PHASE TRANSITIONS OF POTATO STARCH HYDROLYSATES DIFFERING IN BRANCHING CHARACTERISTICS

Rafał Ziobro¹, Halina Gambuś¹, Dorota Gumul¹, Werner Praznik², Tomasz Jankowski^3
1 Department of Carbohydrates Technology, Agricultural University of Cracow, Poland
² Institute of Chemistry, University of Agricultural Sciences, Wien, Austria
³ Department of Biotechnology and Food Microbiology, Poznań University of Life Sciences, Poland

ABSTRACT

In order to establish the influence of branching characteristics on thermal properties of starch hydrolysates, phase transitions of samples differing in content of amylose have been compared. Potato starch dissolved in DMSO/water mixture was hydrolysed by crude pullulanase preparation and fractionated by precipitation. The fractions differed in their branching characteristics and after solubilisation and retrogradation gave endotherms at different temperatures. The results prove that even slight changes in branching pattern can cause significant variation in phase transitions. The solvent system used in the experiments seems to stabilize linear glucans for a long time and promote recrystallisation of branched molecules.

Key words: starch hydrolysates, enzymatic hydrolysis .

INTRODUCTION

Branching characteristics of starch has a significant influence on its properties. Native starch polymers are commonly divided into two groups: linear long chains (amylose) and short chain branched glucans (amylopectin). During acid or enzymatic hydrolysis their degree of polymerisation and branching characteristics may significantly change. The impact of branching on partially hydrolysed starch polymers, such as maltodextrins with low DE is rarely taking into account in technological applications. However it is well known that linear starch glucans are much more prone to retrogradation and have higher complexing ability than their branched counterparts [6]. The networks built up of amylopectin molecules in dilute solutions have been reported to be stable for several months, while amylose yield thick aggregates after a few days [13]. Although there is a common belief that the reason of such behaviour is branching characteristcs, it is worth to remember that branching is not the only difference between amylose and amylopectin. Amylopectin is known to be one of the largest biopolymers, by orders of magnitude larger than amylose. It significantly affects its mobilty in solution. If the branching pattern is the only cause for the mentioned differences, they should also be observed for molecules of comparable hydrodynamic volume for example partially hydrolysed amylopectins. Currently there are no reports on such systems.

Pullulanase is known to selectively cleave a-1,6 bonds in starch. Therefore under its action there is a significant reduction of amylopectin’s size, while amylose remains unchanged. If we control the degree of hydrolysis it is possible to obtain mixture of branched and linear glucans of comparable size. The phenomena occuring in such systems can give more information on the influence of branching characteristics on the stability of glucans in solution.

In the following study we have checked the impact of different branching characteristics on the phase transitions of the potato starch hydrolysates obtained with crude commercial pullulanase preparation and fractionated by precipitation into three groups differing in branching characteristics.

MATERIAL AND METHODS

40 g of commercial potato starch (Superior) was dissolved in 400 cm³ of dimethylsulfoxide (DMSO) at 70°C and stirred on the magnetic stirrer for 24 h. The solution was then diluted with 1 M acetic buffer (pH = 0.5), heated and stirred for the next day. After cooling the solultion to 50°C, 1 cm³ of pullulanase preparation Pulluzyme 750 (ABM Chemicals, England) was added and the hydrolysis was conducted for 3 days. The solution was cooled, and put in the refrigerator at 4°C, where the self-precipitation occured. Fraction I was obtained after 2 days by centrifugation of the obtained precipitate for 10 min at 4000 rpm (centrifuge, MPW 341). The solution was put back in the refrigerator for the next 5 days, after which fraction II was sampled by the same procedure. The supernatant obtained after centrifugation of fraction II was heated to 70°C and treated with butanol-1. After 1 hour of heating the solution was slowly cooled to 20°C and put in the refrigerator for the next 3 days. Fraction III was obtained as described above. The supernatant was discarded. All the fractions were purified directly after preparation by washing twice with 96% ethanol and centrifugation for 5 min. The drying was performed at ambient conditions and fractions were ground in a mortar.

SEC

Preparation of native potato starch
17 mg of potato starch was mixed with 1 cm³ 5 M NaOH and stirred at 60°C (3 days, magnetic stirrer) in a 12 cm³ test tube (closed with a screw cap). After dissolving the solution was neutralized with 0.25 cm³ 2 M HCl (pH 7), centrifuged (10 min at 13 000 rpm) and analyzed on SEC column system II.

Preparation of hydrolysed fractions
50 mg of the fraction was mixed with 3 cm³ 0.5 M NaOH and stirred at 60°C (3 days, magnetic stirrer) in a 12 cm³ test tube (closed with a screw cap). After dissolving 0.8 cm³ of the solution was neutralized with 0.2 cm³ 2 M HCl (pH 7), centrifuged (10 min at 13 000 rpm) and analyzed on SEC column system II.

Derivatization of samples with 2-amino pyridine (2-AP)
60-70 mg of samples (native starch and hydrolysed fractions) were weighed in a 12 cm³ test tube and dissolved in 300 mg 2-AP and 1 cm³ H₂O at 60°C (closed with a screw cap). Dissolution and derivatization period was approx. 20–24 h. Subsequently 1.2 cm³ 2 M HCl (pH 6-7) and 0.2 g dimethylaminoborane were added and stirred for additional 2 days at 60°C. The resulting solution was transfered in 20 cm³ methanol (centrifuge cape). The precipitate was separated at 3500 rpm (15 min), the supernatant was removed, the precipitate washed 3 times with each time 5 cm³ methanol, 2 times with 5 cm³ acetone and, finally, dried at 60°C.

Preparation of AP-samples for SEC-analysis
8-10 mg of derivatized sample dissolved in 1 cm³ 0.5 M NaOH was stirred at 60°C (2 days, magnetic stirrer) in a 12 cm³ test tube (closed with a screw cap). Subsequently the samples were neutralized with 0.25 cm³ 2 M HCl (pH 7), centrifuged (10 min at 13 000 rpm) and analyzed on SEC-column system.

Debranching of samples with pullulanase
For debranching, 0.6 cm³ of AP-sample was incubated with 20 mm³ of Pullulanase M1 (Megazyme Int. Irland LTD., from Klebsiella planticola, 720U/mL, in 3.2M amm. sulph., purified by ultra filtration in the centrifuge, removing of stabilizing sugars and salts) over night at 30°C. The resulting samples were analyzed at SEC column system I.

SEC-systems
System I consisted of pre-column (10 × 100 mm, Fractogel) + Superose 12 (10 × 300 mm) + 2 columns Fractogel HW40 (10 × 300 mm). The eluent was 0.02 M NaCl, pumped at the flowrate 0.6cm³/min. Injected sample volume was 200 mm³. DRI and fluorescence detectors were used. Separation range was 1000–200 000 Dalton (Superose 12). Calibration was based on the absolute determination of molecular weight distribution of 2-amino pyridine endgroup labeled samples of potato and fractions I–III. Molecular weight distributions were calculated from combined fluorescence- and RI-elution profiles. Molecular weight distribution for non-derivatized samples was achieved from AP-starch calibration.

System II included pre-column (10 × 100 mm, Fractogel) + Superose 6 (10 × 300 mm)+ 2 columns Fractogel HW40 (10 × 300 mm). Separation range was 1000–1 000 000 Dalton ( Superose 6). Other data were the same as for System I.

Iodine staining of SEC-fractions
For control the fractions of SEC (1 fraction = 1.8 cm³ ) were stained with 0.5cm³ I/KI reagent (125 mg I and 400 mg KI in small amount of H₂O, diluted with 500 cm³ H₂O and 500 cm³ 0.1 M acetic acid) and investigated with respect to extinction at 525 and 640 nm. Extinction at 640 nm informs about lcb/nb chains and at 525 nm about scb chains of investigated starch. Finally, the ratio of 640 /525 nm informs about the ratio of lcb/nb and scb components in the fractions.

DSC
The gelatinisation of native potato starch was checked by the modified method of Fredriksson et al. [4]. 5 mg of starch was placed in an aluminum pan and thoroughly mixed with identical amount of water with a thin needle. The pan was hermetically closed and left for 2 h to equilibrate. Scans were performed in a Perkin-Elmer DSC-7 at rate 10°C/min. The empty pan was used as blank. Transition temperatures (T, T, T) and enthalpy (DH) were read with the aid of system software.

The endotherms of retrograded starch were done according to Sievert and Wuersch [15]. Hermetically closed aluminum pans with starch water mixture were heated in a calorimeter to 120°C, cooled and stored at 8°C for 24, 48 and 168 h. After this time the sample was again put in the calorimeter and scanned from 20 to 180°C at rate 10°C/min.

RESULTS AND DISCUSSION

Preparation of the samples
The amounts of the fractions obtained from 40 g of potato starch were 13, 1 and 3 g for fractions I - III, respectively. Consequently total yield was about 42%. Comparison of the masses of fraction I and II demonstrates that the observed self-precipitation may be regarded as a rapid process, comparable to normal retrogradation. It is however unusual, that a significant amounts of amylose are stable in a solution for a long period and precipitate only after complexing with butanol. The possible reason for this stability is the interaction of linear glucans with DMSO present in solution.

SEC results
Weight average molecular weight (296 kDa) of native potato starch estimated from the results of SEC on Fractogel/Superose columns (Fig. 1) are almost two orders of magnitude lower than those usually reported for potato starch. It suggests that the applied solubilisation conditions significantly decreases the hydrodynamic volume of starch molecules (especially amylopectin), probably due to the disruption of some weak bonds (hydrogen or van der Waals) between aggregated glucans. The differences between non-derivatized and aminopyridine-labelled samples are minor (in this case M_w = 272 kDa).

Fig. 1. Molecular weight distribution of native potato starch prior and after derivatization with 2-aminopyridine

Fraction I exhibited significantly lower molecular weights especially in case of AP-sample (M_w = 57 kDa, Fig. 2). It may be observed that untreated sample (M_w = 76 kDa) contained significant amount of relatively high molecular weight components, which were dissolved during derivatisation. It is probable, that those components are similar to aggregates existing in the native potato amylopectin. The tendency of glucans present in fraction I to form aggregates may be the reason of their rapid self-precipitation.

Fig. 2. Molecular weight distribution of fraction I prior and after derivatization with 2-aminopyridine

Molecular weights of glucans present in fraction II (Fig. 3) were close to those observed in fraction I. The differences between non-derivatized (M_w = 64 kDa) and AP-labelled samples (M_w = 63 kDa) were in this case much smaller. The lack of aggregated structures is probably due to highly branched character of this fraction, which was observed by iodine staining [5].

Fig. 3. Molecular weight distribution of fraction II prior and after derivatization with 2-aminopyridine

Some aggregation could be observed in fraction III, which was obtained by butanol treatment (Fig. 4). The molecular weights were in this case lowest and did not exceed 10⁵ Da. The low weight average molecular weight of those linear glucans (45 kDa for untreated sample, 40 kDa after derivatisation) reflects their hydrolysis by traces of a-amylase present in enzymatic preparation.

Fig. 4. Molecular weight distribution of fraction III prior and after derivatization with 2-aminopyridine

Debranching analysis
After debranching native potato starch exhibited broad molecular weight distribution (Fig. 5), and weight average DP of 75. Fractions I and II had much lower chain lengths and did not differ in this aspect. The average DP of 22 is similar to the average chain length of potato amylopectin reported by Hanashiro et al. [7]. Thus self-precipitating fractions seem to be formed mainly of partly debranched amylopectin fragments. This is in contrary to the opinions that branching limits aggregation. Most probably some relatively long linear chains are necessary to make crosslinks between branched molecules, because when waxy maize starch was used as a substrate no self-precipitation was observed (unpublished data). This would explain why the more stable fraction II has lower iodine affinity [5]. Other important factor which could affect aggregation is the solvent system. DMSO is known to influence the shape of amylopectin and increase its density [3, 11]. In water solution it may be regarded as complexing agent, which interacts with starch and limits formation of the uniform network. The resulting small micellae containing starch with DMSO can interact with each other until they are too large and precipitate.

Fig. 5. Molecular weight distribution of debranched potato starch and fractions I-III

Fraction III was only slightly degraded by pullulanase. The average molecular weight was about 4 times lower after debranching (M_w = 10 kDa, DP = 63), which should be close to the number of branches per molecule. The low degree of branching suggests that this fractions consists of pure amylose, probably partly degraded due to residual a-amylase activity of pullulanase preparation. The amylose chains consisting of less than 100 anhydroglucose units are normally regarded as most prone to retrograde [12]. This was not the case, as fraction III was stable in solution and precipitated only after butanol complexation. Such strange behaviour seems to be caused by the presence of DMSO, which significantly affects the conformation and flexibility of amylose helix [14, 17].

DSC
At the excess of water potato starch exhibits single endotherm (Fig. 6), because the higly mobile amorphous phase facilitates the disruption of crystalline regions and they melt concurrently with granule swelling [2]. Gelatinisation endotherm displays relatively narrow range of transition temperatures (63–79.4°C), which reflects the uniform structure of crystalline regions in this type of starch [4]. The substantial degree of crystallinity of potato starch amylopectin results in high gelatinisation temperature and enthalpy (17.6 J/g). Potato starch represents B type packing type, which is known to be more stable than A type, characteristic for cereal starches [8]. The double helices in this type of structure are less stretched during granule swelling, so their disruption occurs at higher temperatures. The loss of double helical order in starch is known to be the main factor influencing gelatinisation enthalpy [1].

Fig. 6. Phase transitions of retrograded potato starch

Phase transitions of retrograded native potato starch are shown in the Figure 7. The endotherm observed during dissolving of starch retrograded during 24 hours of storage, was characteristic for at least partly branched glucans, as the range of phase transition temperatures did not exceed 100°C. The large aggregates build up of amylopectin and amylose molecules are thermally more labile than pure amylose crystallites [10]. Low enthalpy (1.72 J/g) and decreased by 13°C onset temperature are characteristics for such crystallites [16]. After 48 h the process continued (DH = 6.73 J/g), and additionally to the described transition (DH = 9.72 J/g) the small endotherm of recrystallised amylose could be observed at approximately 160°C. The enthalpy of this phase transition at the end of the storage period was 0.55 J/g.

Fig. 7. Phase transitions of retrograded potato starch

The retrogradation of solubilized fraction I significantly differed from those observed for native starch (Fig. 8). Although there was some trace of phase transition around 50°C, the main endotherm occured slightly above 100°C, probably because of higher concentration of linear chains in the obtained gel [5]. The lower enthalpy of the transition indicates that the crystallites were also more labile in comparison to native starch. This results are different from those obtained for acid hydrolysed starch, which revealed higher initial recrystallisation rate than native wheat starch [18]. This could be due both to the starch origin and hydrolysis method. Native potato starch has different retrogradation tendency than wheat, although there are contradictory opinions on the direction of this relation [9, 16]. Moreover, unlike acid treatment, the action of pullulanase is highly selective and gives product with significantly changed branching characteristics. This may adversly influence retrogradation tendency or decrease the amount of crystalline phase in retrograded starch.

Fig. 8. Phase transitions of retrograded hydrolysate fraction I

The retrogradation of fraction II (Fig. 9) was different from fraction I, and much more resembled native sample. The ednotherm of crystallites formed after 24 h found at approximately 50°C, and was not affected by the time of storage. Due to the small amount of linear chains in this fraction, the endotherm at 120 was not found.

Fig. 9. Phase transitions of retrograded hydrolysate fraction II

Fraction III was only partly solubilised at the applied conditions. The intense endotherm (Fig. 10) could be observed in the same range as in fraction I. The high enthalpy of the transition is typical for RS, which is insoluble even after cooking. This confirms the hypothesis, that the behaviour of linear glucans in water and DMSO is very different. In the absence of DMSO long linear glucans quickly retrograde and form thermodynamically stable precipitate. During the experiments some ordering of the structure took place, which was reflected by narrower temperature range of the transition observed after 168 h of storage at 8°C.

Fig. 10. Phase transitions of retrograded hydrolysate fraction III

CONCLUSIONS

The observed phase transitions of potato starch hydrolysates correspond to their branching characteristics. In a native starch the recrystallisation of branched molecules is dominating in the first period of storage, pure amylose crystals are formed after longer time. This may look to be in contrary to the common opinion, that amylose retrogradation is a short time process, but should rather be explained by the interaction between branched and linear glucans in the formation of the crystalline or double helical structures. The crystallisation of amylose is obviously a different phenomenon from its retrogradation.

The small changes in DSC endotherms over time indicate, that most of the ordered structures in the retrograded samples were formed during first 24 h. During the next days they were only slightly reorganized.

The applied water/DMSO system obviously hindered amylose precipitation from starch solution. However after complexation with butanol it formed stable, highly ordered structure, that retrograded quickly from water solution. The DSC data clearly indicate that this product is an example of thermostable resistant starch.

In case of starch hydrolysates even small differences in the amount of linear chains have a significant influence on phase transitions. For example fraction I and II recrystallize in a different way, although their molecular weight distribution and average chain length are very similar. The possible difference could be their internal structure – fraction I contains higher amount of long linear fragments, fraction II has probably shorter and more uniform distances between branching points.

The work was supported by KBN grant: PBZ/KBN/021/P06/99/03.

REFERENCES

1. Cooke D., Gidley J., 1992. Loss of crystalline and molecular order during starch gelatinization, origin of the enthalpic transition. Carbohydr. Res., 277, 550–561.

2. Cottrell J. E., Duffus C. M., Paterson L., Mackay G. R., 1995. Properties of potato starch: effects of genotype and growing conditions. Phytochemistry, 40(4), 1057–1064.

3. Durrani C. M., Donald A. M., 2000. Shape, molecular weight distribution and viscosity of amylopectin in dilute solution. Carbohydr. Polym., 41, 207–217.

4. Fredriksson H., Silverio R., Andersson R., Eliasson R., Aman P., 1998. The influence of amylose and amylopectin characteristics on gelatinization and retrogradation properties of different starches. Carbohydr. Polym., 35, 119–134.

5. Gambuœ H., Ziobro R., Jankowski T., Gumul D., Pałasiński M., 2004. Próba zastosowania pullulanazy do otrzymywania liniowych glukanów ze skrobi ziemniaczanej. [An attempt to use pullulanase for the production of linear glucans from potato starch], Żywn.. Nauka Technol. Jakoœć. 38(1), 41–54 [in Polish].

6. Garcia-Alonso A., Jimenez-Escrig A., Martin-Carron N., Bravo L., Saura-Calixto F.,1999. Assessment of some parameters involved in the gelatinization and retrogration of starch. Food Chem., 66, 181–187.

7. Hanashiro I., Abe J., Hizukuri S., 1996. A periodic distribution of the chain length of amylopectin as revealed by high-performance anion-exchange chromatography. Carbohydr. Res., 283, 151–159.

8. Hoover R., 2001. Composition, molecular structure, and physicochemical properties of tuber and root starches: a review. Carbohydr. Polym., 45, 253–267.

9. Jacobson M. R., Obanni M., BeMiller J. N., 1997. Retrogradation of starch from different botanical sources. Cereal Chem., 74(5), 511–518.

10. Klucinec J. D., Thompson D. B., 1999. Amylose and amylopectin interact in retrogradation of dispersed high-amylose starches. Cereal Chem., 76(2), 282–291.

11. Millard M. M., Wolf W. J., Dintzis F. R., Willett J. L., 1999. The hydrodynamic characterisation of waxy maize amylopectin in 90% dimethyl sulfoxide-water by analytical ultracentrifugation, dynamic and static light scattering. Carbohydr. Polym., 39, 315–320.

12. Pfannemueller B., 1992. Struktura i właœciwoœci skrobi. Structure and properties of starch. In Materiały IV Letniej Szkoły Skrobiowej – Problemy modyfikacji skrobi, 63–78.

13. Putaux J., Buleon A., Chanzy H., 2000. Network formation in dilute amylose and amylopectin studied by tem. Macromolecules, 33, 6416–6422.

14. Radosta S., Haberer M., Vorwerg W., 2001. Molecular characteristics of amylose and starch in dimethyl sulfoxide. Biomacromolecules, 2, 970–978.

15. Sievert D., Wuersch P., 1993. Amylose chain association based on differential scanning calorymetry. Food Sci., 58, 1332–1334.

16. Singh N., Singh J, Kaur L., Sodhi N. S., Gill B. S., 2003. Morphological, thermal and rheological properties of starches from different botanical sources. Food Chem., 81, 219–231.

17. de Vasconselos C. L., de Azevedo F. G., Pereira M. R., Fonseca J. L. C., 2000. Viscosity-temperature-concentration relationship for starch-DMSO-water solutions. Carbohydr. Polym., 41, 181–184.

18. Zhang W., Jackson D. S., 1992. Retrogradation behavior of wheat starch gels with differing molecular profiles. J. Food Sci., 57(6), 1428–1432.

---

Rafał Ziobro
Department of Carbohydrates Technology,
Agricultural University of Cracow, Poland
Balicka 122, 30-149 Cracow, Poland
email: rrziobro@cyf-kr.edu.pl

Halina Gambuś
Department of Carbohydrates Technology,
Agricultural University of Cracow, Poland
Balicka 122, 30-149 Cracow, Poland

Dorota Gumul
Department of Carbohydrates Technology,
Agricultural University of Cracow, Poland
Balicka 122, 30-149 Cracow, Poland
Phone: (+48 12) 662 47 71
Fax: (+48 12) 662 47 47
email: rrgumul@cyf-kr.edu.pl

Werner Praznik
Institute of Chemistry,
University of Agricultural Sciences, Wien, Austria
Muthgasse 18, A-1190 Wien, Austria

Tomasz Jankowski
Department of Biotechnology and Food Microbiology,
Poznań University of Life Sciences, Poland
Wojska Polskiego 48, 60-627 Poznań, Poland
Phone: 061 846 60 04

---

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.

---

Main  -  Issues  -  How to Submit  -  From the Publisher  -  Search  -  Subscription