As shown in Figure 2a, the content of free amino acid of HAHp was decreased from 4.72 to 1.91 mg/mL (P < 0.05) after heat sterilization (121 °C, 30 min). In addition, the reducing sugar content of HAHp-H was reduced to 119.3 mg/mL, statistically lower than that of HAHp (150.2 mg/mL) (Figure 2b). During thermal processing, peptides/free amino acids in hydrolysates interacted directly with reducing sugar (cross-linking) to form larger molecules, resulted in the reduced free amino acid and reducing sugar contents. On the other hand, higher molecular weight of peptides could degrade to smaller peptides or free amino acids [2,16]. The balance between peptides/amino acids cross-linking and peptides degradation determined the free amino acids content in the heated products. The result of Figure 2 indicated that reducing sugar or its degraded products attached to free amino acids (cross-linking) played an important role in HAHp-H.

It was observed that the sterilization product of HAHp had slightly higher total amino acid content compared with HAHp, this result was similar to the observations of Oluwaniyi et al. [17] that the processes of boiling (100 °C, 10 min) and roasting (145 °C, 15 min) increased the total amino acid contents of four marine fishes including herring, Atlantic mackerel, horse mackerel and white hake. As seen in Table 2, cysteine was lost completely in HAHp-H. Additionally, the content of lysine was decreased from 0.961 to 0.814 mg/100 mg. Since cysteine has the ability to form a dehydroalanyl residue and bind with the ɛ-amino group of lysine in the Maillard reaction [18], result of Table 2 implied that cysteine and losses of lysine under the heat treatment probably contribute to form cysteine-derived crosslink compounds, e.g., lysinoalanine. The levels of amino acid, i.e., valine, leucine, phenylalanine, histidine and arginine, were all increased. In particular, the histidine level of HAHp-H was dramatically increased from 0.294 to 1.368 mg/100 mg. An interesting finding was that these amino acid levels, except for the losses of cysteine and lysine, were reported to be increased or damaged in different protein, peptide, or free amino acid/sugar heating systems [4,17]. We conjectured that differences in substrates and heating conditions determined amino acid composition in the thermal products.

Table 2, clearly shows that the contents of alanine, valine, leucine, arginine and histidine amino acids accounting for the presence of antioxidant activity [19], were slightly or dramatically increased in HAHp-H. A higher ratio of hydrophobic amino acids was also reported to improve the antioxidant activities of the Maillard reaction products (MPRs) [19,20]. Furthermore, the hydrophobic amino acid residues were important for the formation of hydrophobic tail in the COOH-terminal region, which exhibited a predominant role for anticancer peptides to mediate their cytotoxic effect [21]. The hydrophobic amino acids content of HAHp-H were increased to 8.947 mg/100 mg with a relative percent of 48.74%, indicating that HAHp-H should display higher antioxidant ability and antiproliferative activity than HAHp. Results shown in Figure 1 and Table 1 confirmed this viewpoint.

The UV-visible spectra of HAHp and HAHp-H showed a similar contour. However, the absorption values of HAHp-H were generally higher than that of HAHp under same wavelength (Figure 3a). Sugar caramelization might occur in addition to the Maillard reaction when heated at temperatures above 120 °C or 9 < pH < 3, leading to a browning of the mixture [1,22]. However, provided amino compounds such as amino acids, peptides and proteins were present, the Maillard reaction usually took place [22]. In Figure 3a, HAHp-H had a maximum absorbance appeared in the range of 260–320 nm, which was characteristic of melanoidins [23,24]. The absorbance under 420 nm (browning intensity) was often used as an indicator for browning development in the Maillard reaction [25]. As can be seen in Figure 3b, the browning intensity of HAHp-H was increased to 1.386, significantly higher than that of HAHp (0.081) (P < 0.05), suggesting that the Maillard reaction might play a predominant role in HAHp-H. The results of Figure 3 were in accordance with the decreased free amino acid and reducing sugar contents in HAHp-H (seen in Figure 2).

The vibration regions of the amide bonds of proteins and the chemical fingerprints of carbohydrates were readily identified in FT-IR spectroscopy [26,27]. The most distinctive spectral features for proteins were amide I band at 1600–1700 cm⁻¹ (C═O stretching), amide II band at 1500–1550 cm⁻¹ (N–H deformation) and amide III band at 1200–1300 cm⁻¹ (C–N stretching and N–H deformation). Normally, the amide I band was strong, the amide II band was weak and the amide III band was moderate [26]. A series of overlapping peaks located in the region of 1180–953 cm⁻¹, which were described as the “saccharide” bands resulting from vibration modes such as the stretchings of C–C and C–O and the bending mode of C–H bonds. These absorptions were weak in the spectra of most proteins [27].

As can be seen in Figure 4, the bands of amide I in HAHp and HAHp-H were located at 1659.06 cm⁻¹ and 1645.10 cm⁻¹, respectively. The result meant that the amide I band of HAHp-H was modified by thermal sterilization, meanwhile, the result of Figure 4 also indicated that both HAHp and HAHp-H might contain the second structure of alpha-helical, which has a peak maximum around 1660–1650 cm⁻¹ [28]. The IR spectral features of HAHp and HAHp-H between 1500 cm⁻¹ and 1550 cm⁻¹ showed a clear change and the band intensity of HAHp-H at 1540.87 cm⁻¹ was decreased dramatically, indicating that heat sterilization resulted in a big conformation change of amide II band (N–H deformation). As for amide III bands, which were very complex in proteins, there was slight variation from the range of 1200 cm⁻¹ to 1300 cm⁻¹ region (Figure 4). Nevertheless, the FT-IR spectroscopy would not be sufficiently sensitive to distinguish such small changes.

In the IR region of 1180–953 cm⁻¹, which associated with “saccharide” bands, the absorbances of HAHp and HAHp-H were not weak, implying that saccharides were contained in HAHp and HAHp-H. The result of FT-IR could contribute to explain the reason for MRPs formation in HAHp-H even without addition of sugars. In the mid-infrared spectrum of MRPs, several chemical structures were changed. For example, the group of NH₂, especially from lysine, was lost. While the amount of groups, including the Amadori compound (C═O), Schiff base (C═N) and pyrazines (C═N) associated with Maillard products might be increased [4].

The MW profiles of HAHp and HAHp-H were displayed in Table 3. HAHp was a mixture of small peptides and the percent of peptides below 3000 Da was above 95%, while in HAHp-H it decreased to 60.77%. The decreased fractions were within the ranges of 1000–3000 Da and 500–1000 Da. However, the higher MW of 3000–5000 Da and lower MW below 500 Da in HAHp-H were increased to 38.06% and 19.00%, respectively (P < 0.05). The result was similar to the study of Liu et al. [3]. Because of the presence of peptides/amino acids cross-linking in the Maillard reaction, larger molecules tended to form, while peptides/proteins degradation resulted in the increased smaller peptides or free amino acids. High MW and low MW components in MRPs were accountable for the enhanced antioxidant activities [29], and hydroxyl groups of MRPs were effective electron donors to increase reducing power [30].

The free radical scavenging activity was determined by the 1,1-diphenyl-2-picryl-hydrazil (DPPH) method with few modifications [34]. Briefly, 375 μL of 0.02% DPPH solution in 99.5% ethanol was added to 1.5 mL of HAHp and 1.5 mL of 99.5% ethanol, then mixed thoroughly and kept in the dark at room temperature for 60 min. The absorbance of the mixtures was measured at 517 nm. HAHp and butylated hydroxytoluene (BHT) were tested in the same way for comparison. DPPH radical scavenging activity was calculated as follows: Radical   scavenging   activity   ( % ) = [ ( control − sample + blank ) / control ] × 100where control, replaced 1.5 mL of sample with 1.5 mL of distilled; blank, replaced 375 μL of 0.02% DPPH with 375 μL of 99.5% ethanol.

The presence of reductants of samples could result in reducing Fe³⁺/ferricyanide complex to the ferrous form (Fe²⁺), and the Fe²⁺ could therefore be monitored by measuring the formation of Perl’s Prussian blue at 700 nm. A higher absorbance at 700 nm meant a better reducing power [30]. The reducing power of HAHp-H was measured according to the method of Oyaizu [35] with slight modifications. 1.0 mL of HAHp-H was mixed with 1.0 mL of 0.2 M sodium phosphate buffer (pH 6.6) and 1.0 mL of 1% potassium ferricyanide (K₃Fe(CN₆)). The mixture was incubated at 50 °C for 20 min, followed by addition of 1.0 mL of 10% trichloroacetic acid. Then, the mixture was centrifuged at 3000 g for 10 min and 2.0 mL of supernatant was blended with 2.0 mL of deionized water and 0.8 mL of 0.1% ferric chloride (FeCl₃). The absorbance of the mixture was measured at 700 nm. The reducing power of HAHp and BHT were also determined for comparison.

Three human cancer cell lines, DU-145 prostate cancer cell line, 1299 lung cancer cell line and 109 esophagus cancer cell, were provided by College of Food Science and Pharmacy, Zhejiang Ocean University and cultured in RPMI 1640 medium at 37 °C in a humidified atmosphere with 5% CO₂ incubator (ThermoForma, USA), supplemented with 10% heat-inactivated fetal calf serum, 200 mM l-glutamine, 100 units/mL of penicillin, and 100 units/mL streptomycin.

The inhibition rate of cancer cell lines after treatment with HAHp and HAHp-H was determined by the MTT method [36]. Cancer cell lines were transferred to fresh media every 2–3 days to maintain in an exponential growth and digested by 0.25% trypsin to obtain cell suspension. Then, 200 μL of cell suspension (1 × 10⁴ cells/mL) was transferred into a 96 wells plate and incubated for 24 h. After this period, the suspension was removed and 200 μL of the sample (HAHp and HAHp-H, concentration ranging from 5 mg/mL to 40 mg/mL) was added, followed by incubation for 48 h at 37 °C. Then, 200 μL of MTT (1 mg/mL) was added to these cells respectively and incubated for 4 h at 37 °C. The medium was removed and 150 μL of DMSO was added to dissolve the formazan crystals formed and precipitated. Finally, the absorbance was read at 490 nm with an ELISA reader (BR680, USA). The inhibition rate was calculated using the following equation: Inhibition   rate   ( % ) = [ ( A C − A S ) / ( A C − A B ) ] × 100where A_C, absorbance of control, using 200 μL of RPMI 1640 complete medium instead of 200 μL of HAHp-H; A_S, absorbance of sample; A_B, absorbance of RPMI 1640 complete medium.

The content of free amino acids was determined by the method of ninhydrin reaction [37]. 0.2 mL of sample was mixed with 3.8 mL of distilled water, 1.0 mL of ninhydrin and 1.0 mL of 0.2125 mol/L sodium phosphate buffer (pH 8.2) in a 25 mL of cylinder. The solution was heated at 100 °C for 15 min and cooled down to room temperature quickly. Then, the volume of the mixture was adjusted to 25 mL by addition of distilled water. After being kept at room temperature for 15 min, the absorbance of the mixture was measured at 570 nm. A calibration curve was obtained by using l-alanine as a standard. The content of free amino acids was expressed as mg per mL of sample.

Reducing sugar levels of HAHp and HAHp-H were determined by the Fehling method with some modifications [38]. Briefly, HAHp and HAHp were diluted to 100 mL after removal of proteins by addition of zinc acetate and potassium ferrocyanide solution. 5 mL of Fehling solution A (15 g copper sulfate and 0.05 g methylene blue dissolved in distilled water), 5 mL of Fehling solution B (50 g sodium potassium tartrate tetrahydrate, 75 g sodium hydroxide and 4 g potassium ferrocyanide dissolved in distilled water) and 10 mL of distilled water were added into a 250 mL of conical flask. The mixture was heated to boilling in 2 min, followed by titration with HAHp and HAHp to reach bright yellow colour. The consumed volumes of HAHp and HAHp-H were recorded, respectively. HAHp and HAHp-H were replaced by glucose, which was used as the standard of reducing sugar, to standardize the mixture of Fehling solution A and B under the same condition. The content of reducing sugar was expressed as mg per mL of sample.

HAHp and HAHp-H powders were respectively dissloved in 6 mol/L HCl and hydrolyzed at 110 °C for 24 h. After removal of HCl under vacuum, the dried samples were then redissolved in 0.1 M sodium citrate buffer (pH 2.2). Subsequent amino acid analysis was carried out in an automatic amino acid analyzer (Hitachi 835-50, Tokyo, Japan). Results were determined as mg per 100 mg of sample.

UV-visible spectra of HAHp and HAHp-H were recorded by a Shimadzu 1200UV-vis spectrophotometer (Tokyo, Japan) with the wavelength ranging from 200 to 600 nm. The browning intensity was measured at 420 nm [39].

All IR spectra of HAHp and HAHp-H were detected using an FT-IR spectrometer (NICOLET NEXUS670, DTGS). HAHp and HAHp-H powders were mixed with potassium bromide (KBr) powder separately. Scanning was carried out in the range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹ and 32 scans were calculated to obtain an average for each sample.

The MW distribution profiles of HAHp and HAHp-H were estimated by gel permeation chromatography (GPC) using an ÄKTA purifier system, equipped with a Superdex-75 HR 10/300 column from Amersham Pharmacia, Uppsala, Sweden. The elution consisting of 50 mM sodium phosphate buffer (pH 5.8 containing 0.15 mol/L NaCl) was delivered at a flow rate of 0.4 mL/min. 100 μL of sample (filtered by a 0.22 μm mircofiltre) was injected onto the column. Chromatograms were recorded by UV detection at 280 nm. Four MW standards, including bovine serum albumin (68,000 Da), aprotinin (6512 Da), vitamin B₁₂ (1355 Da) and oxidized glutathinone (613 Da), were used to obtain a MW calibration curve. The data analysis was performed using gel permeation chromatography software.

Data were represented as mean ± standard deviation. The significant differences (P < 0.05) between samples were evaluated through the student’s t-test.