toHan d et al.,th e term "hot-break" has caused a great deal of confusion because of the implication that the necessary heating must be accomplished during or immediately after chopping inorde r topreserv e the pectin in thejuice . Moreover,Smi t &Nortj e (1958) found that a delay of three hours at 21-27°C betweenbreakin g (screen 6.4 mm)an d preheating (82°C)resulte d in higher grossviscosit y at the same soluble solids contentwhe n compared to the grossviscosit y of sampleswhic hwer e immediately preheated after breaking,eve n though pectin,methoxy l content and comparative serum viscositieswer e lower.A s theWI S contentwa sver y much the same they suggested that the insoluble particleswer e structurally different. Crandall &Nelso n (1975) studied the effects ofholdin g on theviscosit y of juice and puree prepared from two tomatovarieties .Crushe d tomatoes (screen 9.5 mm)wer e held at 21°C for 1hou rbefor e heating to 93°C. For onevariety ,holdin g had no effect on the serum and grossviscosit y of the juice,bu t the grossviscosit y of the puree increased; the other variety showed lower serumviscosit y of juice and puree afterholding ,bu t the grossviscosit y of juice and puree increased. Leoni et al. (1979,1980 , 1981)presente d resultswhic h indicate that holding cold break crushed 50 tomatoesa troo mtemperatur efo ru pt o3 hour sbefor efinishin ga t60-62° C mayincreas eth egros sviscosit yo ftomat oconcentrate .I twa sfoun dtha t asa resul to fthi sholding ,th ehexametaphosphat esolubl epecti nincrease d whileth esodiu mhydroxid esolubl epecti nremaine d fairlyunchange dan dth e freegalacturoni caci ddecrease dwhe ncompare dt oa col dbrea kwithou t holding.Th ephenomeno nwa sexplaine db yassumin gtha twithou tholding ,P E andP Gexhibite doptima lactivit ydurin gbrea k (60-62°C)an devaporatio n (48-50°C).Wit hholding ,th ebreakdow no fpecti ni sslowe ran dobviousl y thesaponifie dpecti nprecipitate sa scalcium-pektinate srathe rtha nbein g brokendow nfurthe rb yPG .Holdin gwa sfoun dt ohav ea favourabl eeffec to n thegros sviscosit yo fconcentrate sobtaine d fromvarietie swit ha hig h initialconten to fpecti csubstance s (varietiesfo rmechanica lharvesting ) whereasi tscarcel yaffecte dgros sviscosit y inth ecas eo fconventiona l varieties.I tha st ob ementioned ,however ,tha tho tbrea kconcentrate s werealway shighe ri ngros sviscosit y thaneithe rth ecol dbreak ,o rth e coldbrea kwit hholding ,concentrate .Th eho tbrea kconcentrat econtaine d muchmor ewate ran dsodiu mhydroxid esolubl epectin ,bu tles sfre e galacturonicacid . Allth eliteratur ecite dabov edeal swit hth eeffect so fholdin g crushedtomatoe sa trelativel ylo wtemperatures .I tca nb econclude dfro m thisliteratur etha tth eeffect so fholdin gar einfluence db ycoarsenes so f theparticle si nth ecrushe dtomatoe s (smallparticle sfavou renzym e action),temperatur eo fholding ,duratio no fholdin gan dprobabl yb y variety (amounto fpectolyti cenzymes) .When ,fo rexample ,th ecrushin go f tomatoesresult si ncoars epiece s(large rtha n1 cm )an dth etemperatur e duringcrushin gi skep ta ta lo wleve l(20°C ), th emomen to fheatin gi s probablyles simportant .However ,th eheatin grat eshoul db ea shig ha s possiblebecause ,a sKertes z& Locont istate di n1944 ,i nth ehea t inactivationo fan yactiv esyste mo fenzym ean dsubstrat eth etemperatur e passesthroug ha rang ewher eth eenzym eactivit yi sth ever yhighest . Theliteratur eals ooffer ssom einformatio no nth eeffect so fholdin g crushedtomatoe sa thighe rtemperatures ,o njuic egros sviscosity .Lu h& Daoud (1971)an dGarce s& Lu h (1972)indicate d thatholdin gcrushe d tomatoes (3m mscreen )a ttemperature sinsufficien tfo rcomplet eenzym e inactivationma yresul ti ndecrease dviscosities ;moreover ,the yshowe d thatholdin gcrushe dtomatoe sa ttemperature so f93°C ,104° Can d115° C 51 beforefinishin g increasedgros sviscosity .However ,th eholdin geffec twa s smallertha nth eeffec to fa nincrease dprehea ttemperature .Bette rpecti n retentiona sa resul to fquicke renzym einactivatio nserve da sa n explanationfo rth eincrease dgros sviscosity .Mier se tal . (1970)reporte d thatcrushe dtomatoe syieldin gjuice swit hhig hgros sviscosit yshowe d grossviscosit ygain so fu pt o20 %afte r1 0min .a t100° Cbu tcoul dsho w losseso fu pt o25 %afte r3 0min .a t 100°C.Whe ninitia lgros sviscosit y wasa ta ver ylo wlevel ,th ejuice sfro mth eheate dcrushe dtomatoe sshowe d increasedgros sviscosit yeve nafte r3 0min .a t100°C .Increase san d decreasesi ngros sviscosit ywer esmalle ra t71° Cholding .Th eviscosit yo f serumsusuall ydecline dprogressivel ythroug hth e3 0minute sheatin g period.Th egros sviscosit y increasewa sexplaine db yth esoftenin go f tomatotissu ei nth ecrushe dtomatoe sa sa resul to fcooking ;mor esoli d tissuewen tthroug hth epulpe rscree nint oth ejuic eupo npulping .Thi s increasei ntota lsolid swa saccompanie db ya decreas ei nth e% o f insolublewast efro mth epulper .

Effect of the finisher operation on gross viscosity

Mosto fth einformatio no nth eeffect so fth efinishe roperatio no n thegros sviscosit yo ftomat ojuic eha sbee nobtaine dfro mstudie so fHan d etal .(1955 )an dMoye re tal .(1959) . Moyere tal .(1959) ,studyin gth efactor saffectin gth eyiel do f tomatojuice ,showe dthat ,excludin gth elosse ssuffere ddurin gwashin gan d trimming,bein ga reflectio no fra wproduc tquality ,th egreates tvariatio n inyiel doccurre d inth efinishe roperatio nwher eth eskin san dseed sar e separatedfro mth ejuice .O fthre emethod so ffinishin g (revolvingpaddle , tapered screwan deccentri cdrum) ,th epaddl efinishe rprovide d thewides t rangei nyields ,bu talso ,unde roptima lcondition sfo rhig hgros s viscosity,th ehighes tyields .Han de tal .(1955 )foun dtha tth eus eo fa paddletyp efinishe rru na tintermediat eo rhig hspeed ,resulte di nhighe r juicegros sviscositie stha nth eus eo fa scre wextractor .Thi sresul twa s confirmedb yBel-Ha j (1981).H eals oshowe dtha tdoubl eextractio nwit ha screwtyp eextracto ralway syielde dhighe rgros sviscositie stha nsingl e extraction,althoug hth edifference swer eno tsignificant . 52

Witha paddl etyp efinishe rth eeffect so ffinishin gwer eclosel y relatedt opaddl espee dan dpreheatin gtemperatur e (Hande tal . 1955,Moye r et.a l 1959).A nincreas ei npaddl espee dalway sresulte di nhighe rgros s viscosityregardles so fth epreheatin gtemperature ;th eeffect so n serumviscosit ywer esmaller .No tonl yth egros sviscosit ybu tals oth e yieldwer eshow nt ob eproportiona lt opaddl espeed .A ta lo wpaddl espee d fewsuspende d solids,mos to fwhic hwer espherical ,wer eincorporate di n thejuic ean dth ejuic ewa so flo wgros sviscosity .A ta hig hpaddl espee d moresuspende d solidswer eincorporate di nth ejuice .Thes esolid swer e moreelongate dan dth ejuic eha da hig hgros sviscosity .I ncontrast , Bel-Haj (1981)state dtha tpaddl espee dha dn osignifican teffec to ngros s andseru mviscosit yan do npecti ncontent .I nadditio nt oit seffec to n serumpectin ,preheatin gexert sa softenin geffec to nth etomat otissue , whichi ntur nha sa nimportan tpractica linfluenc eo nth eactio no fth e finisher.Therefore ,lowes tyield san dlo wgros sviscositie swer eobtaine d whensample swer eprepare dwithou tprehea ttreatment ,especiall ywhe nthe y werefinishe da tlo wpaddl espeeds .A searl ya s1944 ,Kertes z& Locont i notedtha tmanufacturer sfinis hcol dbrea kjuice sa ttemperature so f6 0- 70°Ci norde rt oincreas eth eyield .Temperature si nth erang eo f60-72° C gaveth ehighes tyield sbu tlo wviscositie swhe nextractio nwa sa tlo w paddlespeeds .Seru mviscositie sa tthes eprehea ttemperature sappeare dt o bea ta minimu man dtherefor eth eseru mwa seasil yseparate d fromth e insolublematerial ,resultin gi nhig hyields ;however ,gros sviscosit y remainedlo wa sa resul to fth elac ko fsuspende dsolid si nth ejuice .I n fact,th egros sviscosit y ofthes esample swa slowe rtha ntha to fsample s whichwer eno tpreheated .Whe ntomatoe swer epreheate dt o93° Co rhigher , serumviscosit y increasedwhic hmad efinishin ga tlo wpaddl espeed smor e difficult.Th eyiel do ftomat ojuic edroppe dan dgros sviscosit ywa sonl y slightlyhighe rtha ngros sviscosit y obtaineda tintermediat epreheatin g temperatures.A tth ehig hpreheatin gtemperatures ,paddl espee dha dt ob e increasedt omaintai nsatisfactor yyields ;a sth eyiel dincreased ,s odi d thegros sviscosit y (Hande tal .1955 ,Moye re tal .1959) .Thes eresult s arei nagreemen twit hthos eo fCoole r (1962)wh ofoun dtha tstea mblanchin g tomatoesreduce dth eyield .H esuggeste dtha tseru mviscosit y isincrease d byth e"i nsitu "enzym einactivatio nan dtha tthi shig hviscosit yseru m 53 probablystick st oth eparticles .Whittenberge r& Nuttin g (1957)als o showedsom eresult swhic hstres sth eimportanc eo fpreheatin gi n determiningth efinishin geffect .Unheate dgree ntomatoe syielde da juic e lowi ngros sviscosit yan dcel lwal lconten twhil eunheate drip etomatoe s (cellsmor eeasil yseparable )yielde dhighe rgros sviscosity .Preheatin ga t 88°Cchange dth epictur ecompletely ;juice so fgree ntomatoe swer emuc h thickertha nthos eo fre drip etomatoes . Ingeneral ,preheatin gan dfinishin gtemperature sar eapproximatel y thesame ;however ,i twa sfoun dtha tth egros sviscosit yo fjuic edecrease d asa resul to fa decreas ei nfinishin gtemperatur eafte ra prehea t treatmenta t93°C .Gros sviscosit y increasedwhe njuic ewa sfinishe da t 93°Cafte rpreheatin ga t77°C .However ,th eeffec to fa nincreas ei n preheattemperatur ewa smuc hlarge rtha nth eeffec to fa finishe r temperatureincreas e(Han de tal . 1955).Coole r (1962)showe dtha tfinishe r temperature (30°Co r90°C )ha dn oinfluenc eo nth egros sviscosit yo f juiceswhe ntomatoe swer ecrushe da troo mtemperatur e (30°C). BothHan de tal .(1955 )an dMoye re tal .(1959 )observe dincrease d grossviscosit yan dyiel da sa resul to fenlargemen to fth efinishe rscree n openings (0.6,0.8 ,1.1 ,1. 5mm) .However ,the ystate d thatthes eincrease s wereo fn opractica lsignificanc ebecaus etomat ojuic eprepare dusin g screenopening slarge rtha n0. 6m mshowe ddefect sdu et oth eappearenc eo f specksconsistin gchiefl yo fsee dfragments .Smi t& Nortj e (1958)foun d thatth eus eo flarge rscree nopening si na scre wextracto r (0.5-1.0mm ) increasedpast egros sviscosity ,reduce dseru mseparatio nbu tha dlittl e effecto nseru mviscosity .Twig g (1959)observe dtha tth eus eo fa 0. 6m m finisherscree nresulte di nketchup slowe ri ngross ,seru mviscosit yan d solidsconten tbu thighe ri nseru mseparation ,tha nketchup sprepare dwit h a0. 7 or0. 8m mfinishe rscree nsize .Bel-Ha j(1981 )foun dsignifican t differencesi njuic egros sviscosit ybetwee nth eus eo fa 0. 7m man d0. 8m m screenopening .However ,thi sdifferenc ewa sno taccompanie db ysignifican t differencesi nsolubl esolids ,tota lsolids ,tota lpecti no rseru m viscosity. Moyere tal .(1959 )showe dtha ta ninvers erelationshi pexist sbetwee n the% o ffinishe rwast ean dth egros sviscosit yo ftomat ojuic ea ta give n preheatingtemperature .Hig hyield sar eobtaine dwhe nth efinishe ri s adjusted togiv ea relativel ydr ywast ematerial .Lo wyield sresul tfro m 54 losso fth eaqueou sphase ,o rserum ,whic hi sobviou swhe non econsider s thefac ttha t99 %o fth ejuic eb yweigh ti sserum .Th econdition so f finishingwhic hconserv eth egreates tamoun to fseru mals oresul ti nth e highest% o fsuspende dsolids . Fromth eliteratur ei tseem stha tprehea ttreatment ,typ eo ffinishe r aswel la sfinishin gadjustment shav ea ninfluenc eo nth efinishin geffect . Finisheradjustment sstudie di nliteratur ewer epaddl espeed ,scree nsiz e andtemperature .I tseem sreasonable ,however ,t oassum etha tothe r factors,suc ha stwis to rpitc ho fth epaddles ,clearenc ebetwee npaddle s andscree n (Moyere tal . 1951)an djuic einpu twil lhav ethei rinfluenc eo n thefinishin geffect .

Effect of concentration on gross viscosity

Industrially,tomat ojuic ei softe nconcentrate deithe rfo rdirec tus e inproduct sa sketchu pan dothe rsauce s (lowsolid sconcentrates )o rfo r storagepurpose s (highsolid sconcentrates) .Theoreticall yman ytechnique s areavailabl efo rth econcentratio no ftomat ojuices ,fo rexample : evaporationeithe ra tatmospheri cpressur eo runde rvacuu m (Feinberg 1967),reverse dosmosi s (Mersone tal .1980 ,Ishi ie tal .1981) ,th e centrifugeseparation ,seru mevaporatio nmetho da sdescribe db ySul c& Ciric (1968)o reve nfreez econcentratio n (Deshpandee tal .1984) .T o date,however ,commercia ltomat oconcentrate sar emainl yprepare db y evaporation;man yevaporator san dpast eproducin gprocesse shav ebee n describedi nth eliteratur e (Randalle tal .1966 ,Carlso ne tal .1967 ,Anon . 1975,Che ne tal .1979 ,Anon .1981 ,Anon .198 1b ,L eMair e1981 ,Ladwi g 1982,Skierkowsk i1982 ,Dal ee tal .1982 ,Mores i& Liverott i1982 ,Rumse y etal .1984) . Themetho dan ddegre eo fevaporation ,a swel la sstorag econditions , areimportan tfactor sinfluencin gth equalit yo ftomat opast ean d concentrates.Qualit yca nb eevaluate do nth ebasi so fvitami nC content , degreeo fbrowning ,colour ,flavour ,tast ean dgros sviscosity .Lowe r evaporationtemperatur ean dshorte rprocessin gtim ear ethough tt ominimiz e thehea tdamag et otomat oconcentrate san dt oimprov equalit y (Mannheim& Kopelmann 1964).Withi nth escop eo fthi sthesi sonl yth eeffect so f 55

evaporationan dsom eevaporatio ntechnique so nth egros sviscosit yo fth e concentrateswil lb ediscussed . Kopelmann& Mannhei m (1964)an dMannhei m& Kopelman n(1964 )studie d thehea ttransfe rcoefficient so ftw oevaporator s (Luwaswept-surfac ean da atmosphericopen-pa nevaporator )use di ntomat ojuic econcentrat e manufacturefo rtw omethod so fjuic eevaporatio n (wholejuic ean dseru m evaporation);the yals ostudie dth eeffect so fevaporatio no njuic e quality.Juic ewa sprepare db ybreakin ga t60°C ,holdin gfo r5 min .an d finishingthroug ha 0. 8m mscreen .I twa sreporte dtha tth eseru m evaporationmetho dgav econsiderabl esaving si nevaporatio ntime .Th e juicewa sseparate d intofibre san dseru man dth eseru mwa sconcentrate d beforeresuspendin gwit hth efibres .Thi sfindin gwa sexpecte d tofavou r improvedproduc tquality .However ,i twa sfoun dtha tBostwic kvalue so f 22°Bxpast esample swer e5- 6 cmfo rth ejuic eevaporatio nmetho dan d 12-13c mfo rth eseru mevaporatio nmethod ,thi sirrespectiv eo fth e evaporatortype .Th eBlotte rvalue s (fordescriptio no fthi smetho dse e section2.4. )wer eals omuc hlowe rfo rth ejuic eevaporatio nmethod . Kopelmann& Mannhei m (1964)an dMannhei m& Kopelman n (1964)showe dthat , afterdilutio no fth epaste st o5°Bx ,gros sviscositie s (Brookfield)fo r thejuic eevaporatio nmetho devaporate dwit hth eLuw awer etwic ea shig ha s thosefo rth eseru mevaporatio nmetho do rth eatmospheri cevaporator .I n general,i twa sfoun dtha tth edilute dpast esample sha dlo wgros s viscositieswhe ncompare dt ocommercia ljuic esamples .Th elo wgros s viscosity forth eseru mevaporatio nconcentrate swa sexplaine db yreferrin g toth eresult so fEphrai me tal . (1962).The yreporte dtha tcentrifugin ga coldbrea kjuic eresulte d inth ecrushin go fcell san di ndisruptio no fth e solidsuspensio nstructur eo fth ejuice ,causin ga gros sviscosit y losso f 50%.Hea tdamag ea sa resul to fprolonge dheatin ga ta hig htemperatur ema y servea sa nexplanatio nfo rth elowe rviscositie so fth ejuic e reconstituted fromth e"atmospheri cevaporation "past ei ncompariso nt o thoseo fth ejuic efro mLuw apast e (Harper& E lSahrig i1965) . Harper& E lSahrig i (1965)studyin gth eviscometri cbehaviou ro fho t breaktomat oconcentrate sprepare d ina conventiona lvacuu mpa nreporte d thatdilutio no fa 30 %T Spast et o25 %T So r16 %T Sresulte d ina 20 %gros s viscosity losswhe ncompare dt oth egros sviscosit yo fconcentrate sa ta correspondingsolid slevel .The yexplaine dthi sfindin gb yassumin gsom e 56 experimentalerror .Dilutio no fa 30 %T Sreconstitute dpast eobtaine db y theseru mevaporatio nmetho dresulted ,o nth eothe rhand ,i ngros s viscosity losseso fapproximatel y70 %whe ncompare d toth eorigina l vacuumpa nconcentrate so fth esam esolid scontent . Marsh (1977,1982 )reporte d thatth emor ewate rwa srequire d todilut e asampl e (Pfaudlerswept-fil mevaporator )t o12°B xth ehighe rwa sth e Bostwickvalu ereporte dfo rth eproduc tan dth efurthe ri tdeviate dfro m thevalu ebefor econcentration .A pasteurizin ghea ttreatmen tfollowe db y storagecause da nincreas ei nthi seffect .Th eexten to fdesiccatio no fth e waterinsolubl efractio no fth esolid san dthei rinabilit yt oresor b maximally ispropose da sth ecaus eo fth ephenomenon .Th emechanis mo f resorptionwa spartiall yrestore dfollowin ga hea ttreatmen to fth edilute d paste(3 0min . 100°C).A compilatio no fth eresult so fMars he tal .1982 , 1983an d198 4showe da goo dcorrelatio n (r=0.92)betwee nth etru eBostwic k valuea t12°B xan dth eBostwic ko fa 12°Bx ,80° Cheate ddilutio no fa hig h solidspaste .Tomat opure ea t12°B xwit ha Bostwic ko f2- 6 cm (30sec , 20°C)wil llos e1- 2c mBostwic kreadin ga sa resul to fconcentratio nt oa highsolid spast e (30%TS). Luhe tal .(198 4b )foun dtha ttota lpecti n (drybasis )an dseru m viscositydecrease ddurin gconcentratio nan dsubsequen tpasteurization . Theystat etha ti nth eindustria lremanufactur eo ftomat oproduct sfro m pastes,suc ha sketchu pan dsauces ,ther ei salway ssom elos so fgros s viscosity inth ereconstitute dproducts .Th ephenomeno nma yb epartiall y explained,accordin gt oLu he tal .(198 4b) ,b yth echang ei nsiz ean d solubilityo fth epecti nan dcellulos epolymer sdu et oth econcentratio n andheatin gprocesses .

Effect of homogenization on gross viscosity

Iti sclea rfro mth epreviousl ydiscusse d literaturetha ta tomat o juicecontainin glarg equantitie so fsuspende dmatte rwil lhav ea hig h grossviscosity .However ,i ti slikel ytha tgros sviscosit y isno tgoverne d byth equantit yo fcellula rmateria lalone ,bu tals ob yth econfiguratio n orshap eo fth eparticles .I ti swel lknow ntha telongate dparticle shav ea greatereffec to ngros sviscosit ytha nmor espherica lparticles .(Han de t 57 al. 1956,Moye r et al. 1957,Whittenberge r &Nuttin g 1957). Several authors have shown thatmechanica l treatment may increase grossviscosity . In general, these increases in grossviscosit y are attributed to a decrease in cell size,a change in cell shape (more elongated or linear)an d an increase in total surface area.However ,th e effect ofhomogenizatio n seems to depend largely on themetho d of juice preparation,th e typeo f homogenizer and themetho d of grossviscosit y measurement,causin g the data from the literature tob e somewhat confusing. Therefore,literatur e on the effects ofhomogenizatio n ispresente d here in chronological order. Luh et al. (1954) found that homogenization with a laboratory homogenizer increased the grossviscosit y ofho tbrea k (disintegration of tomatoes inboilin gwate r for 15min. )tomat o paste and puree. Hand et al. (1955)presente d photographs illustrating thepositiv e effects of partial and complete homogenizationwit h a piston type homogenizer on grossviscosity . Partialhomogenizatio n resulted in fragmentation of the predominantly spherical cellular particles;photograph s of a completely homogenized juice show abrush-hea p type of structure contributing to the increased grossviscosity . Robinson et al. (1956) studied the factors influencing thedegre e of settling or serum separation in tomato juice.Homogenizatio n reduced the average particle sizebu t increased thevolum e of particles after centrifuging.Th e increase involum ewa s associated with the rupture of the cellular envelope.Th e fragmented cells did not settle as compactly as the intact cells. Whittenberger &Nuttin g (1957)dissecte d tomatoes into four separate tissue fractionswhic hwer e converted tojuices .Th ejuice s containing cellwall s (juices from shell tissue,cente r tissuean d seed envelopes)wer e thickened by increasing the linearity and surface area of thewall s through homogenization ina Warin g blender.Althoug h enzymatic digestion of pectin ina juice lowered grossviscosity ,gros sviscosit y was restored by mechanically changing the structure ofpecti n free cellwalls .Partia l enzymatic digestion of cellulose in the cellwall s irreversibely lowered grossviscosity . However,maximu m grossviscosit y was obtained by treating the original pectin containing juice in theblender .Evidentl y the combined effects of dissolved pectin in raising the serumviscosit y and of insoluble 58

pectini nraisin gth ewate rholdin gcapacit yo fth efragmente dcel lwalls , moretha noffse tth egreate rwal lfragmentatio ni nth epecti nfre ejuice . Theviscosit yo fth eseru mremaine dalmos tconstan tdurin gth eblende r treatment.Whittenberge r& Nuttin g (1958)showe dtha tbot hho tan dcol d breakjuice sthickene dupo nhomogenizatio n (2min .Warin gblender) .Washe d juicesan dcellulosi ccel lwall swer eals osusceptibl et omechanica l break-up;i nmos tcase sth eremova lo felectrolyte saccentuate dth e thickeningeffec to fhomogenization .However ,a ho tbrea kjuic esampl etha t thickenedmos tupo nwashin gdi dno tincreas efurthe ri nthicknes supo n homogenization.Microscopi cexaminatio nshowe dtha tmos to fth ecel lwall s remainedwhol ean dunbroken .Apparentl yi nthi scas eth erelativel ylarg e quantitieso fpecti csubstance si nth ewall sacte da splasticizer san d rendered thecel lwall sles sbrittl ean dmor eresisten tt omechanica l stress. Ephraime tal . (1962)showe dtha tmil do rdrasti chomogenizatio no f unheatedhan dstraine djuic eresulte di ngrea tgros sviscosit y increaseso f 60-70%a sresul to fth eproductio no fmor eparticle so fsmalle rsize . Beckere tal .(1972 )state dtha ti ncol dbrea kjuice s (delayo f1 hou r betweenmaceratio ni na Warin gblende ra tlo wspee dan dheatin gt o95°C ) theplan tenzyme sar eabl et oattac kpecti nan dothe rcel lwal lbindin g polymers.Degradatio no fthes ecementin gpolymer sfree sth ecel lwal l fibers,whic ho nblendin g (Waringblender ,hig hspeed )becom eentangle dan d raiseth egros sviscosity .Sinc eth ecol dbrea kjuic elack sth edispersin g effectso fthes epolymer sther ei sonl ya moderat egros sviscosit y increase.I nho tbrea kjuice s (maceratebrough tdirectl yi nwate ro f93°C , enzymeinactivatio nwithi n9 sec. )th eenzymati cdigestio no fth ebindin g polymersi sprevente dan dth ecel lwal lcemen tpreserved .Therefor eth e cellwal lfibre sremai nentwine dan dth ecel li sgenerall yimperviou st o blending.Treatin gho tbrea kjuice swit haci dsolubilize sprotopecti nfro m thecel lwalls ,increase sseru mviscosit yan ddecrease sth e% o fintac t cells.Th ecel lwall ,ri do fit sintercellula rcement ,disintegrate so n blendingint oa "brus hheap "o ffibrils .Thes estrand sbecom eentangle d withstrand sfro mneighbourin gbrus hheap san djuic egros sviscosit y increasessevera ltimes .Th eviscou sseru mhelp skee pth ebrus hheap s distributedthroughou tth esolution .Th eseru msolubl eviscosit ycomponen t liberatedb yth eaci dtreatmen tmus tten dt oremai nentrappe d inth ebrus h 59 heapsunti lblendin g occurs.I t then escapes to increase the serum viscosity. Individually neither the serum solublepolymer s nor the cell wall fibers can cause grossviscosit y changes.Whe n ajuic e contains both large amounts of dispersed cellwal l fibersan d ahig hviscosit y serum,th e grossviscosit y isgreatl y elevated. "Slowho tbreak "juice s (heating macerate to 93°Cwithi n 45 sec.)gav e slight increases ingros s viscosity uponhomogenization . Crandall &Nelso n (1975)showe d thatmechanica l treatment (commercial high speed rotary mill) decreased grossviscosit y and serumviscosit y of the juice prepared by theboilin g breakmetho d (tomatoes 45 min. inboilin g waterbefor e crushing). The grossviscosit y of "modified hotbreak " juice (macerate heated to 93°Cwithi n 2min.) ,measure d by efflux,i s increased by mechanical treatment.Th e efflux measurements of themodifie d cold break juice (delay of 1hou rbetwee nmaceratio n and heating to 93°C) showed that mechanical treatment had only a slight positive effect on grossviscosity . Serumviscosit y of themodifie d hot-an d cold break juices are not significantly changed whenmechanicall y treated. Bostwickmeasurement s of puree samples (15°Bx)sho w significant increases ingros sviscosit y after homogenization for theboilin g break and themodifie d hotbrea k method; modified coldbrea k puree did not react uponhomogenization .Whe n puree sampleswer e diluted andmeasure d with the efflux pipette itwa s found that all samples,excep t theboilin gbrea k puree of onevariety ,decrease d or remained unchanged in grossviscosit y after homogenization. Ingenera l literature shows that thegros sviscosit y of all typeso f tomato juice increased uponhomogenizatio nwhe n the right typeo f homogenizerwa s chosen.Ther e are indications that cellwall s containing large amounts of pectin are lessbrittl e and need more power or pressure to behomogenize d than cells ofwhic h thepecti nha sbee n removedb y acid or pectolytic enzymes.However ,onc ehomogenized , thepecti n containing walls raise grossviscosit y more thanwall s devoid of pectin.Concentrates , probably due to thehighe r amount of cellspresent ,see m to respond more quickly tohomogenizatio n than dojuices . 60

Effect of additives on gross viscosity

Theadditio no fcompound ssuc ha sacid ,salt san dsugar st otomat o juicesan dproduct sha ssometime sbee nreporte da sinfluencin gviscosity . Acidha sbee nadde dt otomat oproduct sa tthre estage so fprocessing : 1)befor ehea ttreatment ,fo rth epurpos eo fenzym einactivatio n (Wagner& Miers 1967),2 )t ojuice ,fo rth epurpos eo facidifie dbul kstorag e (Dougherty& Nelso n1974 ,Sidh ue tal .1984 )an d3 )i nth eformulatio no f ketchupsan dothe rsauces . Wagnere tal .(1969 )observe d increasesi ngros sviscosit ywhe naci dwa s addedprio rt oth ehea ttreatment .The yattribute d theincrease st oquicke r enzymeinactivation ,increase dextractio no fpecti nan dimprove ddispersio n ofcel lwal lcomponents .Crea n(1969 )foun dtha tth eextractibilit yo f pectinfro mcel lwal lmaterial swa sno tsignificantl yaffecte db ypH ,an d henceconclude d thatchange si ngros sviscosit ywer ecause db yphysica l changesi ncel lwal lmaterial san dsolubl epectins .Th eeffec twa s reversible;th eorigina lgros sviscosit y couldb ecompletel yrestore db y readjustingth ep Ht oit sorigina lvalue .Takad a& Nelso n (1983)reporte d thatviscositie so fa mode lsyste mo ftomat oproduc t (citruspecti n+ bovineseru malbumin) ,o fdilute dtomat opure ean do f10°B xtomat opure e wereaffecte db yp Hchanges .Maximu mviscosit ywa sobtaine da tp H4.2-4.4 , theorigina lviscosit ywa srestore dafte rreadjustin gth epH .The y explainedthes efinding sb yassumin ga reversibl eelectrostati ccomple x formationbetwee npecti nan dprotein . Dougherty& Nelso n (1974)an dSidh ue tal .1984 )state dthat ,i nbul k storage,th eacidificatio no ftomat ojuic ewoul dreduc eth erequire dhea t treatmentan dai di npotentia lspoilag econtro ldurin gstorage .Doughert y& Nelson (1974),however ,foun ddecrease si njuic egros sviscosit ywhe n tomatojuice swer estore da tp Hlevel sfro m2 t o4 fo ra perio do fu pt o3 months.Thes egros sviscosit y losseswer eirreversibl ean dwer eno tcause d byth ene tadditio no fNaC l(0.7 %whe nacidifie d top H2 )afte r neutralizationo fth ejuic et op H4.3 .Hydroge nbondin gan dothe r associationsamon gpectins ,cellulose sand/o rothe rconstituent scoul dhav e beenpermanentl ydisrupte da sa resul to facidification . (Dougherty& Nelson1974 ). Gros sviscosit yo fcol dbrea kjuice swa sno taffecte db y storagefo r3 month sunde raci dcondition s (pH1.3-1.4) .Acidifie djuice s 61 showed a slight off-flavour afterbringin g pH to 4.45;however ,thi swa s not detected inketchup sprepare d from acidified coldbrea kjuice s (Sidhu et al. 1984). Twigg (1959)detecte d increases inketchu p viscosity with a decrease inpH , and contributed this to increases in theviscosit y of low esterpectins . Results presented by Whittenberger and Nutting (1957, 1958) indicate that the cellwal l surfaces are electrically charged,helpin g tomaintai n walls in suspension.Th e addition of 0.2% sodium chloride and calcium chloride towashe d cellwal lmateria l dampened the charges considerably, causing decreased viscosity. Twigg (1959)foun d that the addition of approximately 8.5 g/1calciu m chloride to tomato juice increased ketchup grossviscosit y and decreased serum separation. Thevalue s forwate r soluble pectinwer e greatly reduced,whil e the amounts of oxalate soluble pectin greatly increased. Cooler (1962)showe d that calcium chloride increased the grossviscosit y of steamblanche d tomato juicebu t had no effect on coldbrea kjuices .Magnesiu m chloride decreased the gross viscosity of the steamblanche d juice.H e concluded that electrolytes are important secondary agents imparting electroviscous properties to macromolecules promoting thickening by extension and thinning by contraction. Dougherty &Nelso n (1974)di dno t detect any decrease in juice grossviscosit y asa result of the additiono f 1% sodium chloride. Sherkat & Luh (1977),however ,presente d resultswhic h clearly show that the addition of 0.8% sodium chloride toreconstitute d 5°Bxjuice ,o r 1.6% sodium chloride to reconstituted sauce of 12°Bx,resulte d in lower gross viscosities due to thedecreas e in thehydratio n of cellulose,pectin , proteins and otherpolymers . Bel-Haj (1981)reporte d inhi s literature review that the apparent viscosity of pectin solutions is increased by theadditio n of sugars such asglucose ,maltos e and sucrose.Th e increase inviscosit y isattribute d to the formation ofH-bond s and aggregates of sugar-pectin molecules,a swel l as to the dehydrating effect of the sugar.

Effect of additional heat on gross viscosity

Usually tomato products receive a heat treatment,befor e or after 62 packing,i norde rt opreven tspoilag eb ymicroorganisms .Thi sadditiona l heattreatment ,however ,ma yaffec tth eviscosit yo fth etomat oproduct . Mierse tal . (1970)stresse dtha tprocessor sshoul dkee pheatin gt oa minimumonc eth ecrushe dtomatoe sar efinished ,especiall ywhe nthe yadap t techniquesb ywhic hproduct shavin ga hig hviscosit yar eproduced .Juice s havinghig hgros san dseru mviscositie slos t35-65 %o fth eseru mviscosit y and 10-18%o fth egros sviscosit yafte r1 0min .heatin ga t100°C .Viscosit y losseswer elowe ra tlowe rtemperature san dwit hjuice shavin glowe r initialviscosity .Stephen se tal .(1970 )detecte dgros sviscosit ylosse s of 10-20%afte r 10min .heatin ga t88°C ;especiall yhig hviscosit yjuice s wereheat-sensitive .Han de tal .(1955 )foun dn odifference si njuic e grossviscosit yafte rpasteurizatio na t93 ,12 1o r129°C .However ,the yd o notmentio nth egros sviscosit ybefor epasteurization . Mierse tal .(1970 )explaine d thelos si nviscosit yb yhydrolysi so r othertype so fdegradatio no fth epolymeri ccomponent spresen ti nth e tomatojuic ebot hi nth eseru man dth einsolubl efraction .Mier se tal . (1970)presen textensiv eliteratur edat aindicatin gtha tth ea trando m cleavageo fa limite dnumbe ro fbond si na hig hmolecula rweigh tpecti nma y resulti na viscosit y losso fove r50% .

brookfield viscosity (m Pa. s) ivwv-

5000- y y 7000- y y? A s ' yp 1000- y • s s -;y 500- sy 200- 4'' |2 •---• SS'L 14 o-—o , 12 » i IUU-V canned |,4 4 4 \ 1— i i i 100 50 20 10 2.5 rp.m.

Fig.9 :Effec to fpasteurizatio no ngros s viscosityo ftomat ojuice sreconsti ­ tuted frompaste . Theho tbrea k temperaturewa s100° C forsampl e2 an d64.4° Cfo rsampl e4 (from Sherkat& Lu h 1977). 63

Theyconclud etha ttomat oproduct shavin ghig hviscosit yan dhig hmolecula r weightpectin scoul dlos egros sviscosit yver yquickl ya sa resul to fhea t byth ecleavag eo fonl ya fe wbonds .Indeed ,Lu he tal . (1984a)foun da definitedecreas ei nweigh taverag emolecula rweigh to fpectini cacid san d ofteno fpecti cacid safte rho tbrea k (40sec .107°C) ,concentratio n (30min . 45°C)an dpasteurizatio n (30min .100°C) . Sherkat& Lu h (1976,1977 )showe dtha thea tprocessin go f26°B xpast e for3 0min .i nboilin gwate rresulte d insignificantl ylowe rgros s viscosityafte rdilutio nt o5°B xwhe ncompare dt osample sfroze na t-23°C . Thenegativ eeffec to fpasteurizatio nwa smuc hhighe rfo rsample shavin ga preheattreatmen ta t10 0° Ctha nfo rsample spreheate da t64°C .(Fig . 9). Heatprocesse dpaste scontaine d lowertota lpecti nan dprotopecti ncontent s thanfroze npastes .

Effect of storage on gross viscosity

Thedat ao nth einfluenc eo fstorag etim ean dstorag etemperatur eo n thegros sviscosit yo ftomat oproduct sar erathe rlimited .Coole r(1962 ) foundtha tho tbrea kjuice sthinne ddurin gstorag ewhil ecol dbrea kjuice s wereno taffected .H eassume dtha tho tbrea kjuice sexiste da swea kgels , whichseldo mattai na nequilibriu mstate .Mannhei m& Kopelman n(1964 ) reportedtha tth egros sviscosit yo fcol dbrea kpast edi dno tchang edurin g astorag eperio do f1 4weeks ;ther ewa sn odifferenc ebetwee nstorag ea t roomtemperatur eo rstorag ea t37°C .Mars he tal .(1984 )studie dth e effectso fstorag eo nth egros sviscosit yan dsolid sconten to fho tbrea k tomatojuice san dintermediat eleve lconcentrate sprepare d from1 2 varieties.Storag eo fu pt o1 8months ,a tambien ttemperature ,ha dn o distincteffec to ntota lsolids ,solubl esolid san dWIS/T Srati oo fth e juice.Seru mviscosit ywa sfoun dt odecreas ewhil egros sviscosit y (Bostwick)tende dt oincreas eslightly .Storag eha dalmos tn oinfluenc eo n theBostwic kviscositie so fintermediat eleve lconcentrate s (10-15°Bx). Marshe tal .(1982 )presente dresult swhic hsho wtha tth egros sviscosit y of12°B xdilution so fhig hsolid spaste sdecrease da sa resul to fstorag e ofth epast efo r9 month sa tambien ttemperature ;gros sviscosit yo fth e correspondingjuic esample so nth eothe rhan dincrease d slightlya sresul t ofstorage . 64

2.4. Factors affecting serum separation in tomato products

The degree of serum separation,whic h is the tendency of the insoluble solids to settle out,leavin g apracticall y clear liquid at the top of the tomato product,i s considered to contribute to the overall term consistency. For tomato juice and tomatoketchup ,consistenc y is an important parameter indeterminin g the quality grade.Seru m separation is oftenmeasure d by determining theBlotte rvalu e (distance of travelo f serum in filter papermeasure d from the edge of amoun d of tomato product), thedegre e of settling (volume of serum at the top ina graduated cylinder as % of the totalvolume )o r the rate of filtration (volume of filtrate obtained after filtering a certain amount of tomato product). Robinson et al. (1956)reporte d that,a s a general rule,ther e isa n inverse relationship between the degree of settling and the grossviscosit y of tomato juice.The y state that thephenomeno n of settling in ordinary tomato juice isno t a simple process of sedimentationbu t involves a closer packing together ofparticle s in contactwit h each other. In this context the shapeo f theparticle s seems tob emor e important than average particle size. Itwa s found that thedegre e of settling isdetermine d by the amount of insoluble solids in suspension and by the extent towhic h the intact cells are disrupted. Ruptureb y homogenization reduced the degree of settling. The amount of totalpecti n in tomato juicedi d nothav e amajo r effect on the degree of settling. The formation of a clear serum in theuppermos t region of the juice column canb e explained by twodistinc t processes (Shomer etal . 1984): 1) toomuc h liquid, forwhic h a given concentration of insoluble particles cannot fill thevolum e of the liquid column; thisma y occurwithi n several hours of shelflife;an d 2)a slowly diminishing volume of the precipitate during storage asa result of gradual collapse under gravity stress.Shome r et al. (1984)confirme d the conclusion ofRobinso n et al. (1956)tha t serum hasn o effect on the settling of theparticles . Intact or disrupted cell walls from tomato pericarp could notb e suspended in thejuice ,eve nwhe n theviscosit y and/or density of the serumwa s increased by the addition of high esterified pectin and sucrose.Henc e they considered tomatojuic e to be a swelled precipitate rather than a suspension.Pectinas e addition 65

resulted ina n increased precipitate volume due to swelling of thewal l and partial dispersion of themicrofibrila r system. Cellulase addition led to partial or complete degradation of themicrofibrila r systemwit h a resultant collapse of theprecipitate .Homogenizatio n increased thevolum e of the precipitate.Th e addition of pectinase after homogenization gave an additional increase,th e addition of cellulase resulted in collapse to the level of intact cells treated with cellulase. Studieswit h light and electronmicroscope s led to the supposition that the cellulose microfibrilar system formed and supported the construction of the precipitate. Shomer et al. (1984)therefor e concluded that the inactivation of pectinases isno t the reason for an improved homogeneous appearence as a result of hotbreak . They suggested that if endogeneous enzymes affect the formation of a clear serum at the top of the juice column,thes e enzymes are apparently cellulases.Th e activity of cellulases during the ripening, postharvest and processing stagesma y decrease the ability of the microfibrilar system towithstan d collapse under gravity stress.I n contrast to thewor k of Shomer et al. (1984), Smit &Nortj e (1958) found that the addition of Pectinol A topast e reduced grossviscosit y and serum viscosity and increased the degree of settling.However ,the y did not take into consideration the fact that PectinolA also contains cellulase activity. Factors influencing the degree of serum separation inketchu p were studied by Twigg (1959). Neither grossviscosit y nor serum viscosity correlated wellwit h serum separation asmeasure d with the Blotter-test. Multiple regression analyseswit h the four compositional factors: structural solids,no n structural solids,tota l pectin and pH,yielde d a correlation of 0.69 with serum separation.Th e relative influence of the four factors on explainable serum separation differenceswere : structural solids 57%;no n structural solids 22%;p H 19%an d totalpecti c substances 2%. All factors except pHha d a negative relationship with serum separation. Cooler (1962)conclude d that the Blotter-test measures the mobility of solution,whic h isrelate d to serumviscosity , to the degree of entrapmentwithi n the suspended particles and also to the amount of serum relative to the suspended particles. Marsh et al. (1979b ) showed that grossviscosit y and serum separation were unrelated quality attributes inketchup .Whe n grossviscosit y was 66

standardized to a given Bostwick level,seru m separation varied from none to considerable. In contrast toal l above cited authors they reported that serum separation depends on serumviscosity . Retention of at least 80%o f the tomatoes' original serumviscosit y isrequire d forminima l serum flow. Components responsible forhig h serumviscosit y apparently also control serum separation. Increases inho tbrea k temperature,finishe r paddle speed and finisher screen size,a swel l ashomogenizatio n of the suspended particles,wer e found to decrease the level of serum separation in tomato juice and ketchup (Robinson et al. 1956,Smi t &Nortj e 1958,Twig g 1959,Coole r 1962). Storage of ripemachin e harvested fruits,o n the other hand, resulted in large serum separation increases. (Marsh et al. 1979b) .

2.5. Composition of tomato polysaccharides

Inparagrap h 2.3.4.o f this literature review the importance of tomato polysaccharides as contributors to grossviscosit y of tomato products was discussed. Tomato polysaccharides are isolated by preparing alcohol insoluble solids (AIS)o rwate r insoluble solids (WIS). TheAI S covermor e or less the complete polysaccharide contentwhil e WIS represent only the cellwal l polysaccharides.Larg e differences are found inamount s ofAI S andWI S in tomatoes and tomato products (Table 9);a result of differences invariet y and degree ofripeness . Complete chemical analyses ofAI S orWI S from ripe tomatoes are rarely found in literature. Gross &Wallne r (1979)an d Gross (1984), analysing isolated cellwall s of outer pericarp from ripe tomatoes,foun d high amounts of uronic acid,cellulos e and galactose (Table 10).However ,mos t authorsus e the complex of polysaccharides for fractionation and particularly for the isolation of pectin fractions.Gros s (1984)fo r example fractionated cellwall s in 22.7%ionicall y associated pectin (chelator soluble),9.7 % covalently bound pectin (alkali-soluble),19.3 % hemicellulose and 48.3%cellulose . Jona &Fo a (1979) found that tomato cell wall pericarpwa shighes t inpecti c substanceswhe n compared to other fruits andvegetable s as peach,pepper ,strawberry ,cherry ,appl e and grape. In general,th e pectic substanceswer e themai n component of fruit 67

Table9 :% AI San d% WI S intomat ofrui tan djuic e (freshweigh t basis).

%AI S %WI S Product Reference

1.3-1.9% tomato McColloch etal .195 0

1.96-2.04% tomato Woodmanseee tal .195 9

0 66-•0.90% tomato Stevens197 2

1.01-2.32% tomato Janoria1974 ,Janori a& Rhode s197 4

1.40-1.73% tomato Stein& Brow n197 5

1.20-2.71% tomato Stevens& Paulso n197 6

1.22-1.71% tomato Brown& Stei n197 7

1.2% tomato Malis-Arad198 3

1.08-1.82% juice Cooler196 2 0.80-0.85% juice Wagner et al. 1969 0 54-•1.42% juice Marsh et al. 1979-1984 0 79 •1.07% juice Luh et al. 1982, 1984

Table 10: Composition of isolated cell walls of ripe tomatoes (% w/w).

Gross &Wallne r 1979 Gross 1984

Rham. 1.5 1.2 Ara. 2.7 3.0 Xyl. 3.4 4.3 Man. 2.3 2.4 Gal. 6.3 4.0 Glue. 2.2 2.2 AGA 30 25.4 Cellulose 30.0 46.1 Protein 3.8

82.2% 88.6% 68

cellwalls ,followe db ycellulose ,whil ehemicellulos eaccounte dfo ronl y asmal lfractio no fth etota lpolysaccharides .Fo rtomat ofrui tcel lwall s Jona& Fo a (1979)reporte da compositio no f51 %pectin ,18 %hemicellulos e and31 %cellulose .

Pectin

Table1 1show sa compilatio no fliteratur edat ao nth epecti nconten t inrip etomat ofrui tan dtomat ojuice .Tota lpecti nconten trange sfro m 0.2-0.5%o na fres hweigh tbasis ,whic hwil lequa l3.0-7.5 %o na dr ybasis . Luhe tal . (1984b ,198 4c )foun dtha ttota lpecti no ndr ybasi sdecrease d from3.96-6.75 %t o2.92-5.74 %whe njuic ewa sevaporate d topast ean dfro m 5.46-8.94%t o4.95-6.25 %whe ntomatoe swer eprocesse d topaste .Expresse d as% o fAIS ,tota lpecti nrange sfro m14-26 %i nfres htomatoe s(Woodmanse e 1959,Stephen se tal .1970 ,Stei n& Brow n1975 ,Malis-Ara d 1983).Stephen s etal .197 0showe dtha tprocessin gtomatoe sint ojuic e (hotbrea kmethod ) decreasedthi spercentag efro m18.1-19.2 %fo rtomatoe st o10.8-13.3 %fo r juice.Tabl e1 1als oshow sa larg ediversit yi nth econtributio no fth e differentpecti nfraction st oth etota lamoun to fpectin .Thes edifference s areprobabl ycause db ydissimilarit yi nfractionatio nmethods ,althoug h varietyan dripenes sleve lcertainl ypla ya role . Littleinformatio ni savailabl eo nth echemica lcompositio no fth e pectino rth epecti nfractions .Lu he tal .(195 4a )determine dth e anhydrogalacturonicaci d (AGA)conten tan ddegre eo festerificatio n (DE)o f thepecti nfraction si ntomatoes .The yfoun dAG Avalue so f68 ,78 ,4 7an d DEvalue so f75 ,42 ,5 6fo rwater ,oxalat ean dHC 1solubl epecti n respectively.Othe rD Evalue sfoun di nliteratur ear e5 5fo rwater ,4 0fo r oxalatean d3 5fo rHC 1solubl epecti n (Beckere tal .1968) .Th ewor ko f Brown& Stei n(1977 )indicate stha tgalactos ei sth emos timportan tneutra l sugari nth epecti nfraction s(approximatel y50 %o ftota lneutra lsugars) . Otherneutra lsugar sanalyse dwer exylose ,arabinose ,ribos ean drhanmose ; nosignifican tdifference si nth ecompositio no fth epecti nfractio n betweencultivar swer efound .Th ecompositio no fth eionicall yassociate d 69

S a o «!

3 a

3

3 o

w H H H 3 J) •a , . O i-l rH *-• ct) 0> i 4-» 4J ^ a^ •cH 1-1 a> <0 r-l rH •4J •5 •§ w V3 5 _) J 70 pectin(I.A.P. )an dcovalentl yboun dpecti n (C.B.P.)a sdetermine db yGros s (1984)i sshow ni nTabl e12 .Th emos timportan tneutra lsugar si nthes e pectinfraction sar earabinose ,galactos ean drhamnose .

Table12 :compositio no ftw ocel lwal lpecti nfraction s (%w/w) .

I.A.P. C.B.P.

Rham. 1.8 1.6 Ara. 3.8 3.3 Xyl. 0.5 0.8 Man. 0.2 0.2 Gal. 2.0 1.6 Glue. 0.5 0.3 AGA 69.8 15.9

total 78.6% 23.7%

Iti sobviou stha tonl ya smal lpar to fth eC.B.P .fractio ni sactuall y pectin. Luhe tal . (1984a ,198 4c )establishe dmolecula rweight so fpectin s byviscosit ymeasurements .The yrepor tweigh taverag emolecula rweight so f 7,870-26,800 daltonsfo rpectini cacid san d6,600-24,00 0dalton sfo rpecti c acid,dependin go nvariety .Th eprocessin go ftomatoe st opast edecrease d themolecula rweigh to fpectins .Stei n& Brow n(1975 )repor ta molecula r weighto f >2x 1 0 foral lth epecti nfraction sa sdetermine dwit hge l filtration.Thes evalues ,however ,ar eprobabl yoverestimate ddu et ocharg e effectsa sa resul to felutio nwit hwater .

Hemicellulose

Generallyhemicellulos ei ssolubilize dfro mdepectinize dAIS ,WI So r cellwal lmateria lb yth eadditio no fconcentrate dalkal i( 4M) . Herranze t al. (1981)repor ta hemicellulos econten ti nth eedibl epar to ffres hrip e 71

tomatoes of 2.06% ± 1.06% on drybasi s and 0.13% ± 0.07% on freshweigh t basis. Gross (1984)foun d ahemicellulos e content of 17.2%afte r fractionation of cellwal lmaterial .Fro m findings ofWilliam s & Bevenue (1954) it canb e calculated thathemicellulos e approximates 4.0% ondr y basis and 21%o nAI Sbasis . In their study Gross &Wallne r (1979) conclude thathemicellulos e isno t degraded during tomato fruit ripening;however , Huber (1984)clearl y shows that thehemicellulos e of ripe fruits contains lower quantities ofhig hM W polymers and higher quantities of lowM W (<40,000)polymer swhe n compared to thehemicellulos e ofunrip e fruit. Williams and Bevenue (1954)foun d thehemicellulos e fraction to consist of 37.5%xylos e and glucose containing polymers and 15%protei n while 47.5%wa s not identified. Thehemicellulos e fraction isolated by Gross (1984)containe d 21.9%glucose , 19.5%xylose ,8.4 %mannose ,4.3 % galactose,4.0 % arabinose and 4.3% anhydrogalacturonic acid, leaving37.6 % forunidentifie d components.Hollowa y & Greig (1984)presen t relative compositions ofhemicellulos e fractions of tomato and other fruits and vegetables (Table13) .

Table 13:Relativ e composition ofhemicellulos e of several fruits and vegetables.

Rham. Ara. Xyl. Man. Gal. Glue. AGA Methoxyl Acetyl Protein

tomato 0 3.7 27.0 12.1 6.9 31.9 9.3 0.5 0.4 8.2 pear 1.6 17.2 19.8 3.8 13.4 27.7 8.3 0.7 1.7 5.8 peach 1.8 9.4 18.3 4.2 11.9 19.6 9.5 1.3 2.5 21.4 cabbage 1.8 14.3 18.0 7.5 31.4 22.9 16.9 1.3 3.9 11.6

Noteworthy are the low amounts of arabinose and galactose and thehig h amounts ofmannos e in thehemicellulos e of tomatoeswhe n compared to other fruits andvegetables . 72

Cellulose

Theresidu eo fAIS ,WIS ,o rcel lwal lmateria llef tafte r solubilizationo fpecti nan dhemicellulos ei sconsidere d tob eth e cellulosefraction .Value sreporte dfo rth ecellulos econten ti ntomatoe s andtomat ojuic ear ei nth erang e0.3-0.6 %o nfres hweigh tbasi s (Luhe t al.1960 ,1982 ,198 4b,c ;Herran ze tal .1981 ;Southgat e 1976).O nth e basiso fcel lwal lweight ,Gros s& Wallne r (1979)an dGros s (1984)repor t valueso f30.0 %an d46.1 %respectively .Th ecellulos efractio na sisolate d byGros s (1984)i scompose do f68.5 %cellulose ,11.9 %no ncellulosi c neutralsugar san d0.4 %anhydrouroni cacid .

Protein

Afourt himportan tcomponen ti nAI So rWI Si sprotein ,whic hi sno t isolateda sa separat efractio nbu ti sprobabl yinclude dbot hi npecti nan d hemicellulosefractions .Th eamoun to fprotei ni sgenerall ycalculate dfro m thetota lnitroge nconten ta sdetermine db yth eKjehldah lmethod . Asha sbee nshow nbefor e (Table8) , theprotei nconten ti nAI Srange s between20-30% .Gros s& Wallne r (1979),however ,repor ttha tisolate dcel l wallmateria lcontaine donl y3.8 %protein .Thi sdiscrepanc yi sprobabl y causedb ya differen tmetho do fsampl epreparation . 73

GENERAL ANALYTICAL METHODS

3.1. Viscosity measurements

Cross viscosity Thegros sviscosit yo ftomat oproduct s (juicean dconcentrate sa t several°B xlevels )wa sdetermine dusin gth eHaak erotovisc oRV ,a rotationalviscometer ,an dth eBostwic kconsistometer ,a ninstrumen t commonlyuse di nth etomat oindustry .Th eBostwic kconsistomete rconsist s ofa rectangular ,graduate d troughdivide di ntw osection sb ya sprin g operatedgat eassembly .Th esmalle ro fth etw osection sserve sa sth e reservoirfo rth emateria lt ob etested .Openin go fth egat eresult si n flowo fth esampl ethroug hth etrough .Gros sviscosit ydat ameasure dwit h theBostwic kconsistomete rar ereporte da sc mflo wafte r3 0sec .a t20° C (Davise tal . 1954).Whe nuse dfo rmeasurin ggros sviscosit yth eHaak e

rotoviscowa sequippe dwit hth eM Vsenso rsyste man dth eMV T rotarybob . Grossviscosit ydat awer ecalculate d fromth etorqu esigna lS ,give ni n G*5 scalegraduation s0 t o100 ,usin gth eformula :n« —jj — (mPa.s).I nthi s formulaG i sa constan tinstrumen tfacto ran dn i sth eroto rspee d (min ) (Schramm 1981).Gros sviscosit y isreporte da sapparen tviscosit y (n )a t app 20°Cusin ga roto rspee do f24 3rpm . Fig. 10show sth erelationshi pbetwee nth egros sviscosit yvalue sa s obtainedwit hth eBostwic kconsistomete ran dth eHaak erotovisco .Thi s figureindicate stha tth eBostwic kconsistomete ri sno tsensitiv eenoug h formeasurin ghig hviscosit y tomatoproducts .O nth eothe rhand ,however , thisinstrumen ti sver ysuitabl efo rmeasurin gth egros sviscosit yo f tomatoproduct swit ha Bostwic kflo wo f5 c mo rmore .

Serum viscosity Ca. 30g .o ftomat oproduc twa scentrifuge d for2 0min .a t48,00 0x g . ina Sorval lRC-5 Brefrigerate dsuperspee dcentrifuge .Th eseru mwa s filteredthroug hWhatma nGF/ F (0.7um )an dSchleiche r& Schiil l(0.4 5urn ) membranefilters .Th eviscosit yo fth eseru mwa smeasure dwit hth eHaak e rotoviscousin gsenso rsyste mNV ,an dexpresse da smPa. sa ta shea rrat eo f 873.3sec -1 (20°C). 74

Topp' mPa.s) o

1500

1000-

500-

< < "* *> kS->«4as-«o8Q<_0

-1— 10 15 20 Bosl-wick (cm/30 sec. ,20°C)

Fig.10 :Relationshi pbetwee ngros sviscosit yvalue smeasure dwit hbot h theBostwic k consistometeran dth eHaak erotovisco . Bostwick cm - 237.3 x 1/Al - 5.2 (n - 160, r - 0.995). app

3.2. Determination of solids content

Soluble solids (°Bx) Ca. 30g .o ftomat oproduc twa scentrifuge dfo r2 0min .a t48,00 0x gi na Sorval lRC-5 Brefrigerate dsuperspee dcentrifuge .Afte rfiltratio n ofth eseru mthroug hWhatma nGF/ F (0.7um )an dSchleiche r& SchUl l0.4 5u m membranefilters ,th eamoun to fsolubl esolid si nth efiltere d serumwa s determinedb yreadin gth esuga rscal eo fa nAbb erefTactomete ra t20°C .

Total solids (%TS). Thetota lsolid sconten to fth etomat oproduct swa sdetermine d usingth evacuu move ndryin gmetho ddescribe db yLam b (1977).

Preparation of water insoluble solids (WIS) Ca. 30gram so ftomat oproduc twa scentrifuge dfo r2 0min .a t48,00 0g 75

ina Sorval lRC-5 Brefrigerate dsuperspee dcentrifuge .Th eresidu eobtaine d afterdecantin gth esupernatan twa sextracte dfo ron ehou rwit hwate ro f 50-55°C.Extractio nan dcentrifugatio nwer erepeate dunti lth e°B xi nth e supernatantwa szero .Afte rth elas textraction ,th eresidu ewa s resuspended ina smal lvolum eo fwater ,freeze-drie dan dweighed .Afte r reducingth eparticl esiz eo fth eWI St o± 0. 7m mwit ha hamme rmil l (CulattiDF H48) ,th eWI Swa sstore da t3° Cunti lfurthe ranalysis . Freshtomatoe swer eheate di na microwav eove nt oa hear ttemperatur e exceding90°C .Afte rcooling ,th eskin san dseed swer eremove db yhand . Thefurthe rprocedur efo rth epreparatio no fWI Sfro mfres htomatoe swa s identicalt oth eprocedur edescribe dabove .

Preparation of alcohol insoluble solids (A IS) Twovolume so f96 %ethanol ,containin g0.0 5K HC1 ,wer eadde dt oon e volumeo ftomat oproduct .Afte ra nextractio no fon ehou ra troo m temperature (stirring)th emixtur ewa sfiltere do na Buchne rfunne lunde r suction.Th efiltrat ewa sdiscarde dan dth eprocedur ewa srepeate donc e with60-70 %ethano lan dthre etime swit h96 %ethanol .Afte rth elas t extractionth eresidu ewa sdecoulorize dwit hdiethylethe ran dair-dried . Afterreducin gth eparticl esiz eo fth eresidu ewit ha Culatt iDF H4 8 hammermil lt o± 0. 7mm. ,th eAI Swa sstore da t3° Cunti lfurthe ranalysis .

Preparation of serum-alcohol insoluble solids (s-AIS) Tomatojuic ewa scentrifuge dfo r4 5min .a t25,40 0g i na Sorval lRC-5 B superspeedcentrifuge .Th eseru mwa sweighe dan dfiltere dthroug ha D- l glasssintere dfilte ran dthe nthroug hSchleiche r& SchUl lglas smicrofibr e paperno .8 .Tw ovolume so f96 %ethano lwer eadde dan denoug h37 %HC 1t o maketh esolutio n0.0 5N .Th eprecipitat ewa sfiltere dthroug hline ncloth , washedwit h96 %ethanol ,dissolve di na smal lvolum eo fdistille dwate ran d freeze-dried.

3.3. Chemical analysis of WIS

Neutral sugar analysis Theneutra lsuga rcompositio nwa sdetermine db ygl cafte r 76 hydrolysiso fth epolysaccharide st omonomer san dconversio no fthes e monomerst oaldito lacetates ,accordin gt oth emetho ddescribe db yJone s& Albersheim (1972). Severalmethod shav ebee nuse dfo rth ehydrolysi so fpolysaccharides .2 N TFAhydrolysi s (Jones& Albershei m 1972)an d1 M H,SO ,hydrolysi s (Englyst 1981)hav ebee nuse dfo rth edeterminatio no fnon-cellulosi c polysaccharidesan dbot har ereporte da sgivin gcomparabl eresult s(Englys t 1981).0. 8N H SO ,hydrolysi s (Saeman1963 )an d2 N H SO ,hydrolysi s (Selvendrane tal . 1979),bot hafte rprehydrolysi swit h72 %H SO, ,hav e beenuse dfo rth edeterminatio no ftota lpolysaccharides . Table1 4show sth eeffec to fsom emethod so fhydrolysi so nth eanalyse d sugarcompositio no ftomat oWIS .

Table14 :Suga rcompositio n (%w/w )o ftomat oWI Si nrelatio nt ometho do f hydrolysis.

1hou rprehydrolysi s 3hour sprehydrolysi s

2N TF A 0.8N H.SO . 2N H.SO . 0.8N H SO. 2N H.SO . 2 4 2 4 2 4 2 4

Rhamnose 1.0 0.2 0.4 0.2 0.5 Arabinose 1.3 1.2 1.1 1.2 1.1 Xylose 3.6 3.6 3.3 3.6 3.7 Mannose 1.5 2.7 2.8 2.7 2.9 Galactose 2.6 2.3 2.0 2.2 2.2 Glucose 2.2 20.9 26.6 27.1 28.7

Thedeterminatio no fth eneutra lsuga rcompositio no ftomat oWI Safte r hydrolysiswit h2 N TFA ,result si nunderestimatio no fth emannos econtent . Thereforeth emetho do fdeterminin gth enon-cellulosi cpolysaccharide si n combinationwit hdeterminin gcellulos ecanno tb euse dfo rtomat oWIS . Ingeneral ,th e0. 8N H SO ,an d2 N H SO ,hydrolysi si sprecede db ya prehydrolysiswit h72 %H SO ,fo r1 hou ra t30°C .Tabl e1 4als opresent s resultsfo rth esuga rcompositio no ftomat oWI Swhe nth eprehydrolysi si s extended to3 hour sa t30°C .Th emai neffec to fa nextensio ni n 77

prehydrolysistim ean da nincreas ei nH_SO ,normalit y isa nincreas ei nth e determinedglucos econtent .I ncompariso nt oth e2 N TF Ahydrolysis ,H„SO , hydrolysisresult si nlowe rrhamnose ,galactos ean darabinos econtents .Th e determinationo fth erespons efactor so fth esugars ,relativ et oth e internalstandard ,als oshow stha txylos ei sdegrade dwhe nth eprehydrolysi s stepi sextende dt o3 hours . Noneo fth emethod so fhydrolysi smentione d inTabl e1 4wa sabl et o hydrolyseth ecellulos ecompletely ,a sth edetermine dcellulos econten to f thissample ,accordin gt oth eUpdegraf fmethod ,wa s32.1% .I norde rt o increaseth edegre eo fhydrolysi so fparticularl y thecellulose ,th etomat o WISsampl ewa sprehydrolyse dwit ha nenzym emixtur eprio rt o2 N TF A hydrolysis.Th eenzym emixtur ewa sprepare d fromdesaccharifie dUltrazym e 100(Nov oFermen tAG ,Basel ,Switzerland )an dMaxaym eC L200 0(Gis t Brocades,Delft ,Th eNetherlands) .Ultrazym e10 0wa sdissolve d in0.0 1M sodiumacetat ebuffe rp H5. 0an dultrafiltere di na nultrafiltratio ncel l (AmiconDiafl otyp e50 )equippe dwit ha Y M5 Diafl ofilter .Maxazym eC L 2000wa sdissolve di n0.0 1M sodiumacetat ebuffe rp H5. 0an d desaccharifiedb yge lfiltratio no na Bio-Ge lP-1 0column.(Bio-Ra dLabs. , Richmond,Cal. ,USA ) NovoFermen tA.G .als oprovide du swit hth epurifie d fermentationliquo rfro mwhic hUltrazym ei sprepared .Thi sliquo rlack s maltodextrinsan dtherefor edoe sno tnee ddesaccharification .Maxazym eC L 2000an dUltrazym e 100wer emixe di nsuc ha rati otha tth efina lenzym e mixturecontaine dth efollowin gactivities : -endo-glucanas e (EC3.2.1.4.) :0.5 3Units/m l (onAkz ocarboxymethy l celluloseA F0305) . -exo-glucanas e(E C3.2.1.91) :0.0 1Units/m l (onAvice lcellulose , typeSF) . -cellobias e (EC3.2.1.21) :0.0 9Units/m l (onDifc ocellobiose) . -endo-P G (EC3.2.1.15) :3.5 7Units/m l (onpolygalacturoni cacid , ICN102711) . -P E(E C3.1.1.11) :0.2 1Units/m l(o nappl egree nribbo npectin , DE62.0%) .

Table1 5show sth esuga rcompositio no fth etomat oWI Safte renzymati c prehydrolysisan dfina lhydrolysi swit h2 N TFA . 78

Table15 :Suga rcompositio n (%w/w )o ftomat oWI Safte renzymati c prehydrolysisan d2 N TF Ahydrolysis .

Rhamnose 1.1 Arabinose 1.2 Xylose 4.2 Mannose 3.4 Galactose 2.7 Glucose 31.9

Thismetho do fhydrolysi sobviousl y lacksth edisadvantage so fbot hth e2 N TFAan dth eH-SO ,metho do fhydrolysis ,whil eth edetermine dglucos econten t correspondst oth ecellulos econten ta sdetermine daccordin gt oth emetho d describedb yUpdegraf f (1969). Forroutin eanalysis ,3 m go fsampl ewa sweighe dint oa plasti creactio n vesselan dsuspende d in1 m lo fenzym emixture .A scontrol ,on ereactio n vesselwit henzym emixtur eonl ywa sincluded .Th evessel swer eincubate da t 40°Cfo r4 8hour swit hcontinou sstirring .Subsequentl yth econten twa s transferred toscre wcappe dtube san devaporate d todrynes sunde ra strea mo f aira t40 CC.Th ehydrolysi so fth esampl ewa scomplete dwit h2 N TF A containingmyo-inosito la sinterna lstandard .Aldito lacetate sprepare d accordingt oJone san dAlbershei m(1972 )wer eseparate do na ga s chromatograph (HewlettPackar dmode l575 0G )fitte dwit ha dua lflam e ionizationdetecto ran da colum n (length30 0cm ,i.d .3 mm )packe dwit h3. 7g of3 %OV-27 5o nChromosor bW.A.W .80-10 0mesh .Th etemperatur eo fth e injectionpor twa s250°C ,o fth eflam edetecto r230°C-an do fth ecolum nove n 190°C.Th egaschromatograp hwa sconnecte d toa Laborator yDat aContro lCC M 301compute rwhic hdirectl ycalculate dth eneutra lsuga rconcentratio n relativet oth einterna lstandard .Th eweigh t% compositio no fth eneutra l sugarswa scalculate dan dcorrecte dfo rth ewate ruptak edurin gth e hydrolysiso fth epolysaccharide .A correctio nwa sals omad efo rth eneutra l sugarspresen ti nth eenzym emixture . 79

Anhydrogalacturonic (ACA) content and degree of esterification (DE) TheAG Aconten tan dD Eo fpecti ni ntomat oinsolubl esolid swa s determinedb yth ecoppe rio nexchang emetho ddescribe db yKeybet s& Pilni k (1974)an dKata n& v.d .Bovenkam p (1981).Pecti nwa sprecipitate da scoppe r pectatean dwashe d toremov eth efre ecoppe rions .Th eboun dcoppe rion s wererelease db ywashin gwit h0. 6N HC 1an ddetermine dusin ga natomi c absorptionspectrophotomete r (Perkin-Elmer,mode l2380) . Whenpecti nwa sdegrade dt osuc ha nexten ttha tn oprecipitatio noccurre d withcoppe rsulphat e (Siliha 1985),th eAG Aconten twa sdetermine d colourimetricallyutilizin gm-hydroxydipheny la sdescribe db yAhme d& Labavitch (1977).Correction sfo rcarbohydrat einterferenc ewer emad e accordingt oth emetho ddescribe db yKintne r& v Bure n (1982). Theconcentratio no fAG Ai nth efraction scollecte d fromge lpermeatio nan d ionexchang echromatograph ywa sdetermine db yth eautomate d m-hydroxydiphenylmetho d (Thibault1979) . Theamoun to fpecti nsolubilize dfro mWI Sa sa resul to fenzym etreatmen t wasdetermine db yth eautomate dm-hydroxydipheny lmetho dafte r saponificationfo r3 0min .i n0.0 5N NaOH .Th eD Eo fthi ssolubilize d pectinwa scalculate d fromth eamoun to fmethano la sdetermine do na SpectraphysicsS P800 0H.P.L.C .usin ga 30 0x 7. 8m mAmine xHP X8 7H colum n anda 5 0x 4. 6m mA G5 0W- X4 (Bio-Ra dLabs. ,Richmond ,USA )precolumn . Methanolwa sdetecte dwit ha nERM AER C751 0refractiv einde xdetecto rafte r elutionwit h0. 6ml/mi n0.0 1N H S O at30°C .Sample swer esaponifie dfo r1 houri n0.0 5N NaO Hprio rt oth emeasuremen t (Voragene tal .198 6a) .

Total neutral sugar content Thetota lamoun to fneutra lsugar si nth eeluate so fge lpermeatio n chromatographywa sdetermine dcolourimetricall yusin gpheno lsulphuri caci d asdescribe db yDuboi se tal .(1956) .

Cellulose content Thecellulos econten ti nAIS ,WI San dthei rfraction swa sassesse d accordingt oth emetho ddescribe db yUpdegraf f (1969).

Protein content Theprotei nconten twa sdetermine db yth esemi-automate d 80

micro-Kjeldahlmetho ddescribe db yRooze n& Ouwehan d (1978)usin ga conversionfacto ro f6.25 .Protei nsolubilize dfro mWI Sa sa resul to f enzymeincubatio nwa sdetermine daccordin gt oth emetho do fLowr y (1951).

Moisture content Oneg .o fsampl ewa stransferre d topredried ,weighe daluminiu m dishes.Th eweigh tlos swa sdetermine dafte r1 6hour sdryin ga t106°C .

Starch content Thestarc hconten twa sdetermine denzymaticall yusin gth eu.v.-metho d suppliedb yBoehringe r (Mannheim, FRG).

3.4. Properties of tomato particles

Particle size analysis Particlesiz edistributio nwa sdetermine do na Haave r& Boecke r laboratorysieve-shake rfitte dwit hsieve so f420 ,315 ,250 ,200 ,160 ,125 , 100an d5 6ur npor esize .20 0gra msample so fpast ewer edilute dwit h distilledwate rt o1 litr ean dtransferre dt oth esieves .Afte rshakin gfo r 45min .whil erinsin gwit h4-4. 5 litrewater ,th eretaine dmateria lwa s washedfro mth esieve swit hdistille dwater .Afte rfiltratio no fthes e fractionsthroug hdrie dan dpreweighe d filterpaper ,th efilte rpaper splu s solidswer edrie dfo r2 4hour sa t90° Can dweighed .Th eweigh to fth e fractionswa srecorde da sa % o fth etota lamoun to finsolubl esolid spresen t inth esample .

Microscopic examination of tomato cells Tomatojuices ,concentrate san denzym etreate dsample swer edilute dt o approximately0.5°B xan dexamine dusin ga nAxioma tZeis smicroscop e(phas e contrastillumination )equippe dwit ha camera .Photograph swer etake no n Agfapan2 5fil mexpose da s1 8DIN .Th efina lmagnificatio no nth efil m negativewa s51. 2times .

Degree of sedimentation 50gra mo fa 5°B xtomat ojuic eo rdilute dtomat oconcentrat ewa s weighed ina graduate d 100m lseru mbottl etogethe rwit h10 0m g 81

sodiumbenzoate .Afte r filling thevolum eu p to 100m lwit h distilled water theheigh t of the sedimented particle layerwa s recorded daily for one week.

Density gradient centrifugation 3.5 gram of 5°Bx tomato juice or diluted tomato concentrate was transferred toa centrifuge tube togetherwit h 35 gram Percoll (Pharmacia Fine Chemicals,Uppsala ,Sweden )o f density 1.05 (in0.2 5 M sucrose). Centrifuging the tubesfo r 25min . at40,00 0x g ina SorvallRC-5 B superspeed centrifuge at room temperature resulted in a self-generated density gradient.Afte r centrifugating, the tubeswer e photographed and layerswit h different density were isolated and examined microscopically.

3.5. Conductivity measurement The conductivity of dialysed tomato juice sampleswa smeasure d ina WTWmode lL F 39 conductivity/resistancemeter .

3.6. Serum sugars Serum sugarswer e determined on a Spectraphysics SP 8000H.P.L.C . using a 300x 7.8 mmAmine xHP X 87P column and amixe d bed ion exchange precolumn (AG 50W-X 4 (H+form),A G 3-X4A (0H~form),Bio-Ra d Labs., Richmond, Cal.,USA) .Sugar swer e detected with an ERMAER C 7510 refractive index detector after elutionwit h 0.5 ml/min. distilled water at 85°C (Voragen et al. 1986c).

3.7. Enzymes: purification and activity determination

Endo-polygalacturonase (poly (1,4-a-D-galacturonide) galacturonohydrolase, EC 3.2.1.15) Mould PGwa s isolated fromUltrazym e 20 (NovoFermen t A.G.,Basel , Switzerland) and purified as described by Vijayalakshmi et al. (1978). PG activity was assayed colourimetrically by determining the increase in reducing groups by theNelson-Somogy imetho d as described by Spiro (1966). Polygalacturonic acid (ICNPharmaceutical s Inc.Cleveland ,Ohio ,USA )wa s 82 useda sa substrate .A mixtur econsistin go f0. 2m lo f0. 2M sodiu msuccinat e bufferp H5.2 ,0. 1m lo fpolygalacturoni caci dsolutio n1 %(w/v )an d0. 1m l ofdistille dwater ,wa ssubjecte d toth eenzym eactio nan dincubate da t30° C fora specifie dperio do ftime .On euni to fP Gi sth eamoun to fenzym ewhic h catalysesth ereleas eo f1 ymo lo freducin gsuga rpe rminut eusin gglucos ea s standard.

Pectinesterase (pectin pectylhydrolase, EC 3.1.1.11) MouldP E (Gist-Brocades,Seclin ,France )wa spurifie dusin gth emetho d describedb yBaro ne tal . (1980).P Eactivit ywa sdetermine dtitrimetricall y accordingt oth emetho do fVa se tal .(1967 )b yestimatin gth efre ecarboxy l groupsforme di npecti na sa resul to fenzym eaction .Th eamoun to f0.0 1N NaOHrequire dt omaintai nth ep Ho fth esubstrat esolutio na t4. 0a t30 CCwa s measuredusin ga nautomati ctitrato r (Combititrato r3D ,Metroh mAG ,Herisau , Switzerland).Th eenzym esubstrat ewa sa 0.5 % (w/v)solutio no fa commercia l applepecti n (Obipectin,gree nribbon ,D E62.0% )i n0.1 5M NaCl .On euni to f PEi sth eamoun to fenzym ewhic hliberate s1 umo lo fcarboxylgroup spe r minute.

Endo-pectinlyase (poly(methoxygalacturonide) lyase, EC 4.2.2.10) Endo-PLwa sisolate dfro mUltrazym e2 0an dpurifie daccordin gt oth emetho d describedb yva nHoudenhove n (1975).P Lactivit ywa sdetermine db ymonitorin g theincreas ei nabsorptio na t23 5n mo fa solutio no fhig hmethoxy lpecti n (Voragen1972) .Absorptio na t23 5n mi sspecifi cfo rth edoubl ebond sforme d asa resul to fP Lcatalyse dpecti ncleavage .Th eenzym esubstrat econsistin g of0.2 5m l1 %(w/v )pecti nsolutio n (DE- 92.3%) ,1. 5m lMcllvain ebuffe rp H 6.0an d 1.15m ldistille dwater ,wa spipette d intoa 1 0m mquart zcuvett ean d placedi na wate rbat ha t30°C .Th ecuvett ewa stransferre d toa temperatur e controlledchambe ri nth espectrophotomete ran d0. 1m lenzym esolutio nwa s added.On euni to fP Lactivit yi sth eamoun to fenzym ewhic hcatalyse sth e formationo f1 ymo lo fdoubl ebond spe rminute .Fo rth ecalculatio no fth e amounto fdoubl ebonds ,a mola rabsorptio ncoefficien to f550 0M cm was used.

Cellulose complex Enzymesformin gpar to fth emulti-componen t cellulasesyste mwer eisolate d 83

fromMaxazy mCL ,a commercia lcellulas epreparatio nfro mTrichoderm avirid e origin (GistBrocades ,Delft ,Th eNetherlands) ,an dpurifie da sdescribe d byBeldma ne tal . (1985). Exo-glucanase (1,4-&-D-glucan cellobiohydrolase, EC 3.2.1.91) Theactivit ywa smeasure di na mixtur econsistin go f0. 5m l1 %Avice l cellulose (TypeS.F. ,Serva ,Heidelberg ,FRG )i n5 0m Msodiu macetat e bufferp H5. 0an d0. 5m lenzym esolution .Afte rincubatio nfo r2 0hour sa t 30°Cth emixtur ewa scentrifuged .Th eamoun to fnewl yforme dreducin g endgroupsi nth eclea rsupernatan twa smeasure dusin gth eNelson-Somogy i method. Endo-glucanase (1,4-8-D-glucan glucanohydrolase, EC 3.2.1.4) Theactivit ywa smeasure d ina mixtur econtainin g0.2 5m l1 % carboxymethylcellulose(Akucel lA F0305 ,Akzo ,Arnhem ,Th eNetherlands) , 0.15m l0. 1M sodiumacetat ebuffe rp H5. 0an d0. 1m lenzym esolution . Afterincubatio na t30° Cfo r1 hr .th ereactio nwa sstoppe dan dth e reducingsugar swer emeasure db yth eadditio no fth eNelson-Somogy i reagent.Sample swer ecentrifuge dbefor ereadin gth eabsorbanc ea t52 0nm . 8-glucosidase ($-D-glucoside glucohydrolase, EC 3.2.1.21) Theactivit ywa sassaye db yth eadditio no f0. 1m lenzym esolutio nt o0. 9 mlo f0.1 %p-nitrophenyl-6-D-glucopyranosid e (Koch-Light,Colnbrook , Bucks.,England )i n5 0m Msodiu macetat ebuffe rp H5.0 .Afte rincubatio n for1 hr .a t30° Cth ereactio nwa sstoppe db yth eadditio no f1 m l0. 5M glycinebufferp H9.0 ,containin g2 m MEDTA .Th econcentratio no f p-nitrophenolwa smeasure da t40 0nm ,usin ga nabsorptio ncoefficien to f 13,700M _1cm_1. Cellobiase (a &-glucosidase) Theactivit ywa sdetermine db yth eadditio no f0. 1m lenzym esolutio nt o

0.1m l0.1 %cellobios e (Difco,Detroit ,Michigan ,USA )i nH 20,0. 1m lH 20 and0. 2m l0. 1M sodiumacetat ebuffe rp H5.0 .Afte rincubatio nfo r2 0 hoursa t30° Cth ereducin gendgroup swer emeasure dwit hth eNelson-Somogy i assay.

Hemicellulases Hemicellulolyticenzyme swer eisolate d fromPectinas eno .29 ,a commercia l preparationfro mAspergillu snige rorigi n (Gist-Brocades,Delft ,Th e Netherlands),an dpurifie da sdescribe db yRombout se tal .(1986 , arabinanases)an dVorage ne tal .(198 6d ,galactanases) . 84

Arabinofuranosidase (a-L-arabinofuranoside arabinohydrolase, EC 3.2.1.55) Theactivit ywa smeasure db yth eadditio no f0.0 5m lo fenzym esolutio nt o 0.145m lo f0.1 %p-nitrophenyl-a-L-arabinofuranosid e(Sigma ,S tLouis ,USA ) and0.3 5m l0.0 5M sodiu macetat ebuffe rp H5.0 ,an dincubatio nfo r1 hou r at30°C .Th ereactio nwa sstoppe db yth eadditio no f1 m l0. 5M glycin e bufferp H9.0 ,containin g2 m MEDTA .Th econcentratio no fp-nitropheno lwa s measureda t40 0nm ,usin ga nextinctio ncoefficien to f13,70 0M cm . Endo-arabinanase (1,5-a-L-arabinan arabinohydrolase, EC 3.2.1.99) Theactivit ywa smeasure do nappl ejuic eultrafiltrate ,a slightl ybranche d a-l,5-L-arabinan(haze-arabinan ,Vorage ne tal . 1982).Th ereactio n mixturefo rth eendo-arabinanas eassa yconsiste do f0. 1m l0.5 % ultrafiltratesolutio ni nH„0 ,0.3 5m l0.0 5M sodiu macetat ebuffe rp H5. 0 and0.0 5m lenzym esolution .Afte rincubatio na t30° Cfo r1 hour ,th e reactionwa sstoppe dan dth ereducin gsugar swer emeasure db yth e additiono fth eNelson-Somogy ireagen tusin garabinos ea sstandard . Endo-galactanase (1,4-$-D-galactan galactohydrolase, EC 3.2.1.89) Theactivit ywa sdetermine do na g-l,4-D-galacta nisolate dfro mpotatoe s andi nadditio ntreate dwit harabinofuranosidas e inorde rt oremov e arabinosesid echain sfro mth egalacta nbackbon e (Voragene tal .198 6c) . Thereactio nmixtur econtaine d0. 1m lo f0.5 %potat ogalacta ni nH„0 ,0.3 5 ml0.0 5M sodiu macetat ebuffe rp H5. 0an d0.0 5m lenzym esolution .Afte r incubationa t30° Cfo r1 hour ,th ereactio nwa sstoppe dan dth ereducin g sugarswer emeasure db yth eadditio no fth eNelson-Somogy ireagent .Sample s werecentrifuge dbefor ereadin gabsorbanc ea t52 0nm ;galactos ewa suse da s standard. 85

WIS AS A FACTOR DETERMINING CROSS VISCOSITY

4.1. Introduction

Asurve yo frelevan tliteratur e (chapter2 )show sconflictin g opinionsregardin gth eexten tt owhic hsolubl ean dinsolubl esolid s contributet oth egros sviscosit y oftomat oproducts .I tis ,however , generallyaccepte dtha ta ho tbrea kproces syield sproduct shighe ri n grossviscosit y thana col dbrea kprocess .Therefore ,i na nattemp tt o clarifyth erol eo fsolubl ean dinsolubl esolids ,tomat ojuic ean d concentratesprepare daccordin gt odifferen tbrea kmethod swer eanalyse d forgros san dseru mviscosit yan dchemica lcomposition .Moreove rthi s chapterdescribe sth einfluenc eo fartificiall yinduce dchange si n insolublean dsolubl esolid so nth egros sviscosit yo ftomat oproducts .

4.2. Methods

4.2.1. Preparation of hot break and cold break tomato juice

Pilot plant - Hotbrea kjuice .1 0litre so fwate rwer eheate di na 5 0litr escal d kettle.4 8k gtomatoe s (varietiesH3 0an dH2826 )wer ecrushe di na Ross i & Catellino .5 1choppe ra ta rat eo f1. 3kg/min .an dadde dt oth escal d kettle.Th etemperatur eo fth ecrushe dtomatoe si nth escal dkettl ewa s maintaineda t93-98°C .Te nminute safte rth elas tadditio nt oth escal d kettleth eheate dcrushe dtomatoe swer efinishe dusin ga singl estag eRoss i & Catellipaddl efinishe r (1/100/RE)equippe dwit ha 0. 6m mscreen .Th e finishedtomat ojuic ewa shea tprocesse d for1 hou ra t100° C(PH Bjuice) . -Col dbrea kjuice .2 0k go fchoppe dtomatoe s (varietiesH3 0an dH2826 ) wereadde dt o5 litre so fcol dwate ri nth escal dkettle .Th echoppe d tomatoeswer eheate dt o80 CCi n8-1 0min .Finishin gan dhea tprocessin g wereth esam ea swit hth epreparatio no fho tbrea kjuice .(PC Bjuice) . 86

Laboratory - Hotbrea kjuice .Tomatoe swer equickl yheate di na Philip s201 0C commercialmicrowav eove n(typ eAA H050 ,4. 4kW ,245 0± 2 5MHz )unti la hearttemperatur eo f90° Cwa sreached .Heatin gt othi stemperatur ewa s enoughfo rcomplet eP Ean dP Ginactivation .Afte rcoolin gt oroo m temperaturea correctio nfo revaporatio nwa smad eb yaddin gdistille d water.Th etomatoe swer efinishe d througha Bauknech tKU2- 1tapere dscre w juiceextracto requippe dwit ha 0. 9m mscree nan dhea tprocesse dfo r1 hou r at100°C . (LHBjuice) . -Col dbrea kjuice .Tomatoe swer efinishe da troo mtemperatur eusin gth e Bauknechtjuic eextractor .1 5minute safte rth estar to ffinishin gth e juicewa sheate di nth emicrowav eove ni nth esam ewa ya swit hth e preparationo fho tbrea kjuice .Th etota ltim efo rfinishin gan dheatin g was3 0min .Afte rcoolin gt oroo mtemperatur ea correctio nwa smad efo r evaporatedwater .Th ejuic ewa shea tprocesse dfo r1 hou ra t100°C .(LCB. . juice).Anothe rcol dbrea ktomat ojuic ewa sprepare db yhea tprocessin g finishedtomatoe s 5hhour safte rth efinishin gprocedure .(LCB „juice) .

Factory - Hotbrea kjuice .Ho tbrea ktomat ojuic ewa sprepare d ina Ross i& Catell i 1000L recirculatin gchop/scalde r (Fig. 11).

rough pulp

tomato feed

k££\A;'\A7 ^ temp,senso r ^^-g- I±_ level - _ heat exchanger indicator o chopper J

3um temp U P P sensor

Fig.11 :Th eRoss i& Catell i1000 Lrecirculatin gchop/scalder . 87

Inthi ssyste mtomatoe sar ecrushe da ta rat eo f7 tons/hr .an dimmediatel y immersedi nhot ,previousl y crushedtomatoes .Heatin gt o103° Ci s accomplished inth etubula rheate rsectio no fth echop/scalder .Th eheate d crushedtomatoe swer efinishe dusin ga thre estag eRoss i& Catell ifinishe r equippedwit h1.0-0. 8an d0. 5m mscreens .Th ejuic ewa shea tprocesse dfo r 1hou ra t100°C .(FH Bjuice) . -col dbrea kjuice .Th emetho do fpreparin gcol dbrea kjuic ewa sidentica l totha to fho tbrea kjuic eexcep tfo rth edegre eo fth ebrea ktemperature . Inth ecas eo fth ecol dbrea kproces sth ebrea ktemperatur ewa s40° C insteado f103°C .(FC Bjuice) .

4.2.2. Concentration of tomato juice

Concentration in an industrial evaporator Tomatojuic ewa sconcentrate d ina Ross i& Catell iT. _D.F.F .Ante o evaporator.Thi sdoubl eeffec tevaporato ri scapabl eo fevaporatin g12,00 0 kgwater/hr .Th efres htomat ojuic ei sreceive dint oth esecon deffec to f theevaporato rwhic hconsist so fa nincline dban ko ftubes .Th erelativel y thintomat ojuic ecirculate seasil yb yconvectio ni nthi ssectio nwhic hi s heatedb yth econdensin gvapour sfro mth efirs teffect .Afte r preconcentrationi nth esecon deffec tth econcentrat ei stransferre db y gravitythroug ha nautomati cleve lregulato rint oth efirs teffec to fth e evaporator.Circulatio ni nthi seffec ti scarrie dou tb ya turbin ewhic h pumpsth econcentrat edow nth eheatin gtube s (D.F.F.= Downwar dForce d Flow).Produc ttemperature swer e60° Ci nth efirs teffec t (58-60c mH g vacuum)an d40° Ci nth esecon deffec t (69-70c mH gvacuum) .

Concentration in a laboratory evaporator ABuch irotar yfil mevaporato rwa suse dfo rth elaborator y concentrationo ftomat ojuice .Usuall yth eproduc ttemperatur edurin g concentrationwa s70° C(52-5 3c mH gvacuum )an dth ewate rbat htemperatur e 90°C,however ,occasionall yth eproduc twa sconcentrate da t20° C(7 4c mH g vacuum)usin ga wate rbat htemperatur eo f40°C .Th eevaporato rwa sequippe d witha scrape rt opreven tburnin go fth etomat oconcentrate . 88

WIS concentration 6 tomatojuic esample so f25 0gra mwer ecentrifuge d for4 5min .a t 25,400x g i na Sorval lRC-5 Brefrigerate dsuperspee dcentrifuge .Afte r centrifugating0 ,30 ,60 ,90 ,12 0an d15 0gram so fclea rseru mwer e removed.Afte rresuspendin gth eparticle si nth eremainin gserum , concentrateswer eobtaine dwit hth eorigina lseru mviscosit ybu twit h increasedWI Scontents .

4.2.3. Reduction of serum viscosity and exchange of serum

6x 25 0gram so fho tbrea ktomat ojuic eo fvarietie sH20 8an dH31 8 werecentrifuge dfo r3 0min .a t25,40 0x g i na Sorval lRC-5 Brefrigerate d superspeedcentrifuge .Th econten to ftw ocentrifug etube swa sresuspende d andcombine d (samplecod e c). Toth eseru mo ftw oothe rtubes ,4 4unit s purifiedP Gan d5 1unit spurifie dP Ewer eadded .Afte rincubatio nfo r2 4 hoursa t40° Cth eenzyme swer einactivate db yheatin gfo r2 0min .a t85°C . Theenzym etreate dseru mwa sresuspende dwit hth eresidu ei nth ecentrifug e tubesan dth econtent.o fth etw otube swa scombine d (samplecod ee) .Th e serumo fth ethir dse to fcentrifug etube swa scombine dan dexchange dwit h thecorrespondin gseru mfro mjuic eo fth eothe rvariety .Seru mo fon e varietywa sresuspende dwit hresidu eo fth eothe rvariet y (samplecod e s).

4.2.4. Dialysis of tomato juice

2k go ftomat ojuic ewa sdialysed ,i nbag so fcellulos eacetate ,fo r3 daysa t4° Cagains t6 volume so f2 5litre sdistille dwater .Th edialyse d tomatojuic ewa sconcentrate dan dth eviscosit y ofth econcentrate swa s comparedt onon-dialyse dconcentrates . 89

4.3. Results and Discussion

Pilot plant hot and cold break

In the 1980 season pilot plant hot and coldbrea k tomato juice was prepared from tomatovarietie s H30 andH282 6 as described under 4.2.1. The tomatoeswer e roundish,full y ripe,an d had amea nweigh t of 99 gram and 66 gram forvariet y H30 and H2826 respectively. Thejuic ewa s concentrated using theBiich ilaborator y rotary film evaporator. Figures 12an d 13 show gross and serumviscositie s of the juices and their concentrates as related to the soluble solids content (°Bx). It is clear that therewer e largevarieta l differences .A t a specified soluble solids level avariet y H2826 concentrate has almost double the grossviscosit y of aH3 0variet y concentrate (Fig.12) . Incontras t towha t isgenerall y assumed,n o difference in gross viscosity was found between hotbrea k and cold break juices and concentrates in spite of large differences in serumviscosit y (Fig.13).

log Tl app. log n<

1.5-

4.0-

1.0-

3.0- // ,

0.5

x H30 PHB x H30 PHB » H,„ PCB i H PCB 2.0- 30 ./ • H2, PHB • H2!26 PHB

o H,, PCB O H2B26 PCB

r °"ot T 1.0 1.5 10 20 log °Bx °Bx

Fig.12 :Relationshi pbetwee n °Bxan dgros s Fig.13 :Relationshi pbetwee n°B xan dseru m viscosity forho tbrea k juice,col d viscosity forho tbrea kjuice ,col d breakjuic ean dthei r concentrates breakjuic ean dthei r concentrates (varietiesH30 ,H2826) . (varietiesH30 ,H2826) . 90

Table16 :Solid sconten to ftomat ojuic eprepare d fromvarietie sH3 0an d H2826. sample °Bx TS AIS s-AIS WIS %w/ w %w/ w %w/ w %w/ w

H30PH B 5.8 6.38 1.03 0.25 0.89 H30PCB * 4.6 4.84 0.71 0.10 0.65 H2826PH B 5.9 6.75 1.52 0.40 1.26 H2826PCB * 4.0 4.60 0.96 0.19 0.76

*lo w°B xdu et oadditio no fwate rbefor ebrea k (see4.2.1.) .

Table1 6show stha tth emai neffec to fth ecol dbrea ki sa decreas ei n s-AISamount ,eve nwhe ncompare da tidentica l°Bx ;thi si si naccordanc e withth elo wseru mviscositie sfound .Th elo ws-AI Svalue sar ecause db y thedegradatio no flarg epart so fth ehig hmolecula rweigh tseru msolubl e pectinint olowe rmolecula rweigh tfragment sno tprecipatabl ewit halcoho l (Table 17).

Table17 :AG Aconten to fAIS ,s-AI San dWI Sprepare dfro mho tan dcol d breakjuic eo ftw ovarietie s (weight% o nbasi so ffres hweigh tx 100a t5°Bx) .

H30 H2826

PHB PCB PHB PCB

AIS 21.3 16.2 33.3 24.8 s-AIS 12.8 6.2 18.2 12.1 WIS 7.8 7.9 11.9 9.9

Tables1 6an d1 7als oindicat emino rchange soccurrin gt oth eWIS ;th eWI S pectini saffecte da si sillustrate db yth edecreas ei noveral ldegre eo f esterification (Table 18). 91

Table18 :Overal lD Eo fth epecti ni nAI San dWI Sfraction so fjuic efro m varietiesH3 0an dH2826 .

sample AIS s-AIS WIS

H30 PHB 55 65 35 H30 PCB 46 66 21 H2826 PHB 59 69 40 H2826 PCB 56 66 29

Table19 :Suga rcompositio no fAIS ,WI San ds-AI Sprepare d fromtomat ojuic e

oftw ovarietie s (Mol%).

AIS WIS s-AIS

PHB PCB PHB PCB PHB PCB

H30

Rhamnose 0.7 0.7 0.3 0.3 1.1 1.0

Arablnose 2.7 2.9 4.9 4.5 1.6 3.1

Xylose 7.3 8.3 6.8 6.7 1.2 1.5

Mannose 4.5 5.3 11.1 9.3 1.0 0.6

Galactose 4.3 6.3 5.1 6.2 3.1 7.5

Glucose 33.0 36.1 47.2 45.0 6.7 5.8

AGA 47.6 40.5 24.5 28.1 85.5 80.7

H2826

Rhamnose 0.5 0.6 0.3 0.3 0.6 0.6

Arablnose 5.0 5.7 5.5 5.8 5.5 7.6

Xylose 7.2 7.8 8.5 8.0 1.4 1.1

Mannose 4.3 5.3 8.1 9.6 1.2 0.8

Galactose 5.7 8.2 6.3 7.7 5.7 12.7

Glucose 31.7 34.8 45.7 46.7 7.5 5.2

AGA 45.7 37.7 25.6 21.8 78.3 72.2 92

Although large parts ofth eseru mpecti n aredegrade d during cold breaka s a result ofpectolyti c enzyme action,th eoveral lD Eo fth eremainin g high molecular weight serum pectin isalmos t unchanged.A nexplanatio n forthi s observation canprobabl y befoun d inTabl e 19,whic h shows thesuga r composition ofth eho tan dcol dbrea k samples.Th emola r amountso f arabinose andgalactos e inth es-AI So fcol dbrea k juicehav e increased greatlywhe n compared toho tbrea k juice.Thi sma y indicate that during cold break,pectolyti c enzymes initially attack aneasil y accessible soluble pectin,e.g .a pectinwithou t side chains.Afte r saponification and degradation of thispectin ,th eremainin g serumpecti n isstil lhighl y esterified,probabl y because thispecti n isprotecte d against enzymatic attackb y side chains ofarabinos e andgalactose . 109 Happ. U.0-\

/ 3.0 s /•

x H30 PHB / » H PCB A •*' 30 .0* • H2826 PHB 2.0- / ° H„„ PCB

Logt\a W=2.045 log %WI S •2784 y r= 0.99 8 4 i I 1 1 1 1 -0.2 0.5 log % WIS Fig. 1A: Relationship between %WIS and gross viscosity for hot break juice, cold break juice and their concentrates (varieties H30, H2826). The data show that degradation of serum pectin and the resulting decrease in serum viscosity does not necessarily affect gross viscosity. In the case of the pilot plant cold break, the samples were heated to a temperature of 80°C within 8-10 min. after crushing. Therefore pectolytic enzyme action was probably restricted mainly to the easily accessible pectin in the serum while the insoluble solids were more or less unaffected. Altogether, the 93 resultspoin tt oth eimportanc eo fth ewate rinsolubl esolid si n determininggros sviscosity .Moreover ,thi si silustrate d inFig .14 ,whic h showsgros sviscosit y forho tan dcol dbrea ksample so fbot hvarietie sa sa functiono fth eWI Scontent .Obviousl yth evarieta ldifferenc ei ngros s viscositya sobserve di nFig .1 2i sonl ycause db ya differenc ei nWI S content (Table16) .

Laboratory hot and cold break

ClassI glasshous etomatoe so fth evariet ySonatin ewer eobtaine dfro m theSprenge rInstitute ,Wageningen ,th eNetherland so n11/6/198 1an d immediatelystore da t3° Ci nth edark .Th etomatoe swer efull yrip ean dha d amea nweigh to f6 6g/tomato .Ho tan dcol dbrea kjuic esample swer e preparedbot hafte r4 day san d3 week sstorag ea t3°C .Storag ecause da 5 % decreasei nAI Sconten to fth etomatoes ;however ,th eAI Scompositio ndi d notchange .

Table20 :Solid sconten to fjuic eprepare d fromtomat ovariet ySonatin e after4 day san d3 week sstorag ea t3°C .

Sample °Bx TS AIS s-AIS WIS %w/ w %w/ w %w/ w %w/ w

LCB 4day s 5.0 5.71 0.95 0.32 0.80 LCB 4day s 5.2 5.90 0.96 0.21 0.80 LCB.3 week s 5.0 5.58 0.91 0.19 0.72 LHB 3week s 5.1 5.88 1.09 0.35 0.74

Duringth epreparatio no fpilo tplan tcol dbrea kjuice ,pectolyti cenzyme s actedmaximall y 10min .o nth ecoars ecrushe dtomatoe sbefor ethe ywer e heatinactivated .Durin gth elaborator ypreparatio no fcol dbrea kjuice , however,th eenzyme swer eactiv eo nth efinishe dproduc t (0.9mm )fo ra t least1 5min .an dfo ra slon ga s5* 5hour sbefor ethe ywer ehea t inactivated.Eve nth elon genzym eactio ndi dno tresul ti ndetectabl ewate r 94 insoluble solidsbreakdow n (Table 20),possibl y because the incubation temperature of ± 20°Cwa s unfavourable forhig h enzyme activity.A swit h thepilo t plant samples,th e laboratory hot and coldbrea k samples did not differ in grossviscosit y in spite ofver y large differences in serum viscosity (Fig. 15an d 16). logtUeru m

1.5-

logTic

1.0-

3.0- /* /* / 0.5- ./ 2.0- y x LHB • LCB,

o LCB2

*--r IF 1.5 10 lb log°Bx °Bx Fig. 15: Relationship between °Bxan d Fig. 16: Relationship between °Bx and grossviscosit y forho t break serum viscosity for hot break juice,col d break juicean d juice, cold break juice and their concentrates their concentrates (variety Sonatine). (variety Sonatine). Table 21: AGA content of AIS, s-AIS and WIS prepared from tomato juice of variety Sonatine (weight % on basis of fresh weight x 100 at 5°Bx). In parentheses the degree of esterification of the AGA

LHB LCB LCB2

AIS 24.2(65 ) 17.6(35 ) 14.9(14 ) s-AIS 20.3(69 ) 9.6 (45) - WIS 4.9 (37) 6.4 (6) — 95

Pectindegradatio nb yenzym eactio nwa sagai nrestricte d toth esolubl e serumpectin .However ,i ncas eo fth elaborator ycol dbrea kjuice ,th e enzymesacte dlon genoug ht opartl ysaponif yth eremainin gundegrade d serumpecti n(Tabl e21 )an dt odegrad ehig hmolecula rweigh tpolymer s containing50 %o fth eseru marabinos ean d30 %o fth eseru mgalactos et o alcoholsolubl efragment s (datano tshown) .Th elonge ractio no fth e pectolyticenzyme si sals oindicate db yth ever ylo woveral ldegre eo f esterificationo fth eWIS-pectin .Fro mth edat ai nTabl e2 1i tappear stha t apar to fth esaponified ,undegrade dseru mpecti nha sprecipitate d inth e WIS. Leonie tal . (1979,1980 ,1981 )als oreporte dprecipitatio no f saponified serumpecti na sa resul to fa col dbrea ktreatment .

Factory hot and cold break

Inth e198 1season ,ho tan dcol dbrea kjuice swer eprepare do nfactor y scalefro mth etomat ovariet yH1361 .Th etomatoe swer emature ,re dcoloure d andha da mea nweigh to f12 2g/tomato .Th eRoss i& Catell iT3 0evaporato r wasuse dt oconcentrat eth ejuic et osevera lsolid slevels .

log "n app. logn app

./ Jo 3.0- 3.0- y

x Hi36i FHB

2.0- 2.0- • H1361 FCB

0 H136, LHB » Hi36i LCBi

r^-V -| 1 1 1 1 1 1 1 1— 0.5 1.0 1.5 -0.3 0 0.5 log log %WIS

Fig. 17:Relationshi p between °Bxan dgros s Fig. 18:Relationshi p between%WI San d viscosity forho tbrea kjuice ,col d grossviscosit y forho tbrea k break juicean dthei r concentrates juice,col dbrea kjuic ean d (variety H1361). theirconcentrate s (variety H1361). 96

Laboratoryho tan dcol dbrea kjuice swer eprepare d fromth esam evariet ya s areference .Thes ejuice swer econcentrate dwit hth erotar yfil m evaporator. Fig.1 7show sth egros sviscosit yo fth ejuic ean dconcentrate sa s relatedt oth esolubl esolid s (°Bx).N odifferenc ei ngros sviscosit yca n bedetecte dbetwee nth etw oho tbrea ksamples ;however ,th ecol dbrea k samplessho wa distinctl ylowe rgros sviscosity .Th egros sviscosit yo fth e factorycol dbrea ksampl ei sapproximatel y60 %o fth efactor yho tbrea k sample.Thi sdifferenc ei sobviousl ycause db yth elowe ramoun to fWI Si n thecol dbrea kjuic e (Table 22).Th eenzym eactio ndurin gth efactor ycol d breakresulte di na complet edegradatio no fseru mpecti nan di n saponificationo fWI Spectin .However ,th eresult sd ono tgiv ea decisiv e answert oth equestio nwhethe rth eWI Swa sals odegrade db yth eenzymes .I t iscertai ntha tth ecompositio no fho tbrea kjuic eWI Sdiffer sfro mcol d breakjuic eWIS ,bu ti tcanno tb eexclude dtha tthi sdifferenc ean dth e differencei nWI Samoun tcoul db ea resul to fprocessin gcondition s(e.g . finishingoperation ,se e7.3.2. )othe rtha nbrea ktemperature .

Table22 :Solid scontent ,seru mviscosit yan dcompositio no ffactor yan d laboratorytomat ojuice so fvariet yH1361 . sample °Bx %TS %WIS Compos itiono f WIS* serum at20 ' Bx AGA DE Protein Cellulose

FHB 4.6 5.34 1.05 71 14.5 23 21.6 40.7 FCB 4.4 5.02 0.84 2 10.2 0 17.6 36.9 LHB 4.4 5.20 1.07 188 - - - -

LCB1 4.3 5.07 0.90 4 - - - -

* expresseda sweigh t% o nbasi so ffres hweigh tx 100a t5°Bx .

Figure1 8show stha tal lth eho tan dcol dbrea ksample sprepare dfro m varietyH136 1hav eequa lgros sviscositie sa ta specifi cWI Samoun ti n spiteo fth edifference si nWI Samoun tan dWI Scomposition . 97

Figure1 9show smicroscopi cpicture so findustriall yprepare dGree k hotan dcol dbrea kpaste so funknow nvariety .Fro mthes epicture si ti s cleartha tth eWI Si nthi scol dbrea kpast eha sbee ndegraded .

Fig. 19:Tomat ocell sa spresen ti nho tan dcol dbrea kpaste . a:ho tbrea kpaste ,b :col dbrea kpast e (magnification 61.4x).

Particlesiz eanalysi s (Fig.20 )als oclearl yshow sth edegradatio no fth e tomatocells .

% of WIS 100-

hot break

• cold break 50-

pTl \TT, >420 420-315'315-250 250-200 200- 160' 160 -125'12 5— 10 0 100 - 56 <56 particle size range (^m )

Fig. 20:Differenc ei nsiz eo fparticle s presenti nhot breakan d cold breakpaste . 98

Table23 :Solid sconten tan dviscositie so fdilute dho tan dcol dbrea k Greektomat opaste .

"Bx %TS %WIS %s-AIS %AIS n n app serum hotbrea k 9.1 10.55 1.57 0.38 1.87 230 6.4 coldbrea k 9.2 10.15 1.30 0.20 1.40 136 1.3

Thiscol dbrea kpast econtain s18 %les sWI Stha nth eho tbrea kpast e(Tabl e 23). Serumpecti nha sbee nalmos tcompletel ydegrade da si sindicate db y thelo wseru mviscosity .Analysi so fth eneutra lsuga rcompositio no fth e WISconfirm sth edegradatio no fth eWI Specti n(Tabl e 24). 68%o fth eWI Specti nha sbee nsolubilize d togetherwit h40 %o fth eWI S arabinosean d30 %o fth eWI Sgalactose .Whil eth eWI Specti ni sno tonl y solubilizedbu tals odegraded ,th esolubilize darabinos ean dgalactos e remaina shig hmolecula rweigh tpolymer san dprecipitat ei nth es-AIS . This,togethe rwit hth edegradatio no f72 %o fth eseru mpectin ,cause s thelarg eincreas ei nmola ramount so farabinos ean dgalactos ei nth es-AI S asa resul to fth ecol dbreak .

Table24 :Suga rcompositio n (Mol%)o fGree kho tan dcol dbrea ktomat o paste.

WIS s-AIS

hotbrea k coldbrea k hotbrea k coldbrea k

Rhamnose 0.2 - 0.4 3.3 Arabinose 5.5 4.4 3.1 14.7 Xylose 13.0 16.4 0.9 4.6 Mannose 6.3 8.4 0.6 0.1 Galactose 7.3 6.9 3.8 20.7 Glucose 39.0 52.0 5.0 6.0 AGA 28.7 12.1 86.2 50.5 99

Thepecti n inWI S and serumha sbee n degraded to such an extent that determination of theAG A contentwit h the copper precipitation method was no longer possible,therefor e the reported AGA contentswer e determined with them-hydroxydipheny lmetho d as described by Ahmed & Labavitch (1977). Thedegradatio n of theWI S not only resulted in a lower grossviscosit y at a specific °Bxvalu ebu t also in a lower grossviscosit y at a specific WIS amount. TheWI S of this cold break sample isapparentl y of poorer quality than that of theho tbrea k sample. (Fig. 21 and22) . log'h.Qpp 3.0-1 3.0- x hot break o cold break // // 2.0- 2.0- / x hot break o cold break

0.5 1.0 u 0.5 log°Bx log%WI S Fig.21 :Relationshi pbetwee n Fig.22 :Relationshi pbetwee n °Bxan dgros svisco ­ %WISan dgros svisco ­ sityfo rho tbrea kpast e sityfo rho tbrea kpast e andcol dbrea kpast e andcol dbrea kpast e producedi nGreece . producedi nGreece . WIS concentrates

The use of theWI S concentration technique in studying the influence of soluble and insoluble solids ongros sviscosit y offers the opportunity toprepar e sampleswit h identical serumviscosit y and amount of soluble solidsbu twit h increasing amounts ofWI S ofunifor m quality. Comparison of grossviscosit y of such concentrates with normal rotary film evaporator concentrateswil l show the influence ofWI S content and serumviscosit y on gross viscosity. From the results,presente d graphically inFig . 23,i t canb e concluded that no difference exists in grossviscosit y ofWI S or evaporator concentrates in spite of a large difference in serumviscosity . Serum viscosity againwa s not found topla y a detectable role in the determination of gross viscosity. 100

log tupp. • WISconcentrat e 3.0- x laboratory concentrate /• - x/

/

/ nser™at2.4%WIS Z0 1 • 2mPa.s x 9mPa.s

oU 1j ,1 I I I I I i I -03 0 0.5 log % WIS Fig.23 :Relationshi pbetwee n%WI San d grossviscosit y forconcentrate s preparedaccordin gt otw omethods .

Degradation of serum pectin and exchange of serum

Table25 :Effec to fa chang ei nseru mviscosit yo ngros sviscosit yo ftomat ojuic ean d

concentrate.

tomato juice concentrate variety code °Bx WIS n n Bostwick °Bx WIS n n Bostwick serum app serum app

208 n.c 4.6 1.22 239.3 9.8 9.0 2.45 885.2 2.4 c 4.6 1.15 2.9 203.2 10.9 9.8 2.79 9.6 1040.6 1.6 3 e 4.6 1.17 1.3 216.4 9.5 11.0 3.35 2.2 1278.5 <1 s 4.8 1.20 224.8 9.8 10.7 2.69 - 1022.7 1.6 318 n.c 4.9 1.02 174.9 11.9 10.4 2.11 - 678.2 3.5 c 5.0 0.99 2.7 161.7 12.2 10.6 2.19 8.3 721.1 3.7 e 4.9 1.00 1.2 175.7 11.4 10.8 2.24 2.0 714.6 3.5 s 4.6 0.99 165.1 12.1 10.6 2.35 10.9 820.6 3.1

1)code :n. c= no tcentrifuged ,c ~ centrifuged ,e • enzym etreated ,s • seru mexchang e

2)n at10°Bx :20 8• =10. 1mPa.s ,31 8- 7. 4mPa. s serum 3)to olo wa sresul to fseparatio ndurin gHaak emeasurement . 101

Table2 5show sth eresult so fa nexperimen ti nwhic hth eseru mpecti n isdegrade db yth ecombine dactio no fpurifie dP Ean dP Gwithou tan ychang e inWI So rothe rsolubl esolids .Anothe rpossibilit yo fchangin gth eseru m viscosity ina tomat ojuic ei st oexchang eit sseru mfo rth eseru mo f anothertomat ojuice ,result so fthi sexperimen tar eals oinclude di nTabl e 25.A logaritmi crelationshi pwa sagai nfoun dbetwee nth egros sviscosit y (bothHaak ean dBostwic kviscosity )an dth eWI Sconten to fth esample , independentlyo fth eseru mviscosity .Thes eresult sca nlea dt on oothe r conclusiontha ntha tth egros sviscosit yo fa tomat ojuic eo rconcentrat e ismainl ydetermine db yit sWI Scontent .

Dialysis of tomato juice

Thedialysi so ftomat ojuic eresult si nth eremova lo flo wmolecula r weightsugar ssuc ha sglucose ,fructos ean dsaccharose ,an di nth eremova l ofelectrolyte s (Tables2 6an d 27).Dialysi sdoe sno tchang eth eWI San d serumpecti nconten to fa juice .Whittenberge r& Nuttin g (1958)an dFod a& McCollum (1970)assume dtha tth eelectrolyte shav ea negativ einfluenc eo n grossviscosity . 63%o fth emeasure d °Bxvalu ei sdetermine db yth eneutra lsugar sglucose , fructosean dsaccharos e (Table26) .

Table26 :Influenc eo fdialysi so ncompositio no ftomat ojuice .

°Bx %TS %WIS serumsugar s mg/ml.

glucose fructose saccharose notdialyse d 5.5 6.25 0.94 15.92 17.32 1.26 dialysed 1.0 2.17 0.94 2.92 3.02 0.37

82%o fthes esugar shav ebee nremove ddurin gdialysis ,correspondin g to68 % ofth edecreas ei ntota lsolids .Thi smean stha t1. 3gram/10 0g juic eo f othersubstance shav ebee nremove ddurin gdialysis .Tabl e2 7show sth e effecto fthi sremova lo np Han dconductivity . 102

Table27 :Effec to fdialysi so np Han dconductivity .

'Bx %TS %WIS pH conductivity (umho)

notdialyse d 5.6 6.29 0.94 4.33 7.49x 10 3 dialysed+ suga radde d 5.1 5.13 0.92 4.05 7.41x 10 2

Theinfluenc eo fdialysi so ngros sviscosity ,a smeasure dwit hth eBostwic k consistometer,an do nseru mviscosit yi sshow ni nTabl e28 .Gros sviscosit y ofdialyse d tomatojuic emeasure dwit hth eHaak erotovisc odi dno tdiffe r fromgros sviscosit yo fjuic ewhic hwa sno tdialysed .

Table28 :Influenc eo fdialysi so ngros sviscosit yan dseru mviscosit y

%WIS nonotdialyse tdialyse dd dialysed

Bostwick n Bostwick serum serum

1.0 12.5 2.6 13.4 2.5 1.5 6.5 4.4 6.9 3.8 2.0 3.4 7.7 3.5 5.7 2.5 1.7 13.2 1.8 8.7

However,du et oseparatio ni nth emeasurin gga pth eus eo fth eHaak e rotoviscofo rgros sviscosit ymeasurement so nconcentrate so fdialyse d juicewa sno tpossible .Tabl e2 8doe sno tindicat ea ninfluenc eo fth e dialysed sugarsan delectrolyte so ngros sviscosity .Th eremova lo fthes e componentsdecrease d serumviscosity ,bu tagai nth edecreas ei nseru m viscositydi dno taffec tgros sviscosity .

4.4. Conclusions

Figures2 4t o2 6presen ta summar yo fgros san dseru mviscosit ydat a determineddurin gth eseason s1980-198 4o ntomat ojuic ean dconcentrate s 103

V • . * o a) * •« • 4-i M • ••• % > 4 *:

\

»X o *§W» •« *~ e*

QJ V) O m\ ii a ql i4 • e 5

U CD

~*J|»

0) o jo o M •H > J3 104 fromvarietie s H30,H136 1 and H2826.Apar t from a difference invariety , the analysed samples also differed inmethod s of production and concentration.Figure s 24 and 25 show that in spite of all these differences there isa strong relationship between Haake orBostwic k gross viscosity and WIS contents.Figur e 26, on the contrary,show s aver y large spread in serumviscosit y data at theseWI S contents. The results of the experiments discussed in this chapter prove that thegros sviscosit y of tomatojuice s and concentrates is determined primarily by theirWI S content. Serumviscosity ,whic h is strongly influenced by break temperature,ha sn o effect on grossviscosity . Variety and processing method influence grossviscosit y at a specific °Bxb y affecting the amount ofWIS .A swil lb e discussed in chapter 7,th e processing method canals o affect the quality of theWIS . 105

COMPOSITION AND STRUCTURAL FEATURES OF TOMATO WIS

5.1. Introduction

Theelucidatio no fth echemica lstructur eo fplan tcel lwall sha sbee n thesubjec to fman ystudies .A tthi stime ,however ,littl eca nb esai d aboutth esecondar yan dtertiar ystructur eo fth ecel lwal lpolymer s (McNeile tal .1984 ,De y& Brinso n 1984).Th estructur eo fth ecel lwal l isobviousl ymuc hmor ecomple xtha nKeegstr ae tal .(1973) ,Robinso n (1977) orLampor t& Epstei n (1983)hav eassume di nthei rmodels . Fromth eavailabl eliteratur e (section2.3.2. )i tca nb econclude d thatth efirmnes so ftomat ofrui ti sstrongl yinfluence db yth ecel lwal l polysaccharides (WIS)an dthei rchemica lan dbiochemica lchange sdurin g ripening.Result spresente di nth eprecedin gchapte rstres sth eimportanc e ofth eWI Si ndeterminin gth erheologica lpropertie so ftomat oproducts . Theai mo fth eexperiment sdescribe d inthi schapte rwa st oobtai n moreknowledg eo nbot hth ecompositio no fth etomat ocel lwal lan dth e structuralfeature so ftomat ocel lwal lpolysaccharides .WI Sisolate dfro m freshtomatoe san dtomat oproduct swa sfractionate d eitherb yus eo f solventslik eammoniu moxalate ,HC 1an dNaOH ,o rb yus eo fpurifie d enzymes.Th efraction sobtaine dwer efurthe rcharacterize db yge l permeationan dio nexchang echromatography ,an db yanalysi so fneutra l sugaran dglycosidi clinkag ecomposition .

5.2. Methods

5.2.1. Fractionation of tomato WIS

TheWI So ffres htomatoes ,tomat ojuic ean dtomat opast ewa s fractionated intopectin ,hemicellulos ean dcellulos efraction sa sshow ni n Fig. 27a.I nsom einstance sth epecti csubstance swer efractionate d accordingt othei rsolubilit yi n0.2 %ammoniu moxalate ,0.0 5N HC 1an d0.0 5 NNaO H+ 5 m MEDT A (Fig. 27b). 106

27b [WIS | I 0.2% ammoniumoxalate j 1 hr 60 °C centrifuged

Ipelle t I oxalate - soluble-pectin 005 N HCI , 1 hr 60°C dialysed HCI-soluble 27 a centrifuged - and [wis] freeze- pectin pellet dried 0.05 M NaOH NaOH soluble 5 mM EDTA 005 N NaOH pectin 1 hr 4°C 5 mM EDTA 1 hr i.°C centrifuged •- -«centrifuged- pellet dialysed and 16%NaO H freeze dried 10mg/m l NaBH4 ,, 8h r room temp. total pectin centrifuged hemicellulose extract |pelle t| acidified washed pH 5.2

dialysed and dialysed and freeze dried freeze dried i Icellulos e I !hemicellulos eI

Fig.27 :Extractio no fpolysaccarid efraction so ftomat oWIS .

Each fractionationwa s carried outb y extracting three times,unde r continuous stirring,wit h the extractant (100ml/ gWIS )an d twicewit h distilled water.Afte r each extraction,th emixtur ewa s centrifuged in a SorvallRC-5 B refrigerated superspeed centrifuge for 20min . at 25,400x g (4CC). The combined supernatantswer e dialysed against distilled water for 5 days and freeze-dried. The neutral sugar composition and uronide content of the fractionswer e determined.

5.2.2. Characterization of pectin fractions

10m g quantities of ammonium oxalate soluble andHC I soluble pectin were dissolved in0.0 5 M sodium acetate buffer pH 6.0. In the case of ammonium oxalate soluble pectin a small amount of remaining insoluble materialwa s removedb y centrifugation.Th e clear supernatantwa s applied to a 60x 1.2 cm column of Sephacryl S-500 (Pharmacia Fine Chemicals, Uppsala,Sweden ; fractionation range 40,000 - 2x 10 measured on dextran) 107

andelute dwit h0.0 5M sodiu macetat ebuffe rp H6. 0a ta flo wrat eo f1 5 ml/hr.an da tambien ttemperature .Fraction so f2 m lwer ecollecte dan d assayedfo ruronide san dtota lneutra lsuga rcontent .Th emolecula rweigh t offragment sseparate dusin gge lpermeatio nchromatograph ywa sestimated , usingdextran sa sa referenc ean dreporte da sapparen tmolecula rweight .O n oneoccasion ,fraction sobtaine db ychromatograph yo fHC 1solubl epecti n wereals oanalyse d forneutra lsuga rcomposition . Fractionscontainin gth elowe rmolecula rweigh tfragment swer epooled , concentratedan dagai napplie dt oa Sephacry lS-50 0column .Thi stime , however,th ecolum nwa selute dwit hdistille dwater .Th eseparatio n achievedupo nelutio nwit hwate ri sno tbase dsolel yo nsiz eexclusion ; uronidecontainin gfraction sar eexclude d fromth ecolum nmatrix , presumablydu et ocharg eeffects ,an dtherefor eelut ei nth evoi do fth e column.Th eammoniu moxalat esolubl epecti nwa sals ofractionate do na 6 0x 1.2c mcolum no fSephacry lS-20 0(fractionatio nrang e1000-80,00 0measure d ondextran) .Eluent ,flo wrate ,fractio nsiz ean danalyse so fth efraction s wereth esam ea sfo rfractionatio no fthi specti no nSephacry lS-500 .

5.2.3. Characterization of the hemicellulose fraction

Ionexchang echromatography :15 0m gquantitie so fhemicellulos ewer e dissolved in2 0m ldistille dwate rb ystirrin gfo r2 4hr .a t20°C .Th e insolublemateria lwa sremove db ycentrifugation .Th eclea rsupernatan twa s appliedt oa 15. 0x 2. 2c mcolum no fDEAE-cellulos ei nacetat efor m (WhatmanD E52 )equilibrate dwit hdistille dwater .1 0m lfraction swer e elutedfro mth ecolum n (100ml/hr. )wit ha stepwis egradien to f successively 120m ldistille dwater ;10 0m lportion so fpotassiu macetat e bufferp H7.0 ,respectivel y0.0 5M ,0. 1M ,an d0. 5M ;10 0m lsodiu m hydroxidesolutio n0. 1M ,8 0m lsodiu mhydroxid e0. 2M ,11 0m lsodiu m hydroxide0. 5M an dfinall ywit h29 0m lsodiu mhydroxid e 1M t owas hth e column.Th eexperiment swer eperforme da tambien ttemperature .Suga r containingfraction sformin ga pea kwer epooled ,dialyse dan dfreeze-dried . Fromth etota lsub-fractio nth esuga rcompositio nan dglycosidi clinkag e compositionwer eestimated .Glycosidi clinkag ecompositio nwa sestablishe d bypermethylatio naccordin gt oth emetho ddescribe db yHakomor i(1964 ) 108 followed by 2N trifluoroacetic acid hydrolysis as described by Talmadge et al. (1973).Th e sugar derivativeswer e identified and determined by capillary-glc and by combined glc-ms. For capillary-glc a 10 m x 0.32 mm,wal l coated OV-225 column in a Carlo-ErbaFractova p 4160 gas chromatographwa s used. Sampleswer e injected cold-on-column, the oven temperature washel d for 1min . at 150°C,the n raised to 200°Cwit h a rate of 2cC/min.an d held at 200°C for 2min .Peak swer e recorded and integrated with aLD C 301 computing integrator. Glc-mswa s performed on aV.G.-M M 7070 F mass spectrometer coupled to aPy e 204 gas chromatograph equipped with a 1.5 m column packedwit h OV-225 (3%)o n Chromosorb WHP. To enable differentiation between 1,2 and 1,4 linked xylose residues, reduction to alditolswa s also carried outwit h sodiumborodeuteride .

Gel permeation chromatography: 5-10 mg ofhemicellulos e in0. 5 ml distilled water (clarified by centrifugation)wa s applied to a 60.0x 1.2 cm column of Sephacryl S-500an d elutedwit h distilled water at a flow rate of 10ml/hr . Fractions of 2m lwer e collected and assayed foruronide s and neutral sugars.Suga r containing fractionswer e analysed for sugar composition. To 0.5 ml of a clearhemicellulos e solution (5-10m g in 0.01 M sodium acetatebuffe r pH 5)wa s added 0.3 units of a purified 1,4-g-D-glucan-glucanohydrolase (E.C.3.2.1.4 )isolate d from a technical cellulase preparation (Maxazym CL 2000,Gist-Brocades ,Delft ,Th e Netherlands).Th e reactionmixtur ewa s incubated for2 0hrs .an d fractionated by gel permeation chromatography on Sephacryl S-500. The included fractionwhic h contained themajo r part of theneutra l sugarswa s further chromatographed over a 100x 2.6 cm column ofBio-Ge l P-2 (Bio-Rad, Richmond, Cal.,USA ; fractionation range 100-1800measure d on globular biomolecules)thermostate d at 50°C and eluted with distilled water at a flow rate of 25ml/hr .Fraction s of 2.5 mlwer e collected and assayed for totalneutra l sugars.Th e sugar composition of pooled fractions was established. 109

5.2.4. Enzymatic fragmentation of WIS and characterization of the fragments solubilized

Enzyme incubations:50-15 0m g quantities ofWI Swer e suspended in 5-10 ml 0.05 M sodium acetatebuffe r pH 5.0. Enzymeswer e added (Table 29)an d themixtur ewa s incubated for 48 hours at 37°C.Afte r the incubation, the solubilized fragmentswer e separated from the residueb y means of centrifugation and filtration.Th e residueswer ewashe d twicewit h distilled water;washin gwate r and reaction liquidwer e combined and made up to 10o r 25m lwit h distilled water.Th e solubilized fragments were assayed for neutral sugar composition,uronid e and protein content. The residueswer e resuspended indistille d water and freeze-dried. Some of these residueswer e treated furtherwit h other types of enzymes.

Gelpermeatio n chromatography: Reaction liquids from the control (WIS incubated with 0.05 M. sodium acetatebuffe r pH 5.0) and forPL ,PG ,an d galactanase treated WIS,wer e applied to a 60x 1.2 cm column of Sephacryl S-200an d elutedwit h 0.05 M sodium acetatebuffe r pH 6.0. Fractions of 2 mlwer e collected and assayed for totalneutra l sugar and uronide content. Fractions forming apea kwer e pooled and analysed forneutra l sugar composition.Dependin g on the fractionation pattern thepool swer e further fractionated ona 60x 1.2 cm column of SephacrylS-20 0 (water as eluent), SephacrylS-50 0 (0.05M sodium acetatebuffe r pH 6.0 aseluent ) or on a 90 x 1.0 cm column of Bio-GelP- 2 (water as eluent). 1m l fractionswer e collected in the case of fractionation over Bio-GelP-2 ; in the other cases 2 ml fractionswer e collected. The fractionswer e analysed for total neutral sugarsan d uronide content;poole d peakswer e assayed for neutral sugar composition.

The reaction liquid obtained by incubation ofWI Swit h endo-glucanase + exo-glucanasewa s fractionated on Sephacryl S-500.Th e included peak of this separation was further fractionated onBio-Ge lP-2 .Al l the columns were elutedwit hwater ;th e fractionswer e analysed for totalneutra l sugar and uronide content,th e pooled peaks forneutra l sugar composition. Some of the reaction liquids obtained after the incubation of residues of enzyme treated WISwit h new enzymes,wer e also fractionated on Sephacryl S-200.Th e fractions and peakswer e analysed asmentione d above. 110

CD LQ CD CD P3 p> 3 CB 1 X I-1 X b" + + 1+ 0 0 1 1 pj 0& 1 *0 13 H i- in H- S 3 p> 1 ti 3 CD rt 3 (D 0 K rot ici H r O N n rt Rl PI p- C to 0 CD *• fU & 3 ID 3 H fl P- 0> 0 Po) 1CD i« 3 PI rt CO pl 3 (P O P- 3 CD 3 PJ Hi C > PI li rt 8> CO H CO 0 H- P- la CD Pi (D rt (r)t pi rt m CD H- rt < 1 £ H<- » Hi h rt PI ^< rt H i< o Ln G > CD ^< rt CO ED H- P»

P* puri t ~j o ~J o ID O K* < •2 1 P- >r LJ l-» O O O \ 0 0 Hi *» o 3 rt 3 ^ ^ — — VO CO p* 01 NJ H CD CD -J o Ln to CO ib. Ln Lon P- 0 3 en 3 Hi N -J Ln -J CO X X 1CD CI G Ln 3 CO I-* h-» O I-1 \B 1 \B o o i- (-• 1 1 h-» o o o O O o 3 rt CD *» Co iQ 0 3 <£ U3 I—* NJ K> N c G ID O P» Ln o o o H B 3 3 m CD

1+ 1 + 1 1 1 + xylan (3-1,4)

+ 1 1 + 1 1 1 arabinan (a-1,5, a-1,3)

+ 1 1 1 1 1 + haze arabinan (a-1,5)

1 1 1 arabinogalactan (3-1,3, 3-1,6)

O-i + 1 1 + 1 + 1 + 1 potato galactan (3-1,4) ID +

1 1+ pentosan (arabinoxylan) 0 rt 1 gentobiose (6-1,6 glucose) < G P- P- 3 1 postulan (3-1,6 glucose) (D CO

Hi 1+ pachyman (3-1,3 glucose) 0 % + + laminarin (6-1,3, 3-1,6 glucose) 0. P- 3 OJ pnp-3-D-glucoside CD 3 ^N< J^ h-* carboxymethylcellulose § CO ai K> avicell * ~" ' + 1 1 i 1 P i + polygalacturonic acid + 1 1 I 1 + 1 1 pectin DE 92 (PL) + i 1 1 + 1 pectin DE 62 (PE)

+ brown ribbon pectin Ill

5.3. Results

5.3.7. Quantity and composition of WIS from fresh tomatoes

Thequantit yan dcompositio no fWI Sprepare d fromfres hrip etomatoe s ispresente di nTabl e30 .A larg edifferenc ei nWI Squantit yi sfoun d betweenth etw ovarieties .WI Si scharacterize db yth epresenc eo flarg e amountso fcellulose ,anhydrogalacturoni caci dan dprotein .Th epecti n presenti nth eWI So ffres htomatoe sshow sa nintermediat edegre eo f esterification.

Table30 :Quantit yan dcompositio no fWI Sprepare d fromfres htomatoe s (%w/w).

Variety H30 H1361 %WI S 0.74 1.06 Rhamnose 1.6 1.5 Arabinose 1.5 3.3 Xylose 3.8 4.6 Mannose 2.4 4.0 Galactose 3.6 3.6 Glucose 29.5 32.3 Totaln.s. * 42.4 49.3 AGA 20.6 19.0 DE 58 59 Protein 16.7 14.5 Cellulose 34.1 33.4 Moisture 6.6 11.9

Total 86.3% 94.7%

*n.s .= neutra lsugar s 112

5.3.2. Microscopic appearence of WIS in tomato juice

In tomato juice the fraction ofWI S comprises only 15-20% of all tomato solids.WI S consistsmainl y ofwhol e tomato cellsvaryin g in diameterbetwee n 50ur nan d 500ym . These cells occur alone or sometimes in clusters. Furthermore fragments ofvascula r bundles frombot h placental tissue and radial and innerwall s of thepericar p are present. Finally smallpiece s of skinmateria l canb e found. (Fig.28) .Thes e results confirm data presented by Reeve et al. (1959).

5.3.3. General composition of WIS, pectin, hemicellulose and cellulose prepared from tomato juice

Table 31 shows the chemical composition ofbot hWI S andWI S fractions for hotbrea k juice of thevarietie s 318 and 208.A s fara s theWI S is concerned the twovarietie s mainly differ in amount ofWI S and in the % of arabinose and galactosewithi n thisWIS .Wit h the exception of theAG A and protein contents the composition of tomato juiceWI S isquit e similar to the composition ofWI S from fresh tomatoes. Thepecti n fraction,i n general 40-45% of theWIS ,contain s large amounts ofAG A and protein.Approximatel y 25-35% of the arabinose and galactose present in theWI S is found as a constituent of thepecti n fraction. Thehemicellulos e consistsmainl y of glucose,xylose ,mannos e and protein. Approximately 50%o f thearabinos e and galactose and 15-20% of the glucose present in theWI S forms part of thehemicellulos e fraction.Dependin g on thevariet y thehemicellulos e fraction comprises 25-30% of tomato juice WIS.Mos t of the glucose present inWI S is found in the cellulose fraction. Surprisingly this fraction also contains smallamount s ofAG A and even+ 15% of the arabinose present in theWIS .Moreove r trace amounts of other neutral sugars are found. Ingenera l tomato juiceWI S contains 30-35% cellulose. 113

9M-- *

'.-*#:|p t\4i :

Fig.28 :Th eWI So ftomat ojuice ,a :whol etomat ocell s b:fragment so fvascula rbundles ,c :fragments'o fski n material (magnification61.4x) , 114

o

o

a. -H •a B 4J 0) o -a e

<-i 3d)

O rt «

*M V U o 3 a to «e CM

3 CD

J3

O

3

•a •H a?

O H •a

6

C5 rf O i-l -H 115

5.3.4. Composition and structural features of WIS pectin

Fractionation of WIS pectin Thepecti npresen ti nWI Sca nb efractionate daccordin gt osolubilit y bysuccesiv eextraction swit h0.2 %ammoniu moxalate ,0.0 5N HC1 ,bot ha t 60°C,an dfinall ywit h0.0 5N NaO H+ 5 m MEDT Aa t4°C .Th edistributio no f thepecti nbetwee noxalate ,HC 1an dNaO Hsolubl efraction si sinfluence db y thesuccessiv eprocessin gstep suse di nth eproductio no ftomat opast efro m freshtomatoes :a sindicate d inTabl e32 .A sa resul to fheatin gdurin ga regularho tbrea kproces san dsubsequen thea tprocessing ,muc ho fth eHC 1 solublepecti ni ssolubilize d causinga decreas eo fbot hth e% o ftota l pectina swel la sth epercentag eo fgalacturoni caci di nth eWIS .Th e solubilizedpecti ni shighl yesterifie da si sindicate db yth edro pi n degreeo festerificatio no fth eunsolubilize dpectin .Concentratio ncause s afurthe rincreas ei nsolubilizatio no fWIS-AGA .Comparabl eresult swer e obtainedfo rtomato ,juic ean dpast eo fvariet yH1361 .

Table32 :Effec to ftomat oprocessin go nth efractionatio npatter no f pectin (varietyH30) .

totalpecti n %AG Ai nWI S %o ftota l pectin fraction fraction (%o fWIS ) oxalate HC1 NaOH

tomato 43.1 14.6(58 ) 20.7 25.6 53.7 juice 41.3 10.9(29 ) 28.4 3.3 68.2 paste 40.8 9.3 (28) 34.3 7.0 58.7

inparenthese sth eD Eo fth epecti n

Table3 3show sth eeffec to ftomat oprocessin go nth esuga rcompositio n ofth eHC 1solubl epecti nfraction so ftw ovarieties .Th ehighl yesterifie d pectin,solubilize da sa resul to fprocessing ,obviousl ycontain sonl y minoramount so fth eneutra lsugar sarabinos ean dgalactose .Thi stabl e alsostresse sth elarg edifference swhic hma yexis ti ncompositio no f 116 pectino rpecti nfraction sfro mdifferen tvarieties .

Table33 :Suga rcompositio n (Mol%)o fHC 1solubl epecti nfractio na s influencedb yprocessin g (varietiesH30 ,H1361) .

tomato juice paste

H30 H1361 H30 H1361 H30 H1361

Rham. 7.6 5.4 7.2 8.1 8.4 7.7 Ara. 4.1 15.9 11.1 21.8 9.8 22.4 Xyl. 1.1 1.0 1.5 1.8 2.1 2.5 Man. - 0.1 - 0.4 0.2 0.9 Gal. 4.4 3.6 9.5 9.4 8.0 10.0 Glue. 0.9 1.1 3.5 4.2 4.9 8.7 AGA 81.9 72.9 67.2 54.3 66.6 47.8

Theresult sfo rth efractionatio no fth eWI Specti nfractio no fnon - pasteurizedho tbrea kjuic eo fvariet yH3 0ar epresente d inTabl e34 . Theammoniu moxalat esolubl epecti nfractio ni scharacterize db ya relativelylo wdegre eo festerification .Thi sfractio ncontain s23.6 %o f theAGA ,5.9 %o fth earabinos ean d4.2 %o fth egalactos epresen ti nth e WIS.Th epecti nsolubl ei nHC 1ha sa highe rD Ean di scharacterize db yhig h amountso frhamnose ,arabinos ean dgalactose .32.7 %o fWIS-arabinose ,19.9 % ofWIS-galactos ean d41.0 %o fWIS-AG Aform spar to fth eHC 1solubl epectin . Oxalatesolubl epecti nan dHC 1solubl epecti ntogethe rcompris e approximately65 %o fth eAG Apresen ti nth eWIS . Another30.6 %o fWIS-AG Ai sboun di nsuc ha wa ytha ttreatmen twit h dilutedalkal ii snecessar y toextrac ti tfro mth etomat ocel lwall .O na percentagebasi sth ealkal isolubl efractio ncontain srelativel ysmal l amountso fAGA ,i ncontras tt oth ehig hamount so fprotei n (60%o ftota l WISprotein )whic har efound .Togethe rth epecti nfraction scontai n95 %o f theAGA ,70 %o fth erhamnose ,49 %o fth earabinose ,39 %o fth egalactos e butals o10 %o fth exylos epresen ti nWIS . 117

Table34 :Compositio no fWI San dpecti nfraction sfo rnon-pasteurize dH3 0 hotbrea kjuice .

WIS 0.2% 0.05 N 0.05N ammoniumoxal . HC1 NaOH+ 5 m MEDT A

Rha. 1.2 2.5 (5.7) 3.7 (6.6) 0.8 Ara. 1.6 1.1 (2.8) 4.5 (8.8) 0.7 Xyl. 4.2 1.2 (3.0) 1.0 (2.0) 0.8 Man. 3.3 0.5 (1.0) - - - Gal. 3.7 1.8 (3.7) 6.4 (10.2) 2.1 Glue. 31.7 1.9 (3.9) 1.5 (2.4) 0.7 AGA 13.3 42.2 (79.9) 47.6 (70.0) 15.1 DE 47 34 55 - protein 19.5 26.2 6.5 43.9 moisture 6.7 12.8 10.0 9.3

%o fjuic e 1.08 0.0£ 0.124 0.290 %o ftota l pectin - 16.2 25.1 58.7 %o ftota l AGA — 23.6 41.0 30.6

* inparenthese sMol %suga rcomposition .

Gelpermeation chromatography of pectin fractions Figures29 aan d29 bsho wth eelutio nprofile so nSephacry lS-20 0an d SephacrylS-50 0fo rth eammoniu moxalat esolubl epecti nfraction .Th e pectini nthi sfractio nha sa heterogeneou smolecula rweigh tdistribution . Theapparen tmolecula rweigh to fth epecti nmolecul eelutin gi nth e includedvolum e (S-500column ,Fig .29b )i sestimate da t40,000-80,000 ;th e amounto fAG Ai nthi smolecul eequal s7 %o ftota lWIS-AGA . Elutiono fth eHC 1solubl epecti nfractio no na Sephacry lS-50 0colum n resultsi nit sseparatio nint otw otype so fpecti nmolecules ,o fwhic hon e hasa molecula rweigh tmuc hhighe rtha nth eoxalat esolubl epectin s(Fig . 29c).Th ehighe rmolecula rweigh tfractio n (poolI )contain s15-16 %o fth e WIS-AGA,th eothe rfractio n (poolII )13-18 %o fWIS-AGA .Poo lI Iwa s fractionatedo na S-50 0colum nwhil eelute dwit hwate r (Fig.29d) .Du et o 118 chargeeffect sa pecti nwit hsmal lamount so fneutra lsugar s (poolIII )wa s separatedfro mth eremainin gpectin .

concentratio n (^g/ml)

concentration l^g/ml) K -AGA 200H o. \ 100- f \-AGA A

100- bO- / M ft •/ r\«* R /" /L, I'IJ rNS 0-1—*—i 1 1 0 0.5 1 0 0.5 1

concentration ((j,g/ml) A concentration (ng/ml) concentration (ng/ml)

100- 100-1 d 100-

50- 50 *\ V-AGA /\-NS

nAl 'AGA\

0 0.5 1 0 0.5 1 0 0.5 1

Fig.29 :Elutio npattern sfro mge lpermeatio nchromatograph yo fpecti n fragmentso ftomat ojuice . a.10. 7m goxalat esolubl epectin ;S-20 0elute dwit hbuffe r b. 9.2m goxalat esolubl epectin ;S-50 0elute dwit hbuffe r c. 9.7m gHC 1solubl epectin ;S-50 0elute dwit hbuffe r d.Poo lI IFig .29 c(1.3 1m gAGA) ;S-50 0elute dwit hwate r e.10. 2m gHC 1solubl epectin ;S-50 0elute dwit hbuffe r •Rhamnose, AArabinose, aXylose , oGalactose, &Glucose . 119

Table3 5show sth esuga rcompositio no fthi sfraction .Th eamoun to fAG A presenti nthi smolecul ecorrespond st o10 %o fth eWIS-AGA .

Table35 :Suga rcompositio ni nMol %o fpoo lII I(Fig .29d )

Rham. 3.6 Ara. 3.0 Xyl. 1.1 Man. 0.3 Gal. 5.1 Glue. 1.5 AGA 85.2

Theelutio no fth eHC 1solubl epecti no nSephacry lS-50 0wa srepeated ;fro m eachfractio nth eAG Aconten tan dneutra lsuga rcompositio nwa sdetermined . (Fig.29e) .Mos to fth earabinos ean dgalactos eco-elute dwit hth ehighe r molecularweigh tpecti nfragment .Thi sfragmen tals ocontaine dmino r amountso fxylose .Par to fth earabinos ean dglucos eelute di nth etai lo f theinclude dpeak ;obviousl ythes esugar sar eno tconnecte dt oth eAG Ao f thelowe rmolecula rweigh tpecti nfragment .

5.3.5. Composition and structural features of tomato hemicellulose

Thesuga rcompositio no fth ehemicellulos efractio nwa sfoun dt ob e lesssensitiv et otomat oprocessin gtha nth esuga rcompositio no fth e pectinfraction .Th ecompositio no fhemicellulos efraction si sgive ni n Table36 . Hemicelluloseisolate dfro mAI Sprepare d fromfres htomatoe swa s fractionatedo nDEAE-cellulos e(Fig .30) .Thi sresulte d invariou s fractionswit hdifferen tsuga rcompositio n (Table 37). Sugarcomposition s derivedfro mmethylatio nanalysi scorresponde dwel lwit hthes efigures . FractionsC ,E an dF ar ehig hi nxylos econtent ;fraction sA ,B an dD containa hig hproportio no fglucose ;fraction sA an dB moreove rcontai n relativelyhig hamount so fmannose . 120

Frommethylatio nanalysi s (Table38 )i tappear stha tth exylos eresidue si n fractionsC ,E an dF ar epredominantl y 1,4-linkedan dver ylikel yoriginat e fromg-l,4-xylan .I nthi sconnectio nfractio nE ,containin g42 %o fth e xylosepresen ti nhemicellulose ,i sth emos timportan tfraction .

Table36 :Effec to ftomat oprocessin go nth esuga rcompositio n (Mol%)o f thehemicellulos efraction ,(varietie sH3 0an dH1361) .

H30 H1361

tomato juice paste tomato juice paste

Rham. 0.9 1.0 1.6 0.8 0.8 1.0 Ara. 5.2 5.6 6.0 6.2 5.5 5.6 Xyl. 24.0 24.5 26.8 30.0 30.2 29.3 Man. 16.5 15.8 17.9 17.8 17.9 17.2 Gal. 9.4 9.7 9.6 9.7 9.7 9.7 Glue. 38.4 37.9 33.3 29.5 31.6 31.9 AGA 5.5 5.4 4.7 5.8 4.4 5.3

concentration {|4.g/ml I

800 volume (ml) 10.05M| 0.1 M 10.5 M I0.1N |0.2N|Q5N H20 I KAc (pH 6.0) I NaOH

Fig. 30:Fractionatio no ftomat ohemicellulos eo nDEAE-cellulose . 121

3

o» CT

w w w .f

H r-- in in j* H n H M U

< o in H

o

a

2 122

Table 38: Glycosyl linkage composition of tomato hemicellulose fractions.

Monosaccharide Deduced glycosyl Occurence of type of glycosyl linkage per sugar unit in% linkage ^^^^_^_^^_^______^_^_^ A B C D E F Residue

Araf. terminal 100 100 100 83 100 100 1,2,3,5- 17

Xylp. terminal 51 12 6 42 7 4 45 1,4- 5 T 82 25 81 75 13 35 1,2- 28 JL 24 33 1,2,4- 4 4 8 10 1,2,3 -"^_"S:»1 6 0.3 4 1 1 1,3,4-- - ' 2 3 4 1,2,3,4- 53 7 1 6 9

Hanp. 1,4- 69 74 100 82 80 100 73 1,4,6- 27 26 18 20 27 1,2,4,6- 4

Galp. terminal 70 100 100 100 21 1,4- 26 100 52 1,3,4- 27 1,3,4,6- 4 100

Glucp. terminal 5 23 8 12 1,4- 61 39 39 49 57 74 47 1,4,6- 26 19 25 43 35 14 33 1,2,4- 4 1,3,4- 3 1,2,3,4- 4 1,3,4,6- 1 2 5 1,2,3,4,6- 3 42 13 6 8 p indicates pyranose ringform f indicates furanose ringform

These three fractions also contain most of the AGA present in the hemicellulose. Fractions A, D and the insoluble residue have high proportions of terminal xylose residues, 1,2-linked xylose residues, and 1,4 as well as 1,4,6-linked glucose residues, as is typical for xyloglucans. Further support for the presence of xyloglucans can be derived from the results presented in Fig. 32b. The presence in fraction A of 1,4-linked glucose, 1,4-linked mannose with branching at C-6, and terminal galactose may point to the occurrence of galactoglucomannans. However, due to the complex composition of these fractions the interpretations, are tentative. 123

Ingenera lth etomat ohemicellulos econtain s40-50 %o fth earabinos ean d galactosepresen ti nth eWIS .I ti snoteworth ytha tal lth earabinos e residuesar eterminall ylinked ,indicatin gth eabsenc eo farabinan si nth e hemicellulosefraction .Approximatel y 15%o fth ehemicellulos egalactos ei s 1,4-linked(fractio nA ,insolubl eresidue )an dma yb epresen ta sgalactans . Gelpermeatio nchromatograph y onSephacry lS-50 0(Fig .31) ,wit hwate ra s eluent,reveale da hig hmolecula rweigh turonid econtainin gfractio nric h inxylos e(Fig .31a ,b) , anda ninclude dfractio nwit hlowe rmolecula r weightpolymer sparticularl yric hi nmannos ean dglucos e (Fig. 31c).

concentration l^g/ml

60- C. ft 1\ —-man 40- ! \ , 1 A—glut \ / \ V-gQ' \ 20- - /j /-*> 0- 0 0.5 1 0.5 1

Fig.31 :Ge lpermeatio nchromatograph yo ftomat ohemicellulos eo n SephacrylS-500 ,eluen tdistille dwate r a.Elutio npatter nfo rAG Aan dtota lneutra lsugar s(NS ) b.Elutio npatter nfo rarabinos e (ara),xylos e (xyl)an d galactose (gal) c.Elutio npatter nfo rmannos e (man)an dglucos e(glue )

Aftertreatmen to fth ehemicellulos efractio nwit hpurifie dendo- 8 -1,4-glucanase,th evoi dpea kalmos tcompletel yshift st oth einclude d fraction(Fig .32a).Th eremainin gpolysaccharide si nth evoi dpea kcontai n onlyxylos ean duroni cacid .I ti sno tknow nwhethe rthi suroni caci di s galacturonicaci do rmethyl-O-glucuroni cacid .Th efractionatio npatter no f theS-50 0include dpoo lo nBio-Ge lP- 2i sshow ni nFig .32b .Th e compositiono fth eintermediat eoligomeri cfragment si stypica lfo r xyloglucanfragments .Presumabl ybot hxylan san dxyloglucan sar edegrade d byth eendo-glucanas eused .Th emultisubstrat especificit yo fthi senzym e wasshow nb yBeldma ne tal . (1986).Th ehig hmannos econten ti nth eP- 2 124

void,i nparticula ri nth etai lo fth epea ki sals ostriking .Th edime r fractioni shig hi nxylos ean darabinose .Xylos ean dglucos ear eth emai n monomericsugar sfound .

concentration (ug/m ) cone entration (^g/ml) b. tetra hi di mono a. 150-

400- \\ I- ll 100- Ara 04 1.0 9.3 10.2 3.7 208 - Xyl. 18.7 4.4 28.2 38.1 455 60.1 45.9 Maa 37,5 63.1 2.2 0.5 - r Gal. 14.5 2.0 13.5 5.1 3.0 1.3 200- 50- Glue. 28.9 29,5 46 8 1.6.6 50.8 14.9 52.8

0- «*A ^ 0 05 1l\ 0.2 0.4 0.6 0.8

Fig.32 :Elutio npattern sfo rendo-glucanas etreate dtomat ohemicellulose . a.S-50 0elute dwit hbuffe r b.include dpoo lFig .32a ;P- 2elute dwit hwate r

5.3.6. Composition and structural features of the fragments solubilized from WIS by enzymes

Asecon dapproac ht oth estud yo fth ecompositio nan dstructur eo f tomatoWI Si sth eanalysi so fWI Sfragment ssolubilize db ypurifie d pectolytic,cellulolyti can dhemicellulolyti cenzymes . Thecharacteristic sfo rth epurifie denzyme sar egive ni nsectio n5.2.4 . (Table 29); thechemica lcompositio no fth eWI Si sshow ni nTabl e34 .Tabl e 39present sth eresult so fthes estudies .I ti sobviou stha tnon eo fth e enzymesha sbee nabl et osolubiliz esubstantia lamount so fprotei nfro mth e WIS. Asa contro lexperimen tWI Swa ssuspende di n0.0 5M sodiu macetat e bufferp H5.0 ,th esam ebuffe ra suse dfo rth eenzym eexperiments .Thi s treatmentresulte d inth esolubilizatio no fsom esugars .Ge lpermeatio n chromatography ofthes esugar so nSephacry lS-20 0indicate d thepresenc eo f highmolecula rweigh tpecti nfragment sa swel la slo wmolecula rweigh t neutralsuga rcontainin gfragment s(Fig .33) .Th eonl yneutra lsugar sfoun d inth elo wmolecula rweigh tfragment swer eglucos ean dmannose . 125

Table39 :Amoun to fsugars ,AG Aan dprotei nsolubilize d fromWI Sa sa resul to fenzym eactio n

(%o ftota lamoun tpresent) .

enzyme Rham. Ara. Xyl. Man. Gal. Glue. AGA Protein

Control 4.0 5.4 1.2 1.0 5.0 1.6 7.5 0.0 PL 36.4 45.6 9.2 2.0 23.4 2.7 58.4 4.1 PG 21.8 15.0 7.2 0.2 9.6 0.2 54.9 6.7 PE+ P G 30.1 49.1 10.8 0.0 28.0 0.4 83.9 5.0 exo-glucanase+ endo-glucanas e 21.421. 4 36.2 48.0 45.1 21.8 42.6 22.8 4.5 exo-g+ endo- g+ P E+ P G 42.7 68.7 69.0 46.5 47.2 49.4 77.3 5.5 exo-g+ endo- g+ P L 33.0 91.0 58.5 40.8 44.5 40.2 70.8 0.0 galactanase 18.5 34.7 1.4 0.2 24.9 0.8 43.1 0.0 exo-arabinanase 0.0 38.1 1.3 0.9 0.0 1.2 0.0 2.5 exo-arabinanase+ P L 28.4 52.4 5.3 0.8 16.9 2.0 40.4 0.0 exo-arabinanase+ P G 15.4 47.7 5.5 3.6 7.3 1.5 35.9 3.9

exo-g™ exo-glucanase ,endo- g« endo-glucanase .

concentration (|4,g/ml 40-1 -AGA NS

20-

-NS -JLh LI 0.5 \ KQ V

Fig.33 :Elutio npatter n (S-200,buffe r elution)fo rsugar ssolubilize d from28. 6m gWI Sa sa resul to f incubationwit hbuffer . 126

Fragments solubilized by PL Apartfro mhavin ga sligh tactivit yo npotat ogalacta nth epectlnlyas e usedi nthi sstud yca nb econsidere da pur eenzyme .Larg eamount so fAGA , arabinose,rhamnos ean dgalactos ewer erelease d fromWI Sa sa resul to f incubationwit hpectinlyase .Th erati oAGA/neutra lsugar si nth ereactio n liquidwa s2.0 ;th edegre eo fpolymerisatio n (totalsugar/reducin gsuga r ratio)o fth esolubilize dfragment swa sestimate da t6 . Thefractionatio npatter no fth ereactio nliqui do nSephacry lS-20 0i s showni nFig .34a .Tabl e4 0give sth esuga rcompositio no fth efragmen t voidingth eS-20 0colum ni na sharp ,narro wpeak ,(poo lI) .

concentration (jig/ml

concentration l^g/ml)

100-

50-

*—i—i—*i—•- 0 0.5 1

Fig. 34:Elutio npattern sfo rfragment ssolubilize dfro m58. 9m gtomat o WISa sa resul to fincubatio nwit hPL . a.Tota lreactio nliquid ;S-20 0elute dwit hbuffe r b.Poo lI IFig .34 a (3.7m gAGA) ;S-20 0elute dwit hwate r c.Poo lI VFig .34 b (0.9m gAGA) ;P- 2elute dwit hwate r

Sinceth eenzym ewa spur ei tca nb eassume dtha tth eneutra lsugar sar e boundt oth epecti nfragment .I tca nb econclude dtha tP Lsolubilize sa highmolecula rweigh t (»80,000) ,highl ybranche dpecti nfragmen t containing9 %o fth eAG Apresen ti nWIS .Noteworth yi sth epresenc eo f xylosei nthi sfragment . 127

Table40 :Suga rcompositio n (Mol% )o fpoo lI (Fig . 34a).

Mol% %o ftota lamoun tpresen ti nWI S

Rham. 10.5 21 Ara. 18.7 26 Xyl. 9.9 5 Gal. 21.3 16 AGA 38.3 9

Thefractionatio npatter no fth ereactio nliqui do nSephacry lS-20 0 (Fig.34a )als oindicate stha tth ebul ko fth esolubilize dAG Ai sno tboun d toth elo wmolecula rweigh tneutra lsuga rcontainin gfragments .Poo lI Iwa s concentratedan dfractionate do nSephacry lS-20 0whil eelute dwit hwate r (Fig.34b) .Du et ocharg eeffect suronid econtainin gfragment swer e separated fromneutra lsuga rcontainin gfragments .Th eelutio nprofile s showni nFigure s34 aan d34 bindicat etha tapproximatel y40 %o fth e AGApresen ti nWI Si ssolubilize db yP Li nth efor mo fintermediat esize d (MW± 10,000-40,000 )polymers ,containin gsmal lamount so fneutra lsugar s (Table41) .

Table41 :Suga rcompositio n (Mol%)o fpoo lII Iafte rfractionatio no n SephacrylS-20 0(Fig . 34b).

Mol% %o ftota lamoun tpresen ti nWI S

Rham. 1.7 Ara. 1.8 Xyl. 1.5 Man. 0.2 Gal. 2.0 Glue. 3.5 AGA 89.2 40 128

Theneutra lsuga rcontainin gfragment s (poolIV ,Fig .34b )wer e fractionatedo nBio-Ge lP- 2 (fig.34c) .Onl y5 %o fth eAG Apresen ti nWI S isdegrade d tomonomer/dime ra sa resul to fP Laction ;8 %o fth eAG Ai s degraded tooligomeri csize dfragments .Th emonomeri cfractio ncontain s glucosean darabinos e (3%o ftota lpresen ti nWIS) ; thedimeri cfractio n onlyglucose .Mos to fth eneutra lsugar swer efoun di na fractio nwit ha degreeo fpolymerisatio nbetwee n5 an d 10.Thi sfractio nconsist slargel y ofglucos e (56.9Mol% ,1.2 %o ftota lWIS-glucose )an dxylos e (23.3Mol% ,3 % oftota lWIS-xylose) .

Fragments solubilized by PC Asha sbee nshow ni nTabl e2 9th epolygalacturonas euse di nthi s studywa spur eexcep tfo rhavin ga sligh tactivit yo n8-1,4-xylan .Th e amounto fAG Asolubilize da sa resul to fP Gactio ni scomparabl et oth e amountsolubilize db yP Ltreatmen t (Table39) .

concentration ((ig/ml)

i concentration (^g/ml) 300- 150

200- 100-

100- t II \ \\

V—,—.i^ - 0.5 K„

Fig. 35:Elutio n patterns forfragment s solubilized from 56.4m g tomato WIS asa result of incubation withPG . a.Tota lreactio nliquids ;S-20 0elute dwit hbuffe r b. PoolI , Fig.35 a (2.3m gAGA) ; S-500elute dwit hbuffe r c.Poo lII ,Fig .35 a (0.7m gAGA) ; P-2elute dwit hwate r 129

However,th eamount so farabinos ean dgalactos esolubilized ,respectivel y 15.0%an d9.6 %o fth equantitie spresen ti nWIS ,ar econsiderabl ylowe r thanth equantitie srelease db yP Laction . Therati oAGA/neutra lsugar si nth ereactio nliqui dwa s5 ;th eaverag e degreeo fpolymerizatio no fth esolubilize dfragment swa sestimate da t5 . Figure35 ashow sth efractionatio npatter no fth ereactio nliqui do n SephacrylS-200 .35 %o fth eAG Apresen ti nWI S (67%o ftota lsolubilize db y PG)i ssolubilize da shig hmolecula rweigh tfragments . PoolI wa sconcentrate dan dfractionate do nSephacry lS-50 0(Fig .35b) . The elutionprofile sshow ni nFigure s35 aan d35 bindicat etha t24 %o fth e WIS-AGAi ssolubilize da sfragment swit ha napparen tmolecula rweigh t between40,00 0an d80,000 .Th emola rsuga rcompositio no fthes efragments , giveni nTabl e42 ,closel yresemble dth esuga rcompositio no fth e unfractionated reactionliquid ,excep tfo rth eamoun to fxylos epresent .

Table42 :Suga rcompositio n (Mol% )o ffragment swit ha M Wo f± 40,000-80,000release db yP G (Fig. 35b).

Mol% %o ftota lamoun tpresen ti nWI S

Rham. 3.9 10 Ara. 4.3 8 Xyl. 1.6 1 Man. - Gal. 5.3 5 Glue. 1.8 AGA 83.2 24

Thedegre eo festerificatio no fth epecti ni nth ereactio nliqui dwa sfoun d tob e58 ,whic hi shighe rtha nth eD Eo fth etota lWIS-pectin .Obviousl y thishig hD Ei sunfavourabl efo rfurthe rP Gcatalyse ddegradatio no f solubilizedpectin . Fig.35 ashow stha tonl y 10%o fth eWIS-AG Ai sdegrade dt olo wmolecula r weightfragment s (poolII )a sa resul to fP Gaction .Fractionatio no fthes e fragmentso nBio-Ge lP- 2 (Fig.35c )reveale dtha t4 %o fthi sAG Ai s 130 degraded presumably tomonome r and dimer and 3% to oligomers. The only neutral sugars determined in theBio-Ge lP- 2 fractionswer exylose ,glucos e and aver y small amount ofmonomeri c galactose. The presence of xylose and glucosema y be explained by the residualxylanas e activity of theenzyme .

Fragments solubilized by PE + PC By the simultaneous action of PE and PG,84 %o f theAG A present inWI S could be solubilized; only 9% remained unsolubilized. Compared to the effect ofP G action,th e combination ofP E and PG induced the solubilization of large amounts of arabinose and galactose.Th e ratio AGA/neutral sugars forP E+ PGwa s 3.3,compare d to 2.0 forP L and 5.0 for PG.Apparentl y the arabinose and galactose containing fragments solubilized by PE+ PG,bu t not by PG alone,ar e part of highly esterified regions in theWI S pectin. Since thesehighl y esterified regionswer e also released by PLbu t notb y PG alone they are apparently not connected to the cellwal l matrix bymean s of low esterified pectinregions .

Fragments solubilized by endo-glucanase + exo-glucanase The combination of endo-an d exo-glucanase (fractions III.an d exo III,Beldma n et al. 1985)release d large amounts of glucose,xylos e and mannose containing polymers.Als o noteworthy is the solubilization of22.8 % of theWIS-AGA . Because of the absence of pectolytic enzyme activity (not found in the crude technical enzyme preparation), these solubilized high molecular weight pectin fragments (as indicated by S-500ge l permeation) were probably bound to solubilized parts ofhemicellulos eand/o r cellulose. Due to small impurities of arabinanases and galactanases and the activity of endo III on arabinoxylan it isdifficul t to draw conclusions about the occurrence of arabinose and galactose molecules in these solubilized pectin fragments. Solubilized neutral sugars (eluting in the included of the S-500 fractionation)wer e fractionated onBio-Ge lP- 2 (Fig.36) .Th e elution pattern shows four peaks:monomers ,dimer san d twopeak swit h a degree of polymerisation between 6an d 10 (K 0.4 and 0.1 respectively). Glucose sugarswer e mainly present as cellobiose (27%o fWIS-glucose) , indicating cellulose degradation and a low cellobiase activity. 6% of the WIS-glucose was present asmonomer ; the remaining part of the solubilized glucose was found in the fragmentswit hK 0.1 and 0.4. 17%o f theWIS-xylos e is 131 degraded tomonomers ;21 %t ooligomers .

concentration (ng/ml) di

1500

1000

Fig. 36: Elution pattern for neutral sugar containing fragments solubilized from 57.2m g tomato WIS as a result of incubation with endo- and exo-glucanase. P-2 column, eluted with water. The peak with K 0.4 contains 13% of the WlS-xylose, the peak with K 0.1 r av av 8%. Mannose was found as oligomers and dimers; 25% of the WIS-mannose in the peak with K 0.4, 9% in the peak with K 0.1 and 6% as dimer. r av av

Fragments solubilized by PE + PC + endo-glucanase + exo-glucanase In generaln osynergisti c effectswer e observed when endo-glucanase, exo-glucanase,P Ean dP Gwer e added simultaneously toWIS . Only forxylos edi dth ecombinatio n ofenzyme s releasemor e sugar thanth e sumo fth eseparat e actions.Thi s could mean thatxylan sar einvolve di n thebindin g ofpecti n tocelluloses .Th eamoun t ofgalactos e solubilizeda s 132 aresul to fcombine dpectolyti can dcellulolyti cenzym eactio ni sth esam e asth esu mo fth eseparat eeffects .Obviousl yth epectolyti can d cellulolyticenzyme sac to ndifferen tgalactos econtainin gpolymers .Th e amountso frhamnos ean darabinos esolubilize da sa resul to fth ecombine d additiono fcellulases ,P Ean dP Gwer eeve nles stha nth ecombine dseparat e effects.Part so fth erhamnos ean darabinos esolubilize db yglucanase sar e alsosolubilize db yth ecombinatio no fP Ean dPG .Th eadditio no f glucanasest oP Ean dP Gdoe sno tincreas eth esolubilizatio no fAGA ; glucanasesapparentl y solubilizepart so fWI Specti nals osolubilize db yP E +PG .

Fragments solubilized by PL + endo-glucanase + exo-glucanase Thecombine dactio no fglucanase san dP Lsolubilize s91 %o fth e WIS-arabinose;thi si smor etha nth esu mo fth eamount so farabinos e solubilizedb yglucanase san dP Lseparately .I tappear stha tarabina n chainsfor ma connectio nbetwee ncellulos ean dpectins .Thi sfindin gi s supportedb yth epresenc eo f15 %WIS-arabinos ei nth ecellulos efraction . Thearabina ni nthi sparticula rarabina ncontainin gpecti nsegmen ti sno t solubilizedb yP Lalon eno rb yth ecombine dactio no fglucanase s+ P E+ PG . Theamount so fxylos ean dgalactos esolubilize db yth ecombine dactio no f glucanasesan dP Lwer eth esu mo fth eindividua lamount ssolubilized .Th e amounto fAG Asolubilize db yth ecombine dactio no fglucanase san dP Lwa s lesstha nth esu mo fth eseparat eeffects ;par to fth eamoun tsolubilize d byglucanase si sals osolubilize db yPL .

Fragments solubilized by pure galactanase Theadditio no fpur egalactanas eresulte donl yi nth esolubilizatio n ofsignifican tamount so fAGA ,arabinose ,galactos ean drhamnose .Th e purifiedgalactanas ewa sfre eo fPG ,PL ,xylanas ean darabinanas eactivit y andshowe dn oactivit yo na g-1,3 ,B-1, 6linke darabinogalacta n (Table29) . Fractionationo fth ereactio nliqui do nSephacry lS-20 0(Fig .37a )show s thatth eAG Ai ssolubilize da shig hmolecula rweigh tfragments .Poo lI wa s concentratedan dfractionate do nSephacry lS-50 0(Fig .37b) .Figure s37 a and37 bsho wtha t21 %o fth eWIS-AG Ai ssolubilize da sfragment swit ha n apparentmolecula rweigh tbetwee n40,00 0an d80,000 .Th eresult sindicat e 133

concentration (^,g/ml) concentration (fig/ml 200- 100-j b.

AOA NS A 100- 50- AGA

NS

0-<-"*n 1 r* -*n 1 *r 0 05 1 05 1

Fig. 37: Elution patterns, for fragments solubilized from 37.1 mg tomato WIS as a result of incubation with pure galactanase. a. Total reaction liquid; S-200 eluted with buffer b. Pool I Fig. 37a (1.3 mgAGA) ; S-500 eluted with buffer

that galactans are important in the binding of pectin to the cell wall matrix. The arabinose solubilized as result of pure galactanase action (34.7% of the WIS-arabinose) is not bound to the high molecular weight AGA. It is found in the included fraction and in the intermediate fraction of the S-200 fractionation (Fig. 37a). Solubilized galactose is mainly present in the included fraction of the S-200 fractionation but is also found in the void. The fact that the arabinan is not found associated with the high MWpecti n fragments implies that it was originally part of the galactan polymer.

Fragments solubilized by exo-arabinanase Purified exo-arabinanase is very specific in solubilizing arabinose from WIS (Table 39). The absence of any solubilized galactose, rhamnose and AGA points to the presence of at least 38% of the WIS-arabinose as side chains on other cell wall polymers. 134

Fragments solubilized from the residues of enzyme treated WIS The WIS residues left after the enzyme incubations were freeze-dried and again incubated with enzymes. The results are shown in Table 43. The addition of PG to the residue of endo-glucanase + exo-glucanase treated WIS results in the solubilization of 23% of the WIS-AGA. This result confirms the conclusion that endo-glucanase + exo-glucanase solubilize a part of the WIS-AGA which is also solubilized by PG. Gel permeation of the reaction liquid on Sephacryl S-200 shows that approximately 60% of the solubilized AGA is present in the void peak, together with most of the solubilized rhamnose, arabinose and galactose. PG is not able to solubilize large amounts of AGA from the residue of PL treated WIS; however, when compared to the amount of solubilized AGA, the amounts of solubilized rhamnose, arabinose and galactose are rather high (compare with Table 39: effect of PG alone).

Table 43: Amounts of sugar and AGAsolubilize d from WIS residues as a result of enzyme action (.% of total amount originally present in WIS). residue of treatment with: second enzyme Rham. Ara. Xyl. Man. Gal. Glue. AGA endo-g + exo-g PG 11.3 8.2 5.5 2.9 3.7 2.8 23.0 PL PG 13.4 10.0 6.8 1.0 6.2 0.4 15.4 endo-g + exo-g PL 36.5 36.3 5.7 2.5 17.8 3.3 46.6 PG PL 22.8 30.5 9.6 2.4 12.6 2.7 31.3 galactanase PL* 47.5 41.6 18.8 8.5 18.8 5.3 25.9

* 0.1 ml enzyme added to 50 mgresidu e instead of 0.2 ml (Table 29).

Obviously PL has released those parts of the WIS pectin which are low in rhamnose, arabinose and galactose content and which are also liable to PG action. In contrast to PG, PL is able to solubilize large amounts of AGA from the residue of endo-glucanase + exo-glucanase treated WIS. The molar sugar composition of the reaction liquid and the fractionation pattern on Sephacryl S-200 are very similar to those for the reaction liquid of WIS treated with PL alone. The glucanases and PL obviously attack unrelated 135

partso fth eWIS . Theincubatio no fth eresidu eo fP Gtreate dWI Swit hP Lresulte di n thesolubilizatio no f31 %o fth eAGA ,31 %o fth earabinose ,23 %o fth e rhamnosean d13 %o fth egalactos epresen ti nth eorigina lWIS .I tappear s thatpar to fth eAG Arelease db yP Lca nals ob erelease db yPG .Th epecti n segmentssolubilize db ybot henzyme sar elo wi nrhamnose ,arabinos ean d galactosecontent . Theadditio no fP Lt oth eresidu eo fWI Streate dwit hpurifie d galactanaseresult si nth esolubilizatio no f26 %o fth eWIS-AGA ,whic hi s accompaniedb ysolubilizatio no famount so frhamnose ,arabinos ean d galactosecomparabl et oth eamount ssolubilize db yP Lalone .Noteworth yi s thesolubilizatio no f19 %o fth eWIS-xylos ean d5 %o fth eWIS-glucose ; thesesugar swer eals ofoun di nth ereactio nliqui dafte rtreatmen twit hP L alone(Fig . 34c).

5.4 Discussion

TheWI Squantitie spresen ti ntomatoe san dtomat ojuic ea sdetermine d inou rstudie srange dfro m0.74 %t o1.14 %o na fres hweigh tbasis ;thes e dataar ecomparabl et oth evalue sreporte di nth eliteratur e (seesectio n 2.5).Apar tfro mth edifference si nWI Squantity ,th esample smainl y differed in% o farabinos ean dgalactos epresen ti nth eWI San dWI S fractions (Tables30 ,31 ,33) .I ti swel lknow ntha tth eamount so fWIS , arabinosean dgalactos ear estrongl yinfluence db yvariet yan dleve lo f ripeness (section2.3.2.) . Basedo nou rdat atomat oWI Sca nb efractionate d into40-45 %pectin ,25-30 % hemicellulosean d30-35 %cellulose .Jon a& Fo a (1979)reporte dvalue so f respectively51% ,18 %an d31% .Gros s (1984),wh ofractionate d thecel l wallso fth eoute rpericarp ,foun d48.3 %cellulos ei nadditio nt o32.4 % pectinan d19.3 %hemicellulose .Jon a& Fo a (1979)ha dalread yreporte dtha t thecel lwall so fth eoute rpericar pwer eric hi ncellulose . Thepecti nfractio ncontaine dalmos tal lo fth eAGA ,th emajo rpar to fth e protein,bu tles stha n50 %o fth earabinos ean dgalactos epresen ti nth e WIS.Surprisingly ,th epecti nfractio ncontaine d 10%o fth eWIS-xylose .Th e presenceo fxylos ei ncel lwal lpecti nha sals obee nreporte db yothe r 136 authors(d eVrie s1981 ,1983 ;Steven s& Selvendra n 1984).Maimos ean d xylosewer emainl yfoun di nth ehemicellulos efraction .Thi sfractio nals o containedu pt o50 %o fth earabinos ean dgalactos epresen ti nWIS ,15-20 % ofth eWIS-glucos ean dsubstantia lamount so fth eprotein . Themajo rpar to fth eglucos epresen ti nWI Swa sfoun di nth ecellulos e fraction.Thi sfractio nals ocontaine d 15%o fth eWIS-arabinose .Th e presenceo farabinos ei ncellulos efraction si sconfirme db ySteven s& Selvendran (1980)an dVorage ne tal . (1983). Theinsolubl epecti no fnon-pasteurize dho tbrea kjuic ecoul db e fractionatedint o16.2 %oxalat esoluble ,25.1 %HC 1solubl ean d58.7 %NaO H solublepecti n (Table34) .Thes efraction scontaine drespectivel y23.6% , 41.0%an d30.6 %o fth eWIS-AGA .Thes evalues ,whe nbase do njuic eweight , arecomparabl et oth edat areporte di nth eliteratur e (Table 11).Th e degreeo festerificatio no fth eAG Apresen ti nth epecti nfraction s correspondswel lwit hth evalue sreporte db yLu h (1954a).A sexpected ,th e ammoniumoxalat esolubl epecti nha da lo wdegre eo festerification ,whil e theHC 1solubl epecti nwa smor ehighl yesterified . Theammoniu moxalat esolubl epecti nwa scharacterize db ylo wquantitie so f arabinosean dgalactose ;rhamnos ewa sth emajo rneutra lsuga ri nthi s fraction.7 %o fth eAG Apresen ti nWI San disolate dfro mth eoxalat e extractha da napparen tmolecula rweigh to f40.000-80.000 .Th eremainin g oxalatesolubl eAG Awa sfoun di nlarge rfragment swit ha wid eM W distribution (Fig. 29b). TheHC 1solubl epecti nfractio ncontaine d67 %o fth earabinos ean d51 %o f thegalactos epresen ti nth etota lWI Spectin .15-16 %o fth eWIS-AGA , solubilizedb yHC1 ,wa spresen ti na fragmen twit ha nestimate dM Wo f1 0. Thishig hmolecula rweigh tpecti nfragmen tcontaine dlarg eamount so f neutralsugars ,primaril yrhamnose ,arabinos ean dgalactos e (Fig.29c,e) . 10%o fth eAG Apresen ti nth eWI San dsolubilize db yHC 1wa sisolate da sa n intermediatesize dfragmen t (figures29c ,d )wit hlo wamount so farabinos e andgalactos e (Table35) ; incontras tt oth eoxalat esolubl efragment swit h alo wamoun to farabinos ean dgalactos eth epecti ni nthi sfragmen t probablyi smor ehighl yesterified . Allisolate dpecti nfragment scontaine d smallbu tsignifican tamount so f xylose. TheNaO Hsolubl epecti nfractio nconsiste do f44 %protei nbu tcontaine d 137 onlya smal lproportio no fAGA .Thi sresul tconfirm sdat afoun db yGros s (1984).Thi sfractio nwa sno tcharacterize d furtherbecaus eth epecti nwa s saponifiedan dpartl ydegrade da sa resul to ftreatmen twit halkali . Hemicellulosesfro mvegetable san dfruit ssho wa ver ydivers e composition (Voragene tal . 1983,Silih a1985 ,Hollowa y& Grei g 1984).Th e tomatohemicellulos ei scharacterize db yhig hamount so fmannos e(lik e mangoan d cherry),almos tn ofucose ,an drelativel y lowamount so f arabinose (incontras tt ocherry ,pineapple ,appl ean dcarrot )whe n comparedt ohemicellulose so fothe rsources .Io nexchang echromatograph yo f tomatohemicellulos eo nDEAE-cellulos ean danalysi so fglycosidi clinkag e compositiono fisolate d fractionsshowe dtha t60 %o fth ehemicellulos e xylosei sprobabl ypresen ta sa 1,4-linke dxyla nwit hsom ebranch-point sa t C-2 (fractionsC, Ean dF ,Tabl e37) .32 %o fth ehemicellulos exylos emigh t bepresen ta sxylogluca n (fractionsA an dD ,Tabl e37) .Thes eresult sar e incomplet econtras tt oth eresult sobtaine dwit happl ehemicellulos e (Voragene tal .198 6b) .N oindication swer efoun dfo rth epresenc eo f fucogalactoxyloglucansa sreporte db yVorage ne tal .(198 6b) . Thefac ttha txyla nco-elute swit hAG Aupo nio nexchang echromatograph y (Table37 )an dge lpermeatio nchromatograph y (Fig.31a ,b) ,indicate sa linkagebetwee nAG Aan dxylan .Al lth earabinos ean dth emajo rpar to fth e galactosewa spresen ta stermina lresidues .Ther ewer en oindication sfo r thepresenc eo fpecti carabinan si nth ehemicellulose ;approximatel y 15%o f thegalactos eresidue swer e1,4-linke d andwer epossibl ypresen ta s galactans.Onl y 10%o fth earabinos efoun di nhemicellulos ewa sdetecte di n thesam efraction sa sth exylan ;therefor eth ebranchin go fxyla nwit h arabinofuranosidewa slimited .55 %o fth ehemicellulos earabinos ewa sfoun d inth esam efraction sa sth exylogluca nfragments .Par to fth etermina l arabinosemolecule sar econnecte d toxylos eresidues ,a si sindicate db y thecompositio no fth edime rfractio nobtaine dafte rfractionatio no f endo-glucanasetreate dhemicellulos eo nBio-Ge lP- 2 (Fig. 32b). Theresult sals oindicat eth eexistenc ei ntomat ohemicellulos eo fa 1,4 bound (gluco)mannanwit hbranc hpoint sa tC- 6whic hma yb elinke dt o terminalgalactose . Asa resul to fth eincubatio no ftomat oWI Swit hpurifie dP La hig h molecularweigh thighl ybranche d pectinfragmen t (MW^80,000 )wa s solubilized.Thi sfragmen tcontaine d 9%o fth eAGA ,26 %o fth earabinose , 138

16%o fth egalactos ean dI nadditio n5 %o fth exylos epresen ti nWIS .I ti s likelytha tthi sfragmen tform spar to fth ehig hM WHC 1solubl epecti n fragmentric hi nneutra lsugar s (poolI ,Fig .29c) .Therefore ,th epecti n segmentsi nthi sHC 1solubl epecti nfragmen toutsid ethes ebranche dregion s consisto falmos tunbranche dAG Achain swit ha hig hDE .Thes echain swer e degradedb yPL .40 %o fth eWIS-AGA ,solubilize d asresul to fPL ,wa sfoun d insevera lfragment swit ha nestimate dM Wo f10,00 0- 40,000 .Thes e fragmentsonl ycontaine dsmal lamount so farabinos ean dgalactos e(Tabl e 41); sincethes efragment scoul dno tb efurthe rdegrade db yPL ,thei rD E wasprobabl ylow . Apecti nfractio nwit ha M Wo f40,00 0- 80,000 ,containin g24 %o fth e AGApresen ti nWI San dlo wi narabinos ean dgalactos econten twa sisolate d fromth eP Gdiges to ftomat oWIS .(Fig .35a ,b) . Withrespec tt omola r sugarcomposition ,M Wan dDE ,thi specti nfragmen tclosel yresemble da par t ofth eHC 1solubl epecti n (Tables35 ,42 ,Figur e29 dpoo lIII) ;however , thetota lAG Aquantit yi nth eP Gsolubilize dfragmen twa stwic ea shigh . 20%o fth eAG Apresen ti nth eWI Swa sdegrade dt ointermediat ean dsmal l sizedfragment sa sa resul to fP Gaction .I ti sprobabl etha tth emajo r parto fthes edegradatio nproduct soriginat efro mth eammoniu moxalat e solublepecti nfraction . Simultaneousactio no fP Ean dP Grelease dlarg eamount so farabinos ean d galactose;thi si si ncontras tt oth eactio no fP Galone .Th esolubilize d arabinosean dgalactos etherefor eoriginate d frompecti nfragment s connected toth ecel lwal lmatri xthroug hhighl yesterifie dpecti n molecules.Thes eresult sconfir mth eresult sobtaine dafte rP Laction . Theadditio no fP Gt oth eresidu eo fP Ltreate dWI San dth eadditio no fP L toth eresidu eo fP Gtreate dWI Sshowe dsom edegre eo foverla pi n solubilizedpectin .Bot henzyme sapparentl y solubilizea pecti nfragmen t lowi nrhamnose ,arabinos ean dgalactos econtent .Thi svie wi sfurthe r supportedb yth eresult so fth ege lpermeatio nchromatograph y ofth ediges t obtainedafte rP Lan dP Gaction . Thefac ttha tpurifie dendo-galactanas erelease dpecti nfragment s(M W 40,000- 80,000 ,21 %o fWIS-AGA )a swel la sfragment scontainin ggalactos e andarabinose ,indicat etha tarabinogalactan spla ya rol ei nconnectin g pecticpolymer st oth ecel lwal lmatrix .Thes earabinogalactans ,degrade d byendo-galactanase ,ar eno ta par to fth ehighl ybranche dpecti nfragmen t 139

released by PL.Thi s is indicated by the observation that the amounts of arabinose and galactose solubilized by PL from the endo-galactanase treated WIS (Table43 )wer e similar to theamount s released by treatment of theWI S with PL (Table39) . The exo-arabinanase,whic h isknow n toprefe r highly branched arabinan asa substrate (Voragen et al. 1986c), solubilize d 38%o f the arabinose present in theWIS . Simultaneous addition of exo-arabinanase and PL indicated that the exo-arabinanase also acts on thehighl y branched PL solubilized fragment. Theadditio n of endo-glucanasean d exo-glucanase to tomatoWI S resulted in the solubilization of 23%o f theWIS-AGA ,whic h iso f highMW . This implies the existence of apecti n fraction connected to cellulose or to hemicellulosepolymer s degradable by glucanases.I tappeare d that this fraction isdegradabl eb y PGbu t only partly by PL (Tables 39,43) ; therefore theD E of this pectinmus t be relatively low.Th e points of attachment may be xylanpolymer s in thehemicellulos e and/or the arabinan polymerswhic hwer e found to exist in the cellulose fraction. In this connection it is interesting tonot e that xylose solubilization is enhanced by the simultaneous action of PE,PG ,endo-glucanas e and exo-glucanase, while arabinose solubilization is enhanced by simultaneous action ofPL , endo-glucanase and exo-glucanase.Th e existence of arabinan chains as connectionsbetwee n pectin and the cellwal lmatri x isals o indicated by theobservatio n that a pure endo-arabinanase was able to release considerable amounts ofAG A from tomatoWI S (A.G.J.Voragen ,persona l communication). Pectin solubilization was found tohav en o synergistic effect on cellulose degradation,whic h is in contrast to the data reported by Voragen etal . (1980)fo r the solubilization of cellulose from appleWI S and AIS.

5.5. Conclusions

The results of the studies on thepolysaccharid e fragments extracted chemically and enzymatically led to the following conclusions regarding the structural features of the tomato cellwall . A largeproportio n of thepecti n occurs as fragmentswit h a low content of 140 arabinose and galactose side chains.Thes e fragments are esterified in such awa y that they are susceptible tobot h PG and PL action.P G degradation results in relatively highly esterified fragmentswit h an apparentM W of 40,000-80,000; degradationwit h PL results in fragmentswit h a lowerD E and anapparen t MW of 10,000-40,000. Approximately 9% of theWIS-AG A formspar t of ahighl y branched pectin segment containing 53%o f the arabinose and 41%o f the galactose present in the totalpecti n (respectively 26%an d 16%o fWIS-arabinos e and WIS-galactose).Thes e fragmentsar e solely linked toth e cellwal l matrix throughhighl y esterified AGA chains,whic h can onlyb e degraded by PL. Endo-galactanase isno t able to degrade the (arabino)galactan present in this fragment;exo-arabinanas e isabl e to release arabinosemolecules . 23% of theWIS-AG A is attached only to thehemicellulos eand/o r cellulose. The points of attachment may be thexylan s present in the hemicellulose fraction or thearabinan swhic h are found in the cellulose fraction.Th e DE of thepecti n in these fragments isprobabl y low.

21% of theWIS-AG A ispresen t in the cellwal l as fragmentswit h an apparent MW of 40,000-80,000,connecte d to the cellwal lmatri x through (arabino)galactans. 7%o f theWIS-AGA ,als oboun d to (arabino)galactans,i s ofmuc hhighe r apparent MW. Since this (arabino)galactan isdegrade d by endo-galactanase, its structure isno t identical to the arabinogalactan present in thehighl y branched pectin fragments described above. Xylans, arabinans and inparticula r (arabino)galactans are involved in the attachment of pectic substances to the cellwal lmatrix .A substantial amount of arabinose and galactose,however ,i spresen t inneutra l sugar chains linked to pectic substances but not to other cellwal lpolymers . 40-50% of the arabinose present inWI S is found as terminally linked residues inhemicellulos e polysaccharides. 35%o f theWIS-arabinos e is connected togalactan s other than those found in thehighl y branched pectin fragment.A n unknownpar t of the arabinose isprobabl y present as slightly branched arabinan chains.

The tomatohemicellulos e contains approximately 80%o f the cellwal l xylose. 60%o f this amount ispresen t in 6-1,4-xylans.Thes exylans , branched with only small amounts of arabinose,ar e probably bound to pectin. 30%o f thehemicellulos e xylosewa s found tob e present in fragmentswit h a composition typical forxyloglucans . 141

Thepresenc e ofglucomannan s and a limited amount of galactan (±7 % of WIS-galactose) in thehemicellulos e was also indicated. Themajo r part of the galactose found in thehemicellulos e (40-50% ofWIS-galactose )wa s present as terminal residues. 142

6. CONTRIBUTION OF WIS CONSTITUENTS TO CROSS VISCOSITY

6.1. Introduction

The results presented in chapter 4 of this thesis show the important contribution of theWI S to the grossviscosit y of tomato products. Gross viscosity at a specific WIS levelwa s not influenced by serum viscosity, variety or concentration method. In chapter 5 the composition and structural features of tomatoWI S were discussed. The tomatovarietie s differed mainly inWI S content and in amounts of arabinose and galactosewithi n thisWIS .Tomat oWI S was fractionated into 40-45% pectin,25-30 % hemicellulosean d 30-35% cellulose. Moreover tomatoWI Swa s found to contain approximately 20%protein ,mainl y present aspar t of thepecti n fraction.Th e amount and composition of the pectin fractions was found to change as a result of processing. The relative contribution ofWI S constituents to the grossviscosit y of tomato products remains unclear. Inorde r to answer this question the effect of technical aswel l aspurified ,wel l characterized enzyme preparations on the grossviscosit y of diluted tomato paste and a 10°B x tomato concentratewa s studied. The data found in the existing literature (Whittenberger &Nuttin g 1957,Smi t &Nortj e 1958,Fod a &McCollu m 1970, Brown & Stein 1977)ar e almost impossible to interpret because of the use of commercial,uncharacterize d enzymes for the selective degradation of tomato constituents contributing to grossviscosity .

6.2. Methods

6.2.1. Enzyme treatment of diluted hot break paste

One part ofvariet y H30ho tbrea k paste (unpasteurized FHB juice concentrated in aR& C T,_ Anteo evaporator to 28-30° Bx,sectio n4.2.1 )wa s thoroughly mixedwit h two parts distilled water. 1500 grams of this dilution (9.4°Bx,1.83 %WIS )wa s transferred into awate r thermostated (40°C),doubl e jacket reactionvesse l as shown in Fig. 38.Enzyme s were 143 addedwhe n theproduc t had reached a temperature of 40°C (+^ hr.). A sample of 150 gramswa s takenbefor e the enzyme addition; other samples were taken after 5, 10,20 ,30 ,60 ,120 ,18 0 and 300minute s of enzyme action.

puree

sample withdrawal

Fig.38 :Reactio nvesse luse dfo r enzymatictreatmen to f dilutedtomat opaste .

All sampleswer e immediately heated inboilin gwate r for 15minute s and then cooled to room temperature.Th e enzyme treated sampleswer e analysed for grossviscosity , serumviscosity , °Bx,WI S content andWI S composition; the resultswer e compared to the results of ablan k treatment. Changes in the appearance of the cellwall s caused by enzyme action were determined by microscopic examination. Table 44 shows the amounts of enzyme added to the substrate inweigh t % (on thebasi s of freshweigh t andWI Sweight ) for the technical enzyme preparations and inunits/k g freshweigh t and units/g WIS for the purified enzymes. Table 45 shows some data on the enzyme activities of the technical enzyme preparations.Purifie d pectolytic enzymes did not show activities other than on their substrates. 144

Table44 :Enzym eaddition st o2: 1dilute dH3 0paste .

freshweigh t WIS

%w/ w U/kg %w/ w U/g

MaxazymCL200 0 0.125 6.8 RapidaseC8 0 0.002 0.11 Maxazym+ Rapidas e 0.125+ 0.00 2 6.8+ 0.1 1 RohamentP 0.01 0.55 ProteaseCG A5647 0 0.1 5.5 PG (Polygalacturonase) 36.7 2.01 PE (Pectinesterase) 18.5 1.01 PL (Pectinlyase) 25.8 1.41 PE+ P G 18.5+ 36.7 1.01+ 2.0 1

MaxazymCL2000 :Gist-Brocades ,Delft ,Th eNetherland s RapidaseC80 :Gist-Brocades ,Seclin ,Franc e RohamentP :R6hm ,Darmstadt ,FR G ProteaseCG A56470 :Ciba-Geig yAG ,Basel ,Switzerlan d

Table45 :Enzym eactivitie so ftechnica lenzym epreparation s(U/g )

enzymeactivit y RapidaseC8 0 MaxazymCL200 0 RohamentP Protease CGA5647 0

Pectinesterase 996 - - Pectinlyase 68 - 1.3 Polygalacturonase 421 - 760 0.44 Exo-glucanase - 39 5.6 Endo-glucanase 118 750 94 Cellobiase 210 32 Arabinanase 44 0.5 Galactanase 38 0.4 Protease + _

1:++ +stron gactivit yo ngelati nlayere dfil mstri p( 5hour s40°C ), + activity,- n oactivity . 145

VarietyH3 0ho tbrea ktomat opast e (28-30°Bx)wa sdilute di nsuc ha waythat ,afte renzym eaction ,th eWI Sconten twa sapproximatel y1% . Theenzyme s(Tabl e46 )wer eadde dt oth eheate d (40°C),dilute dpast ei n thereactio nvesse lan dth emixtur ewa sincubate dfo r7 hours .Th eenzym e treatedsampl ewa sdivide dint o6 portion so f25 0gram san dstore d overnighti na refrigerator .Th enex tda yth esample swer e "WIS-concentrated";0 ,14% ,25% , 35%,41 %an d48 %o fth eweigh twa sremove d asclea rseru mafte rcentrifugatio nfo r3 0minute sa t24,50 0x g .Th e particleswer eresuspende di nth eremainin gseru man dth egros sviscosit y wasmeasure dwit hbot hth eHaak erotovisc oan dth eBostwic kconsistometer . WISwa sprepare di nduplicat efro ma non-centrifuged ,enzym etreate d sample.Th eWI Samount si nth e"WIS-concentrated "sample swer ecalculate d fromthi sdetermine dvalue .

Table46 :Dilutio no ftomat opast ean damount so fenzym eadde dt oth e dilutedpast esamples .

enzymeadditio n enzymetreatmen t gramspast e gramswate r %w/ w units

MaxazymCL200 0 400 1200 0.125 350 1200 0.01 RapidaseC8 0 350 1200 0.002 Maxazym+ Rapidas e 400 1200 0.01+ 0.00 2 RohamentP 350 1200 0.01 ProteaseCG A5647 0 400 1200 0.10 PE 300 1200 36.8 PG 350 1200 11.2 PE+ P G 350 1200 36.8+ 11. 2 PL 350 1200 38.6 Blank 300 1200 - -

TheWI Swa sanalyse dfo rAGA ,degre eo festerificatio no fth eAGA ,protein , cellulosean dneutra lsuga rcontent .Th eexac tprocedur efo rpast edilutio n andenzym eadditio ni sshow ni nTabl e46 . 146

6.2.2. Enzyme treatment of a 10°Bx tomato juice concentrate

1.5kg .o fa 10°B xconcentrat e (1.99%WIS )o fa variet y31 8ho tbrea k juice (hotbrea ktemperatur e93°C ,finishe rscree nsiz e1. 5mm )wa sdivide d intoportion so f10 0grams .0.1 %sodiu mbenzoat ewa sadde da sa preservative.Th esample swer eheate d to40° Ci na wate rbat hbefor eth e appropriateamount so fenzym e (Table47 )wer eadded .

Table47 :Enzym eaddition st o10°B xtomat oconcentrat e (onbasi so fWIS) .

enzymeadditio n

enzymetreatmen t %w/ w U/g

MaxazymCL200 0 6.44 RapidaseC8 0 0.10 Maxazym+ Rapidas e 6.44+ 0.1 0 RohamentP 0.52 UltrazymM10 * 0.13 ProteaseCG A5647 0 4.26 PG 1.91 PE 1.03 PE+ P G 0.85+ 0.7 2 PL 1.29

*Ultrazy mM1 0 (NovoFermen tAG ,Basel ,Switzerland )i scomparabl et o RohamentP bu tapproximatel y4 time sa sconcentrated .

Exactly2 minute safte raddin gth eenzym ea continuou sgros sviscosit y measurementwa sstarte di nth eHaak erotovisc ousin grotar ybo bMVII Ia ta speedo f5 4rp man da ta temperatur eo f40°C .Th emeasuremen twa sstoppe d after5 hours ;th eenzym etreate dsampl ewa sstore dovernigh ti na refrigeratorbefor e°Bx ,seru mviscosit yan d% WI Swer edetermined .Th eWI S wasanalyse d forAGA ,D Eo fAGA ,cellulose ,protei nan dneutra lsuga r content. 147

6.3. Results and discussion

6.3.1. Effect of enzyme treatment on gross viscosity and composition of diluted hot break paste

Table4 8show sdat ao ngros san dseru mviscosit yan dcompositio no f the2: 1dilute dtomat opast euse da sa substrat efo rth eenzymes .Th edat a area naverag eo f1 1individua ldeterminations ,on efo reac henzym e treatment.Th egros sviscosit yo fthi sdilute dpast esampl ei sonl y70 %o f thegros sviscosit yo fa tomat ojuic eafte rconcentratio nt o1.83 %WI S (compareFig .24 ,chapte r 4).Th ephenomeno no fa los si ngros sviscosit y asa resul to fconcentratio nan dsubsequen tdilutio n ("dilution-loss")wil l bediscusse dextensivel y inchapte r7 .

Table48 :Gros sviscosity ,seru mviscosit yan dcompositio no fdilute d tomatopast eo fvariet yH30 .

n (mPa.s) 423.1 + 25.5 app (mPa.s) 8.2 + 0.3 n °Bx 9.4 + 0.1 serum WIS (mg/100g ) 1830 + 33 AGA (mg/100g ) 230 + 14 DE 34 + 3 Protein (mg/100g ) 379 + 12 Cellulose (mg/100g ) 578 + 26 Rhamnose (mg/100g ) 11.0 + 1.0 Arabinose (mg/100g ) 22.9 + 1.4 Xylose (mg/100g ) 63.9 + 5.6 Mannose (mg/100g ) 51.5 + 4.1 Galactose (mg/100g ) 52.3 + 3.6 Glucose (mg/100g ) 536.8 + 40.0

Figure3 9show sth eeffect so fenzym eadditio no nth egros sviscosit y ofth edilute dtomat opaste .I ti sapparen tfro mthes efigure stha tal lth e 148

enzymes or combinations of enzymes,excep t PE,ar e able to decrease gross viscosity.

Voresidua l gross viscosity % residual gross viscosity o A

* Blank • RapidaseC o Maxazym CL200 0 • Rap.* Max. t Protease CGA 56470

3 4 5 3 4 F incubation time (hr) incubation time (hr)

Fig.39 :Chang ei ngros sviscosit y(Haake )o fdilute dtomat opast ea sa resul to ftreatmen t withenzymes .

The largest decrease ingros sviscosit y (86%)i sobtaine d with a combination of a technical pectolytic (Rapidase C80)an d cellulolytic (Maxazym CL2000) enzyme preparation. The role of pectin in determining grossviscosit y is indicated by the action of technical and purified pectolytic enzyme preparations.A reduction of 35% to 60%i n gross viscosity canb e obtained by degrading certain parts of theWI S pectin. The role of cellulose in determining grossviscosit y must not be underestimated. Fig. 39 clearly shows that degradation of cellulose may result in a grossviscosit y loss as great as that obtained with the pectolytic enzymeRapidas e C80.Th e smallest effects on grossviscosit y were observed with PE and protease. PE deesterifies highly esterified pectinwithou t affecting the glycosidic bonds in thepecti n chain. Therefore onewoul d not expect a large change in grossviscosit y asa result of PE enzyme action.Althoug h theWI S containsmor e than 20%protei n (1.5 times theAG A content) the results dono t indicate that protein makes an important contribution to gross viscosity. 149

Some interesting initial effects are observed when pectolytic enzymes are added to the diluted tomato paste (Fig.39) .A sharp initial decrease in grossviscosit y is found after the addition ofP G and Rohament P; the addition of PL,P Ean d PE+ PG results ina n initial increase in gross viscosity which reaches amaximu m 20minute s after the addition of the enzyme.Afte r the initial decrease,th e grossviscosit y of PG treated diluted tomato paste increases to amaximu m after 30minute s of enzyme incubation. The chemical changes inWI S accompanying the sharp initial decrease in grossviscosit y caused by PG and Rohament P action are presented in Table 49.

Table49 :Decreas ei ngros sviscosity ,i nWI Sconten tan di namoun to fWI Sconstituent safte r 10min .enzym eactio n(i n %).

n WIS AGA DE Protein Cellulose Rham. Ara. Xyl. Man. Gal. Glue, app

RohamentP 22 7 39 48 n.s. n.s. 30 17 n.s. n.s. 19 n.s. PG 20 2 27 49 n.s. n.s. n.s. n.s. n.s. n.s. 13 n.s.

1:calculate dD Eo fsolubilize dAGA . n.s.= no tsignificant .

30-40% of theAG A present inWI S is rapidly solubilized by the action of PG (purified or inRohamen t P), resulting in agros sviscosit y loss of up to 22%. The loss inWI S is completely covered by the solubilization of AGA with a degree of esterification of 48-49%.Result s presented in the preceeding chapter showed that incubation ofWI Swit h a purified PG resulted in the solubilization of pectin fragmentswit h an apparent MW of 40,000-80,000,containin g 24%o f theAG A present inWI S (DE=58). The solubilization of these pectin fragments obviously caused the rapid decrease in grossviscosit y of the diluted tomato paste found after the addition of PG orRohamen t P. The addition of PE+ PG or PL resulted only in the solubilization of 10-12% ofAG A in the first 10minute s of enzyme action. 150

Table 50 shows the chemical changes in theWI S accompanying the initial increase ingros sviscosit y as a result of the addition of PL,P E and PE + PG. The results for theadditio n of PG are also included because this enzyme causes agros sviscosit y increase after the initial decrease. The gain ingros sviscosit y asmeasure d after 20minute s of enzyme action may be ashig h as 35% (PE+ PG). Itwa s observed that gross viscosity increased simultaneously withWIS ,xylose ,mannose ,glucos e and cellulose as determined by theUpdegraf f method.

Table50 :Change si ngros sviscosity ,i nWI Sconten tan di nWI Scompositio nafte r2 0min .(PE , PE+ PG ,PL )an d3 0min .(PG )enzym eactio n(i n %).

n WIS AGA Protein Cellulose Rham. Ara. Xyl. Man. Gal. Glue. app

PE +22 +11 +8 +9 +18 +15 +18 +11 +26 n.s. +12

PE+ P G +35 +11 -13 +9 +21 n.s. n.s. +25 +11 +19 +22

PL +23 +6 -13 +6 +11 n.s. n.s. +15 +19 n.s. +14

PG -8 * +2 -26 n.s. +6 n.s. n.s. +15 +14 n.s. +6

*wa s-20 %afte r1 0min .incubatio ntime , n.s.:no tsignificant .

Moreover,th e amounts of solubilized AGA did not increase between 5 and 20 minutes of enzyme incubationwit h PE,P E+ PG and PL;n o increases in solubilized AGAwer e found between 5 and 60minute s of incubationwit h PG. The increase inWI S content may be attributed to the formation of aggregates as a result of thehea t treatment necessary for the inactivation of the enzymes. Inanothe r experiment itwa s observed that theheatin g of PE+ PG treated serum resulted in the formation of a precipitate;n o precipitationwa s found after theheatin g of untreated serum. The increase in grossviscosit y may be caused by apartia lnegatio n of the "dilution-loss" (see 7.3.3); no increase ingros sviscosit y was detected when a 10° Bx tomato concentratewa s treatedwit h enzymes (see6.3.2. ,Fig . 42). After 5hour s of enzyme actionbot hRohamen tP and PG had degraded large parts of theWI S pectin,causin g adecreas e ingros sviscosit y of 151

40-50% (Table 51). No degradation of other cell wall polymers was detected.

Table 51: Decrease in gross viscosity, in WIS content and in amount of WIS constituents as a result of enzyme action for 5 hours on a 2:1 diluted tomato paste (in %).

Enzyme WIS AGA DE3 Rham. Ara. Gal. Cellulose Xyl. Han. Protein napp

RohamentP 54 15 61 4 57 21 26 5 n.s. n.s. n.s.

PG 40 12 43 20 22 20 20 5 n.s. n.s. n.s.

PE1 n.s. n.s. n.s. 25 n.s. 11 16 n.s. n.s. n.s. 7

PE+ P G 47 12 57 0 47 35 22 5 n.s. n.s. n.s.

PL 35 12 42 8 39 29 23 8 n.s. n.s. n.s.

RapidaseC8 0 57 18 62 0 68 43 38 n.s. 11 n.s. 5

MaxazymCL200 0 61 24 11 36 24 35 21 44 52 53 n.s. 2 Rap.+ Max . 86 39 64 2 73 62 47 39 48 36 11

Protease 25 16 18 32 29 34 11 15 30

1: PE reduced DE of soluble AGAfro m 51 to 18. 2: Determination of neutral sugar content after 180 min. instead of 300 min. 3: DEo f unsolubilizedAG A n.s. not significant

The differences found between the effects of Rohament P and purified PG are probably due to a difference in the amount of units of PG added to the diluted tomato paste. Both enzymes released the major part of the total solubilized AGA in the initial stage of the reaction; smaller amounts were released as result of the subsequent period of enzyme action (Table 52). The DE of the AGA solubilized after the initial stage is higher than that of the pectin solubilized initially. From the results it can be concluded that PG solubilizes pectin with an overall DE of 50-60% (DE of total WIS pectin = 34%). The solubilization of this type of pectin greatly influences the gross viscosity of the diluted tomato paste. Although the total effect of PL on gross viscosity is comparable to the effects of PG and Rohament P, the mechanism of PL action is quite 152

different. PL solubilizes 10% of the AGA with a DE of 94 in the first 10 minutes of enzyme incubation (Table 52).

Table 52: Effect of pectolytic enzymes on pectin degradation and gross viscosity loss (in %).

AGA decrease DE solubilized AGA gross viscosity loss

0-10 min. 10-300 min. 0-10 min. 10-300 min. 0-10 min. 10-300 min.

RohamentP 38 23 48 66 22 32 PG 27 17 49 52 20 20 PL 10 31 94 50 rease1 4 35

1D Eincrease st o6 4afte r1 hou ro fenzym e incubation. 2D Edecrease st o3 8afte r1 hou ro fenzym e incubation. 3calculated .

This did not result in a decrease in grossviscosity ,presumabl y due to the simultaneous increase inWI S (Table 50)an d to the partialnegatio n of the "dilution-loss".Afte r 1hou r of enzyme action,whe n 21%o f theAG A (average DE=83)ha d been solubilized, the grossviscosit y returned to the same level as found at the start of the incubation. Between the 2nd and the 5thhou r of enzyme action another 21%o f theAG A (averageDE=38 )wa s solubilized; the totalAG A solubilization of 42% (averageDE=62 ) finally resulted in a grossviscosit y loss of 35%.A swil l be shown in 6.3.2., the effect of PL on grossviscosit y depends on theD E of the totalWI S pectin. In chapter 5 pectin fragments susceptible todegradatio n by both PL and PGwer e identified. The results presented in this chapter indicate that these fragments contribute to agrea t extent togros sviscosity . The results obtained by the simultaneous addition of PE and PG and by the addition of Rapidase C80 confirm the important contribution of the higher esterified parts in theWI S pectin togros sviscosity . The contribution of the lower esterified parts in theWI S pectin,however ,remain sunclear . From theresult s obtained after theadditio n ofRohamen t P it canb e concluded that approximately 40%o f theWI S pectinha s a low degree of 153

esterification (Table 51).Unfortunatel y the added enzymeswer e not able to solubilize this type of pectinunde r the reaction conditions used.Th e fact that even PG isunabl e to solubilize this 40%o f low esterified pectinma y be attributed to ahig h degree of acetylation of thispecti n or to an inaccessibility of thispecti n for theenzymes . The grossviscosit y decrease as a result of theadditio n of Maxazym CL2000 is comparable to the decrease obtained with Rapidase C80 (Table51) . Inorde r toreac h this decrease,Maxazy m CL2000need s to degrade 24%o f the WIS compared to 18%fo rRapidas e C80.Cellulos e degradation accounts for almost 60%o f theWI S solubilization found afterMaxazy m CL2000 incubation. This enzyme degraded not only cellulosebu t also hemicellulose (xylose, mannose)an d parts of theWI S pectin connected tohemicellulos e and/or cellulose. In the first hour of enzyme incubation no solubilization of pectinwa s found. Cellulose,however ,solubilize d for 20% (=6.3 % ofWIS) , xylose and mannose for 25% (= 1.4% of WIS), and grossviscosit y decreased by 36%.I t seems likely that the decrease in grossviscosit y as a result of Maxazym CL2000 action ismainl y caused by thedegradatio n of cellulose. The combined action of Rapidase C80 and Maxazym CL2000 greatly decreased grossviscosity . The cell structurewa s severely attacked aswa s shown by microscopic examination (Fig.40) .Rapidas e C80 alone,lik e the other pectolytic enzymes,ha d no distinct effect on the microscopic appearence of the cells;Maxazy m CL2000 treated cellsha d thinner walls when compared tountreate d cells,bu twer e unbroken.

Fig.40 :Effec to fenzyme so nth estructur eo ftomat ocell s a:control ,b :Maxazy mCL2000 ,c :Rapidas eC8 0+ MaxazymCL200 0(magnificatio n61.4x) . 154

The solubilization of 30%o f theprotei n as a result of the action of protease resulted ina grossviscosit y loss of 25% (Table 51).I tis , however,questionabl e whether this grossviscosit y loss is caused by protein solubilization orb y the solubilization of 18%AG A and 8% cellulose.Th e role ofprotei n as a contributor to grossviscosit y ismor e clearly shown inFig .41 .

n Blank D Blank 1 Maxazym CL200 00.01 % • PL Maxazym CL 2000 0.125% » PE o Rapidase C 80 » PG cm • Max.*Ra p cm x PE'PG 3i • Rohament P 3 4 Protease CGA 56470 2- .^»

10 10

\

2- 2-

-i—i | i i—i—i—i—i—i—i—i—i—n— n—i—i—i—i- 1.0 2.0 1.0 2.0 %WIS %WIS

Fig.41 :"WIS-concentration "curve s(Bostwick )o fenzym etreated ,dilute dtomat opaste .

This figure shows the"WIS-concentration " curves of samples of diluted tomato paste incubated with enzymes for 7hours .Th e slope and intercept of these straight lines indicate thepotentia l of enzymatically modified WIS to"build " grossviscosit y (WIS-quality). Similar curveswer e obtained after themeasuremen t of grossviscosit y with theHaak e rotovisco instead of the Bostwick consistometer. The incubation of the diluted pastewit h protease resulted in the solubilization of 50%o f theprotein ,22 %o f theAG A and 8% of the cellulose present inWIS .Th e solubilized protein obviously does not contribute to grossviscosity . 155

Proteasetreate ddilute dpast econcentrate d toa specifi cWI Sconten tshow s increasedgros sviscosit ywhe ncompare dt ountreate ddilute dpast ea tth e sameWI Sconcentration .I tmus tb erealize dtha ta tthi sWI Sconcentratio n theproteas etreate dpast econtain smor etomat ocell san di stherefor e enrichedi nth econstituent sdeterminin ggros sviscosity . Anincreas ei ngros sviscosit ya ta specifi cWI Samoun twa sals ofoun dwhe n thedilute dpast ewa sincubate dwit hPE .Du et oth eincreas ei ncalciu m reactiveCOOH-groups ,th eP Etreate ddilute dpast etransforme dint oa ge l uponstorage . Thepoores tWIS-qualit ya t1 %WI Swa sobtaine dafte rtreatmen to fth e dilutedpast ewit hRapidas eC80 ,o rwit hamount so fpurifie dP Ean dP G comparablet oth eamount spresen ti nRapidas eC80 .(Fig . 41). Theincubatio no fth edilute dpast ewit h0.125 %Maxazy mCL200 0resulte di n thesolubilizatio no f78 %o fth ecellulose ,30 %o fth eAG Aan d11 %o fth e proteinpresen ti nth eWIS ;moreove r80-90 %o fth eWIS-xylos ean d WIS-mannosewa sdegraded .Th eWI Sconcentratio ncurv eo fthi senzym e treated,dilute dpast eshow sa completel ydifferen tslop etha nth eothe r curves(Fig .41) . Atth elowe rWI Samount s (+1% )th elos si ncellulos e isapparentl ycompensate d forb yrelativel yhig hamount so ftomat ocell s andtherefor eo fAGA .A tth ehighe rWI Samount sth edecrease drigidit y ofth ecel lwalls ,cause db yth elos si ncellulose ,seem st obecom e increasinglyimportan ti nmaintainin ggros sviscosit y (Table53) .Shotne r (1984)showe dtha ta partia ldegradatio no fth emicrofibrilla rcellulos e structureb y0.2 %o fa nunspecifie d cellulaseresulte di na collaps eo fth e cellwal lprecipitat eunde rgravit y stress.Thi slos si nabilit yt o withstand collapseobviousl yals oaffect sth egros sviscosity . Theincubatio no fdilute dtomat opast ewit hRapidas eC80 ,P E+ P Gan d RohamentP showe dtha tapproximatel y40 %o fth eWIS-AG Acanno tb e solubilizedb ythes eenzyme s(Tabl e51) .Moreover ,result sobtaine dafte r theadditio no fRohamen tP (noP Eactivity )indicate d thatthi s unsolubilizedpecti ni salmos tunesterified .I tca nb eassume dtha tP Lan d PGwil lals ob eunabl et osolubiliz ethi styp eo fpectin .Therefor ei tca n becalculate dfro mth eresult sshow ni nTabl e53 ,tha tth eWI Si nth eP G treatedtomat opast econtain sa certai namoun to fhighl yesterifie dpectin ; thisi ncontras tt oth eWI So fP Ltreate dtomat opaste .Thi samoun to f highlyesterifie dpecti n (44mg/10 0g )seem st ob eth ereaso nfo rth e 156 highergros sviscosit yo fth eP Gtreate dsampl ewhe ncompare dt oth eP L treatedsampl e (Table53) .

Table53 :Gros sviscosit yan dcompositio n (mg/100g )o fenzym etreate d dilutedtomat opast ea t2 %WIS .

Enzyme Bostwickc m AGA DE Cellulose Protein

MaxazymCL200 00 125% 9.4 357 29 263 644 PE+ P G 7.7 114 0 743 509 RapidaseC8 0 7.6 113 0 717 471 Max.0.01 %+ Ra p 6.8 131 0 709 526 RohamentP 6.8 110 0 783 480 PL 6.7 147* 4* 718 503 MaxazymCL200 00 01% 6.5 264 35 632 454 PG 6.1 143** 19** 674 470 Blank 5.8 286 32 660 404 PE 4.9 259 15 628 450 Protease 3.7 274 34 745 254

* probably9 8m gwit hD E0 an d4 9m gwit hD E1 2(calculate dfro m RohamentP ) **probabl y9 9m gwit hD E0 an d4 4m gwit hD E6 4 (calculatedfro m RohamentP )

Itmus tb erealize dtha tth edifference si nWI Scompositio na scause d byth eadditio no fenzyme sar emuc hlarge rtha nth edifference swhic hoccu r naturally intomat oWIS . Differencesi ngros sviscosit yar ecause drathe rb ydifference si nWI S amount (amounto ftomat ocells )tha nb yth esmal ldifference si nWI S composition.Th esmal ldifference si ngros sviscosit y founda ta specifi c WISamoun tma yb eexplaine db ya differenc ei nconten tan dcharacteristic s ofAG Ao rb ya differenc ei nrigidit yo fth ecel lwalls . 157

6.3.2. Effect of enzyme treatment on gross viscosity and composition of a W°Bx tomato juice concentrate

A10°B xtomat ojuic econcentrat ewa sals osubjecte d toenzym etreatment . Itscompositio n (Table54 )i scomparabl et oth ecompositio no fth edilute d paste (Table48) ,excep tfo rcellulos econten tan dD Eo fth epectin .A majordifferenc ebetwee nth etw osubstrate si sth eleve lo fth egros s viscosity.Th ehighe rgros sviscosit yo fth e10°B xconcentrat eca nb e attributed toth eabsenc eo fa "dilution-loss" .

Table54 :Gros sviscosity ,seru mviscosit yan dcompositio n ofa 10°B xconcentrat eo fvariet y318 .

(mPa.s) 614.7 ^app (mPa.s) 8.6 1serum °Bx 9.9 WIS (mg/100g ) 1990 AGA (mg/100g ) 245 DE 44 Protein (mg/100g ) 373 Cellulose (mg/100g ) 663 Rhamnose (mg/100g ) 12.7 Arabinose (mg/100g ) 27.5 Xylose (mg/100g ) 77.0 Mannose (mg/100g ) 51.7 Galactose (mg/100g ) 66.5 Glucose (mg/100g ) 571.5

Changesi ngros sviscosit yan di nth ecompositio no fth etomat o concentratea sa resul to fincubatio nwit hth eenzyme sar epresente di n Table5 5an dFig .42 .Th ecurve sshow ni nFig .4 2ar ecorrecte dfo ra blan k treatment.I ngeneral ,th eadditio no fenzyme st oth e10°B xtomat o concentratele dt oth esam efina lresult sa sobtaine db yth eadditio no f enzymest oth edilute dpaste .A shar pinitia ldecreas ei ngros sviscosit y wasagai nfoun dafte rincubatio nwit hRohamen tP ,P Gan dals owit h 158

UltrazymM10 .Th e initial decreases as a result of Rapidase C80,P L and PE

+ PG actionwer e smaller; the final effects,however ,wer emuc h greater.

% residual gross viscosity % residual gross viscosity 100 100-te°-°—°- • [

60-

o Rohament' P a PE A Ultrazym M10 0 PG * Rapidase C • PL 20- • Maxazym CL 2000 1 PE*PG D Rap. C8 0+ Max. CL 2000 » Protease CGA 56470

4 5 0 4 5 time (hr) time(h r]

Fig.42 :Change s ingros s viscosity (Haake)o fa 10°B x tomato concentrate asa resul to f treatment with enzymes.

Table 55: Reduction ingros s viscosity, inWI Sconten t andi namoun t ofWI Sconstituent sa s

result ofenzym e action for5 hour so na 10°B x tomato concentrate (in %).

1 n WIS AGA DE Rham. Ara. Gal. Cellulose Xyl. Man. Protein app

RohamentP 45 15 60 20 52 32 29 n.s. n.s. +22 7

UltrazymM1 0 54 21 70 6 72 51 38 n.s. n.s. +16 9

PG 36 13 47 35 21 17 22 9 n.s. n.s. 6

PE +3 +3 +6 25 n.s. n.s. n.s. 6 n.s. n.s. 5

PE+ P G 67 14 63 5 44 35 31 5 n.s. n.s. n.s.

PL 54 14 53 18 36 30 25 8 n.s. n.s. 5

RapidaseC8 0 68 20 70 0 80 53 46 5 n.s. +16 6

Maxazym CL200 0 68 36 25 43 25 29 32 67 71 76 8

Rap.+ Max . 87 45 70 0 89 70 68 62 62 61 10

Protease CGA 5647 24 12 17 42 11 18 35

1:D Eo funsolubilize dAG A

n.s.:no t significant 159

Thelarges tgros sviscosit y losswa sobtaine dwit hth ecombinatio nRapidas e C80an dMaxazy mCL2000 ;th esmalles tlos si ngros sviscosit ywit hProteas e CGA56470 .Th eadditio no fP Et oth etomat oconcentrat eha dn oinfluenc eo n grossviscosity . Someinterestin gdifference si nth eeffec to fenzyme so nth egros s viscosityo fth etw osubstrate swer enoticed .I ncontras tt oresult s obtainedfo rth edilute dpaste ,n oincreas ei ngros sviscosit ywa sfoun d afterth eadditio no fPL ,P Ean dP E+ P Gt oth etomat oconcentrat e(Fig . 42).Thi sdifferenc ema yb ecause db yth eabsenc eo fa "dilution-loss "i n the10°B xtomat oconcentrat e (see6.3.1. ,7.3.3.) . Pectolyticenzyme scapabl eo fdegradin gesterifie dpecti n (PL,P E+ PG , RapidaseC80 )hav ea greate rinfluenc eo ngros sviscosit yo fth e10°B x concentrate (DEo fth eAG A= 44) ,tha no ngros sviscosit yo fth edilute d paste (DEo fth eAG A= 34).PL ,fo rexample ,solubilize d53 %o fth eWIS-AG A presenti nth e10°B xconcentrate ,wit ha noveral lD Eo f68% ,resultin gi na grossviscosit y losso f54% .42 %o fWI SAG A (DE=62)wa ssolubilize dfro m thedilute dpaste ,resultin gi na gros sviscosit ylos so f35% .

6.1. Conclusions

Allresults ,irrespectiv eo fth etomat oproduc tused ,poin tt oa n importantrol efo rth emor ehighl yesterifie dpart so fth eWI Specti ni n contributingt ogros sviscosity .I ti slikel ytha tthes epectin sar e identicalt oth epectin ssusceptibl et odegradatio nb ybot hP Lan dP Ga s describedi nchapte r5 .Th edegradatio no fthes epectin sb yP Gresult si n theimmediat esolubilizatio no ffragment swit ha hig hM Wan da hig hDE , givingris et oa rapi ddecreas ei ngros sviscosity .P Linitiate sth e degradationo fthes ehighl yesterifie d fragments;pecti nsolubilize smor e slowlyan dtherefor egros sviscosit ydecrease sgradually . Theimportanc eo fth elo westerifie dpecti nremain sunclear ;non eo fth e enzymeswa scapabl eo fselectivel ysolubilizin g thistyp eo fpectin .Th e samehold sfo rth econtributio no fhemicellulos epolymer st ogros s viscosity.Th erol eo fth eprotei npresen ti nth eWI Si sprobabl y negligible.Cellulos eseem st ob ea nimportan tWI Sconstituen twit hregar d togros sviscosity .It srelativ econtributio nt ogros sviscosit yi s 160

smaller than the contribution of the esterified pectin; solubilization of 67% of the cellulose (22.3% of theWIS ) resulted in the same gross viscosity loss as solubilization of 70%o f theAG A (8.6%o f the WIS). The rigidity of the cellwalls ,cause d by thepresenc e of cellulosic structures, seems to influence grossviscosit y athighe r WIS concentrations. Natural differences inWI S composition,whic h aremuc h smaller than the differences brought about by theadditio n of enzymes,canno t cause large differences in grossviscosity . Grossviscosit y isprimaril y determined by the amount ofWI S (amount of tomato cells). The presence of large amounts ofhig h ester pectin in theWI Sma y have some influence on grossviscosit y at a specificWI S amount.Th e rigidity of the cellwall sma y influence grossviscosit y at higher WIS concentrations. 161

INFLUENCE OF PROCESSING FACTORS ON THE WIS CONTENT, WIS COMPOSITION AND CROSS VISCOSITY OF TOMATO JUICE AND CONCENTRATES

7.7 Introduction

The grossviscosit y oftomat o juicean dconcentrate s isdetermine d mainlyb yit sWI Scontent ; serumviscosit y doesno tcontribut e togros s viscosity (chapter 4). Thedifference s observed between varieties were explained interm so fa differen tWI Scontent ; large differences inWI S compositionwer eno tfoun d (chapter 5). Data showni nchapte r6 indicate the importanceo fmor e highly esterified partso fth eWI Specti nan do f cellulose ascontributor s togros sviscosity . Cellulosic structures also determine therigidit y ofth ecel lwalls .Thi s rigidity seemst o influence grossviscosit y athighe rWI Sconcentrations .Th eamoun tan d compositiono fth epecti n fractionswer e found tochang ea sa resul to f processing (chapter5) . This chapter dealswit h theinfluenc eo fprocessin g factors sucha sho t break temperature,finishin g operation,concentratio n andpasteurizatio no n theWI Scharacteristic s andgros sviscosit y oftomat o juicean d concentrates.

7.2. Methods

7.2.1. Production scale equipment used to produce samples for studying the effects of processing conditions on gross viscosity

The samplesuse d tostud y theinfluenc eo fprocessin g conditionso n thegros sviscosit y oftomat o juicean dconcentrate swer e produced mainly ina Portugues e factory inth eregio no fLisbon . Samplesprepare da t differentho tbrea k temperatureswer e produced both inthi s factory aswel l asi na nItalia n factory inth eregio no fParma .Th eequipmen t usedi n thesetw ofactorie si sliste d below. 162

Portugal hotbreak : Rossi& Catell i100 0L recirculatin gchop/scalder .(se e section4.2.1.) • finishing: Rossi& Catell ithre estag epaddl efinishe r13 3b-14 . (capacity 14tons/hr.) . evaporation:Ross i& Catell idoubl eeffec tT_ .D.F.F .Ante oevaporator , (seesectio n4.2.2.) .

Italy hotbreak : Rossi& Catell i"Eldorado "scalde r (Fig.44) .Thi sscalde r isbasicall y composedo fa vacuu mtan kan da pressurize d hotsectio ntan kcouple dt oa tubula rhea texchanger .Du e toth epresenc eo fth epressurize dsection ,ho tbrea k temperaturesi nexces so f100° Cca nb eobtained .Th e capacityo fthi sscalde ri s3 5tons/hr . finishing: Rossi& Catell itw ostag e"Butterfly "finisher ,maximu m capacity2 0tons/hr . evaporation:Ross i& Catell idoubl eeffec tT6 0Califf oevaporator .Thi s evaporatorproduce s± 5 ton s23°B xpaste/hr .a ta n evaporationrat eo f22,00 0k gwater/hr .Produc t temperatureswer e68° C (54c mH gvacuum )i nth efirs t effectan d46° C (68.5c mH gvacuum )i nth esecon deffect .

^ exchanger h u M o

Fig.44 :Th e Rossi& Catell i "Eldorado"scalder . 163

7.2.2. Production of tomato juice and paste at various hot break temperatures

Tomatojuice sprepare da tho tbrea ktemperature srangin gfro m90° Ct o 103°Cwer eproduce di nPortuga li nth e198 3an d198 4seasons .Th ebrea k temperatureswer erecorde da tth eoutle to fth etubula rheate ran d thereforerepresen tfina lproduc ttemperatures .Th etemperature smeasure d inth echoppe rwere ,i ngeneral ,4-6° Clower . Screenopening suse di nth epulper ,refine ran dsuperrefine runit so fth e finisherwer erespectivel y 1.0mm ,1. 0m man d0. 8mm .Juic esample swer e heatprocesse dfo r4 5min .a t100°C . In198 4an d198 5tomat opast ewa sproduce di nItal yfro mtomat ojuic e prepareda tho tbrea ktemperature srangin gfro m90° Ct o115° C(fina l producttemperature) .I norde rt ominimiz eth eresidenc etim eo fth ebroke n tomatoesi nth evacuu mchambe ro fth eEldorad oscalde r (enzymeaction )th e levelo ftomatoe si nthi stan kwa skep ta t1- 2tons . Screenopening suse di nth eButterfl ypulper/finishe rwer erespectivel y1. 2 mm- 0. 6m mi n198 4an d0. 8m m- 0. 8m mi n1985 .Th etomat ojuic ewa s concentratedt oa leve lo f23°Bx ,sterilize dfo r2 min .1 5sec .a t109° C andthe npacke dasepticall yi nmetalize dfil mbags .Th esamplin go fth e tomatopast ewa sstarte dn oearlie rtha n4 hour safte ra chang ei nho t breaktemperature .Eac hexperimen twa scontinue dfo ra tleas t1 0hours .

7.2.3. Production of tomato juice and paste using different finishers and finisher conditions

Inorde rt ostud yth eeffect so ftyp eo ffinisher ,finishin g temperaturean dfinishe rscree nopenin go ngros sviscosity ,th efollowin g sampleswer eprepared . Tomatoeso fth evariet yH3 0wer ecrushe da tho tbrea ktemperature so f90° C and103°C .A par to fth ecrushe dtomatoe swa sprocesse dusin gth ethre e stagepaddl efinishe r (1.0m m- 1. 0m m- 0. 8m mscreens) .Th ejuic ewa s cannedan dhea tprocesse dfo r4 5min .a t100°C .Anothe rpar to fth ecrushe d tomatoeswa sfinished ,eithe ra troo mtemperatur eo rafte rreheatin gt o 90°C,usin ga laborator yBauknech tK U2- 1tapere dscre wjuic eextracto r equippedwit ha 0. 9m mscree n (batchsiz e2. 4kg) .Thes ejuice swer eno t heatprocessed . 164

Tomatojuic ewa sprepare d fromtomatoe so fth evariet yH3 0a ta ho t breaktemperatur eo f103°C .Differen tcombination so fscreen swer euse dt o finishth ecrushe dtomatoes .(Tabl e57) .

Table57 :Scree ncombination suse dfo rfinishin gho tbrea kcrushe d tomatoes.

pulper refiner superrefiner

1.5m m 1.5m m - 1.5m m 1.5m m 1.0m m 1.5m m 1.5m m 0.8m m

Thejuic ewa sconcentrate d to29°B xan dpacke dasepticall y inmetalize d bags. 5.5k go ffres htomatoe so fth evariet yH3 0wer eheate d (20min . 120°C)i nclose dsterilizatio nbag si na retor tt oa hear ttemperatur eo f 90°C. Theho ttomatoe swer efinishe dusin gth etapere dscre wjuic e extractorequippe dwit ha 0. 5m mo ra 1. 5• screen.Th ejuice swer eno t heatprocessed .

7.2.H. Concentration of tomato juice

Preparation of juice samples Allth ejuice suse dt ostud yth einfluenc eo fth econcentratio n processo nth egros sviscosit yo fconcentrate san ddilute dconcentrate s wereproduce di nPortuga la ta ho tbrea ktemperatur eo f103°C .Th epulper , refineran dsuperrefine rscree nopening suse dt ofinis hth ecrushe d tomatoeswer erespectivel y 1.0mm ,1. 0m man d0. 8mm .Th ejuice swer e cannedan dhea tprocesse d for4 5min .a t10 0°C ,unles sthe ywer euse d directlyfo rconcentratio ni nth eT3 0doubl eeffec tevaporator ,o rfo r separationi na decante rcentrifuge . 165

Concentration in an industrial evaporator Tomato juicewa s concentrated in aRoss i & Catelli T30D.F.F . Anteo evaporator (see section4.2.2. )t o solids levels of 10,15 ,20 ,2 5 and 30°Bx. The concentrateswer e heat processed for 45min . at 100°C.

Concentration in a laboratory evaporator A Buchirotar y film evaporatorwa suse d to concentrate juice samples on laboratory scale.Fo r details see section 4.2.2. Concentrates were heat processed for 15min .a t 100°C.

Concentration by evaporation under atmospheric conditions Juicewa s concentrated ina beake r at a temperature of approximately 100°C.Th e juicewa s stirred continuously during concentration.

Water removal by gel filtration media A dialysisba g (cellulose acetate) filled with tomato juice was dredged with CM-Sephadex (cation-exchanger with sodium as counter-ion, Pharmacia Fine Chemicals,Uppsala ,Sweden )an d placed in a dessicator at room temperature.Th e amount of CM-Sephadexwa s 10%o f the juiceweight . The concentration processwa s continued for two days.

Centrifugation-serum-concentration

Laboratory 6 juice samples of 250 gramwer e centrifuged for 45min .a t 25,400 xg in a Sorvall RC-5B refrigerated superspeed centrifuge. The clear serum,6 x 200 gram,wa s collected and concentrated in a Buchi rotary film evaporator toapproximatel y 30°Bx; the condensatewa s collected. Equal amounts of serum concentratewer e added to the fibres in the centrifuge tubes.T o each centrifuge tube different amounts of condensatewer e added. The fibres, serum concentrate and evaporated waterwer e thoroughly mixed.

Pilot plant Tomato juice aseptically packed inbag swa s centrifuged at room temperatureusin g a Sharpless P600 Super-D-canter centrifuge at a capacity of 200kg/hour .Th ebow l speed of thedecante r centrifugewa s 6000 rpm 166

(=3000x g), thespee ddifferenc ebetwee nscre wan ddru mwa s4 0rpm ,th e ringda mwa sse ta tth epositio nneares tt oth escre wshaft .Th ejuic ewa s separated into25 %fibre san d75 %serum .Th eseru mwa sconcentrate dt o 30-35°Bxusin ga vacuu mpan .Fibre san dconcentrate d serumwer erecombined .

Factory In198 4a Westfali aC A220.01 0decante rcentrifug ewa suse dfo rth e centrifugationo ftomat ojuic ea ta temperatur eo f90-92° Can da capacit y of2 tons/hr .Th ebow lspee do fth edecante rwa s500 0rp m (=3100x g),th e ringda mwa sse ta tth epositio nneares tt oth escre wshaft .Th ejuic ewa s separatedi n77 %seru man d23 %fibre .Th eseru mwa sconcentrate di na stirredvacuu mpa n (vacuum6 6c mHg ,temperatur e52°C ,190 0k gwate r evaporated/hr.)t o55-60°Bx .Th econcentrate d serumwa spacke di n5 k gcan s at92°C ;th efibre sa t94-95°C .

Concentration of dialysed tomato juice Approximately2 k go ftomat ojuic ewa sdialyse dfo r3 day sa t4° C against6 portion so f2 51 .distille dwater .Th edialyse djuic ewa s concentratedusin ga Btich ilaborator yrotar yfil mevaporato reithe rwit ho r withoutth eadditio no fth epreviousl ydialyse d sugars.Th esam etomat o juicewa sals oconcentrate dwithou tprio rdialysis .

Concentration of PE treated tomato juice 1.5k go ftomat ojuic ewa sincubate d for2 4hour sa t37° Cwit h6 4 unitso fpur emoul dPE .P Etreate da swel la suntreate djuic ewa s concentratedusin ga Btich ilaborator yrotar yfil mevaporator .

7.3. Results and Discussion

7.3.1. Effect of hot break temperature on gross viscosity

Theinfluenc eo fth eho tbrea ktemperatur eo nth egros sviscosit yo f tomatojuic ean dit sconcentrate swa sstudie dusin gtw otype so fscalders : 1)a conventiona lRoss i& Catell i100 0L scalde ran d2 )a moder nRoss i& Catelli"Eldorado "scalder . a) to o i-t 3 -* /-N r-l in• CO j* tu o "8-J8 c (3 •H co O (U •H •U o • 4-1 o CM •H u to p. O 00 Bfr a o

00 ON CM ON

on m 0) 1 o CO• CO• C © CO o •—' 1-4 o to o P. & o CJ CM r~ iri -3- CM r^ o 0\ m -tf 0> U1 00 %o -* »—' r-~ l/"> •—* r-~

o M o 00 •rl CO CO i-H CO —1 CM c o o a CM V to id u tO a to —I O o -< O H O 4-1 r- o O 00 C M a> o c 6-5 o « o u

o r- CO ON 0) X in ua m 10

J3 Cd PQ PS 168

Table 58 and Figures 45 and 46 present the results of experiments carried out for two succesive years (1983,1984 )wit h theR& C 1000L scalder.Fo r bothyear s thegros sviscosit y increased significantly upon an increase in hotbrea k temperature from 97-98°C to 103-105°C. In 1983 the increase in hotbrea k temperature from 90°C to 97CCwa s also found to affect the gross viscosity.

log t] app log!] app. &

3.0- 3.0- s ,ODD / & / £ •/ ^° tlserumQt10°Bx 2.0- 2.0- / a 92 °C D 92° C 4.1 mPa.s • 98 °C • 98° C 4.2 mPa.s o 105 °C o 105° C 4.7 mPa.s

0.5 0.7 1.5 -Q3 1.0 log %WIS log °Bx

Fig.45 :Effec to fho t break Fig. 46:Effec to fho tbrea ktemperatur e temperatureo ngros s ongros sviscosit yi nrelatio nt o viscosityan dseru m theWI Sconten t(1984) . viscosity inrelatio n to °Bx (1984).

This increase in grossviscosit y as a result of an increase inho t break temperature has oftenbee n attributed toquicke r enzyme inactivation (Luh et al. 1981,1982 ,1984a,b,c )an dbette r removal of insoluble solids from skin and seed material (Whittenberger &Nuttin g 1957,Twig g 1959). Indeed, the data shown in Table 58 on the serumviscosit y of the samples and the DE of theWI S pectin seem to indicate that aquicke r enzyme inactivation is involved in the increase in grossviscosity . Considering thedat a presented in chapter 4 (comparison of hotbrea k (PHB)wit h coldbrea k (PCB))an d chapter 6 (action of PE+ PG on tomato concentrate), it ishar d to believe that the tomato enzymeswer e able todegrad e 20%o f theWI S in the relatively short timebetwee n crushing and heating to90°C . Microscopic examination of thejuice s did not show any changes in the 169

appearenceo fth ecel lwall sa sa resul to fth eho tbrea ka tdifferen t temperatures;thi si si ncontras tt oresult spresente db yX ue tal . (1986). Moreoverresult spresente d intabl e5 8sho wtha tth equalit yo fth eWI Si s notaffecte db yth eho tbrea ktemperature ;al ljuice sshowe dcomparabl e grossviscosit ya t2 %WIS .Apparentl yth eenzyme sdegrade d theseru mpecti n andpartl ysaponifie dth eWI Spectin ,however ,n oindication swer efoun d forWI Sdegradatio na tth elowe rho tbrea ktemperature . In198 4i twa sfoun dtha tth eseru mviscosit y increasedonl yslightl ya sa resulto fth eincrease si nho tbrea ktemperature ;thi sindicate stha tth e pectolyticenzyme swer eeffectivel y inactivatedeve na ta ho tbrea k temperatureo f92°C .(Fig .45) . Thereforeth ehighe rgros sviscosit yfoun d asa resul to fa nincreas ei nho tbrea ktemperatur efro m92-98° Ct o105° C ismos tlikel ycause db yincrease d tissuesoftenin gan dbette rremova lo f WISfro mski nan dsee dmaterial .Agai nth equalit yo fWI Swa sno taffecte d byho tbrea ktemperature .(Fig . 46). Inth eho tsectio ntan ko fth eR& C"Eldorado "scalde rth erati o betweenrecirculatio nrat eo fheate dtomatoe san dfee drat eo ffres h tomatoesi sapproximatel y20 .Fres htomatoe sca nb eheate dalmos t instantaneously totemperature shig henoug hfo renzym einactivation .I nth e experimentsthi stemperatur ewa smeasure da tth einle to fth etomatoe si n theho tsectio ntank .Th efina lproduc ttemperatur ewa smeasure da tth e pointwher eth erecirculate dheate dtomatoe sentere dth eho tsectio ntank .

Table59 :Effec to fho tbrea ktemperatur eo nviscosit yan dWI Sconten to f tomatopast e (1984).

temperatureho tsectio ntan k analysiso fpast edilute dt o12.5°B x

%WI S n n Bostwick inlet aftertub .heate r app serum

87 91 2.70 969 6.9 2.8 95 101 2.46 815 6.5 2.0 98 107 3.09 1443 7.6 1.2 104 109 2.81 1181 7.8 1.3 170

Tables 59 and 60 show the results of the experimentswit h this type of scalder.Th e Bostwickvalue swer e determined directly in the factory on paste sampled at theoutle t of theR& C T60 Califfo evaporator. °Bx,%TS , WIS, n and n were determined in the laboratory on aseptically app serum J r filled paste.

Table60 :Effec to fho tbrea ktemperatur eo nsolid sconten tan dgros sviscosit yo ftomat opast e (1985). temperature hotsectio n tank solids content paste Bostwickdilute dpaste * n serum inlet aftertub . heater °Bx %rs %WIS 12.5°Bx 8.3°Bx 10°Bx

90 92 23.5 24.78 4.30 2.7± 0. 3 7.2 ±0. 5 9.0 94 98 23.4 24.87 4.52 2.3± 0.3 6.5 ± 0.6

100 105 23.5 24.85 4.30 2.1± 0.3 6.2 ±0. 5

105 108 23.6 25.42 4.85 1.7± 0.3 5.8± 0. 4

109 115 23^0 24.73 4.70 1.6 ± 0.1 5.4± 0. 2 10.6

*dat aar ea mea nvalu eo f8-1 0measurements .

With the "Eldorado" scalder no large differences in grossviscosit y were found between theus e of hot break temperatures of 91-92°C and 101-105°C (final temperatures),whic h is in contrast to results obtained with theR& C 1000L scalder.Ho t break temperatures inexces s of 105°C,however ,di d result in increased WIS content and grossviscosity . Itha s tob e noted that in Italy firm,hig hviscosit y varieties areuse d for processing. Apparently it ismor e difficult to remove theWI S from the skinmateria l of thesevarietie s than from the skinmateria l of themediu m viscosity varieties used inPortugal .Microscopi c examination did not indicate cell wallbreakdow n at low orhig hho tbrea k temperatures,WI S qualitywa s not affected by thedegre e of theho tbrea k temperature.Moreover ,dat a on serumviscosit y showed that pectolytic enzymeswer e effectively inactivated. The results obtained with theR& C "Eldorado" scalder also point to the improved removal ofWI S from skin and seed material as a cause of increased grossviscosit y athig hho tbrea k temperatures. 171

7.3.2. Effect of finishing on gross viscosity

Crushed tomatoes of thevariet y H30wer e prepared atho t break temperatures of 90°C and 103°C.Par t of the crushed tomatoeswa s finished with an industrialR& C paddle type finisher at approximately theho t break temperature,anothe r partwa s finished using a laboratory Bauknecht KU 2-1 tapered screwjuic e extractor at 90°C. Fig. 47 shows thathighe r grossviscositie s of juice and concentrates were obtained when crushed tomatoeswer e finished with the industrial paddle finisher.Thi sdifferenc e canb e attributed toa 23% (hotbrea k 103°C)an d 36% (hotbrea k 90°C)highe r WIS content in thejuic e finished with the paddle finisher. (Table61) .

n 103-1 industrial paddle • 90- T finisher o 103-1 lab. tapered screw 4 90-Tfinis

-0.3 0.5 log% WIS

Fig. hi: Effecto ftyp eo ffinishe r Fig.48 :Effec to ftyp eo ffinishe ro n ongros sviscosit yi nrelatio n grossviscosit yi nrelatio nt o to°Bx . WIScontent .

Similar resultswer e foundwhe n nonhea t processed samples,prepare d at a hotbrea k temperature of 103°C,wer e used in these experiments. Obviously the paddle finisher ismuc hmor e efficient in removing WIS from the skinmaterial .Th e chemical composition of theWI S injuice s finished with the tapered screw finisherwa sno t significantly different from the chemical compositions shown inTabl e 58 forjuice s finished with the paddle finisher. The increase of theWI S content at higher hot break temperatures 172

(34%fo rtapere dscre wextractor ,20 %fo rpaddl efinisher )canno tb e attributed toa nincrease d finishingtemperature .Coolin gth etomatoes , crusheda t103°C ,t oa temperatur eo f90° Co r65° Cbefor efinishin gwit h thepaddl efinishe rdi dno tresul ti ndecrease dgros sviscosity .Moreover , differencesi njuic egros sviscosit ywer eals ofoun dafte rfinishin g tomatoes,crushe da teithe r90° Co r103°C ,wit hth escre wtyp efinishe ra t atemperatur eo f90°C .

Table61 :Viscosit yan dsolid sconten to fjuice s (at5°Bx )finishe dwit h twotype so ffinisher . hotbrea ktemp . finisher %TS %WIS n n Bostwick app serum

90°C taperedscre w 5.40 0.56 20.2 1.4 >24 90°C paddle 5.70 0.77 106.7 1.6 18.7 103°C taperedscre w 5.74 0.75 78.5 2.4 20.4 103°C paddle 5.78 0.92 157.3 2.4 13.3

Preheattreatmen texert sa nimportan tinfluenc eo nth efinishe roperation , especiallywhe na les sefficien tfinishe r (tapered screw)i sused .Whe n tomatoes,crushe da t90°C ,wer efinishe dwit hthi stapere dscre wjuic e extractorth ejuic eno tonl ycontaine d lessWI S (Table61 )bu tals oWI So f aninferio rqualit y (Fig.48) .I ti sinterestin gt onot etha tth e heatprocessin ga t100° Co ftomatoe scrushe da t90 CC,doe sno tresul ti na n improvedremova lo fWI Sfro mski nmateria ldurin gfinishing .Apparently ,a combinationo fth ehighe rtemperatur ean dmechanica lshea r (recirculation inth echop/scalder )i sneede dfo rth eloosenin go fWI Sfro mski nmaterial .

Table62 :Viscosit yan dsolid sconten to fjuic e (at5°Bx )finishe da t20° C usingth etapere dscre wfinisher . hotbrea ktemp . %TS %WIS n n Bostwick r app serum

90°C 5.40 0.68 37.2 1.3 >24 103°C 5.66 0.80 99.6 2.2 17.5 173

Surprisingly,th eefficienc yo fth etapere dscre wfinishe rincrease dwhe n thecrushe dtomatoe swer efinishe da t20° Cinstea do f90°C . (Tables61 , 62).Th equalit yo fth eWI Si njuic eprepare da ta ho tbrea ktemperatur eo f 90°Cdi dno tincreas ehowever . FreshH. _tomatoe swer eheate di nclose dsterilizatio nbag si na retortt oa hear ttemperatur eo f90°C .Th eho ttomatoe swer efinishe dusin g thetapere dscre wfinishe requippe dwit ha 0. 5m mo r1. 5m mscreen .Th e totalyiel dincrease d from93 %t o96 %whe nth e1. 5m mscree nwa sused .A ta screensiz eo f1. 5m ma 22 %increas ei nWI Sconten twa sfoun d (Table63) , resultingi nincrease dgros sviscosit y (Fig. 49).

Table63 :Viscosit yan dcompositio no fjuice s (at5°Bx ) finishedusin gdifferen tscree nopenings .

finisher screensiz e

0.5m m 1.5m m

78.7 97.2 n app 2.3 2.3 ^serum Bostwick 21.2 19.1 %TS 5.71 5.64 %WIS 0.73 0.89 AGA1 15.7 16.2 DE 32 35 Cellulose 33.8 31.0 Protein 17.1 17.0

1:expresse da s% o fWI S

Inspit eo fthi sincrease ,th equalit yo fWI Sobtaine dusin ga 1. 5m m screensiz ewa sno ta sgoo da sth equalit yo fth eWI Sobtaine dusin ga 0. 5 mmscreen :thi swa ssee npredominantl ya ta highe rWI Sconten t (Fig. 50). Thechemica lcompositio no fth eWI Swa sno taffecte db yscree nsize .(Tabl e 63). Usingth emor eefficien tthre estag epaddl efinisher ,n oeffect so f 174 finisher screen size (0.8-1.0-1.5 mm final screens)o n theWI S content and grossviscosit y of tomato pastewa s found (Fig.51) .

3A-

3.0-

2.6-

2.2-

D 0.5 mm 1.8- • 15 mm

1.4- i i i - i i i i i i i n 1 1 1 1 r 0.5 1.0 1.5 -0.3 0.5 log°Bx log %WIS

Fig.49 :Effec to ffinishe rscree n Fig.50 :Effec to ffinishe rscree n openingo nth egros sviscosit y openingo nth egros sviscosit y inrelatio nt o°B x(tapere d inrelatio nt oWI Sconten t screwfinisher) . (taperedscre wfinisher) .

The effects of preheat treatment,finishe r temperature and finisher screen opening on the finisher operationwer e shown todepen d on the type of finisher. In confirmationwit h results presented by other authors (Hand et al. 1955,Bel-Ha j 1981) itwa s found that apaddl e type finisher ismor e efficient in removing WIS from skinmateria l than a screw type finisher. The positive effect of ahighe r preheating temperature on the gross viscosity andWI S content of juicewas ,therefore ,muc hmor e pronounced for the less efficient screw type finisher. Hand et al. (1955)reporte d a decrease injuic e grossviscosit y as a result of a decrease in finishing temperature after aho tbrea k at 93°C. Our results show that,wit h an efficient paddle finisher,a decrease in finishing temperature does not influence theWI S content and gross viscosity of tomato paste.Wit h the less efficient screw type finisher,a finishing temperature of 20°C resulted inhighe rWI S content and gross viscosity of juicewhe n compared to finishing at 90°C. The stiffness of the particles at 20°C obviously influenced the removal ofWI S from skin material. 175

The gross viscosity andWI S content of paste wasno t influenced by the size of thefinishe r screen openings (range 0.8-1.5 mm) when crushed tomatoes were finished using thepaddl e finisher. The size of thefinishe r screen openings did,however , influence thegros s viscosity andWI Sconten t of juice when thetapere d screw finisher was used. loglapp. 3.5

3.0-

2.5 a0. 8m m o1. 0m m •1. 5m m T" 0.9 1.1 1.3 log°B x 1 r 0.1 0.3 0.5 -log% WIS Fig.51 :Effec to ffinishe rscree nopenin g onth egros sviscosit y (paddle finisher)o fdilution so ftomat o paste.

In general,th edat a inth eliteratur e point toa n increase inth e gross viscosity of juice, paste andketchu p as result ofa n increase in size of the finisher screen opening (Hand et al. 1955,Smi t & Nortje 1958,Moye r et al. 1959,Twig g 1959,Bel-Ha j 1981).

7.3.3. Effect of concentration on gross viscosity

In order toestablis h thepotentia l effect of the concentration process on tomato paste gross viscosity, tomato juice of thevariet y H30 was concentrated using different methods or different types of equipment. Figure 52 shows that themetho d of concentration orth etyp e of equipment 176 had no influence onth egros s viscosity ofth econcentrates . From thedat a in Fig.2 3i twa salread y concluded that "WIS-concentration" (section 4.2.2.) resulted inth esam e gross viscosity of concentrates as concentration with thelaborator y rotary film evaporator. Obviously the recovery of gross viscosity isno t improved by using a relatively mild laboratory concentration technique instead ofa n industrial evaporator. In spite of some solubilization ofWI Spectin , ingenera l the composition of WISan dth eWIS/T S ratio were found to remain constant during concentration. Thesolubilizatio n ofpecti n seems todepen d onth edegre e of esterification of theWI Spectin : juice samples ofth evarietie s 722an d 318 with aD Eo f52 %los t upt o25 %o fth eAG Aconten t (equal to4 %o fth e WIS) upon concentration with a Pfaudler wiped film evaporator. Onth e contrary, juice samples of thevariet y H30wit h aD Eo fth eWI Specti n of 29 lost almost noAGA .A s reported above (Table 36)th ehemicellulos e present inWI Sdi dno t change during concentration.

logtlapp . 3.4H ./

3.0-

2.6-

type of concentrate: 2.2- industrial R&C OFF TJJ laboratory 20°C laboratory 70 °C atmospheric concentration centrifugation serum concentration

dilutions of 29 °Bx R&C DFF T30- 1.8- concentrate

0.2 0.4 0.6 log%WI S

Fig.52 :Gros sviscosit yo fjuice san dconcentrate s prepared accordingt odifferen tmethods ; grossviscosit yo fdilution so fa nindustria l concentrate. 177

Figure 52 also showsa significant differencebetwee n the gross viscosity of juice and concentrates on the onehan d and the grossviscosit y of diluted paste at the same solids level on the other hand. This' difference in grossviscosit y willb e called the "dilution-loss". There were no indications that concentration affects the serumviscosit y of diluted concentrates. As canb e seen inFig .52 ,a "dilution-loss" of 65-70%wa s found when tomatojuic e of thevariet y H30wa s concentrated to 29°Bx using theR& C T,. Anteo evaporator and then diluted to the original strength. Comparable "dilution-losses"wer e found after the dilution of tomato paste samples concentrated in theR& C T60 Califfo evaporator (samples decribed in section 7.3.1.). The dilution of 23°Bxpast e samples concentrated using the laboratory rotary film evaporator (n of concentrates is shown inFig . 52)resulte d indilutio n losses of only approximately 30%irrespectiv e of the temperature of concentration (20°C or 70°C). Evenunde r atmospheric conditions during evaporation the "dilution-loss"di d not increase to a levelhighe r than35% . Theeffec t of the centrifugation-serum-concentrationproces s on the "dilution-loss"wil lb e discussed later.Th e results indicate that the magnitude of the "dilution-loss" strongly depends on themetho d of concentration; the temperature of evaporation seems tohav e no influence. The"dilution-loss " isals o influenced by the solids content of the concentrate.

% dilution loss

T)apP.(mPa.s) 100H

160- *f

50- 80

a diluted to 4.9°Bx,not heated o heated 15min . 100° C

-i 1 1 r —i r -| 1 1 r 50 100 10 20 30 °Bx before dilution % water evaporated

Fig.53 :Gros sviscosit yafte rdilutio no fconcentrate s Fig.54 :% "dilution-loss "a sfunctio no fth e to4.9°Bx ;effec to fadditiona lheatin go f %wate revaporate ddurin gconcentration , dilutionso ngros sviscosity . 178

Figures5 3an d5 4sho wth e"dilution-loss "o ftomat ojuic econcentrate dt o varioussolid slevel susin gth eR& CT3 0Ante oevaporato r ( of app concentratesi sshow ni nFig .52) . Usingthi sevaporato rth e "dilution-loss"wa sfoun dt oh eproportiona lt oth eamoun to fwate r evaporated (Fig.54) .Additiona lheatin go fth edilute dsample sfo r1 5min . inboilin gwate rdi dno tnegat eth e"dilution-loss" ;i tresulte di nonl ya smallgros sviscosit y increase (Fig.53) . Thesam ewa sfoun dwhe ndilution s oflaborator yconcentrate d sampleswer eheated . Thedegre eo fsedimentatio n (section3.4. )o fWI Sparticle si n relationt oWI Sconcentratio nfo runconcentrate d tomatojuic euse di nthi s studyi sshow ni nFig .55 .Th ejuic ewa sdilute dwit hdifferen tamount so f wateran dtransferre d to10 0m lgraduate dseru mbottles .Afte r7 day sth e volumeo fth esedimen twa srecorded .Th eWI Sparticle sfille dth ecomplet e volumea ta concentratio no f± 0.6 5 g/100ml .Sample so fth ejuic e concentrated tosevera lsolid slevel swit hth eR& CT3 0Ante oevaporato r (Fig.5 3an d54 )wer edilute dt o4.9°B x (0.95%WIS) ,whic hi sth e refractiveinde xo fth ejuic ebefor econcentration . sediment volume (ml) 100- so

/ u / 50 f

0 0.4 0.8 WIS (g/100ml) Fig. 55: Degree of sedimentation of WIS particles in relation to WIS concentration.

50 grams of each dilution was weighed into the serum bottles and adjusted with water to a final volume of 100 ml. The sedimentation volume measured after 7 days clearly decreased as the concentration of the paste before dilution increased (Fig. 56). 179

sediment volume (ml) 100 90

78

63 63

50-

49 10.2 153 21.2 23.9 °Bx before dilution ~~0 54 71 81 83~ % water evaporated Fig. 56: Sedimentation volume of WIS particles (0.475 g/100 ml) in relation to concentration of tomato juice.

Comparison of Fig. 55 and Fig. 56 shows that the sedimentation volume obtained with 0.475 g WIS/100 ml as present in the 23.9°Bx concentrate, equals the sedimentation volume of only 0.24 g WIS/100 ml as present in unconcentrated juice. Microscopic examination of the dilution of the 23.9°Bx concentrate showed a considerable amount of collapsed and crumpled cells (Fig. 57).

Fig. 57: Collapsed and crumpled cells as present in diluted industrial tomato paste (magnification 61.4x). 180

Apparently those cellswer e not able to reabsorb thewate rwhic h was evaporated initially. This effect could alsob e demonstrated by theus e of the density gradient centrifugation technique (section 3.4.).

ABC H^ juice H3(,laborator y H^ industrial paste paste Fig.58 :Densit ygradien tcentrifugatio n oftomat ojuic ean dpast esamples .

Figure 58 shows typical centrifugation patterns for:ho tbrea k juice(A) , a diluted laboratory concentrate of A (B)an d a diluted industrial concentrate of A (C).A s a result of the concentration and dilution steps, a band X appears in sample Cwhic h isno t present in sampleA . Thisban d consists almost entirely of collapsed and crumpled cells.Ban d Xwa s found tob e barely present in sampleB . The "dilution-loss"o f this sample (+ 30%), however,wa s rather smallwhe n compared to the 60%"dilution-loss " of sample C. It isobviou s that there isa positive correlation between the magnitude of the"dilution-loss " and the presence of collapsed and crumpled cells. This also explainswh y concentration curves (n vs.%WIS )ar e not dependent onmetho d of concentration.Durin g concentration, tomato cells collapse and crumple irrespective of themetho d of concentration. In this situation grossviscosit y isdetermine d only by the amount ofWIS . 181

It isno t knownb y whichmechanis m the collapsed cells have lost the ability to reabsorbwate r and to expand. The influence of vacuum, shear and water removal on the determined "dilution-loss"wa s studied using the rotary vacuum film evaporator (Table 64).Juic e of thevariet y H1361 was concentrated at a product temperature of 20°Can d at two levels of shear. Moreover the juicewa s treated in the evaporator under such conditions that no evaporation occurred. In thiswa y the influence of shear,vacuu m and the combination ofbot h factors onjuic e grossviscosit y was studied. The gross viscosity of the juices and diluted paste sampleswa smeasure d two days after the treatment.

Table 64:Influenc e of different treatments in a laboratory rotary film evaporator ongros sviscosit y of tomato juice and diluted tomato paste.

sample product vacuum evapora tion rotation % dilution/ temp. (rpm) vis cosity loss

1 70 + + 63 42 2 70 + + 158 51 3 65 + - 63 21 4 65 + - 158 22 5 65 - - 63 17 6 65 - - 158 19 7 65 + - — 12

From the data it iseviden t that the loss in grossviscosit y due to evaporation,measure d as "dilution-loss",i s causedb y a complex of factors such aswater-removal ,mechanica l shear andvacuum . Results shown in Table 64 also indicate that apar t of the "dilution-loss" isno t caused by evaporationbu t solely by vacuum and mechanical shear. InFig .5 9 the gross viscosity of dilutions of laboratory concentrates isplotte d against the WIS content before dilution.Apparentl y the real "dilution-loss" startsa t 1.5-1.6%WIS ;a t thispoin t approximately 40%o f thewate rwa s evaporated. A loss in grossviscosit y atWI S levels lower than 1.5-1.6%wa s also found 182 when tomato juicewa s stirred in the rotary film evaporator without being concentrated.Thes e resultswer e confirmed by repeating the experiment using a different tomato juice.Figur e 59 also reveals that concentrates of juice fromwhic h ions and lowmolecula r weight sugarswer e removed by dialysis showed "dilution-losses" comparable to theviscosit y loss induced by vacuum andmechanica l shear, (see also Tables 26 to 28).Replenishmen t of the sugars removed by dialysis prior to concentration did not result in an increase of the"dilution-loss" .Thi s result indicates that the ions removed during dialysis are potentially important with regard to "dilution-loss". Itwa s found that approximately 50%o f the calcium present injuic e could be removed by dialysis. The treatment of tomato juicewit h purified PE (noP G or PL activity present) resulted ina n increase in the grossviscosit y of thejuic e and in gel formation upon storage.

'lapp. imPa.s) 160-

140-

120-

100- —i 1 1 1 r 0.5 1.5 2.5 3.5 %WIS before dilution

Fig.59 :Gros sviscosit y afterdilutio no fconcen ­ tratest o0.95 %WIS .Effec to fstirrin gi n evaporatoro ngros sviscosity . D dilutions laboratory concentrate • juice stirred inevaporato r without concentration Odilution so flaborator y concentrates ofdialyse d juice •dilutio no flaborator y concentrateo f dialysed juice+ remove d sugars 183

Concentrateso fP Etreate djuic ewer eals ohighe ri ngros sviscosit ytha n concentrateso funtreate djuic e (Table65) .

Table65 :Influenc eo ftreatmen twit hP Eo nviscosit yan d solidsconten to ftomat ojuic ean dconcentrates .

°Bx %TS %WIS n app n serum

Blank 5.3 6.16 0.97 169 2.4 9.9 11.26 1.74 569 5.1 16.4 18.32 2.85 1543 15.6 PE 5.3 6.23 0.97 202 1.2 9.8 11.35 1.74 664 1.9 17.2 19.48 3.12 2771 2.6

Dilutiono fth emos tconcentrate dsample sresulte di na "dilution-loss "o f 25%i nth euntreate d samplean d75 %i nth eP Etreate dsample .Microscopi c examinationshowe da larg eamoun to fcollapse dcell si nth edilutio no fth e PEtreate dsample .Th esedimentatio nvolum eo fthi ssampl ewa s5 1m la t 0.475g WIS/10 0m lcompare d to8 1m lfo rth euntreate d sample (seeals o Fig.55) .Obviousl yth eincreas ei ncalciu mreactiv ecarboxy lgroup scause d byP Eactio nresulte d ina ver ylarg e"dilution-loss" .I ti sinterestin gt o notetha tseru mviscosit ydecrease da sresul to ftreatmen twit hPE . Apparentlya larg epar to fth esaponifie d serumpecti nprecipitate d inth e WIS.Th eprecipitatio no fsaponifie d serumpecti nha salread ybee nshow n insectio n4.3 . (samplesLHB ,LCB.) . Inth ecours eo fthes estudie sa nattemp twa smad et oconcentrat e tomatojuic ewithou tmechanica lshea ran dvacuu mb ywate rremova lwit h ionican dnon-ioni cge lfiltratio nmedia .(Tabl e66) . 1.5gra mo fSephade x wasdredge do na dialysi sba gfille dwit h1 0gram so fjuice .Th edecreas e inweigh to fth ejuic ean ddialysi sba gwa sdetermine dafte r2 days . Thewea kcatio nexchange rCM-Sephade xwa sfoun dt ohav eth ehighes t "concentrationcapacity" .I na secon dexperimen t60 0gram so fjuic ewa s concentrated,agai nb ydredgin gCM-Sephade x (sodiuma scounter-ion )o na dialysisba gfille dwit hth ejuice . 184

Table66 :Concentratio no ftomat ojuic ewit hioni can dnon-ioni c gelfiltratio nmedi a (Sephadex,Pharmaci aFin eChemicals , Uppsala,Sweden) .

degreeo fconcentratio n

SephadexG5 0 1.5 SephadexG10 0 5.3 SephadexG15 0 3.6 SephadexG20 0 4.5 Sephadex-DEAE 2.0 Sephadex-QAE 2.1 Sephadex-CM 14.6 Sephadex-SP 4.7

Aftertw oday sth ejuic ewa sconcentrate d 11.1times ;thi smean stha tthi s pastewa scomparabl et oa 60°B xho tbrea kpaste .

l °9 ^1 QPp. 3.4-

3.0-

2.6

Bostwick at 12.5 °Bx 2.2- laboratory concentrated juice 3.0 • dilutions Sephadex concentrated juice 3.0 o dilutions industrial concentrated juice 4.0

~I— -0.1 0.6 log %WIS

Fig. 60:Gros sviscosit yo fdilution so findustriall y concentrated juicean dSephade x concentrated juice comparedt ogros sviscosit yo fjuic e concentrates. 185

Uponth edilutio no fsample sconcentrate daccordin gt othi stechnique ,n o "dilution-loss"wa sobserve d (Fig.60) .Moreover ,microscopi cexaminatio n showedn osignifican tdifferenc ei namoun to fcollapse d cellsbetwee n untreatedjuic ean ddilution so fSephade xconcentrate djuice .Thi s observationlead st oth econclusio ntha twate rremova la ssuc hdoe sno t causea "dilution-loss" .Anothe rindicatio nfo rthi sconclusio nwa sfoun d afterth ecentrifugatio no ftomat ojuic ei na Sorval lRC-5 Brefrigerate d superspeedcentrifug efo r4 5min .a t25,40 0x g .Resuspensio no fth eseru m andfibre s(respectivel y75 %an d25 %o ftota lweight )resulte di nonl y smallgros sviscosit y losseso fbetwee n0 %an d10% . Table6 7show sth eresult so fa laborator ycentrifugation-serum - concentrationexperiment .N o"dilution-losses "wer efound .Th egros s viscosity ofthes esample sbefor edilutio nwa sth esam ea sth egros s viscosityo fothe rtype so fconcentrate s (Fig. 52).

Table67 :Gros sviscosit yo fconcentrates ,prepare daccordin gt oth e centrifugation-serum-concentrationtechnique ,befor ean dafte r dilution*.

concentrates dilutionso f4. 9°B x

°Bx %WIS n n app app

4.9 0.92 145.0 152.6 7.1 1.38 355.5 148.0 9.5 1.74 627.4 148.0 11.6 2.12 912.1 148.5 13.4 2.44 1192.8 146.7 15.0 2.74 1490.8 151.2

*dilution swer emad e1 da yafte rpreparatio no fth econcentrates .

Insummar yi tca nb econclude d that"dilution-losses "ar eobserve dwhe n certainseru mcomponent s (ions)ar econcentrate db yevaporatio ni nth e presenceo ftomat ocells .Th emagnitud eo fth e"dilution-loss "i sfurthe r influencedb ymechanica lshea ran dvacuum . 186

The results obtained indicated that the centrifugation-serum- concentration technique could be a solution for thepreventio n of "dilution-loss".Tw o types of decanter centrifugeswer euse d to test the usefulness of this technique on a larger scale.Th e results obtained with the different types of decanterwer e comparable;a typical result is shown inFig .61 .

109llapp . 3.4-

laboratory concentrates dilutions decanter concentrate

Fig. 61: Gross viscosity of dilutions of industrially concentrated juice and juice concentrated according to the centrifugation - serum - concentration technique compared to gross viscosity of juice concentrates.

Dilutions of the concentrate prepared in the R&C T30 Anteo evaporator show a "dilution-loss" of 28% when diluted to a soluble solids level of 12.5%. At this solids level a concentrate prepared using the decanter showed a "dilution-loss" of 14%. Bostwick values at 12.5% soluble solids for evaporator paste and decanter paste were respectively 5.5 cm and 4.4 cm. Under specified conditions the use of a decanter in the centrifugation-serum-concentration process resulted in a significantly 187 reduced "dilution-loss",although ,a complete prevention of the "dilution-loss" could not be achieved. Obviously the centrifugation of tomato juice ina decanter cannot be directly compared to the centrifugation of tomatojuic e in abenc h top centrifuge.A n explanation for the differencebetwee n the two centrifuges canb e found in thehighe r mechanical shear on the insolubleparticle s during centrifuging in a decanter. The results presented in this study on the effect of the centrifugation-serum-concentration technique on thegros sviscosit y of tomato paste are contradictory to results published by other authors (Ephraim et al. 1962,Kopelman n& Mannhei m 1964,Mannhei m &Kopelman n 1964, Harper & El Sahrigi 1965) .Harpe r & El Sahrigiwer e the only authorswh o used aho tbrea k juice for the centrifugation but unfortunately they did not describe the centrifugation process indetail . The "dilution-loss" is influenced not only by the concentration process but alsob y other processing factors,a swil lb e shown in section 7.3.4.an d in the following example.Whol e tomatoeswer e heated in closed sterilizationbag s ina retort for 20min . at 120°C (control). A heart temperature of 90°Cwa s measured.

Ticpp(mPa.s ) 140H

n retort 20 min. 120 °C o retort 20 min. 120°C • 20min. 100°C 120 • retort 20 min. 120 °C +juic e pasteurized

100-

80- -1—I— n 1 1 r 10 20 30 °Bx before dilution

Fig.62 :Effec to fadditiona lhea t treatmento ngros s viscosityo fjuic ean ddilute d concentrates. 188

Theheate dtomatoe swer efinishe dthroug ha Bauknech tK U2- 1tapere dscre w juiceextracto requippe dwit ha 1. 5m mscreen .Th echemica lcompositio no f thisjuic ewa sshow ni nTabl e63 .Hal fo fth efinishe djuic ewa shea t processedfo r10 0min .a t100°C .I nanothe rexperiment ,tomatoe swer e heatedfo r2 0min .a t120° Can dthe nth eretor ttemperatur ewa sreduce dt o 100°C;heatin gwa scontinue dfo r2 0min .Th eheate dtomatoe swer efinishe d inth esam ewa ya sdescribe dabove . Bothadditiona lhea ttreatments ,o nto po fth e2 0min .a t 120°C,gav eris e toa reductio ni ngros sviscosit yo f± 25 %(Fig .62) ,cause db yth e solubilizationo f20-25 %o fth eAG Apresen ti nth eWIS .Compare dt oth e "dilution-loss"fo rth econtrol ,thes eadditiona lhea ttreatment scause d onlysmal l"dilution-losses" ;thi swa sdu et oth ever ylo wgros sviscosit y ofth ejuic ebefor eevaporation . Therewer en oindication stha tth edegre eo fho tbrea ktemperatur e influencesth emagnitud eo fth e"dilutio nloss" .Tomat opast esample s prepareda tdifferen tho tbrea ktemperature s (Table59 )showed ,afte r dilution,identica lgros sviscosit ya ta specifi cWI Scontent ,irrespectiv e ofth eho tbrea ktemperatur e (Fig. 63).

3.4-

concentration 3.0 curve

2.6-

2.2-

1.8 T 1 1 r -0.1 0 0.2 0.4 ' o!6 log %WIS

Fig. 63:Gros sviscosit yo fdilution so f23°B x tomato paste prepareda tdifferen tho t break temperatures compared togros s viscosityo fconcentrates . 189

7.3.4. Effect of heat processing and freezing on gross viscosity

Table 68 shows the effect of pasteurization on the gross viscosity of non-pasteurized hot break tomato juice and concentates. A 10% reduction in gross viscosity of a 15.8°Bx tomato paste was observed due to heat processing for 100 min. at 100°C. Upon dilution of this sample a "dilution-loss" of 48% was found; in the case of non-pasteurized paste, a "dilution-loss" of only 29% was found.

Table 68: Effect of pasteurization of tomato juice and concentrates on gross viscosity before and after dilution*.

before pasteurization after pasteurization

°Bx WIS n °Bx %WIS n app app

concentrates 4.8 0.95 143.1 4.8 0.91 124.8

9.4 1.79 573.2 9.4 1.78 530.2

15.8 3.00 1609.9 15.8 2.89 1471.3

18.7 3.49 2211.8

dilutions 9.4-> • 4.8 0.90 102.8 4.7 0.91 82.0

15.8 + 4.8 0.92 101.4 4.7 0.91 74.3

18.7• * 4.8 0.88 99.0

* Concentrates were obtained in a rotary film evaporator on laboratory scale; an aliquot of the concentrates was heat processed (100 min. at 100°C); measurements were done either directly on diluted non-pasteurized samples or on pasteurized samples which were diluted. Juice of the variety H30 was used.

Similar effects of pasteurization on the gross viscosity of tomato paste were observed when hot break paste was produced in Italy for the purpose of the experiments described in section 7.2.2. Tomato paste was sampled at the outlet of the Rossi & Catelli T60 Califfo evaporator (non-sterilized paste) and at the aseptic filler (paste sterilized for 2 min. 15 sec. at 109°C). Comparison of gross viscosity data for non-sterilized and sterilized paste 190

indicated that Bostwickvalue s at 12.5°Bxwer e 0.5-1.0 cm.highe r for the sterilized paste.Hea t processing seems to increase the "dilution-loss"o f a tomato concentrate. These results confirm the results reported by Marsh et al. (1977, 1982). The loss in grossviscosit y of a juiceupo n pasteurization (Fig. 62, Table 68) canb e related to adecreas e inWI S content,cause d mainly by the solubilization ofWI S pectin.Thi s quantity of solubilized WIS seems to depend on the amount ofAG A present in theWI S relative to the other constituents (e.g.AGA/Cellulos e ratio). Industrially prepared samples of juice, ingeneral ,sho w aAGA/Cellulos e ratio of0.30-0.3 5 andwil l lose approximately 4-5% of theWI S as a result ofhea t processing. The laboratory prepared juice sample shown inFig .6 2ha d a relatively high AGA/Cellulose ratio of 0.52 due to the effective "in-situ" inactivation of enzymes in the retort.Thi s sample lost almost 10%o f theWI S and showed a 25% decrease in grossviscosit y asa result ofhea t processing; the AGA/Cellulose ratio decreased to 0.40.

Table69 :Effec to ffreezin go ftpmat ojuic ean dconcentrate so ngros sviscosit ybefor ean d afterdilution* .

beforefreezin g afterfreezin g

°Bx %WIS °Bx %WIS app app

concentrates 4.5 1.04 165.4 4.4 1.01 132.5 8.0 1.82 652.8 7.9 1.77 568.5 12.7 2.90 1543.0 12.6 2.82 1437.9 18.4 4.22 - 18.4 4.03 -

dilutions 8.0 -»• 4.6 1.04 130.2 4.5 1.04 123.5 12.7 -> 4.5 1.04 122.1 4.4 0.99 109.8

18.4 •*• 4.6 1.06 120.7 4.4 0.98 103.8

* Concentrateswer eobtaine d ina rotar yvacuu mevaporato ro nlaborator y scale;a naliquo to f

theconcentrate swa sfroze n(-20°C )an d stored for6 weeks ;measurement swer edon eeithe r

directlyo ndilute dnon-froze n sampleso ro nfroze nsample swhic hwer edilute dupo nthawing .

Juiceo fth evariet yH136 1wa sused . 191

Table 69 shows the effect of freezing on the grossviscosit y of tomato juice and concentrates.Freezin g the tomato juice caused a gross viscosity loss of 20%;concentrate s also showed significant grossviscosit y losses upon freezing.Freezin g hardly increased themagnitud e of the "dilution-loss" (unfrozen juice as reference). In this connection it should be noted that comparisons should only bemad e at corresponding WIS levels. It is obvious that frozen storage ofbot h juice and concentrates affects the ability to detect a"dilution-loss" .

7.4. Conclusions

The increase in the grossviscosit y of tomato juices and tomato juice concentrates as a result of theus e ofho tbrea k temperatures higher than thoseneede d to inactivate the enzymes canb e attributed to abette r removal ofWI S from the skinmaterial .WI S "quality" and the microscopic appearence of the tomato cells are not affected by higher hot break temperatures.Th e effect of a specific hot break temperature on the removal of theWI S from the skinmateria l probably depends on the tomato variety. The effects ofho tbrea k temperature,finishe r temperature and finisher screen opening on the finisher operation are closely related to the effectiveness of the finisher.

Themetho d of concentration or the type of evaporator used for concentration have no influence on the grossviscosit y of concentrates. In general theWIS/T S ratio remains fairly constant during concentration. A "dilution-loss" is observed when certain serum components (ions)ar e concentrated by evaporation in thepresenc e of the tomato cells.Th e magnitude of the"dilution-loss " is influenced bymechanica l shear,vacuu m and heat processing of the concentrated juice.Th e quality of the juice before concentration also influencesth e "dilution-loss".Th e centrifugation-serum-concentration technique results in thepreventio n of the"dilution-loss "whe n abenc h top centrifuge isused . Theus e of a decanter centrifuge in this technique results ina decreased "dilution-loss".

Bothhea t processing and freezing result indecrease d grossviscosit y of tomato juice and tomato juice concentrates. 192

REFERENCES

Ahmed,A.E .& J.M .Labavitch ,1977 .A simplifie d method foraccurat edeterminatio n ofcel lwal l uronidecontent .J .Foo dBlochem .1:361-365 . Anonymous,1975 .Concentrate sho tbrea ktomat opast et o32 %tota l solids.Foo dProcessin g 36:68. Anonymous,1981a .Rosell aupgrade s tomatoprocessin gplan ta tTatura .Foo dTechnol .Aust . 33:437. Anonymous,1981b .Cerebo sdecentralis e tomatoprocessing .Foo dTechnol .Aust .33:486 . Baron,A. ,F.M .Rombouts ,J.F .Drillea u& W .Pilnik ,1980 .Purificatio nan d propertieso fth epectinesteras eproduce d byAspergillu sniger .Lebensm .Wiss .u . -lechnol.13:330-333 . Bartholin,G .& H .Koubaa ,1981 .Aptitud e desju sd etomate sa l a concentration.Repor tC.T.C.P.A.-statio nexperimentale-13540-Puyricard . Bartolome,F.M. , 1972.A studyo fth efactor san dmechanism s affectingconsistenc y ofprocesse d tomatojuice .Ph.D .Thesis ,Purdu eUniv. ,Lafayette ,Indiana ,USA . Becker,R. ,J.R .Wagner ,J.C .Miers ,D.W .Sanshuc k& W.C .Dietrich ,1968 .Consistenc yo ftomat o products.Il lEffect so fp Hadjustmen t during tomatojuic epreparatio n onpecti ncontent s and characteristics.Foo dTechnol .22 :503-506 . Becker,R. ,J.C .Miers ,M.D .Nutting ,W.C .Dietric h& J.R .Wagner ,1972 .Consistenc yo ftomat o products.VI IEffect so facidificatio n oncel lwall san d cellbreakage .J .Foo dSc .37 : 118-125. Beldman,G. ,M.F .Searle-va nLeeuwen ,F.M .Rombout s& F.G.J .Voragen ,1985 .Th ecellulas eo f Trichodermaviride .Purification ,characterization ,an dcompariso n ofal ldetectabl e endoglucanases,exoglucanase s andg-glucosidases .Eur .J .Biochem .146 :301-308 . Beldman,G. ,1986 .t ob epublished . Bel-Haj,H.M. ,1981 .Effec t ofcultivars ,brea ktemperature ,pulpin gan dextractio nmethod so n theviscosit yo ftomat ojuice .Ph.D .Thesis ,Ohi oStat eUniv. ,Columbus ,USA . Besford,R.T .& G.E .Hobson ,1972 .Pecti c enzymesassociate dwit h thesoftenin go ftomat o fruit.Phytochemistr y 11:2201-2205 . Brady,C.J. ,G .Macalpine ,W.B .McGlasso n& Y .Ueda ,1982 .Polygalacturonas e intomat ofruit s and theinductio no fripening .Aust .J .Plan tPhysiol .9:171-178 . Brown,H.E .& E.R .Stein ,1977 .Studie so nth ealcohol-insolubl e solidso fChic oII Ian d Homestead-24tomatoes .J .Agric .Foo dChem .25 :790-793 . Buescher,R.W .& E.C .Tigchelaar ,1975 .Pectinesterase ,polygalacturonase ,C- x cellulase activitiesan d softening ofth eri ntomat omutant .Hort .Sci .10 :624-625 . Buescher,R.W .& G.E .Hobson ,1982 .Rol eo fcalciu m andchelatin gagent s inregulatin gth e degradation oftomat ofrui t tissueb ypolygalacturonase .J .Food .Biochem .6 :147-160 . Carlson,R.A. ,J.M .Randall ,R.P .Graha m& A.I .Morga nJr. ,1967 .Th erotar y steam-coilvacuu m evaporator.Foo dTechnol .21 :194-196 . Chen,C.S. ,V.M .Lim aHo n& A .Marsaiol i jr.,1979 .A hea t transfercorrelatio n forrotar y steam-coilvacuu mevaporatio n oftomat opaste .J .Foo d Sc.44 :200-202 . Cooler,F.W. , 1962.Rheologica l studieswit hcanne d tomatojuice .Ph.D .Thesis ,Universit yo f Maryland,USA . Crandall,P.G .& P.E .Nelson ,1975 .Effect so fpreparatio nan dmillin g onconsistenc yo ftomat o juicean dpuree .J .Foo d Sc.40 :710-713 . Crean,D.E. ,1969 .A studyo fth econsistenc yo ftomat ojuic ea sinfluence db yp Hchange san d cellwal l components.Dissert .Abst .30 :1742-B . Crookes,P.R .& D .Grierson ,1983 .Ultrastructur e oftomat ofrui tripenin g and therol eo f polygalacturonase isoenzymes incel lwal ldegradation .Plan tPhysiol .72 :1088-1093 . Dale,M.C. ,M.R .Oko s& P .Nelson ,1982 .Concentratio no ftomat oproducts :analysi so fenerg y savingproces salternatives .J .Foo dSci .47 :1853-1858 . Davies,J.N .& G.E .Hobson ,1981 .Constituent so ftomat ofruit-th e influenceo fenvironmen t nutritionan dgenotype .C.R.C .Crit .Rev .Foo dSci .Nutr .15 :205-280 . Davis,R.B. ,D.DeWees e& W.A .Gould ,1954 .Consistenc ymeasurement so ftomat opuree .Foo d Technol.8 :330-334 . 193

Dekazos,E.D. ,1972 .Callos e formationb ybruisin gan dheatin go ftomatoe san d itspresenc ei n processedproducts .J .Foo dSci .37 :562:567 . Delincee,H .& B.J .Radola ,1970 .Som esiz ean dcharg epropertie so ftomat opecti n methylesterase.Biochim .Biophys .Act a214 :178-189 . Delincee,H. ,1976 .Thin-laye r isoelectricfocusin go fmultipl eform so ftomat opectinesterase . Phytochem.15 :903-906 . Deshpande,S.S. ,S.K .Sath e& D.K .Salunkhe ,1984 .Freez econcentratio no ffrui tjuices .C.R.C . Crit.Rev .Food .Sci .Nutr .20 :173-248 . Dey,P.M .& K .Brinson ,1984 .Plan tcell-walls .In :Advance si ncarbohydrat echemistr yan d biochemistry,R.S .Tipso n& D .Horto n (Eds).Academi cPress ,London ,42 :265-382 . Dickinson,D.B .& D.P .McCollum ,1964 .Cellulas e intomat ofruits .Natur e203 :525-526 . Dongowski,G. ,W .Bock ,0 .Markovi c& A .Slezarik ,1980 .Einflus smolekulare rParamete rde s Galakturonasubstrats aufdi eAktivitS teine rPolygalakturonas eau sTomaten .Di eNahrun g 24:619-626 . Dougherty,R.H .& P.E .Nelson ,1974 .Effect so fp Ho nqualit yo fstore d tomatojuice .J .Foo d Sci. 39:254-256 . Draetta,I. ,T .Kuroda ,M .Iaderoz a& C .Ciampi ,1979 .Inativaca otermic ad apectinesteras e dotomate :relaca otempo ,temperatur ae pH .Colet .Inst .Technol .Aliment .10 :115-139 . Dubois,M. ,K.A .Gilles ,J.K .Hamilton ,F.A .Reber s& F .Smith ,1956 .Colorimetri cmetho d for determinationo fsugar san drelate dsubstances .Anal .Chem .28 :350-356 . El-Sayed,M.N.K .& H.T .Erickson ,1966 .Pecti c substancesi ntomatoe sa srelate d towhol efrui t firmnessan dinheritance .Proc .Am .Soc .Hort .Sci .89 :528-531 . Englyst,H. ,1981 .Determinatio no fcarbohydrat ean d itscompositio n inplan tmaterials ,Basi c Clin.Nutr .3 :71-93 . Ephraim,A. ,H.C .Mannheim ,J.B.S .Braverma n6 .L .Shelef ,1962 .Th eeffec to fprocessin g conditionso nth econsistenc yo ftomat ojuice .Bull .Res .Counc .o fIsrael .1 1C3 : 286-294. Feinberg,B. ,1967 .Concentratio nb yevaporation .In :Fundamental so fFoo dprocessin g operations,J.L .Hei d& M.A .Josly n (Eds).AV IPublishin gCo .Westport ,Connecticut ,USA . Foda,Y.H. ,1957 .Pecti cchange sdurin gripenin ga srelate d tofles h firmnessi nth etomato . Ph.D. Thesis,Univ .Illinois ,Urbana ,Illinois ,USA . Foda,Y.H .& J.P .McCollum ,1970 .Viscosit ya saffecte db yvariou sconstituent so ftomat o juice.J .Foo dSci .35 :333-338 . Garces,M.B .& B.S .Luh ,1972 .Characteristic so fpectinesteras ean dpolygalacturonas ei n VF-145tomatoes .Confruct a 17:360-370 . Goodenough,P.W. , 1981.Compositio n andbiochemistr yo ftomat ofruit .In :Proceedings :Nort ho f Englandbranc h symposium tomatoesan dtomat oproducts ,Institut eo fFoo dScienc e& Technology (UK)14 :28-39 . Goose,P. ,1981 .Development si nth eproductio no fprimar ytomat oproducts .In :Proceedings : Northo fEnglan dbranc hsymposiu m tomatoesan d tomatoproducts ,Institut eo fFoo dScienc e 6.Technolog y (UK),14 :15-27 . Gould,W.A. ,1983 .Tomat oProduction ,Processin gan dQualit yEvaluation .AV IPublishin gCo. , Westport,Connecticut ,USA . Grierson,D .& G.A .Tucker ,1983 .Timin go fethylen ean dpolygalacturonas e synthesisi n relationt oth econtro lo ftomat ofrui tripening .Plant a157 :174-179 . Gross,K.C .& S.J .Waliner ,1979 .Degradatio no fcel lwal lpolysaccharide sdurin g tomatofrui t ripening.PI .Physiol .63 :117-120 . Gross,K.C .& M.E .Saltveit ,1982 .Galactos econcentratio nan dmetabolis m inpericar ptissu e fromnorma lan dnon-ripenin g tomatofruit .J .Am .Soc .Hortic .Sci .107 :328-330 . Gross,K.C ,1983 .Change si nfre egalactose ,myo-inosito lan dothe rmonosaccharide s innorma l andnon-ripenin gmutan ttomatoes .Phytochemistr y22 :1137-1139 . Gross,K.C , 1984.Fractionatio nan dpartia lcharacterizatio n ofcel lwall s fromnorma lan d non-ripeningmutan ttomat ofruit .Physiol .Plan t62 :25-32 . 194

Gross,K.C .& C.E .Sams ,1984 .Change s incel lwal lneutra l sugarcompositio ndurin gfrui t ripening:a specie ssurvey .Phytochemistr y23 :2457-2461 . Hakomori,S. ,1964 .A rapi dpermethylatio no fglycolipi d andpolysaccharid ecatalyze db y methylsulfinylcarbanio n indimethy lsulfloxide .J .Biochem .55 :205-208 . Hall,C.B .& R.A .Dennison ,1960 .Th erelationshi po ffirmnes san dpectinesteras eactivit yo f tomatofruit .Proc .Am .Soc .Hort .Sci .75 :629-631 . Hall,C.B. ,1963 .Change si nth ealcoho l insolublesolid so fportion so fdetache d tomatofruit s duringripening .Proc .Am .Soc .hort .Sci .83 :717-720 . Hall,C.B. ,1964 .Cellulas eactivit y intomat ofruit saccordin g toportio nan dmaturity .Bot . Gaz.125 :156-157 . Hall,C.B .& J.T .Mullins ,1965 .Cellulas eactivit yi nsteril etomat ofruits .Natur e206 : 638-639. Hand,D.B. ,J.C .Moyer ,J.R .Ransford ,J.C .Henin g& R.T .Whittenberger ,1955 .Effec to f processingcondition so nth eviscosit yo ftomat ojuice .Foo dTechnol .9 :228-235 . Harper,J.C .& A.F .E lSahrigi ,1965 .Viscometri cbehavio ro ftomat oconcentrations .J .Foo d Sci.30 :470-476 . Hermann,K. ,1979 .Ubersich flbert di eInhaltstoff ede rTomaten .Z .Lebensm .Unters .u .-Forsch . 169:179-200 . Herranz,J. ,C .Vidal-Valverd e& E .Rojas-Hidalgo ,1981 .Cellulose ,hemicellulos ean dligni n contento fra wan dcooke dSpanis hvegetables .J .Food .Sci .46 :1927-1933 . Hinton,D.M .& P .Pressey ,1980 .Glucanase si nfruit san dvegetables .J .Amer .Soc .Hort .Sci . 105:499-502 . Hobson, G.E., 1962 Determinationo fP Gi nfruits .Natur e195 :804-805 . Hobson, G.E., 1963 Pectinesterase innorma lan dabnorma l tomatofruit .Biochem .J .86 : 358-364. Hobson, G.E., 1964 Polygalacturonase innorma lan dabnorma ltomat ofruit .Biochem .J .92 : 324-332. Hobson, G.E., 1965 Thefirmnes so ftomat ofrui ti nrelatio nt opolygalacturonas eactivity .J . Hort.Sci .4 0 66-72. Hobson, G.E., 1968 Cellulaseactivit ydurin gth ematuratio nan dripenin go ftomat ofruit . J.Foo dSci .33 :588-592 . Hobson,G.E. ,1980 .Effec to fth eintroductio no fno nripenin gmutan tgene so nth ecompositio n andenzym econten to ftomat ofruit .J .Sci .Foo dAgric .31 :578-584 . Hobson,G.E. ,1981 .Enzyme san d texturechange sdurin gripening .Annu .Proc .Phytochem .soc . Eur.1 9n orecen tadv .biochem .fruit sveg. :123-132 . Holloway,W.D .& R.I .Greig ,1984 .Wate rholdin gcapacit yo fhemicellulose sfro mfruits , vegetablesan dwhea tbran .J .Foo d Sci.49 :1632-1633 . Houdenhoven,F.E.A .van ,1975 .Studie so npectinlyase .Ph.D .Thesis ,Agric .Univ .Wagenlngen , TheNetherlands . Huber,D.J. ,1983 .Polyuronid edegradatio nan dhemicellulos emodification s inripenin gtomat o fruit.J .Amer .Soc .Hort .Sci .108 :405-409 . Ishii,K. ,S .Konomi ,K .Kojima ,M .Kai ,N .Uka i& N .Uno ,1981 .Developmen to fa tomat ojuic e concentration systemb yrevers eosmosis .AC SSymp .Ser .154 .Synth ,membr .vol .II :1-16 . Janoria,M.P .& A.M .Rhodes ,1974 .Juic eviscosit ya srelate d tovariou sjuic econstituent san d fruitcharacter s intomatoes .Euphytic a 23:553-562 . Janoria,M.P. ,1974 .Samplin gvariatio n inalcoho l insoluble solidconten tan dviscosit yo f tomatojuices .J .Hort .Sci .49 :305-310 . Johannessen,G.A. ,1981 .Th esoli dreview .Californi atomat ogrowe r24 :4-5 . Jona,R .& E .Foa ,1979 .Histochemica lsurve yo fcell-wal lpolysaccharide so fselecte dfruits . ScientiaHorticultura e10 :141-147 . Jones,T.M .& P .Albersheim ,1972 .A ga schromatographi cmetho d forth edeterminatio no faldos e anduroni caci dconstituent so fplan tcel lwal lpolysaccharides .PI .Physiol .49 :926-936 . Katan,M.B .& P .va nd eBovenkamp ,1981 .Determinatio no ftota ldietar y fiberb ydifferenc e ando fpecti nb ycolorimetr y orcoppe rtitration .Basi cclin .Nutr .3 :217-239 . 195

Keegstra,K. ,K.W .Talmadge ,W.D .Baue ran dP .Albersheim ,1973 .Th estructur eo fplan tcel l walls.Il lA mode lo fth ewall so fsuspension-culture d sycamorecell sbase d onth e Interconnectionso fth emacromolecula rcomponents .PI .Physiolog y51 :188-196 . Kertesz,Z.I .an dJ.D .Loconti ,1944 .Factor sdeterminin g theconsistenc yo fcommercia lcanne d tomatojuice .N.Y .Stat eAgr .Expt .Sta .Geneva ,Techn .Bull ,n o272 . Keybets,M.J .an dW .Pilnik ,1974 .Som eproblem si nth eanalysi so fpecti ni npotat o tubertissue .Potat oRes .17 :169-177 . Kintner,P.K .& J.P .v .Buren ,1982 .Carbohydrat e interferencean dit scorrectio n inpecti n analysisusin gth em-hydroxydipheny lmethod .J .o fFoo d Sci.47 :756-759 ,764 . Knee,M .& I.M .Bartley ,1981 .Compositio nan dmetabolis mo fcel lwal lpolysaccharide si n ripening fruits.In :Recen tadvance s inth ebiochemistr yo ffruit san dvegetables .J . Friend& N.J.C .Rhodes .(Eds) .Academi cPress ,London ,p p133-148 . Knee,M. ,1982 .Frui tsoftening .I IPrecurso r incorporation intopecti nb ypea r tissueslices . J.Exp .Bot .33 :1256 -1262 . Kohn,R. ,0 .Markovi c& E .Machova ,1983 .Deesterificatio nmod eo fpecti nb ypecti nesterase s ofAsp .foetidus ,tomatoe san dalfalfa .Collect .Czech .Chem .Commun .48 :790-797 . Kopelmann,J .& H.C .Mannheim ,1964 .Evaluatio no ftw omethod so ftomat ojuic econcentration .I Heat-transfercoefficients .Foo dTechnol .18 :907-910 . Lackey,G.D. ,K.C .Gros s& S.J .Wallner ,1980 .Los so ftomat ocel lwal lgalacta nma yinvolv e reduced rateo fsynthesis .PI .Physiol .66 :532-533 . Ladwig,H. ,1982 .Di eKonzentrierun gvo nlomatensafte ndurc hdi eAnwendun gmoderne r Verarbeitungstechnologien.Fluss .Obs t49 :211-213 . Lamb,F.C. ,1977 .Tomat oproducts .Bull .27- L5th .ed. ,Nationa lCanner sAssociation , Washington. Lamport,D.T.A .6 cL .Epstein ,1983 .A ne wmode l forth eprimar ycel lwall :a concatenate d extensin-cellulosenetwork .Curren tTopic si nPlan tBiochemistr yan dPhysiolog y 2:73-83 . Lee,M .& J.D .Ma cMillan ,1968 .Mod eo factio no fpecti cenzymes .1 Purificatio nan dcertai n propertieso ftomat opectinesterase .Biochemistr y 7:4005-4010 . Lee,M .& J.D .Ma cMillan ,1970 .Mod eo factio no fpecti cenzymes .Il lSit eo finitia laction s oftomat oP Eo nhighl yesterifie d pectin.Biochemistr y 9:1930-1934 . LeMaire ,W.H. ,1981 .Th ebes t littletomat opast eplan ti nItaly .Foo dEngineerin g International 6:28-33 . Leoni,C , L.Bolzoni ,R .Aldini ,1979 .Holdin gtim eo fcrushe d tomatoesbefor e finishingan d concentration:effect so nth econsistenc y ofth econcentrat eproduced .Industri aConserv e 54:199-203 . Leoni,C , 1980.Storag eo fcrushe d tomatoesbefor econcentration .Effect so nth econsistenc y ofth econcentrate .Int .Soc .Hort .Sci .(tomat osymposium) .Act aHorticultura e100 : 315-232. Leoni,C , L.Bolzoni ,R.Aldini ,1981 .Holdin go fcrushe d tomatoesbefor e finishingan d concentration.:effect so nth econsistenc y ofth econcentrat eproduce d asa functio no f thetomat ovarietie sused .Industri aconserv e56 :10-13 . Lowry,O.H. ,N.J .Rosebrough ,A.L .Far r& R.J .Randall ,1951 .Protei nmeasurement swit hth e folinpheno lreagent .J .Biol .Chem .193 :265-275 . Luh,B.S. ,S .Leonar d& W.Dempsey ,1954a .Pecti csubstance so fPearso nan dSa nMarzan o tomatoes.Foo dRes .19 :146-153 . Luh,B.S. ,W.H .Dempse y& S .Leonard ,1954b .Consistenc yo fpaste san dpure efro mPearso nan d SanMarzan otomatoes .Foo dTechnol .8 :576-579 . Luh,B.S. ,S.J .Leonar d& H.J .Phaff ,1956 .Hydrolysi so fpecti cmaterial san doligouronide sb y tomatopolygalacturonase .Foo dRes .21 :448-455 . Luh,B.S. ,F .Villarreal ,S.J .Leonar d& M .Yamaguchi ,1960 .Effec to fripenes sleve lo n consistency ofcanne d tomatojuice .Foo dTechnol .14 :635-639 . Luh,B.S .& H.N .Daoud ,1971 .Effec to fbrea ktemperatur ean dholdin g timeo npecti nan dpecti c enzymesi ntomat opulp .J .Foo dSci .36 :1039-1043 . Luh,B.S. ,M .Mushref i& Z .Wang ,1981 .Carbohydrat epolymer si nprocessin gtomatoes .Projec t T8,progres srepor t1 t oth eCaliforni aLeagu eo fFoo dProcessors . 196

Luh,B.S. ,M .Mushrefi ,M.A .Sarha n& Z .Wang ,1982 .Carbohydrat epolymer si nprocessin g tomatoes.Projec t18 ,Progres srepor t2 t oth eCaliforni aLeagu eo fFoo dProcessors . Luh,B.S. ,M.A .Sarha n& Z .Wang ,1984a .Pectin san d fibresi nprocessin gtomatoes .Foo d Technol.Aust .36 :70-73 . Luh,B.S. ,M .Mushrefi ,R.Y .Fen g& M.C .Gancedo ,1984b .Carbohydrat epolymer si nprocessin g tomatoes.Projec tT8 ,Progres srepor t3 t oCaliforni aLeagu eo fFoo dProcessors . Luh,B.S. ,M.C .Gancedo ,R.Y .Fen g& J.M .Tsiang ,1984c .Carbohydrat epolymer s inprocessin g tomatoes.Projec t18 ,Progres srepor t4 toth eCaliforni aLeagu eo fFoo dProcessors ,+ supplement febr.21 ,1985 . Malis-Arad,S. ,S .Did i& Y .Mizrahi ,1983 .Pecti csubstances :change si nsof tan d firmtomat o cultivarsan d innon-ripenin gmutants .J .Hort .Sci .58 :111-116 . Mannheim,H.C .& J .Kopelmann ,1964 .Evaluatio no ftw omethod so ftomat ojuic econcentratio n II.Produc tcomparisons .Foo dTechnol .18 :911-914 . Markovic,0 .& A .Slezarik ,1977 .Isolatio nan dcharacterizatio n ofon efor mo ftomat o endo-D-Galacturonase.Collect .Czech .Chem .Commun .42 :173-179 . Markovic,0. ,R .Koh n& 0 .Luknar ,1981 .Preparatio no fhigh-molecular-weigh t pecticaci db y tomatopecti nesterase .Collect .Czech .Chem .Commun .46 :266-269 . Markovic,0 .& A .Slezarik ,1981 .Th emod eo factio no fth eisolate d fromo ftomat o endo-D-galacturonanase.Collect .Czech .Chem .Commun .46 :525-531 . Markovic,0. ,E .Machov a& A .Slezarak ,1983 .Th eactio no ftomat oan dAspergillu sfoetidu s pectinesteraseso noligoraeri csubstrate sesterifie dwit hdiazomethane .Carbohydr .Researc h 116:105-111 . Marsh,G.L. ,J .Buhler t& S .Leonard ,1977 .Effec to fdegre eo fconcentratio nan do fhea t treatmento nconsistenc y oftomat opaste safte rdilution .Journa lo fFoo dProcessin gan d Preservation1 :340-346 . Marsh,G.L. ,S.J .Leonard ,J.E .Buhler t& M.R .Vilas ,1979a .Yiel d andqualit yo fcatsu p produced toa standar d solidsan dconsistenc y level.I Metho do fdeterminin gth eamoun to f tomatosolid srequired .Journa lo fFoo dProcessin gan dPreservatio n3 :189-193 . Marsh,G.L. ,S.J .Leonar d& J.E .Buhlert ,1979b .Yiel d andqualit yo fcatsu pproduce d toa standard solidsan dconsistenc y level.I IInfluenc eo fhandlin gpractices ,brea k temperaturean dcultivar .Journa lo fFoo dProcessin gan dPreservatio n 3:195-212 . Marsh,G.L. ,J.E .Buhler t& S.J .Leonard ,1980 .Effec to fcompositio nupo nBostwic kconsistenc y oftomat oconcentrate .J .Foo dSci .45 :703-706 ,710 . Marsh,G.L. ,J.E .Buhlert ,T.K .Wolcott ,J.R .Hei l& S.J .Leonard ,1982 .198 1tomat ovariet y processing trials.Univ .California ,Davis ,USA . Marsh,G.L. ,S.J .Leonard ,T.K .Wolcot t& J.R .Heil ,1983 .198 2tomat ovariet yprocessin g trials.Univ .California ,Davis ,USA . Marsh,G.L. ,T.K .Wolcott ,J.R .Hei l& R.L .Merson ,1984 .198 3tomat ovariet yprocessin g trials.Univ .California ,Davis ,USA . McColloch,R.J .& Z.I .Kertesz ,1948 .A nunusuall yhea tresistan tpectolyti c factorfro m tomatoes.Arch .Biochem .17 :197-19 9 McColloch,R.J. ,B.W .Nielse n& E.A .Beavens ,1950 .Factor sinfluencin gth equalit yo ftomat o paste.I IPecti cchange sdurin gprocessing .Foo dTechnol .4 :339-343 . McColloch,R.J. ,G.J .Kelle r& E.A .Beavens ,1952 .Factor sinfluencin gth equalit yo ftomat o products.I Surfac elocalize dpecti cenzyme sinactivate d byblanching .Foo dTechnol .6 : 197-199. McCready,R.M. ,E.A .McCom b& E.F .Jansen ,1955 .Th eactio no ftomat oan davocad o polygalactaronase.Food .Res .20 :186-191 . McNeil,M. ,A.G .Darvill ,S.C .Fr y& P .Albersheim ,1984 .Structur ean d functiono fth eprimar y cellwall so fplants .Ann .Rev .Biochem .53 :625-663 . Merson,R.L. ,G .Parede s& D.B .Hosaka ,1980 .Concentratin g fruitjuice sb yrevers eosmosis . Polym.Sci .Technol .13 :405-413 . Miers,J.C. ,J.R .Wagne r& D.W . Sanshuck,1967 .Consistenc y oftomat oproducts .I IEffec to fp H duringextractio no ntomat ojuic econsistency .Foo dTechnol .21 :923-926 . 197

Miers,J.C. ,J.R .Wagner ,M.D .Nutting ,W.G .Schultz ,R .Becker ,H.J .Neumann ,W.C .Dietric h & D.W. Sanshuck,1971 .Fiel dprocessin go ftomatoes .2 Produc tqualit yan dcomposition .J . Food Sci.36 :400-404 . Mizrahi,Y ,R .Zoha r& S .Malis-Arad ,1980 .NaCl-induce dripenin go fth enon-ripenin gtomat o mutantnor .PI .Physiol .(Suppl. )65:83 . Moresi,M .& C .Liverotti ,1982 .Economi c studyo ftomat opast eproduction .J .Foo dTechnol . 17:177-192 . Moshrefi,M .& B.S .Luh ,1984 .Purificatio nan dcharacterizatio n oftw otomat o polygalacturonase isoenzymes.J .Foo dBiochem .8 :39-54 . Moyer,J.C. ,W.B .Robinson ,J.R .Ransford ,R.L .L aBell e& D.B .Hand ,1959 .Processin g conditionsaffectin g theyiel do ftomat ojuice .Foo dTechnol .13 :270-272 . Nakagawa,H. ,Y .Yanagaw a& H .Takehana ,1970 .Studie so nth epectolyti cenzyme .Par tIV . Purificationo ftomat ofrui tpectinesterase .Par tV .Som epropertie so fth epurifie d tomatopectinesterase .Agr .Biol .Chem .34 :998-1003 . Nelson,P.E .& D.K .Iressle r (Eds),1980 .Frui tan dvegetabl ejuic eprocessin gtechnology ,AV I PublishingCo. ,Westport ,Connecticut ,USA ,p p548-573 . Orbell,J.A. ,1980 .Asepti cbag-in-bo xpackaging .Foo dTechnol .9 :56-57 . Patel,D.S .& H.J .Phaff ,1960a .Studie so nth epurificatio no ftomat opolygalacturonase .Foo d Res.25 :37-46 . Patel,D.S .& H.J .Phaff ,1960b .Propertie so fpurifie d tomatopolygalacturonase .Foo dRes .25 : 47-57. Pharr,D.M .& D.B .Dickinson ,1973 .Partia lcharacterizatio n ofC xcellulas ean dcellobias e fromripenin gtomat ofruit .PI .Physiol .51 :577-583 . Pharr,D.M. ,H.N .So x& W.B .Nesbitt ,1976 .Cel lwall-boun d nitrophenylglycosidases oftomat o fruit.J .Amer .Soc .Hort .Sci .101 :397-400 . Polacsek-Racz,M .& K .Pozsar-Hajnal ,1976 .Determinatio no fpectinmethylesterase , polygalacturonase andpecti csubstance si nsom efruit san dvegetables .Par tI I- Stud y intoth epectolyti c tissueenzyme so fJonatha napple san d someothe rfruit san d vegetables.Act aAlimentari a5 :189-204 . Poovaiah,B.W .& A .Nukaya ,1979 .Polygalacturonas ean dcellulas eenzyme s inth enorma lrutger s andmutan tri ntomat ofruit san d theirrelationshi p toth erespirator yclimacteric .PI . Physiol.64 :534-537 . Pozsar-Hajnal,K .& M .Polacsek-Racz ,1975 .Determinatio no fpecti nmethylesterase , polygalacturonase andpecti csubstance si nsom efruit san dvegetables .Par tI - Stud yint o thepectolyti cenzym econten to ftomatoes .Act aAlimentari a4 :271-280 . Pressey,R .& J.K .Avants ,1971 .Effec to fsubstrat e sizeo nth eactivit yo ftomat o polygalacturonase.J .Foo dSci .36 :486-489 . Pressey,R .& J.K .Avants ,1972 .Multipl e formso fpectinesteras e intomatoes .Phytochemistr y 11:3139-3142 . Pressey,R .& J.K .Avants ,1973 .Tw o formso fpolygalacturonas e intomatoes .Biochem .Biophys . Acta309 :363-369 . Pressey,R. ,1977 .Enzyme sinvolve d infrui tsoftening .AC SSymposiu m series47 :172-191 . Pressey,R .& J.K .Avants ,1982 .Solubilizatio no fcel lwall sb ytomat opolygalacturonases : effectso fpectinesterases .J .Foo dBiochem .6 :57-74 . Pressey,R. ,1983 .(3-galactosidase si nripenin g tomatoes.PI .Physiol .71 :132-135 . Pressey,R .& D.S .Himmelsbach ,1984 .1 3C-n.m.r .spectru mo fa D-galactose-ric hpolysaccharid e from tomatofruit .Carbohydr .Res .127 :356-359 . Randall,J.M. ,R.A .Carlson ,R.P .Graha m& A.I .Morga njr .1966 .Yield stomat opast ewit h 50% solids.Foo dEngineerin g 38:168-169 . Reeve,R.M. ,V.F .Kaufma n& F.F .Wong ,1959 .Microscopi c structurean dreconstitutio no f vacuum-dried tomatopowders .Foo dTechnol .13 :561-566 . Rexova-Benkova,L. ,0 .Markovi c& M.J .Foglietti ,1977 .Separatio no fpecti cenzyme sfro m tomatoesb yaffinit ychromatograph yo ncross-linke d pecticacid .Collect .Czech ,chem . Commun.42 :1736-1741 . 198

Rick,CM. , 1978.Th eTomato .Scientifi cAmerica n239 :66-76 . Rigney,C.J .& R.B.H .Wills ,1981 .Calciu mmovement ,a regulatin gfacto ri nth einitiatio no f tomatofrui tripening .Hor tScienc e16 :550-551 . Robinson,W.B. ,E .Stot z& Z.I .Kertesz ,1954 .Th eeffec to fmanufacturin gmethod so nth e ascorbicaci dconten tan dconsistenc y characteristicso ftomat ojuice .J .Nutr .30 :435-442 . Robinson,W.B. ,L.B .Kimbal l& J.R .Ransford ,1956 .Factor saffectin g thedegre eo fsettlin gi n tomatojuice .Foo dTechnol .10 :109-112 . Robinson,D.C. ,1977 .In :Advance s inbotanica lresearch ,H.W .Woolhous e (Ed.),5 :89-151 . AcademicPress ,London . Rombouts,F.M. ,1986 .T ob epublished . Roozen,J.P .& R .Ouwehand ,1978 .Half-automatisch ebepalin gva nstiksto f indrog e levensmiddelen.Voedingsmiddelentechnologi e11 :23-25 . Roy,S.K .& B .Choudhury ,1972 .Studie so nphysico-chemica l characteristicso fa fe wvarietie s oftomat o(Lycopersico nesculent ™mill. )i nrelatio nt oprocessing .J .Foo dSci . Technol.,Indi a9 :151-153 . Rumsey,T.R. ,T.T .Conant ,T .Fortis ,E.P .Scott ,L.D .Pederse n& W.W .Rose ,1984 .Energ yus e intomat opast eevaporation .Journa lo fFoo dProces sEngineerin g 7:111-121 . Rushing,J.W .& D.J .Huber ,1984 .I nvitr ocharacterizatio n oftomat ofrui tsoftening .Th eus e ofenzymicall yactiv ecel lwalls .PI .Physiol .75 :891-894 . Rushing,J.W .& D.J .Huber ,1985 .Initiatio no ftomat ofrui tripenin gwit hcopper .J .Amer . Soc.Hort .Sci .110 :316-318 . Saeman,J.F. ,W.E .Moor e& M.A .Millett ,1963 .Suga runit spresent .In :Method s incarbohydrat e chemistry,II ICellulose ,sectio nII .chemica lanalysis ,R.L .Whistle r (ed.),Academi c Press,London ,p p54-70 . Sawamura,M. ,E .Kneg t& J .Bruinsma ,1978 .Level so fendogenou sethylene ,carbo ndioxid ean d solublepectin ,an dactivitie so fpectinmethylesteras ean dpolygalacturonas e inripenin g tomatofruits .Plan tan dCel lPhysiol .19 :1061-1069 . Schramm,G. ,1981 .Introductio n topractica lviscometry .Gebrude rHaak eGmbH ,Karlsruhe ,F.R.G . Selby,K .6 ,C.C .Maitland ,1967 .Th ecellulas eo fTrichoderm aviride .Separatio no fth e components involved inth esolubilizatio no fcotton .Biochem .J .104 :716 . Selvendran,R.R. ,J.F .Marc h& S.G .Ring ,1979 .Determinatio no fAldose san duroni caci d contento fvegetabl e fiber.Analyt .Biochem .96 :282-292 . Sherkat,F .& B.S .Luh ,1976 .Qualit y factorso ftomat opaste smad ea tsevera lbrea k temperatures.J .Agric .Food .Chem .24 :1155-1158 . Sherkat,F .& B.S .Luh ,1977 .Effec to fbrea ktemperatur eo nqualit yo fpaste ,reconstitute d juice,an d saucemad efro mM-3 2tomatoes .Canadia nInstitut eo fFoo d ScienceTechnolog y 10:92-96 . Shomer,I. ,P .Lindne r& R .Vasiliver ,1984 .Mechanis mwhic henable sth ecel lwal l toretai n homogenousappearanc eo ftomat ojuice .J .Foo dSci .49 :628-633 . Sidhu,J.S. ,V.K .Bhumbl a& B.C .Joshi ,1984 .Preservatio no ftomat ojuic eunde raci d conditions.J .Sci .Foo dAgric .35 :345-352 . Siliha,H.A.I. ,1985 .Studie so nclou d stabilityo faprico tnectar .Ph.D .Thesis ,Agricultura l University,Wageningen ,Th eNetherlands . Sims,W.L. ,1980 .Histor yo ftomat oproductio naroun d theworld .Act aHorticultura e 100:25 , 26. Skierkowski,K .1982 .Trend s inth edevelopmen to ftomat oconcentrat emanufacture .Przemys l spozywczy36 :135-137 . Smit,C.J.B .& B.K .Nortje ,1958 .Observation so nth econsistenc y oftomat opaste .Foo d Technol.12 :356-358 . Sobotka,F.E .& A.E .Watada ,1971 .Cellulas e inhig hpigmen tan dcrimso ntomat ofruit .J .Amer . Soc. Hort.Sci .96 :705-707 . Sobotka,F.E .& D.A . Stelzig,1974 .A napparen t cellulasecomple xi ntomat o (Lycopersicon esulentumL. )fruit .PI .Physiol .53 :759-763 . 199

Southgate,D.A.T. ,B .Bailey ,E .Collinso n& A.F .Walker ,1976 .A guid et ocalculatin g intakes ofdietar y fibre.J .Huma nNutritio n30 :303 . Spiro,R.G. ,1966 .Analysi so fsugar sfoun d inglycoproteins .Method s inEnzymol .8 :4-5 . Stein,E.R .& E .Brown ,1975 .Ge l filtrationan ddis cge lelectroforesi so ftomat opecti c substance.J .Agr .Foo dChem .23 :526-529 . Stephens,T.S. ,G .Saldana ,F.P .Griffith s& P.W .Leeper ,1970 .Pecti c substancesan d viscositieso fjuice sprepare d fromChico ,Chic oGrande ,L aBonit aan dHomestead-2 4 varietieso ftomatoes .J .Ri oGrand eValle yHortic .Soc .24 :139-145 . Stevens,M.A .& K.N .Paulson ,1976 .Contributio no fcomponent so ftomat ofrui talcoho l insolublesolid s togenotypi cvariatio n inviscosity .J .Amer .Soc .Hort .Sci .101 :91-96 . Stevens,B.J.H .& R.R .Selvendran ,1980 .Th eisolatio nan danalysi so fcel lwal lmateria lfro m thealcohol-insolubl e residueo fcabbag e (Brassicaolerace avar .capitata) .J .Sci .Foo d Agric.31 :1257-1267 . Stevens,B.J.H .& R.R . Selvendran,1984 .Structura l featureso fcel lwal lpolysaccharide so f thecarro tDaucu scarota .Carbohydr .Res .128 :321-333 . Stier,E.F. ,CO . Ball& W.A .Maclinn ,1956 .Change si npecti c substanceso ftomatoe sdurin g storage.Foo dTechnol .10 :39-43 . Strand,L. ,C .Rechtori s& H .Mussel ,1976 .Polygalacturonase s releasecell-wall-boun d proteins.PI .Physiol .58 :722-725 . Sulc,D .& D .Ciric ,1968 .Herstellun gvo nFrucht -un dGeraiisemar kKonzentraten .Flussige sObs t 35:230-236 . lalmadge,K.W. ,K .Keegstra ,W.D .Baue r& P .Albersheim ,1973 .Th estructur eo fplan tcel l wallsI .Th emacromolecula rcomponent so fth ewall so fsuspension-culture d sycamorecell s witha detaile d analysiso fth epecti cpolysaccharides .PI .Physiol .51 :158-173 . Takada,N .& P.E .Nelson ,1983 .Pectin-protei n interaction intomat oproducts .J .Foo dSci .48 : 1408-1412. Takahena,H. ,T .Shibuya ,H .Nakagaw a& N .Ogura ,1977 .Purificatio nan dsom epropertie so f endo-PG fromtomat opericarp .Techn .Bull .Fac .Hort .Chib aUniv .n o25 :29-34 . Themmen,A.P.N. ,G.A .Tucke r& D .Grierson ,1982 .Degradatio no fisolate d tomatocel lwall sb y purifiedpolygalacturonas e invitro .PI .Physiol .69 :122-124 . Thibault,J.F. ,1979 .A nautomate dmetho d forth edeterminatio no fpecti csubstance si ncherr y fruits.Lebensm .-Wiss .u .-Technol .12 :247-251 . Tighelaar,E.C. ,W.B .McGlasso n& R.W .Buescher ,1978 .Geneti cregulatio no ftomat ofrui t ripening.Hortscienc e13 :508-513 . Tucker,G.A. ,N.G .Robertso n& D .Grierson ,1980 .Change si npolygalacturonas e isoenzymes duringth eripenin go fnorma lan dmutan ttomat ofruit .Eur .J .Biochem .112 :119-124 . Tucker,G.A. ,N.G .Robertso nan dD .Grierson ,1981 .Th econversio no ftomat ofrui t polygalacturonaseIsoenzym e2 int oIsoenzym e1 i nvitro .Eur .J .Biochem .115 :87-90 . Tucker,G.A. ,N.G .Robertso n& D .Grierson ,1982 .Purificatio n andchange si nactivitie so f tomatopactinesteras e isoenzymes.J .Sci .Foo dAgric .33 :396-400 . Tucker,G.A .& D .Grierson ,1982 .Synthesi so fpolygalacturonas edurin gtomat ofrui tripening . Planta155 :64-67 . Twigg,B.A. ,1959 .Consistenc yan d serumseparatio no fcatsup .Ph.D .Thesis ,Univ .o f Maryland,Colleg ePark ,USA . Updegraff,D.M. ,1969 .Semimicr odeterminatio no fcellulos ei nbiologica lmaterials .Analyt . Biochem.32 :420-424 . Vas,K ,M ,Nedbalek ,H .Scheffe r& G .Kovacs-Proszt ,1967 .Methodologica l investigationso nth e determinationo fsom epecti cenzymes .Fruchtsaft-Ind .12 :164-184 . Vijayalakshmi,M.A. ,E .Segard ,F.M .Rombout s& W .Pilnik ,1978 .Conditio n forcross-linkin go f sodiumpectat ewit hepichlorohydrin .Lebensm .-Wiss .u .-Technol .11 :288-290 . Voragen,A.G.J. ,1972 .Characterizatio n ofpecti nlyas eo npectin san dmethy l oligogalacturonates.Ph.D .Thesis ,Agric .Univ. ,Wageningen ,Th eNetherlands . Voragen,F.G.J. ,R .Heutin k& W .Pilnik ,1980 .Solubilizatio no fappl ecel lwall swit h polysaccharide-degrading enzymes.J .Appl .Biochem .2 :452-468 . 200

Voragen,A.G.J. ,F .Geers t& W .Pilnik ,1982 .Hemicellulase si nenzymi c fruitprocessing .In : Useo fenzyme s infoo d technology p497 .Lavoisier ,Pari sCede x08 . Voragen,F.G.J. ,J.P.J .Timmers ,J.P.H .Linsen ,H.A .Schol s& W .Pilnik ,1983 .Method so f analysis forcell-wal lpolysaccharide s of fruitan dvegetables .Z .Lebensm .Unters . Forsch.177 :251-256 . Voragen,A.G.J. ,H.A . Schols,H.A.I .Silih a& W .Pilnik ,1986 .Enzymi c lysiso fpecti c substances incel lwalls .Som e implications forfrui tjuic e technology.AC S Symposium Series (inpress) . Voragen,A.G.J. ,H.A . Schols& W .Pilnik ,1986a .Determinatio n ofdegre eo fmethylatio nan d acetylation ofpectin sb yH.P.L.C .Accepte d forpublicatio n inFoo dHydrocolloids . Voragen,A.G.J. ,H.A .Schol s6 .W .Pilnik ,1986b .Structura l featureso fth ehemi-cellulosi c polymerso fapples .Accepte d forpublicatio n inZ .Lebensm .Unters .Forsch . Voragen,A.G.J. ,H.A .Schols ,M.F .Searle-v .Leeuwen ,G .Beldma n& F.M .Rombouts ,1986c . H.P.L.C.analysi so foligomeri can dmonomeri csaccharide s fromenzymaticall ydegrade d polysaccharides.Accepte d forpublicatio n inJ .Chromatography . Voragen,A.G.J. ,1986d .T ob epublished . Vries,J.A .de ,A.G.J .Voragen ,F.M .Rombout s& W .Pilnik ,1981 .Extractio n andpurificatio no f pectins fromalcoho l insoluble solidsfro mrip ean dunrip eapples .Carbohydr .Polymer s1 : 117-127. Vries,J.A .de ,C.H .de nUijl ,A.G.J .Voragen ,F.M .Rombout s& W .Pilnik ,1983 .Structura l featureso fth eneutra l sugarsid echain so fappl epecti c substances.Carbohydr .Polymer s 3:193-205 . Wagner,J.R .& J.C .Miers ,1967 .Consistenc y oftomat oproducts .I Th eeffect so ftomat oenzym e inhibitionb yadditives .Foo dTechnol .21 :920-923 . Wagner,J.R. ,J.C .Miers ,D.W . Sanshuck& R .Becker ,1968 .Consistenc y oftomat oproducts .I V Improvement of theacidifie dho tbrea kprocess .Foo dTechnol .22 :1484-1488 . Wagner,J.R. , J.C.Miers ,D.W .Sanshuc k& R .Becker ,1969 .Consistenc y oftomat oproducts .V Differentiation ofextractiv ean denzym e inhibitory aspectso fth eacidifie dho tbrea k process.Foo dTechnol .23 :247-250 . Wallner,S.J .& J.E .Walker ,1975 .Glycosidase s incel lwall-degradin g extractso fripenin g tomatofruits .PI .Physiol .55 :94-98 . Wallner,S.J .& H.L .Bloom ,1977 .Characteristic so ftomat ocel lwal ldegradatio n invitro . Implications forth estud yo ffruit-softenin g enzymes.PI .Physiol .60 :207-210 . Whittenberger,R.T .6 cG.C .Nutting ,1957 .Effec to ftomat ocel lstructure so nconsistenc yo f tomatojuice .Foo dTechnol .11 :19-22 . Whittenberger,R.T .& G.C .Nutting ,1958 .Hig hviscosit y ofcel lwal l suspensionprepare d from tomatojuice .Foo dTechnol .12 :420-424 . Williams,K.T .& A .Bevenue ,1954 .Som ecarbohydrat ecomponent s oftomatoes .J .Agr .Food . Chem.2 :472-474 . Wills,R.B.H .& G.J .Rigney ,1979 .Effec t ofcalciu mo nactivit yo fmitochondri a andpecti c enzymes isolated fromtomat ofruits .Foo dBiochem .3 :103-110 . Woodmansee,C.W. ,J.H .McClendo n& G.F .Somers ,1959 .Chemica lchange sassociate dwit hth e ripeningo fapple san d tomatoes.Foo dRes .24 :503-514 . Xu,S-Y. ,C.F .Shoemake r& B.S .Luh ,1986 .Effec to fbrea ktemperatur e onTheologica l propertiesan dmicrostructur eo ftomat ojuice san dpastes .J .Foo dSci .51 :399-407 . York,O.K. ,M .O'Brien ,D .Trombropoulos ,F.H .Winte r& S.J .Leonard ,1967 .Relatio no ffrui t damaget oqualit yan dconsistenc yo ftomat oconcentrates .Foo dTechnol .21 :69-71 . Zainon,M.A .& G.J .Brady ,1982 .Purificatio n andcharacterizatio no fth epolygalacturonas eo f tomatofruits .Aust .J .Plan tPhysiol .9 :155-169 . SAMENVATTINC

Het onderwerp en de doelstellingen van deze studie zijnnade r uitgewerkt inhoofdstu k 1. Inhoofdstu k 2word t de relevante literatuur uitgebreid besproken. Dit literatuuroverzicht omvat devolgend e onderwerpen: deproduktie ,algemen e samenstelling enverwerkin gva n tomaten (2.1.); definities en kwaliteitsparametersvoo r tomatenprodukten (2.2.); de faktoren die bijdragen aan de schijnbareviscositei t van tomatenprodukten (2.3.); de faktoren die de serumseparatie in tomatenprodukten belnvloeden (2.4.)e nd e samenstelling van depolysaccharide n in tomaten (2.5.). Gezien de lengte van 2.3. isee n samenvatting van de,i ndez eparagraaf ,besproke n literatuur weergegeven in 2.3.1. De algemene analytischemethode n diemeerder ekere n tijdenshe t onderzoek zijn gebruiktworde nbesproke n inhoofdstu k 3. Inhe t onderzoek werd extra aandacht besteed aan eenvergelijkin g tussen tweemethode n van viscositeitsmeting enaa n debepalin gva n deneutral e suikers aanwezig in dewateronoplosbar e bestanddelen (WIS)va n tomatenprodukten. DeBostwic k consistometerblee knie t gevoelig genoeg voor hetmete nva n hoog visceuse tomatenprodukten. Voor hetmete nva n tomatenprodukten met een Bostwickwaarde groter dan 5 cmblee k dit instrument echter zeer geschikt. Dehydrolys eva n depolysaccharide n inhe tWI S van tomatenprodukten met 0.8 N H SO, of 2N H.SO,,beide nvoorafgegaa n door eenhydrolys eme t 72%H SO,, resulteerde in een onbevredigende bepalingva n deneutral e suikers. Hydrolyseme t eenmengse lva n pektolytische en cellulolytische enzymen voorafgaand aan eenhydrolys eme t 2N TFA leverde daarentegen goede resultaten op. Het onderzoek naar de invloed van de oplosbare bestanddelen (serumviscositeit)e nd e onoplosbarebestanddele n op de schijnbare viscositeit van tomatensap en concentraten isbeschreve n inhoofdstu k 4.D e schijnbareviscositei t van tomatensap en concentratenwer d voornamelijk bepaald door het WIS gehalte.D e invloed van de tomatenvarieteit,d e "break" temperatuur en demethod eva n concentreren op de schijnbare viscositeit bij eenbepaald e concentratie oplosbare bestanddelen (°Bx)ko n geheelverklaar d worden doorverschille n inWI S gehalte.D e schijnbare viscositeit bij eenbepaald eWI S concentratiewer d nietbelnvloe d door de tomatenvariSteit,d e serumviscositeit,d e"break"temperatuu r en demethod e van concentreren tenzij dezeproduktiemethode n leidden tot afbraakva n het WIS. De resultaten vanhe t onderzoeknaa r de samenstelling en strukturele kenmerkenva nhe tWI Sworde nbesproke n inhoofdstu k 5.WI S fragmenten,i n oplossing gebracht metbehul pva n extraktiemiddelen zoalsammoniumoxalaat , HC1 enNaO H ofme tbehul p van gezuiverde enzymenwerde n gekarakteriseerd d.m.v.gelpermeatie -e n ionenwisselingschromatografie,analys eva n de neutrale suikers enbepalin g van deglycosidisch e bindingen. Inhe t algemeenko nhe tWI S gefraktioneerd worden in 40-45% pektine, 25-30% hemicellulose en 30-35% cellulose.D e pektinefraktie bevatte grote hoeveelheden galacturonzuur (AGA)e n eiwit enbovendie n ongeveer 25-35% van het arabinose engalactos e aanwezig inhe tWIS .He tmerendee lva n dit arabinose en galactosebevon d zich in deHC 1 oplosbare pektinefraktie en maakte deelui tva n een sterkvertak t pektinesegment dat alleenvi a hoog veresterde AGA ketens aan de celwandmatrixgebonde nwas . Een groot gedeelteva nhe t pektine kwamvoo r in fragmenten die slechts in geringemat e vertaktware n met arabinose en galactose bevattende zijketens. Deveresteringsgraa d van deze fragmenten en deverdelin g van de estergroepen was zodanig dat zowelP G alsoo kP L in staatware n deze fragmenten af tebreken . Xylanen,arabinane n en (arabino)galactanen blekenbetrokke n te zijnbi j de binding vanhe t pektine aan de celwandmatrix. De tomatenvarieteitendi e bestudeerd werdenverschilde nvoora l inhe t WIS gehalte en ind e hoeveelheid arabinose en galactose aanwezig inhe tWIS .D everschille n in arabinose-e n galactosegehalte kwamenhe t duidelijkst tot uiting in de HCl-oplosbarepektinefraktie .D e samenstelling van deze pektinefraktie bleek sterk teverandere n tijdens deprodukti eva n tomatensap.He t tomatenhemicellulosebevatt e ongeveer 80%va n hetWIS-xylose . 60%va n deze hoeveelheid was aanwezig in devor mva n g-1,4-xylanen,di e waarschijnlijk verbondenware nme t het pektine.30 %va n dehemicellulose-xylos e hoeveelheid bleekvoo r tekome n in devor mva n fragmentenme t een samenstelling kenmerkend voorxyloglucanen .He t hemicellulose bevatte verder ± 50%va nhe tWIS-arabinos e enWIS-galactos e enbovendie n 15-20% van hetWIS-glucose . Alle arabinose enhe t merendeelva n de galactose residuen bleken eindstandig inhe t hemicellulose voor tekomen .D emogelijk e aanwezigheid inhe themicellulos eva n (gluco)mannanen en een geringe hoeveelheid galactanenwer d aangegeven. De cellulosefraktiebeston d voornamelijkui t glucose,maa rbevatt e tevens ongeveer 15%va nhe t WIS-arabinose. Inhoofdstu k 6worde n de resultatenweergegeve nva nhe t onderzoek naar de relatievebijdrag eva n deWI S bestanddelen aan de schljnbareviscositel t van een tomatensapconcentraat en eenverdund e tomatenpasta. Commerciele en zuivere,goe d gekarakteriseerde enzymenwerde n gebruikt om de schljnbareviscositei tva n eenverdund e tomatenpasta en een 10°Bx tomatensapconcentraat teverlagen . De resultatenwezen , onafhankelijkva nhe t gebruikte tomatenprodukt,o p een belangrijke bijdrageva n dehoge rveresterd e gedeelten inhe t WIS pektine aan de schljnbare viscositeit.He twer d waarschijnlijk geacht dat deze pektinefragmenten identiekware n aan depektinefragmente ngevoeli g voor afbraak door zowelP G alsPL ,zoal sbeschreve n inhoofdstu k 5.D e bijdrage van het laagveresterde pektine enhe t hemicellulose aan de schljnbare viscositeit bleef onduidelijk; debijdrag e van het eiwit was te verwaarlozen. Celluloseblee k een relatief kleinere bijdrageaa n de schljnbare viscositeit te leveren danhe t veresterde pektine.D e starheid van de celwanden,veroorzaak t door de struktuur van cellulose scheen de schljnbare viscositeit bij hogere WIS gehaltes tebelnvloeden . De conclusiewer d getrokken dat deklein everschille n inWI S samenstelling, zoals dieva n nature gevondenwerde n in debestudeerd e tomatenvariSteiten, geen aanleiding konden geven tot groteverschille n in de schljnbare viscositeit. De schijnbare viscositeit van tomatensap en concentraten werd daarom voornamelijk bepaald doorhe tWI S gehalte,da twaarschijnlij k evenredig is met dehoeveelhei d tomatencellen. De invloed van procesomstandigheden zoals:"ho t break" temperatuur werd toegeschreven aanhe t effektiever verwijderenva n hetWI S van de schil.D e kwaliteit vanhe t WIS enhe tmicroscopisc h uiterlijkva n de tomatencellenwer d niet beinvloed doorhoger e hotbrea k temperaturen. De invloed ophe t passerenva n de "hotbreak " temperatuur,d e temperatuur tijdenshe t passeren en de diameterva n de openingen in de passeerzeef, hing nauw samenme t de effectiviteitva n de passeermachine. De schijnbare viscositeit van concentraten van tomatensap bleek niet belnvloed teworde n door demethod eva n concentreren noch doorhe t type verdamper dat gebruiktwerd . Inhe t algemeenveranderd e de verhouding WIS/TSnauwelijk s tijdenshe t concentreren.He t gelijktijdig concentreren vanbepaald e serumcomponenten (waarschijnlijk ionen) en tomatencellen tijdenshe tverdampe n bleek tot eenviscositeitsverlie s te leiden na verdunning van depast a ("dilution-loss"). De grootteva nhe t "dilution-loss"wer d belnvloed doormechanisch ekrachte n envacuu m tijdens hetverdampe n en door eenhittebehandelin g vanhe t geconcentreerde sap.D e "hotbreak " temperatuur en deverdamptemperatuu rbleke n geen invloed op de grootteva n het "dilution-loss" tehebben .He t "dilution-loss"blee k tevens belnvloed teworde n door dekwalitei t vanhe t sapvoorafgaan d aan het concentreren.He t concentrerenva n tomatensap metbehul p van de "centrifugeren serum concentreren"method e voorkwam het "dilution-loss" wanneer gebruikwer d gemaakt van een tafelcentrifuge en reduceerde het "dilution-loss"wannee r gebruikwer d gemaaktva n een decanter. De schijnbare viscositeit van tomatensap en concentratenwer d nadelig belnvloed zoweldoo r een conserverende hittebehandeling als ook door bevriezing.