Determination of amylose content in different starches using modulated differential scanning calorimetry

^a,b Moorthy, S.N, cAndersson, Lena, ^a*Eliasson, Ann-Charlotte. ^a,c,d Santacruz, Stalin, ^dRuales, Jenny. ^a Department of Food Technology, Center for Chemistry and Chemical Engineering, University of Lund, P.O. Box 124, S-22100 Lund, Sweden. ^b Division of Crop Utilisation and Biotechnology, Central Tuber Crops Research Institute, Sreekaryam, Trivandrum-695917, India. cDepartment of Food Science, Swedish University of Agricultural Sciences, P.O. Box 7051, S-750 07 Uppsala, Sweden. ^d Department of Food Science and Biotechnology, Escuela Politécnica Nacional, P.O. Box 17 01 2759 Quito, Ecuador. ^* corresponding author Abstract A simple calorimetric method for determination of amylose content in starch is reported. Modulated differential scanning calorimetry (MDSC) was used since there are continuous heating-cooling cycles taking place, which can help in release of amylose from starch granules, and therefore more accurate results can be expected. Two common surfactants viz. sodium dodecyl sulphate (SDS) and cetyltrimethylammonium bromide (CTAB) were used. A large number of native starches including cereal, root and pea starches were examined. The results were compared with those obtained by iodimtery and gel permeation chromatography (GPC) for all the starches. There was a good match between the values from iodimetry and GPC with those obtained using the surfactants. Both surfactants seemed to work equally well, even for those starches of around 40% in amylose content. However, in case of SDS higher standard deviations were usually obtained than for CTAB in the determination of transition enthalpies. Keywords : Amylose content, starch, Modulated DSC, surfactants, enthalpy of transition.

1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

0. First Page
1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

1. Introduction

Starch characteristics are influenced to a large extent by the amylose fraction present in it ^1-3. Widely varying amylose content is observed in starch isolated from various botanical sources, depending on the variety and environmental factors [2, 4-6]. The amylose content can vary from nearly zero in waxy maize to over 50% in some pea and maize varieties. In addition, the breeders have been able to develop lines with very high or very low amylose contents, which are suitable for special applications in food as well as non-food industry. The wide variability in the amylose content makes it imperative to have accurate methods for its determination. The original colorimetric procedure ⁷ has been modified by many workers in order to make it more accurate ^8-11. Another method, which has been used extensively, is potentiometric titration of bound iodine ¹². Near Infrared Reflectance (NIR) Spectroscopy has recently been tried in the determination of amylose content in rice starch ¹³. Gel permeation chromatography (GPC) is being widely used to estimate the amylose content ¹⁴. However, all these methods suffer from some drawbacks. Whereas the iodometric method is simple, it suffers from the difficulty in solubilising the starch completely, which can lead to erroneous results. Use of various solvents and treatments have been suggested to overcome this problem ^8-11. Interference from amylopectin, though small, can also give unreliable results. The amperometric method is slow and the broad inflection point can lead to inaccuracies. NIR procedure necessitates standardisation for each material. The GPC method is very reliable, but it uses costly enzymes and is time consuming. Differential Scanning Calorimetry (DSC) is becoming more and more important in studying starch characteristics, especially gelatinisation and retrogradation. Complexation of starch with lipids and surfactants has been well studied ^15-16, and most of the results indicate that it is the amylose fraction that forms the complex that can be studied using DSC ^17-19. When a starch suspension is heated in presence of a lipid or surfactant two endothermic peaks are observed, the first one corresponding to the gelatinisation of starch, and the second one due to the melting of the amylose-lipid complex. Whereas starch gelatinisation is an irreversible process, the latter is reversible and can be observed during reheating of the sample ^20-21. It has also been found that gelatinisation of starch helps in rendering the amylose molecules more accessible to the lipid /surfactant to form a complex. Further, a second heating leads to higher enthalpy of melting of the complex accompanied by a perceptible shift in the temperature at onset of melting, T_o, ²². All these results indicate that the complex is stronger in the second cycle. Since the starch gelatinisation is not reversible, no peak in that region is observed during the second cycle. The ability of the amylose to complex with lipids, and the thermal transition of this complex, offer a possibility for using DSC for amylose determinations. This possibility was first investigated by Kugiyama and Donovan to determine the amylose content in starch ¹³. The amylose content in eight starch samples was determined based on the enthalpy of melting of the complex of amylose with lysolecithin, compared to standard amylose. Their results indicated that the amylose content determined by this method agreed well with the results obtained by iodine staining for five of the starches. For the remaining three starches, the values were close to that obtained by GPC. The DSC-method has been further improved by Sievert and Holm ²⁴ and Mestres et al ²⁵. Gough et al ²⁶ briefly mentioned the use of sodium dodecyl sulphate (SDS) in a calorimetric study of amylose in rice and wheat starches. The enthalpy (H) measured in the second heating can safely be taken as the enthalpy change due to the melting of amylose-lipid complex. This is the basis of the determination of amylose content in starch using lipids and surfactants ^{25, 27}. The aim of the present study was to investigate whether different types of surfactants can be used for the amylose determination using DSC. It has been clearly established that heating-cooling-reheating processes can help in the release of amylose from starch granules and, thereby, amylose-lipid complex formation ²⁴. Since modulated DSC (MDSC) makes use of a sinusoidal modulation (oscillation) that is overlaid on the conventional linear heatiang or cooling ramp ²⁸, it was presumed that it could be exploited for amylose determination in starches. Two ionic common surfactants, SDS and cetyltrimethylammonium bromide (CTAB) were used for complexation. A wide range of starches, including cereal, tuber, pea and genetically modified starch samples, was studied. The results determined in the DSC-method were compared to the values obtained using a iodometric method and GPC. All the amylose content values were determined in our laboratories for the same batches of starches to eliminate variability, which may be occurring if the comparisons were based on literature data.

0. First Page
1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

2. Results and discussion

The results obtained with MDSC were compared with the results obtained with DSC for three of the starches: A. xanthorriza, C. edulis and O. tuberosa. It was found that the onset of melting (T) and the melting enthalpy (H) of the transition of the amylose-lipid complex were higher, approximately 1.5 times higher for H, with MDSC compared to the ordinary DSC (data not shown). The increase in these values can be attributed to the higher complexation achieved by the continuous temperature cycling. The efficacy of MDSC in differentiating reversible and non-reversible peaks in different systems has been highlighted ²⁹. Thus, MDSC seems to be more suitable for studying starch-lipid systems and determining amylose content based on the enthalpy of starch-lipid complex melting. The enthalpy values of melting the complexes (H), and the amylose contents calculated based on complexation with SDS or CTAB, iodine binding and GPC are given in Table 1. The starches thus represent both A-starches (6 starches) and B-starches (17 starches), whereas only one C-starch was included. There will thus be both lipid-containing and lipid-free starches in the study. The amylose content was expected to be in the range from 0 to about 40%. The amylose content, determined using CTAB complexation, is plotted against the amylose content determined using GPC in Fig. 1. The corresponding plot for results obtained with complexation with SDS is shown in Fig. 2. Equations and correlation coefficients (R²) obtained with linear regression analysis are given in Table 2. Corresponding calculations were done for the cereal starches, and the root and tuber starches, separately. High correlation coefficients were obtained in all the cases (Table 2). These results show that there is no big difference in the amount of amylose determined related to the type of surfactant used. The length of the hydrophobic chain also is very important in deciding the complex formation ³⁰. It has been found that 12 to 18 carbon chains are optimal for complexation with amylose. Both surfactants are within this range and there does not seem to be significant differences between these two chain lengths. Other factors like size of the hydrophilic group may also have a role which need to be examined. A correlation between the results obtained by SDS and CTAB gave the correlation coefficient R² = 0.852, and the equation: %A_SDS = 0.759 + 0.895 * %A_CTAB where %A_SDS is the amylose content determined from complexation with SDS, and %A_CTAB is the amylose content determined from complexation with CTAB. The equation showed that amylose contents above 7.2% the CTAB method tend to give a higher value compared to SDS, whereas below 7.2% the SDS method tend to give higher values compared to CTAB. The results in Table 2 further show that there is no big difference in the results in relation to the type of starch, which might have been expected. The correlation coefficient is somewhat better for A-starches compared with B-starches. It has been speculated that in case of iodine binding the amylose content could be overestimated because of long amylopectin chains contributing to the iodine binding. However, there is no support for such an effect in case of lipid complexation. If amylopectin should form a complex with the added surfactant this complex could not be expected to show a thermal effect in the temperature range corresponding to the transition of the amylose-lipid complex ³¹. There was good agreement between the values for amylose obtained by GPC and iodine staining, and those obtained using CTAB or SDS complexation. The correlation coefficients were all above 0.81, with generally lower values for SDS. For cereal starches the apparent amylose content was lower than the total, which is expected because of the lipid content in the cereal starches (Table 1). For most of the tuber starches, the apparent amylose was higher than total amylose, as determined from iodine binding. A similar trend has earlier been found for potato starch ³² and barley starch ³³, and it may be due to the blue value of the amylopectcin ^34-38. The MDSC method thus appears to be quite efficient since it gives acceptable values for a wide range of starches. The surfactants used are cheaper than lysolecithin, and the source of surfactant seems not to be critical. Unlike iodine staining, which often requires defatting to get more accurate results, the present method does not need any such step.

0. First Page
1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

3. Experimental

Materials

Starch was extracted from cassava, Colocasia esculenta, Dioscorea alata, D. rotundata, D esculenta, Xanthosoma saggittifolium, Arrowroot, Amorphophallus paeoniifolius, Pacchyrhizus, Canna edulis (purple and dark purple selections) by standard procedure from tubers harvested at CTCRI Farm under standard package of practices. Starch was also extracted from tubers of Arracaha xanthorhiza cultivar FB-001 and Canna edulis cultivar MH-1173 obtained from International Potato Centre, Quito, Ecuador and from Oxalis tuberosa purchased in the local market at Quito, Ecuador. The cereal and pea starches were obtained from Svalöf-Weibull AB, Landskrona, Sweden and starch extracted according to Meredith and Dengate ³⁹. The three barley starches were pearled prior to isolation. Waxy maize and the three potato starches were supplied by Lyckeby-Stärkelsen, (Kristianstad, Sweden). Potato starch amylose was obtained from Sigma, and iodine from Merck, Darmstadt, Germany. Isoamylase (EC. 3.2.1.68, 59 000 U/mg protein) was purchased from Hayashibara Biochemical Laboratories, Inc., Japan.

Methods

Colorimetric determination of the amylose content was carried out according to the procedure described by Morrison and Laignelet ⁸. Six replications were used for each sample. GPC of the starch samples was done by the following procedure, described earlier ³². The starch was de-branched using isoamylase and passed through a Sepharose CL-6B column. The relative carbohydrate content in collected fractions was determined by the phenol-sulphuric acid method ⁴⁰. Two runs were made for each sample. DSC was run on Seiko 6200 DSC (Seiko Instruments Inc., Shizouko, Japan) equipment provided with an inbuilt software. The samples were directly weighed into coated aluminium pans (TA Instruments, USA). SDS- (5% w/w) or CTAB- (5% w/w) solution was added to get a water-starch ratio of 2:1 (10% surfactant-w/w on starch basis). The pans were prepared and allowed to stand for 1 hour before the measurements were carried out. An empty aluminium pan was used as reference. The heating cycle used was as follows. First heating from 15⁰C to 150⁰C at the rate of 3⁰C min^-1, cooling to 30⁰C at 30⁰C min^-1, second heating 30⁰C to 130⁰C at 3⁰C min^-1, and final cooling to 30⁰C at 30⁰C min^-1. The modulation frequency was 0.17Hz and the amplitude 2⁰C. The melting enthalpy of the starch-surfactant complex (H) was determined from the thermogram giving the total value, and calculated on dry weight basis. The results given are the average value of three measurements. The statistical evaluation was done using Microsoft Excel software.

0. First Page
1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

4. Acknowledgement

The financial support from the Cerealia Foundation R&D is acknowledged.

References

1. C. Mestres, P. Colonna, A. Buleon, J. Food Sci., 53 (1988) 1809-1812.
2. A. C. Eliasson, M. Gudmundsson, Starch: physicochemical and functional aspects, in: A.-C. Eliasson (Ed.), Carbohydrates in Food, Marcel Dekker, New York, 1995, pp. 431-503.
3. C. G. Biliaderis, Starch Polymers, in: H. Walter (Ed.), Polysaccharide Association Structures in food, Marcel Dekker, Inc., New York, 1998, pp. 57-168.
4. I. Defloor, I. Dehling, J. A. Delcour, Starch/Stärke, 50 (1998) 58-64.
5. W. R. Morrison, M. Nazir Azudin, J. Cereal Sci., 5 (1987) 45-50.
6. S. N. Moorthy and T. Ramanujam, Starch/Stärke 38 (1986) 58-61.
7. R. M. McCready, W. Z. Hassid, J. Amer. Chem. Soc., 65 (1943) 1154-1157.
8. W. R. Morrison, B. Laignelet, J. Cereal Sci., 1 (1983) 9-20.
9. C. M. Sowbhagya, K. R. Bhattacharya, Starch/Stärke, 23 (1971) 53-56.
10. C. A. Knutson, M. J. Grove, Cereal Chem., 71 (1994) 469-471.
11. S. J. McGrane, H. J. Cornell, C. J. Rix, Starch/Stärke, 50 (1998) 158-163.
12. F. L. Bates, D. French, R. E. Rundle, J. Amer. Chem. Soc., 65 (1943) 142-148.
13. R. D. Delwiche, M. M. Bea, R. E. Miller, B. D. Webb, P. C. Williams, Cereal Chem., 72 (1995) 182-187.
14. S. Kobayashi, S. J. Schwartz, D. R. Lineback, J. Chromatogr., 319 (1985) 205-214.
15. F. F. Mikus, R. M. Hixon, R. E. Rundle, J. Am. Chem. Soc., 68 (1946) 1115-1123.
16. N. Krog, Starch/Stärke, 23 (1971) 206-210.
17. J. W. Donovan, C. J. Mapes, Starch/Stärke, 32 (1980) 190-193.
18. A. C. Eliasson, Thermochim. Acta, 95 (1985) 369-374.
19. R. Stute, G. Konieczny-Janda, Starch/Stärke, 35 (1983) 340-347.
20. C. G. Biliaderis, C. N. Page, L. Slade, R. R. Sirett, Carbohydr. Polym., 5 (1985) 367-389.
21. A. C. Eliasson, Carbohydr. Res., 172 (1988) 83-95.
22. A. C. Eliasson, Thermochim. Acta, 246 (1994) 343-356.
23. M. Kugimiya, J. W. Donovan, J. Food Sci. , 46 (1981) 765-770,777.
24. D. Sievert, P. Würsch, Cereal Chem., 70 (1993) 333-338.
25. C. Mestres, F. Matencio, B. Pons, M. Yajid, G. Fliedel, Starch/Stärke, 48 (1996) 2-6.
26. B. M. Gough, P. Greenwell, P. L. Russell, On the interaction of sodium dodecyl sulphate with starch granules, in: R. D. Hill, L. Munck (Eds.), New Approaches to Research on Cereal Carbohydrates, Elsevier, Amsterdam, 1985, pp. 99-108.
27. D. Sievert, P. Würsch, J. Food Sci., 58 (1993) 1332-1334, 1345.
28. M. F Lai, Cheng-Yi Lii, Cereal Chem., 76 (1999) 519-525
29. L. N. Bell, D. E. Touma, J. Food Sci., 61 (1996) 807-810, 828.
30. M. Kowblansky, Macromolecules, 18 (1985) 1776-1779.
31. V. K. Villwock, A.-C. Eliasson, J. Silverio, J.N. BeMiller, Cereal Chem., 76 (1999) 292-298.
32. H. Fredriksson, J. Silverio, R. Andersson, A.-C. Eliasson, P. Åman, Carbohydr. Polym., 35 (1998) 119-134.
33. H. Tang, H. Ando, K. Watanabe, Y. Takeda, T. Mitsunaga, Carbohydr. Res., 330 (2001) 241-248.
34. Y. Takeda, C. Takeda, H. Mizukami, I. Hanasuiro, Carbohydr. Polym., 38 (1999) 109-114
35. Y. Takeda, S. Hizukuri, B. Juliano, Carbohydr. Res., 168 (1987) 79-88
36. S. Hizukuri, Y. Takeda, N. Maruta, B. Juliano, Carbohydr. Res., 189 (1989) 227-235
37. Y. Takeda, T. Shitaozono, S. Hizukuri, Starch/Stärke, 40 (1988) 51-54
38. A. Schulman, S. Tomooka, A. Suzuki, P. Myllarinen, s. Hizukuri, Carbohydr. Res., 275 (1995) 361-369
39. P. Meredith, H. N. Dengate, W. R. Morrison, Starch/Stärke, 30 (1978) 119-125.
40. M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F. Smith, Analytical Chemistry, 28 (1956) 350-356.

0. First Page
1. Introduction
2. Results and discussion
3. Experimental
4. Acknowledgement
5. Tables

5. Tables

Table 1 : Amylose content determined by modulated DSC, iodine binding and GPC

^aAverage and standard determination of triplicate measurements of DH. ^bAmylose content calculated according to the following formula: % amylose = (DH/16.5)*100, where 16.5 ± 0.5 was determined as the melting enthalpy for potato amylose-SDS complex. ^cAmylose content calculated according to the following formula: % amylose = (DH/15.3)*100, where 15.3 ± 0.5 was determined as the melting enthalpy for potato amylose-CTAB complex. ^dND = not detected Table 2 : Linear correlation (p<0.001) between different subgroups in determination of amylose content.