Pineapple and Banana Pectins Comprise Fewer Homogalacturonan

Mar 4, 2009 - ... in CWMs from other dicots such as apple (Malus malus), pawpaw (Carica papaya), and ambarella (Spondias cytherea) on a dry-matter bas...
0 downloads 0 Views 100KB Size
Download PDF
Biomacromolecules 2009, 10, 717–721

717

Pineapple and Banana Pectins Comprise Fewer Homogalacturonan Building Blocks with a Smaller Degree of Polymerization as Compared with Yellow Passion Fruit and Lemon Pectins: Implication for Gelling Properties Beda M. Yapo* Unite´ de Formation et de Recherche en Sciences et Technologie des Aliments, Universite´ d’Abobo-Adjame´, 02 BP 801 Abidjan 02, Coˆte d’Ivoire Received December 23, 2008; Revised Manuscript Received February 13, 2009

Pectins are viewed as multiblock cobiopolymers of different pectic polysaccharides, notably, homogalacturonan (HG) and rhamnogalacturonan I (RG I). Furthermore, on the basis of HGs isolated from different (pectins from) dicot cell walls, HG is supposed to have an average degree of polymerization (DP) of ∼100 irrespective of the plant source. To validate or invalidate these suppositions, pectins from both monocot (pineapple and banana) and dicot (yellow passion fruit and lemon) cell walls were examined. The results show that all the extracted pectins comprise HGs as well as type I and II arabinogalactan side chain-containing RGs I, but of significantly (p < 0.05) different relative proportions; lemon pectin being the richest in HGs, followed by yellow passion fruit pectin. The HG building blocks of each pectin are homogeneous with respect to the molecular size but have a significantly (p < 0.05) reduced length in monocot pectins (59-67) compared to dicot ones (93-102). Lemon pectin displayed the highest degree of esterification (DE), viscosity-average molecular weight (Mv), and gelling ability, whereas with similar DEs and a higher Mv, banana pectin exhibited a lower gelling ability than yellow passion fruit pectin. It is concluded that both the HG amount and DP strongly influence the gelling properties of pectin.

1. Introduction Pectins are complex nonstarchy polysaccharides present in the middle lamellae and the primary cell walls of higher plants. Even though the singular term of “pectin” is most often used in literature, it is worth noting that there is not a unique pectin but several pectin types given the high heterogeneity of their structural composition. The fine structure of pectins, although not completely unravelled, could constitute at least 17 kinds of monosaccharides of which galacturonic acid is typically the major one, followed by either galactose or arabinose. Several studies showed that these sugars were not randomly distributed in the pectin structures but rather concentrated in different structural elements (or domains) also known as pectic polysaccharides. At least seven sorts have been described in literature, namely, homogalacturonan (HG), type I rhamnogalacturonan (RG I), type II rhamnogalacturonan (RG II), xylogalacturonan (XGA), apiogalacturonan, galactogalacturonan, and galacturonogalacturonan.1-3 These different pectic polysaccharides did not stem from a sole plant origin but from various plant cell walls. However, the bulk of cell walls hitherto examined consisted of at least HG and RG I, including sometimes RG II as the most frequent substituted galacturonan. The structural elements other than HG and RG I are frequently gathered under the generic term of substituted galacturonans, which means that all of them are branched galacturonans, as opposed to HG, which is a nonbranched (linear) polymer of 1,4-linked R-D-GalpA units. RG I has a backbone made of repeating [f4)-R-D-GalpA-(1f2)-R-L-Rhap-1f] disaccha* To whom correspondence should be addressed. Tel.: +225-015-51150. E-mail: [email protected].

ride units, bearing generally neutral side chains of type I and type II arabinogalactans. RG II is a highly complex macromolecule with a backbone composed of at least seven 1,4linked R-D-GalpA units, which is branched with four structurally conserved side chains, consisting of at least 12 different monosaccharides, including six specific sugars.4,5 These are apiose, 2-O-methyl-fucose, 2-O-methyl-xylose, 3-C-carboxy-5-deoxy-L-xylose (aceric acid), 2-keto-3-deoxyD-manno-octulosonic acid (Kdo), and 3-deoxy-D-lyxo-heptulosaric acid (Dha). Even if it may still be a matter of debate, it seems that a pectin macromolecule is the result of the assembling, through covalent bonds, of different structural elements (or building blocks), with HG and RG I being the two common basic components. HG is currently supposed to have an average length of ∼100 contiguous GalA units, whereas the length of the backbone of RG I remains somewhat elusive, ranging from ∼30 GalA-Rha repeats in citrus pectins,5,6 to ∼200 GalA-Rha repeats in sycamore cell ones.7 Furthermore, the backbones of the RG I building blocks of a pectin macromolecule may have substantially different lengths,5 inasmuch as two backbones of ∼15 and 35 GalA-Rha repeats and three backbones of ∼15, 28, and 100 GalA-Rha repeats have been reported in lemon and soybean pectins, repectively.8 In contrast, HG isolated, by mild acid hydrolysis, from the cell walls of Rosa glauca was found to have a degree of polymerization (DP) at least >100.9 Some time after, similar observations were made in carrot cell wall pectin using enzymatic degradation10 and in commercial (acid-extracted) apple, beet, and citrus pectins using mildacid hydrolysis.11 Recently, HGs isolated by both enzymatic and mild acid hydrolyses from a commercial (acid-extracted) lime pectin have exhibited similar DPs in the range of 121

10.1021/bm801490e CCC: $40.75  2009 American Chemical Society Published on Web 03/04/2009

718

Biomacromolecules, Vol. 10, No. 4, 2009

( 5, indicating that the chemical (mild acid hydrolysis) procedure does not lead to an underestimation of the length of HGs.12 Later, HGs that have been isolated, by mild acid hydrolysis, from water-, oxalate-, dilute acid-, and dilute baseextracted citrus pectins, have displayed similar DPs in the order of 100,5 indicating that mild acid extraction of pectins does not cause any significant modification of the length of HGs. All these observations of similar HG lengths irrespective of their source and isolation method led to the supposition that the DP of HG is independent of the plant origin and is ∼100 on average.5,11 However, these assumptions were made on the basis of studies of few pectins from dicot cell walls and may not be true for other sources of pectins, especially monocot ones. The present study, therefore, aims at examining the structural compositions and some hydrodynamic and functional (gelling) properties of pectins from monocot cell walls in comparison with pectins from dicot cell walls.

2. Experimental Section 2.1. Raw Materials and Purified Cell Wall Materials (CWMs). Four plant species including two monocots (pineapple (Ananas comosus) and banana (Musa AAA cv. Poyo, Cavendish subgroup)) and two dicots (yellow passion fruit (Passiflora edulis var. flavicarpa) and lemon (Citrus limon)) were used. Freshly harvested banana, lemon, pineapple, and yellow passion fruits at maturation stage were purchased from a cooperative of local producers (Bacon, Ivory Coast). Pineapple flesh (PF), banana peel (BP), yellow passion fruit rind (YPFR), and lemon peel (LP) were produced from them as the raw materials and then purified cell walls, referred to as cell wall materials (CWMs), were prepared from these raw materials as described previously.13,14 2.2. Pectin Extraction and Purification. Pectins were extracted with hot citric acid (a natural weak acid extractant) under the following conditions; dry CWM to citric acid extractant weight ratio 1:25, pH 1.8, temperature 75 °C, duration 60 min, and number of extractions 2, which was much less depolymerizing and de-esterifying of extracted pectins than a conventional (dilute strong mineral) acid extraction under similar conditions.15 The pectin extracts from PF and BP cell walls were purified by a two-step purification process using first acidified methanol, followed by cation (cupric ion) binding precipitation methods and those from YPFR and LP cell walls were purified only by the latter method. According to the plant source, the different pectin isolates were referred to as PF, BP, YPFR, and LP pectins. Pectin extraction was carried out at least in three independent runs. 2.3. Partial Acid Hydrolysis of Purified Pectins. Different pectin samples were first cold alkali-de-esterified to prevent β-elimination and differences in acid susceptibility.5 De-esterified samples were submitted to a nongalacturonan degrading partial acid hydrolysis under the following optimum hydrolytic conditions: 0.2 M nitric acid at 100 °C and for 45 min, which was more suitable for all the purified pectin isolates than the hydrolytic conditions of 0.25-1 M sulfuric acid, at 100-120 °C, and for 1 h or 0.25 M hydrochloric acid, at 100 °C, and for 3 h that had been employed to extract galacturonans from various purified cell walls.9,16,17 The resulting acid-insoluble fractions were recovered by centrifugation, washed, solubilized in distilled water upon raising the pH with the aid of 0.1 M sodium hydroxide, extensively dialyzed against distilled water, and freeze-dried for further analysis. 2.4. Characterization. The galacturonic acid content of CWMs, pectin samples, and insoluble products from the partial acid hydrolysis was determined by a modified m-hydroxydiphenyl assay.18 The difference in response of glucuronic and galacturonic acids in the presence and absence of tetraborate was used to distinguish between the two uronic acids and actually quantify GalA present in isolated fractions. The glycosyl residue composition was determined by gas-liquid chromatography (GC) after reduction and conversion of acid-released monosaccharides to corresponding alditol acetate derivatives. In this connection, CWMs were pretreated with 12 M H2SO4

Communications (23 °C, 1 h) and then treated with 1 M H2SO4 (100 °C, 3 h). The isolated pectins and resulting insoluble products from the partial acid hydrolysis were hydrolyzed with 1 M H2SO4 (100 °C, 3 h) or 2 M TFA (120 °C, 90 min). The glycosyl linkage composition was determined by a slightly modified methylation analysis procedure.14,19 Briefly, samples were first per-O-methylated with sodium methylsulfinyl carbanion and methyl iodide in dimethyl sulfoxide and the methyl-esterified carboxyl groups were reduced with lithium triethylborodeuteride (Superdeuteride). After dialysing and freeze-drying, the samples were submitted to acidhydrolysis (2 M TFA, 120 °C, 90 min), reduced with NaBH4, and acetylated. This procedure allowed the galactose derivatives originating from galactosyluronic acid to be identified unambiguously since they were O-acetylated at C-6, whereas the galactosyl residues all gave derivatives that were O-methylated at C-6. The partially methylated alditol acetates (PMAAs) were identified by their GC retention times and electronic impact mass spectra (GC-EIMS) and quantified by GC relative to the used internal standard (myo-inositol hexaacetate). The viscosity-average molecular weight (Mv) and molecular weight distribution (MWD) were analyzed by viscometry and high-pressure size exclusion chromatography (HPSEC), as described elsewhere.14,20 The degree of esterification (DE) of pectin was assessed by titration and gelling properties were evaluated by determining the setting time of gel preparations (65.0% sucrose, 0.70 wt % pectin, pH 2.3) on cooling at 30 °C in an incubator and the gel strength after 24 h of aging at 30 °C with a Ridgelimeter, as described previously.15 The nitrogen (N) content of CWM samples was determined by the Kjeldahl procedure using a Foss Tecator Kjeltec 2300 Analyzer fitted with an automatic distilling and titration unit (Foss Tecator, Ho¨gana¨s, Sweden) and the protein content was calculated using a conversion factor of 6.25 (N × 6.25).14 2.5. Statistical Analysis. All the collected data were statistically evaluated by the global test of a one-way analysis of variance (ANOVA), followed by the Bonferroni’s posthoc test for multiple comparisons whenever applicable using a GraphPad Prism V.3 software (GraphPad software Inc., San Diego, CA). Differences in the means of treatments were evaluated at a significant level of 0.05.

3. Results and Discussion The galacturonic acid contents of CWMs from PF, BP, YPFR, and LP were 4.5, 10.8, 16.9, and 27.8% (w/w), respectively (Table 1). These contents were significantly (p < 0.05) different from one another, indicating that the two dicot (YPFR and LP) cell walls were richer in galacturonic acid than the two monocot (PF and BP) cell walls. This suggested that the dicot cell walls contained higher amounts of pectins than the monocot cell walls, in corroboration with others.17,21 However, PF cell wall could contain much less pectin than BP, and this could also be the case of YPFR cell wall comparatively to LP cell wall. It has also been reported a higher amount of pectin in banana pericarp cell wall than in pineapple leaf cell wall on the basis of the galacturonan amount.17 In terms of the galacturonic acid content, the four CWMs ranged as follows: PF < BP < YPFR < LP, suggesting that the galacturonic acid (and probably the pectin) contents of cell walls depended upon the plant source. On the basis of its rather low galacturonic acid content, pineapple could be considered as a commelinoid species, whereas banana was unlikely. Indeed, even though the galacturonic acid content of CWM from BP was lower than the galacturonic acid contents of CWMs from both YPFR and LP, it is, however, comparable to those reported in CWMs from other dicots such as apple (Malus malus), pawpaw (Carica papaya), and ambarella (Spondias cytherea) on a dry-matter basis.22,23 The yields of isolated PF, BP, YPFR, and LP pectins were 1.3, 2.8, 4.1, and 10.7% (w/w), respectively (Table 1). All the pectin yields were significantly (p < 0.05) different and reflected

Communications

Biomacromolecules, Vol. 10, No. 4, 2009

719

Table 1. Total Recovery, Sugar Composition, Macromolecular, and Gelling Characteristics of Cell Wall Material, Pectin, or Homogalacturonan Isolates from Different Monocot and Dicot Sourcesa PF

BP

YPFR

LP

16.9 ( 2.3c 12.7 ( 1.4c

27.8 ( 2.9d 5.6 ( 1.1d

CWM galacturonic acid (%, w/w) protein (%, w/w)

4.5 ( 0.9a 3.1 ( 0.7a

10.8 ( 1.8b 8.4 ( 0.9b pectin

total recovery (%, w/w) sugar composition (%, w/w) galacturonic acid rhamnose arabinose galactose xylose neutral sugar degree of esterification (mol%) Mv (kDa) jelly setting time (s) jelly grade (° sag)

1.3 ( 0.1a

2.8 ( 0.2b

4.1* ( 0.2c

10.7 ( 0.3d

62.3 ( 1.5a 2.8 ( 0.2a 5.6 ( 0.5a 5.1 ( 0.5a tr 13.6 ( 0.9a 57 ( 3a 60.4 ( 1.7a 1426 ( 4a 91 ( 0.2a

68.4 ( 2.1b 2.6 ( 0.1ab 3.1 ( 0.3b 7.2 ( 0.2b tr 12.9 ( 0.6a 55 ( 4a 87.3 ( 2.9b 1069 ( 5b 137 ( 0.4b

79.2 ( 2.2c 2.4 ( 0.1b 2.7 ( 0.2bc 2.9 ( 0.2c 1.7 ( 0.1 9.7 ( 0.5b 54 ( 2a 79.6 ( 1.2c 853 ( 3c 168 ( 0.3c

84.9 ( 3.1d 1.4 ( 0.1c 2.2 ( 0.2c 1.7 ( 0.3d tr 5.3 ( 0.5c 64 ( 3b 120.7 ( 4.1d 605 ( 7d 226 ( 0.7d

HG total recovery (%, w/w) sugar composition (%, w/w) galacturonic acid rhamnose degree of polymerization

58.6 ( 1.7a

67.3 ( 2.2b

82.3 ( 1.4c

90.2 ( 2.7d

94.2 ( 2.3 tr 59 ( 4a

92.8 ( 1.4 tr 67 ( 3a

93.7 ( 1.8 tr 93 ( 7b

93.1 ( 1.2 tr 102 ( 5b

a Data are means ( standard deviations (n ) 5, except *n ) 3). Mean values in the same line with different letters are significantly different (p < 0.05) using the Bonferroni’s posthoc test. CWM, cell wall material; PF, BP, YPFR, and LP, pineapple flesh, banana peel, yellow passion fruit rind, lemon peel, respectively; HG, homogalacturonan; tr, traces YPFR > BP > PF. These differences in the gelling abilities of the four pectins could be due to the existing significant differences in DE and Mv, with higher values favoring the gel formation15 and increasing the gel formed strength under equal conditions.15,27 It cannot be ruled out, however, that possible differences in intermolecular distribution patterns of the methoxy groups esterifying galacturonic acid residues of the HG building blocks of pectins could also influence their gelling properties to some extent. To gain further structural information for elucidation of their differing hydrodynamic and gelling properties, all the pectins were submitted to a galacturonan-resistant mild acid hydrolysis. The sugar composition and proportion of resulting acid-insoluble fractions are shown in Table 1. Galacturonic acid (92.8-94.2%) was practically the sole glycosyl residue that these acid-insoluble fractions contained. Rhamnose was indeed the only additional neutral glycosyl residue present in traces. As a result, it is inferred that these acid-insoluble fractions corresponded to HGs. Their relative proportions (58.6-90.2%, w/w) within the isolated pectins were, however, significantly (p < 0.05) different. LP pectin had the highest proportion of HGs, followed by YPFR pectin; the lowest proportion being recorded for PF pectin. Furthermore, the molecular weight distribution analysis revealed that each of the different HG fractions exhibited a Gaussianlike distribution pattern (Figure 1), indicating that the HG building blocks of each of the four pectins formed homogeneous populations with respect to the molecular size. These observations are consistent with previous reports,5,11 and probably with the assumption that within a given plant origin, HGs exhibit a very homogeneous molar mass.28 In contrast, calculations of the DP of HG, on the basis of its Mv and sugar molar

Communications

composition, show that the DPs (59-67) of HGs isolated from PF and BP were significantly (p < 0.05) smaller than the DPs (93-102) of the HGs obtained from YPFR and LP pectins (Table 1), as also unambiguously evidenced by their sizeexclusion chromatography elution profiles, showing a higher retention time, and therefore, a peak shifting toward lower molecular sizes (Figure 1). The fact that LP pectin comprised the highest proportion of HG building blocks of a longer DP in addition to having the highest DE could explain its displaying of the strongest gelling ability. Although BP pectin had a higher Mv than YPFR pectin, its gelling capacity, by contrast, appeared inferior, which could be ascribed to its much lower proportion of HG elements having additionally a significantly reduced DP. The rather low gelling ability of PF pectin was demonstrated by it consisting of the lowest proportion of HG building blocks of a shorter size. It is known that many pectin-related factors such as the pectin neutral sugar content, solution concentration, Mv, and chain length influence its gelling properties and that its galacturonan domains play a key role in the building up of junction zones.27 Hence, by considering the classical model of pectin,1,27 it could be suggested, on the basis of the results obtained, that LP pectin formed the longest macromolecular chain, wherein long HG building blocks were interspersed with relatively short neutral sugar containing-RG I blocks, whereas PF pectin formed the least long macromolecular chain in which rather short HG building blocks were interspersed with relatively long neutral sugar containing-RG I ones. As a result, LP pectin displayed by far the highest gelling ability under equal (gel preparation) conditions (pH, soluble solids content, and pectin concentration). In other words, the preparation of gels of the same strength required much higher concentration of PF pectin than the remainder, especially LP pectin. These observations corroborate the reports that the preparation of pineapple fruit jams and preserves required higher addition of (extracted) pectin compared to orange and apple fruit ones.29 Although the neutral sugar content of BP was higher than that of YPFR, it could have contributed to a little-to-no extent the difference in gelling ability, considering that the extra neutral sugar content of BP pectin was not exaggeratedly high as it could be found in other extracted pectins such as beet pectin.30 Furthermore, typical commercial apple pectin generally contains a much higher total neutral sugar (27%) than citrus (lemon) pectin (8.5-9.2%),27 and yet one can behave as a good gelling agent as the other. The DPs of the HGs isolated from the dicot (YPFR and LP) pectins were comparable to the reported values of ∼100 for HGs and galacturonan-like polysaccharides that had been isolated from (pectins from) cell walls of other dicots such as Rosa glauca,9 apple,11 Arabidopsis thaliana,28 and sugar beet (Beta Vulgaris).30 These results indicate for the first time that HG building blocks of monocot pectins have a reduced length compared with HG building blocks of dicot pectins, in contradiction with the previous supposition of an average length of 100, irrespective of the plant (pectin) source. Besides, they show how the HG amount as well as its DP can have a significant impact on pectin gelling properties.

4. Conclusions The present study shows that pectins from different monocot and dicot species can differ by their macromolecular structures and hydrodynamic and gelling properties. The two examined monocot sources, namely, pineapple and banana cell walls,

Communications

contain pectins that are richer in rhamnogalacturonan I, but less rich in homogalacturonan building blocks than the dicot cell wall pectins. Homogalacturonan building blocks of pectins are homogeneous with regard to the molecular size but appear significantly less long in the monocot pectins than in the dicot ones. Few xylogalacturonans are present in yellow passion fruit pectin, thus permitting to view the latter as comprising at least three different structural domains. Acknowledgment. V. Besson is gratefully acknowledged for valuable assistance.

References and Notes (1) Schols, H. A.; Voragen, A. G. J. In Pectins and Pectinases; Visser, J., Voragen, A. G. J., Eds.; Elsevier: Amsterdam, Netherlands, 1996; pp 3-19. (2) Albersheim, P.; Darvill, A. G.; O’Neill, M. A.; Schols, H. A.; Voragen, A. G. J. In Pectins and Pectinases; Visser, J., Voragen, A. G. J., Eds.; Elsevier: Amsterdam, Netherlands, 1996; pp 47-55. (3) Ovodova, R. G.; Popov, S. V.; Bushneva, O. A.; Golovchenko, V. V.; Chizhov, A. O.; Klinov, D. V.; Ovodov, Y. S. Biochemistry (Moscow) 2006, 71, 538–542. (4) O’Neill, M. A.; Ishii, T.; Albersheim, P.; Darvill, A. G. Ann. ReV. Plant Biol. 2004, 55, 109–139. (5) Yapo, B. M.; Lerouge, P.; Thibault, J.-F.; Ralet, M.-C. Carbohydr. Polym. 2007, 69, 426–435. (6) Prade, R. A.; Zhan, D.; Ayoubi, P.; Mort, A. J. Biotechnol. Genet. Eng. ReV. 1999, 16, 361–391. (7) McNeil, M.; Darvill, A. G.; Albersheim, P. Plant Physiol. 1980, 66, 1128–1134. (8) Nakamura, A.; Furuta, H.; Maeda, H.; Nagamatsu, Y.; Yoshimoto, A. Biosci. Biotechnol. Biochem. 2001, 65, 2249–2258. (9) Chambat, G.; Joseleau, J.-P. Carbohydr. Res. 1980, 85, C10–C12. (10) Konno, H.; Yamasaki, Y.; Katoh, K. Phytochemistry 1986, 25, 623– 627. (11) Thibault, J.-F.; Renard, C. M. G. C.; Axelos, M. A. V.; Roger, P.; Cre´peau, M.-J. Carbohydr. Res. 1993, 238, 271–286.

Biomacromolecules, Vol. 10, No. 4, 2009

721

(12) Hellı´n, P.; Ralet, M.-C.; Bonnin, E.; Thibault, J.-F. Carbohydr. Polym. 2005, 60, 307–317. (13) Yapo, B. M.; Koffi, K. L. J. Agric. Food Chem. 2008, 56, 5880– 5883. (14) Yapo, B. M.; Koffi, K. L. J. Sci. Food Agric. 2008, 88, 2125–2133. (15) Yapo, B. M. J. Agric. Food Chem 2009, 57, 1572–1578. (16) Chambat, G.; Barnoud, F.; Joseleau, J. P. J. Plant Physiol. 1984, 74, 687–693. (17) Jarvis, M. C.; Forsyth, W.; Duncan, H. J. Plant Physiol. 1988, 88, 309–314. (18) Filisetti-Cozzi, T. M. C. C.; Carpita, N. C. Anal. Biochem. 1991, 197, 157–162. (19) Pellerin, P.; Doco, T.; Vidal, S.; Williams, P.; Brillouet, J. M.; O’Neill, M. A. Carbohydr. Res. 1996, 290, 183–197. (20) Yapo, B. M.; Koffi, K. L. J. Agric. Food Chem. 2006, 54, 2738– 2744. (21) Carpita, N. C.; McCann, M. In Biochemistry and Molecular Biology of Plants; Buchanan, B., Gruissem, W. Jones, R., Eds.; American Society of Plant Physiologists: Rockville, MD, 2000; pp 52-108. (22) Nwanekezi, E. C.; Alawuba, O. C. G.; Mkpolulu, C. C. M. J. Food Sci. Technol. 1994, 31, 159–161. (23) Koubala, B. B.; Mbome, L. I.; Kansci, G.; Tchouanguep Mbiapo, F.; Crepeau, M.-J.; Thibault, J.-F.; Ralet, M.-C. Food Chem. 2008, 106, 1202–1207. (24) De Vries, J. A.; Voragen, A. G. J.; Rombouts, F. M.; Pilnik, W. Carbohydr. Polym. 1981, 1, 117–127. (25) Aspinall, G. O.; Craig, J. W. T.; Whyte, J. L. Carbohydr. Res. 1968, 7, 442–452. (26) Ros, J. M.; Schols, H. A.; Voragen, A. G. J. Carbohydr. Polym. 1998, 37, 159–166. (27) Voragen, A. G. J.; Pilnik, W.; Thibault, J. F.; Axelos, M. A. V.; Renard, C. M. G. C. In Food polysaccharides and their applications; Stephen, A. M., Ed.; Marcel Dekker: New York, 1995, pp 287-339. (28) Ralet, M.-C.; Cre´peau, M.-J.; Lefe`bvre, J.; Mouille, G.; Ho¨fte, H.; Thibault, J.-F. Biomacromolecules 2008, 9, 1454–1460. (29) May, C. D. Carbohydr. Polym. 1990, 12, 79–99. (30) Yapo, B. M.; Robert, C.; Etienne, I.; Wathelet, B.; Paquot, M. Food Chem. 2007, 100, 1356–1364.

BM801490E