. : 4 . . . ...1.. 5: .. .5.... .523! £232.22: Ext—Sang Ruhr. 1.. 3113-... zfi:.:.::.’zl?azsaxa:rsakzpaiathns lmyfisx Axflg:li~u¢..t. 5L I:,.)..‘DI3 tab!) in . 3. 1...... . ._ tn 2.5.5.1...)z:r|>~.$:u:l§l.§’ 3 a: 3.3 .l.x.:....fi..:l .;S.:.llatas§5!.:i . :5: '15:. 1. .21.! alarfiit..395r.vl..a.. .391..¢L{u5>._afull—s:5;,iisa..£ut.xt)..l 065} t i ............ K .5 agr)1vxd.:;:ah9ra:.fl anti... .Jthvtaaaazs...‘ taillalgi:!§§z$~hflri . . .. . :3... . . .13. 5”; I... Q .Akg..1sa..lal.vfila$ $1. at. 2.212. £13 iifiiii {at}!!! I)? ‘21.:3?.iiv:xnxxx>-yi..11... H. ~ 3 I... C tibidnursvldclt - (If... “753535193533 )DI..:9 Anti): .. . 533} «1.39. it! 353.43an..{Euteaaosxfi‘fcufi .IAS..|fxltab.a-t.. .3...» «(Warluhi ;l.lu)!vb.§¢flfl ..... s I 3. .5:..$:J.:z$v.o,.l31:: 3 a :35. :Iaijlkt‘. 1333.03.92 .2 ibis!!! I. 92’... x. J . 3 ............ L13. ..... I Yahvlk.12aas...i:: .Eltxll::lk3zt iit._a.wNv.-kfiz..rnti5§ a :9? "3.2.2.“: ha..)¢.€.~«.st.inaia iazx3zxasxs>iilu4§3 5.4..)32.«a%.):>3 tulnaiaf. .9-‘|.AAI¢3 9.! . iii a. 12-33.? anur’wwliahvfiwflurcnih.‘;1~5.Ru.astu.a..at}b £82:A3I!.§933I§?.3§ 35.}..Siagii :. 9.5. 3.3.. . . u u .2 .. . «fittufilNfiik 5)..u.33|2‘1VWfl.M-15:rtt\.lvo.t Shirl]??? : “132.1333... 1.2.1.3.... 5:13.05 . a. It on!!!) llsifiidi :35 .1. 35.31.51,! i?.¥N.}:IISOI $3.1: 55:55.): . t . . .. .v. 3 All My ‘13.... t. a la . . 1N... svfisbvn.»bayrfiufi-rat)fla.§‘fi§f xx a 1 us .3: . z I. 3.379.151.2525‘71v! :ailuitvaAVQE321z 3...! 15:11: a: 111.73... 351....) 6666666 .wlllaiit’as 11:35.13 .2 a 3 . 53...: 3:31....)1 ...Z.§..)l§1§§.r.§) : 1:151lag3 2 s. I, Ottaaiil 11:0 {zajgzta‘l .1. 9.1.53) .. iv... : a 3.1. I is. I: z . l...“- .. ai’..._:ahw§ ILA)..~:.. it». OBI" gfila‘btsail :3 b 9 . 1:011 .12: t5..ipl...:fil¢..{.:: giin. Iifxlzstltatisi’. tat-991...... :K...:;..3:5 .u...l§,...;..,3zuflfi.xu!fivfi 233.... 21.2. :39» "$2.... .5... a . xxxxxx .. l. . . 3):; ..... 15.71.3351 I..ac~qli §§I§3§SA¢).'335I$2D£:\5) t‘atifiufl : .:..!.§v.)l:1.9 1.3:?) x:.-: gyu!..t|tlfiai txt3fl..fla}fiizd’za )Qallt... s. .2 1r... .. Jaw-0.5... std... f. .5.. . :I x: s abdyift. furrzifi . :5...:._,..arab.1!.b :..s....l.~.lsb>...ifi L. L... um“... .4 than . . 3 692.321.)... :: .. I I: it: I) a. x 3“. .3f::':.: :3 1"! a) . 53...... at. at 1;..5...ufluu.r aladduvs-i¢gttatlx~uol l nail $3.0“..lirlti‘l 3 a I: I. :9 332......301. 2:. tacit 3' If). .. :itlbisz .19): .1273: . ESQJPJ. ? z. 3-. U: u ‘35. JV!!!) 3):}..1‘932. mirntklb. £11.55! sglaaai’fiuflu. 1 . . . . 3.927 ’7 )Nr‘vfivt Iratllirlaa ..1:’1:I§‘.1I}%§§§ 4;- 3| v.12 .IAiIfiIa 9‘5. 7.. 333A . . . :g . .. 38:?{hfiu‘d‘ahete rpflflcxg§§§.zvs D. a: 9) J a). .3. k.§:3|...oiflcs Digit: . .330“ :0... Iii). . w k. lilliligta . u .. . . a .53.." \. AI:§.§1) .. I: k .- 99.“...finuuqqzfivfin kg... his... $.2.n.3.:..§4.....b.2s::3... n m . 1... $95.31.... . 8E Lfitswz.2~§2\3:¢9§ al>1~u1hr~l‘ra It: . . asgrgé . ‘gfllhi . I8...- . in 4:32;: :. A Izltu...r~.t.$.s§fl¢i£§fl‘}:adlo II .. it}... n I 1“: § ilk, groin"... lhgfi’légaali . . .. :‘s‘ buixiiiiilflki )5! . gigs): g ...... ulorg 9.53.]: J3»)Ixtl\lb.ifl.lb)$t>§|hu‘zv.g1. «5.8.. . .1. {3.53323; t2:- ~.l..v.. t:£ii:dfitni~5.§{it"rllbtiflfiaeu‘alg: :v..\. 5232...: iiiii ’73)?!) II: I): 151......le 331.1... .73339Ib3313hfil illeilgzlll 333‘... .11 iv.L~UBuiD\sl\3519713-7€ IIIIIIII It! 2319! 3k!!! in". gtkligl1153§3gz2~tx :::::::: v . t. lxfizlia:3;lil 3125.1. :32!§u‘l..2§ It} . fgzglzig’: . LEI): :11 iv-l.a..Is.._‘:..nr:‘i)|l’slt.lf.|:‘nézatila . . .00" . I’llablaDtiil}‘l’5£\.ui§!0>flfl}rlk§l.2 xiii-339.10)... .5315) 1.3.0? All... 1:3,) I) 'a guru; l .p . > a :zgluzzcagioailzgliIJ at; 15 o- 2153 {Vthflvbi ’ahtvia. Ai?!!53)§:£.\3§i6.‘31)133.53:...) 35...”... 3’ .71 1...... 23.,‘5- 3.1%.. 19: i-J-ID..§’II#.§..9U,:5.9 $3.33,}. 3 . 3.3.x}... 3.3 3333.1.hf. u§.£is.2.. 71,3731}!!! 2. 15.]... igézfi’! 1+£=2f£$1115§2§i§r gag . ill 3):]! . .5; . 5345..-: ritual... bit-.. 2. :90!!! .1..!..:(5 .2: stanfntag’al-fia. $3.! a Sagan: 152310.”; . |§I§¥ al.11: .a 1 2... f . i. .u. . . . .. . . a . . . .1 . é . . . fig . A : . legggf .10 x g ‘1; “rot; :‘iggatgkgrpii.étl§..;pu....&21:l..t. 51.x! ; 3.5151355 {Joggfljloéluu .. . . I . ,.. .u...§.§.3... .7 .. .r.....1u.. :: b y 1.2 1.. t 1. MICHIGAN STATE UNIVERSITY LIBRARIES I|I|||I|I||| II 3 1293 10668 3661 ideal» LIBRARY numgan State University rHESIS This is to certify that the thesis entitled Effects of Cooking in Solutions of Varying pH on the Dietary Fiber Components of Vegetables presented by Laura Ann Morabito Brandt has been accepted towards fulfillment of the requirements for L degree in FOOd Science wfiéw L7 Lajor profear DateFebruary 11, 1983 0-7639 MSUiJ an ‘m'nm'i-w ‘ ‘ " ' " ', Institution 1 rr MSU LIBRARIES —r_—_ s. \- zI" V~~e"}g FE§.. f“ ‘W “ox hi3 % 9 $45.11.)“ ‘lAR 1 2199‘ MW i903: “I ans RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. EFFECTS OF COOKING IN SOLUTIONS OF VARYING pH ON THE DIETARY FIBER COMPONENTS OF VEGETABLES By Laura Ann Morabito Brandt A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1983 6/9057? ABSTRACT EFFECTS OF COOKING IN SOLUTIONS OF VARYING pH ON THE DIETARY FIBER COMPONENTS OF VEGETABLES By Laura Ann Morabito Brandt Green beans, peas, potatoes, cauliflower and corn were cooked in buffers of pH 2, 4, 6, and 10. Tenderness of the raw and cooked vege- tables was determined. Individual dietary fiber components were measured on the raw and cooked vegetables as well as in the cooking media. These included water-soluble pectin and hemicellulose, water- insoluble pectin and hemicellulose, cellulose, and lignin. In general, cooked vegetables were toughest at pH 4 and softest at pH 10. Soluble and insoluble pectin decreased at pH 4. At pH 10, soluble hemicellulose increased with a corresponding decrease in insoluble hemicellulose. Fiber components in the cooking solution increased with pH. Cellulose showed no change. The most dramatic effects for fiber were found in cauliflower and beans; corn showed no significant effects. Neutral sugars were quantitated in two non-fibrous and two fibrous fractions. The vegetables showed similar trends in their sugar com- position. A prediction equation for shear force using fiber components as the independent variables showed that insoluble pectin, lignin, and soluble hemicelluloses may be used to predict tenderness in corn, peas, potatoes, green beans and cauliflower cooked at various pH levels. To Kerryn 11 ACKNOWLEDGMENTS I wish to thank my advisor, Dr. Melissa Jeltema, for her support and friendship. I would like to acknowledge the members of my thesis committee: Dr. Mary Ellen Zabik, Dr. Robert Herner, and Dr. Theodore Wishnetsky. Special appreciation goes to Brian Jeltema for his computer expertise and genial cooperation and to Dr. Cress for his statistical assistance. I would like to express my appreciation to my co—workers and to WKAR radio, both of whom have enriched my stay at MSU. I would especially like to thank my mother, for always encouraging me to reach higher goals. Most of all, I thank my husband and best friend, Kerryn, for everything. TABLE OF CONTENTS page LIST OF TABLES ......................... vi LIST OF FIGURES ........................ viii INTRODUCTION .......................... 1 REVIEW OF LITERATURE ....................... 3 Dietary Fiber ........................ 3 Fiber Components: Location and Structure ........ 3 General Effects of Cooking ................. 11 Effects of Cooking in Relation to Vegetable Fiber Components ...................... 12 Pectin .......................... 12 Hemicellulose . . . ................... l6 Cellulose . . ...................... 17 Lignin .......................... 18 Phytin. . . . . . . . .................. 19 Starch .......................... 19 Sensory Aspects of Vegetable Texture ............ 23 EXPERIMENTAL .......................... 25 Vegetable Preparation and Cooking Procedure ......... 25 Analysis of Fibrous and Non-fibrous Vegetable Components .................... 28 Moisture. .-. . . ........ . ........... 28 Ash ........................... 28 Crude Fat ........................ 28 Protein ......................... 29 Fiber Extraction ...................... 30 Sugar Quantitation ..................... 34 Data Analyses ........................ 37 HPLC Preparation and Analysis ................ 38 Preparation of Fractions ................. 40 Materials ........................ 41 Quantification ...................... 43 Equipment ........................ 43 RESULTS AND DISCUSSION ..... . . . . . ..... . . . . . . 45 Rationale for Determining Composition of Fractions from Separation Scheme .............. . . . . . 45 Results of Extraction Procedure ................ Shear Press Data. . . . . ................. Cooking Solution Fiber ................... Water Soluble Fiber Components ............... Non-fibrous Components ................. . . Hater Insoluble Pectin ................... Water Insoluble Hemicelluloses. . . . . . . . . . ..... Cellulose ......................... Lignin ........................... Non- fibrous Components in Raw Vegetables ........... Crude Protein ....................... Crude Fat and Ash ..................... Moisture .......................... Recoveries ......................... Relationship Between Fiber Components and Shear Force .......................... Simple Correlations .................... Shear Prediction Equations for Cooked Vegetables at Various pH Values ..................... HPLC: Problems and Comments ................. Comparisons with Literature Values ............. Sensory Aspects of Vegetable Texture ............. SUMMARY AND CONCLUSIONS ..................... RECOMMENDATIONS ......................... LIST OF REFERENCES ........................ APPENDIX I - Standard Curves for Hexoses, Pentoses, and Uronic Acids at Various Wavelengths ....... APPENDIX II - Average Results of Fiber Extraction Procedure ..................... APPENDIX III - HPLC Standard Curves of Five Sugars and HPLC Chromatogram ............... 102 104 111 116 123 Table 12 13 14 LIST OF TABLES Preparation and cooking methods for fresh vegetables (200 9 lots) Scheme for analysis of variance and orthogonal comparisons of vegetable fiber data Summary of hexose, pentose, and uronic acid combinations Neutral sugar contents of fiber fractions in raw vegetables F values for analysis of variance and orthogonal comparisons of the vegetables in the raw and cooked states for shear data F values of orthogonal comparisons for raw vs. cooked vegetables in the shear data F values of orthogonal comparisons for cooking solution fiber F values of orthogonal comparisons for water soluble fiber components Starch contents of vegetables F values of orthogonal comparisons for starch F values for analysis of variance and orthogonal comparisons of the vegetables in the raw and cooked states for starch F values of orthogonal comparisons for insoluble pectin F values of orthogonal comparisons for insoluble hemicellulose F values of orthogonal comparisons for lignin vi Page 27 39 46 47 57 58 61 65 68 69 70 73 77 82 IIa IIb IIc IId IIe IIf IIg Proximate analysis of raw vegetables Moisture analysis of raw and cooked vegetables Average percent recovery for vegetables Simple correlations of fiber components with shear force Comparison of HPLC and colorimetric methods A comparison of total neutral sugars based on the four experimental fiber fractions with literature values Average results Average results Average results dialysis Average results dialysis Average results Average results extraction Average results for the starch extraction for protein extraction for the water solubles before for the cooking solution after for the insoluble pectin extraction for the insoluble hemicellulose for the cellulose extraction 83 85 87 89 94 97 116 117 118 119 120 121 122 Figure 11 12 13 14 15 16 LIST OF FIGURES Proposed model for the primary plant cell wall (Keegstra et al., 3) Distribution of cell wall components Structure of cellulose showing cellobiose repeating unit Xylan and some of its branching molecules a-1,4-polygalacturonic acid and some of its branching side chains Basic units of lignin Extraction scheme for fiber components Preparation of raw vegetable fractions for HPLC analysis Firmness of vegetables in raw and cooked states Cooking solution fiber in vegetables as a function of pH Percentage water soluble hemicellulose in vegetables as a function of pH Percentage water soluble pectin in vegetables as a function of pH Percentage water insoluble pectin in vegetables as a function of pH Percentage water insoluble hemicellulose in vegetables as a function of pH Percentage cellulose in vegetables as a function of pH Percentage lignin in vegetables as a function of pH viii Page 10 35 42 56 6O 63 64 72 79 81 Ia Ib Ic IIIa IIIb IIIc IIId IIIe IIIf 1119 Standard curve for hexoses at 620 nm using the anthrone reaction Standard curve for hexoses at 600 nm using the orcinol reaction Standard curve for hexoses at 670 nm using the orcinol reaction Standard curve for pentoses at 600 and 670 nm using the orcinol reaction Standard curve for uronic acids at 525 nm using the carbazole reaction HPLC HPLC HPLC HPLC HPLC HPLC HPLC standard curve of glucose (4X attenuation) standard curve of glucose (2X attenuation) standard curve of xylose (2X) standard curve of arabinose (2X) standard curve of mannose (2X) standard curve of galactose (2X) Chromatogram of pectin fraction (cauliflower) 111 112 113 114 115 123 124 125 126 127 128 129 INTRODUCTION The awareness of fiber in the diet has increased drastically in the past few years. The reduction of certain ailments such as diverticulosis and cancer of the colon has been attributed to increased levels of dietary fiber. Inadequate fiber intake has been hypothesized to cause disease by decreasing stool bulk, which causes increased transit time and concentration of toxic substances. Researchers are beginning to realize that dietary fiber should be measured on a variety of foods as well as under different cooking conditions. Little information exists concerning the relationship between dietary fiber components, pH, and texture in vegetables, an important source of fiber in the diet. To study the effect of pH on dietary fiber components, vegetables commonly cooked in the home (beans, cauliflower, potatoes, peas, and corn) were cooked at four different pH levels encompassing high, low, and near neutral pH ranges. Individual fiber components of cooked and raw vegetables were correlated with tenderness measurements. Previous workers have shown that vegetables are toughest at pH 4—4.5 and softest in alkaline media, mainly due to changes in the water soluble and insol- uble pectin components and to the soluble and insoluble hemicellulose components. This study was conducted to determine the dietary fiber components mainly responsible for affecting tenderness in raw vegetables, compared to those cooked at different pH's. The exact nature of the dietary fiber components themselves as well as dietary fiber methodology still remain to be elucidated. For this 1 2 reason, this study also included the measurement of sugars contained in various dietary fiber fractions. Vegetables are most commonly encountered at different pH levels in canned goods. Canned goods are classified according to pH. The four major categories include: low-acid, pH535.3; medium—acid, pH 4.5-5.3; acid, pH 3.7-4.5; and high acid, pH 53.7 (Frazier and Westhoff, 1978). Some common canned vegetables and their corresponding pH values are as follows: peas, pH 5.6—6.5; green beans, pH 4.9-5.5; brine-packed corn, pH 5.8-6.5; potatoes, pH 5.4-5.9; and spinach, pH 4.8—5.8 (Lopez, l98l). Some produce such as sauerkraut (from cabbage), celery, carrots, olives, and pickles may undergo lactic acid fermentation in which case the final pH of these products will be from pH 3.3 to pH 3.8 (Frazier and Nesthoff, 1978). Since softening usually results from canning procedures, final pH, or both, firming agents such as calcium or aluminum salts are usually added to canned goods to avoid undesirable textural qualities. REVIEW OF LITERATURE Dietary Fiber Trowell (l976) defined dietary fiber as those plant polysaccharides and lignin not digested by the endogenous secretions of the human upper gastrointestinal tract. However, some dietary fiber components can be digested by the microflora of the colon. At present, the extent of di- gestion of various polysaccharides in the small and large intestine of humans is unknown. Until the physiological significance of the complex mixture termed "dietary fiber" is known, Trowell's definition should suffice. Currently, dietary fiber includes hemicelluloses, cellulose, pectic substances, gums, mucilages, and lignin. Fiber Components: Location and Structure The three main sections of the cell wall are the middle lamella, primary cell wall, and secondary cell wall. The middle lamella is rich in pectic substances which act to cement the plant cells together. Pectin is also found in the primary cell wall. The primary cell wall is formed very early in the cell's development. Keegstra et al. (1973) proposed a model for the primary cell wall as shown in Figure 1; note that protein is covalently linked to the pectic substances by arabinogalactan. When cell growth stops, certain mature cells form a secondary cell wall (North, T967). During this time, lignin is synthesized and depos- ited in the cell walls. The greatest concentration of lignin is in this secondary cell wall as shown in Figure 2, although it is also contained 3 mm arabinan and A-linked /galactan rhamnogalacturonan ////seryl residue 000 5000 <><><><><><><> a firm“ 0 O O O~\tetra-arabinoside 3,6-linked arabino- galactan attached to serine OOOOOOOOOOOOU xyloglucan HHIW I HIWIHHIMIHHIWIHHIW IFFIWIHHP’ Cellulose fibril WWW Figure 1: Proposed model for the primary plant cell wall (Keegstra et al., 1973). a-CELLULOSE > HEMICELLULOSE K PECTIN LUMEN LIGNIN ~~~~~~ ,.———"” ““~~~y-""" ~"m"""''"“"1""""‘,""""" MIDDLE PRIMAR SECONDARY LAMELLA WALL WALL Figure 2. Distribution of cell wall components. in the middle lamella and primary cell wall. The cell wall performs three major functions (Albersheim, 1965): 1) to physically counteract the osmotic pressure exerted by the cell wall contents, 2) to provide structure, and 3) to physically impede pathogens. The secondary cell wall, thicker than the primary cell wall, acts as a plant skeleton, although the primary cell wall also contributes to plant rigidity. Cellulose and hemicellulose, both located in the primary and secondary cell walls, show opposite distribution trends: cellulose is mainly concentrated in the secondary cell wall, while hemicellulose lies mainly in the primary cell wall. Cellulose. Cellulose is a polymer containing 84,4-linked D-glucopyran- ose residues containing the repeating unit of cellobiose as shown in Figure 3 (Northcote, 1972). The individual glucan chains are held to- gether in fibrils by hydrogen bonds. The crystalline microfibrils are embedded in an amorphous gel (Albersheim, 1965). Some claim that cellu- lose is not a homopolymer in the plant cell wall; rather it is coval- ently bonded to noncellulosic sugars (Albersheim, 1965). According to Pigman and Horton (1970), cellulose has been found in conjunction with lignin and hemicellulose. HZOH HZOH cellobiose unit Figure 3. Structure of cellulose showing cellobiose repeating unit. 7 Hemicellulose. Hemicellulose is composed of three main groups of poly- saccharides: l) the xylans, 2) the mannans and glucomannans, and 3) the galactans and arabinogalactans (Bauer et al., I973). Xylans, polymers of B—l,4—linked xylopyranose units, are the most widely distributed hemicelluloses in higher plants. Both B-3-linked arabinofuranose and a-1,3-linked glucuronic acid have been found attached to the xylan chain as shown in Figure 4. Bauer et al. (I973) found hemicellulose covalently linked to pectic substances and noncovalently linked to cellulose fibrils. They proposed that hemicelluloses be redefined as those plant polysaccharides that are noncovalently bound to cellulose. The mannans and glucomannans are derivatives of B-l,4—linked manno— pyranose and B-l,4—linked glucopyranose units containing various amounts of glucose and mannose whose ratio depends on the plant source (Aspinall, 1970). Galactans and arabinogalactans are highly branched derivatives of 1,3- and 1,6-linked D-galactopyranose polymers containing arabinose side groups. Most of the arabinose is in the form of the disaccharide 3—O-L-arabinopyranosylarabinofuranose linked by the C—6 of the galacto— pyranose chain. The major residues of galactan are galactose; if large amounts of L-arabinose are present, the polymer is called arabinogalactan (Aspinall, l970). Xylose, glucose, mannose, galactose, arabinose, and rhamnose have been detected in many vegetables such as spinach, onions, red cabbage, kale, watercress, cauliflower, leek, parsley, potatoes, beets, and radishes (Englyst, 1981). OH H XYLAN (repeating unit) HOHZC H L-ARABINOFURANOSE COOH HO 0” H OH D-GLUCURONIC ACID Figure 4. Xylan and some of its branching molecules. Pectin, It was previously thought that pectin was composed solely of aeD—l,4-linked galacturonic acid residues (Kertesz, 1951). Today it is generally regarded that other sugars such as xylose, arabinose, rhamnose, glucose, and galactose are covalently attached to the polygalacturonide backbone and other sugars may be an important part of the main chain (Albersheim, l965). Rhamnogalacturonan is the main polysaccharide component of pectic substances; its backbone consists of a-1,4-linked D-galacturonopyranose residues with some oel,2-linked rhamnopyranose residues. Up to 65% of the carboxyl groups of poly—D—galacturonic acid may be esterified by methyl groups, with positions 2 and 3 acylated in some instances. Galac— tans (B-l,4-linked D-galactopyranose units) and arabinans (a-l,5- and a—1,3-linked L-arabinofuranose units) are the other two major pectic substances (Aspinall, 1970). Figure 5 shows some of the structures mentioned above. Lignin, Lignin, a complex polymer of phenylpropanoid units (C6 - C3), plays an important role in internal transport of water, nutrients, and metabolites in plants by decreasing water permeation across cell walls in conducting xylem tissues (Freudenberg and Neish, 1968). Lignin also helps in preventing fungal penetration of the epidermal cell walls and provides mechanical strength for plant cell walls. Lignin (Figure 6) is derived from the enzymatic dehydrogenation and polymerization of coumaryl, coniferyl, and sinapyl alcohols. Although the presence of sinapyl and p-coumaryl alcohols varies with the variety of plant, coniferyl alcohol is found in all plants (Freudenberg and Neish, 1968). ARABINAN Figure 5. oel,4-polygalacturonic acid and some of its branching side chains. 11 H H éoml?‘ 1: 1“ I H H CHZOH H2011 HZOH TRANS-p-COUMARYL TRANS-CONIFERYL TRANS-SINAPYL ALCOHOL ALCOHOL ALCOHOL Figure 6. Basic units of lignin. General Effects of Cooking Information concerning the physical and chemical changes occurring in vegetables during cooking still remains incomplete and many contra— dictory results exist. The effects on individual fiber components due to cooking media pH also remain uncertain due to the paucity of invest— igations in this area. Before attempting to understand changes in vegetables due to pH differences, one should understand cooking effects in general. A series of structural changes occurs in vegetables during cooking beginning with starch gelatinization (Reeve, 1970). Although the temperature of gelat- inization depends upon the size of the starch granule, most granules begin to swell between 60°C and 700C (Reeve, 1967). The smaller the starch granule, the higher the gelatinization temperature and the slower the reaction. After complete starch gelatinization, the pectic substances begin to break down, resulting in initial tissue softening by cell separation and loss of rigidity in individual cell walls (Day, 1909; Personius and 12 Sharp, 1939; Reeve, 1954; Sterling, 1955; Sterling and Bettelheim, 1955; Reeve, 1970). Any agent or process which decomposes the pectic sub- stances in the middle lamella gives rise to cell separation (Sterling, 1955). Considering that pectic materials also occur in the primary wall, one may assume that degradation of pectic substances contributes to cell wall softening (Sterling, 1955). Because of their weak nature, the hydrogen bonds of pectic and other cell wall polysaccharides become dissociated by elevated cooking temp- eratures, which facilitates cell separation (Sterling, 1963). Thus, cooking not only leads to depolymerization but also results in hydrogen bond rupture of pectic substances, with their subsequent swelling and solubilization (Bettelheim and Sterling, 1955b; Sterling, 1963). In addition to the above cooking effects, the protein in the cyto— plasm and cell membranes becomes denatured (Charley, 1972). Water dif- fuses from the vacuoles once the cell is destroyed and the cooked tissue loses its turgor, becoming limp and flabby (Charley, 1972). Cooking also causes the expansion and escape of intercellular gases which are then replaced by steam. Effects of Cooking in Relation to Vegetable Fiber Components Textural changes in fruits and vegetables during cooking are attrib- uted mainly to alterations of pectic substances which constitute the intercellular cement. Degradation of these substances involves hydrogen bond disruption and ion-bridge rupture, as well as depolymerization of the carbohydrate backbone. These changes result in weakening and eventual solubilization of pectic substances in all parenchymatous 13 fruits and vegetables, causing a decrease in firmness (Reeve, 1970). This softening, the result of cell separation, is strongly affected by variations in both pH and temperature during cooking (Sterling, 1955). EH effects. Both high and low pH affect the firmness of cooked vege- tables. Plant tissues boiled at pH 4.0 to 4.5 were firmer than those cooked at higher or lower pH, which correlated well with the increase in soluble pectin contents (Doesburg, 1961). Since potatoes, beets, tur- nips, apples, and cauliflower showed the same trends, this demonstrated that cell walls in parenchymatous tissues from various organs of edible plants had identical structure. In a more acidic milieu (pH 3-4), formation of soluble pectin may be caused by the exchange of calcium bridges in protopectin by hydrogen ions (Mattson, 1946), therefore, more pectin should be found in acidic media (Bettelheim and Sterling, 1955b). Doesburg (1961) claimed that during cooking in acidic media, the increase in soluble pectins cannot be due to depolymerization of pectin. He expected that the molecular weight of soluble pectin should be rather high after cooking, which has been shown by its high viscosity. On the other hand, Bettelheim and Sterling (1955b) stated that the depolymerization occurred by breakage of primary valence bonds in acidic media, resulting in a lowered intrin- sic viscosity; they also pointed out the unlikelihood that soluble pectin is formed mainly by an increase in its solubility at high temperatures. However, hydrolysis of other cell wall constituents to which pectic sub— stances are attached, such as hemicelluloses, may contribute to soluble pectin formation during cooking (Doesburg, 1961). Personius and Sharp (1939) showed that cooking plant tissue in lactic acid solutions from pH 4 to pH 2.5 resulted in decreased tensile 14 strength as the pH decreased. On the other hand, Sterling (1968) demon- strated decreased firmness as the pH increased from pH 3 to pH 8 with a plateau between pH 4 and 6, the tissue being hardest at pH 3. At a low pH, the cells were well-rounded and close, but as the pH increased, more cell separation and collapse was evident. Sterling (1968) used Kertesz's (1951) theory to explain his results: the higher the pH (above 3), the weaker the hydrogen-bonded pectin gel. At a pH of less than 2.8, the pectin gel is weakened again. Higher pH's are also effective in increasing soluble pectin. According to Doesburg (1961), pectin solutions showed a decreased jelly- ing power during boiling at high pH values due to a shortening of the pectin chains. This was verified by a decrease in pectin at a pH greater than 5.8. When he measured the pH of solutions of higher pH values after cooking, a decrease in pH had occurred due to formation of carboxyl groups by de-esterification of the carbomethoxy groups. When Doesburg (1961) boiled pectic solutions at a pH exceeding four, severe depolymerization of pectin molecules occurred. He proposed that cooking at a pH just above four weakened the gel structure of protopectin by a beginning depolymerization, not quite strong enough to cause pectin solubilization. With further pH increases, continued depolymerization may cause larger amounts of soluble pectin. At lower temperatures in alkaline media, pectins are saponified to pectates and esterified chains of polygalacturonic acid decompose. The same phenomena occur during cooking in weak acid or neutral media (Doesburg, 1961). Both Albersheim (1959) and Neukom and Deuel (1958) agreed that during cooking in low acid or neutral media, pectates are much more stable than pectin. Doesburg (1961) advised that in order to determine pectins in plant materials, 15 the extraction should not entail high temperatures nor high pH. Simpson and Halliday (1941) found that due to the hydrolysis of protopectin to pectin upon cooking, the protopectin content decreased in carrots and parsnips while pectin increased in the cooking water and decreased in the vegetables themselves. The increase of pectin and pec- tic acid and the decrease of protopectin and total pectic substances during cooking is analogous to softening during ripening (Monte and Maga, 1980). These authors point out that only half as much insoluble pectin was extracted from cooked pinto beans as from the raw beans. Cooking also weakens the cell walls, with which insoluble pectin is thought to be associated (Aspinall, 1970). I Hughes et al. (1975b) proposed that pectic substances degrade in potatoes during cooking by chelation of calcium and magnesium between polyuronide chains by naturally occurring phytin. However, it was agreed later that for pectic breakdown and loss of intercellular adhesion, sev- eral factors should be considered (Hughes et al., 1975b,c): the struc- ture of pectic substances in vegetables, degree of methoxylation possibly affecting trans—elimination, degree of carboxylation, pH of the tissue and ions present, and the method of cooking, rather than chelation of calcium by phytin. The role of phytin in peas and potatoes is discussed later in this paper. ‘Temperature effects. Solutions of pectin are very sensitive to increased temperature in neutral and weakly acidic solution. Under these condi- tions, pectin breaks down to lower molecular weight products by breakage of glycosidic bonds adjacent to methoxyl groups by trans-elimination (Albersheim, 1959; Albersheim et al., 1960). Thus, with conditions 16 occurring during the usual cooking period, i.e. increased temperature and near neutral pH (pH 6, approximately) of plant tissue, the presence of monovalent ions, ion-exchange, and hydrogen bond breakage, trans—elimina- tion of 40% esterified polyuronide substances is favored, as shown in potatoes (Hoff and Castro, 1969; Warren and Woodman, 1973). Hemicellulose Heller et al. (1977) showed that acid or base and heat resulted in the solubilization of hemicellulose. The analysis was by the NDF (neu- tral detergent fiber) procedure of Goering and Van Soest (1970), i.e., indirect measurement of hemicellulose (neutral detergent fiber minus acid detergent fiber). In red wheat bran and purified corn pericarp, the hemicellulose content generally increased from pH 2.2 to pH 6.6, plateauing at pH 6.6, and decreasing from pH 6.6 to pH 11.5 after a 5 minute or 60 minute reflux under these conditions. In general, the lowest amount of hemicellulose was observed at pH 11.5 under reflux con- ditions. However, using the same pH range of pH 2.2 to pH 11.5 with 24 hour shaking at 25°C, little or no change was shown in hemicellulose values, indicating the importance of heating conditions. Peanut hulls showed no significant changes in hemicellulose either at room temperature or upon reflux, but the initial hemicellulose content of peanut hulls was much lower than the other two samples. Mattheé and Appledorf (1978), again using the Goering and Van Soest (1970) procedure, showed that hemicellulose was unaffected, going from the raw to the cooked state in vegetables, which agreed with the results of Heller et al. (1977). Herranz et al. (1981) demonstrated a decrease in hemicellulose for 17 carrots, but an increase for other vegetables; however, these increases were insignificant (P< 0.05) for cauliflower, spinach, turnip tops, spring white cabbage, potatoes, broad beans and chard. Monte and Maga (1980) found that cooked pinto beans showed a 33% decrease in extractable hemicellulose A and complete depletion of hemi- cellulose 8. They statedthat these results are opposite of those ex— pected because the concentrating effect of cooking should have resulted in increased hemicellulose A and B of the cooked bean as compared to the raw bean. Cellulose Cellulose shows no alterations in its nature under normal cooking conditions, with the cell walls retaining their integrity (Day, 1909; Sterling, 1955; Sterling and Bettelheim, 1955). However, starchy vege- tables such as potatoes undergo a great deal of cell rupture due to ex- pansion of gelatinized starch (Reeve, 1954). Cell wall rupture in pota- toes during cooking has also been reported by Hutchison (1908) and by Sherman (1948). .Monte and Maga (1980) reported a 50% decrease in lignocellulose and and crude cellulose in pinto beans rationalized by enzymatic release of oligosaccharides in the beans. The cell wall disintegration and cellu- lose increase described by Simpson and Halliday (1941) has been greatly disputed. Herranz et al. (1981) as well as Mattheé and Appledorf (1978) reportdincreased cellulose upon boiling vegetables, explained by in- creased hydrolysis and liberation of cellulose upon cooking, making it more available for analysis. In contrast, Sterling (1955 ) demonstrated increased softening and 18 cell separation in cooked carrots with no indication of cell wall rupture or broken cells, even after 60 minutes of steaming. He showed the same results in apples steamed for 60 minutes, accompanied by starch gelatin- ization. Potatoes, either boiled, steamed, or baked for one hour also showed no evidence of cell rupture. Thus, Sterling (1955) concluded that cell walls are not broken by heat and moisture during cooking. Although cellulose showed no change upon cooking, pectin showed a great deal of change, resulting from dis- solution of pectic substances. Tauss (1889) observed little if any cellulose dissolution in plant materials cooked under 5 atmospheres of pressure; only at 10 atm and higher did he nohacellulose breakdown, with 14% of the cellulose being decomposed. Sterling (1955) suggested that the results of Simpson and Halliday (1941) may have been affected by mis- handling in preparation for microscopic sections. Liam Only strong chemicals degrade lignin, thus it is unlikely that it is affected by mild cooking conditions. It also has little effect on the textural qualities of vegetables (Isherwood, 1955). However, lignin may affect the properties of other compounds in the cell wall even in small amounts. Lignin, a very reactive material, will polymerize and combine with many compounds, rendering a cell wall consisting of polysaccharides completely insoluble (Isherwood, 1955). The nature of lignin varies with age, cultural conditions, and plant variety, so it is difficult to elu- cidate interactions of lignin with other components of the cell wall. Herranz et al. (1981) found that lignin decreased with boiling ex- cept in summer white cabbage, which showed a slight increase; however, 19 the lignin changes were only significant for Brussels sprouts and aspar- agus. Tannin-protein condensation products and products of the Maillard reaction caused inflated lignin values for Van Soest and Robertson (1976). Mattheé and Appledorf (1978) demonstrated these trends in broccoli. mm According to Mattson et al. (1951), the normal softening of peas during cooking is caused by a reaction between intact phytin (inositol hexaphosphate) in the cell and calcium pectate of the middle lamella, I which loses calcium to the phytin, forming the very insoluble calcium phytate. Removal of calcium causes dissolution of the middle lamella, weakening the adhesion between cells, thus softening the pea tissue. When phytase breaks down the phytin (e.g., through storage), the peas will not soften, creating the “uncookable" phenomenon. If the phytin is destroyed, the peas will become uncookable and they will remain very firm. In potatoes, Wager (1963) showed that during cooking, some of the middle lamella calcium becomes bound to the tuber by phytic acid. He theorized that softening of the potatoes during cooking depended mostly on the amount of phytic acid. However, Sterling (1966) proposed a dif- ferent hypothesis: that phytin is present in potato cell walls and may be involved in ionic crosslinking in the same way as pectic substances. Starch When starch granules reach their gelatinization temperature, swelling and solubilization occurs (Sterling, 1963); most starch granules 20 swell at approximately 650C in cooked potatoes (Reeve, 1970). After com- plete starch gelatinization, some breakdown of the middle lamella pectins begins between cell walls of neighboring cells. Swelling of starch granules also distends cell walls, tending to make the polyhedral shapes rounder. This makes the cells distended and push apart, causing a "mealy” texture (Reeve, 1954; Sterling, 1955; Sterling and Bettelheim, 1955). Many viewpoints exist concerning the influence of starch on cooked vegetable texture. Those changes which some authors have attributed to starch, others have ascribed to pectin, and vice-versa. Textural dif- ferences in cooked potatoes are caused by differences in internal pres— sures developed in the cells during swelling of gelled starch (Whitten- burger and Nutting, 1950; Whittenburger, 1951). However, Hoff (1972) claimed that starch does not appear to cause cell separation on cooking. He stated that starch gelation pressure exists only when mass transfer occurs in tissue and does not coincide with gelation of the starch gran- ule. Hoff (1972) cited Rathsack's theory as more probable: under cer- tain conditions, the limit of elasticity (varying with temperature) of the weakest point will be exceeded when a potato is heated to boiling. Mealy potatoes are weakest in the middle lamella, and this results in cell separation. The weakest point in non-mealy potatoes is the cell wall itself, which results in cell rupture. I Hughes et al. (1975a) seem to agree with the above reasoning. In potatoes, they found that as the cooking time increased, the compressive strength decreased and release of pectic substances and starch increased. They attributed changes in compressive strength to degradation of pectic substances and not to differences in starch swelling. They claimed that 21 cell rupture is not important during boiling since the most dramatic effects were shown by the pectic substances; if the cell walls were weak, a larger amount of starch would have been released into the cooking media. Sharma et al. (1959) found intercellular adhesion positively related to polyuronide; Linehan and Hughes (1969a) expected cell adhesion to be inversely proportional to starch content, however, a direct relationship was found to exist (Barmore, 1937; Sharma et al., 1959). Linehan and Hughes (1969a) supposed that starch influences texture in ways other than cell distension. Differences in cookability of several varieties of potatoes, i.e., "mealiness" or "sogginess", could not be ascribed to the behavior of the pectic substances or starch (Personius and Sharp, 1939; Freeman and Ritchie, 1940; Bettelheim and Sterling, 1955a,b). According to Sterling (1963), if the pectic cement is strong (rich in calcium pectinate) and the amount of starch is low, less separation occurs in cooked potato and a soggy texture results. Nonaka (1980) and Personius and Sharp (1939) found results in pota- toes contradictory to the "swelling theory“, i.e., they found that the separation of potato cells was not dependent on starch gelatinization. Van Buren (1970) proposed that the most likely change is in the pectic material in the middle lamella. Sterling and Bettelheim (1955) claimed that although pectic materials swell in hot water, enlarging the middle lamella and increasing cell separation, the calcium bridges counteracted this effect, thus tightening the bonds. Regardless of theory, cell separation appears to be related to change in pectic substances, causing a decrease in cellular adhesion and compressive strength. 22 Linehan and Hughes (1969b) suggested that intercellular adhesion in cooked potatoes depends mostly on the levels of amylose and pectic sub- stances, and to a lesser degree on calcium and magnesium levels. They postulated that high amylose levels mask effects on intercellular adhe- sion by pectic substances, so that this adhesion is evident only at low amylose levels. A parallel between intercellular adhesion and starch was evident only when the starch level was high (Linehan and Hughes, 1969b). Amylose was shown to be most important because the relation between intercellular adhesion and amylose was always better that that between intercellular adhesion and starch. Amylose increased intercellular adhesion, while amylopectin had little or no effect (Linehan and Hughes, 1969b). They theorized that a small amount of starch may "leak" out through the weakened cell walls during cooking. The linear amylose chains act as a cement between tuber cells by hydrogen bond formation between these chains and polysaccharide of the cell walls (Linehan and Hughes, 1969b). According to Whistler and Smart (1953), hydrogen bonds form between amylose, but not amylopectin and other polysaccharides such as cellulose. Owing to the relatively large size of the amylose molecule, its diffusion through the cell wall would be expected to be extremely slow. Even so, Bettelheim and Sterling (1955a) found that up to 50 mg of starch per gram of tissue were lost from potatoes during boiling in water, which showed that the potato cell wall presented no real barrier to starch diffusion. 23 Sensory Aspects of Vegetable Texture In a study by Odland and Eheart (1975), frozen broccoli prepared by three different methods (water, steam, and steam-NH3 blanched) was eval- uated for textural and flavor characteristics by a sensory panel of five trained members. The pH values for broccoli blanched by the three methods and cooked after 6 months storage at -180C were as follows: water-blanched = pH 6.56, steam-blanched = pH 6.50, and steam-NH3 blanch- ed = pH 6.86. The texture of steam—blanched broccoli was preferred to that of water-blanched broccoli, although the pH values did not differ greatly. The steam-NH3 blanched (highest pH) broccoli received the highest score for color and overall acceptability, however, it received the lowest score for flavor. This showed that all three components were important in judging these vegetables and neither flavor, texture nor color should be examined independently. In another study by Eheart and Odland (1973), peas, green beans, lima beans, broccoli, and Brussels sprouts were blanched with different levels of NH4HC03 in the blanch water. Six trained panelists evaluated the vegetables for color, flavor, texture and overall acceptability. The panelists were able to discern a significant increase in greenness of color for all vegetables blanched in NH4HCO3 with the best color rated for vegetables containing the greatest amount of NH4HC03 (0.3% in blanch water). The flavor was unaffected by NH4HCO3, except for peas, where flavor scores were decreased by the highest concentration of base. The most significant adverse affect of NH4HC03 on texture occurred at the highest level of the additive. These vegetables were considered "too soft" by most of the judges. Overall acceptability of peas and lima beans was unaffected by NH4HCO3, which the authors rationalized by 24 attributing this to the small differences in color scores for these two vegetables. Beans, broccoli, and Brussels sprouts showed differences in acceptability due to color improvement under more alkaline conditions, but at the same time, adverse softening. Because vegetables differ in their inherent pH, buffer systems, and textural properties, the authors concluded that the optimum concentration of NH4HC03 varies depending on the type of vegetable. The studies of Jeon et al. (1975) involved the correlation of instru- mental and sensory measurements on fresh—pack whole cucumber pickle tex- ture. They found significant correlations between sensory and objective tests. The authors pointed out that the texture of the pickles depended on the variety. In a study by Wager (1963), a sensory panel evaluated the texture of potato strips cooked at various pH values, ranging from pH 4.5 - 7.0. Although this was only a portion of the pH scale, the panel was able to detect distinct differences. The most favorable texture was found to be pH 6 - 7. At lower pH values, the panel scored the texture much lower. EXPERIMENTAL Vegetable Preparation and Cooking Procedure Common lots of fresh corn, peas, potatoes, green beans, and cauli— flower were purchased from local supermarkets and prepared according to the following procedure. The methods represent typical home cooking procedures. After dividing each vegetable into common lots, 200 9 samples of each vegetable were weighed and prepared accordingly: peas — whole, shelled; cauliflower - cut into 1 to 1% inch cubes; small kernels of corn - cut off the cob (1 small ear = 75 9, approximately); green beans - cut into approximately 1 inch pieces; and potatoes (U.S. No. 1), peeled and cut into 1 inch cubes. In all cases, any spoiled portions were trimmed from the vegetables beforehand. The vegetables were cooked in covered 1% quart Pyrex saucepans at medium burner setting on an electric stove in phosphate buffers of pH 2, pH 4, pH 6, and pH 10. These buffers were obtained from Mallinckrodt and the buffer for pH 6 served as the control. The vegetables were analyzed in triplicate in both the raw and cooked states according to a confounding plan in which a portion of the two factor interaction was confounded. Each batch analyzed consisted of the five vegetables each at a different pH plus one vegetable in the raw state; for example, batch 1 represented: peas - raw, green beans — pH 2, corn - pH 4, cauliflower - pH 6, and potatoes - pH 10. Each vegetable was assigned a random number. 25 26 The vegetables were cooked in this scheme until all five vegetables were prepared in all five states, thus, 15 batches of vegetables were prepared (5 groups of vegetables prepared at one time - triplicate analysis), or 75 total samples. As a preliminary test, the volume occupied by 200 g of sample was measured in a beaker to determine the volume of buffer required as a cooking medium. Half of this volume (number of milliliters of buffer) was used to cook peas, corn, and green beans. For the bulkier vegetables, potatoes and cauliflower, the same volume occupied by the samples in beakers was used for the volume of buffer (wet volume buffer = dry volume sample height), as recommended by Charley (1970). Optimum cooking time was determined by piercing each vegetable with a fork and noting the time when the vegetable's texture felt ten- der (ranging from 10 minutes for potatoes to 15 minutes for corn and cauliflower). Once determined, this optimum cooking time was kept con- stant for each vegetable. The cooking medium was allowed to boil before the vegetable samples were placed into it. After cooking, each vege- table was immediately drained through a strainer into a bowl and rinsed with approximately 60 to 100 ml of deionized water. The total volume of cooking solution was measured in a graduated cylinder and approximately 100 ml of it was saved for further analysis and stored at -250C. The cooking conditions are summarized in Table 1. Two 50 9 samples of each vegetable were weighed to the nearest 0.1 g and sheared with the Allo-Kramer Shear Press, Model TR3 (Food Technology Corp. Texturecorder) at a 14 sec downstroke. The samples were first allowed to cool to room temperature after cooking and sheared as soon as possible. The tenderness was expressed as pounds force per gram using 27 Table 1. Preparation and cooking methods for fresh vegetables (200 9 lots) Vegetable Preparation Cooking Solution Cooking Time Method Before Cooking (min) (ml buffer) Potatoes Peeled, cut 350 10 into 1 inch cubes Green Beans Cut into 1 250 11 inch pieces Peas Shelled, 200 12 whole Corn Cut off cob 175 15 Cauliflower Cut into 1- 550 15 1% inch cubes the formula: Tenderness = Reading x transducer x range Sample wt. x 100 Adequate samples were saved for moisture determination. The samples were cut into tiny pieces and frozen in 9 inch, round, aluminum pie pans before placing them in a freeze—drier. The remaining samples were fro- zen at -25°C and lyophilized for 48 hr in a Virtis Unitrap II freeze- drier with an 8 liter capacity, equipped with a clear drum of dimensions 43.2 cm height x 30.5 cm diameter (17 x 12 in). The samples were dried at a pressure of 4—6 x 10'2 Torr and tray temperatures of 400-500C. Samples were then ground in a cyclone sample mill (UD Corp., Boulder, Colorado) and stored in polyethylene sample bags in a desiccator con— taining Drierite (CaSO4)o 28 Analysis of Fibrous and Non-Fibrous Vegetable Components All samples were weighed to :_0.0001 9. Moisture The moisture content of each raw and cooked vegetable was determined by AACC procedure 44-40 (AACC, 1962). Approximately 2 g duplicate sam- ples of each vegetable were dried for 6 hr at 900C under a vacuum of 27 in of mercury in a Hotpack vacuum oven, Model 633. The following formula was used to calculate percentage moisture: Wt. of moisture lost (9) x 100 Sample wt. (9) % Moisture = Ash Approximately 0.5 9 samples were ashed in a Barber-Coleman Oven, Model 293C at 525°C for 48 hours. Only the raw, freeze-dried vegetables were ashed. After cooling in a desiccator, the samples were reweighed and percentage ash was determined as follows: % Ash = (wt. of crucible + ash) - (wt. of crucible) x 100 Sample wt. Crude Fat The raw, freeze—dried vegetables were analyzed for crude fat by using approximately 3.0 9 samples. These were placed into extraction thimbles (Cellulose, single thickness, 33 mm inner diameter x 800 mm exterior length, Whatman, England), covered with glass wool, and extract- ed with hexanes for 6 hr in a Soxhlet apparatus. After cooling, the ex- tracts were filtered through glass fritted filter funnels of medium por- osity. The solvent was evaporated on a steam bath in tared 150 ml beakers until no visible solvent remained. The beakers were dried for 29 13 hr at 70°C under a vacuum of 27 in of mercury. After cooling in a desiccator and reweighing the beakers, the percentage crude fat was calculated as follows: a _ wt. of fat ( ) X 100 [7 Crude fat - samp1e Wt- g Protein The raw, freeze-dried vegetables were analyzed for crude protein by a revision of AACC method 46-13 (AACC, 1962) using a micro-Kjeldahl ap- paratus. Approximately 30 mg of sample was digested with 1.30 g of po- tassium sulfate, 2.0 ml of sulfuric acid, and 40 mg of mercuric oxide. After cooling the flask, about 5 ml of water were added to the samples. The flasks were transferred to a micro-Kjeldahl distillation unit. Eight ml of sodium hydroxide - sodium thiosulfate solution were added along with a few zinc granules. The sodium hydroxide - sodium thiosulfate solution was prepared by dissolving 50 g of NaOH and 5 g of Na25203'5 H20 in water, and diluting to 100 ml. Approximately 40 ml were distilled into a 50 ml beaker containing 5 ml of 4% boric acid solution and 4 ml of methyl red + 30 mg methylene blue indicator (100 mg methyl red + 30 mg methylene blue dissolved in 60 ml of 95% ethanol made to 100 ml with boiled, distilled water). The solution was titrated with 0.0230 N hydrochloric acid to a pur- ple endpoint. Blank determinations were made using the same procedure. Total nitrogen as per cent protein was calculated as follows: = (Sample - blank)(Normality of HCl)(meq. wt. N)(6.25)(100) Sample wt. (9) % Protein 30 Fiber Extraction All samples were analyzed according to the same scheme as described for cooking; however, 10 samples were analyzed at one time, i.e., two groups of vegetables. The procedure used is essentially the same as ex— plained by Jeltema and Zabik (1980) with minor modifications. Chlorophyll Removal. Since chlorophyll would interfere with the color- imetric procedures described later in this paper, this substance was removed from the green beans and peas. Samples weighing approximately 0.2 g were transferred to fritted glass Gooch crucibles of coarse porosity and placed on top of a glass funnel in a beaker. Approximately 50 ml of diethyl ether were poured into the crucibles and allowed to drain freely, with occasional stirring. This procedure was repeated two additional times until little or no green color remained; a total of about 150 ml of solvent were used. The material on the crucible was allowed to air dry before proceeding further with the extraction steps. Total Water Solubles. The remaining vegetables (corn, peas, and cauli- flower) also weighing 0.2 g were placed into 50 ml stainless steel cen- trifuge tubes to which 25 ml of deionized water were added. They were shaken on a Burrell shaker for two hours at room temperature. The tubes were centrifuged at 15,000 rpm for 15 minutes (Damon/IEC B—20A centrifuge, Model 3444) and the supernatant filtered through the same Gooch crucible into 100 ml bottles. Five ml of water were added to the samples and they were mixed, recentrifuged for 15 minutes, and decanted. Twenty ml of deionized water were added to the centrifuge tubes and the above pro- cedure was repeated. Finally, the filters were rinsed with a few ml of water. The collected filtrate was heated at 1000C for 30 minutes on a 31 steam bath to inactivate natural enzymes. After cooling, the total vol- ume was measured in a graduated cylinder. This material was frozen until hexose, pentose, and uronic acids were measured. Water Soluble Fiber Components. Nine ml of the total water solubles and the cooking solution media were pipetted into glass 50 ml centrifuge tubes. One ml of 1 mg/ml amyloglucosidase (glucoamylase, 1,4—a—D-glucan gluco- hydrolase; E.C. No. 3.2.1.3, Sigma Chemical Co.) and 1.2 ml of 2M acetate buffer pH 4.5 were added. The tubes were shaken on a Burrell shaker at room temperature for 18 hours. After starch extraction, the solutions were dialyzed against deionized water in 0.39 inch tubing (MW cutoff = 12,000) at room temperature for 24 hrs to remove solubilized starch and water soluble sugars. The water was changed once after about 8 hrs. The total volumes were measured after removing the solutions from the tubing. The solutions were frozen until hexose, pentose, and uronic acids were measured. Water Insoluble Starch. Nine ml of water were used to wash any small particles remaining on the Gooch crucibles into the tubes containing the unextracted material. This was heated at 1000C on a steam bath to gel- atinize the starch. After cooling, one ml of 1 mg/ml amyloglucosidase and 1.2 ml of 2M acetate buffer pH 4.5 were added. Seven drops of tol- uene were added to prevent microbial growth; the tubes were mixed and placed in an incubator at 37°C for 18 hrs. After adding approximately 32 ml of 95% ethanol, the tubes were centrifuged at 15,000 rpm for 15 min and filtered through the same Gooch crucibles into 100 ml bottles. The material was remixed with about 20 ml of ethanol, centrifuged for 15 min and filtered. Approximately 5 ml of ether were added and mixed. 32 The tubes were centrifuged for 5 min, and the supernatant again filtered. After measuring the total volume, the solutions were frozen for further analysis of hexose, pentose, and uronic acids. The ether extracted mater- ial was air dried, and the material remaining on the filter was scraped off: and readded . Protein Removal. Ten ml of water, ten ml of 0.2 N hydrochloric acid, approximately 30 mg of pepsin (from hog stomach mucosa, 2 times crystal— Ii zed and lyophilized E.C. No. 3.4.23.1, Sigma Chemical Co.), and 5 drops of toluene were added to the precipitate in the metal centrifuge tube and mixed. The mixture was incubated at 37°C for approximately 20 hrs. 1 After adding 20 ml of ethanol, centrifugation, washing, and filtration were carried out as described for starch using the same Gooch filter for ea ch sample throughout the extraction procedure. Pectin Extraction. The method used for pectin extraction was similar to that of McCready and McComb (1952) as modified by Jeltema and Zabik (1980). Forty ml of 5% (wt/vol) EDTA were added to and mixed with the Extracted material to sequester divalent ions. The pH was adjusted to 11 - 5 with 1 N NaOH to demethylate pectin molecules. After constant St‘i rring for 30 min on a Corning Hot Plate Stirrer, Model PC-351, the pH was adjusted to pH 5.0-5.5 with glacial acetic acid. One—half ml of 96. 15 units/ml pectinase or 0.54 ml of 88.50units/ml pectinase (poly— gal acturonase, poly-a-1,4-galacturonide glycanohydrolase, from Aspergillus m, E.C. No. 3.2.1.15, Sigma Chemical Co.) was added with continued stirring for one hour. The entire mixture was poured onto the appropriate filter over a volumetric flask and each tube was rinsed with a few mls of Water to facilitate transferring the material. After removing the 33 filters, they were rinsed with two small portions of 95% ethanol and an- hydrous ethyl ether. This was allowed to air dry at room temperature, the solutions were made to volume in the volumetric flasks, and portions were frozen for later sugar analysis. Hexose, pentose, and uronic acids were me as ure . Water Insoluble Hemicellulose Fraction. The material remaining on each filter was scraped into 50 ml glass centrifuge tubes. The particles from the filter were rinsed into the tubes with 10 ml of 5% sulfuric acid (vol/vol). The centrifuge tubes were placed on a Kjeldahl digestion unit and heated for 10 min at low setting with constant turning and occasional mixing. The material in the tubes was not allowed to boil. The contents were then filtered through the appropriate filter into vol- umetric flasks. A few ml of sulfuric acid were used to rinse the filters. Two additional extractions were performed, repeating the above procedure. After the last extraction, the filters were rinsed with about 3 ml of deionized water. The samples were frozen until later sugar analysis. Hexose, pentose, and uronic acid content were measured. The material remaining on the filters was rinsed with ethanol and anhydrous ether and allowed to air dry. Cellulose Extraction. The filters were placed in 83 x 48 mm beakers, and approximately 25 ml of cold 72% sulfuric acid (wt/wt) were added. A glass stirring rod was placed in each beaker and the mixture was stirred to disperse the material; the filters were covered with Parafilm and refrigerated at 2-40C for 48 hrs. The mixture was stirred each day and if necessary, more acid was added to cover the material. After 48 hrs, the acid solution remaining on the Gooch filter crucibles was filtered 34 into flasks and rinsed with deionized water. Any acid remaining in the beakers was also rinsed and filtered. The material remaining on the filters was rinsed with 95% ethanol and ethyl ether, then air dried. The samples were frozen until hexose, pentose, and uronic acid content was measured. Lignin Quantitation. The filters were rinsed with acetone, then dried under vacuum at 1000C for 30 min. They were then cooled in a desiccator and weighed, after which they were ashed at 525°C in a Barber—Coleman Oven for 24 hrs, cooled and reweighed. Percentage lignin was calculated as the difference in weight before and after ashing, divided by the sample weight, multiplied by 100. Figure 7 outlines the entire extrac- tion scheme described above. Sugar Quantitation The concentration of hexoses, pentoses, and uronic acids was deter- mined colorimetrically on a Beckman Spectrophotometer Model DB-G at 620 nm, both 600 and 670 nm, and 525 nm respectively after derivatization as described below. Standards between 0.0 and 80 ug/ml were run with each set of hexose and uronic measurements; hexose standards consisted of glucose and uronic standards consisted of galacturonic acid. Standards containing between 0.0 and 30 ug/3 ml were run with each set of pentose measurements using a xylose standard. Hexoses were also run with the pentose measurements to account for hexose interference. The sugars were desiccated over CaSO4 at room temperature. All standards were diluted from a fresh 1% solution and were frozen until needed. The standard curves are contained in the appendix. 35 Chlorophyll Extraction (Ethyl ether; beans and peas only) I Total Water Solubles Water Soluble Hemicelluloses (Water, room temp; two, 2 hr ———————————>~(Starch extraction with extractions) amyloglucosidase; dialysis) I Starch Extraction o (Amyloglucosidase, 37 C/18 hr, 2M acetate buffer 0 pH 4.5) I Protein Extraction (Pepsin, 37 C/20 hr, 1:1 water/0.2 N HCl) I Pectin Extraction (Sequester divalent ions w/ EDTA; demethylate pH 11.5-—>5.5; Pectinase, room temp/1 hr) I Water Insoluble Hemicellulose o (5% v/v H2304, three, 10 min extractions, temp below 100 C) Cellulose 0 (72% w/w H2504, 48 hrs/2-4 c) Lignin (Wt. after cellulose -wt. after ashing) Figure 7. Extraction scheme for fiber components. 36 Hexoses. Hexoses were determined using the anthrone reaction (Hassid and Newfeld, 1964). Test tubes containing 1 ml of sample at an appropri— ate dilution were taken from the freezer and thawed. Five ml of anthrone solution were added and the solution was mixed with a test tube mixer. The anthrone, which was not held beyond 36 hrs, was made by dissolving 0.1 g of anthrone in 76 ml of sulfuric acid and 24 ml of deionized water. The test tubes were heated at 100°C for 20 min, then cooled to room temperature in tap water before measuring the transmittance. Pentoses. Pentoses were measured using the orcinol method of Albaum and Umbreit (1947). Test tubes containing 3 ml of sugar solution were I removed from the freezer and thawed. Three ml of ferric chloride sol- ution and 0.3 ml of orcinol solution were added and the mixture stirred. Ferric chloride solution was prepared by dissolving 0.1 g of FeCl3 in 100 ml of HCl. Orcinol solution was prepared by dissolving 5 g of orci— nol in 50 ml of 95% ethanol. The test tubes were heated at 100°C for 45 min, then cooled to room temperature, after which the transmittance was read immediately. Uronic Acids. Uronic acid concentration was measured by the carbazole reaction of Dische (1947). Test tubes containing 1 ml of solution were taken from the freezer and warmed to room temperature. Six ml of con— centrated sulfuric acid were added and mixed. The test tubes were heated at 100°C for 25 min and cooled to room temperature in tap water. Two tenths ml of 0.1% carbazole were added quickly and the test tubes were agitated. Carbazole solution was prepared by dissolving 0.1 g of carbazole in 100 ml of 95% ethanol. The transmittance was read promptly after two hours. 37 Data Analyses The data were analyzed with a PDP 11/45 computer. Standard curves of glucose, xylose, and galacturonic acid were obtained by plotting concentration versus absorbance using a composite of all individual standard values. A least square fit was used to determine the slope and y-intercepts. Total micrograms of hexose and galacturonic acid were calculated by determining the micrograms in each test tube ((absorbance - y—intercept) + slope) and multiplying by the appropriate dilution factor. Total micrograms in the pectinase blank were subtracted for the pectin extrac— tion to account for the color reaction with pectinase. No other compo- nents were found to interfere with color reactions. Total micrograms of pentoses were determined using simultaneous equations because of the hexose absorption at the same wavelength. Absorbances were calculated at 600 and 670 nm. The absorptivity of glucose and xylose was determined at both wavelengths and the following equations were solved: _ = pent pent y hex hex ABS blank CONC (E670) . CONC (E670) 670 _ _ pent pent + hex hex ABS600 blank — CONC (E600) CONC (£600) The appropriate factor was again used and total micrograms due to enzyme interference for pectin again subtracted out. The percentage of each substance was calculated as follows: % = total u x dilution x polymerization x 100 sample wt (pg) factor factor 38 Polymerization factors were calculated to convert monosaccharide to polysaccharide by taking the molecular weight of hexose and pentose, subtracting the weight of one water molecule which is added upon hydrol- ysis, and dividing by the sugar's molecular weight. Since galacturonic acid was in the monohydrate form, two water molecules were subtracted. The following polymerization factors were used: hexose = 0.90, pentose = 0.88, uronic acids = 0.83. Statistical Analyses. Orthogonal comparisons were made for differences due to cooking, differences among the five vegetables themselves, and differences due to media pH, divided into linear, quadratic, and cubic effects. The data were analyzed to determine whether the vegetables reacted the same to media pH (media pH x vegetable). Each vegetable was analyzed for the effect of cooking, which was divided into linear, quadratic, and cubic effects. These analyses were calculated for the hexose, pentose, and uronic components in each fraction as well as the total of these components. Table 2 shows the statistical scheme fol- lowed in this experiment. The LSSTEP statistical program was run on the CDC 6500 to obtain prediction equations for shear, predicted from fiber components, the independent variables. HPLC Sample Preparation and Analysis The raw vegetable samples were analyzed by high-pressure liquid chromatography (HPLC) in order to quantitate the following neutral sugars: glucose, galactose, xylose, arabinose, and mannose, and to examine the extracted fractions for completeness of separation in the 39 Table 2. Scheme for analysis of variance and orthogonal comparisons of vegetable fiber data. Source of variation df Total 74 Batch 14 Media pH 4 Raw vs. Cooked 1 Linear 1 Quadratic 1 Cubic 1 Raw vs. Cooked 0 pH 6 1 Vegetable 4 Media pH x Vegetable1 12 (Individual Vegetable) Raw vs. Cooked 1 Linear 1 Quadratic 1 Cubic 1 Error 40 1Includes Media pH x Vegetable4 confounding which is not part of the interaction; therefore, df # 16. 4O extraction scheme. The pectin, hemicellulose, starch, and protein frac- tions were analyzed for each vegetable using two replicate samples from each raw vegetable. Hydrolysis reactions for pectin and hemicellulose according to Cole (1967) and Seckinger et al. (1960) and for starch according to Southgate (1976) were attempted initially, but extremely poor chromato- grams were obtained characterized by increased tailing and poor resolu- tion. Since the amounts of sugars could not be calculated from these chromatograms after the hydrolysis procedures, all hydrolysis reactions were eliminated from the sample preparation. The sugars were quantitated by the method of triangulation by comparing sample areas to standard curves in which the percentage area versus concentration was plotted. Preparation of Fractions General Procedure. Appropriate aliquots, depending on the amount of sample available, were pipetted into beakers and neutralized with solid barium hydroxide to a pH of approximately 6.8 to 7.2 using a Beckman pH meter. The samples were centrifuged at 3400 rpm for five min using a Sorvall centrifuge, Model GLC—l. Fine porosity fritted glass funnels were used to filter the samples. The filtrate was decanted into 100 ml round bottom flasks. Ethanol/water was evaporated using a BUchi Roto- vapor—R rotary evaporator, equipped with a water bath at 45-500C for protein and starch fractions, and at 550C for hemicellulose and pectin fractions. Water aspiration was the source of vacuum. The samples were concentrated until 0.7 ml of less remained in the flask, depending on the fraction. 41 Protein Fraction. Aliquots of'25-70 ml were used for the analysis. The solvent was evaporated until 0.3 ml or less remained in the flask. Starch Fraction. The starch fraction was prepared with the omission of the neutralization step. The samples were first evaporated to approx- imately 5 to 13 ml, then injected into the HPLC at the 4X differential refractometer attenuation setting in order to accomodate the relatively large glucose peak. After this analysis, each sample was further evap- orated to a volume of 0.5 ml or less and run at the 2X attenuation set- ting to analyze for the other sugars. Hemicellulose Fraction. Aliquots containing 20-70 ml were used for the analysis. Final sample volumes of 0.5 to 0.7 ml were obtained. Pectin Fraction. After the filtration step, the solution was passed through a 1.5 cm x 5 cm column containing analytical grade mixed-bed ion exchange resin AG 501-X8 (D) from Bio-Rad Laboratories, in order to remove uronic acids. After washing the column with 10 ml of deionized water, the solution was evaporated until 0.5 to 0.7 ml remained. Figure 8 shows the preparation scheme described above. Materials Glucose, galactose, xylose, arabinose, and mannose standards as well as other chemicals were analytical reagent grade. The solvent, water, was HPLC grade (J.T. Baker Chemical Co., Phillipsburg, NJ), Prior to injection, all samples and solvent were filtered through 0.45 pm membrane filters (13 mm) (Gelman Instrument Co., Ann Arbor, MI). Ten ml Teflon—capped vials were used for the samples. The water was degassed by water aspiration. Duplicate 20 ul injections were made. 42 Aliquot (Protein, Starch, Hemicellulose, Pectin) Neutralize with Ba(0H)2 (omit for starch) Centrifuge @ 3400 rpm/5 min Filter ., Pass pectin extract through mixed-bed resin to remove uronic acids Concentrate A I Inject into HPLC I Quantitate: glucose galactose xylose arabinose mannose Figure 8. Preparation of raw vegetable fractions for HPLC analysis. 43 Quantification Standard curves were prepared as described previously for all five sugars under investigation, using the 2X detector setting. An additional curve was plotted for glucose (in starch) at the 4X detector setting. The amount of each sugar was read from the standard curve in mg/ml. To obtain the percent sugar in each of the four fractions (starch, protein, hemicellulose, and pectin) on a dry basis, the following equations were used: .Standard curve Final volume, ml Total volume of each concentration value X aliquot volume, ml X fraction, ml in mg/ml = Total mg in each fraction (A) 2. Percent sugar on a dry basis = A x 100 Initial freeze—dried sample wt, before fiber extraction procedure (approx. 200 mg Equipment The system consisted of a solvent reservoir heated to 45°C, a pump (Waters Associates, M-45 Solvent Delivery System), a Model U6K injector (Waters Associates, Milford, MA) a refractive index detector (Waters R401 Differential Refractometer), and an OmniScribe recorder (Houston Instruments). The HPLC carbohydrate analysis column (300 mm x 7.8 mm) was packed with Aminex HPX-87P, heavy metal (lead) form, 8% crosslinked 9 pm polystyrene cation exchange resin obtained from Bio- Rad Laboratories (Richmond, CA). The column was maintained at 850 : 0.010C by an Exacal water bath, Model EX-lOO (Neslab Instruments, Inc.). An Alltech Associates (Deerfield, IL) water jacket surrounded the column. A 25 pl syringe (Hamilton Co., Reno, NV) was used for all 44 injections using a sample volume of 20 ul. Operating conditions were as follows: Eluant: Water Flow Rate: 0.4 ml/min Temperature: 85°C Detector: Refractive index 0 2X (setting at 4X for glucose peak of starch) Chart Speed: 0.5 in/min RESULTS AND DISCUSSION Rationale for Determining Composition of Fractions from Separation Scheme Table 3 shows a summary of the data interpretation for combining hexose, pentose, and uronic acids in various fractions. The rationale behind these combinations stems mainly from the HPLC results shown in Table 4. HPLC was performed specifically to pin-point the exact nature of hexose and pentose sugars. Starch, The vegetables all show glucose as the major component of this fraction according to the HPLC results in Table 4. No galactose or xylose were seen, although minor amounts of arabinose and mannose were evident. Obviously, one would expect glucose to predominate in the starch fraction, since the samples were incubated with amyloglucosidase. Most researchers have not usually analyzed the sugars found in each fraction, rather, they have measured the sugars found in the combined water soluble fraction versus the sugars found in the combined water insoluble fraction. Since the hexoses were found to be composed solely of glucose, the hexose sugars measured were assumed to be glucose from starch with possibly some glucomannans contributing since mannose was detected. The pentose sugars detected in the extraction procedure were added to the insoluble hemicellulose fraction because arabinans are known to be a hemicellular component (Nelson, 1960; Cole, 1969; Collings and Yoko- yama, 1979; Jones et al., 1979). 45 46 Table 3. Summary of hexose, pentose, and uronic acid combinations Fraction hexose pentose uronic acids Starch Starch Insol. Hemi Insol. Pectin Protein -—— Insol. Hemi ND Soluble Hemi -—- Soluble Hemi Soluble Pectin Insol. Hemi ——- Insol. Hemi —— I Insol. Pectin ——- Insol. Hemi Insol. Pectin Cellulose Cellulose Insol. Hemi -—— Cooking Solution —— Cooking Soln. Cooking Solution Fiber Fiber Fiber _ ND not determined -—-not added to any fraction 47 mwzpm> czow do cowumm>oc ucmucmpm 8:8 :mm: H I S. H 5.0 mm. H 8.0 I mo. H 85 $32538 I 8. H 3.0 om. H £5 84 H a: 8. H 8.0 mag I 8. H 88 8. H 2.0 8. H 8.0 S. H a: 8838 I 8. H 2.0 I I S. H 2.0 $8 I 8. H 8.0 mm. H 88 8. H was 8. H 8.0 :8 :HwHoca 8. H 85 S. H 85 I I cm. H 8.0 330538 8. H 85 8. H 38 I I 8. H w: 29.3 S. H 35 8. H 8.0 I I See H 8.8 88.8.8 8. H 88 I I I New H 8.: ES mm. H Rd 8. H 2.0 I I we. H 3m 28 eaempm wmoccme mmocvnecm ewe—Ax wmoHompem amou:_m :owHomcw idmwmwn zany Hzmwmz m—mamm :o uwmma & :oHHomcme H mw—nmuwmw> :8; 2w meowpomce Leave to mpcmucoo cmmzm pmcpzwz .v mpnmp 48 H0..H 00.0 00. + 00.0 00..H 00.0 00. + 00.0 00. + 00.0 0030_000000 00..H 00.0 H0..H 00.0 H0..H 00.0 .1. H0..H 00.0 00000 0H..H 00.0 H0..H 00.0 00..H 00.0 00. H.00.0 00..H 00.0 00000000 00..H 00.0 00 .H 00.0 00..H 00.0 00..H 00.0 00..H 00.0 0000 H0..H 00.0 00..H 00.0 00..H 00.0 00..H 00.0 00..H 00.0 0000 wmo—:__wUHEo: 00..H 00.0 00..H 00.0 00..H 00.0 11. 00.0.H 00.0 0020_00_000 00..H 00.0 0H..H 00.0 00..H 00.0 .11 00..H 00.0 00000 00.0.H 00.0 00..H 00.0 00..H 00.0 .11 00..H 00.0 00000000 00..H 00.0 00..H 00.0 00..H 00.0 .11 00..H 00.0 0000 00..H 00.0 00..H 00.0 00..H 00.0 .1. H0..H 00.0 0000 000000 0000005 000500.00 000?? 0080200 00803 00.50009, “0.0000 00.03 £3.53 2300 :0 00000 .0 00500.35. A.0_00000 0 00000 49 The uronic acids detected in the starch fraction were added to the insoluble pectin fraction. Potatoes contain a relatively high proportion of pectin compared to other vegetables. However, more pectin (as uronic acids) was extracted in the starch step than in the pectin step. This has also been found by Jones (1951). The values for the uronic acids contained in the starch fraction are listed in the appendix. The ratios of hexoses to pentoses are greater in the HPLC method than for the colorimetric method. The hexoses from the HPLC data agree with the colorimetric data whereas the pentose values do not. Although the samples were not hydrolyzed for HPLC analysis, a starch- rich vegetable such as potato shows a 30% difference between HPLC versus the colorimetric method, while corn and peas show only 8% and 17%, respectively. Protein. The protein extraction contained measurable hexoses and pen— toses (appendix). The HPLC analyses show galactose as well as glucose comprising the hexose sugars. Keegstra et al. (1973) found that 3,6- linked arabinogalactans are attached to the serine portion of the cell wall protein. They also found that arabinosyl tetrasaccharides are glycosidically attached to hydroxyproline residues of the wall protein. Mod et al. (1981) found that rice hemicellulose fractions also contain protein. The glucose found by HPLC was probably leftover from the starch fraction. The pentose sugars contained mostly xylose together with some arabinose. These sugars comprise insoluble hemicellulose and were added to this fraction. Uronic acids were not measured for the protein fraction, however, if they had been measured, they would have been added 50 to the water insoluble pectin fraction since Keegstra et al. (1973) showed that protein is probably bound also to pectic polysaccharides. The galactose detected may have originated from uronic acids, but hemi- cellulose should not be ruled out. Water Insoluble Pectin. Upon looking at the results for insoluble pec— tin (Table 4), it is obvious that galactose, the major neutral sugar of pectin, was not detected chromatographically. The reasons behind this omission will be discussed at the end of this section. Perhaps galactose was inadvertently combined with another sugar because of its inexact elution properties. Konno et al. (1980) detected 72.7% galac- tose in lemon pectin, followed by 14.6% arabinose, 4.1% rhamnose, 3.9% xylose, and 1.0% glucose; they indicated that neither mannose nor fucose could be detected. Hoff and Castro (1969), using oxalic acid- ammonium oxalate to solubilize pectic substances, indicated that the pectin polysaccharides from potatoes were also composed mainly of gal- actose (86.0%), followed by rhamnose (6.0%), arabinose (5.6%), xylose (1.8%), and fucose (0.6%). Although Hoff and Castro (1969) found 7.1% glucose in the pectin fraction, they corrected for this sugar by attrib- uting its presence to starch contamination. All of the vegetables in this study showed glucose as the predom- inant sugar in the pectin fraction. This was probably due to unextracted starch or possibly, hemicellulose. Potatoes showed the largest amount of residual glucose, which agrees with the fact that potatoes contain more starch than any of the vegetables under analysis. Mannose is not generally regarded as part of the pectic substances, but a considerable amount was detected in all of the vegetables under investigation. 51 Mannose is usually considered to be associated with the hemicellulose fraction; perhaps glucomannans were part of this fraction. Due to the uncertainties of the exact nature of the hexoses, they were not combined with any of the other fractions. Although rhamnose determinations were not made, it is probable that this sugar was contained in the pectin fraction. The galactose included in hemicellulose is probably from the galactans in pectin, since the aforementioned researchers have found mainly galactose in pectin. These findings indicate the need for better methods of separating pectin and hemicellulose, since they are closely associated with one another within the cell wall. The pentose sugars were composed mainly of xylose. However, this finding does not agree with those mentioned above. This also supports the theory that hemicelluloses were probably extracted along with the pectin. Therefore, the pentoses were added to the hemicellulose fraction. Polygalacturonic acid constitutes the major portion of pectin, thus the uronic acids were assumed to be pectin. However, more uronic acids were extracted in the starch step than in the pectin fraction. Water Insoluble Hemicelluloses. Hemicelluloses contain a xylan back- bone with covalently attached glucomannans, arabinogalactans, and galactomannans. However, in this study, xylose predominated only in potatoes while corn and cauliflower contained mainly galactose, peas showed mostly glucose, and beans contained mostly arabinose. These results are very inconsistent with the results of others such as Hoff and Castro (1969). They found mostly glucose (56.7%) in the hemicel- lulose polysaccharides of potatoes, followed by xylose (23.1%), galactose 52 (12.0%), mannose (12.0%), arabinose (2.1%), and rhamnose (0.5%). In the present study, xylose was identified as the predominating sugar in pot- atoes, with glucose following. The ratio of glucose to xylose in the Hoff and Castro (1969) study was 2.5, whereas in the current study, the ratio of 0.6 was obtained. However, as stated in the pectin discussion, Isome pectin was included in the hemicellulose fraction. By adding to- gether the results of the hemicellulose and pectin fractions, it is evident that the results more closely approximate those of Hoff and Castro (1969), for example, the glucose to xylose ratio would be equal to 2.3 for potatoes rather than 0.9 which better resembles the liter- ature value of 2.5. One must bear in mind that Hoff and Castro (1969) used alkali to solubilize hemicellulose which differs from the proce— dure used in this study. It appears that with the extraction technique used for vegetables in this study, the hemicellulose was extracted in earlier extraction steps; however, Jeltema and Zabik (1980) did not find this true for cereals. According to Jones et al. (1979), xylose was the major neutral sugar in the hemicellulose fraction of corn, followed by glucose, arabinose, galactose, and mannose in decreasing order. The ratio of glucose to xylose was 0.2, as opposed to 0.5 in the present study. Galactose predominated, followed by arabinose, xylose, mannose, and glucose. Uronic acids have also been found in hemicellulose prepara- tions (Seckinger et al., 1960; Cole, 1969). From the results of the aforementioned investigators, it seems reasonable that the water insoluble hemicelluloses in this study may be composed of the sum of the hexoses + pentoses + uronic acids. 53 However, only the pentoses were added to the hemicellulose fraction since the hexoses were shown to contain a larger proportion of galactose as compared to glucose in this study; perhaps the galactose may be from the pectin fraction as this sugar has been shown to constitute a major por- tion of pectin (Hoff and Castro, 1969; Konno et al., 1980). The glucose detected in this fraction is probably from hemicellulose, although Theander and Aman (1981) have indicated that the more accessible and amorphous regions of cellulose microfibrils may be hydrolyzed by dilute sulfuric acid. They stated that this depends largely upon the plant material under consideration. However, Aspinall (1963) stated that con- clusive evidence for this phenomenon has not been reported; the results may be attributed to changes in the physical state of the molecules such as variations in degree of hydrogen bonding. Due to these uncer- tainties, hexoses were not added to any fraction. The uronic acids were not combined with any fraction since it could not be determined which portion was due to galacturonic acid found mainly in pectin, and which was due to glucuronic acid, contained largely in hemicellulose (Selvendran, 1981). Total Water Solubles. Total water soluble material includes water sol— uble starch, sugars, and hemicellulose. Total water soluble results are found in the appendix. Cooking Solution Fiber. Cooking solution fiber represents the material extracted into the cooking media. Cooking solution fiber values were obtained by adding the pentose and uronic measurements obtained after dialysis which represent soluble hemicellulose plus pectin. The hexose values were not added since it was not known to what extent soluble 54 starch and pectic or hemicellular sugars contributed to this measurement. Water SOlUble'HemiceIIUTOSe and Pettin. The pentose values of the water soluble fraction after dialysis represent the water soluble hemicelluloses, while the uronic values of this fraction were taken as pectin. The divi- sion between soluble pectin and hemicellulose was not absolute, since uronic acids are found in hemicellulose fractions and components such as arabinans are found in pectin. Again, this demonstrates the need for a better method of separating hemicelluloses and pectins. The hexose values were not added to either fraction for the same reasons previously described in the cooking solution fraction. Cellulose. The hexose values obtained from the cellulose extraction represent cellulose. Pentose and uronic acids were also measured in the cellulose fraction. The pentoses were attributed to remnants from the hemicelluloses and were added to this fraction. Jones et al. (1979) found mostly glucose (82%) in the cellulose fraction of corn, followed by xylose (10.5%), arabinose (3.3%), mannose (3.0%), and galactose (1.6%). The ratio of hexoses to pentoses was 6.3; in the current study this ratio was 3.5 in the raw corn sample. The Keegstra et al. (1973) model of the plant cell wall (Figure 1) shows cellulose fibrils closely asso- ciated with xyloglucans and arabinans which probably contribute to the pentoses found by Jones et al. (1979). Results of Extraction Procedure Shear Press Data Figure 9 shows a plot of the force to shear versus pH of the five vegetables. In the raw state, the firmness was found to decrease in the 55 following order: cauliflower, beans, potatoes, peas, and corn. All of the vegetables were firmest at pH 4 in the cooked state. At pH levels above and below 4, all of the vegetables showed a decrease in firmness. In general, the vegetables were softest at pH 10; at pH 2, the firmness was only surpassed by the results at pH 4. These results agree favor- ably with those of Doesburg (1961) who showed that plant tissues possess maximum firmness at pH 4.0-4.5. He observed decreased firmness at pH values both higher and lower than pH 4. The reasons behind these pheno- mena and their relation to fiber components will be discussed further. Statistical analyses of the shear data are shown in Tables 5 and 6. All vegetables were affected by cooking and pH. The vegetables did not react the same to media pH, as reflected by the significance of media pH x vegetable interaction. The terms linear, quadratic and cubic refer to the shape of the curves in a plot of pounds force versus pH of vegetable or percent fiber component versus pH of vegetable. A linear plot represents a straight line, where the x and y components may be either directly or inversely proportional to each other. A quadratic plot appears like a symmetric peak. A cubic plot looks somewhat like a double sine curve. Any combination of these three types of curves is possible. Beans and cauliflower showed the most dramatic changes in comparing raw and cooked vegetables. They exhibited significant linear effects since their firmness decreased as pH increased. They showed significant quadratic and cubic effects since at pH 4 they were firmer than either pH 2 or 6. Corn and potatoes showed no significant effects, except in comparing the raw and cooked vegetables. Peas were significantly firmer in the raw state rather than in the cooked state and showed a significant 56 —O—POTATOES —D——-—-CAULIFLOWER ———v——BEANS —00~—————CORN ——©—-PEAS 9 8 7i Force to Shear (lb/g) O 25‘ o 1— - ° 0 I ——O 0 \ RAW 2 4 6 8 10 pH Figure 9. Firmness of vegetables in raw and cooked states. 57 Table 5. F values for analysis of variance and orthogonal comparisons of the vegetables in the raw and cooked states for shear data Source F value Media pH 575.13** Raw vs. cooked 2061.03** Linear 188.47** Quadratic 33.54** Cubic 20.93** Raw vs. cooked @ pH 6 1303.28** Vegetable 257.34** Media pH x Vegetable 66.79** **P<0.01 58 H0.0Va*r mo.ona amm.m aawm.mH mm.H Ho.o enm.¢ ownzu amH.n ramm.mm 00.0 om.H ow.m ovumevwzo *emm.ofi eeqo.¢fim mm.o mn.m armm.wm wacvg ramm.mN¢H armm.omfi ar-.m¢m ermm.om ermm.mm¢ vmxooo .m> 30m LGZO C... szu mcmwm mwOHwaba CLOQ mama 0H0u c0050 00H :0 00000Hmmm> 000000 .0> 300 cow 00000000500 F0como;uco to mwzpm> a .o m—nmp 59 linear effect for the same reason as for cauliflower and beans; they also showed a significant cubic effect due to the firmness at pH 4. Cooking Solution Fiber This fraction represents the material extracted into the cooking media, plotted in Figure 10. The individual values for the pentose and uronic acid components comprising this fraction are listed in the appen- dix. In general, the most cooking solution fiber was extracted at pH 10, the lowest amount at pH 4. Cauliflower and peas showed the most material extracted into the cooking media, followed by beans, potatoes, and corn. Corn showed very little material extracted at all pH levels, which re- lates well to the shear data, also showing little change with pH. Upon comparing the individual pentose values and uronic acid values comprising cooking solution fiber, in general, the uronic components outweighed the pentose components for all vegetables at all pH levels, except for corn. The trends for the pentose fractions were similar in all vegetables: the amount of soluble hemicellulose did not vary sig- nificantly in pH levels 2, 4, and 6, however, a distinct increase was exhibited from these three pH levels to pH 10. In general, the uronics showed small increases in soluble pectin from pH 2 to pH 6, with a much more rapid increase up to pH 10. According to the statistics in Table 7, peas showed a significant (p<:0.01) linear effect since percent cooking solution fiber increased as the pH increased, and a significant quadratic effect due to the increase in fiber at pH 6. Potatoes, beans and cauliflower showed similar trends while corn showed no significant changes. According to Doesburg (1961) higher pH's are effective in increasing 60 -—-Cr——-POTATOES -C}——CAULIFLOWER ——V—BEANS -Cr-—--CORN -—-——43-—PEAS 2.01 1.5- 1.0 % Cooking Solution Fiber 0.5r Figure 10. Cooking solution fiber in vegetables as a function of pH. 61 Table 7. F values of orthogonal comparisons for cooking solution fiber. Vegetable F value Peas Raw vs. cooked 194.15** Linear 579.92** Quadratic 150.49** Cubic 3.36 Corn Raw vs. cooked 1.73 Linear 0.37 Quadratic 0.08 Cubic 0.04 Potatoes Raw vs. cooked 11.27** Linear 40.47** Quadratic 4.90* Cubic 0.33 Beans Raw vs. cooked 25.55** Linear 37.69** Quadratic 9.17** Cubic 0.07 Cauliflower Raw vs. cooked 487.22** Linear 1470.35** Quadratic 230.43** Cubic 0.58 *P <0.05 **P <0.01 62 soluble pectin as demonstrated in Figure 10. This is a result of the shortening of the pectin chains with the formation of carboxyl groups through de-esterification of the carbomethoxy groups. By pH 10, severe depolymerization occurs, causing increased soluble pectin. Simpson and Halliday (1941) found that the cooking water from both carrots and parsnips contained increased levels of pectic substances after 20 minutes of steaming. After 45 minutes of steaming, the pectic substances increased 5- to 7-fold. These increases were due to the hydrolysis of protopectin to form pectin, although some of the pectin decomposed during the cooking process. Heller et al. (1977) found that insoluble hemicellulose content plateaued at pH 6.6 and decreased from there until pH 11.5. Selvendran et al. (1979 ) found increased xyloglucans in alkaline media. In the present study both the hexose and pentose values showed large increases in soluble hemicellulose at pH 10 which would agree with these two inves- tigators. Water Soluble Fiber Components Figures 11 and 12 show water soluble fiber components. Table 8 shows statistical analyses of this data. Water Soluble Hemicellulose. Water soluble hemicellulose increased with pH except for corn, which showed no change (Figure 11). Cauliflower and beans showed that water soluble hemicelluloses degrade as the pH increased, evinced by the rising slope of the graph, proceeding from pH 2 to pH 10, with the most degradation at pH 10. This agrees with the finding of Van Soest and Robertson (1977). However, they also found increased 63 ——o—-POTATOES —D-——CAULIFLOWER -——v——BEANs —<>—-c0RN 3.5“ ——©—PEAS I 3.0 2.5 a 2.0 O 3 E E 0’ 1.5 I OJ 3 3 ,g I N I 1.0 V O 0 w Figure 11. Percentage water soluble hemicellulose in vegetables as a function of pH. 64 —O——-POTATOES —D—-—CAULIFLOWER ———V—BEANS —<>——c0RN I 3.0 2.5— 2.0 .E I .p 8 D. Q) E 1.5- I ' B U') 33 I V fr:— 0 ° : ’ of #0 0L . 4 RAW 2 6 8 16 pH Figure 12. Percentage water soluble pectin in vegetables as a function of pH. 65 Table 8. F values of orthogonal comparisons for water soluble fiber components Vegetable Water Soluble Water Soluble Hemicellulose Pectin Peas Raw vs. cooked 0.17 1.88 Linear 0.26 0.11 Quadratic 4.57* 0.43 Cubic 0.42 0.03 Corn Raw vs. cooked 0.05 0.05 Linear 0.02 0.00 Quadratic 0.02 0.06 Cubic 0.01 0.00 Potatoes Raw vs. cooked 6.05* 4.63* Linear 0.11 0.15 Quadratic 0.03 0.00 Cubic 0.69 0.01 Beans Raw vs. cooked 15.79** 10.60** Linear 25.90** 9.36** Quadratic 0.28 0.94 Cubic 0.08 0.20 Cauliflower Raw vs. cooked 175.32** 22.62** Linear 383.67** 25.77** Quadratic 55.61** 18.20** Cubic 1.03 0.10 *P< 0.05 **P<0.01 66 soluble hemicellulose in acidic media, but this may not be true for all plant tissues. In this study, cauliflower, peas and potatoes showed slight increases in soluble hemicellulose at pH 2. Potatoes, beans and cauliflower had significantly more soluble hemicellulose extracted in the cooked vegetable than in the raw (Table 8). Neither peas nor corn showed these trends, although peas exhibited a significant quadratic relationship due to the increased soluble hemicellulose at pH 2. Caul- iflower and beans showed the most striking differences, both exhibiting a highly significant linear relationship because of the rapid rise in soluble hemicellulose from pH 4 to pH 10; a quadratic relationship was evident for the cauliflower, caused largely by the increase at pH 2. These results relate directly to the cooking solution fiber trends which also showed increased hemicellular material, proceeding from pH 2 to pH 10. The greatest incline in both cases occurred from pH 6 to pH 10 in all vegetables except corn. Water Soluble Pectin. The soluble pectin values were generally higher than those of the soluble hemicelluloses. Cooking increased water soluble pectin, except in the corn, which showed no change (Figure 12). This correlated well with the shear data, which also showed no significant effects for corn. Cauliflower, beans, and peas contained more soluble pectin at pH 2 than at pH 4. Cauliflower and beans showed the most dramatic increase, proceeding from pH 6 to pH 10. This data, which correlates inversely with the shear values, agrees favorably with Doesburg's (1961) results. The increase in soluble pectins in acidic media may be due to the ex— change of calcium bridges in protopectin by hydrogen ions (Mattson, 1946). 67 Doesburg (1961) did not see any reason for the occurrence of pectin depolymerization in acidic media, but this is most likely to occur in alkaline media as shown by the large increase in soluble pectin at pH 10. These results relate well to the pectic components of the cooking solution fiber data which also show a decrease in pectin at pH 4 and dramatic increases by pH 10. The orthogonal comparisons showed the exact same trends as the hemicelluloses, with the exception of peas, which showed no significant effects (Table 8). Non-fibrous Components Starch. The data in Table 9 show the starch contents of the vegetables in the raw and cooked states. Potatoes contained the most starch in all of the five cooking conditions, followed by corn, peas, beans, and caul- iflower. The raw vegetables all showed less starch than the cooked vege- tables except for cauliflower; however, the raw value showed a higher standard deviation for cauliflower. Peas and potatoes showed signifi- cantly more starch in the cooked vegetables than in the raw (p<:0.01, Table 10). Potatoes were also significant (p<:0.05) in the linear rela- tionship because of the rapid increase in starch from the raw state to pH 2, and the linear decline thereafter. All other cooked vegetables did not show significant differences in starch content, as expected. The vegetables did not react the same to media pH, as reflected by the highly significant media pH x vegetable interaction (Table 11). The starch values appear to be somewhat lower than expected; for example, most workers have shown the starch contents of potatoes to be about 80% (Englyst, 1981) on a dry basis. The methods of measuring 68 0:00H000_000 m :0 00000 “00000000000 00v :00H00>m0 0000:0H0 0:0 :00: H A00.0v Aa0.00 A00.0v A00.00 000.00 00.0 00.0 00.00 00.00 00.00 00 :0 “00.00 A00.00 A00.0v A00.00 A00.00 00.0 00.0 00.00 00.00 00.00 0 :0 A00.00 A00.00 000.00 A00.0v A00.Hv 00.0 00.0 00.00 00.00 00.00 0 :0 000.00 000.00 Aa0.00 A00.00 A00.00 00.0 00.0 00.00 00.00 00.00 0 :0 A00.0 000.0 A00.00 “00.00 “00.00 00.0 00.0 00.00 00.00 00.0 300 c030000030o mcmwm mmOH0Hoa :Lou 0000 “00000 000v :000H0 0 0_000000> 00 000000000 H00000H000> mo mpcmpcoo nucmum .m 0_000 69 Tabie 10. F va1ues of orthogona] comparisons for starch Vegetabie F vaiue Peas Raw vs. cooked 7.43** Linear 0.00 Quadratic 0.56 Cubic 0.33 Corn Raw vs. cooked 2.10 Linear 0.01 Quadratic 0.86 Cubic 0.74 Potatoes Raw vs. cooked 54.03** Linear 7.30* Quadratic 1.63 Cubic 0.36 Beans Raw vs. cooked 0.06 Linear 0.00 Quadratic 0.00 Cubic 0.01 Cauiifiower Raw vs. cooked 0.01 Linear 0.02 Quadratic 0.03 Cubic 0.02 *P<0.05 **P < 0.01 70 Tabie 11. F vaiues for analysis of variance and orthogonai comparisons of the vegetabies in the raw and cooked states for starch Source F vaiue Media pH 7.71** Raw vs. cooked 28.07** Linear 1.56 Quadratic 0.18 Cubic 1.03 Raw vs. cooked 18.76** @ pH 6 Vegetabie 650.21** N Media pH x Vegetable .71** **P (0.01 71 starch are very controversial. Many dietary fiber methods fail to remove all of the starch prior to analysis of the remaining components. Englyst's (1981) procedure involved the use of amyloglucosidase with a preliminary end-over-end rotation of the samples, three times per minute, at 900C for 5 hours to gelatinize the starch and free the sugars which contributed to the increased starch value as compared to the results in the current study. The starch values for the cooked vegetables were better than for the raw vegetables although they were still low compared to others' results. The problem lies in the method of measuring starch. According to Schweizer (1981), it may take one hour or more for adequate starch digestion. He claimed that autoclaving samples for 15 minutes at 1300C ensured complete starch gelatinization, but this may degrade pectins as well. It is noteworthy that freeze drying has been reported to increase the difficulty of hydrolyzing starch (James and Theander, 1981). Because of this, some have used Termamyl (Novo) enzyme in combination with amyloglucosidase. If the sample is autoclaved, there is no reported problem with freeze—dried materials. If the sample is not autoclaved, however, problems may ensue. No one has attempted to examine which enzymes extract less hemicellular or pectic material. Water Insoluble Pectin The water insoluble pectin which is contained in the vegetable as opposed to water soluble pectin which has leached out into solution should show opposite and inverse trends, but this is not totally the case (Figure 13). Beans contained the most pectin (as uronic acids) in the pectin fraction; however, potatoes contained the largest amounts of ‘——CF--POTATOES -—D—CAULIFLOWER 70 —V—-BEANS ‘-<>---CORN -———CF-PEAS 6... S 5l ‘1. C .5 U 0'.“ ‘1’ 3 :3 3 '8 E 2. 3 1 \U 0 1 RAW 2 4 6 8 16 pH Figure 13. Percentage water insoluble pectin in vegetables as a function of pH. 73 Table 12. F values of orthogonal comparisons for insoluble pectin Vegetable F value Peas Raw vs. cooked 0.10 Linear 0.15 Quadratic 0.03 Cubic 0.03 Corn Raw vs. cooked 0.01 Linear 0.00 Quadratic 0.00 Cubic 0.01 Potatoes Raw vs. cooked 0.45 Linear 0.76 Quadratic 1.08 Cubic 0.99 Beans Raw vs. cooked 0.40 Linear 5.85* Quadratic 1.45 Cubic 1.88 Cauliflower Raw vs. cooked 1.69 Linear 9.37** Quadratic 0.57 Cubic 0.62 *P< 0.05 **P< 0.01 74 pectin (as uronic acids) in the starch fraction. Doesburg (1961) demon- strated that at pH 4, hardnessmeter readings correlated directly with the amount of pectin in the vegetables. However, this plot resembles the water soluble pectin trends, proceeding from raw to pH 6. From pH 6 to pH 10, the trends of the two fractions do correlate inversely. By pH 10, potatoes, beans, and cauliflower show a decrease due to increased degra— dation of pectin in basic pH. Pectin is usually thought to be stable in acid and unstable in alkali (Kertesz, 1951). This was found in potatoes, beans, and cauli- flower, which showed elevated levels of pectin at pH 2 with a decline at pH 10. Albersheim (1959) demonstrated the dependence of citrus pectin solutions on pH. He found that pectin was unstable in neutral and weakly acidic conditions at elevated temperatures due to degradation of the polyuronide chain. This was evident in cauliflower and peas, which showed declines in pectin by pH 6. Proceeding from raw to pH 2 (Figure 13), one would expect to see a decline; from pH 2 to pH 4 an incline would be expected. After pH 4, one should see a decline. Perhaps the protopectin was bound in the acidic environment making it less available to the pectinase enzyme. In Table 12 one can see that peas, corn and potatoes showed no sig- nificant effects in comparing the raw and cooked vegetables. However, beans showed a linear effect (p< 0.05) because the amount of insoluble pectin decreased as the pH increased from pH 6 to pH 10. Cauliflower also showed a linear trend (p<:0.01) as the amount of insoluble pectin decreased as the pH progressed past pH 2. Corn and peas do correlate directly with the shear data. Corn 75 shows no significant effects for pectin and peas show a maximum at pH 4. Individual hexose, pentose, and uronic values for pectin are listed in the appendix. Water Insoluble Hemicelluloses Figure 14 shows the graph of water insoluble hemicellulose data; Table 13 shows statistics for this fraction. The hexose, pentose and uronic acid composition is listed in the appendix. Corn and potatoes displayed no significant changes in insoluble hemicellulose from pH 2 to pH 10, which is consistent with the shear data. More obvious changes were evident in cauliflower, beans and peas, each showing a decrease from pH 6 to pH 10. This phenomenon correlates well with the shear data, which shows a decrease in firmness at basic pH. Van Soest and Robertson (1977) and Heller et al. (1977) found that most plant tissues show increased soluble hemicellulose at both high and low pH values; at neutral pH, hemicellulose is the least soluble. This is the basis for the detergent fiber method. The vegetables con- tained less insoluble hemicellulose at basic pH, indicating decreased hemicellulose in the vegetable tissue which correlates inversely with the increase in soluble hemicellulose. However, the vegetables should demonstrate a corresponding decrease in hemicellulose in acidic condi- tions, which was not evident except in cauliflower. The increase in insoluble hemicellulose from raw to pH 2 (Figure 14) in beans, peas and corn, which one would not expect, may have occurred because the insolu- ble hemicellulose was more available for analysis to the sulfuric acid. An increase in insoluble hemicellulose should be evident at pH 6-8, but only peas and beans showed increases at pH 6. Since the insoluble 76 --43--POTATOES -43——-——-CAULIFLOWER ——-—4§F—BEANS -<>—-——-—CORN --—43——PEAS % Insoluble Hemicellulose 'RAw 2 4 6 8 W pH Figure 14. Percentage water insoluble hemicellulose in vegetables as a function of pH. Table 13. F values of orthogonal comparisons for insoluble hemicellulose Vegetable F value Peas Raw vs. cooked 0.08 Linear 1.86 Quadratic 0.06 Cubic 1.29 Corn Raw vs. cooked 0.96 Linear 0.14 Quadratic 0.31 Cubic 0.17 Potatoes Raw vs. cooked 1.00 Linear 0.00 Quadratic 0.02 Cubic 0.04 Beans Raw vs. cooked 3.79 Linear 7.08* Quadratic 2.17 Cubic 1.09 Cauliflower Raw vs. cooked 6.19* Linear 19.08** Quadratic 6.32* Cubic 0.05 *P< 0.05 **P< 0.01 78 hemicellulose values were comprised of five fractions, perhaps all of the fractions do not contain true hemicelluloses. For example, the starch pentoses were solely arabinose as indicated by the HPLC results, which may stem from either hemicellulose or pectin. Both protein and pectin pentoses were mainly xylose, but also contained arabinose. The opposite was true for insoluble hemicellulose pentoses. Again, one has the prob- lem of correctly attributing each fraction to one specific fiber compo- nent. Cellulose The cellulose values are plotted in Figure 15. The hexose, pentose, and uronic acids are listed in the appendix. The only significant (p<:0.01, F=15.95) term was the "Vegetable" component which means that there were differences among the five vegetables themselves. No significant changes in cellulose were noted due to cooking or pH, as expected. Proceeding from pH 6 to pH 10, beans showed an increase in cellulose while peas showed a decrease in cellulose. However, the value of the beans at pH 10 is questionable because of the high standard deviation. The increased cellulose values obtained by Simpson and Halliday (1941), Mattheé and Appledorf (1978), and Herranz et al. (1981), explained by increased hydrolysis and liberation of the material were not congruent with the data presented in this study. The data showed no significant changes in cellulose with cooking which correlated with the strongly acidic conditions needed to degrade cellulose. This agrees with the findings of Tauss (1889), Day (1909), Sterling (1955), and Sterling and Bettelheim (1955). 79 —O——-POTATOES -—D—CAULIFLOWER 71 _-V—BEANS —<>——c0RN —©—PEAS % Cellulose 2—4 0'246816 RAW Figure 15. Percentage cellulose in vegetables as a function of pH. 80 0901 Lignin, the material remaining after a 48 hour extraction with cold 72% sulfuric acid, is plotted in Figure 16; Table 14 shows that there were significant (p (0.01) differences among the five vegetables and that the vegetables reacted differently to media pH. Beans (p <0.01) and cauliflower (p (0.05) displayed significantly greater cellulose in the cooked vegetables compared to the raw. Beans also showed a signif— icant (p (0.01) cubic relationship as a result of the increased levels of cellulose at pH 2 and pH 6, and the decrease at pH 4. Beans and potatoes showed a sharp decline in lignin from pH 2 to pH 4. These results agree with those of Herranz et al. (1981); the authors gave no reasons for their results. Beans also showed a sharp increase in lignin from pH 4 to pH 6. This agrees with the studies of Van Soest and Robertson (1976) and Mattheé and Appledorf (1978), who attributed lignin increases mainly to tannin-protein condensation products during cooking. Non-fibrous Components in Raw Vegetables Table 15 shows the results of crude protein, crude fat, and ash determinations on only the raw vegetable reported on a dry basis. The moisture analysis shown in Table 16 was determined on each vegetable in the raw and cooked states. Crude Protein The trend in the crude protein values agrees with the results of Joslyn (1950). However, Joslyn listed cauliflower as containing the most protein, followed by peas. The maturity of the peas is inversely 81 —O—POTATOES —D———CAULIFLOWER —-——V—BEANS —<>—c0RN ———©—-PEAS 3.0- 2.5 2.0 ;. 1.5- 1'0 .‘- .. 1- o . '0 . O 0.5 ’ 0 I RAM 2 4 8 16 pH Figure 16. Percentage lignin in vegetables as a function of pH. Table 14. F values of orthogonal comparisons for lignin Vegetable F value Peas Raw vs. cooked 0.81 Linear 3.75 Quadratic 0.21 Cubic 1.32 Corn Raw vs. cooked 0.80 Linear 1.35 Quadratic 0.28 Cubic 0.65 Potatoes Raw vs. cooked 0.04 Linear 1.88 Quadratic 2.25 Cubic 1.49 Beans Raw vs. cooked 13.57** Linear 1.64 Quadratic 0.03 Cubic 23.93** Cauliflower Raw vs. cooked 4.24* Linear 0.00 Quadratic 0.57 Cubic 0.06 *P (0.05 **P (0.01 83 Table 15. Proximate analysis of raw vegetables1 0 2 o 3 O 2 Vegetable A Crude A Crude A Ash Protein Fat Peas 30.19 :_1.37 0.88 :_0.01 4.17 :_0.30 (26.8)4 Beans 19.04 :_0.41 0.81 :_0.00 7.36 :_0.26 (21.6) Corn 12.73 i 0.42 3.52 :_0.38 3.49 :_0.18 (14.2) Potatoes 10.93 :_0.01 0.23 :_0.02 5.58 :_0.53 (9.0) Cauliflower 25.33 :_1.74 1.08 :_0.23 7.64 :_0.40 (28.9) 1Reported on a dry basis 2Mean and standard deviation based on 3 replications (N x 6.25 crude protein) 3Mean and standard deviation based on 2 replications 4Values in parentheses taken from Joslyn (1950) N x 6.25 84 related to protein content. Perhaps the peas that were compared were not of the same maturity; the comparison used in this table was for "medium" maturity peas. The other three vegetables followed Joslyn's trends with beans, followed by corn and potatoes. All vegetables ex- cept potatoes showed a difference of 10 to 13% as compared to the lit- erature values; potatoes differed by 21%. All of the protein values are based on the assumption that the vegetables contain 16% nitrogen. However, this assumption is false (Joslyn, 1950). For example, corn contains the proteins zein, glutelin, and globulin, while peas contain legumin, legumelin, and vicilin, cited by Joslyn, from Jones (1931). The conversion factor would be less than 6.25 if the actual nigrogen contents were considered. Jones stated that potatoes may only contain half as much protein as is commonly sup- posed, since only one—third to one-half of the nitrogen in potatoes has been accounted for as protein. Crude Fat and Ash The highest amount of crude fat was extracted from corn, followed by cauliflower, peas, beans, and potatoes. Beans and cauliflower showed similar values for ash. The percentage ash was obtained in the following decreasing order: cauliflower, beans, potatoes, peas, and COM. Moisture Moisture was shown to increase with cooking as demonstrated in all cases except peas cooked at pH 4 (Table 16). However, this value indicates a much higher standard deviation than the others. These 85 000>wuu00000 .000000000000 0 000 0 00 00000 000000>00 00000000 000 0002 00000v 0000> 000 0000000 5000 00000 00000000000 00 00=P0> m.¢ m 000000000000 0 00 00000 000000>00 00000000 000 000: N 00000 00000 00 00000000 0000000000 H 000V 0000 00.0...0 00.00 00.0 0.. 00.00 00.0 H 00.00 00.0 .+. 00.00 00.0 .0 00.00 0088.00 0000 :00 00.0 H 00.00 00.0 H 00.00 00.0 H 00.00 000.0 0 00.00 00.0 H 00.00 e000 000V 0000 00.0 0 00.00 00.0 0 00.00 00.0 H 00.00 00.0 H 00.00 00.0 .+. 00.00 003000.038 000V 0000 00.0 H 00.00 00.0 H 00.00 000.0 0 00.00 00.0 .10 00.00 00.0 H 00.00 0000 mfiamv 00.0 H 00.00 00.0 H 00.00 00.0 H 00.00 00.0 0 00.00 00.0 H 00.00 0:000 00 :0 0 :0 0 :0 0 I0 300 000000000 0000000 00000000> 000000000> 000000 000 300 00 00000000 00000002 .00 00000 N.H 86 results agree with those of others (Simpson and Halliday, 1941; Mattheé and Appledorf, 1978) who also found more moisture after cooking due to increased hydration of the plant tissue. No significant differences were seen among the vegetables cooked at different pH's, although the moisture contents were greatest at pH 10 for all vegetables except cauliflower. McConnell et al. (1974) found that when vegetables were soaked in 0.1M NaOH versus 0.1M HCl, the water-holding capacity in— creased by 21% which agrees with the results in this study. They found the water-holding capacity of acetone—dried raw vegetables to decrease in the following order: green beans, cauliflower, peas, potatoes, and corn. This compares favorably with the results obtained for the vegetables in this research study with raw beans possessing the most moisture, followed by cauliflower, potatoes, peas, and corn, the only difference occurring between the order of peas and potatoes. The experimental moisture values also agree favorably with the values shown in parentheses, obtained from Osborne and Voogt (1978). All of the raw values are within standard error of the literature values except for corn. Perhaps a different type of corn was used in the literature. Although the literature values for cooked vegetables were not within standard error of the experimental values, a general trend towards increased moisture upon cooking was evident. Recoveries Recoveries comprise the sum total of hexoses, pentoses, and uronic acids measured in each extraction, as well as the fat, protein, ash and lignin values. Although fat, protein, and ash were measured only for the raw vegetables, the results were assumed to be the same for the 87 Table 17. Average percent recovery for vegetables Vegetable % Recovery Peas Raw 82.2 pH 2 94.1 pH 4 94.1 pH 6 94.2 pH 10 95.4 Corn Raw 97.0 pH 2 103.9 pH 4 107.7 pH 6 102.2 pH 10 102.1 Potatoes Raw 90.8 pH 2 ' 100.8 pH 4 96.4 pH 6 96.3 pH 10 94.1 Beans Raw 81.7 pH 2 98.0 pH 4 97.9 pH 6 98.9 pH 10 99.0 Cauliflower Raw 92.7 pH 2 95.8 pH 4 98.3 pH 6 98.7 pH 10 102.0 88 cooked vegetables. The recoveries are shown in Table 17. The values ranged from 81.7% to 107.7%. The raw vegetables all showed the lowest values be- cause there were no cooking solution fiber values for the raw vege- tables. Low starch values were obtained for raw peas and potatoes which contributed to their low recovery values. In general, corn ex— hibited the highest recoveries while peas showed the lowest recoveries. Relationship Between Fiber Components and Shear Force Simple Correlations A few significant simple correlations were evident between fiber components and shear force as shown in Table 18. The two major com- ponents that correlated with shear were insoluble pectin and water soluble hemicellulose. .Pectin was correlated most positively with shear (p< 0.01), fol- lowed by the pectin pentoses. The pentoses and uronics of soluble hemicellulose were negatively correlated with shear (p< 0.05), as well as the sum of the soluble hemicelluloses. This agrees with the find- ings of Doesburg (1961) and Reeve (1970) who agreed that changes in pectin influence texture in plants. Lignin showed a positive correla— tion with shear, although not significant. However, the pectin x lig- nin interaction was significant at p< 0.05. These findings indicated that both pectin and lignin contribute to increased firmness in vege- tables. Other terms involving pectin were highly significant such as: pectin squared, cellulose x pectin, and pectin x hemicellulose. It is well known that the loss of hemicelluloses contribute to the softening of vegetables. This was reflected by the negative 89 Table 18. 'Simple correlations of fiber components with shear force Component Shear Cellulose Hexoses (CELL) 0.05 Pectin Uronics (PECT) 0.61** Lignin (LIG) 0.14 Water Soluble Hemicellulose Hexoses -0.21 (WSHEMHEX) WSHEMPENT -0.37* WSHEMURO (NS Pectin) -0.37* SUMWSHEM (Zlhexose, pentose, uronic) -0.35* CELLPENT 0.17 Protein Pentoses (PROPENT) 0.00 Starch PENT -0.01 PECTPENT 0.56** Water Insoluble Hemicellulose Hexoses -0.19 (HEMHEX) HEMPENT 0.08 HEMURO - 0.01 SUMHEM (Izhexose, pentose, uronic) -0.14 CELL SQ 0.03 PECT SQ O.59** LIG SQ 0.17 SUMWSHEMSQ -O.31 *P<0.05 **P<0.01 Table 18 (cont'd.) 90 Component Shear SUMHEM SQ -0.02 CELL x PECT 0.54** CELL x LIG 0.14 CELL x SUMWSHEM —0.30 CELL x SUMHEM -0.04 PECT x LIG 0.44* PECT x SUMWSHEM 0.32 PECT x SUMHEM 0.53** LIG x SUMWSHEM -0.25 LIG x SUMHEM -0.03 SUMNSHEM x SUMHEM -o.44* 91 correlations found in all terms involving water—soluble hemicellulose. Shear Prediction Equationsfor Cooked Vegetables at Various pH Values The prediction equation, calculated using fiber components as the independent variables, showed that shear can be predicted for vegeta— bles cooked at different pH values, by insoluble pectin, lignin, and water-soluble fiber (sum of hexoses, pentoses and uronics): SHEAR=1.5174 PECTIN - 2.5967 LIGNIN — 0.3727 SUMWSHEM + 4.4454 The regression coefficient shows how well the equation predicted shear. Since R2=0.66, 66% of the variation in tenderness can be pre- dicted by this equation. The above equation included both raw and cooked vegetables. When only the raw vegetables were included in the equation, the regression coefficient increased to 0.87, indicating that 87% of the variation in tenderness could be predicted in the raw vege- tables, but only pectin and the total water soluble hemicelluloses ‘ figured into the equation. The lignin values were too similar in the raw vegetables and consequently were not included in the prediction equation; SHEAR=0.6980 PECTIN + 1.6758 SUMWSHEM + 2.8041. HPLC: Problems and Comments Initially, HPLC was undertaken to examine the fibrous and non- fibrous fractions for two reasons: 1) to compare chromatographic re- sults with colorimetric findings, for the purpose of determining the validity of the extraction procedure, and 2) to quantitate hexose and pentose sugars in terms of individual sugars comprising these two 92 general categories. No fiber extraction procedure is without problems arising from the nature of the fibrous and non-fibrous components themselves. Al- though there is a better understanding of the nature of such plant components as pectins and hemicelluloses, much work still remains to pinpoint the exact substances comprising them. The same holds true for the methods of determining dietary fiber. Many methods exist, each measuring different combinations of dietary fiber components. Until there exists a universal method applicable to a wide range of foods, the pro and con arguments of each method will continue. With the advent of modern instrumentation came new ideas for quantitating fiber components. It is currently the trend to compare chromatographic techniques with the older colorimetric methods. Al- though HPLC is claimed to be a type of "panacea" for quantitating car- bohydrates as well as other substances due to the lack of need for a derivitization step, problems were encountered in this study which bear directly on the overall results shown in Tables 4 and 19. First, high standard deviations are evident in Table 4 which indi— cates poor agreement between the two replicate samples used for analysis. Two of the three replicate samples were used for HPLC analysis per frac- tion, each being injected in duplicate. Although there was good agree- ment between the duplicates on the same sample, there was very poor agreement between the two replicates themselves. Perhaps different combinations of sugars were extracted in each replicate. Also, the method of triangulation was used to quantitate the sugars, which in itself is not as precise as modern computerized methods of integration. Some of the fractions investigated did not show baseline resolution, 93 especially in the protein fraction. Schweizer (1981) indicated that hydrolyzed proteins often interfere with the analyses in HPLC, however, this does not hold true for GLC. Perhaps, using a larger sample size in the dietary fiber analysis would have reduced the variability some- what, as suggested by Rasper (1981). It seems that the HPLC results in this study are of greatest value in a qualitative rather than quan- titative aspect. Again, one must bear in mind that HPLC was undertaken mainly to determine which sugars comprised the hexoses and pentoses. Only the raw vegetables were examined by HPLC; although the same results were assumed for cooked vegetables, this may not be true. In comparing chromatographic values with those obtained colorimet- rically, one must keep in mind that although interferences from hexoses were accounted for in the pentose fraction, other interferences exist. For example, it has been reported that uronic acids interfere in the pentose reaction to the extent of 9.4% of the corrected uronic acid value (Southgate, 1981). Xylose has been shown to interfere in the anthrone reaction for hexoses to the extent of 10% (Laine et al., 1981). However, no further adjustments were made for hexoses or pentoses. The hexoses to pentose ratios shown in Table 19 also exhibit marked differences. In general, the colorimetric determinations were higher than the chromatographic ones. Laine et al. (1981) found that sugars quantitated by HPLC in the starch and hemicellulose fractions were approximately 22% lower that those obtained by colOrimetry. In the present study, all of the fractions analyzed showed a decrease in values by the chromatographic method, some differences being very large. All of the standards showed excellent reproducibility (see stand- ard curves in appendix) with the exception of the galactose standard. 94 000000000> 300 00 000003 000500 000 00 00000 0 0 0.0 0.0 00.0 00.0 wo.m mw.o 00300000000 0.0 0.0 00.0 00.0 mm.o mm.0 00000 00 0.0 00.0 00.0 00.0 00.0 003800 0.0 0.0 00.0 00.0 0m.o mH.o 0000 0.0 m.0 00.0 00.0 w0.0 00.0 0000 0000000 0.0 00 00.0 mo.o mm.0 00.0 00300000000 0.0 0m 00.0 0o.o om.m 0m.0 . 00000 ~00 0000 00.0 mo.o Nw.mm 0m.om 00000000 000 i 00.0 - 00.00 00.00 0000 0 00 00.0 00.0 00.0 00.0 0000 000000 000000000000 Q00: 000000000000 0001 000002000009 Q00: 00000000 00000 00000000000x0: A0v 00000000 00 Ezm Axv 0000x0: 00 Sam 0000000pxm 0 0000005 000005000000 000 U00: 00 0000000009 .00 00000 95 0.0 0.0 00.0 00.0 00.0 00.0 00200000300 0.0 m.o mm.0 00.0 mm.0 00.0 00000 00 0.0 00.0 00.0 00.0 00.0 00000000 m.o «.0 No.m 00.0 mm.o 00.0 0000 m.m N.v 0o.0 0N.o 0w.m mm.o 0000 000000000000: 0.0 m.0 N0.N 00.0 00.m 00.N 00300000000 0.0 m.o mm.m m0.0 00.0 ww.0 00000 00 0.0 No.0 0m.0 00.00 00.m 00000000 m.0 «.0 mm.o mo.0 00.0 00.0 0000 o.m 0.0 00.0 00.0 No.0 00.0 0000 000000 000000000000 0000 000000000000 0000 000000000000 0000 00000000 00000 00000000000x0: 000 00000000 00 000 A00 0000x0I 00 000 00000000xm 0.0.0000V 00 00000 96 Although several sources of galactose were prepared and injected into the HPLC, each one had its OWn characteristic peaks, indicating either impure sample or poor column performance. Finally, one galactose sample was chosen for this study based on the least number of peaks eluted which was three. The three peaks were summed up and plotted as one value. According to the results shown in Table 4, no galactose was pre— sent in the pectin fraction. This can hardly be correct, as it is well documented that galactans are an integral part of the pectin fraction (Hoff and Castro, 1969; Keegstra et al., 1973; Konno et al., 1980). Perhaps galactose is more susceptible to hydrolysis and decomposition that the other sugars, although this cannot be supported by the lit- erature. Another problem stemmed from sample preparation. Most workers hydrolyze their samples before measuring individual sugars. This was attempted unsuccessfully. Not only was there considerably less agree— ment between replications, but quantitation could not even be attempted because of total lack of baseline resolution. Theander and Aman (1981) reported problems with hydrolyses related to the formation of degrad- ation products, particularly with pectin—rich samples. The furans, color-producing carbonyl compounds, and many phenols can interfere with both colorimetric and chromatographic measurements. Hexuronic acids are most susceptible to these types of reactions followed by pentoses and hexoses. Uronic acids condense with sugars during acid treatment and will not ordinarily be detected by chromatographic methods. A sample Chromatogram is shown in the appendix. 97 000000 00000000 000 00000 0000 00x00 000000000 000000000 000000000 00 000 00 000 0000000 0003 000000000 0000000000000 000 .000000 .0000000 .000000 00 Sam ”00:00> 0000000000xm "00000> 0000000000N 0000000000 000 I 02 0000000 000000 0 m.0 m.o 0.0 m.0 0.0 00 0000000 0.0 0.0 0.0 0.0 N.0 oz .00xm 00300000000 00000 N.0 m.o N.0 0.0 0.0 oz .00xm 00000 0.0 00 0.0 0.0 m.o 0.0 0000 000 00000000 N.m 0.0 m.00 m.N 0.0 s 000002 0 00N00300m 0.0 m.o m.M0 0.0 0.0 0.0 .00 00 000000>000 00.0 - 0.0 0.0 0.0 0.0 002000 0.0 0.0 0.m m.0 0.0 oz .00xm 00000000 m.0 m.o m.0 0.0 0.0 02 .00xm 0000 0.0 00 0.0 0.0 0.0 - 0003000 ~.0 0.0 0.0 N.0 0.0 02 0.00xm 0000 000000 0000 0000 00x 0000 00000 000000 0000 000002 000500 00 00000 0 000000 000000m0> 300 0003 000000000 00000 0000000000x0 0:00 000 00 00000 000000 00:00> 0000000000 0000000 00000 00 0000000000 < .om 00000 98 Comparisons with Literature Values Table 20 shows the percentage of neutral sugars from HPLC analysis based on the total. Glucose values do not include the glucose from the starch fraction. One can clearly see the large differences between literature values and the experimental values caused by the problems encountered in the HPLC analyses. The literature values were obtained by GLC anal— ysis, using alditol acetate derivatives. Not only do the absolute values show great differences, but the trends also show large varia- tions. Clearly, the HPLC analyses must first be perfected before any cogent conclusions can be drawn. ‘ Sensory Aspectsof Vegetable Texture Although sensory analyses were not undertaken due to the nature of the commercial buffers, it is of interest to speculate upon the outcome of such tests based upon others' findings. The question is whether or not a sensory panel would be able to detect the small textural differences, e.g., between vegetables cooked at pH 2 to pH 4 or between pH 4 and pH 6. From the findings of Odland and Eheart (1975) and Eheart and 0d- land (1973), one may surmise that panelists judging the five vegetables in this study would have found the vegetables cooked at pH 2 to be too firm and of an objectionable olive-green color due to pheophytin in the beans and peas. At pH 4, the vegetables probably would have been rated better than those at pH 2 since they would contain less pheophytin, although they would have been rated firmest. They would still be infer- ior to vegetables cooked at pH 6, probably the optimum. Wager (1963) 99 found potato strips to possess the most favorable texture at pH 6-7 as detected by a sensory panel. The vegetables cooked at pH 10 would have been rated most inferior in texture, because of the large degree of softening. Sloughing would have been evident in most of the vegetables, also contributing to the inferior quality. It is very probable that a sensory panel would have detected subtle differences in texture for the vegetables at various pH levels based on the studies of Jeon et al. (1975) who found excellent instrumental and sensory correlation for pickle texture. Odland and Eheart (1975) had excellent sensory panel results; their panelists detected textural differences among broccoli samples at different pH levels, although the levels were quite close to each other. The levels between the pH values in the current study are wide enough that distinct differences would have been detected in texture. SUMMARY AND CONCLUSIONS Green beans, potatoes, peas, cauliflower, and corn were cooked in buffers of pH 2, 4, 6, and 10 to determine the effect of cooking in solutions of varying pH on dietary fiber components. Tenderness was determined on the raw and cooked vegetables. The individual dietary fiber components were measured in the raw and cooked vegetables and in the cooking media. These included water soluble pectin and hemicel- lulose, water insoluble pectin and hemicellulose, cellulose, and lignin. The cooked vegetables were firmest at pH 4 and softest at pH 10. Although increased firmness at pH 4 is usually attributed to an increase in insoluble pectin in the vegetable itself and a corresponding decrease in soluble pectin, both fractions showed a decrease at pH 4. The soft- ening at pH 10 was attributed mainly to the decreased insoluble hemi- cellulose and pectin corresponding to increased soluble hemicellulose and pectin. No problems were encountered with the extraction procedures, how- ever, the values for starch were low, possibly because some of the starch was extracted in the pectin fraction. Perhaps a wet ball milling step may have increased the amount of starch extracted. Overall, cauliflower and beans showed the most dramatic effects on fiber components. Since corn displayed little change in shear force at different pH values, no significant effects on fiber components were evident for this vegetable. In general, all of the vegetables except 100 101 corn showed significant changes in the amounts of the fiber components contained in raw and cooked vegetables. In other fractions, cooking solution fiber increased with pH be- cause of the increased solubility of the pectic and hemicellular com- ponents at basic pH. Cellulose showed no significant effects with pH as expected because of the strongb/acidic conditions necessary for cellulose degradation. Lignin showed no significant effects except for beans and cauliflower. Beans had increased lignin values from pH 4 to pH 6 ascribed to tannin-protein condensation products. A prediction equation relating shear force and fiber components, the independent variables, showed that fiber components may be used to predict tenderness in vegetables cooked at different pH values. Insol- uble pectin, lignin, and soluble hemicelluloses appeared to contribute the most to the texture of the vegetables. Neutral sugars were quantitated by HPLC in the protein, starch, insoluble pectin, and insoluble hemicellulose fractions of the raw vegetables in order to determine the major hexoses and pentoses in these fractions, which are not differentiated in the fiber extraction procedure. Also, this served as a check on the extraction scheme. However, the results were not comparable to those obtained by others because of problems encountered in sample preparations and with the galactose standard. Glucose was the major hexose detected in the four fractions and xylose comprised most of the pentoses. RECOMMENDATIONS The problems associated with HPLC should be examined in the fol- lowing instances: a) Other sugars should be investigated such as rhamnose and fucose. b) Hydrolysis procedures should be re-tested and the best method v V v applied to fiber extraction procedures. The problems associated with galactose quantitation should be further investigated. Perhaps a guard column or a purification procedure could be applied to the various fractions to eliminate background inter- ference Vegetables by the same extraction method should be looked at by HPLC and GLC concurrently. The pH of the vegetables could have been measured before and after cooking in the buffer systems. The cooking solution pH could have been measured to see if the naturally occurring organic acids had any effects on these sol- utions. Sensory panelists could have examined the vegetables for color, flavor, and texture in order to correlate objective shear-press data with sensory analyses. 102 103 Histological examinations of the vegetables could have been per- formed to examine changes in the vegetable tissues, The method for quantitating fiber components has not yet been per- fected by anyone; this area can always use improvement, e.g., dif- ferent enzymes could be used, wet ball milling and/or autoclaving could be examined in order to remove starch more thoroughly, and a better method of distinguishing pectin from hemicellulose is desperately needed. LIST OF REFERENCES LIST OF REFERENCES Albaum, H. G. and Umbreit, N. w. 1947. Differentiation between ribose- 3- -phosphate and ribose- 5- -phosphate by means of the orcinol- -pentose reaction. J. Biol. Chem. 167. 369. Albersheim, P. 1959. Instability of pectin in neutral solutions. Biochem. and Biophys. Res. Comm. 5: 253. Albersheim, P. 1965. The substructure and function of the cell wall. In "Plant Biochemistry." Eds. J. Bonner and J. Varner, Academic Press, New York. Albersheim, P., Neukom, H., and Deuel, H. 1960. Splitting of pectin chain molecules in neutral solutions. Arch. Biochem. Biophys. 90: 46. American Association of Cereal Chemists. 1962. AACC approved methods, 7th ed. The Association. St. Paul, MN Aspinall, G.O. 1963. Hemicelluloses, gums, and pectic substances. Rec. Adv. Food Sci. 3: 282. Aspinall, G.O. 1970. ”Polysaccharides." Pergamon, Oxford. Barmore, M.A. 1937. Potato mealiness and changes in softness in cook- ing. Food Res. 2: 377. Bauer, H.D., Talmadge, K.w., Keegstra, K., and Albersheim, P. 1973. The structure of plant cell walls. 11. The hemicellulose of the walls of suspension—cultured sycamore cells. Plant Physiol. 51: 174. Bettelheim, F.A. and Sterling, C. 1955a. Factors associated with potato texture. I. Specific gravity and starch content. Food Res. 20: 71. Bettelheim, F.A. and Sterling, C. 19556. Factors associated with potato texture. II. Pectic substances. Food Res. 20: 118. Charley, H. 1970. "Food Science." John Wiley and Sons, New York. Charley, H. 1972. Fruits and vegetables. In "Food Theory and Appli— cations," ch. 6, Eds. P.C. Paul and H.H. Palmer, John Wiley and Sons, Inc., New York. 104 105 Cole, E.W. 1967. Isolation and chromatographic fractionation of hemi- celluloses from wheat flour. Cereal Chem. 44: 411. Cole, E.W. 1969. Some physicochemical properties of a wheat flour hemicellulose in solution. Cereal Chem. 46: 382. Collings, G.F. and Yokoyama, M.T. 1979. Analysis of fiber components in feeds and forages using GLC. J. Agric. Food Chem. 27: 373. Day, E.D. 1909. The effect of cooking on cellulose. J. Home Econ. l: 177. Dische, Z. 1947. A new specific color reaction for hexuronic acids. J. Biol. Chem. 167: 189. Doesburg, J.J. 1961. Relation between the behavior of pectic sub- stances and changes in firmness of horticultural products during heating. Qualitas Plant. et. Material Vegetal. 8: 115. Eheart, M.S. and Odland, D. 1973. Quality of frozen green vegetables blanched in four concentrations of ammonium bicarbonate. J. Food Sci. 38: 954. Englyst, H. 1981. Determination of carbohydrate and its composition. In "The Analysis of Dietary Fiber in Food," ch. 6, Eds. W.P.T. James and O. Theander, Marcel Dekker, Inc., New York. Frazier, W.C. and Westhoff, D.C. 1978. "Food Microbiology," 3rd ed. McGraw-Hill Book Co., Inc., New York. Freeman, M.E. and Ritchie, W.S. 1940. Pectins and the texture of cooked potatoes. Food Res. 5: 167. Freudenberg, K. and Neish, A.C. (Eds.) 1968. "Constitution and Biosynthesis of Lignin.” Springer-Verlag, New York. Goering, H.K. and Van Soest, P.J. 1970. Forage fiber analysis. Agric. Handbook No. 379, ARS/USDA, Washington, D C Hassid, W.Z. and Newfeld, F.F. 1964. Quantitative determination of starch in plant tissues. Methods Carbohydrate Chem. 4: 33. Heller, S.N., Rivers, J.M., and Hackler, L.R. 1977. Dietary fiber: The effect of particle size and pH on its measurement. J. Food Sci. 42: 436. Herranz, J., Vidal-Valverde, C., and Rojas-Hidalgo, E. 1981. Cellul- ose, hemicellulose, and lignin content of raw and cooked Spanish vegetables. J. Food Sci. 46: 1927. Hoff, J.E. 1972. Starch "swelling pressure" of cooked potatoes. J. Agric. Food Chem. 20: 1283. 106 Hoff, J.E. and Castro, M.D. 1969. Chemical composition of potato cell wall. J. Agric. Food Chem. 17: 1328. Hughes, J.C., Faulks, R.M., and Grant, A. 1975a. Texture of cooked potatoes: Relationship between the compressive strength of cooked potato disks and release of pectic substances. J. Sci. Food Agric. 26: 731. Hughes, J.C., Faulks, R.M., and Grant, A. 1975b. Texture of cooked potatoes: Relation between compressive strength, pectic substan- ces and cell size of Redskin tubers of different maturity. Potato Res. 18: 495. Hughes, J.C., Grant, A., and Faulks, R.M. 1975c. Texture of cooked potatoes: The effect of ions and pH on the compressive strength of cooked potatoes. J. Sci. Food Agric. 26: 739. Hutchison, R. 1908. "Food and the Principles of Dietetics," 2nd ed. William Wood and Co., New York. Isherwood, F.A. 1955. Texture in fruits and vegetables. Food Manu- facture. 30: 399 James, W.P.T. and Theander, O. (Eds.) 1981. "The Analysis of Dietary Fiber in Food." Marcel Dekker, Inc., New York. Jeltema, M.A. and Zabik, M.E. 1980. Revised method for quantitating dietary fibre components. J. Sci. Food Agric. 31: 820. Jeon, I.J., Breene, W.M., and Munson, S. T. 1975. Texture of fresh- -pack whole cucumber pickles. Correlation of instrumental and sensory measurements. J. Texture Studies. 5: 399. Jones, D. B. 1931. Factors for converting percentages of nitrogen in foods and feeds into percentage of proteins. U.S. Dept. Agric. Circ.183:1—21. Jones, J.K.N. 1951. The chemical composition and properties of pectins. Chem. and Ind. 1951: 430. Jones, R.W., Krull, L.H., Blessin, C.W., and Inglett, G.E. 1979. Neutral sugars of hemicellulose fractions of pith from stalks of selected plants. Cereal Chem. 56: 441. Joslyn, M.A. 1950. “Methods in Food Analysis." Academic Press, Inc., New York. Keegstra, K., Talmadge, K.W., Bauer, W.D., and Albersheim, P. 1973. The structure of plant cell walls. 111. A model of the walls of suspension-cultured sycamore cells based on the interconnections of the macromolecular components. Plant Physiol. 51: 188. 107 Kertesz, Z.I. 1951. II-The Pectic Substances." Wiley Interscience, New York. Konno, H., Yamasaki, Y., and Ozawa, J. 1980. Fractionation of poly- saccharides and acidic oligosaccharides forned on partial acid hydrolysis of pectin. Agric. Biol. Chem. 44: 2195. Laine, R.A., Varo, P., and Koivistoinen, P.E. 1981. Observations on the analysis of dietary fiber. In "The Analysis of Dietary Fiber in Food,” ch. 2. Eds. W.P.T. James and O. Theander, Marcel Dekker, Inc., New York. Linehan, D.J. and Hughes, J.C. 1969a. Texture of cooked potato. II. Relationships between intercellular adhesion and chemical com- position of the tuber. J. Sci. Food Agric. 20: 113. Linehan, D.J. and Hughes, J.C. 1969b. Texture of cooked potato. III. Intercellular adhesion of chemically treated tuber sections. J. Sci. Food Agric. 20: 119. Lopez, A. 1981. "A Complete Course in Canning," 11th ed. The Can- ning Trade, Baltimore, MD. Mattheé, V. and Appledorf, H. 1978. Effect of cooking on vegetable fiber. J. Food Sci. 43: 1344. Mattson, S. 1946. The cookability of yellow peas. Act Agric. Suecana. 2: 185. Mattson, S., Akerberg, E., Eriksson, E., Koutler—Andersson, E., and Vahtras, K. 1951. Factors determining the composition and cook- ability in peas. Acta Agric. Scand. O. McConnefl, A.A., Eastwood, M.A., and Mitchell, W.D. 1974. Physical characteristics of vegetable foodstuffs that could influence bowel function. J. Sci. Food Agric. 25: 1457. McCready, R.M. and McComb, E.A. 1952. Extraction and determination of total pectic substances in fruits. Anal. Chem. 24: 1986. Mod, R.R., Ory, R.L., Morris, N.M., and Normand, F.L. 1981. Chemical properties and interactions of rice hemicellulose with trace min- erals. J. Agric. Food Chem. 29: 449. Monte, W.C. and Maga, J.A. 1980. Extraction and isolation of soluble and insoluble fiber fractions from the pinto bean (Phaseolus vulgaris). J. Agric. Food Chem. 28: 1169. Nelson, R. 1960. Fractionation of hemicelluloses. TAPPI 43: 313. Neukom, H. and Deuel, H. 1958. Alkaline degradation of pectin. Chem. and Ind. 1958: 683. 108 Nonaka, M. 1980. The textural quality of cooked potatoes. 1. The relationship of cooking time to the separation and rupture of potato cells. Am. Potato J. 57: 141. Northcote, D.H. 1972, Chemistry of the plant cell wall. Annu. Rev. Plant Physiol. 23: 13 Odland, D. and Eheart, M.S. 1975. Ascorbic acid, mineral and quality retention in frozen broccoli blanched in water, steam and ammonia- steam. J. Food Sci. 40: 1004. Osborne, D.R. and Voogt, P. 1978. "The Analysis of Nutrients in Foods," Academic Press, New York. Personius, C.J. and Sharp, P.F. 1939. Adhesion of potato tissue cells as influenced by pectic solvents and precipitants. Food Res. 4: 299. Pigman, W. and Horton, D. (Eds.) 1970. ”The Carbohydrates,” Vol. 2A, Academic Press, New York. Rasper, V.F. 1981. Fractionation of the insoluble residue in dietary fiber analysis. In ”The Analysis of Dietary Fiber in Food," ch. 3, Eds. W.P.T. James and O. Theander, Marcel Dekker, Inc., New York. Reeve, R.M. 1954. Histological survey of conditions influencing tex— ture in potatoes. 1. Effects of heat treatment on structure. Food Res. 19: 323. Reeve, R.M. 1967. A review of the cellular structure, starch, and texture qualities of processed potatoes. Econ. Botany 21: 294. Reeve, R.M. 1970. Relationships of histological structure to texture of fresh and processed fruits and vegetables. J. Texture Studies 1: 247. Schweizer, T.F. 1981. General Discussion. In "The Analysis of Dietary Fiber in Food," ch. 15, Eds. W.P.T. James and O. Theander, Mar- cel Dekker, Inc., New York. Seckinger, H.L., Wolf, M.J., and MacMasters, M.M. 1960. Hemicelluloses of the cementing layer and of some cell walls of the corn kernel. Cereal Chem. 37: 121. Selvendran, R.R., Ring, S.G., and DuPont, M.S. 1979. Assessment of procedures used for analysing dietary fibre and some recent devel— opments. Chem. and Ind. 1979: 225. Selvendran, R.R., Ring, S.G., and DuPont, M.S. 1981. Determination of the dietary fiber content of the EEC samples and a discussion of the various methods of analysis. In "The Analysis of Dietary Fiber in Food,” ch. 7, Eds. W.P.T. James and O. Theander, Marcel Dekker, Inc., New York. 109 Sharma, M.K., Isleib, D.R., and Dexter, S.T. 1959. The influence of specific gravity and chemical composition on hardness of potato tubers after cooking. Am. Potato J. 36: 105. Sherman, H.C. 1948. "Food Products," 4th ed. Macmillan Co., New York. Simpson, J. I. and Halliday, E. G. 1941. Chemical and histological studies of the disintegration of cell-membrane materials in vege- tables during cooking. Food Res. 6: 189. Southgate, D.A.T. 1976. "Determination of Food Carbohydrates." Science Publishers, LTD, London. Southgate, D.A.T. 1981. Use of the Southgate method for unavailable carbohydrates in the measurement of dietary fiber. In "The Anal- ysis of Dietary Fiber in Food," ch. 1, Eds. W.P.T. James and 0. Theander, Marcel Dekker, Inc., New York. Sterling,C . 1955. Effect of moisture and high temperature on cell walls in plant tissue. Food Res. 20: 474. Sterling, C. 1963. Texture and cell —wall polysaccharides in foods. Rec. Adv. Food Sci. 3' 9. Sterling, C.J. 1966. Anatomy and histology of the tuber with respect to processed quality. Proc. Plant Sci. Symp. Campbell Inst. Agric. Res., Camden, NJ. pp. 11-25. Sterling, C. J. 1968. Effect of solutes and pH on the structure and firmness of cooked carrot. J. Food Technol. 3: 367. Sterling, C. and Bettelheim, F.A. 1955. Factors associated with potato texture. III. Physical attributes and general conclusions. Food Res. 20: 130. Tauss, H. 1889. Verhalten von Holz und Cellulose gegen erhbhte Temp- eratur underhohten Druck bei Gegenwart von Nasser Dinglers Poly- tech. J. 273: 276. Theander, 0. and Aman, P. 1981. Analysis of dietary fibers and their main constituents. In "The Analysis of Dietary Fiber in Food," ch. 5, Eds. W.P.T. James and 0. Theander, Marcel Dekker, Inc., New York. Trowell, H., Southgate, D.A.T., Nolever, T.M.S., Leeds, A.R., Gassull, M.A., and Jenkins, D.J.A. 1976. Dietary fiber redefined. Lancet. 1: 967. Van Buren, J. 1970. Current concepts on the texture of fruits and vegetables. Crit. Rev. Food Technol. ' 5. 110 Van Soest, P.J. and Robertson, J.B. 1976. Analytical problems of fiber. Presented at the 36th Annual Meeting of the Institute of Food Technologists, June 6-9, 1976 at Anaheim, CA. Van Soest, P.J. and Robertson, J.B. 1977. What is fibre and fibre in food? Nutr. Revs. 35: 12. Wager, H.G. 1963. The role of phytin in the texture of cooked pota- toes. J. Sci. Food Agric. 14: 583. Warren, D.S. and Woodman, J.S. 1973. Distribution of cell wall com- ponents in potato tubers: A new titrimetric procedure for the estimation of total polyuronide (pectic substances) and its de- gree of esterification. J. Sci. Food Agric. 24: 769. Whistler, R.L. and Smart, C.L. 1953. "Polysaccharide Chemistry." Academic Press, New York. Whittenberger, R.T. 1951. Changes in specific gravity, starch content and sloughing of potatoes during storage. Am. Potato J. 28: 738. Whittenberger, R.T. and Nutting, 0.0. 1950. Observations on sloughing of potatoes. Food Res. 15: 331. Worth, H.G.J. 1967. The chemistry and biochemistry of pectic substan- ces. Chem. Rev. 67: 465. APPENDICES 111 APPENDIX I. Standard curves for hexoses, pentoses, and uronic acids at various wavelengths Absorbance O 0'! l 0.4 _ 0.3 ~ 0.2 _ x X 0.1 _ i i i I l I I 0 20 40 60 80 100 ,ug/ml Figure Ia. Standard curve for hexoses at 620 nm using the anthrone reaction. 112 Absorbance I 1 l I T o 20 4o 60 80 100 Fg/ml Figure Ib. Standard curve for hexoses at 600 nm using the orcinol reaction. 113 Absorbance 0 U1 l 0.4- x 0.3- 0.2 - 0.1— l I I I T 0 20 40 6O 80 100 pQ/ml Figure Ic. Standard curve for hexoses at 670 nm using the orcinol reaction. 114 Pentose 670 Absorbance Pentose 600 0 3.33 6.67 10.00 Pg/ml Figure Id. Standard curve for pentoses at 600 and 670 nm using the orcinol reaction. Absorbance 115 XX XX XX Figure Ie. I T l 20 4O 60 pg/ml Standard curve for uronic acids at 525 nm using the carbazole reaction. 80 1 100 116 APPENDIX II. Average results of fiber extraction procedure IIa. Average results for the starch extraction Vegetable % hexose % pentose % uronic Peas Raw 7.66 1.80 0.66 pH 2 14.23 1 79 0.88 pH 4 11.92 1.66 1.35 pH 6 12.82 1.87 0.89 pH 10 13.47 1 64 0.96 Corn Raw 15.90 0 15 1.33 pH 2 18.43 0 30 1.29 pH 4 18.22 0 23 1.50 pH 6 20.74 0 44 1.42 pH 10 17.83 0 30 1.47 Potatoes Raw 39.82 0 28 4.00 pH 2 58.98 0 12 6.56 pH 4 54.20 0 26 5.02 pH 6 53.26 0.11 5.61 pH 10 51.67 0.22 5.24 Beans Raw 2.36 1.20 1.12 pH 2 2.97 0.98 0.56 pH 4 2.73 0.82 0.44 pH 6 2.91 0.82 0.46 pH 10 2.78 0.72 0.49 Cauliflower Raw 1 35 2.01 0.41 pH 2 1.25 1.75 2 24 pH 4 1.34 2.16 0.48 pH 6 1.87 2.41 0.61 pH 10 l 54 1.96 0.56 117 IIb. Average results for protein extraction Vegetable % hexose % pentose Peas Raw 1.18 1.59 pH 2 1.55 1.86 pH 4 1.42 1.72 pH 6 1.63 1.93 pH 10 1.32 1.49 Corn Raw 0.51 0.26 pH 2 0.50 0.37 pH 4 0.76 0.42 pH 6 0.43 0.39 PH 10 0.58 0,44 Potatoes Raw 4.81 0.38 PH 2 2.83 0.44 pH 4 1.82 0.35 pH 6 2.47 0.37 pH 10 2.81 0.28 Beans Raw 0.95 0.80 pH 2 1.18 0.94 pH 4 1.38 1.22 pH 6 1.22 1.17 pH 10 1.70 1.52 Cauliflower Raw 2.08 2.99 pH 2 2.17 2.98 pH 4 2.43 2.95 pH 6 2.20 2.73 pH 10 1.49 1.58 118 11c. Average results for the water solubles before dialysis Vegetable % hexose % pentose % uronic Peas Raw 13.78 0.97 2.41 pH 2 13.14 1.05 2.97 pH 4 13.67 0.92 2.39 pH 6 13 37 0 98 2.75 pH 10 11.92 1 53 3.39 Corn aw 45.08 0.26 6.45 pH 2 42.34 0.64 5.88 pH 4 45.88 0.30 6.94 pH 6 40.07 0.26 5.41 pH 10 42.94 0.53 5.43 Potatoes Raw 2.63 0.19 0.50 pH 2 2.24 0.39 0.81 pH 4 2.50 0.34 0.86 pH 6 3.11 0.43 1 10 pH 10 2.85 0.50 1 10 Beans Raw 26.44 0 85 4.06 pH 2 28.07 0 66 4.77 PH 4 28.89 0 65 5.34 pH 6 28.09 0 88 5.48 pH 10 29.79 1 54 7.24 Cauliflower Raw 29.21 1.07 4.40 pH 2 20.90 1.63 4.73 pH 4 25.21 1.39 4.63 pH 6 22.96 2.07 6.14 pH 10 24.21 4.52 7.79 119 IId. Average results for the cooking solution after dialysis Vegetable % hexose % pentose % uronic Peas Raw - - - pH 2 0.05 0.08 0.13 pH 4 0.09 0.07 0.11 pH 6 0.05 0.09 0.14 pH 10 0.25 0.43 0.60 Corn Raw - - - pH 2 0.04 0.02 0.01 pH 4 0.10 0.02 0.02 pH 5 0.03 0.02 0.01 pH 10 0.03 0.03 0.02 Potatoes Raw - - - pH 2 0.03 0.01 0.02 pH 4 0.04 0.01 0.02 pH 5 0.03 0.03 0.05 pH 10 0.25 0.09 0.16 Beans Raw - - — pH 2 0.05 0.04 0.06 pH 4 0.05 0.03 0.06 pH 6 0.09 0.04 0.06 pH 10 0.08 0.09 0.21 Cauliflower Raw - - - pH 2 0.12 0.13 0.16 pH 4 0.07 0.09 0.16 pH 6 0.12 0.18 0.31 pH 10 0.46 0.84 0.74 120 IIe. Average results for the insoluble pectin extraction Vegetable % hexose % pentose % uronic Peas Raw 1 92 0.96 0 79 pH 2 2.67 1.57 0 84 pH 4 2.21 1.00 0 52 pH 6 2 93 1.01 0 77 pH 10 2 76 0.84 0 50 Corn Raw 1 16 0.88 0 04 pH 2 1 14 1.11 0 16 pH 4 l 09 1.43 O 00 pH 6 0 94 1.21 0 00 pH 10 1 12 1.05 0 00 Potatoes Raw 10.08 0.62 1.24 pH 2 3.39 0.53 0.08 pH 4 4.35 0.43 0.14 pH 6 3.43 0.62 0.01 pH 10 4.85 0.43 0.27 Beans Raw 4 11 2.59 4.02 pH 2 4 69 3.60 4.87 pH 4 5 27 3.61 4.20 pH 6 4 90 3.90 5 10 pH 10 3 93 2.28 2 57 Cauliflower Raw 5 16 2.92 2 86 pH 2 4 65 2.10 1.69 pH 4 4.21 2.32 1 79 pH 6 5.21 2.04 1.51 pH 10 3 63 1.07 0.37 lAll negative numbers were assumed to be zero by the computer 121 IIf. Average results for the insoluble hemicellulose extraction Vegetable % hexose % pentose % uronic Peas Raw 3.84 1.01 0 60 pH 2 3.02 0.53 0.58 pH 4 2.89 0.79 0.71 pH 6 2.97 0.63 0.70 pH 10 3.05 0.54 0.60 Corn Raw 0 68 2.02 0.35 pH 2 0 82 1.91 0.35 pH 4 0 59 1.59 0.24 pH 6 0 90 1.82 0.35 pH 10 0 94 1.65 0 1 Potatoes Raw 5.40 0.14 1 34 pH 2 2.09 0.03 0 32 pH 4 4.11 0.03 0.52 pH 6 4 24 0.03 0.57 pH 10 3 29 0.04 0 32 Beans Raw 0.39 0.33 0 24 pH 2 0.64 0.37 0 23 pH 4 0.42 0.35 0 22 pH 6 0.46 0.30 0 17 pH 10 0.40 0.34 0 16 Cauliflower Raw 0.14 0.10 0.06 pH 2 0.33 0.21 o 11 pH 4 o 21 0.16 0 06 pH 6 o 15 0.13 0.08 pH 10 o 28 0.20 0.09 122 I19. Average results for the cellulose extraction Vegetable % hexose % pentose % uronic Peas Raw 4,92 0.56 1.67 pH 2 5.00 0.40 1.62 pH 4 6.13 0.39 1.56 pH 6 5.48 0.42 1.49 pH 10 3.46 0.55 1.29 Corn Raw 1.38 0.35 0.33 pH 2 1.63 0.31 0.35 pH 4 1.42 0.31 0.31 pH 6 1.40 0.30 0.24 pH 10 1.20 0.37 0.27 Potatoes Raw 3.92 0.21 0.58 pH 2 2.33 0.14 0.34 pH 4 2.93 0.17 0.33 pH 6 2.22 0.17 0.29 pH 10 2.73 0.19 0.40 Beans Raw 3.87 0.61 1.48 pH 2 3.93 0.61 1.11 pH 4 4.16 0.29 0.90 pH 6 3.73 0.39 0.84 pH 10 5.88 0.41 0.96 Cauliflower Raw 2.27 0.32 0.46 pH 2 2.64 0.65 0.47 pH 4 2.76 0.44 0.62 pH 6 2.86 0.32 0.61 pH 10 3.30 0.48 0.57 123 APPENDIX III. HPLC standard curves of five sugars and HPLC chromatogram 16 « 14 _ 12_ 10 - 2) Area (cm Concentration (mg/ml) Figure IIIa. HPLC standard curve of glucose (4X attenuation). 124 2) Area (cm 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Concentration (mg/ml) Figure IIIb. HPLC standard curve of glucose (2X). Area (cmz) (D Figure 125 T _. O I I I I I I | I | I I I I 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Concentration (mg/ml) IIIc. HPLC standard curve of xylose (2X). 126 Area (cmz) Concentration (mg/ml) Figure IIId. HPLC standard curve of arabinose (2X). Area (cmz) Figure IIIe. 12 127 I I I | I l I I I I I I I l 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Concentration (mg/ml) HPLC standard curve of mannose (2X). 128 2) Area (cm I I | I I I I I I I O 1 2 3 4 5 Concentration (mg/ml) Figure IIIf. HPLC standard curve of galactose (2X). 129 .AgmzomeFzmuv cowuumxw :wuuma we Emgmopmaoggo 04¢: . mmuzcwz mm mm em mm om mH ma efi NH oH w .v N o e S e 0 S n 0 n n e a .1 S M .D 0 a 1|: m. m. :WE\:?m.o xm Loam: uo~o.o + omw :wmmg mmcmgoxm 8.58 mcmgbmboa :1 m vmchPmmogu xw .Ego4 Acmmgv Papas >>mmc .aumnxmz xmcvs< .11111L "mwmxpmc< mpmguxsoagmo umzuowm Glucose 55:: To .3 ON .3: 95ml "ummam “Lego "Lovempmo "omega m_waoz ”.aEmH :Ezpou ”:E:_ou ”mung 30pm um_nsmm MICHIGAN STRTE UNIV. LIBRARIES IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 31293106683661