Physicochemical and chemical properties of mung bean protein isolate affected by the isolation procedure

Abstract

The effects of different mung bean protein isolation methods on the chemical composition, the physicochemical properties, and selected antinutritional factors of mung bean protein isolates were investigated. Six protein isolates were prepared by isoelectric precipitation at different extraction pH levels (pH 8 and 9), by micellization, and by hybrid isolation at varying salt concentrations (0.25 M, 0.50 M, 0.75 M). The extraction conditions affected the amount of antinutritive compounds of the isolates. Compared to mung bean flour, micellization reduced phytic acid content by approximately 48% and trypsin inhibitor activity by around 88%. The remaining phytic acid concentration of the isolates influenced their re-solubility, particularly under acidic conditions. The protein isolates exhibited significant differences in surface hydrophobicity and thermal characteristics, indicating structural modifications caused by the extraction methods. Micellization and extraction at pH 8 were identified as mildest isolation methods, as evidenced by the highest enthalpy values. SDS-PAGE analysis demonstrated an enrichment of globulins and comparable protein profiles among the isolates, suggesting that the observed differences arise from conformational changes rather than variations in protein composition. The product yield in protein extraction from mung beans ranged from 8% to 19%, emphasizing the importance of enhancing overall extraction efficiency or exploring the utilization of by-products obtained during the protein isolation process.

Graphical abstract

Highlights

• •Conventional alkali extraction resulted in the highest product yield up to 19%.
• •Extraction at pH 8 and micellization yielded the most native protein isolates.
• •Hybrid isolation and micellization reduced trypsin inhibitor activity the most.
• •Micellized protein isolate revealed the highest protein solubility.

Abbreviations

ANS

8-Anilinonaphthalene-1-sulfonic acid

ΔH

enthalpy of denaturation

dw

dry weight

FT-IR

Fourier transform infrared spectroscopy

H

hybrid isolates

H0.25

H prepared with 0.25 M

H0.50

H prepared with 0.5 M

H0.75

H prepared with 0.75 M

IP

isoelectric mung bean protein isolate

IP8

IP extracted at pH 8

IP9

IP extracted at pH 9

MP

micellized mung bean protein isolate

NR

non-reducing conditions

OAC

oil absorption capacity

R

reducing conditions

SDS

sodium dodecyl sulfate

TCA

trichloroacetic acid

TIA

trypsin inhibitor activity

TIU

trypsin inhibitor units

T_D

denaturation temperature

WAC

water absorption capacity

1. Introduction

The world population is growing steadily. According to forecasts, more than nine billion people will be living on our planet in 2050 (United Nations, 2022). The associated overexploitation of natural resources is contributing to the global ecological crisis and climate change and impeding the nutrition of the entire world population. Excessive consumption of land and other resources for the production of animal proteins and an increasing consumer awareness with respect to health, nutrition, and the ecosystem has paved the way for growth in the alternative protein market. Pulses are an excellent source for the production of protein products due to their high protein content and their highly extractable proteins. The good extractability of legume proteins can be attributed to the protein storage anatomy in these crops, facilitating separation (Tenorio et al., 2018). Today, the most commonly used protein source among legumes is still soy. It is, therefore, of utmost importance to explore new alternative protein sources in order to provide a wider range of proteins with distinct functional properties for the replacement of animal protein. These include other legumes. The European Food Safety Authority (EFSA) has recently accepted mung bean protein isolate as a novel food (EFSA Panel on Nutrition et al., 2021). This opens up the possibility of using mung bean protein isolate as a plant-based protein alternative for numerous food products.

Mung bean (Vigna radiata (L.) Wilczek) is used worldwide as a nutrient-rich food as well as an important animal feed. Mung beans are a rich source of proteins, fibers, and minerals (Mubarak, 2005; Shrestha et al., 2023). The protein content of mung bean ranges from 24% to 29% and is rich in essential amino acids, especially the aromatic amino acids leucine, isoleucine, and valine (Mubarak, 2005; Li et al., 2010). The amino acid content of mung bean protein isolate largely meets World Health Organization (1991) requirements, except for sulfur-containing amino acids and tryptophan (Kudre et al., 2013). Mung bean protein isolate was found to have good gelling and foaming properties and is well suited for texturization comparable to commercial soybean products (Li et al., 2010; Brishti et al., 2017). However, the presence of antinutrients can limit the applicability of mung bean and its isolate, since their occurrence can influence the bioavailability of proteins and minerals (Mubarak, 2005). Among others, phytic acid and trypsin inhibitors have been detected in mung bean (Mubarak, 2005). Several protein extraction techniques have been described that reduce the levels of antinutrients in various protein isolates from different sources such as soy or Moringa oleifera (Ali et al., 2010; Illingworth et al., 2022). Therefore, the choice of extraction technique can be an important factor to reduce the amount of antinutritional factors in the resulting protein isolate.

Besides the chemical composition, isolation parameters strongly influence the technofunctional quality of a protein isolate. Well-established, conventionally used techniques for the production of protein isolates are alkaline extraction followed by isoelectric precipitation and salt extraction followed by dilutive precipitation, titled as micellization. Both procedures have already been performed on various beans and other legumes (Muranyi et al., 2013, 2016; Tanger et al., 2020; Illingworth et al., 2022). It was found that different pH values during alkaline extraction led to variations in protein yield, purity, and in the functional properties of the resulting protein ingredients (Das et al., 2021). In addition, high alkaline extraction values promote protein denaturation due to the strong pH-shift using isoelectric precipitation and, thus, affect protein solubility. Compared to isoelectrically prepared protein isolates, micellization resulted in protein isolates with less protein denaturation and higher protein purity, but lower protein yield, which was observed for various raw materials (Rahma et al., 2000; Muranyi et al., 2016; Illingworth et al., 2022). However, limited research has been conducted on the effects of micellization on the chemical and physicochemical properties of mung bean protein isolates (Shrestha et al., 2023), which is essential for the assessment of potential applications.

It is already known that mung beans contain some antinutritional factors that can be reduced by dehulling. Therefore, dehulled mung beans were used in the present study. Until now alkaline extraction combined with isoelectric precipitation was studied on mung beans mainly at one extraction pH (Li et al., 2010; Brishti et al., 2017). Considering the relevance of milder isolation pH for the physicochemical and functional properties of protein isolates, a comparison of mung bean protein isolation at two different pH values was conducted in the present study. In addition, salt extraction was performed in combination with dilute precipitation to determine the effects of a milder isolation method on mung bean protein isolate. Salt extraction at 0.5 M NaCl has been studied previously (Rahma et al., 2000), but determination of the ideal NaCl concentration for mung bean protein extraction may enhance protein solubility, leading to improved extraction efficiency. Considering the distinct benefits of the two conventional isolation methods and their impact on technofunctional properties, the combination of both isolation techniques could provide hybrid isolates with high quality and yield. Therefore, salt extraction in combination with dilutive and isoelectric precipitation (= hybrid isolation) of mung bean proteins was performed in the present study for the first time. Different salt concentrations were utilized to produce hybrid isolates to examine the effect of NaCl during protein isolation. To evaluate the obtained protein isolates, the chemical composition and selected technofunctional properties and antinutritional factors were comprehensively investigated in our study. Thus, this study contributes to the advancement of isolation techniques for mung bean proteins by considering multiple factors such as pH, salt concentration, and hybrid isolation methods.

2. Material and methods

2.1. Mung beans

Commercial dehulled mung beans (Vigna radiata (L.) Wilczek) from Rapunzel Naturkost were purchased in Germany. The beans came from Chinese organic farming and were already mechanically shelled and halved. The beans were not heated above 40 °C at any time.

2.2. Chemicals and standards

Ultrapure water was obtained from a Direct-Q® Water Purification System (Merck, Darmstadt, Germany). For sample defatting, n-hexane from Merck (Darmstadt, Germany) was used. Sodium chloride (Sigma Aldrich, Taufkirchen, Germany), hydrochloric acid (VWR, Darmstadt, Germany), and sodium hydroxide (Thermo Fisher Scientific, Waltham, MA, USA) were used for protein solubilization and isolation. For the measurement of the protein content, TRIS (Merck, Darmstadt, Germany) and EDTA standards and DT Superabsorber were used (Gerhardt, Königswinter, Germany). Magnesium acetate from Merck (Darmstadt, Germany) and nitric acid from VWR (Darmstadt, Germany) were utilized for determination of ash content. For determination of fat content, hydrochloric acid (VWR, Darmstadt, Germany) and petroleum benzine (Merck, Darmstadt, Germany) were used. Total Starch Kit from Megazyme (Wicklow, Ireland) was used for starch determination. Phytic acid Assay Kit (Megazyme, Wicklow, Ireland), ammonium molybdate, ascorbic acid, sulfuric acid, and trichloroacetic acid (TCA) from Sigma (Taufkirchen, Germany) were used for phytic acid determination. For determination of trypsin inhibitor activity, calcium chloride dihydrate, dimethyl sulfoxide, tris(hydroxymethyl)aminomethane from Merck (Darmstadt, Germany), acetic acid, N-α-Benzoyl-DL-arginine-4-nitroanilide hydrochloride from VWR (Darmstadt, Germany), and trypsin bovine pancreas from Sigma Aldrich (Taufkirchen, Germany) were utilized. Sodium dodecyl sulfate (SDS), β-mercaptoethanol, and bromophenol blue from Merck (Darmstadt, Germany), Tris-HCl and glycerol from Serva (Heidelberg, Germany), acetic acid from VWR (Darmstadt, Germany), ethanol from Julius Hoesch (Düren, Germany), and Precision Plus Protein™ all blue prestained protein standard, electrophoresis buffer, and coomassie brilliant blue R-250 from Biorad (Feldkirchen, Germany) were used for SDS-PAGE. Disodium hydrogen phosphate, sodium dihydrogen phosphate (VWR, Darmstadt, Germany), and 8-Anilinonaphthalene-1-sulfonic acid (ANS) from Alfa Aesar (Ward Hill, USA) were used for surface hydrophobicity measurements.

2.3. Preparation of mung bean flour

The seed kernels were ground with a blade-grinder (Grindomix GM200, Retsch, Haan, Germany) and sieved using a 300 μm mesh sieve. The flour was defatted with n-hexane at a 1:3 (w/v) ratio for 2 h and air-dried for 24 h. The flour was stored at 10 °C until further use.

2.4. Determination of isolation parameters by protein solubility curves

To maximize protein yield for each isolation procedure, the protein solubility of mung bean flour was determined. The mung bean flour was diluted 1:10 (w/v) with water and the pH was adjusted from 2 to 10 with HCl (1 M, 5 M) or NaOH (1 M, 5 M). For salt extraction, the same procedure was done by dilution of the mung bean flour in salt solutions of 0.25 M, 0.50 M, and 0.75 M NaCl, respectively. The solutions were centrifuged at 1,000 g for 20 min (Heraeus™ Biofuge™ Stratos™, Thermo Fisher Scientific, Waltham, MA, USA). Protein content of the flour and the supernatant was analyzed via Dumas method (Dumatherm, Gerhardt, Germany). A nitrogen to protein conversion factor of 6.25 was used. The protein content of the samples was related to their respective dry weight (dw). The protein solubility was determined according to the Eq. (1):

$$\mathrm{Protein}\mathrm{solubility}\left[\%\right]=\frac{\mathrm{Protein}\mathrm{content}\mathrm{in}\mathrm{supernatant}\left[\%\right]}{\mathrm{Protein}\mathrm{content}\mathrm{in}\mathrm{sample}\left[\%\right]\mathrm{dw}}\times 100$$ (1)

2.5. Preparation of protein isolates

2.5.1. Preparation of alkali extracted and isoelectric precipitated mung bean protein isolates (IP)

The protein isolation was performed according to the method of Brishti et al. (2017) with slight modifications. The defatted mung bean flour was diluted 1:10 (w/v) with water and the pH was adjusted to 9 (IP9) and 8 (IP8) using 1 M and 5 M NaOH. The suspension was stirred at room temperature for 1 h and centrifuged at 1,000 g for 30 min. The supernatant was acidified to pH 5 with 1 M and 5 M HCl and stirred for 30 min. The precipitate was recovered by centrifugation at 4,000 g for 25 min, was washed once with deionized water, and lyophilized using a freeze-drier (Model Alpha 2–4 LSC, Christ GmbH, Osterode, Germany).

2.5.2. Preparation of micellized mung bean protein isolate (MP)

The preparation of micellized protein isolate was done by the method of Muranyi et al. (2016) with slight modifications. The mung bean flour was diluted 1:10 (w/v) in an aqueous salt solution of 0.75 M NaCl with a native pH value of pH 6.1. After 1 h of stirring, the solution was centrifuged at 1,000 g for 30 min. The supernatant was diluted with deionized water at a ratio of 1:5 (to approx. 0.125 M NaCl) and stirred for 30 min at room temperature. The dispersion was allowed to stand at 4 °C overnight and was centrifuged at 12,000 g for 25 min. The precipitate was washed with deionized water and lyophilized.

2.5.3. Hybrid isolates (H): combination of salt induced extraction and dilutive isoelectric precipitation

The defatted mung bean flour was diluted 1:10 (w/v) in aqueous salt solutions of 0.25 M (H0.25), 0.5 M (H0.50), and 0.75 M (H0.75) NaCl and the pH was adjusted to pH 6 with 1 M HCl according to the protein solubility curves. After stirring for 1 h, the dispersion was centrifuged at 1,000 g for 30 min. To induce dilutive precipitation, the supernatant was diluted again to approximately 0.125 M NaCl by dilution with deionized water at a ratio of 1:1, 1:3, and 1:5, respectively. The supernatant was precipitated isoelectrically by adjusting the pH to 4 with 1 M and 5 M HCl and stirred for 30 min. The dispersion was centrifuged at 12,000 g for 25 min, washed with deionized water, and freeze-dried.

2.6. Protein yield and purity

To assess the efficiency of the different isolation methods, the protein yield was determined according to Das et al. (2021) by the following Eq. (2) based on the dry weight content:

$$\mathrm{Protein}\mathrm{Yield}\left(\%\right)=\frac{\mathrm{protein}\mathrm{content}\mathrm{isolate}\left[\%\right]\times \mathrm{weight}\mathrm{of}\mathrm{isolate}\left[\mathrm{g}\right]}{\mathrm{protein}\mathrm{content}\mathrm{flour}\left[\%\right]\times \mathrm{weight}\mathrm{of}\mathrm{flour}\left[\mathrm{g}\right]}\times 100$$ (2)

2.7. Chemical composition

The defatted mung bean flour and the protein isolates were analyzed for protein, starch, fat, moisture, and ash content. Protein content was determined by the Dumas method (Dumatherm, Gerhardt, Germany) based on the nitrogen content (N x 6.25). Starch content was measured enzymatically according to AOAC Method 996.11. The ash content was measured by combustion method at a temperature of 600 °C. To enhance the ashing process, a solution of magnesium acetate (12% w/v) was added and the mass of the magnesium acetate solution was subsequently measured and subtracted from the total ash of the samples. Fat content was determined according to AOAC Method 922.06 with slight modifications, instead of ethanol demineralized water was used. The moisture content was determined by heating the solid samples up to 100 °C (MA-30 Sartorius, Göttingen, Germany) until mass constancy and the dry weight was calculated gravimetrically. All values were expressed as a percentage of the respective dry weight of the samples.

2.8. Antinutritional factors

2.8.1. Determination of phytic acid

The colorimetric determination of phytic acid was done enzymatically (Cat. No. K-PHYT, Megazyme International Bray, County Wicklow, Ireland) by measuring free phosphor and total phosphor content, respectively, resulting from acid extraction and treatment with Megazym phytase enzyme preparation. Depending on the predicted phytic acid content of the samples, 0.7–1.0 g sample material was weighted and stirred overnight in 20 mL HCl (0.66 M). An aliquot of 1 mL of sample extract was centrifuged and 0.5 mL of the supernatant was mixed with 0.5 mL NaOH (0.75 M). For the determination of total phosphor content, 0.06 mL of distilled water, 0.2 mL of phytase assay buffer, and 0.02 mL of phytase enzyme preparation were added to 0.05 mL of sample extract and mixed accurately followed by incubation in a water bath at 40 °C for 10 min. Alkaline phosphatase buffer (0.2 mL) and alkaline phosphatase enzyme preparation (0.02 mL) were added, mixed well and placed in a water bath at 40 °C for 15 min. After 15 min, the reaction was stopped by adding 0.30 mL TCA (50% w/v) and a subsequent centrifugation followed (11,300 g, 10 min). The supernatant was used for colorimetric determination. The free phosphor content was determined by measuring the sample in absence of the addition of phytase and alkaline phosphatase enzyme preparation. Five phosphor standard solutions were prepared (0 μg, 0.5 μg, 2.5 μg, 5 μg, 7.5 μg) to calculate the mean value of phosphor standards. For the colorimetric determination, 1.0 mL of the samples supernatant and phosphor standards were mixed with 0.5 mL color reagent, incubated at 40 °C for 1 h and measured at 655 nm. Phytic acid content was calculated according to Eqs. (3), (4), (5) and related to the respective dry weight of the samples:

$$\mathrm{M}\mathrm{V}\left[\frac{\mathrm{\mu }\mathrm{g}}{\Delta {\mathrm{A}}_{\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{s}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{r}}}\right]=\frac{{\mathrm{M}}_{0.5}+{\mathrm{M}}_{2.5}+{\mathrm{M}}_{5.0}+{\mathrm{M}}_{7.5}}{4}$$ (3)

with MV = mean value of phosphor standards; M = corresponding phosphor standard [μg]/ΔA_phosphor

$${\mathrm{c}}_{\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{s}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{r}}\left[\mathrm{g}/100\mathrm{g}\right]=\frac{\mathrm{M}\mathrm{V}\times \mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\mathrm{e}\mathrm{x}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{v}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{m}\mathrm{e}\times \mathrm{d}\mathrm{i}\mathrm{l}\mathrm{u}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}{10.000\times \mathrm{m}\mathrm{g}\mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\times \mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\mathrm{v}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{m}\mathrm{e}}\times \Delta {\mathrm{A}}_{\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{s}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{r}}$$ (4)

$${\mathrm{c}}_{\mathrm{p}\mathrm{h}\mathrm{y}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{c}\mathrm{i}\mathrm{d}}=\frac{\mathrm{P}\mathrm{h}\mathrm{o}\mathrm{s}\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{r}\left[\mathrm{g}/100\right]}{0.282}$$ (5)

2.8.2. Determination of trypsin inhibitor activity

The trypsin inhibitor activity (TIA) was measured according to the Trypsin Inhibitor Assay half volume method of Liu (2019). Enzyme extract was prepared by stirring 1 g sample material in 50 mL NaOH (10 mM) at room temperature for 3 h. Mung bean extract was diluted with deionized water to the extent that 1.0 mL of the extract caused trypsin inhibition of 30–70%. The test was performed in a water bath at 37 °C. An aliquot of 1.0 mL of the diluted extract was mixed with 2.5 mL Nα-benzoyl-DL-arginin-4-nitroanilid-hydrochlorid, 1.0 mL trypsin enzyme preparation, and 0.5 mL acetic acid (30%). Once the trypsin solution was added, the reaction was stopped after 10 min by the addition of 1.0 mL of 30% acetic acid solution. The mixture was centrifuged (14,100 g, 5 min) and measured at 410 nm (A410S). The reference value (A410R) was determined by measuring the reaction in absence of inhibitors by replacing the sample extract with water. Furthermore, reagent blank values for the sample measurements (A410SB) and reagent blank values for the reference values (A410RB) were carried out by addition of the acetic acid solution prior to the trypsin solution. For all absorbance measurements, deionized water was used as reference. A trypsin unit (TU) is defined as the increase in absorbance by 0.02 at 410 nm. Trypsin inhibitor units (TIU) are expressed per mg protein (dw) and calculated as follows:

$$\frac{\mathrm{TIU}}{\mathrm{mg}\mathrm{protein}}=\frac{\left[\left({\mathrm{A}}_{410\mathrm{R}}-{\mathrm{A}}_{410\mathrm{B}}\right)-\left({\mathrm{A}}_{410\mathrm{S}}-{\mathrm{A}}_{410\mathrm{SB}}\right)\times 100\right]\times \mathrm{mL}\mathrm{diluted}\mathrm{mung}\mathrm{bean}\mathrm{extract}}{\left(\mathrm{mg}\mathrm{sample}\mathrm{per}\mathrm{mL}\mathrm{diluted}\mathrm{mung}\mathrm{bean}\mathrm{extract}\times \mathrm{protein}\mathrm{content}\right)}$$ (6)

2.9. Color parameters and browning index

For the color measurement, a Chroma Meter CR-400/410 (Konica Minolta, Osaka, Japan), equipped with a measuring head CR-400, a data processor DP-400, a CR-A44 white calibration plate, and glass cuvettes CR-A502 was used. The calibration was performed using a white standard with Y = 93.8, x = 0.3157, y = 0.3322 prior to color determination. The browning index was calculated using the equation of Brishti et al. (2020):

$$\mathrm{Browning}\mathrm{index}=\frac{\mathrm{x}-0.31}{0.172}\times 100$$ (7)

with x = (a* + 1.75L*)/(5.645L* + a* - 0.3012b*)

2.10. Protein solubility of the protein isolates

A 1% (w/v) suspension was produced by dispersing the respective isolate in distilled water. The solution was stirred for 1 h until a homogenous mixture was obtained. The pH was set from 2 to 10 with 1 M and 5 M NaOH or HCl. After centrifugation at 1,000 g for 30 min, the supernatant was analyzed for soluble protein content via Dumas method. A conversion factor of 6.25 was used.

2.11. SDS-PAGE analysis

SDS-PAGE analysis under reducing conditions (R) was conducted using a 12% precast polyacrylamide gel (Biorad Laboratories GmbH, Feldkirchen, Germany). A protein solution of 5 g/L was prepared by dispersing the samples in 500 μL of distilled water and by the addition of 500 μL Tris-HCl (0.13 M) buffer (pH 6.8) containing 1 g/L bromophenol blue, 20% glycerol (v/v), 60 g/L SDS, and 10% β-mercaptoethanol. The samples were mixed well, heated at 95 °C for 10 min, and were centrifuged at 14,000 g for 10 min. Aliquots of 5 μL of the supernatant were loaded into the gel chambers. The electrophoresis buffer (Biorad Laboratories GmbH, Feldkirchen, Germany) consisted of 25 mM Tris, 192 mM glycine, and 0.1% SDS (pH 8.3) was diluted 1:10 with water. The electrophoresis was carried out at 120 V for approximately 1 h, until the tracking dye reached the bottom of the gel. The gel was stained with Coomassie Blue R-250 for 45 min and destained with a mixture of ethanol, acetic acid, and water (4:1:5, v/v). SDS-PAGE under non-reducing (NR) conditions were prepared by the same procedure, but without the addition of β-mercaptoethanol.

2.12. Thermodynamic properties

Thermodynamic properties were analyzed using a differential scanning calorimeter (DSC Q2000; TA Instruments, Lukens Drive, New Castle, DE, U.S.). The samples were prepared according to Devkota et al. (2023) with slight modifications. The isolates were suspended in demineralized water 20% (w/v), stirred for 2 h, and hydrated overnight. The measurement was done according to the method of Muranyi et al. (2016). The protein suspensions (10 mg) were heated in sealed standard aluminum pans. An empty pan was used as a reference. The pans were heated from 30 to 110 °C at a linear heating and cooling rate of 2 K/min. Enthalpy (ΔH) and denaturation temperature (T_D) were calculated by the TA Universal Analysis V.4.5A software (TA Instruments). The enthalpy values were related to the protein content (dw) of the samples.

2.13. Surface hydrophobicity

The surface hydrophobicity (S₀) was determined by the method of Nakai (2003) using 1-anilino-8-naphthalenesulfonate (ANS) with some adjustments. A 0.1% stock solution of protein isolate was prepared in phosphate buffer (10 mM, pH 7) and stirred at 4 °C overnight until the isolate had completely dissolved. The stock solution was diluted to five concentrations ranging from 0.004 to 0.06% (w/v). Then, 20 μL ANS solution was added (8.0 mM in 10 mM phosphate buffer, pH 7) to 1 mL of each solution. After stirring, the samples were left in the dark for 15 min. Fluorescence measurement was done with a microplate reader (FLUOstar Omega, BMG Labtech, Ortenberg, Deutschland) at 355 nm λ_excitation and 460 nm λ_emission. The results were shown as the slope of the curve of protein concentration (dw) and fluorescence intensity. Measurement was performed with two independent samples measured in triplicate and corrected by a blank measured without ANS.

2.14. Fourier transform infrared (FT-IR) spectroscopy

The secondary structures of mung bean proteins were measured using a FT-IR spectrometer (Cary 630 FTIR, Agilent Technologies, California, USA). The spectra of the dry powder samples were recorded in the region of 4000–650 cm⁻¹ with a spectral resolution of 4 cm⁻¹ and 32 scans. The FT-IR spectra were analyzed according to Fevzioglu et al. (2020) using OriginLab software Version 2023b (OriginLab Corporation, Massachusetts, USA). The empty crystal was used as background. For analysis, the original absorption spectra were normalized. The normalized spectra and the second derivative spectra were smoothed using a seven-point, third-degree polynomial Savitzky-Golay function. The normalized protein spectra within the amide I region (1600–1700 cm⁻¹) were used for the Gaussian curve-fitting process. Ideal alignment was achieved through iterative peak adjustment.

2.15. Water (WAC) and oil absorption capacity (OAC)

The water and oil absorption capacity were determined by the method of Brishti et al. (2017) with slight modifications. For the determination of WAC and OAC, 0.5 g sample material was dispersed in water at a ratio of 1:20 (w/v) and in corn oil in a ratio of 1:10 (w/v), respectively. The samples were left for 30 min, with remixing after 15 min to achieve a complete dispersion. After centrifugation at 5,000 g for 30 min, the water was decanted and the samples were weighted immediately. For OAC, the supernatant was decanted and the tube was left hanging at a 45-degree angle for 20 min to allow the oil to drain off. The WAC/OAC was calculated as follows:

$$\mathrm{OAC}\mathrm{or}\mathrm{WAC}\left[\mathrm{g}/\mathrm{g}\right]=\left({\mathrm{W}}_2-{\mathrm{W}}_1\right)/\left({\mathrm{W}}_0\times \mathrm{protein}\mathrm{content}\mathrm{sample}\left[\%\right]\right)$$ (8)

where $W_1$ is the weight of the tube plus dry sample [g], $W_2$ is the weight of the tube plus oil or water saturated sample, and $W_0$ is the weight of the dry sample [g]. Protein content was related to the dry weight. The WAC or OAC is expressed as gram of water or oil absorbed per gram of protein isolate, respectively. To determine the samples relative water and oil absorbance, the water-oil absorption index (W/O index) was calculated based on the method of De Kanterewicz et al. (1987), calculated as follows:

W/O Index = WAC/OAC (9)

2.16. Statistics

All results are shown as means of three replicates and the data were expressed as mean ± standard deviation. Statistical analysis of the results was performed using XLSTAT software (version 2022.3.1, Addinsoft Technologies, Paris, France). For pairwise comparisons, an ANOVA with Bonferroni post-hoc test (selected significance level p ≤ 0.05) was used.

3. Results and discussion

3.1. Determination of extraction parameters by protein solubility curves

Analysis of the protein solubility curves. Protein solubility of mung bean flour at different pH values and different NaCl concentrations (Fig. 1) was evaluated to determine the isolation parameters. The extractability of proteins increased with higher pH and NaCl concentration. It is known that proteins have a negative net charge at alkaline pH, which increases their solubility due to stronger electrostatic repulsion. Therefore, the solubility without any NaCl addition was highest at pH 10 (81.6%). Lowest solubility was found at pH 5 (10.9%). These findings were slightly higher compared to those provided by Wang et al. (2011) who found lowest solubility of mung bean protein at pH 4.4.

Fig. 1.

Protein solubility profile of mung bean flour at different pH-values and different NaCl concentrations. Values are means ± standard deviation (n = 3).

The protein solubility further increased with rising salt concentrations (Fig. 1). This phenomenon is attributed to the salting-in effect, wherein proteins interact with salt ions and water molecules. The chloride and sodium ions interact with the oppositely charged protein groups causing the neutralization of their surface charges. Consequently, the electrostatic interactions between the proteins are reduced and their solubility is increased (Schröder, 2017). The addition of 0.75 M NaCl within the pH range of 6–7 resulted in the highest protein extractability, reaching levels of 83.5%–84.0%. As a result, protein solubility was significantly (p ≤ 0.05) increased by more than 15% (pH 7; 0.75 M NaCl) compared to alkaline extraction without NaCl (pH 7; 0 M NaCl). In this way, high alkaline pH values can be avoided during extraction in order to achieve milder extraction conditions and high protein yields. Since the effects of salting-in and salting-out are depending upon the protein composition, charge differences, hydrophobic properties, and structure of the samples (Schröder, 2017), different protein materials may exhibit different behaviors.

In addition, a shift of the isoelectric point from pH 5 to 4 was detected for salt supported extraction. Due to their smaller hydration radius negatively charged chloride ions can bind more strongly to positive charged protein groups than positively charged sodium ions to negatively charged protein groups. Therefore, the proteins become more negatively charged and repel each other more strongly at their original isoelectric point (Ockerman, 1996).

Determination of extraction parameters. To analyze the influence of the extraction pH on the properties of the protein isolates, the extraction of mung bean protein was performed without any NaCl addition at pH 8 and pH 9 for IP8 and IP9, respectively, and precipitation was carried out at the point of lowest solubility, pH 5 for both. Salt extraction of MP was performed with 0.75 M NaCl at a native pH of 6.1 and dilutive precipitation was done by adding water 1:5 (to approx. 0.125 M NaCl). The resulting pH after dilution was 6.5. Hybrid isolates were extracted at different salt concentrations (0.25 M, 0.50 M, and 0.75 M) at pH 6. To induce dilutive precipitation, all salty protein extracts were diluted to a NaCl concentration of about 0.125 M by adding deionized water 1:1, 1:3, and 1:5, respectively. Considering the protein solubility curves at the different NaCl concentrations (0.25 M–0.75 M), the point of lowest solubility shifted from pH 5 to 4. Therefore, the solubility of mung bean proteins was determined at the salt concentration after dilutive precipitation (0.125 M) in order to maximize protein yield. The solubility at 0.125 M NaCl was found to be 19.0% and 20.4% at pH 4 and 5, respectively. Therefore, a precipitation pH of 4 was selected for the hybrid isolates.

3.2. Chemical composition

The chemical composition of mung bean flour and protein isolates is shown in Table 1. The protein content of the mung bean flour was 26.8% dw. Protein concentrations of mung bean protein isolates varied from 95.9% to 99.5% dw, for IP9 and MP, respectively, which were higher than reported in other studies of mung bean, where protein content ranged from 81.5% to 87.8% (Kudre et al., 2013; Brishti et al., 2017).

Table 1.

Major components and antinutritional factors of mung bean flour and protein isolates. Different letters indicate significant differences (p ≤ 0.05). Values are means ± standard deviation (n = 3).

Sample	Flourb	IP8	IP9	MP	H0.25	H0.50	H0.75
Chemical composition
Asha[%]	3.29 ± 0.06	2.61 ± 0.01c	2.76 ± 0.11c	3.61 ± 0.14a	3.26 ± 0.09b	3.38 ± 0.06ab	3.29 ± 0.04b
Proteina[%]	26.78 ± 0.15	97.58 ± 0.30b	95.91 ± 0.09c	99.47 ± 0.80a	96.99 ± 0.13bc	97.68 ± 0.18b	98.02 ± 0.06b
Moisturea[%]	9.52 ± 0.74	2.33 ± 0.12abc	2.03 ± 0.13bc	2.48 ± 0.24a	2.28 ± 0.07abc	2.38 ± 0.03ab	1.91 ± 0.16c
Starcha[%]	54.28 ± 0.69	0.12 ± 0.00c	0.46 ± 0.01a	0.18 ± 0.00b	0.46 ± 0.02a	0.18 ± 0.00b	0.18 ± 0.01b
Fata [%]	1.09 ± 0.03	1.05 ± 0.03c	1.29 ± 0.02b	1.68 ± 0.03a	0.32 ± 0.01f	0.58 ± 0.00d	0.52 ± 0.01e
Antinutritional factors
Phytic acida [g/100g]	1.05 ± 0.01	2.05 ± 0.03c	2.22 ± 0.06c	0.54 ± 0.01d	3.06 ± 0.02a	2.93 ± 0.06ab	2.84 ± 0.12b
TIAa [TIU/mg protein]	20.51 ± 0.71	12.21 ± 0.14a	11.63 ± 0.24b	2.42 ± 0.07c	2.47 ± 0.09c	2.19 ± 0.05c	2.39 ± 0.04c

Abbreviations: IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl; TIA = Trypsin inhibitor activity; TIU = Trypsin inhibitor units.

^aAll values are related to the dry weight of the respective sample.

^bFlour was not included in the Anova to focus on the variations between the different protein isolates.

Highest ash contents were found for MP (3.6% dw) and hybrid isolates (3.3%–3.4% dw) which is most likely due to the addition of salt during extraction (Table 1). Furthermore, the ash content increased with pH during alkaline extraction due to the addition of NaOH and HCl (Muranyi et al., 2013), as shown by the comparison of IP8 and IP9. The determined fat content of defatted mung bean flour was low (1.1% dw). A slight accumulation of fat was observed in IP8, IP9, and MP, while hybrid isolation technique marginally reduced the fat content.

3.3. Protein and product yield

Particularly with regard to economic aspects, product and protein yield are important factors for selecting the appropriate isolation method. Protein yields ranged from 30.4% to 67.5% dw for MP and IP9, respectively (Fig. 2). These results are comparable to those of Rahma et al. (2000), who found protein yields between 40.9% and 66.5% for micellized and isoelectric mung bean protein isolates, respectively. Product yield was lowest for MP (8.3%) and increased to a maximum of 18.8% for IP9. However, the product yield for all hybrid isolates was increased significantly (p ≤ 0.05) about 1.8 times compared to MP isolate.

Fig. 2.

Effect of extraction techniques on protein purity, protein yield, and product yield. Different letters indicate significant differences (p ≤ 0.05). Values are means ± standard deviation (n = 3). Abbreviations: IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl.

While protein yield increased with increasing pH, protein purity decreased slightly but significantly (p ≤ 0.05) with increasing pH in case of IP (Fig. 2). This decline in protein purity can be attributed to the higher solubility of non-protein components in alkaline pH regions (Das et al., 2021). Thus, there is a negative relationship between protein yield and protein purity, which has also been found in studies that have examined other protein sources like lupin, rapeseed, or amaranth (Tenorio et al., 2018; Das et al., 2021). Accordingly, micellar extraction provided the isolate with the highest protein purity but also with the lowest protein and product yield. In addition, an increase from 0.25 M to 0.75 M NaCl during extraction resulted in improved protein purity of hybrid isolates and slightly increased protein yields, consistent with Pickardt et al. (2009). Protein purity of H0.75 (0.75 M NaCl) is comparable to MP isolate, which makes this isolation procedure a potential method to obtain isolates with high purity.

3.4. Antinutritional factors

The mung bean flour exhibited a phytic acid content of 1.05 g/100 g dw sample (Table 1). This finding falls within the range of phytic acid contents observed in various mung bean genotypes and varieties, which typically range from 0.57 to 1.90 g/100 g (Dhole and Reddy, 2015). The isolation methods employed in this study had a strong impact on the phytic acid content. In particular, isolates IP8, IP9, and H0.25 to H0.75 showed an accumulation of phytic acid, while the MP isolate exhibited a reduction. Among the isolates, the highest phytic acid content was observed for H0.25 to H0.75, which had approximately three times higher phytic acid content compared to the defatted mung bean flour (Table 1).

Phytic acid is soluble in a wide range of pH and mostly negatively charged at pH values found in foods because of the six phosphate groups with different pK_a values (Wang and Guo, 2021). With increasing pH, the negative charge of phytic acid increases, so that the interaction with negatively charged proteins is reduced in alkaline pH ranges. However, indirect linkage of protein and phytate can still occur through complex formation via multivalent cations (Wang and Guo, 2021). When the pH is decreased to the isoelectric point, the proteins negative net charge is neutralized, leading to a reduction in repulsive forces. As a result, a greater number of electrostatic interactions can occur, leading to the formation of more protein-phytate complexes. By comparing MP and hybrid isolates it can be seen, that mainly the precipitation method had an influence on the phytate content (Table 1). As MP was consistently extracted and isolated in the region of pH 6, where the net charge of the proteins is more negative, less phytic acid was bound. Precipitation of hybrid isolates at pH 4 favored the interaction between the proteins and phytic acid. Considering IP8 and IP9, alkali extraction and isoelectric precipitation at pH 5 resulted in lower phytic acid levels than hybrid isolation.

Furthermore, it seems that an increase in salt concentration from 0.25 M to 0.75 M NaCl had a slightly reducing effect on the phytate content in the different isolates. An increase in salt concentration generally hinders the protein-phytate complexation (Wang and Guo, 2021). Therefore, a small quantity of phytic acid could be separated after the first centrifugation step during hybrid isolation. However, the phytate reducing effect of NaCl at 0.75 M was marginal (Table 1) compared to H0.25 and H0.50 for effective phytic acid reduction.

The mung bean flour exhibited a trypsin inhibitor activity (TIA) of 20.51 TIU/mg protein dw (Table 1). This value was higher compared to literature data obtained by Mubarak (2005) where TIU/mg protein of dehulled mung bean was 14.6 TIU/mg protein. The isolation method had a strong influence on the TIA of the isolates. Compared to the flour, alkali-extracted protein isolates showed a reduction in TIA by almost half. The increase from pH 8 to pH 9 in the extraction medium had a slightly decreasing effect on TIA. Micellization and hybrid isolation were able to reduce TIA by up to 90% compared to the flour. It can be assumed that the extractability of trypsin inhibitor at pH 6 is reduced and, therefore, trypsin inhibitor was successfully separated after the first centrifugation step for micellar and hybrid isolates. In a study of Chrispeels and Baumgartner (1978), the isoelectric point of trypsin inhibitor from mung bean was defined at pH 5.05, which supports the previous thesis.

3.5. Color analysis

With the exception of MP and the flour, all samples showed positive values for a* and b* (Table 2). The high b* values indicate a yellow hue with a slight reddish hue due to small a* values, which is consistent with the findings of Brishti et al. (2020). MP exhibited a slightly greenish hue expressed by the negative a* value, most similar to the flour. The highest L* values were observed for MP and the hybrid samples. Under alkaline conditions, the extraction of polyphenols is promoted, leading to the formation of brown quinones that can react with proteins, resulting in brown coloration (Xu and Diosady, 2002). This is possibly the reason for the higher browning index of the IP isolates compared to MP and the hybrid isolates (Table 2). Therefore, the lower extraction pH of micellar and hybrid isolation method had a positive impact on the brightness of the mung bean protein isolates. Similar effects were observed in sunflower protein isolate, where lower pH values in NaCl-assisted extraction prevented the covalent binding of phenolic compounds to proteins (Pickardt et al., 2009).

Table 2.

Color parameters of mung bean flour and protein isolates. Different letters indicate significant differences (p ≤ 0.05). Values are means ± standard deviation (n = 3).

Sample	Floura	IP8	IP9	MP	H0.25	H0.50	H0.75
L*	89.71 ± 0.12	74.81 ± 0.05d	74.40 ± 0.01d	82.31 ± 0.16a	79.30 ± 0.11c	80.50 ± 0.02b	79.49 ± 0.52c
a*	−1.49 ± 0.08	2.60 ± 0.05b	2.83 ± 0.02a	−0.11 ± 0.01e	2.24 ± 0.03c	1.84 ± 0.02d	1.88 ± 0.08d
b*	16.06 ± 0.55	30.44 ± 0.04a	29.41 ± 0.02b	25.06 ± 0.25c	24.40 ± 0.22cd	23.97 ± 0.05d	24.83 ± 0.53c
Browning index	0.55 ± 0.03	6.49 ± 0.05a	6.60 ± 0.01a	2.88 ± 0.04d	5.03 ± 0.04b	4.55 ± 0.02c	4.75 ± 0.16c

Abbreviations: IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl.

^aFlour was not included in the Anova to focus on the variations between the different isolates.

3.6. SDS-PAGE analysis

Mung bean storage proteins mainly consist of 60% globulins and 25% albumins (Yi-Shen et al., 2018). The predominant storage protein found in mung beans is the 8S globulin, also known as vicilin. It is estimated to make up approximately 89% of the total globulin content in mung beans. The remaining globulin fraction is composed of the 11S globulin (8%) and the 7S globulin (3%) (Mendoza et al., 2001). SDS-PAGE (Fig. 3) was performed to analyze the changes in protein composition among the different protein isolates. Conventional alkali extraction and isoelectric precipitation is known to accumulate globulins in the resulting protein isolate, since the isoelectric point of globulins is mostly between pH 4 to 5, whereas albumins have a broad solubility profile over a pH range from 2 to 12 (Tanger et al., 2020, Yang et al., 2023). The protein isolates examined in this study demonstrated an accumulation independent of the isolation technique applied and showed comparable band patterns. Six bands were predominantly present under non-reducing conditions: two bands with a molecular weight of 24 and 28 kDa, respectively, a high intensity band ranging from 43 to 50 kDa, and three bands at 57, 62 and 70 kDa, respectively. Four bands (24, 28, 43–50, and 62 kDa) showed no alterations when compared to the SDS-PAGE under reducing conditions, indicating that these bands likely correspond to the 8S globulins, which lack in disulfide bridges (Mendoza et al., 2001). Under reducing conditions, the band at 70 kDa disappeared, while a band at 20 kDa emerged. This band at 20 kDa can be assigned to 11S globulin, which is consistent with the findings of Liu et al. (2015). Mendoza et al. (2001) identified another subunit of 11S globulin at 40 kDa, which could be included in the high-intensity band around 43 kDa in the present study (Fig. 3). Micellization can result in protein isolates consisting of a mixture of globulins and albumins, since the proteins are isolated by micelle formation rather by their isoelectric point (Tanger et al., 2020). Under reducing conditions, characteristic albumin bands from mung bean were identified by Yang et al. (2023) at 25 kDa and between 70 and 95 kDa. In the present samples, bands at 25, 80, and 95 kDa were detected indicating the presence of albumin, although they were less dominant compared to the globulin bands. This demonstrates that some albumins were present in the different isolates. However, despite isolation by micellization, an albumin accumulation could not be observed in MP. The protein patterns in all isolates were comparable, suggesting that the techniques did not exhibit selectivity for specific globulin or albumin fractions.

Fig. 3.

SDS-PAGE of the mung bean protein isolates obtained by different isolation methods under reducing and non-reducing conditions. Abbreviations: Alb = albumin; R = reducing conditions; NR = non-reducing conditions; IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl.

3.7. Protein solubility of the protein isolates

Protein solubility of the final protein ingredients influences their emulsifying activity, foaming capacity, and water holding properties (Deng et al., 2011). Therefore, solubility characteristics are highly important for technological application of protein isolates in food products. As illustrated in Fig. 4, the isolation procedure had an impact on the protein solubility. In the alkaline range above pH of 8, all isolates exhibited a solubility higher than 87%. The lowest solubility of all isolates was observed in the range from pH 4 to pH 6, which is consistent with previous studies on mung beans (Rahma et al., 2000; Brishti et al., 2017). The comparison between IP8 and IP9 revealed that IP8 had a significantly (p ≤ 0.05) higher solubility, particularly between pH 3 and 6, than IP9, which can be attributed to the lower degree of denaturation, due to less harsh extraction conditions (Tanger et al., 2020). IP8 showed the lowest protein solubility at pH 5 (17.0%) and IP9 between pH 4 and 6 (12.3–19.0%) (Fig. 4).

Fig. 4.

pH-solubility curves of mung bean protein isolates obtained by different isolation methods. Values are means ± standard deviation (n = 3). Abbreviations: IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl.

Comparing the solubility profiles of MP, IP, and hybrid isolates, MP revealed the highest protein solubility profile, with an isoelectric point at pH 6, where solubility was the lowest at 14.2% (Fig. 4). There was a shift in the isoelectric point from pH 4 to pH 6, for H0.25, H0.50, H0.75 to MP respectively. This shift can be attributed to the decreasing phytic acid concentration in these samples, as discussed in Section 3.4. Ali et al. (2010) detected a relationship between reduced phytic acid content and enhanced solubility of soybean protein within the pH range from 2 to 4.5. This improved solubility can be attributed to a reduction in the formation of insoluble protein-phytate complexes. The width of the solubility curves for hybrid isolates was comparable to that of the IP isolates. Among the different hybrid isolates, solubility improved with increasing salt concentration during extraction. In particular, at acidic pH values of 3 and 6, H0.75 showed significantly (p ≤ 0.05) higher solubility compared to H0.25 (Fig. 4). However, at pH 2 and 7, the solubility of H0.75 was equivalent to that of H0.25.

3.8. Thermal properties

The effects of the different isolation techniques on the thermal stability and denaturation state of the protein isolates were determined via DSC measurement. Thermal properties like denaturation temperature (T_D) and enthalpy of denaturation (ΔH) are presented in Table 3. The denaturation temperature (T_D) represents the thermal stability of the protein isolate, while ΔH represents the energy required to initiate the denaturation process. These parameters are influenced by factors such as protein structure and conformation, protein content, pH, and ionic strength (Arntfield and Murray, 1981; Murray et al., 1985).

Table 3.

Physicochemical and structural properties of mung bean protein isolates obtained by different isolation methods. Different letters indicate significant differences (p ≤ 0.05). Values are means ± standard deviation (n = 3).

Sample	IP8	IP9	MP	H0.25	H0.50	H0.75
Thermal characteristics
TDa [°C]	86.13 ± 0.07a	86.13 ± 1.20a	82.88 ± 0.28b	84.23 ± 0.67ab	85.43 ± 1.09ab	84.43 ± 0.34ab
ΔHa [J/g protein]	10.89 ± 0.12a	3.77 ± 1.18c	11.72 ± 1.65a	6.94 ± 0.15b	4.63 ± 0.53bc	7.21 ± 1.09b
Surface hydrophobicity
S0a	4209.10 ± 8.77b	4613.15 ± 124.10a	3804.30 ± 16.55c	3267.65 ± 107.13d	4452.65 ± 57.06ab	3762.75 ± 110.95c
Secondary Structure (FT-IR)
β-Sheet [%] (Peak)	42.31 ± 3.08b (1632 cm−1)	40.38 ± 0.46b (1632 cm−1)	48.38 ± 0.20a (1632 cm−1)	39.43 ± 1.34b (1632 cm−1)	43.75 ± 1.36ab (1632 cm−1)	41.30 ± 2.72b (1628 cm−1)
Random structure and α-helix [%] (Peak)	50.93 ± 2.17a (1658 cm−1)	52.67 ± 0.14a (1658 cm−1)	46.13 ± 0.23b (1658 cm−1)	54.04 ± 0.58a (1658 cm−1)	51.03 ± 1.08a (1658 cm−1)	52.56 ± 1.55a (1658 cm−1)
β-Turn [%] (Peak)	6.76 ± 0.92a (1688 cm−1)	6.95 ± 0.32a (1684 cm−1)	5.49 ± 0.13a (1688 cm−1)	6.53 ± 0.78a (1684 cm−1)	5.22 ± 0.30a (1688 cm−1)	6.15 ± 1.18a (1688 cm−1)
Water and Oil binding capacity
WACa [g water/g protein]	2.04 ± 0.01b	2.33 ± 0.09a	1.62 ± 0.02c	1.60 ± 0.06c	1.71 ± 0.06c	1.73 ± 0.08c
OACa [g oil/g protein]	1.90 ± 0.01a	1.93 ± 0.06a	1.85 ± 0.07a	1.39 ± 0.01d	1.52 ± 0.05c	1.65 ± 0.02b
W/Oa index	1.08 ± 0.01b	1.21 ± 0.05a	0.88 ± 0.03c	1.15 ± 0.03ab	1.13 ± 0.01ab	1.05 ± 0.04b

Abbreviations: IP = alkali extracted and isoelectric precipitated mung bean protein isolate at pH 8 (IP8) and pH 9 (IP9); MP = micellized mung bean protein isolate; H = hybrid isolate prepared with 0.25 M (H0.25), 0.50 M (H0.50), 0.75 M (H0.75) NaCl.

^aValues are related to the dry weight of the respective sample.

The T_D values ranged from 82.9 to 86.1 °C for IPs and MP, respectively, whereas enthalpy of denaturation was significantly (p ≤ 0.05) highest for MP and IP8 (Table 3). Higher enthalpy values indicate a more native state of protein (Tanger et al., 2020). Micellization is known as a milder isolation method than alkaline extraction and isoelectric precipitation, leading to less denaturation and more ordered globular structures, which can be associated with the higher ΔH values (Muranyi et al., 2013). Protein isolates from cowpea and pigeonpea demonstrated that micelle isolates had higher enthalpy values than isoelectric prepared protein isolates, but not necessarily higher denaturation temperatures (Mwasaru et al., 1999). High T_D values are associated with high thermal stability and more hydrophobic interactions (Arntfield and Murray, 1981). The enhanced thermal stability observed in IP8 and IP9 could potentially be attributed to partial unfolding that occurred during the isolation process. Furthermore, the enthalpy of denaturation for IP8 and IP9 increased as the extraction pH decreased, aligning with findings from previous studies on amaranth, cowpea, and pigeonpea (Mwasaru et al., 1999; Das et al., 2021). In general, an increase in extraction pH lowers ΔH as described by Arntfield and Murray (1981) due to increased unfolding and denaturation of proteins. The transition enthalpy of the hybrid isolates was slightly higher than for IP9, but significantly (p ≤ 0.05) lower than for IP8 and MP ranging from 4.6 to 7.2 J/g protein dw (Table 3). However, no significant (p ≤ 0.05) differences were observed among the hybrid samples regarding T_D and ΔH. Consequently, it can be concluded that increasing the salt concentration from 0.25 to 0.75 M did not provide additional stabilization of the protein structure and that precipitation at pH 4 led to structural and conformational changes of the hybrid isolates.

3.9. Surface hydrophobicity

Surface hydrophobicity (S₀) represents the number of hydrophobic groups located on the surface of the proteins. On the one hand, denaturation of proteins leads to partial unfolding and exposes hydrophobic groups on the protein surface, which increases surface hydrophobicity (Kato and Nakai, 1980). On the other hand, increasing denaturation leads to the association of these hydrophobic groups, which decreases surface hydrophobicity (Wang et al., 2014). Therefore, S₀ helps to visualize the state of aggregation and denaturation of protein samples. In addition, surface hydrophobicity can be related to technofunctional properties such as protein solubility, emulsifying capacity, as well as oil absorption capacity (Deng et al., 2011).

The surface hydrophobicity of mung bean protein isolates at pH 7 ranged from 3267.7 to 4613.2 for H0.25 and IP9, respectively (Table 3). When comparing the alkaline extracted and isoelectric precipitated protein isolates (IP 8 and IP9) with the micellar protein (MP), it becomes evident that MP had a significantly (p ≤ 0.05) lower surface hydrophobicity compared to IP8 and IP9. This suggests that the hydrophobic groups of MP are primarily located inside the protein structure. Muranyi et al. (2013) reported the formation of micellized lupin protein with a hydrophilic shell and a hydrophobic core during dilutive precipitation, which is achieved by arranging hydrophobic regions inward to attain the most energetically favorable state. In contrast, the extraction conditions were harsher for IP8 and IP9. The abrupt transition from the extraction pH to the precipitation pH of 5 did not provide enough time for the proteins to properly position their hydrophobic groups towards the interior. For this reason, IP9 may have more hydrophobic groups exposed on its surface than IP8 and MP. The surface hydrophobicity values of the hybrid isolates were affected significantly (p ≤ 0.05) by the different salt concentrations (Table 3), but did not appear to follow a consistent trend. Muranyi et al. (2013) reported that the protein structures resulting from the combination of salt extraction and isoelectric precipitation of lupin protein isolates consist of a mixture of micellized proteins and isoelectric irregular protein aggregates. Therefore, the surface hydrophobicity of H0.25, H0.50, and H.075 could be attributed to the presence of both micellized structures and isoelectrically agglomerated structures. The isolation method and varying NaCl concentrations had a significant (p ≤ 0.05) impact on the surface hydrophobicity of the isolates, thereby impacting the conformation of the proteins.

3.10. Fourier transform infrared spectroscopy

The composition of the secondary structure is presented in Table 3. Supplementary information (Fig. S1) provides comprehensive details on the second deviation and the Gaussian curve-fitting. The FT-IR spectra of the protein isolates were similar with slight variations in the absorption intensities. Peaks were observed in the following ranges: 1628-1632 cm⁻¹, linked to β-sheet structures; 1658 cm⁻¹, assigned to α-helix and random structures; and 1684-1688 cm⁻¹, defined as β-turn structures (Fevzioglu et al., 2020). The detected peak in the frequency range of 1660–1640 cm⁻¹ was assigned to α-helix and random structures, due to the overlapping peaks of α-helix and random structures in this region (Fevzioglu et al., 2020). All samples had high contents of β-sheet, α-helix, and random structures. MP isolate exhibited the significantly (p ≤ 0.05) highest percentage of β-sheet (48.4%) and lowest percentage of α-helix and agglomerated structures (46.1%) compared to all other isolates, which is mostly attributed to the milder extraction process and is consistent with the findings for surface hydrophobicity and denaturation enthalpy in the present study. According to Mune Mune et al. (2017), the β-turn structures are product of protein unfolding of higher ordered structures. The highest percentage of β-turn structures was found for IP9 (6,9%) which is also in line with the findings for denaturation enthalpy. However, percentages of secondary structure of IP8, IP9, and hybrid isolates showed no significant (p ≤ 0.05) differences, indicating that the different pH values and salt concentrations during extraction did not have a significant effect on the secondary structure.

3.11. Water and oil absorption capacity

The water and oil absorption capacity of a protein isolate determines its technofunctional properties and plays an important role in food formulations. Both water and oil absorption capacity influence food texture due to fat and water retention and are correlated to emulsion capacity and stability (De Kanterewicz et al., 1987; Elizalde et al., 1988). High WAC is important to bind water in food formulations, which in turn influences thickening and viscosity in dough, whereas OAC is important for fat entrapment in meat and flavor retention (Kinsella, 1979).

WAC ranged from 1.6 to 2.3 g water/g protein dw for H0.25 and IP9, respectively (Table 3). The highest WAC was observed for IP9 and IP8, indicating a greater availability of hydrophilic amino acids for interaction with water. A higher WAC has also been associated with a higher degree of protein denaturation due to the increased surface to mass ratio of denaturized protein and, thus, with a lower protein solubility (Muranyi et al., 2016). As discussed in Section 3.8, isolation at pH of 8 instead of 9 resulted in a less denaturized sample which had a slight effect on the WAC of the both IP isolates. The same effect was seen for MP, which had lower WAC compared to IPs (Table 3). However, WAC of the hybrid isolates did not differ significantly (p ≤ 0.05) from MP isolate, indicating that the WAC is influenced by multiple factors.

The OAC of the isolates ranged from 1.4 to 1.9 g oil/g protein dw for H0.25 and IP9/IP8 (Table 3), respectively. OAC was the highest and quite similar for IPs and MP, while a significant (p ≤ 0.05) reduction was observed in hybrid isolates. A high content of β-sheet is related to higher OAC values, which is in line for MP (Mune Mune et al., 2017). However, also a low degree of denaturation enhances protein flexibility and improves the ability of proteins to align their hydrophobic domains to the oil phase (De Kanterewicz et al., 1987; Muranyi et al., 2016). This could be the reason for higher OAC values of IP8. However, IP9 was the most denaturized sample with comparable OAC values to IP8 and MP. As discussed in Section 3, 3.9.10, sample IP9 had the highest values for surface hydrophobicity and the highest percentage of β-turn structures. Therefore, the OAC value of IP9 could be explained by the partial unfolding and exposure of previously hidden hydrophobic regions to the proteins surface. This in turn promoted the interaction between lipids and proteins and resulted in a higher OAC value (Mune Mune et al., 2017). Increasing NaCl concentration during salt-induced extraction from 0.25 M to 0.75 M resulted in a slight significant (p ≤ 0.05) increase in OAC.

A balanced ratio of hydrophilic and lipophilic properties of protein isolates contribute to improved emulsifying properties. Therefore, De Kanterewicz et al. (1987) calculated the W/O index and found a correlation to the emulsion capacity. The emulsion capacity is higher in the range of two, when a protein can absorb twice as much water as oil. It has to be considered that the W/O index is not directly attributed to the emulsion stability, which is more influenced by the amount of hydrophobic and hydrophilic groups of the protein. The W/O index for analyzed samples in our study ranged from 0.9 for MP to 1.2 for IP9 and H0.25 isolates (Table 3), indicating that the balance between hydrophilic and lipophilic properties is more on the hydrophilic side, except for the MP isolate. The W/O indices of our samples are comparable to bean protein isolate analyzed by De Kanterewicz et al. (1987) with a W/O index of 1.3. Therefore, IP9 and H0.25 would probably be the best choice to achieve a good emulsion capacity. The different salt concentrations during preparation of hybrid isolates did not significantly (p ≤ 0.05) change the W/O index.

4. Conclusions

Alkali extraction of mung bean proteins at pH 9 and their isoelectric precipitation at pH 5 resulted in protein products with the highest protein yield of 67.5%. The addition of NaCl during extraction significantly increased the extractability of mung bean protein in the salt-assisted isolation procedures. However, the resulting salt concentration after dilution decisively determines the yield of precipitated protein, since part of the proteins is kept in solution by the salt. Therefore, dilution in combination with isoelectric precipitation did not increase the protein yield of the hybrid isolates beyond 54.8%. The product yields obtained from the different isolation methods ranged from 8 to 19%, highlighting the need for identifying potential utilization of the side products. This is crucial for developing sustainable and economically viable protein isolation methods. All isolation methods yielded protein-rich isolates with protein contents exceeding 95%. Micellization had the highest potential to generate a highly functional protein isolate, as it effectively reduced the amount of antinutritive compounds and improved the re-solubility of the protein isolate. However, the low product yield of micellized protein isolate is an economic drawback for large-scale applications. Lowering the extraction pH from 9 to 8 resulted in a less denaturized protein structure, as determined by DSC measurements, and improved solubility behavior of the alkali extracted proteins. As hypothesized, salt extraction in combination with dilutive and isoelectric precipitation, combined some attributes of both conventional isolation techniques. Therefore, hybrid isolation increased protein yield compared to micellization while producing an isolate with lighter color and lower trypsin inhibitor activity than the alkali extracted protein isolate. The surface hydrophobicity of the hybrid isolates suggested the occurrence of structural modifications due to extraction performed at varying NaCl concentrations. However, only micellization showed significant differences in the secondary structure. Regarding antinutritional factors, alkaline extraction with isoelectric precipitation and hybrid isolation were not sufficient to reduce the phytic acid content. Additional steps, such as acid pre-extraction or germination, should be considered to produce mung bean protein isolates with low phytic acid content. This study highlights the importance of considering structural changes, antinutritive factors, solubility and preservation of the proteins native state when selecting an appropriate extraction method. Further investigations of emulsifying, foaming, and gelling properties are necessary to comprehensively evaluate the technological applications of the protein isolates.

CRediT authorship contribution statement

Christina Wintersohle: Conceptualization, Methodology, Formal analysis, Writing – original draft, Visualization. Inola Kracke: Formal analysis, Investigation. Laura Melanie Ignatzy: Formal analysis, Investigation. Lara Etzbach: Conceptualization, Writing – review & editing, Supervision. Ute Schweiggert-Weisz: Conceptualization, Writing – review & editing, Supervision, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the supplementary data to this article:

Data availability

Data will be made available on request.