Study of thermal properties and heat-induced denaturation and aggregation of soy proteins by modulated differential scanning calorimetry

International Journal of Biological Macromolecules 40 (2007) 96–104

Study of thermal properties and heat-induced denaturation and aggregation of soy proteins by modulated differential scanning calorimetry Chuan-He Tang a,∗ , Siu-Mei Choi b , Ching-Yung Ma b a

Department of Food Science and Technology, South China University of Technology, Guangzhou 510640, China b Food Science Lab, Department of Botany, The University of Hong Kong, Hong Kong, China Received 23 April 2006; accepted 13 June 2006 Available online 20 June 2006

Abstract The thermal properties and heat-induced denaturation and aggregation of soy protein isolates (SPI) were studied using modulated differential scanning calorimetry (MDSC). Reversible and non-reversible heat flow signals were separated from the total heat flow signals in the thermograms. In the non-reversible profiles, two major endothermic peaks (at around 100 and 220 ◦ C, respectively) associated with the loss of residual water were identified. In the reversible profiles, an exothermic peak associated with thermal aggregation was observed. Soy proteins denatured to various extents by heat treatments showed different non-reversible and reversible heat flow patterns, especially the exothermic peak. The endothermic or exothermic transition characteristics in both non-reversible and reversible signals were affected by the thermal history of the samples. The enthalpy change of the exothermic (aggregation) peak increased almost linearly with increase in relative humidity (RH) in the range between 8 and 85%. In contrast, the onset temperature of the exotherm decreased progressively with increase in RH. These results suggest that the MDSC technique could be used to study thermal properties and heat-induced denaturation/aggregation of soy proteins at low moisture contents. Associated functional properties such as water holding and hydration property can also be evaluated. © 2006 Elsevier B.V. All rights reserved. Keywords: Modulated differential scanning calorimetry (MDSC); Soy proteins; Thermal property

1. Introduction Differential scanning calorimetry (DSC) has been widely used to characterize the thermal properties of food proteins, including heat-induced denaturation [1–6] and glass transition [7–9]. However, the DSC technique suffers from limitations with respect to its ability to describe complex transitions and individual contributing components. In many cases, in order to characterize the thermal properties of proteins, a high protein concentration or a large volume of the samples are needed (since enthalpic change of protein unfolding is very small), making the experiments difficult to control. The use of a high-sensitivity DSC (HSDSC) is required which is capable of detecting small changes in enthalpy or heat capacity when proteins unfold. However, the cost of such an instrument is high. In previous reports studying the glass transition of proteins by conventional DSC, it is necessary to eliminate the influence of

∗

Corresponding author. Tel.: +86 20 87114262; fax: +86 20 87114263. E-mail address: [email protected] (C.-H. Tang).

0141-8130/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ijbiomac.2006.06.013

the thermal history of samples, and the glass transition temperatures (Tg ) are usually determined in the second run [9]. However, these observed phenomena on the second run may not reflect the real situation occurred during the thermal transition of proteins, since the protein conformation and hydration of proteins are inevitably changed after the first scanning run. This is a major reason for the diversity in the results and interpretation about the thermal events of proteins. Modulated DSC (MDSC) offers a solution of overcoming many of the aforementioned analytical limitations. MDSC is an extension of conventional DSC, in which a sinusoidal wave modulation is applied to the standard linear temperature program. One of the advantages of the MDSC method is its ability to separate the reversing processes from the non-reversible thermal events. This is accomplished by subtracting reversing heat flow signal from the total heat flow signal to obtain the non-reversing heat flow signal. All heat flow associated with processes capable of following the temperature is captured in the reversing heat flow signal, while any remaining heat flow attributable to processes not capable of following the temperature modulation is captured in the non-reversing heat flow signal [10].

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

Although MDSC has been available for more than a decade, its use has mainly been restricted to the study of synthetic polymers and thermoplastics (non-food materials), as well as glass transitions associated with carbohydrates [11]. With the exception of formulations in the pharmaceutical industry [12–14] there was only a limited number of studies on food proteins using MDSC [15–17]. In most of these studies, MDSC analysis was used to characterize the glass transitions, and there was nearly no report of using MDSC to study the thermal properties related to hydration and heat-induced aggregation of food proteins. The aim of this study was to investigate the use of MDSC to characterize soy proteins at low moisture contents, and to study the thermal properties. The effects of heat denaturation, prior thermal treatments, and relative humidity (at which the protein is stored) on these properties will be evaluated, since very little information is available on this aspect of MDSC application to food protein systems. Soy protein isolates (SPI) was chosen for this investigation, firstly because this protein product is widely used in food and non-food formulations as a major ingredient, and secondly because the thermal properties, including glass transition, have been extensively investigated by conventional DSC. 2. Materials and methods 2.1. Materials Whole soybean powder was obtained from Henan Hebei Co. (China). SPI was prepared according to the method of Ref. [18] with minor modification. Defatted soybean meal was prepared from ground soybean powder by solvent extraction with n-hexane. Defatted soybean seed meal was then extracted at room temperature with 20-fold 0.03 mol/L Tris–HCl buffer (pH 8.0) containing 5 mmol/L ␤-mercaptoethanol (2-ME) for more than 2 h, and the resulting dispersion was centrifuged at 8000 × g for 15 min to remove the insoluble material. The pH of the supernatant was adjusted to 4.8 at 4 ◦ C with 2N HCl, and the precipitate or curd was collected by centrifugation (8000 × g, 10 min). The curd was dispersed in water at 4 ◦ C, and the dispersion was adjusted to pH 7.5 with 2N NaOH, then centrifuged to yield the supernatant. The supernatant was dialyzed three times at 4 ◦ C against deionized water (1:100, three times), and then lyophilized. The protein content of the SPI was 97.0% (dry basis), determined by Kjeldahl method (N × 6.25). The salts used for conditioning of the humidity-controlled containers and other chemicals were analytical reagents. 2.2. Preparation of moisture-heated SPI A 100 mL of SPI dispersions in distilled water (2%, w/v) were placed and sealed in 200 mL beakers. Then, the beakers were placed in water bath at 80 or 90 ◦ C for 45 min. After the heat treatments, the beakers were immediately cooled by immersing in an ice bath, and then equilibrated to room temperature. Last, the preheated SPI dispersions were freeze-dried to produce the moisture-heated SPI samples.

97

2.3. Conditioning of the protein samples SPI samples at different relative humidity levels were prepared by conditioning in humidity-controlled hermetic containers at room temperature according to the method of Ref. [9]. Each of these containers held a saturated salt solution, imposing a relative humidity (RH) determined at the head space. The salts used were KOH, MgCl2 , K2 CO3 , Na2 Cr2 O7 , NH4 NO3 , NaCl, (NH4 )2 SO4 , KCl, KNO3 , and K2 SO4 , giving RH of 8, 33, 43, 54, 62, 75, 80, 85, 92 and 97%, respectively. The protein samples were dried at room temperature for more than 2 weeks in a dessicator containing P2 O5 , and then placed in different humidity-controlled containers for 2–3 weeks before DSC analysis. 2.4. DSC measurements The thermal characteristics of SPI samples were assessed using a TA Q100-DSC thermal analyzer (TA Instruments, New Castle, DE, USA) according to the procedure of Ref. [5] with some modification. Approximately 2.0 mg of protein samples were weighed into the aluminum liquid pan (Dupont), and 10 ␮L of 0.05 M phosphate buffer (pH 7.0) was added. The pan was hermetically sealed and equilibrated at 25 ◦ C for more than 6 h, then heated from 25 to 110 ◦ C at a rate of 5 ◦ C/min. A sealed empty pan was used as a reference. Onset temperature (Tm ), peak transition or denaturation temperature (Td ), and enthalpy of denaturation (H), were computed from the thermograms using the Universal Analysis 2000 software, Version 4.1D (TA Instruments). All experiments were conducted in duplicates or triplicates. Some of the thermograms were comprised of over-lapping peaks, and only the H of the combined transitions were measured due to difficulties in accurately estimating the partial areas of individual peaks. 2.5. Modulated DSC measurements Modulated DSC experiments on dry SPI samples (with low moisture contents) were carried out using the same instrument (TA Q100-DSC) in standard temperature-modulation mode. Protein samples (3–4 mg, with 0.1 mg accuracy) were packed down and sealed in aluminum solid pans, and an empty pan was used as reference. The pans were heated at a scan rate of 5 ◦ C/min according the following program: equilibrated at −70 or −50 ◦ C, modulated at ±1.0 ◦ C every 60 s, kept isothermal for 5 min, and then heated to 250 ◦ C. Reversible and non-reversible heat flow signals were separated from the total heat flow signals. The glass transition temperature (Tg ) was determined from the reversing heat flow signal. Each sample was repeated three times and data were presented as means ± S.D. 2.6. Pre-scan treatment The non-treated SPI samples (stored at RH 23%) were sealed in the same aluminum solid pans, heated in the DSC at a scan rate of 5 ◦ C/min from 25 ◦ C to various temperatures (75, 150 and

98

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

Fig. 1. Conventional DSC thermograms of non-treated and preheated SPIs. (a) Non-treated SPI (control); (b) preheated (80 ◦ C, 45 min) SPI; (c) preheated (90 ◦ C, 45 min) SPI. The scanning rate was 5 ◦ C/min, and the protein concentration was about 20% (w/v).

Fig. 2. Comparison of heat flow profile from conventional DSC and total heat flow profile from MDSC scan of non-treated SPI (stored at 50% RH). The SPI samples were sealed in aluminum solid pans.

250 ◦ C), and then cooled immediately to 25 ◦ C. The scanned SPI samples in pans were placed at the room temperature over night to sufficiently reabsorb the released water. Last, these scanned samples were re-scanned at the same heating rate 25 to 250 ◦ C.

ment was carried out at a temperature much lower than the Td of glycinin (94.1 ◦ C), its protein conformation was perturbed to certain extent, and the hydrophobic cores initially buried in the interior could be partially exposed. As the result, the partially dissociated glycinin components would refold to form more stable aggregates with higher Td [22]. Thus, it can be reasonably inferred that there are two types of aggregates in preheated (80 ◦ C for 45 min) SPI, those formed from completely denatured ␤-conglycinin (insoluble) and those formed from partially denatured glycinin (soluble). In contrast, both glycinin and ␤-conglycinin were completely denatured after thermal pretreatment at a temperature (90 ◦ C) close to the Td of glycinin (Fig. 1c). In this case, the aggregates seem to be simultaneously composed of completely denatured glycinin and ␤-conglycinin.

3. Results and discussion 3.1. Thermal characteristics of SPI determined by conventional DSC Fig. 1 shows the conventional DSC thermograms of native and preheated SPI, and the DSC characteristics (including the Td of glycinin and ␤-conglycinin components and total H) are summarized in Table 1. The two major endothermic peaks with Td of 76.7 and 94.1 ◦ C were attributed to ␤-conglycinin and glycinin components, respectively, and are consistent with previous reports [19–21]. After thermal pretreatment at 80 ◦ C for 45 min, the endothermic peak of ␤-conglycinin disappeared (Fig. 1b), indicating complete denaturation of the ␤-conglycinin component. The H of the glycinin component declined slightly by 15% after this thermal treatment, and the Td was increased from 94.1 to 97.6 ◦ C (Table 1). The results suggest that although this pretreat-

3.2. Assignment of the signals of MDSC thermograms Fig. 2 shows the heat flow profiles of non-treated SPI (stored at 50% RH), measured by conventional and modulated DSC. The two profiles were found to be in good agreement, showing one major and one minor endothermic peak. Ideally, the total MDSC heat flow profile should be identical to the conventional

Table 1 Summary of DSC characteristics of denaturation and MDSC characteristics of non-treated and preheated SPI at 50% RH Soy protein samples

Td1 b (◦ C)

Non-treated (control) Wet-heated (80 ◦ C, 45 min) Wet-heated (90 ◦ C, 45 min)

MDSC characteristicsa

DSC characteristics

76.7 –d –

Td2 (◦ C)

94.1 97.6 –

Hc (J/g)

15.4 7.2 0.0

Reversible heat flow

Non-reversible heat flow

To

Ta

Ha

T1

H1

T2

H2

72.7 74.3 70.8

156.1 – 136.9

40.0 – 39.9

103.0 125.9 97.6

271.3 247.8e 268.6

224.5 204.0 211.9

25.3 10.9

The data are presented as the means of duplicate measurements. a T , T and H indicate the on-set, peak temperatures, and the enthalpy change respectively of the exothermic peak in the reversible signal. T , T , H , and o a a 1 2 1 H2 indicate the peak temperatures and the enthalpy changes of first and second endotherms in the non-reversible signal, respectively. b T and T indicate the peak temperatures of endotherm of ␤-conglycinin and glycinin components, respectively. d1 d2 c Total enthalpy changes of glycinin and ␤-conglycinin components. d No observable endothermic or exothermic peak. e Total enthalpy change of both overlapping endothermic peak.

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

DSC heat flow profile. However, Td of the major peak measured by MDSC (98.6 ◦ C) was markedly lower than that measured by conventional DSC (115 ◦ C). The discrepancy could be due to the choice of modulation period, since insufficient cycles through the peak can be detrimental to the deconvolution process [11]. The major endothermic peak (from 0 to 180 ◦ C) observed by other researchers in soy proteins, gelatin, sodium caseinate, and corn gluten meal [21,23–25], has been attributed to the loss of residual water [24] or hydrogen bond disruption within the protein molecules [23]. The minor endotherm (from 180 to 240 ◦ C) seems to be associated with the loss of the most stable ‘immobilized water’, which may reflect molecular mobility of proteins. A third minor peak (about 52 ◦ C) was observed only in the conventional DSC heat flow profile (Fig. 2). Cuq and LcardVerni`ere [16] observed a similar endothermic peak (between 41 and 54 ◦ C) in the initial scan of the total heat flow thermogram of semolina (at 11% moisture content). In sunflower proteins, it was also observed that the DSC thermograms of samples at 6–17% RH showed an endothermic peak at about 60 ◦ C in the first scan [9]. This endothermic event is irreversible (absent in the second scan), and considered to result from the breakdown of weak energy interactions of protein, such as denaturation and secondary molecular relaxation [9,16]. Fig. 3 shows a typical MDSC analysis of non-treated SPI (at 50% RH) with total, reversing and non-reversing heat flow signals. The major endothermic peak associated with loss of residual water (from 0 to 180 ◦ C) and the relaxation endotherm (from 180 to 240 ◦ C) were well separated in the non-reversing signal. The enthalpy change of the first endotherm (H1 ) in the non-reversible signal should be closely associated with the loss of residual water, since enthalpy changes due to protein denaturation or relaxation are much less than that caused by disruption of hydrogen bonds or loss of residual water. Based on this assumption, H1 value may reflect water holding capability of the protein sample at a constant RH. Similarly, enthalpy

99

change of the second endotherm (H2 ) is related to inherent hydration ability of the protein sample. Thus, at a constant RH, a protein with high H1 and H2 values may present strong water-holding capability or hydration ability. Besides, the interactions between residual water and protein molecules may affect the loss of residual water during the heating, and the proteins with stronger hydrogen bond interactions with water will release the residual water at higher temperature during the heating. Thus, the endothermic peak temperature (Td1 ) is indicative of the extent of hydrogen bond interactions between protein and water. In the reversible heat flow signal, a distinct exothermic peak with the onset temperature (To ) of 72.9 ◦ C and peak temperature (Ta ) of 150 ◦ C was observed (Fig. 3). Interestingly, the To value of this exothermic peak is slightly higher than that (56 ◦ C) of the endothermic peak attributed to loss of residual water. This suggests that the exothermic event may be associated with the disruption of hydrogen bonds between residual water and proteins during heating. The native conformation of a protein is mainly maintained by hydrogen bonds and electrostatic interactions between water and the protein molecules, whereas thermal stability is closely related to hydrophobic interactions. If the hydrophilic interactions maintaining the tertiary structure of the proteins are ruptured by heating, hydrophobic regions initially buried in the interior of proteins will be exposed to the protein surface, and associate with hydrophobic regions in other protein molecules to form aggregates. Thus, the exothermic peak in the reversible signal can be attributed to protein aggregation which has been shown to be an exothermic event [1,26]. 3.3. MDSC analysis of non-treated and wet-heated SPI Fig. 4 shows the reversible and non-reversible heat flow profiles of non-treated and preheated SPI samples at 50% and 23%

Fig. 3. A typical MDSC scan showing total, reversing and non-reversing heat flow signals of non-treated SPI (at 50% RH). H1 and H2 are enthalpy changes of the two endothermic peaks in the non-reversible heat flow curve.

100

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

RH, respectively, and some thermal characteristics of the two profiles are summarized in Table 1. As expected, the heat flow signals of preheated SPI samples were different from that of nontreated SPI. At 50% RH, preheated (80 ◦ C, 45 min) SPI showed two major endothermic peaks in the non-reversible heat flow profile with Td of about 125 and 200 ◦ C, respectively (Fig. 4A and Table 1). The marked increase in the Td1 and appearance of a new endotherm at 200 ◦ C suggest that this thermal treatment dramatically promoted the hydrogen bond interactions between water and protein. The presence of different forms of protein aggregates, caused by selective denaturation of glycinin and ␤-conglycinin components at 80 ◦ C (Fig. 1), may account for this improvement. In contrast, the non-reversible heat flow profile of preheated (90 ◦ C, 45 min) SPI was almost identical to that of non-treated SPI, except for the relaxation endotherm (200–240 ◦ C) (Fig. 4A). In addition, the combined enthalpy change of the endothermic peaks was slightly different for nontreated and preheated SPI samples (Table 1), indicating the water

holding capability was kept almost constant before and after the thermal treatment. Similarly, the reversible heat flow profile of preheated (80 ◦ C, 45 min) SPI was different from that of the non-treated one (Fig. 4B). In this SPI, the exothermic peak attributed to thermal aggregation was not obvious. This difference may be attributed to much higher Td1 and even Td2 of the endothermic peaks for this SPI compared to that of non-treated one (Fig. 4A), since the thermal aggregation only occurs after the disruption of hydrogen bond or other hydrophilic interactions. The preheated (90 ◦ C, 45 min) SPI exhibited an exothermic peak with To , Ha and Ta values similar to that of the non-treated one (Fig. 4B). From these results, it is suggested that the thermal aggregation of SPI (at low moisture) be mainly associated with the extent of hydrophilic interactions (particularly the hydrogen bond interactions) between the proteins and water. Actually, the proteins in the preheated SPI samples are mainly in the aggregate forms, insoluble or soluble. In the preheated (80 ◦ C, 45 min) SPI, insol-

Fig. 4. Non-reversible heat flow signals (A and C) and reversible heat flow signals (B and D) of non-treated and preheated SPIs at 50% RH (A and B) and 23% RH (C and D). The preheated SPI samples were prepared after the moisture-preheating at 80 or 90 ◦ C for 45 min.

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

101

Fig. 4. (Continued ).

uble and soluble aggregates may be present together, while the aggregates in the preheated (90 ◦ C, 45 min) SPI are insoluble. The interactions maintaining the structure of these initial aggregates are different, and dependent upon the denaturation extent of its protein components. Thus, the thermal aggregation of SPI can be controlled by adjusting the denaturation extent of its protein components. The non-reversible and reversible heat flow profiles of different SPI samples equilibrated at 23% RH were also examined. At this RH, the H ratio of the first endotherm to the second endotherm in non-reversible signals decreased distinctly when compared to that at 50% RH (Fig. 4A and C). This was expected since only the proportion of free water with high water activity (aw ) in the protein samples could be affected by the decline in RH. At 23% RH, the To of the exothermic peak in the reversible heat flow signal of non-treated SPI was observed at a higher temperature (about 90 ◦ C) than at 50% RH (75 ◦ C) (Fig. 4B and D). This is consistent with the fact that thermal stability of

native soy proteins is higher at lower RH. In contrast, the onset of the exothermic peak in preheated (90 ◦ C, 45 min) SPI was nearly unaffected by the decrease in RH (Fig. 4B and D). These data suggest that the aggregation (an exothermic event) of nontreated SPI occurred only after complete protein denaturation. In addition, the appearance of a minor endothermic peak (at about 135 ◦ C) (Fig. 4D) also suggests that, the absence of an exothermic peak in preheated (80 ◦ C, 45 min) SPI is partially attributed to the counteracting effect of endothermic event, possibly associated with the disruption of initial aggregates. In a previous study on the influence of cross-linking on glass transition of native SPI (stored at about 15% RH), two glass transition peaks around 46 and 180 ◦ C respectively were identified in the reversible heat flow profile [21]. In the present study, glass transition peaks were also observed in the reversible heat flow signals from both non-treated and preheated SPI samples, more obviously at 50% RH (Fig. 4B and D). However, there were distinct differences among SPI samples subjected to different level of preheat treatments. For example, two peaks with

102

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

glass transition temperatures (Tg ) of −26 and 50 ◦ C, respectively, were observed in preheated (80 ◦ C, 45 min) SPI, while for both non-treated and preheated (90 ◦ C, 45 min) SPI, only one glass transition was observed at around 50 ◦ C (Fig. 4B). The differences may be attributed to the differences in the size of protein or aggregates, as well as the plasticizing effect of water. The present results show that the extent of protein denaturation is also an important factor affecting glass transitions of soy proteins. 3.4. Effect of first-scan Fig. 5 shows the non-reversible and reversible heat flow signals of SPI samples (at 23% RH), which had been initially scanned to various temperatures (75, 150 and 250 ◦ C). The H of endothermic peak associated with the loss of residual water for the non-treated SPI was unaffected by initial heating to 75 ◦ C, but markedly decreased after initial heating to 150 ◦ C (Fig. 5A). This result suggests that the water holding capability of SPI decrease with increasing the extent of thermal (dry heat) treat-

ment. The apparent relaxation endotherm corresponding to the loss of ‘immobilized’ water was eliminated by initial heating to 250 ◦ C, suggesting that the inherent hydration property or the molecular mobility of SPI only be changed by heating at high temperature. In the reversible signals, the H and To of the exotherm gradually decreased with increasing the final temperature of initial scan in the range from 75 to 250 ◦ C (Fig. 5B). The data suggest that the thermal aggregation of SPI with low moisture is influenced by the thermal history of the samples, depending on the extent of protein denaturation. 3.5. Effect of relative humidity RH at which a protein is stored can affect its heat-induced denaturation. Proteins stored at high RH have high moisture content and low thermal stability, and the proteins would denature and aggregate more easily. Fig. 6 shows some thermal characteristics of the exotherm (Ha and To ), calculated from the reversible heat flow profiles of native SPI, as a function of

Fig. 5. The non-reversible (A) and reversible (B) heat flow signals of non-treated SPI (at 23% RH), dry-heated to various temperatures in the first scan.

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

103

Fig. 6. Effect of relative humidity on enthalpy change (A) and on-set temperature (B) of non-treated SPI.

Fig. 7. Reversible heat flow signals of non-treated SPI at high RH conditions (75–97%).

RH. At RH less than 75%, the Ha increased almost linearly when RH was increased from 8 to 75% (Fig. 6A), while the To was negatively related to RH (Fig. 6B). These results further confirmed that thermal stability of SPI is closely associated with RH at which the protein is stored, and heat-induced denaturation and aggregation of SPI occur more readily at higher RH. At RH higher than 80%, Ha value declined with increasing RH (Fig. 6A), probably due to an endothermic event which was observed in the reversible heat flow profiles at RH higher than 92% (Fig. 7). The appearance of the endothermic peak suggests that at high RH, protein may undergo two phases of denaturation and aggregation. In the first phase, the small size aggregates are formed at relatively low temperatures (e.g., 110 ◦ C), and the protein conformation is still stabilized by residual water. In the second phase, with further increase in heating temperature, hydrogen bonds between initially formed aggregates and water are disrupted. This event is endothermic, thus, resulting in the appearance of an endotherm. Further confirmation of this mechanism will be required.

4. Conclusions Modulated DSC has been successfully applied to study the thermal properties of freeze-dried soy protein samples. The nonreversible heat flow signals may be used to evaluate water holding capability and inherent hydration properties of soy proteins, while the reversible signals are indicative of the heat-induced aggregation and the glass transitions of protein. MDSC analysis showed that soy proteins with different extent of denaturation exhibited different patterns of reversible and non-reversible heat flow profiles. The thermal characteristics of the aggregation exotherm were influenced by the thermal history of the protein samples and the relative humidity at which the proteins were stored. Acknowledgments This work is part of the research projects of Chinese National Natural Science Fund (serial numbers: 20306008 and 20436020), sponsored by the NSFC. The author also gratefully

104

C.-H. Tang et al. / International Journal of Biological Macromolecules 40 (2007) 96–104

acknowledges the financial support from the Natural Science Fund of Guangdong Province, China (serial number: 05006525). References [1] S.D. Arntfield, E.D. Murray, Can. Inst. Food Sci. Technol. J. 14 (1981) 289–294. [2] V.R. Harwalkar, C.-Y. Ma, J. Food Sci. 52 (1987) 394–398. [3] C.-Y. Ma, V.R. Harwalkar, J. Food Sci. 53 (1988) 531–534. [4] A.A. Scilingo, M.C. A˜no´ n, J. Agric. Food Chem. 44 (1996) 3751–3756. [5] G.-T. Meng, C.-Y. Ma, Food Chem. 23 (2001) 453–460. [6] S.W. Ellepola, C.-Y. Ma, Food Res. Int. 39 (2006) 257–264. [7] A. Morales, J. Kokini, Biotechnol. Progr. 13 (1997) 624–629. [8] F. Castelli, S.M. Gilbert, S. Caruso, D.E. Maccarrone, S. Fisichella, Thermochim. Acta 346 (2000) 153–160. [9] A. Rouilly, O. Orliac, F. Silvestre, L. Rigal, Polymer 42 (2001) 10111–10117. [10] Z.P. Lu, Y. Li, S.C. Ng, Y.P. Feng, Thermochim. Acta 357/358 (2000) 65–69. [11] V.L. Hill, D.Q.M. Craig, L.C. Feely, Int. J. Pharmaceut. 161 (1998) 95–107.

[12] S.R. Rabel, J.A. Jona, M.B. Maurin, J. Pharmaceut. Biomed. Anal. 21 (1999) 339–345. [13] S. Nazzal, Y. Wang, Int. J. Pharmaceut. 230 (2001) 35–45. [14] A. Badkar, P. Yohannes, A. Banga, Int. J. Pharmaceut. 309 (2006) 146– 156. [15] V. Micard, S. Guilbert, Int. J. Biol. Macromol. 27 (2000) 229–236. [16] B. Cuq, C. Lcard-Verni`ere, J. Cereal Sci. 33 (2001) 213–221. [17] V. Micard, M.-H. Morel, J. Bonicel, S. Guilbert, Polymer 42 (2001) 477–485. [18] S. Iwabuchi, F. Yamauchi, J. Agric. Food Chem. 35 (1987) 200–205. [19] S. Damodaran, J. Agric. Food Chem. 36 (1988) 262–269. [20] A.M. Hermansson, J. Texture Study 9 (1978) 33–58. [21] C.H. Tang, Z. Chen, L. Li, X.Q. Yang, Food Res. Int. 39 (2006) 704– 711. [22] C.C. Bigelow, J. Theor. Biol. 16 (1967) 187–211. [23] L.N. Bell, D.E. Touma, J. Food Sci. 61 (1996), pp. 807–810, 828. [24] L.D. Gioia, B. Cuq, S. Guilbert, Int. J. Biol. Macromol. 24 (1999) 341– 350. [25] C.H. Tang, X.Q. Yang, Z. Chen, H. Wu, Z.Y. Peng, J. Food Biochem. 29 (2005) 402–421. [26] P.L. Privalov, Adv. Protein Chem. 35 (1982) 1–104.