Antioxidative properties and macrochemical composition of five commercial mungbean varieties in Australia

Mungbeans are growing in popularity among Australian consumers, driven by their beneficial nutritional and phytochemical composition. However, data on the antioxidative, mineral, and phytochemical content of Australian mungbeans at the point of consumer purchase remains scarce. Here, five commercial mungbean samples were analysed for total antioxidant capacity, total phenolics, and total monomeric anthocyanins. Attenuated total reflectance midinfrared spectroscopy was utilised as a rapid and reliable method of obtaining information about the macrochemical composition of the mungbean hulls. Total antioxidant capacity ranged from 170 to 570 mg Trolox equivalents per 100 g, total phenolic content from 130 to 240 mg gallic acid equivalents per 100 g, and anthocyanin content from 10 to 40 mg cyanidin‐3‐glucoside equivalents per 100 g. There was a significant difference between varieties in all measures of antioxidant, phenolic and anthocyanin contents. Using principal component analysis, the midinfrared spectra for the five mungbean varieties could be isolated, highlighting the differences in their phytochemical composition. In general, whole Australian mungbeans appear to have the highest antioxidant, phenolic, and anthocyanin contents. Midinfrared spectroscopy appears to be a valuable method of obtaining and comparing the macrochemical composition of mungbeans. This technology is likely to be of increasing use in the future.

Several studies have been conducted on the nutritional qualities of various mungbean varieties in China (Liu, Liu, Yan, Cheng, & Kang, 2015;Shi, Yao, Zhu, & Ren, 2016;Zhang, Shang, Qin, Zhou, Gao, Huang et al., 2013), India (Dahiya, Linnemann, Nout, Van Boekel, & Grewal, 2013), Myanmar (Kywe, Finckh, & Buerkert, 2007), and Pakistan (Ullah, Ullah, Al-Deyab, Adnan, & Tariq, 2014). These included the quantification of amylose, starch, protein, fibre, ash, carbohydrate, and minerals and the characterisation of constituent amino acids and fatty acids. Varietal influences on the antioxidant properties of mungbeans have also been studied in China (Shi et al., 2016;Zhang et al., 2013) and Myanmar (Ullah et al., 2014). However, information on the phytochemical or antioxidant composition of Australian mungbeans remains very limited. Moreover, all previously cited antioxidant studies, except for the Chinese study by Zhang et al. (2013) and Mamilla and Mishra (2017), who studied a single, unidentified variety, have used mungbeans obtained from farms or agricultural/government agencies, rather than from the point of supply to the consumer. Due to postharvest influences including handling, packaging, storage times, splitting, and other processing methods, the chemical composition of freshly harvested mungbeans does not necessarily represent the composition when they are consumed (Dahiya et al., 2015). Therefore, this study aimed to characterise some of the variation in the nutritional and antioxidant composition of mungbean varieties available to Australian consumers at the point of purchase. It is hoped that this work will highlight the potential of important phytochemical compounds, such as antioxidants, to further differentiate and substantiate the quality of mungbean samples in a manner beyond that afforded by traditional visual and physical attributes.

| Mungbean samples
Five commercial mungbean samples were purchased from two local supermarkets, as available to consumers. Table 1 gives the brand details and country of origin for each sample.
Mungbean samples were ground to a homogenous flour using a Retsch (Sydney, Australia) ZM1000 centrifugal grinding mill with a 1.0-mm mesh. This flour was used for all subsequent analyses except for midinfrared spectroscopy attenuated total reflectance (MIR-ATR).

| Reagents
All reagents used were analytical-grade. Methanol was purchased from Fisher Scientific Australia (Sydney, Australia). Hydrochloric acid and sodium carbonate were purchased from Chem Supply (Adelaide, Australia). All other reagents were purchased from Sigma-Aldrich Australia (Sydney, Australia). Unless otherwise specified, all dilutions and assay preparations were made using Milli-Q ® water (Merck Millipore; Sydney, Australia). All solutions were stored in the refrigerator at 4 C until usage.

| Extraction of antioxidant compounds
Following protocols recently developed in our laboratory (Johnson, Collins, Skylas, & Naiker, 2019), duplicate extracts were prepared from around 2.5 g of sample and 45 ml of 90% v/v aqueous methanol. After vortexing briefly, the extracts were mixed for 60 min using an end-over-end shaker (Ratek RM4; Melbourne, Australia) operating at 50 rpm. The samples were then centrifuged at 1000 g for 10 min using a Heraeus Multifuge (Thermo Fisher Scientific; Sydney, Australia) prior to collecting the supernatant. The extraction was repeated on the pellet with another 45 ml of 90% methanol and end-over-end mixing for 20 min to extract any remaining antioxidative compounds. The combined supernatants were vacuum filtered using a 0.45-μm Advantec ® filter paper (Toyo Roshi Kaisha; Tokyo, Japan) and volumetrically made up to 100 ml with 90% methanol. Extracts were stored in the dark at 4 C until required for further analyses.

| Cupric reducing antioxidant capacity
The total antioxidant capacity was determined using a modification of the cupric reducing antioxidant capacity (CUPRAC) method developed by Apak et al. (2013). To perform the CUPRAC analysis, 1 ml of 10-mM aqueous copper (II) chloride, 1 ml of 1-M aqueous ammonium acetate, 1 ml of Milli-Q ® water, and 1 ml of freshly prepared 7.5-mM neocuproine ethanol solution were combined with 100 μL of the sample extract. After vortexing for 30 s, the samples were incubated in a covered water bath at 50 C for 30 min. The resulting absorbances were read at 450 nm using an ultraviolet spectrophotometer (Thermo Scientific Genesys 10S UV-Vis; Sydney, Australia) blanked with Milli-Q ® water. The CUPRAC was derived as a function of the equivalent absorbance of Trolox (6-hydroxy-2,-5,7,8-tetramethyl-chromane-2-carboxylic acid), a vitamin E analogue (Scott, Cort, Harley, Parrish, & Saucy, 1974) in ethanol solution in the range of 50 to 600 mg L −1 (R 2 = .99).

| Ferric reducing antioxidant power
As it is recommended to use at least two methods for measuring total antioxidant capacity (Apak et al., 2013;Bartosi nska, Buszewska-Forajta, & Siluk, 2016), the ferric reducing antioxidant power (FRAP) assay developed by Benzie and Strain (1996) was also performed on the sample extracts.
The FRAP was derived as a function of the equivalent absorbance of Trolox in ethanol solution in the range 10 to 175 mg L −1 (R 2 = .99).

| Total phenolics
Total phenolics were determined through a modification of the Folin-Ciocalteu method developed by Singleton and Rossi (1965). To perform the total phenolic assay, 2 ml of a 1:10 aqueous Folin-Ciocalteu reagent dilution was combined with 400 μL of sample extract, followed by incubation in darkness at room temperature for 10 min.
After adding 2 ml of 7.5% w/v aqueous sodium carbonate solution and vortexing for 30 s, the samples were incubated in a covered water bath at 40 C for 30 min. Their absorbance at 760 nm was then measured. The total phenolic concentration was derived as a function of the equivalent absorbance of gallic acid in the range 20 to 120 mg L −1 (R 2 = .99).

| Total monomeric anthocyanins
Total monomeric anthocyanins were determined using a modification of the pH differential method described by Cheng and Breen (1991). Aqueous buffer solutions at pH 1 and 4.5 were prepared with 0.025-M KCl and 0.4-M sodium acetate, respectively. Sample extract (500 μL) was combined with 2 ml of buffer and mixed. After equilibrating at room temperature in the darkness for 15 min, the absorbance at 510 and 700 nm was then measured.
The monomeric anthocyanin concentration was calculated using the following formula (Cheng & Breen, 1991 The molecular weight (449.38 g mol −1 ) and molar extinction coefficient (26,900 M −1 cm −1 ) of cyanidin-3-glucoside were used, it being the most abundant anthocyanin in nature (Markakis, 1989) and also the most abundant anthocyanin present in mungbeans (Yao et al., 2013).
Results from all antioxidant tests were expressed as milligrams of the equivalent standard per 100 g of oven-dried mungbean sample.

| Carbon, nitrogen, and protein content analysis
Nitrogen and carbon contents were determined using a LECO TruMac Series Carbon and Nitrogen Analyser (Sydney, Australia). Following previous work on mungbeans (Skylas et al., 2017), protein content was estimated by multiplying the nitrogen content by a conversion factor of 6.25. Results were expressed on an oven-dry weight basis.

| Minerals and trace element analysis
Approximately 1.0 g of each mungbean sample was microwave digested (Biobase BMD-2; Jinan, China) in duplicate with 5 ml of 70% nitric acid. At a pressure of 400 PSI, the temperature was ramped to 90 C over 5 min followed by a 5-min hold time. The temperature was then ramped to 170 C over another 5 min then held for a final 10 min. After cooling, digests were made up to 50 ml with Milli-Q ® water. Trace elements and minerals contents were determined through inductively coupled plasma-mass spectroscopy on a Shimadzu inductively coupled plasma-mass spectrometer 2030 (Sydney, Australia). Each duplicate was analysed in triplicate on the inductively coupled plasma-mass spectrometer. MIR spectra between 4,000 and 400 cm −1 were recorded using the OPUS software Version 7.5 (Bruker Optics Gmbh, Ettlingen, Germany) as the average of 24 scans at a resolution of 4 cm −1 . Three replicates were performed on each mungbean variety.

| Statistical analysis
Statistical tests were performed using IBM SPSS (New York, USA). All error bars show ±2 standard deviations.
MIR spectra were analysed with the Unscrambler X software Version 10.5 (Camo ASA, Oslo, Norway). Based on previous work on barley seed (Gordon et al., 2019), the spectra were preprocessed to the second derivative using a Savitzky-Golay algorithm at a polynomial number of 2 and a smoothing window of 41 points (Savitzky & Golay, 1964). Using the second derivative removes spectral variations in the baseline and slope (Savitzky & Golay, 1964), minimising differences due to noncompositional variables such as the pressure and contact with the reflection module. Principal component analysis was performed on the second derivative of the MIR spectra in the Unscrambler X, using full cross-validation.

| Carbon, nitrogen, and protein contents
The protein contents of the mungbean samples ranged from 25.8 to 28.2% (Table 2), similar to those found by Skylas et al. (2017) for Australian-grown mungbeans. There was minimal variation in the carbon contents between varieties. All values are given as mean concentration (mg/kg) ± SD (n = 3 replicates from 2 digests for each). Samples with different letters in the same row were significantly different according to a one-way ANOVA with post-hoc Tukey testing at α = 0.05. variation in sodium, iron, and copper contents. Other elements, such as zinc, showed minimal differences in concentration between varieties. Copper and zinc values were comparable to those found by Anwar et al. (2007). The values of some other elements, in particular calcium, sodium and magnesium, varied considerably from that reported by Anwar et al. (2007). However, this may be due to Anwar et al. (2007) using only hull-less mungbean seeds, suggesting that these elements may be concentrated in the hull.

| Total antioxidant capacity (CUPRAC and FRAP)
The total antioxidant capacity, measured by both the CUPRAC and FRAP methods, is shown in Figures 1 and 2, respectively. There was a significant difference in the mean CUPRAC and mean FRAP of the samples (P < .001 for both). The Pattu whole variety showed the highest antioxidant capacity followed by Katoomba. The FRAP of the mungbeans tested here was lower than that of acetone-water extracts from Chinese mungbeans (Zhang et al., 2013) and ethanol-HCl extracts from South Korean mungbeans (Lee et al., 2011). This may be due to the varying polarity and extraction efficiency of the solvents used. No previous measurements of CUPRAC in mungbeans were found in the literature.

| Total phenolic content
There was a significant difference in phenolic contents between varieties ( Figure 3; P < .001). The total phenolic content followed the same general trend as total antioxidant capacity, with the whole Pattu and Katoomba samples containing the highest phenolic levels. Overall, the five mungbean varieties analysed here showed much higher total phenolic levels than those reported for methanolic extracts of mungbeans grown in China but slightly lower than total phenolic levels from acetone-water extracts from the same country (Zhang et al., 2013).
F I G U R E 1 Average total antioxidant capacity of the five mungbean varieties, obtained using the CUPRAC assay. Samples with different letters were significantly different according to a oneway analysis of variance with post hoc Tukey testing at α = 0.05. DW, dry weight; TE, Trolox equivalent F I G U R E 2 Average total antioxidant capacity of the five mungbean varieties, obtained using the FRAP assay. Samples with different letters were significantly different according to a one-way analysis of variance with post hoc Tukey testing at α = 0.05. DW, dry weight, TE, Trolox equivalent The phenolic levels were equal to or lower than those found by Shi et al. (2016) using 70% ethanol extracts from 20 Chinese mungbeans.

| Total monomeric anthocyanin content
There was a significant difference in the average total monomeric anthocyanin content of the varieties analysed (Figure 4; P < .001). The anthocyanin content did not display any statistically significantly correlation with the total antioxidant capacity or total phenolic content of the variety (P > .05). The Cock Brand mungbeans had the highest anthocyanin concentration, followed by the Food Sing variety.

| Varietal differences in antioxidant levels
Although  et al. (2014) found that exposing antioxidant-rich radish seedlings to F I G U R E 3 Average total phenolic concentrations of the five mungbean varieties. Samples with different letters were significantly different according to a one-way analysis of variance with post hoc Tukey testing at α = 0.05. DW, dry weight; GAE, gallic acid equivalent F I G U R E 4 Average total monomeric anthocyanins of the five mungbean varieties. Samples with different letters were significantly different according to a one-way analysis of variance with post hoc Tukey testing at α = 0.05. cyd-3-glu, cyanidin-3-glucoside; DW, dry weight light had no effect on the total antioxidant capacity, total phenolic, or α-tocopherol concentration. This hull-loss hypothesis is also supported by Dahiya et al. (2014), who found that compared to dhal made from whole mungbeans, the polyphenol concentrations in split dhal were only 2% lower, but they were 33% lower in dehulled split dhal. Although this analysis was undertaken on a prepared food product, the general trend of antioxidant levels may hold true for raw mungbeans.
Incidentally, the whole Pattu and Katoomba mungbeans were of Australian origin, whereas the split Pattu and Cock Brand were grown in Myanmar and Thailand, respectively. The origin of the Food Sing mungbeans was not specified. Without a larger sample size, it is not possible to determine whether commercially available mungbeans grown in Australia have higher total antioxidant and phenolic levels than imported mungbeans. However, a previous study on barley has shown that prolonged storage typically decreases antioxidant levels (Do, Cozzolino, Muhlhausler, Box, & Able, 2015). Imported mungbeans would likely have greater storage times than Australian-grown mungbeans, due to transport time and different growing seasons in the Northern Hemisphere. Thus, it is possible that this could potentially lower their antioxidant contents.

| MIR spectral analysis
The MIR spectra of the mungbeans before processing are shown in The wide band between 3,600 and 3,020 cm −1 is mainly due to the O─H stretching vibrations of water molecules (Cozzolino, Roumeliotis, & Eglinton, 2014;Gnanasambandam & Proctor, 2000;Gordon et al., 2019;Lai, Wen, Li, Wu, & Li, 2010). The smaller 2,935-2,910 cm −1 peak is due to the antisymmetrical stretching of the C─H bond, from lipids or other compounds (Gnanasambandam & Proctor, 2000;Lai et al., 2010). Indeed, Cozzolino et al. (2014) reported that spectral patterns between 3,000 and 2,720 cm −1 are related to the unsaturated lipid content of the germ of cereals. There was some visible variance between samples in the band of 2,990-2,840 cm −1 , a slightly narrower peak width than that reported by Cozzolino et al. (2014) for barley.
The only obvious feature between 2,840 and 1,770 cm −1 was a tiny but distinguishable peak at 2,370-2,320 cm −1 , likely from aliphatic C─H bonds (Lai et al., 2010).
The largest peak, between 1,130 and 940 cm −1 , has been attributed to O─C stretch vibrations in glucose (Cozzolino et al., 2014) or other pyranose rings (Lai et al., 2010). These rings could be constituents of glycosides, starch, cellulose, or other polysaccharides (Barron & Rouau, 2008;Cozzolino et al., 2014). The C─O─C and C─OH link bonds could also contribute to absorbance in this band (Lai et al., 2010).
There was a reasonable amount of visible differences between mungbean varieties in the band of 1,670-900 cm −1 , very similar to the band of 1,600-800 cm −1 that Cozzolino et al. (2014) reported to display spectral variation between barley varieties. The spectra band on the rightmost side of the graph, between 400 and 800 cm −1 , corresponds to the fingerprint region of the spectra and is difficult to attribute to specific bonds (Gordon et al., 2019;Lai et al., 2010).
Overall, the MIR spectra of the mungbean hulls was very similar to that obtained by Lai et al. (2010) using purified water-soluble polysaccharide fractions extracted from mungbean hulls. Although the amplitude of the peaks obtained here was much lower than that of Lai et al. (2010), the sensitivity of the MIR-ATR instrument used allowed all peaks above 1,000 cm −1 recorded by Lai et al. (2010) to be reliably distinguished in this study. This has an important bearing on future research, as it suggests that particularly for spectral peaks above 1,000 cm −1 , the macrochemical composition of mungbeans can be determined rapidly and reliably using reflectance spectra of the hull rather than performing the laborious process of extracting, isolating, and purifying the desired polysaccharides and other water-soluble constituents before MIR analysis.
F I G U R E 6 Second derivative of the midinfrared spectra of five commercial mungbean varieties F I G U R E 7 Scores plot showing the principal component analysis of the five mungbean spectra. PC, principal component

| Processing of MIR spectra and principal component analysis
As can be seen in Figure 6, the second derivative successfully The loading plot for Principal Component 1 showed that most of the separation was due to spectral differences in the 800-1,600 cm −1 range (Figure 8), similar to the range found by Cozzolino et al. (2014) for barley.

| CONCLUSION
This study found significant variation in the antioxidant composition of commercially available mungbean varieties in Australia. Of the five varieties tested, whole mungbeans grown in Australia had the highest total antioxidant capacity and phenolics concentrations, whereas imported split mungbeans had the lowest. Total monomeric anthocyanins did not appear to correlate with total antioxidant capacity or phenolic levels. MIR spectra obtained using ATR