Seed Coat Color-Mediated Differences in Nutritional Composition and Antioxidant Activity of Mung Bean

Abstract

The mung bean (Vigna radiata) is rich in nutrients and bioactive compounds and is valuable for its antioxidant content in functional food development. However, mung bean seed coats are discarded or used as a low-value feed owing to their coarse texture. Here, 12 homozygous mung bean lines with different seed coat colors were selected from six recombinant inbred lines. The seed coats and cotyledons were separated and quantitatively analyzed for protein, starch, dietary fiber, polyphenols, flavonoids, vitexin, isovitexin, and antioxidant activities using standard chemical assays and HPLC, followed by statistical analysis and principal component analysis. The cotyledons contained more protein (26.97–28.34%) and starch (50.40–56.25%), whereas the seed coat contained more dietary fiber (74.17–79.93 g/100 g) and bioactive compounds. Polyphenolic compounds were significantly higher in the seed coat than in the cotyledons (p < 0.05) and were positively correlated with seed coat darkness, indicating that the black mung bean had higher bioactive functions. This study provides evidence for mung bean variety improvement and functional food development.

1. Introduction

Mung bean (Vigna radiata) is an annual herbaceous legume belonging to the genus Vigna (Fabaceae) and has been cultivated in China for over 2000 years [1]. Their wide adaptability, stress tolerance, and nitrogen-fixing capacity make mung beans an important crop for rotation, green manure, animal feed, disaster relief, and sustainable agricultural systems [2]. The mung bean is rich in protein, vitamins, minerals, and bioactive compounds, including flavonoids and polyphenols [3]. Due to its high nutritional, medicinal, and processing value, the mung bean is widely utilized in the production of sprouts, beverages, and confectionery products. According to traditional Chinese medicine, the mung bean possesses multiple therapeutic effects, including clearing heat, relieving summer heat, promoting diuresis, reducing swelling, moistening the throat, quenching thirst, improving eyesight, and lowering blood pressure [4]. Modern nutritional and pharmacological studies have also demonstrated the hypoglycemic [5], anti-inflammatory [6], antihypertensive, and antioxidant properties [7] of mung beans. Flavonoids, particularly vitexin and isovitexin, have been identified as the primary antioxidant constituents of mung beans [8].

Mung bean cotyledons contain a significantly higher protein content (19.78%) than the seed coat (10.33%) and are abundant in starch and other carbohydrates. Conversely, the seed coat has a higher dietary fiber (65.85%) than the cotyledons (10.75%) and is enriched with polyphenols, flavonoids, and anthocyanins [9]. However, during the processing of mung bean sprouts, cakes, and vermicelli, the seed coat is typically discarded or used as low-value feed because the seed coat has a negative effect on taste and texture, wasting valuable nutrients [10].

Mung bean has different seed coat colors, and there are different dietary preferences among groups of people [11]. However, few studies have examined differences in nutritional and functional components among mung bean varieties with different seed coat colors. In this study, we investigated the differences in protein, starch, flavonoid, and polyphenol components and DPPH and ABTS⁺ radical scavenging capacities in the seed coat and cotyledons among 12 mung bean lines with different seed coat colors. The differences in nutritional factors among mung bean varieties with different seed coat colors were determined to provide useful information for the efficient use of this crop.

2. Materials and Methods

2.1. Materials

Twelve mung bean lines from six recombinant inbred lines were used, and one wild accession (P340225025) was used as a control (

Table 1. Source and Characteristics of Mung Bean Lines.

Trial Material	Original Combination	Seed Color	Hundred-Seed Weight (g)
B1	Zhonglv28 × 21XJ76	Black	6.7
G1	Zhonglv28 × 21XJ76	Green	6.9
B2	Zhonglv26 × Jilv 9	Black	6.5
G2	Zhonglv26 × Jilv 9	Green	6.6
B3	Zhonglv28 × Zhonglv13	Black	7.4
G3	Zhonglv28 × Zhonglv13	Green	7.1
Y1	Zhonglv26 × Huanglvdou	Yellow	5.4
G4	Zhonglv26 × Huanglvdou	Green	6.6
Y2	Zhonglv27 × Huanglvdou	Yellow	6.1
G5	Zhonglv27 × Huanglvdou	Green	5.2
Y3	Zhonglv28 × Huanglvdou	Yellow	5.6
G6	Zhonglv28 × Huanglvdou	Green	7.2
CK	P340225025	Brown	2.3

). All accessions were planted in summer 2024 in Shunyi District, China (Latitude: 116.32° E, longitude: 39.96° N), and harvested in early autumn 2024.

2.2. Sample Preparation

Each mung bean sample (30 g) was weighed and soaked in distilled water at 25 °C for 24 h; then the seed coats and cotyledons were separated manually. The separated materials were dried at 36 °C for 12 h in a forced-air oven (DHG-9055A, Yiheng Technical Co., Shanghai, China), ground into a powder using a high-speed grinder (FW-100, Tianjin Taisite Instrument Co., Tianjin, China), sieved through a 60-mesh stainless-steel sieve (250 μm), and stored in airtight containers at −20 °C.

For each seed coat or cotyledon powder, 0.2 g was accurately weighed and put into a 50 mL centrifuge tube containing 70% (v/v) HPLC-grade aqueous methanol (8 mL, Sigma-Aldrich, St. Louis, MO, USA). The samples were vigorously mixed for 30 s and ultrasonically extracted (250 W, 40 kHz, 25 °C) for 15 min using an ultrasonic cleaner (KQ3200, Kunshan Ultrasonic Instrument Co., Kunshan, China). The samples were centrifuged at 10,000× g for 10 min at 4 °C. The supernatant was filtered through a 0.22 μm nylon membrane filter (Merck Millipore, Burlington, MA, USA). The filtrate, was stored in amber glass vials at −80 °C.

2.3. Crude Protein Content

Crude protein content was assessed according to the previous method [12]. The sample (0.5 g) was digested with 98% H₂SO₄ (10 mL) and K₂SO₄/CuSO₄ catalyst (6.0 g, 10:1 w/w) at 420 °C for 90 min using a digestion system (KDY-9830, Hanon Instruments, Jinan, China). The digested product was distilled with 40% (w/v) aqueous NaOH into 4% (w/v) boric acid (Merck, Darmstadt, Germany) containing a mixed indicator (methyl red/bromocresol green, 1:5), followed by titration with standardized 0.1 M HCl (daily calibration).

2.4. Total Starch Content

Total starch content was determined using the second acid hydrolysis method in the national standard (GB 5009.9-2016) [13]. The sample (2.0 g) was moistened, mixed with HCl, and heated in a boiling water bath for 2 h. After cooling, the sample was neutralized with NaOH, diluted, and filtered. An aliquot of the filtrate was re-hydrolyzed with HCl for 15 min, then cooled, neutralized, and diluted. TS content was determined by anthrone colorimetry (absorbance measured at 620 nm after boiling water bath treatment) using the following glucose standard curve with a conversion factor of 0.9.

$$\mathrm{Starch} \mathrm{content}\left(\%\right)=\left[{\displaystyle \frac{\left(V_S-V_b\right)\times C\times 162.14\times 0.9}{m\times 1000}}\right]\times 100$$

(1)

Here, V_S (mL) is the HCl volume in the sample, V_b (mL) is the HCl volume in the blank, C (mol/L) is the HCl concentration, m (g) is the sample mass, 162.14 is the molar mass of anhydrous glucose (g/mol), 0.9 is the glucose-to-starch conversion factor, and the division by 1000 converts milliliters to liters.

2.5. Total Dietary Fiber Content

Total dietary fiber content was determined using a modified enzymatic–gravimetric method with LC detection based on Phillips et al. [14]. Briefly, a sample (1.0 g) was defatted with petroleum ether, dried at 105 °C for 2 h and then sequentially enzymatically hydrolyzed with heat-stable α-amylase, protease, and amyloglucosidase. Following filtration, washing, and drying, the residue was weighed. Protein (Kjeldahl method) and ash (via incineration at 550 °C for 5 h) contents in the residue were subtracted for correction.

2.6. Total Phenolic Content and Total Flavonoid Content

The samples were extracted with the same procedure as in Section 2.2. The sample (0.2 g) was extracted with 80% methanol (10 mL, Sigma-Aldrich) ultrasonically (400 W, 30 °C) for 30 min, followed by centrifugation at 4000× g for 10 min (Hettich Universal 320R, Andreas Hettich, Tuttlingen, Germany), and the supernatant was collected and filtered. The filtrate was used for determining total phenolic content (TP) and antioxidant activity.

TP was determined by a modified version of the Folin–Ciocalteu method [15]. The sample (1 mL) was added to Folin–Ciocalteu reagent (0.25 mL) at 10% concentration under light-proof conditions for 3 min and then 7% (w/v) Na₂CO₃ (0.75 mL) was added. After 1 h in the dark, the absorbance was recorded at 760 nm using a UV-VIS analyzer (Shimadzu, Kyoto, Japan). The TP measurements are expressed as gallic acid equivalents (GAE) based on a calibration curve (GAE/g).

Total flavonoid content (TF) was quantified by using a method from Shraim et al. [16]. First, 5% (w/v) NaNO₂ (0.75 mL) was added to the sample (1 mL), and the mixture was kept under light-proof conditions for 5 min. Next, 7% (w/v) AlCl₃ (0.75 mL) was added, and the mixture was stirred for an additional 5 min in the dark. Subsequently, 1 mol/L NaOH (5 mL) was added, and the mixture was kept under dark conditions for 5 min. Absorbance readings were taken at 510 nm using a UV-VIS spectrophotometer (Shimadzu). The results are expressed as rutin equivalents, calculated from a standard rutin curve (mg CAE/g).

2.7. Determination of Vitexin and Isovitexin Content

Vitexin and isovitexin were analyzed by HPLC [17]. A sample (1.0 g) was ultrasonically extracted (300 W, 35 °C) with 80% methanol (20 mL) for 40 min, centrifuged (6000× g, 15 min), and the supernatant filtered through a 0.22 μm membrane. HPLC analysis (1260, Agilent Technologies, Santa Clara, CA, USA) used a Symmetry C18 column, Waters Corporation, Milford, MA, USA (4.6 mm × 150 mm, 5 μm) with a mobile phase of methanol:0.3% phosphoric acid (35:65, v/v) at 1 mL/min. Detection was at 260 nm, column temperature was 25 °C, and injection volume was 10 μL. Quantification was via external standards. Vitexin and isovitexin contents were calculated as

$$\mathrm{Content} \left(\mu \mathrm{g}/\mathrm{g}\right)={\displaystyle \frac{Y\times V\times D}{W_S}}$$

(2)

where Y is the concentration from the calibration curve, V_b (mL) is the extraction volume, D is the dilution factor, and W_S (g) is the sample mass. All analyses were performed in triplicate.

2.8. Determination of DPPH Scavenging Activity

The ability to scavenge DPPH radicals was assessed by using a modification of the method of Silva et al. [18]. The sample (0.5 mL) was mixed with 0.1 mM DPPH-methanol solution (0.95 mL) and allowed to react under light-resistant conditions for 30 min. The reduction in absorbance was measured at 517 nm using a UV-VIS spectrophotometer (Shimadzu). Based on the following equation, the percentage of DPPH radical inhibition was calculated to quantify the antioxidant activity.

$$\mathrm{DPPH} \mathrm{activity} \left(\%\right)={\displaystyle \frac{{\mathrm{A}}_{\mathrm{ctrl},\mathrm{DPPH}}-{ \mathrm{A}}_{\mathrm{smpl},\mathrm{DPPH}}}{{\mathrm{A}}_{\mathrm{ctrl},\mathrm{DPPH}}}}\times 100$$

(3)

Here, A_ctrl,DPPH is the absorbance of the DPPH solution with distilled water, while A_smpl,DPPH is the absorbance of the DPPH solution with the mung bean sample.

2.9. Determination of ABTS Radical Scavenging Activity

The ability to scavenge ABTS radicals was assessed using a modification of the method of Guedes et al. [7]. The sample (0.5 mL) was mixed with ABTS working solution (0.95 mL) and incubated at 25 °C for 6 min. The reduction in absorbance was measured at 734 nm using a UV-VIS spectrophotometer (Shimadzu). Based on the following equation, the percentage of ABTS radical inhibition was calculated to quantify the antioxidant activity.

$$\mathrm{ABTS} \mathrm{activity} \left(\%\right)={\displaystyle \frac{A_{\mathrm{ctrl},\mathrm{ABTS}}-A_{\mathrm{smpl},\mathrm{ABTS}}}{A_{\mathrm{ctrl},\mathrm{ABTS}}}}\times 100$$

(4)

Here, A_ctrl,ABTS is the absorbance of the ABTS solution with distilled water, while A_smpl,ABTS is the absorbance of the ABTS solution with the mung bean sample.

2.10. Total Antioxidant Capacity

Total antioxidant capacity (was assessed by the ferric reducing antioxidant power (FRAP) assay with modifications to the method described by Zhu et al. [19]. The sample (0.5 g) was ultrasonically extracted (300 W, 30 °C) with 80% methanol (10 mL, Sigma-Aldrich) for 25 min, centrifuged (5000× g, 10 min), and the supernatant (10 mL) was mixed with FRAP working solution (3.0 mL, preheated to 37 °C). After incubation at 37 °C for 10 min, absorbance was measured at 593 nm (UV-2600, Shimadzu).

2.11. Statistical Analysis

All experiments were performed in triplicate, with results expressed as mean ± standard deviation. One-way ANOVA with Tukey’s honestly significant difference post hoc test (p < 0.05) and Pearson correlation analysis were conducted using SPSS 27.0 (IBM, Armonk, NY, USA) and R software version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Principal component analysis (PCA) was performed via the prcomp function in R, with visualization using R packages (ggplot2 and factoextra). Data visualization (bar charts and heatmaps) was conducted using GraphPad Prism 10.1.2 (GraphPad Software, Boston, MA, USA) and ggplot2 in R.

3. Results

3.1. Nutritional Composition in Mung Bean Lines with Different Seed Coat Colors

A significant difference in the distribution of primary nutrients was observed between the seed coat and cotyledons. Crude protein (CP) content and total starch (TS) were mainly accumulated in the cotyledons. The average CP in the cotyledons (26.97–28.34%) was significantly higher (p < 0.05) than in the seed coat (6.20–8.47%) (Figure 1A). Similarly, the TS in the cotyledons (50.40–56.25%) was significantly higher than that in the seed coat (21.82–23.52%) (p < 0.05; Figure 1B). Conversely, total dietary fiber (TDF) was mainly concentrated in the seed coat (74.17–79.93 g/100 g), and was almost 9-fold higher than that in the cotyledons (7.57–8.85 g/100 g) (p < 0.05; Figure 1C).

Seed coat color had a statistically significant (p < 0.05) effect on nutritional composition. The black seed coat had the highest TS (55.03%) and TDF (8.85 g/100 g) contents in the cotyledon, whereas the wild-type possessed the highest CP content (28.34%). The highest TDF (79.44 g/100 g) and TS (23.52%) were also detected in the black seed coat, but with the lowest CP (6.71%). No significant difference in CP was found between the yellow and green seed coat (p > 0.05).

3.2. Bioactive Compounds in Mung Bean Lines with Different Seed Coat Colors

The distribution of bioactive compounds, which are key secondary metabolites, differed substantially between the seed coat and cotyledons. The total phenolic (TP) in the seed coat (4.77–6.40 mg GAE/g) was significantly higher (p < 0.05) than that in the cotyledons (0.55–1.56 mg GAE/g) (Figure 2A). A parallel trend was observed for total flavonoid (TF), with the seed coat (2.29–4.50 mg CAE/g) having substantially higher TF than the cotyledons (0.43–0.69 mg CAE/g) (p < 0.05; Figure 2B). The major flavone glycosides, vitexin and isovitexin, were almost exclusively localized in the seed coat, with maximum concentrations of 5.39 and 9.67 mg/g, respectively. These values were much higher than the minimal levels detected in the cotyledons (both 0.29 mg/g) (Figure 2C,D).

The lines with black seed coat consistently exhibited the highest TP (6.19 mg GAE/g), TF (4.03 mg CAE/g), vitexin content (5.18 mg/g), and isovitexin content (9.76 mg/g) in the seed coat. Pairwise comparisons within recombinant inbred lines showed that the seed coat always contained much higher levels of bioactive compounds than the corresponding cotyledons. For instance, in line G1.1, the isovitexin content in the seed coat (8.85 mg/g) was nearly 50-fold greater than that in the cotyledons (0.18 mg/g).

3.3. Antioxidant Capacity in Mung Bean Varieties with Different Seed Coat Colors

A strong positive correlation was observed between seed coat color depth (quantified as 100 − L*) and all antioxidant assays (p < 0.01; Figure 3A). Linear regression confirmed that color depth could explain 82–88% of the variability in antioxidant activity (R² = 0.85 for DPPH, R² = 0.88 for ABTS, R² = 0.82 for TA).

Consistent with this correlation, the seed coat exhibited significantly higher (p < 0.01) antioxidant capacity than the cotyledons across all assays (Figure 3B). Furthermore, a clear color-dependent hierarchy of antioxidant capacity was established of black > green > yellow > wild-type. Black seed coat varieties, which had the greatest seed coat color depth (63.85), also had the most potent seed coat antioxidant activities, significantly exceeding (p < 0.05) those of all other varieties. The DPPH scavenging rate (91.67%) was 8.4% higher than that of the yellow varieties (84.55%).

3.4. Correlations Between Bioactive Compounds and Antioxidant Capacity

Heatmap analysis showed that bioactive compounds (TP, TF, vitexin, and isovitexin) exhibited highly significant positive correlations (p < 0.01; r = 0.82–0.94) with all antioxidant capacity measures (Figure 4A). In contrast, macronutrients (CP, TS, and TDF) showed no significant correlations (p > 0.05) with either bioactive compounds or antioxidant capacity.

4. Discussion

4.1. Compositional Differences Between Mung Bean Seed Coat and Cotyledons

There were significant differences in components between the cotyledons and seed coat. This differential distribution may be related to the molecular mechanisms by which gibberellin and abscisic acid regulate seed embryo germination and development. For instance, Kozaki & Aoyanagi [20] reported that gibberellin promotes the decomposition of storage compounds (e.g., starch in cotyledons) to supply energy for embryo growth, whereas abscisic acid maintains seed dormancy by inhibiting nutrient mobilization. These findings are consistent with our observation that the cotyledons were enriched in storage compounds and the seed coats in defensive bioactive compounds.

In this study, the cotyledons were the main storage site for proteins and starches. The protein content was 3–4 times higher than that of the seed coat, and the starch content was 2–2.5 times higher than that of the seed coat. In contrast, the seed coat was enriched in TDF and bioactive compounds, which were significantly higher than in the cotyledons. Additionally, vitexin and isovitexin are almost exclusively present in the seed coat, reaching maximum contents of 5.39 and 9.67 mg/g, respectively. These findings are consistent with the results of Deng & Wang [9] regarding the nutritional components in different mung bean tissues. Furthermore, the present study supports the regulatory role of seed coat color in this distribution pattern and also explains why discarding the seed coat during processing leads to the loss of functional components, such as polyphenols and flavonoids [10].

4.2. Seed Coat Color Regulates Bioactive Components and Antioxidant Capacity

Seed coat color is closely associated with the accumulation of secondary metabolites [21]. The present study demonstrated that the depth of seed coat color showed a significant positive correlation with the bioactive component content (p < 0.01) in the order black > green > yellow > wild type. The black seed coat samples exhibited the highest TP (6.19 mg GAE/g), TF (4.03 mg CAE/g), vitexin content (5.18 mg/g), and isovitexin content (9.39 mg/g). These findings are consistent with previous studies on functional components in black bean [22] and mixed beans [23].

Among the secondary metabolites linked to seed coat color, anthocyanins are a key regulatory factor. Pal et al. [24] demonstrated that higher expression of anthocyanin biosynthesis genes in seed coats correlates with darker coloration. In the present work, the black seed coat varieties had the highest TP and TF in the seed coat. These compounds are synergistically synthesized with anthocyanins [25], suggesting that active anthocyanin metabolism may drive the accumulation of other bioactive components in dark seed coats. The antioxidant activity of seed coats also varied with color. The DPPH scavenging rate of the black seed coats reached 91.67%, which was 8.4% higher than that of the yellow seed coats. The black seed coats also showed significantly higher ABTS scavenging rates and TAC (p < 0.05). The high antioxidant capacity related to color was not limited to the seed coat itself; even the cotyledons of mung bean varieties with dark seed coats exhibited higher antioxidant activity. Pankaj et al. [26] proposed that there may be polyphenol precursor transport from the seed coat to the cotyledons; however, the details of this transport mechanism, such as the key transporters involved and the regulatory signals, require further experimental verification.

4.3. Nutritional Components Show No Correlation with Seed Coat Color

The nutritional components in mung bean are mainly determined by the variety’s physiological characteristics and growth environment [27]. Proteins and starch, which are the main storage compounds, are synthesized and accumulated mostly in the cotyledons. These components are regulated by specific gene networks, which are largely independent of pigment synthesis pathways [28]. By contrast, dietary fiber is enriched in seed coats, and its content is related to seed coat structure and weakly correlated with pigments [29]. Consistent with these prior findings, the results of this study showed no significant correlation between nutritional components and seed coat color (p > 0.05). The protein content in cotyledons varied minimally across different-color varieties (27.38–28.34%), which is consistent with the assumption that protein synthesis is independent of pigment pathways. Although the black seed coats contained the highest dietary fiber (79.44 g/100 g), the seed coats of the yellow, green, and wild-type varieties showed no statistically significant differences in this component, which agrees with the findings of Yan et al. [30] that dietary fiber is weakly correlated with pigments. However, the specific mechanism by which seed coat color might affect these nutrients remains unclear and requires further research.

5. Conclusions

This study systematically investigated the distribution of nutritional components, bioactive substances, and antioxidant capacity in the seed coats and cotyledons of mung bean varieties with different seed coat colors and elucidated the regulatory effect of seed coat color on these traits. The results revealed a significant tissue-specific distribution pattern in mung bean seeds: crude protein and total starch, as the main storage nutrients, were concentrated in the cotyledons, whereas total dietary fiber, total phenolics, total flavonoids, and characteristic flavone glycosides were markedly enriched in the seed coats. The seed coats also exhibited significantly higher antioxidant capacity. Seed coat color exerted a notable regulatory influence on bioactive substances and antioxidant capacity in a color-dependent manner, following a general hierarchy of black > green > yellow > wild-type, with black-seeded varieties showing the highest values across these indicators. Notably, primary nutritional components showed no significant correlation with seed coat color, as their synthesis is governed by a storage metabolism gene network independent of the pigment synthesis pathway. Dietary fiber, in contrast, displayed only a weak association with seed coat pigmentation. Overall, this study clarifies the distribution patterns of nutritional and functional components in mung beans and their relationship with seed coat color, providing a theoretical basis for developing mung bean functional foods and breeding high-quality varieties.