A model system based on glucose–arginine to monitor the properties of Maillard reaction products

Abstract

An arginine–glucose mixture (1:2 mol ratio) with pH 9 was heated at 53–100 °C for 10–350 min, and the effects of process parameters were determined during the Maillard reaction (MR). The heating temperature and time were selected as process conditions and were studied with central composite design. The model system was tested based on the values obtained from antioxidant capacity, browning intensity, pH, acrylamide (AC), and hydroxymethylfurfural (HMF) concentrations. Higher temperatures and longer time resulted in higher antioxidant capacity and browning intensity while lowering the pH values. HMF concentration of MR products was found higher in lower temperatures with longer processing time, whereas AC concentration was found higher in high temperatures.

Introduction

The Maillard reaction (MR) is a non-enzymatic reaction occurring between reducing sugars and amino groups (Cha et al. 2019). The Amadori products are formed after the reaction between amino groups and reducing sugars, which then progress to enolization. This reaction series produces an attractive color, aroma, and flavor in foods, which are known as essential quality parameters. However, over-processing may degrade specific vitamins, influence trace element metabolism, and cause loss of nutritional value (essential amino acids) (Rufian-Henares and De la Cueva 2008). Besides, the other important issue about MR products (MRPs) is the formation of toxic compounds during the MR such as acrylamide (AC), known carcinogen, and hydroxymethylfurfural (HMF), having an adverse effect on health at very high concentrations (Mogol and Gökmen 2016). HMF is also used as an indication for thermal processes, leading to being formed other harmful constituents (Kukurová et al. 2013).

Although the MRPs have been reported as pro-oxidant and carcinogens, these compounds also showed scavenging activity on free radicals occurring in foods (Liu et al. 2020). Various mechanisms could be responsible for the antioxidant capacity of MRPs, such as free radical scavengers, metal chelators, radical chain disruptors, and hydrogen peroxide reducers (Gu et al. 2009). The biological activities of MRPs were extensively studied in various model systems containing d-allose–α-lactalbumin (Sun et al. 2006a); d-aldohexoses-ovalbumin (Sun et al. 2006b); glucose–casein (Gu et al. 2009); fructose/ribose–lysine (Vhangani and Van Wyk 2013); chitosan–glucose/fructose/maltose/lactose (Phisut and Jiraporn 2013); sea cucumber gut hydrolysates–ribose (Han et al. 2017); glucose–glycine model (Yu et al. 2018); xylose–bovine casein hydrolysate (Chen et al. 2019); histidine–fructose (Liu et al. 2019); lysine/arginine/histidine–ribose (Hemmler et al. 2019); glucose–lysine (Yu et al. 2020). Non-enzymatic glycation reactions in foods are strongly related to lysine and arginine, which are the main contributors. Nuts, seeds, legume, seaweed, oat, buckwheat, brown rice, and chocolate are one of the high-arginine food products. Thermally processed arginine rich-foods such as roasting nuts with or without sugar expose to MR (Losso 2016). Therefore, in this study, the combination of glucose and arginine was chosen to model the MR.

Response surface methodology (RSM) allows to reduce large amounts of experimental data and to evaluate their interactions for statistically acceptable results. RSM could be used to assess the experimental data with the aim of the optimization of the processing parameters. The formation of MR depends on types of reaction products and conditions such as molar concentration, pH, temperature, water activity, and time. To control the detrimental effect of potentially harmful compounds and the positive impact of MRPs, the processing conditions required to be determined quantitatively (Knol et al. 2010). RSM would make it possible to predict the formation of AC and HMF in the highest concentration as well as giving insight into how time–temperature combination influence the desired taste, aroma. There were very few studies relating RSM application to the antioxidant activity of MRPs (Gu et al. 2009), while some of them are about to evaluate how to process parameters affect only one response (De Vleeschouwer et al. 2008).

Due to the potential toxicity of heat-induced constituents, AC and HMF, formed during the MR, efforts have been carried out to find the production mechanism and strategies to minimize their formation in food processing by using model systems. The optimum processing conditions were determined using RSM for the achievement of MRPs with maximum level of antioxidant properties and minimum browning intensity (BI) values, HMF, and AC concentration. Particularly, thermally processed foods have high AC and HMF levels; thus, a model based on arginine–glucose will be useful to determine the process conditions that MRPs are increasing progressively. Therefore, this study was aimed to assess the effect of heating time and temperature on the antioxidant properties, BI, HMF, and AC contents in the arginine–glucose model system.

Materials and methods

Materials

l-Arginine monohydrate and D- (+)-Glucose were supplied from Sigma Aldrich (St. Louis, Missouri, USA). 5-(Hydroxymethyl)furfural (HMF), acrylamide (AC), 1,1-diphenyl-2-picryl-hydrazyl (DPPH), 2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), 6-hydroxy-2,5,7,8 tetramethylchromon-2-carboxylic acid (Trolox), 2,2′-Azobis (2-amidinopropane) dihydrochloride (AAPH), fluorescein, potassium persulfate, sodium phosphate, ethanol, methanol, and acetonitrile were also purchased from Sigma Aldrich (St. Louis, Missouri, USA).

Preparation of Maillard reaction products (MRPs)

An arginine–glucose mixture (1:2 M ratio) was prepared to obtain a Maillard reaction model, following the method of Vhangani and Van Wyk (2013) with slight modifications. Glucose (1.4 M) and arginine (0.7 M) were dissolved by using a carbonate buffer (pH 9). Different molar ratios were tested for glucose–amino acid model systems in the literature. In this study, the same molar ratio of glucose and arginine with the study of Wedzicha et al. (1994) was used to see the role of glucose in reaction. Equal volumes of glucose and arginine solutions were transferred into sealed bottles to obtain MRPs. All solutions were heated at predetermined temperatures (53–100 °C) for each time interval (10–350 min). The pH of the resultant MRPs was measured (Schott Instruments Lab8, Mainz, Germany) and then were immediately cooled down to store at − 18 °C until analysis.

Browning intensity of MRPs

The browning intensity (BI) of MRPs was determined spectrophotometrically. Each sample was suitably diluted, and the absorbance values were recorded at 420 nm (Shimadzu, UV-1601, Tokyo, Japan).

The antioxidant activity of MRPs

The antioxidant activity of MRPs was determined by three different tests to understand the antioxidant ability of MRPs. The selected methods were based on H-transfer such as oxygen radical absorbance capacity (ORAC), which gives an activity related to a physiological significant oxidizable substrate and mixed mechanisms such as the Trolox equivalent antioxidant capacity (TEAC) and the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging ability, which give an efficacy against quenching generated free radicals.

The DPPH radical scavenging ability of MRPs was determined by the method described by Sánchez-Moreno (2002). Briefly, a suitable dilution of each MRP solution was treated with 0.1 mM DPPH solution, and all test tubes were left in the dark for 40 min before absorbance measurement at 517 nm (Shimadzu, UV-1601, Tokyo, Japan). The results were expressed as the percentage of reduced DPPH.

The TEAC of MRPs was evaluated by measuring the scavenging of the ABTS⁺ radical (Re et al. 1999). A suitable dilution of each MRP solution was treated with the diluted ABTS⁺ radical solution prepared with potassium persulfate (2.45 mM). The absorbance of samples at 30 °C after 6 min was recorded with a microplate reader (Bio-Tek Instruments, Inc., Winooski, VT). The Trolox solution was used to obtain the calibration curve, and then the degree of decolorization induced by MRPs was measured as TEAC value (μmol TE/g).

The ORAC of MRPs was studied following the method of Huang et al. (2002). Reaction solutions (AAPH and fluorescein) were prepared with phosphate buffer (pH 7.4). Suitable dilutions of MRPs were firstly treated with AAPH as the peroxyl radical generator to start the reaction, and then fluorescein was introduced to the system. The experiment continued until zero fluorescence occurred, and a fluorescence reader (Bio-Tek Instruments, Inc., Winooski, VT) was used to record the absorbance at 37 °C. Trolox (0–200 μM) was used as standard, and results were expressed as micromole Trolox equivalents (TE) per gram (μmol TE/g).

Determination of 5-(hydroxymethyl)furfural and acrylamide content of MRPs

The HMF and AC concentration of MRPs were performed using reverse-phase high-performance liquid chromatography (RP-HPLC) (Shimadzu LC-10A, Kyoto, Japan) with a diode array (DAD SPD-M20A Shimadzu). MRPs were diluted with distilled water and filtered (0.45 μm syringe filter, Millipore Merck, Darmstadt, Germany) for chromatographic separation.

An isocratic elution pattern was adopted for the separation of HMF using a mobile phase of methanol/water mixture (10:90, v/v), including 0.5% (v/v) formic acid. The injection volume was 20 μL while the column (Phenomenex Gemini 110-A, 5 μm) temperature was set at 30 °C with a flow rate of 0.8 mL/min.

The AC content of MRPs was determined by using an isocratic mixture of 50 mM o-phosphoric acid in water/acetonitrile (95/5, v/v) as a mobile phase with a column (Phenomenex Gemini C6-Phenyl, 3.5 μm) at 25 °C. The flow rate was maintained at 0.4 mL/min with an injection volume of 20 μL.

Peak identification was based on the retention time using a comparison of HMF and AC standard compounds. Predetermined amounts of each analyte (HMF and AC) were also added to extracted MRPs and analyzed by HPLC to measure the spiked recoveries. The corresponding calibration curves were used to measure the amounts of detected. The spike recoveries were calculated as the ratio of the quantity spiked to differences in total amount detected and the original amount.

Experimental design

RSM was used to study the effect of process conditions (temperature and time) on the antioxidant capacities of MRPs, and the content of potentially harmful compounds originated from MR. The experimental design was based on a two-factor composite design, including 13 experimental runs and five replications of the center point. The coded values of independent variables and their correspondence with actual values are given in Table 1. The dependent variables were pH, BI, antioxidant activities, HMF, and AC concentration, and the results were analyzed with second-order polynomial by fitting each response using the “Design Expert” software (Stat-Ease, Inc., Minneapolis, USA) statistical package. The main effects of the independent variables on the properties of MRPs were defined with a level of significance at p < 0.05. RSM can be useful to investigate the optimum conditions of the MR model system and help to improve the properties of products, having MR during processing and storage. The advantage of RSM is to reduce the experiments required to evaluate multiple parameters and their interactions with acceptable collective statistical results from sufficient information. Thus, RSM was used to investigate the optimum conditions of temperature and heating time, especially targeting to determine the degree of browning, antioxidant, HMF, and AC formation during MR.

Table 1.

Coded and uncoded levels of the independent variables

Coded levels	Independent variables
	Time (min)	Temperature (°C)
+ 1.41 (+ α)	350	100
+ 1	300	92
0	180	76
− 1	60	60
− 1.41 (− α)	10	53

Results and discussion

Fitting the model

Table 2 shows the pH, BI values, antioxidant activities (TEAC, ORAC, and DPPH values), HMF, and AC concentrations of MRPs at determined design points. The lowest pH and the highest BI, TEAC, and ORAC values were provided by heating at 92 °C for 300 min. The highest DPPH radical scavenging activity was observed in a 100 °C–180 min sample, followed by a 92 °C–300 min sample. Also, lower temperatures presented higher HMF concentrations up to 92 °C, and after this point, AC concentration increased substantially.

Table 2.

Independent variables and results obtained under different conditions based on 2-factor central composite design for response surface analysis

Run	Independent variables		Responses*
	F1	F2	pH	BI	TEAC	ORAC	DPPH	HMF	AC
	Temp (°C)	Time (min)		Abs	μmol TE/g	μmol TE/g	%	ppm	ppm
1	76	180	7.77 ± 0.02	0.91 ± 0.02	0.32 ± 0.01	0.73 ± 0.01	35.50 ± 0.05	0.27 ± 0.01	0.96 ± 0.12
2	100	180	7.25 ± 0.23	1.86 ± 0.04	1.12 ± 0.14	7.83 ± 0.24	84.96 ± 0.06	ND	337.39 ± 12.52
3	92	300	6.95 ± 0.06	1.94 ± 0.01	1.27 ± 0.02	8.20 ± 0.20	66.93 ± 0.15	ND	140.88 ± 3.06
4	60	60	8.36 ± 0.01	0.05 ± 0.01	0.78 ± 0.03	0.25 ± 0.01	4.08 ± 0.04	0.45 ± 0.10	ND
5	60	300	8.29 ± 0.02	0.24 ± 0.01	0.92 ± 0.15	0.68 ± 0.01	4.32 ± 0.02	0.22 ± 0.02	ND
6	76	180	7.77 ± 0.02	1.68 ± 0.01	0.32 ± 0.01	0.33 ± 0.02	35.50 ± 0.0.05	0.27 ± 0.01	ND
7	76	180	7.92 ± 0.03	1.68 ± 0.01	0.89 ± 0.02	0.72 ± 0.02	25.08 ± 0.16	0.17 ± 0.01	0.96 ± 0.12
8	76	10	8.35 ± 0.01	0.04 ± 0.01	0.10 ± 0.08	0.35 ± 0.01	3.30 ± 0.02	0.28 ± 0.02	ND
9	92	60	7.55 ± 0.56	1.59 ± 0.01	0.87 ± 0.04	1.33 ± 0.15	24.40 ± 0.02	ND	11.10 ± 0.05
10	53	180	7.86 ± 0.64	0.04 ± 0.01	0.80 ± 0.19	0.28 ± 0.01	2.79 ± 0.04	2.82 ± 0.12	ND
11	76	350	7.50 ± 0.29	1.93 ± 0.07	1.05 ± 0.01	1.88 ± 0.08	40.30 ± 0.18	0.13 ± 0.01	3.67 ± 0.50
12	76	180	7.92 ± 0.03	1.68 ± 0.01	0.89 ± 0.02	0.73 ± 0.01	25.08 ± 0.16	0.17 ± 0.01	0.96 ± 0.12
13	76	180	7.92 ± 0.03	0.91 ± 0.02	0.89 ± 0.02	0.72 ± 0.02	25.08 ± 0.16	0.17 ± 0.01	0.96 ± 0.12

*All results are the mean ± SD (n = 3)

ND not detected

The analysis of variance (ANOVA) results of the selected model is shown in Table 3. The effects of heating temperature and processing time were analyzed, and their significance was determined by F-values and p values. The variables F₁ (temperature) and F₂ (time) were significant (p < 0.05), while the variables F₁F₂, F²₁, and F²₂ were insignificant for pH and BI values of MRPs. F₂ was found as the only significant variable for ABTS radical scavenging activity (TEAC value), and similarly, F₁ was the single significant variable for DPPH radical scavenging activity. The variables F₁ and F₂ showed a significant interaction for only ORAC values of MRPs due to the synergistic effect between temperature and time on oxidative degradation. It was found that the variables F₁ and F²₁ were significant for HMF and AC concentration of MRPs. In contrast, no significant interaction was observed between the temperature and time for those harmful compounds (p < 0.05).

Table 3.

Analysis of variance (ANOVA) for the fitted model

Source	df	SS	MS	F value	p value prob > F
pH (R2 = 0.87)
Model	5	1.80	0.36	9.11	0.0057 s
Lack of fit	3	0.25	0.083	11.54	0.0194 s
Pure error	4	0.029	7.16 × 10−3
BI (R2 = 0.82)
Model	5	6.02	1.20	6.18	0.0167 s
Lack of fit	3	0.63	0.21	1.14	0.4331 ns
Pure error	4	0.73	0.18
DPPH (R2 = 0.69)
Model	5	7928.11	1585.62	3.11	0.0858 ns
Lack of fit	3	274.74	91.58	0.11	0.9491 ns
Pure error	4	3297.95	824.49
TEAC (R2 = 0.61)
Model	5	0.83	0.17	2.12	0.1776 ns
Lack of fit	3	0.15	0.051	0.51	0.6947 ns
Pure error	4	0.39	0.099
ORAC (R2 = 0.96)
Model	5	89.31	17.86	29.97	0.0001 s
Lack of fit	3	3.98	1.33	27.82	0.0039 s
Pure error	4	0.19	0.048
HMF (R2 = 0.87)
Model	6	5.79	0.97	6.95	0.0163 s
Lack of fit	2	0.82	0.41	128.31	0.0002 s
Pure error	4	0.013	3.196 × 10−3
AC (R2 = 0.83)
Model	5	95,273.58	19,054.72	6.80	0.0129 s
Lack of fit	3	19,626.08	6542.03	23,661.84	< 0.0001 s
Pure error	4	1.11	0.28

df degrees of freedom, SS sum of squares, MS mean square, ns not significant, s significant; significance level of p < 0.05

The effect of temperature and time on pH and BI values of MRPs

Figure 1a shows the changes in pH values during the MR associated with heating temperature and processing time. The formation of complex brown compounds and the Maillard reaction rate depends on several parameters, such as higher pH that may facilitate the initial condensation step (Lertittikul et al. 2007). The highest reduction in pH values was observed at a higher temperature and longer reaction time (p < 0.05). The pH reduction was significantly affected by time and temperature (Table 3). The formation of organic acids, such as formic and acetic acids, might decrease the pH (Vhangani and Van Wyk 2013). Similarly, Gu et al. (2010) found a reduction in pH with the increasing time and temperature.

Fig. 1.

The effect of processing temperature and time on pH (a) and BI (b) values of MRPs

In a Maillard reaction, carbonyl groups react with amino groups producing colorless compounds and then reaction progress to form melanoidins, which give maximum absorbance at 420 nm (Wijewickreme et al. 1997). The nature of reducing sugar influences BI values of MRPs, because the reactivity of these sugars decreases from aldopentoses to hexoses and disaccharides. Thus, glucose with its more reactive aldehyde carbonyl group was selected to determine the effects of processing conditions. Figure 1b shows the BI values of MRPs during the reaction based on temperature and time. Time and temperature both significantly affected the BI values of MRPs, which increased with an increase in processing time and temperature (p < 0.05) (Table 3).

Moreover, samples having lower pH values resulted in higher BI values. This result might be due to the faster Amadori product degradation and carbohydrates dehydration in acidic medium, which then produced furaldehyde and cyclic compounds that could react to form intense brown compounds (Tressl et al. 1995). Similarly, Zhuang and Sun (2011), and Phisut and Jiraporn (2013) reported an increase in brown color as the temperature and time increased for lysine–glucose and chitosan–sugar solutions, respectively.

The effect of temperature and time on the antioxidant properties of MRPs

The antioxidant activities of MRPs were determined by ABTS⁺ and DPPH free radical scavenging activity, and ORAC tests. Figure 2 presents the changes in antioxidant activities of MRPs throughout the Maillard reaction based on time and temperature.

Fig. 2.

The effect of processing temperature and time on antioxidant activity (a ABTS radical scavenging activity; b ORAC value; c DPPH radical scavenging activity) of MRPs

The highest TEAC value was in the case of heating at 92 °C for 300 min, and the antioxidant effect of these compounds increased when temperature and time increased except heating at 100 °C for 180 min. It was also observed that processing time significantly affected the ABTS⁺ scavenging capacity, whereas this behavior was not changed by temperature significantly (Table 3).

Both the increasing reaction times and the temperature had a significant effect and resulted in higher ORAC values (p < 0.05). However, the ORAC value declined after heating at 100 °C, which could be related to the acids formed at higher temperatures with longer processing times, as was observed in ABTS free radical scavenging tests. Similarly, Knol et al. (2010) reported reduced peroxyl radical scavenging activity for fructose–asparagine model systems having longer times at high temperatures. DPPH radical scavenging activity was significantly increased with the increasing temperature (p < 0.05). However, a non-significant increase was found for longer processing times (p > 0.05). It can be concluded that MRPs strongly inhibited DPPH activity at higher temperatures than 76 °C. DPPH radical scavenging values are consistent with those obtained from the ABTS radical scavenging activity and ORAC tests. Similarly, Xu et al. (2017) observed that histidine–glucose model systems resulted in the most potent free radical scavenging activity when heated at high temperatures and the longer time leading a detrimental effect with further increases.

Generally, it was found that treatments at higher temperatures with longer processing time showed higher antioxidant activities along with the higher browning intensity values. The formation of chromophoric compounds could explain the conversion of Amadori products (Sun et al. 2006a). This result is also related to the process conditions, along with the physicochemical properties of reactants. For example, Mogol et al. (2010) reported higher free radical scavenging activity for an arginine–glucose model system when compared to histidine and lysine–glucose model systems.

The effect of temperature and time on the HMF and AC concentration of MRPs

In the early stages of the MR, the Amadori rearrangement reaction occurs, and the products of this reaction are accepted as indicators of heat damage in food products (Delgado-Andrade et al. 2010). HMF is formed under acidic conditions through the 1,2-enolisation of Amadori compounds or caramelization reaction, which is an acid-catalyzed degradation of sugars (Morales and Van Boekel 1997).

Figure 3a shows the HMF values during the MR based on time and temperature. The highest HMF concentration was found in 53 °C–180 min samples, followed by a decrease as the temperature and time increased. Also, HMF was not detected in 92 °C–60 min, 92 °C–300 min, and 100 °C–180 min samples, possibly due to the reversible isomerization of glucose followed by formation to AC or due to the faster degradation of glucose to organic acids (Knol et al. 2005).

Fig. 3.

The effect of processing temperature and time on HMF (a) and AC (b) concentration of MRPs

The change in AC concentration during the Maillard reaction based on temperature and time is presented in Fig. 3b. An increase was observed in AC concentration with the increasing temperature and processing period. The maximum AC formation was reached in samples heated at 100 °C for 180 min, which could be the result of low pH leading the protonation of the amino groups (Mogol and Gökmen 2016). At lower temperatures than 76 °C, AC was not detected regardless of the processing time. It was observed that AC concentration and pH values and AC concentration and BI values were both positively correlated.

Similarly, Knol et al. (2010) observed higher AC concentrations at higher temperatures (160–200 °C) for fructose–asparagine model systems. On the other hand, it was found that AC concentration was inversely related to HMF content. 53 °C–180 min samples were found to have the highest HMF, whereas the decrease in the pH of the model system significantly decreased HMF formation (p < 0.05). The reduction of HMF formation continued as the temperature and processing time increased, which then resulted in a rise in AC concentration. HMF is known as a potent carbonyl compound accelerating the AC formation with increasing temperatures. Therefore, the increase in AC concentration might be due to the use of HMF as a carbonyl source for AC formation (Gökmen et al. 2012). Mogol and Gökmen (2016) also found a negative correlation between AC and HMF concentrations, as the temperature increased for the asparagine–glucose model system containing chitosan.

Determination of process conditions with desirability function

The overall desirability function examined the process conditions that maximized pH and antioxidant activities and minimized BI values, HMF, and AC concentrations. HMF and AC are harmful compounds, and the maximum levels are chosen as 0.5 mg/kg body weight and 0.4 μg/kg body weight (for daily intake of a 70 kg body weight) for HMF and AC, respectively, according to the reports of Joint FAO/WHO Expert Committe on Food Additives (JECFA) (2017). The highest rate of desirability (d) function was obtained as 0.72 for 300 min of processing time and process temperature at 75.08 °C. Thus, these experimental conditions with a d value of 0.72 would result in the obtaining of MRPs with pH at 7.70, 1.53 absorbance for BI value, 0.89 μmol TE/g of ABTS radical scavenging activity, 1.81 μmol TE/g of ORAC value, 41% of DPPH radical scavenging activity, 0.002 ppm of HMF and 0.01 ppm of AC. The predicted values were compared with experimental values. The differences between experimental and predicted values were found lower than 15% that could be considered as low in the desired region (Rezende et al. 2017).

Conclusion

In this study, the desired experimental conditions were found as 75.08 °C–300 min with a d-value of 0.72. The results obtained for optimum conditions were pH of 7.70, 1.53 of BI value, 0.89 μmol TE/g of ABTS radical scavenging activity, 1.81 μmol TE/g of ORAC value, 41% of DPPH radical scavenging activity, 0.002 ppm of HMF and 0.01 ppm of AC. Generally, the MRPs derived at higher temperatures and longer processing time had lower pH values and higher BI values. Both the increasing reaction times and temperature significantly increased the antioxidant activities of MRPs. However, the ORAC value declined after heating at 100 °C, which could be due to the formation of acidic compounds during heating at very high temperatures for longer processing time, as was observed in ABTS⁺ free radical scavenging tests. At higher temperatures and heating times, HMF concentration decreased while AC concentration tends to increase the possible use of HMF as a carbonyl source for AC formation. The results obtained in RSM would be beneficial to find the potential relationship between processing time and temperature for controlling the oxidation in foods due to the high antioxidant capacity of MRPs even they have been accepted as potentially toxic compounds.