Selective green leafy vegetables and their synergistic combination approach as natural anti-diabetic agents: therapeutic potential

doi:10.21203/rs.3.rs-3015714/v1

This study examined the antidiabetic potential and antioxidant properties of ten green leafy vegetables (GLVs) using in-vitro tests for α-glucosidase, α-amylase, and lipase inhibition (AGI, AAI, and LPI, respectively). To extract metabolites, 80% ethanol was used, and the resulting crude extract was examined for total phenolic content (TPC) and bioactivities. Of all the samples tested, A. cepa had the highest AGI activity at 595.28 mg ACE/g extract, 25 times greater than the sample with the lowest activity. During the AAI activity, A. fistulosum exhibited the highest inhibition of 36.73 mg ACE/g extract, surpassing all other samples. Meanwhile, P. crispumdemonstrated the highest LPI with an inhibition of 31.07%. Among antioxidant activities, DPPH activity was substantial, while metal chelating and FRAP activities were moderate. The results of studies combining A. cepa, A. fistulosum, and A. graveolens L. in equal proportions revealed the most effective combination for inhibiting all enzymes, even though the TPC remained uniform in all combinations. The mechanism of inhibition observed with A. cepa was non-competitive, whereas the combination of A. cepa, A. fistulosum, and A. graveolens L. (combination-1)displayed competitive inhibition, similar to that of acarbose. FTIR was used to identify the functional groups from all the samples. GC-MS results indicated that mannitol, myo-inositol, succinic acid, and propanoic acid were likely responsible for the antidiabetic activity. This research about the potential of GLVs as oral agents for treating T2DM could be critical in managing diabetes and developing functional food to prevent T2DM.

Biological sciences/Biochemistry/Enzymes

Biological sciences/Biochemistry/Kinases

Biological sciences/Biochemistry/Metabolomics

Health sciences/Molecular medicine

Globally, non-communicable diseases (NCDs), such as diabetes mellitus, hypertension, and heart disease, have increased susceptibility to COVID-19 disease¹. Much evidence has linked COVID-19 disease with NCDs as a comorbidity that can potentially increase mortality². Type 2 diabetes mellitus (T2DM) is one of the most prevalent metabolic diseases caused due to two main factors: impaired insulin secretion and insulin resistance³. T2DM is known to be managed by an appropriate diet intake, moderate exercise, or by using oral agents or drugs relating to hypoglycemia, anti-obesity, antihypertensive, lipid-lowering agents, and antioxidants as treatment options⁴.

Synthetic antidiabetic drugs to treat T2DM have been known to show promising results but exhibit several disadvantages, such as abnormal bacterial fermentation of carbohydrates in the large intestine, diarrhea, distention, flatulence, meteorism, and other³. To overcome these drawbacks, oral treatment using natural antidiabetic agents has gained colossal significance, owing to their advantages, such as efficiency, negligible side effects, low toxicity, and most minor interferences with other treatments or therapies, compared to synthetic drugs^4,5. Moreover, nutritional intervention is essential to manage this risk, and the most effective prevention strategy is lifestyle changes related to diet, physical activity, and control of metabolic disorders⁶. Moreover, various NCDs have been associated with low fruit and vegetable intake⁷. A high intake of especially green leafy vegetables is highly recommended to prevent NCDs⁸.

Given their potential, screening plant sources and identifying relevant metabolites is crucial to managing T2DM. Several plant species belonging to medicinal, vegetables, fruits, mushrooms, spices, etc., have been shown to possess strong antidiabetic properties, mainly owing to their unique chemical composition, and are used to treat diabetes⁴. Among all these sources, green leafy vegetables (GLVs) have particular importance, as they are consumed every day, in various combinations, eaten raw or with minimal processing. GLVs are commonly referred to as “protective foods” in the human diet due to their numerous health advantages attributed to their high levels of vitamins, essential fatty acids, minerals, amino acids, and dietary fiber, as well as a variety of critical bioactive compounds⁹. Several literatures report that consuming GLVs in moderate quantities can significantly reduce the risk of T2DM, primarily exhibiting preventive effects^10,11. In addition, the high magnesium content of GLVs is known to have a beneficial impact on glucose homeostasis and lower the risk of T2DM¹⁰. In addition, dietary supplements of a few GLVs have been known to improve the complications associated with T2DM, including advanced stages³.

The antidiabetic property of plants is mainly associated with their metabolites, which are known to inhibit α-glucosidase enzyme activity, notably in patients with non-alcoholic fatty liver disease and steatohepatitis³. A few kinds of literature have also reported that a few α-glucosidase inhibitors from plants have shown additional health-promoting properties such as antioxidant activities¹², anti-obesity¹³, anti-Alzheimer's disease¹⁴, antimicrobial¹⁴, anti-tyrosinase activity³, etc. E.g., metabolites from commercial antidiabetic polyherbal formulations extracts have shown α-glucosidase inhibitory (AGI) activity with α-amylase inhibitory (AAI) and lipase inhibitory (LPI) activities¹³. In this view, combining particular plant species with desirable properties can be advantageous in formulating a novel oral agent as a potential antidiabetic treatment option.

The present study investigated the solvent-based extraction of metabolites from selective commonly used GLVs to characterize and determine the biological activities associated with T2DM. The total phenolic contents (TPC), antioxidant, AAI, AGI, and LPI activities were evaluated. The samples with relatively highest activity were subjected to combination studies to achieve optimal actions. Nevertheless, this is the first study on experimenting with dietary sources in a synergistic combination. FTIR was also characterized to identify functional groups affecting the activities. The inhibition kinetics were conducted to investigate how the extracts inhibit α-glucosidase and gain insights into the inhibition mechanism. Functional groups and compounds involving the activities were identified through characterization using FTIR and GC-MS.

2.1 Moisture content

The present study investigated ten different GLVs. Figure 1a depicts the moisture contents of fresh and dehydrated GLVs. The moisture content of the fresh GLVs ranged from 86.68% to 92.23%, indicating a marginal difference among all the samples. While the moisture content of dehydrated GLVs varied greatly (p < 0.05), the values ranged from 9.84% to 16.97%. The reduction in moisture content was from 74.20% to 79.86%. The variation in moisture content of dehydrated GLVs could be due to the drying process effects and the plant sample’s physical properties, such as the thickness of leaves, porous structures, etc.¹⁵. A dehydration dryer is known to minimize the chance of losing functional properties, such as bioactivity, antioxidant capacity, and retention of metabolites and energy consumption, due to the milder processing conditions¹⁵.

The high reduction in moisture content of dehydrated I. aquatica Forsk and S. oleracea L. could be due to the highly porous nature of stems with more extensive vascular tissue as well as the samples having thinner leaves have shown high drying efficiency^16,17, while the samples such as B. perviridis and A. fistulosum baring observed to have dense stems and thicker leaves showed low drying efficiency¹⁵.

2.2 Extraction yield

The sample A. fistulosum had the highest extraction yield, up to 23.86% (dry wt. basis), while the yield from O. basilicum was the lowest (5.74%, dry wt. basis) (Fig. 1b). Other plant samples showed a moderate extraction yield. An efficient solvent system is primarily required to extract different metabolites with varying chemical properties and polarities¹⁸. Several reports have studied the effect of different solvents on the extraction yield of plant metabolites and have reported that 80% ethanol renders the highest yield. E.g., Khatib, et al.¹⁹ studied the fruit of M. charantia and said that 80% ethanol solvent had the highest yield. Likely, the present study used 80% ethanol as the extraction medium. In addition to the solvent system, the present study found that the moisture content positively correlated with the extraction yield (r = 0.611, p < 0.05). Moreover, it was also observed that the sample A. fistulosum with the highest yield had high bioactivities, including α-amylase and α-glucosidase inhibitory activities. Likely, Eruygur, et al.¹⁴ reported that the metabolite of A. cucullata resulting from 80% ethanol extraction had shown more excellent α-glucosidase inhibitory activity than acarbose. Compared with other studies, the present work has also found similarities in extraction yield and bioactivity when using 80% ethanol as a solvent.

2.3 TPC

The TPC of GLVs crude extracts is presented in Table 2. The highest TPC was recorded for A. graveolens L. crude extract, 23.78 mg GAE/g extract (p < 0.05), and three times higher than A. fistulosum. The results differ slightly from previously reported in A. graveolens L. 18-49 mg GAE/g extract²⁰ and A. fistulosum 5.91 mg GAE/g extract²¹. This could be due to the differences in genetic (genotype), environmental (weather, soil type, geographical location), and processing conditions factors¹⁹. Moreover, several literatures have reported that 80% ethanol can extract the maximum or highest TPC from various plant samples, E.g., oyster, garlic, onion, ginger, aloe vera, thyme, and oak^22,23.

2.4 Antioxidant activity

2.4.1 DPPH

The DPPH activity of GLVs crude extracts is presented in Table 2. The sample A. graveolens L. exhibited the highest DPPH activity of 125.57 mg AEAC/g extract (p <0.05), 13 folds higher than A. fistulosum, which showed the lowest activity of 8.99 mg AEAC/g extract. Similarly, Chandrashekharaiah²⁰ reported that leaves and bark of A. graveolens L. had a higher DPPH inhibition ratio, i.e., 85-95%. The other samples, C. sativum, I. aquatica Forsk, and B. perviridis, showed relatively high DPPH activity, i.e., 119.42, 93.35, and 46.77 mg AEAC/g extract, respectively. Our study noted that the DPPH activity was in alignment with TPC. On the other hand, a negligible correlation of DPPH activity was noticed with AAI, AGI, and LPI activities (r = -0.394, -0.408, and 0.277, respectively). Tunnisa, et al.²⁴ reported that samples with high antioxidant activity only sometimes had high AGI activity, as the compounds responsible for the activity can differ.

2.4.2 Metal chelating

Table 2 presents the metal chelating activity; among all the samples, C. sativum crude extract had the highest, and the B. chinensis L. crude extract was the lowest activity, i.e., 92.85 and 2.63 mg EECC/g extract (p < 0.05). Similarly, Wong and Kitts²⁵ documented that C. sativum leaf extract had higher chelating activity than P. crispum leaf extract. This could be because the polyphenols can act as metal ion chelators and interfere with oxidation and peroxidation reactions by donating protons²⁵. On the other hand, B. chinensis L. has been known to have low TPC (Table 2), which matches its low chelating ability (Table 2). The metal chelating activity was known to have a moderate correlation with TPC, DPPH, FRAP, and LPI activities (r = 0.595, 0.574, 0.436, and 0.564, respectively, respectively) (p < 0.05).

2.4.3 FRAP

In the present study, the highest FRAP activity was recorded for C. sativum crude extracts (Table 2), which was six times higher than A. fistulosum crude extracts (p < 0.05). Based on TPC, it could be suggested that polyphenols could scavenge radicals due to their potential to reduce transition metals such as iron (Fe)²⁶. In addition, FRAP activity was correlated with other antioxidant activities but not with AAI, AGI, and LPI activities (r = -0.220, -0.264, and 0.151, respectively) (p ≥ 0.05). Similarly, Paul and Majumdar¹³ reported a negative correlation between FRAP with enzyme inhibitory activities.

Overall antioxidant activities and correlation results suggested that the sample having antioxidant activities did not directly correlate with enzyme inhibitory activities. This indicates that compounds with antioxidant potential did not contribute to enzyme inhibitory activities (Table 2). At the same time, there was a strong correlation between antioxidant activities and TPC.

2.5 α-Amylase inhibitory (AAI) activity

AAI activity of GLVs crude extracts is presented in Table 2. The results indicated that most plant samples exhibited high to moderate activity, including A. fistulosum, O basilicum, P. Crispum, B. chinensis L., C. sativum, and I. aquatica Forsk. Based on previously published reports, A. fistulosum leaf extracts had AAI activity of about 75% at a concentration of 25 µg/mL²⁷. A. graveolens L. was observed to have the lowest AAI activity and high TPC, and A. fistulosum crude extract had a high AAI and low TPC. The results suggested that AAI activity did not agree with TPC, and the compounds with AAI activity may not be from phenolic compounds. Vinodhini, et al.²⁷ reported that allicin (essential oil) is an important compound in A. fistulosum. This compound possesses excellent antidiabetic properties²⁸.

2.6 α-Glucosidase inhibitory (AGI) activity

The AGI activity of GLVs crude extracts is shown in Table 2. The highest AGI activity was in A. cepa crude extracts with 595.28 mmol ACE/g of extract (p < 0.05). Similarly, Masood, et al.²⁸, reported that A. cepa bulb ethanolic extracts showed approximately 80% of AGI activity. A. fistulosum and B. perviridis exhibited moderate activity (282.97 and 210.42 mmol ACE/g of extract, respectively). The rest of the samples showed AGI activity at lower levels. Like AAI and AGI activity did not correlate with the TPC (r = -0.139) (p ≥ 0.05), this was similar to the findings of Paul and Majumdar¹³, who observed that the TPC of commercial antidiabetic polyherbal formulation extracts did not correlate with the AGI activity.

Overall, the AAI and AGI activity of GLVs did not correlate with TPC. This could be due to metabolites other than phenolic compounds contributing to the AGI and AAI activity, also known to bring several other beneficial biological effects²⁸. The metabolites are known to act as AAI and AGI inhibitors via non/mixed-competitive mechanisms, which are mainly driven by non-covalent interactions (i.e., hydrogen bonding, van der Waals forces, π-effects, etc.)⁴. Based on present GC-MS results and also the published literatures, the high antidiabetic activity of A. cepa extracts correlated with the high abundance of several compounds that have been reported to have an antidiabetic activity, such as propanoic acid²⁹, succinic acid³⁰, and myo-inositol³¹. Besides, hexadecanoic acid with low abundance could be correlated with antidiabetic activity. This compound has been reported to possess an antidiabetic activity³².

2.7 Lipase inhibitory (LPI) activity

A total of 8 GLVs have shown high to moderate LPI activity (Table 2). The most increased activity was from P. crispum, which was six folds higher than A. fistulosum (p < 0.05). There are no reports on LPI for P. crispum, but the A. fistulosum leaves reported low LPI activity of 33.76%²¹. LPI activity significantly correlated with TPC (r = 0. 774, p < 0.01). Likely, the LPI activity of a commercial antidiabetic polyherbal formulation extract has been reported to have a direct correlation with TPC with r = 0.985¹³. Due to the vast number of molecular structures of phenolic compounds, the precise mechanism of action on inhibition of lipase is known to vary significantly. The mode of action could be competitive, uncompetitive, or mixed inhibition mechanism³³. In addition, LPI activity did not correlate with DPPH and FRAP activities (r = 0.277 and 0.151, respectively) (p ≥ 0.05) but was only associated with metal chelating activity (r = 0.564, p < 0.05). Based on the correlation between LPI, TPC, and antioxidants activities, it can be conferred that the polyphenols other than showing antioxidant activity are exhibiting LPI activity^13,27. Additionally, LPI activity was not correlated with AAI and AGI activities (r = -0.543 and -0.095, respectively, p ≥ 0.05), similar to as reported by Paul and Majumdar¹³.

2.8 Synergistic combination effect

The combination studies were done by mixing three different GLVs, mainly exhibiting high enzyme inhibitory and antioxidant activities, resulting in seven possible combinations, as presented in Table 1.

TPC of GLVs combination crude extracts is presented in Table 3. The results demonstrated that all seven combinations had higher TPC content, almost comparable to the highest TPC of an individual sample, i.e., A. graveolens L. In addition, no significant difference was observed in the TPC of all seven combinations (p ≥ 0.05). Similarly, Khongrum, et al.³⁴ reported that combining celery and Chinese kale synergized the TPC. However, there was a significant difference in other activities.

Table 3 presents the antioxidant potential. The synergistic combination of GLVs had shown high antioxidant potential, where the highest DPPH activity was observed to be five folds higher (for combination-4) than individual crude extracts (A. graveolens L.) (p < 0.05). Similarly, Jiang, et al.³⁵ reported that the synergistic effect of the eggplant and carrot combination increased DPPH activity approximately seven times. On the other hand, the DPPH activity did not correlate with the TPC (r = 0.057, (p ≥ 0.05). Zhang, et al.³⁶ reported that terpenes, betalains, organosulfides, indoles/glucosinolates/sulfur compounds, protein inhibitors, and other organic acids compounds might contribute to antioxidant activity. The increased antioxidant activity of combination GLVs could be possible due to the: a) synergism of phytochemicals can act differently primarily by protecting phytochemicals from oxidants, b) formation of potent antioxidants with high radical quenching ability, and c) one antioxidant gets oxidized to protect the second antioxidant³⁶.

The results for AAI, AGI, and LPI activities are presented in Table 3. The results for all three-enzyme inhibition were in a different pattern. The increase in AGI activity was the highest when compared to the samples without a combination (p < 0.05). The highest activity was from combination-1 (947.46 mmol ACE/g of extract). Moreover, all seven combinations had considerable AGI activity. The synergistic effect in AGI activity matched the results from Vinholes and Vizzotto¹², who documented that the combined ethanolic extract of C. sinensis and E. uniflora showed a synergistic effect in AGI activity and lipid peroxidation. The subsequent increase in activity was seen for LPI, with a moderate increase in activity (p < 0.05); the highest activity was from combination-7 (43.06%). Similarly, Khongrum, et al.³⁴ reported that the combination of celery and Chinese kale extract showed more significant LPI activity (IC₅₀ 0.048 mg/mL).

Furthermore, the results for AAI activity showed a slight increase in the combinations (p < 0.05), and combination-7 showed the highest activity (39.69 mmol ACE/g of extract). In addition, the combination of chrysanthemum, mulberry, bael, and roselle has been reported to show more significant synergistic inhibition of the α-amylase⁵. Overall, the GLVs combinations demonstrated synergistic activity, in which all the activities increased compared to those from individual samples. Likely, Yang, et al.³⁷ reported that the multi-component of medicinal plants offered great potential for synergistic action on antidiabetic activities. Based on GC-MS data from the combination-1 extract, the high antidiabetic activity was supported by the high abundance of mannitol, myo-inositol, succinic acid, and propanoic acid, which have been reported to have antidiabetic activity^29-31,38. Moreover, compounds with low abundance could be correlated with antidiabetic activity, such as 9,12-octadecadienoic³⁹ and hexadecanoic acid³². Our study noted that in combination-1, mannitol and 9,12-octadecadienoic were not from A. cepa extract but could be from A. graveolens L. based on previous reports, which that compounds were found in A. graveolens L.^40,41. The combined action of these compounds could lead to a synergistic effect that promotes the high antidiabetic activity of this combination. The possible mechanisms for synergistic action could be due to a) various compounds regulating either different or the same target pathways and b) regulating compounds and enzymes that are involved in enhancing the bioavailability³⁷. Thus, consuming GLVs with synergistic interaction can improve postprandial hyperglycemia, significantly preventing and treating diabetes, obesity, and other complications.

2.9 In Inhibition kinetics of α-glucosidase against A. cepa, combination-1, and acarbose

The inhibitory impact at various concentrations of A. cepa, combination-1, and acarbose on α-glucosidase was analyzed by calculating the initial reaction rate (v) through p-NPG consumption for varying times (0–30 min) in the presence of α-glucosidase at various concentrations. The increasing concentrations of A. cepa, combination-1, and acarbose made lowering the v, irrespective of p-NPG concentration (Fig. 2, a1-a3). The inhibitory action was indicated in a concentration-dependent fashion. The Lineweaver-Burk plots of double-reciprocal for determining the inhibition type on α-glucosidase by samples and positive control at various concentrations are shown in Fig. 2, b1-b3. Based on the Michaelis-Menten plot, the K_m and v_max of samples and positive control were computed and shown in Table 4. The linear fitting extension lines crossed at a close position on the x-axis, in which constant K_m values and lower v_max values with increased A. cepa concentration (Fig. 2, b1) were associated with non-competitive inhibition. The mechanism action was non-competitively binding with the neighboring position of the α-glucosidase active site and inducing changes in the conformation of the enzyme⁴². No literatures have reported the inhibition mechanism of A. cepa on α-glucosidase, but previous research regarding thiosulfinate, which is one of the active compounds in A. cepa showed non-competitive inhibition of enzyme activity⁴³.

In contrast to A. cepa, the combination of A. cepa + A. fistulosum + A. graveolens L. (combination-1) was a competitive mechanism with an intersection close y-axis. This inhibition had higher K_m values and constant v_max values attained with increasing concentrations. The interaction between the compounds in the three mixed extracts affected the inhibition mode. The result was similar to Son, et al.⁴⁴, who reported that combining a mixture of glyceollin and luteolin changed the mode of action to competitive inhibition. Competitive inhibitors are the most relevant and generally optimized as drugs⁴⁵. It was noted that a combination of the extracts had synergistic effects on the inhibition action of α-glucosidase. Similarly, the α-glucosidase inhibition type by acarbose was a competitive type, which increased in K_mand constant v_max values with the augmenting concentration of acarbose. The inhibition mode of α-glucosidase matched a report from Son, et al.⁴⁴, who documented that acarbose is a competitive inhibitor. The inhibitor with a competitive type can prevent substrate binding by binding the inhibitor with the free enzyme to produce an enzyme–inhibitor (EI) complex⁴⁴. Generally, to evaluate the inhibition type, Lineweaver-Burk plots have been widely applicated. Overall, acarbose showed the most potent inhibition toward α-glucosidase followed by combination-1 and A. cepa. The result showed that combination-1 had the potential to inhibit the activity of α-glucosidase via a synergistic mode of action, which acted as a natural inhibitor toward α-glucosidase.

2.10 FTIR

The FTIR spectra of GLVs with individual and combinations are shown in Supplementary Fig. S1 and Fig. 3, respectively, and the wavenumbers of identified peaks are shown in Supplementary Tables S1 and S2. Five peaks were identified in the functional group region. The spectra of all the GLVs showed a significant peak at the wavenumber 3100 to 3300 cm⁻¹, corresponding to O─H stretching, indicating the presence of alcohols and phenolic compounds⁴⁶. However, a slight difference was seen in wavenumber, but a significant difference was noticed in the amplitude of all the samples, suggesting the difference in their chemical compositions. Wavenumber at 2921-2924 cm⁻¹attributed for C─H stretching was found in all the samples, indicating the carbonyl group was present⁴⁷. The samples O. basilicum and P. crispum showed the highest amplitude. Further, only the sample O. basilicum had C═O stretching observed at 1728 cm^{−1 46}. The N─H bond corresponding to primary amides or proteins arising from carbonyl stretch was situated at 1579-1614 cm⁻¹, observed in all the samples⁴⁶. The C─N stretching corresponding to amide bond II was seen in two samples, i.e., B. chinensis L. and P. crispum, at wavelengths 1503 and 1500 cm⁻¹. It has been previously reported in B. chinensis L.⁴⁸ and P. crispum⁴⁷.

Five peaks were detected in the figure print region (Supplementary Table S1 and Fig. S1). The CH₃ bending was observed at wavenumber 1348-1397 cm⁻¹ in all the samples⁴⁸. The wavenumber at 1032-1044 cm⁻¹ indicated the presence of C─O stretching corresponding to anhydrides, ethers, carboxylic, and saponins groups was observed in all the samples. The functional groups C═C and COO^-situated at 791-871 cm⁻¹ indicated that they are associated with the vibration of the aromatic ring or ethanol⁴⁷. It was noted that wavenumber at 517-772 cm⁻¹ was detected for alkyl-halogen or alkyl-sulfur bonds or produced by deformation of the ─COO─ group in acetate esters^47,49.

The IR spectra of the highest AGI and AAI activities samples (A. cepa and A. fistulosum) showed the major characteristic peak wavenumber 3264-3273 cm⁻¹ (C─H), 2923-2924 cm⁻¹ (C═CH), 1032-1033 cm⁻¹ (S─O stretch), and 517-515 cm⁻¹ (S─S), which could be of compound allicin⁴⁹. The highest LPI activity in P. crispum was also supported by the peak at 1590 cm⁻¹ (C═O) stretching bond of the ester group and 1348 cm⁻¹ (C─H) bends attributing to the aromatic group, which indicated the presence of apigenin⁴⁷.

In FTIR spectra of GLVs combination crude extracts, the results showed that the peak was almost in the same pattern as those of individual crude extracts, namely A. cepa, A. fistulosum, and A. graveolens L. (Supplementary Fig. S1). This FTIR band pattern was also seen in combination-1 (A. cepa + A. fistulosum + A. graveolens L.) (Fig. 3). The high value of AGI and TPC activity in combination-1 was also in agreement with the peaks with a higher amplitude that appears at 3179-3299 cm⁻¹, which indicated the presence of C─H and O─H bonds of a diallyl disulfide (thiosulfinates)⁴⁹ and phenolic compounds⁴⁶.

2.11 GC-MS

The chromatogram of A. cepa and combination-1 extracts is shown in Supplementary Fig. S2. The metabolites of A. cepa identified were primarily fructose, glucose, propanoic acid, 2,3,4-trihydroxybutyric acid, myo-inositol, and succinic acid (Supplementary Fig. S2a). The main compound of combination-1 was fructose, glucose, mannitol, myo-inositol, succinic acid, and propanoic acid (Supplementary Fig. S2b). All metabolites analyzed in both extracts are shown in Table 5. The compounds of A. cepa from GC-MS analysis have also been reported previously in A. cepa, including fructose, glucose, hexadecanoic acid, succinic acid, isoleucine, phenylalanine, proline, threonine, valine⁵⁰, arabinofuranose, myo-inositol, ribofuranose, sorbopyranose, galactose, fructose oxime, galactose oxime, 2,3,4-trihydroxybutyric acid, gluconic acid⁵¹, and propanoic acid⁵².

In combination-1, several compounds that appear in A. cepa were also seen in combination-1 and insignificant compounds in A. cepa were not detected in combination-1, such as ribofuranose, isoleucine, phenylalanine, gluconic acid, and 2,3,4-trihydroxybutyric acid. In addition, several different compounds from A. cepa appeared in combination-1, which might have come from A. graveolens L., which has been previously reported in A. graveolens L., including 2,3-butanediol³⁴, mannitol, alanine, asparagine⁴¹, galactitol⁵³, and 9,12-octadecadienoic acid⁴⁰. A synergistic combination of the compounds promoted the higher bioactivity of the GLVs. Therefore, these GC-MS-identified individual compounds could synergism and might be responsible for the synergistic antidiabetic and antioxidant effects.

3.1 Materials

Enzymes including α-amylase from porcine pancreas (Type IV-B), α-glucosidase, and lipase from porcine pancreas (Type II) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Chemicals such as p-nitrophenyl-α-D-glucopyranoside (p-NPG), acarbose, and 4-methylumbelliferyl oleate (4MUO) were purchased from Sigma-Aldrich (St. Louis, MO, USA). 2,2-Diphenyl-1-picrylhydrazil (DPPH) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Folin–Ciocalteu reagent, 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), 3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine-p,p-disulfonic acid monosodium salt hydrate (Ferrozine), ethylenediaminetetraacetic acid (EDTA), L-ascorbic acid, and gallic acid monohydrate were purchased from Acros Organics (Fair Lawn, NJ, USA). All chemicals were of analytical grade.

Fresh green leafy vegetables (GLVs), as listed in Table 1, were procured from the Hua Takhe market situated at Latkrabang, Bangkok, Thailand. The collection of these plant materials and the subsequent experimental research conducted on the GLVs followed the guidelines established by the Thailand national authorities. The identification of the GLVs was done by Dr. Karthikeyan Venkatachalam and Dr. Ali Muhammed Moula Ali using LeafSnap, the initial mobile application designed for automatic identification of plant species. Ten kilograms of each GLV were placed in the vegetable crates and transported to the School of Food-Industry, King Mongkut’s Institute of Technology Ladkrabang, Bangkok, Thailand. Upon arrival, all the GLVs were carefully washed using tap water to remove dirt, and drained on the screen to remove the residual water.

3.2 Sample preparation

The cleaned GLVs were dehydrated using a dehydration drier (WRH-100, IKE, Guangdong, China) at 45 °C until a constant weight was achieved. The dehydrated GLVs were ground to powder using a blender (Philips HR 2222, Philips Healthcare, Best, Netherlands). The powder sample was used for extraction.

3.3 Preparation of extracts

Powdered GLVs (100 g) were mixed in 80% ethanol at the ratio of 1:10 (w/v) and homogenized using a hand homogenizer (T25 ULTRA-TURRAX, IKA, Staufen, Germany) at 13000 rpm for 2 min at room temperature (25 ± 2 °C). The mixtures were then macerated at 4 °C for 24 h. The samples were centrifuged (5910 R, Eppendorf, Hamburg, Germany) at 2,500 xg for 15 min at 25 °C and filtered using Whatman No. 1 filter paper (Maidstone, Kent, UK). The obtained filtrates were concentrated using a vacuum rotary evaporator (Rotavapor R-300, Buchi, Flawil, Switzerland) at 40 °C and then freeze-dried (Model LD0.5, Kinetic, Bangkok, Thailand) for 72 h at -50 °C. The obtained extract in the form of powder was used for analysis and characterization.

3.4 Extraction yield

The dry weight determined the extraction yield in percentage (%). The extraction yield was calculated using the formula:

where w₁ is the obtained crude extract's dry weight after freeze-drying and w₂ is the dry weight of initial samples powder after dehydration drier.

3.4.1 Moisture content

The moisture content of fresh and dehydrated GLVs was determined following the AOAC method 930.04⁵⁴. The moisture content was expressed in percentage (%).

3.5 Analysis

The crude extracts in powder, obtained after freeze-drying, were used for analysis.

3.5.1 Determination of total phenolic content (TPC)

TPC was determined as described by Bavisetty and Venkatachalam⁵⁵ with slight modifications. The reaction mixture was prepared by mixing 100 μL samples (1 mg of crude extract were mixed in 1 mL of 50% methanol) with 200 μL 10% Folin–Ciocalteu reagent and 800 μL of 700 mM Na₂CO₃. The mixture was vortexed thoroughly and incubated at room temperature for 2 h. The absorbance was recorded at 765 nm, the standard curve was constructed using gallic acid, and TPC was expressed as gallic acid equivalents in mg GAE/g extract.

3.5.2 Antioxidant

To determine the antioxidant activities, samples were prepared by dissolving the crude extracts in absolute methanol at 1 mg/mL (w/v). The absorbance was measured using a microplate reader. DPPH radical scavenging and ferric reducing antioxidant power were expressed as L-ascorbic acid equivalent antioxidant capacity in mg AEAC/g extract. The metal chelating activity was defined as EDTA equivalent chelating capacity in mg EECC/g extract.

3.5.2.1 DPPH radical scavenging assay

DPPH radical scavenging assay was measured as described by Ali, et al.⁵⁶ with some modifications. The reaction was initiated by mixing 100 μL of prepared samples with 100 μL 0.2 mM DPPH solution and incubating in the dark for 30 min at room temperature. The absorbance was recorded at 517 nm.

3.5.2.2 Metal chelating assay

The metal chelating activity was determined as described by Masood, et al.²⁸ with slight changes. The reaction was initiated by mixing 100 μL prepared samples with that 25 μL of 0.6 mM FeCl₂ solution and 25 μL of 5 mM Ferrozine solution. The mixture was vortexed and incubated for 10 min at room temperature, and absorbance was recorded at 562 nm.

3.5.2.3 Ferric reducing antioxidant power (FRAP) assay

FRAP assay was performed as described by Ali, et al.⁵⁶. To the 25 μL of prepared samples, 175 μL of FRAP reagent [ten volumes of 300 mM acetate buffer pH 3.6, one volume of TPTZ solution (10 mM TPTZ in 40 mM in HCl), and one volume of 20 mM FeCl₃·6H₂O solution] was added. The reaction was incubated in the dark for 30 min, and absorbance was measured at 593 nm.

3.5.3 α-Amylase inhibitory (AAI) assay

The reaction was performed by mixing 20 μL of samples (extracts dissolved in 0.2 M phosphate buffer saline, pH 6.9 at the concentration 1 mg/mL) with that 20 μL of 0.2 M phosphate buffer saline (pH 6.9) containing 20 U/mL of α-amylase enzyme and incubated at 37 ± 2 °C for 10 min. Then, 30 μL of 0.5% starch solution (substrate) was added to the mix and set at 37 ± 2 °C for 8 min. The reaction was stopped by adding 20 μL of HCl (1 M) and 100 μL of iodine solution (0.25 mM). Then the absorbance was recorded at 565 nm. A standard curve was plotted using acarbose, and the inhibitory activity was expressed as acarbose equivalent in mmol ACE/g extract.

3.5.4 α-Glucosidase inhibitory (AGI) assay

AGI assay was performed by mixing 10 μL of samples (extracts were dissolved in DMSO: 10 mg/mL) with 50 μL of 0.1 M phosphate buffer (pH 6.9) and 25 μL of α-glucosidase enzyme solution (0.1 U/mL). Then the substrate 25 μL of p-NPG was added and incubated for 30 min at 37 ± 2 °C. The enzyme reaction was terminated by adding 100 μL of 0.2 M sodium carbonate. Finally, absorbance was recorded at 410 nm. A standard curve was plotted using acarbose, and the inhibitory activity was expressed as acarbose equivalent in mmol ACE/g extract.

3.5.5 Lipase inhibitory (LPI) assay

LPI assay was performed as described by Chatsumpun, et al.⁵⁷. The samples 25 µL (extracts were dissolved in a buffer containing 13 mM Tris-HCl, 1.3 mM CaCl₂, and 150 mM NaCl (pH 8.0), at the concentration 1 mg/mL) were mixed with substrate 50 µL of 0.5 mM 4-methylumbelliferyl oleate (4MUO). Then, 25 µL of lipase enzyme solution (50 U/mL) was added and incubated at 37 ± 2 °C for 30 min. The reaction was terminated by adding 100 µL of 0.1 M sodium citrate (pH 4.2). Fluorescence intensity was recorded at 355 nm (excitation) and 460 nm (emission). The percentage inhibition of enzyme activity was calculated using the formula:

where A₀ is the fluorescence intensity without extract, and A₁ is the fluorescence intensity with the extract.

3.5.6 Inhibitory kinetic assay of α-glucosidase against inhibitors

The α-glucosidase enzyme inhibition was performed as prior mentioned. The kinetic studies indicating inhibition mode against α-glucosidase was evaluated at different concentrations of positive control (acarbose with concentration of 0, 10 and 20 mg/mL), samples (A. cepa and combination-1 with concentration of 0, 5 and 10 mg/mL), and p-NPG various concentrations (0, 1.25, 2.5, 5, and 10 mM). The initial reaction rate (v) was computed at the substrate and sample various concentrations. The models were used as described by Mittal, et al.⁵⁸ as follows:

Lineweaver-Burk equation:

where v is the initial reaction rate of the substrate at varying concentrations, v_max is the maximum reaction rate, K_m is the Michaelis inhibition constants, and S refers to the substrate concentration.

3.5.7 Fourier-transform infrared (FTIR) spectroscopy

FTIR analysis was carried out as described by Ali, et al.⁵⁹. The 5 mg of powder of the crude extracts were subjected to a FITR spectrophotometer (Invenio-S 100183, Bruker, Ettlingen, Germany). The spectra ranging from 4000–400 cm^–1, 32 scans/min in absorbance mode at 4 cm–1 resolution was recorded and analyzed using OPUS software (OPUS ver. 4.2, Bruker, Ettlingen, Germany).

3.5.8 Gas chromatography-mass spectrometry (GC-MS)

Derivatization of extracts was performed as described by Murugesu, et al.⁶⁰ with minor modifications. The extract (25 mg) was dissolved in 500 µL pyridine, vortexed, and sonicated for 10 min using a bath sonicator CP200T (Crest Ultrasonics, Cortland, NY, USA). 100 µL methoxyamine HCl (20 mg/mL pyridine) was mixed and incubated for 2 h at 60 °C and followed by mixing with 300 µL of MSTFA (N-Methyl-N-(trimethylsilyl) trifluoroacetamide) containing 1% trimethylchlorosilane and incubating for 30 min at 60 °C. The derivatized samples were filtered into an autosampler GC-vial using a PTFE syringe filter (0.2 µm, 13 mm) (Agilent Technologies, Santa Clara, CA, USA) and left overnight at room temperature. The derivatized sample (1µL) was injected into the Agilent GC-MS system consisting of a 6890-chromatography gas and an HP-5973 mass selective detector. The GC column and carrier gas used was a 5% DB-5MS phenyl methyl siloxane column (inside diameter of 250 µm, thickness of 0.25 µm), and helium was used as a carrier gas (rate of 1 mL/min). The initial temperature was set at 50 °C for 3 min and increased to 315 °C for 10 min (rate of 10 °C/min). The injector and ion source temperatures were set at 330 °C and 250 °C, respectively. After a solvent delay of 7 min, mass spectra were acquired in full scan and monitoring mode over a mass scan range of 50 to 550 m/z. Each chromatogram peak and retention time of the metabolites were compared with those of Wiley7n.l database. Data were expressed in terms of the abundances of each peak and the percentage of the compound similarity from the library data.

3.5.9 Synergistic combination approach

The GLVs exhibiting the highest AGI, AAI, and LPI activities were combined with GLVs showing the most increased antioxidant activity. For the combination, three types of GLVs were mixed in equal proportions based on weight (1:1:1; w:w:w). Recovery of crude extract, characterization, and bioactivities was determined as prior mentioned for individual samples. The combinations of GLVs were done as noted in Table 1.

3.5.10 Statistical analysis

All the experiments were performed in triplicates, and the results were expressed as means ± SD. One-way ANOVA and significance levels were tested (Duncan Multiple Comparison tests) at the confidence level of p < 0.05. Pearson's correlation coefficient “r” (for normally distributed data) was calculated to express the correlation between parameters. SPSS software was used to analyze the data (SPSS for Windows 28.0, SPSS Inc., IL, USA).

This study highlighted the importance of GLVs and their combinations as a source of bioactive compounds with excellent anti-diabetic, anti-obesity, and antioxidant activities. A. graveolens L., C. sativum, and P. crispum showed potential antioxidant properties due to their high total phenolic content. Samples, A. cepa and P. crispum, exhibited potential antidiabetic and anti-obesity properties. The enzyme inhibition did not directly correlate with that TPC or antioxidant activities. Based on the combination studies, it was observed that all the activities, including antioxidant and enzyme inhibition, were enhanced significantly. Overall, combination-1 containing A. cepa + A. fistulosum + A. graveolens L. showed the potential to be considered an oral antidiabetic agent exhibiting a significant inhibition mechanism like acarbose. This work concludes that GLVs and their combinations have the potential as additives in developing functional foods to prevent and treat T2DM. Future studies on identifying and characterizing individual metabolites and their application in the development of functional food will be carried out to be applied in the food industry with specific health benefits.

Data availability

The first author can provide the datasets used and/or analyzed in the present study upon a reasonable inquiry.

CRediT authorship contribution statement

W.H.M.: conducting research activities, investigating, and writing some manuscript sections. K. V.: Revise the manuscript with critical inputs. A.K.R.: writing some sections and improvising the MS. S.K.: Revise the manuscript with critical inputs. S.C.B.B.: conduct some lab experiments with documenting and improvising the manuscript. A.M.M.A.: Work plan and design the experiments, correcting and critically improvising the manuscript.

Declaration of competing interest

The authors declare no conflict of interest.

Acknowledgment

This work was supported by King Mongkut’s Institute of Technology Ladkrabang. A.M. Moula Ali and S.C.B. Bavisetty would like to acknowledge the support provided by KRIS (King Mongkut’s Institute of Technology Ladkrabang, Research and Innovation Services) (Grant # 2564-02-07-003) and “National Science, Research and Innovation Fund,” Thailand (NSRF, #RE-KRIS/FF65/011).

Fauci, A. S., Lane, H. C. & Redfield, R. R. Covid-19—navigating the uncharted. N. Engl. J. Med. 382, 1268-1269, https://doi.org/10.1056/NEJMe2002387 (2020).
Leung, C. Clinical features of deaths in the novel coronavirus epidemic in China. Rev. Med. Virol. 30, e2103, https://doi.org/10.1002/rmv.2103 (2020).
Blahova, J. et al. Pharmaceutical drugs and natural therapeutic products for the treatment of type 2 diabetes mellitus. Pharmaceuticals 14, 806, https://doi.org/10.3390/ph14080806 (2021).
Papoutsis, K. et al. Fruit, vegetables, and mushrooms for the preparation of extracts with α-amylase and α-glucosidase inhibition properties: A review. Food Chem. 338, 128119, https://doi.org/10.1016/j.foodchem.2020.128119 (2021).
Adisakwattana, S., Ruengsamran, T., Kampa, P. & Sompong, W. In vitro inhibitory effects of plant-based foods and their combinations on intestinal α-glucosidase and pancreatic α-amylase. BMC Complement. Altern. Med. 12, 1-8, https://doi.org/10.1186/1472-6882-12-110 (2012).
Na, L. et al. The Harbin cohort study on diet, nutrition and chronic non-communicable diseases: study design and baseline characteristics. PLoS One 10, e0122598, https://doi.org/10.1371/journal.pone.0122598 (2015).
Kjøllesdal, M. et al. Consumption of fruits and vegetables and associations with risk factors for non-communicable diseases in the Yangon region of Myanmar: a cross-sectional study. BMJ open 6, e011649, https://doi.org/10.1136/bmjopen-2016-011649 (2016).
Phillips, T., McMichael, C. & O’Keefe, M. “We invited the disease to come to us”: neoliberal public health discourse and local understanding of non-communicable disease causation in Fiji. Crit. Public Health 28, 560-572, https://doi.org/10.1080/09581596.2017.1329521 (2018).
Natesh, H., Abbey, L. & Asiedu, S. An overview of nutritional and antinutritional factors in green leafy vegetables. Hortic. Int. J. 1, 58-65, https://doi.org/10.15406/hij.2017.01.00011 (2017).
Randhawa, M. A., Khan, A. A., Javed, M. S. & Sajid, M. W. in Handbook of Fertility (ed Ronald Ross Watson) 205-220 (Academic Press, 2015).
Mamluk, L. et al. Fruit and vegetable intake and risk of incident of type 2 diabetes: results from the consortium on health and ageing network of cohorts in Europe and the United States (CHANCES). Eur. J. Clin. Nutr. 71, 83-91, https://doi.org/10.1038/ejcn.2016.143 (2017).
Vinholes, J. & Vizzotto, M. Synergisms in alpha-glucosidase inhibition and antioxidant activity of Camellia sinensis L. Kuntze and Eugenia uniflora L. ethanolic extracts. Pharmacogn. Res. 9, 101, https://doi.org/10.4103/0974-8490.197797 (2017).
Paul, S. & Majumdar, M. Comparative study of six antidiabetic polyherbal formulation for its multimodal approaches in diabetes management. 3 Biotech 12, 1-11, https://doi.org/10.1007/s13205-022-03166-7 (2022).
Eruygur, N. et al. Screening the in vitro antioxidant, antimicrobial, anticholinesterase, antidiabetic activities of endemic Achillea cucullata (Asteraceae) ethanol extract. S. Afr. J. Bot. 120, 141-145, https://doi.org/10.1016/j.sajb.2018.04.001 (2019).
Babu, A., Kumaresan, G., Raj, V. A. A. & Velraj, R. Review of leaf drying: Mechanism and influencing parameters, drying methods, nutrient preservation, and mathematical models. Renewable Sustainable Energy Rev. 90, 536-556, https://doi.org/10.1016/j.rser.2018.04.002 (2018).
Lawal, U. et al. α‐Glucosidase inhibitory and antioxidant activities of different Ipomoea aquatica cultivars and LC–MS/MS profiling of the active cultivar. J. Food Biochem. 41, e12303, https://doi.org/10.1111/jfbc.12303 (2017).
Ribera, A., Bai, Y., Wolters, A.-M. A., van Treuren, R. & Kik, C. A review on the genetic resources, domestication and breeding history of spinach (Spinacia oleracea L.). Euphytica 216, 1-21, https://doi.org/10.1007/s10681-020-02585-y (2020).
Stramarkou, M., Papadaki, S., Kyriakopoulou, K. & Krokida, M. Effect of drying and extraction conditions on the recovery of bioactive compounds from Chlorella vulgaris. J. Appl. Phycol. 29, 2947-2960, https://doi.org/10.1007/s10811-017-1181-8 (2017).
Khatib, A., Perumal, V., Ahmed, Q., Uzir, B. & Murugesu, S. Low inhibition of alpha-glucosidase and xanthine oxidase activities of ethanol extract of Momordica charantia fruit. J. Pharm. Negat. Results. 8, https://doi.org/10.4103/0976-9234.204906 (2017).
Chandrashekharaiah, K. Antioxidant and type II diabetes-related enzyme inhibition properties of few selected medicinal plants. Biomed. Pharmacol. J. 6, 341, https://doi.org/10.13005/bpj/423 (2013).
Suttisansanee, U. et al. Phytochemicals and in vitro bioactivities of aqueous ethanolic extracts from common vegetables in Thai food. Plants 10, 1563, https://doi.org/10.3390/plants10081563 (2021).
Abdul Qadir, M., Shahzadi, S. K., Bashir, A., Munir, A. & Shahzad, S. Evaluation of phenolic compounds and antioxidant and antimicrobial activities of some common herbs. Int. J. Anal. Chem. 2017, 3475738, https://doi.org/10.1155/2017/3475738 (2017).
Al-Rimawi, F. et al. Anticancer activity, antioxidant activity, and phenolic and flavonoids content of wild Tragopogon porrifolius plant extracts. Evid.-based Complement. Altern. Med. 2016, 9612490, https://doi.org/10.1155/2016/9612490 (2016).
Tunnisa, F. et al. Antioxidant and antidiabetic compounds identification in several Indonesian underutilized Zingiberaceae spices using SPME-GC/MS-based volatilomics and in silico methods. Food Chem.: X 14, 100285, https://doi.org/10.1016/j.fochx.2022.100285 (2022).
Wong, P. Y. & Kitts, D. D. Studies on the dual antioxidant and antibacterial properties of parsley (Petroselinum crispum) and cilantro (Coriandrum sativum) extracts. Food Chem. 97, 505-515, https://doi.org/10.1016/j.foodchem.2005.05.031 (2006).
Sadeer, N. B., Montesano, D., Albrizio, S., Zengin, G. & Mahomoodally, M. F. The versatility of antioxidant assays in food science and safety—chemistry, applications, strengths, and limitations. Antioxidants 9, 709, https://doi.org/10.3390/antiox9080709 (2020).
Vinodhini, S., Vithiya, B. S. M. & Prasad, T. A. A. Green synthesis of silver nanoparticles by employing the Allium fistulosum, Tabernaemontana divaricate and Basella alba leaf extracts for antimicrobial applications. J. King Saud Univ. Sci. 34, 101939, https://doi.org/10.1016/j.jksus.2022.101939 (2022).
Masood, S. et al. Antioxidant potential and α-glucosidase inhibitory activity of onion (Allium cepa L.) peel and bulb extracts. Braz. J. Biol. 83, https://doi.org/10.1590/1519-6984.247168 (2021).
Heimann, E., Nyman, M. & Degerman, E. Propionic acid and butyric acid inhibit lipolysis and de novo lipogenesis and increase insulin-stimulated glucose uptake in primary rat adipocytes. Adipocyte 4, 81-88, https://doi.org/10.4161/21623945.2014.960694 (2015).
Ives, S. J. et al. The effect of succinic acid on the metabolic profile in high‐fat diet‐induced obesity and insulin resistance. Physiol. Rep. 8, e14630, https://doi.org/10.14814/phy2.14630 (2020).
Chhetri, D. R. Myo-inositol and its derivatives: their emerging role in the treatment of human diseases. Front. Pharmacol. 10, 1172, https://doi.org/10.3389/fphar.2019.01172 (2019).
Gori, M. et al. Palmitic acid affects intestinal epithelial barrier integrity and permeability in vitro. Antioxidants 9, 417, https://doi.org/10.3390/antiox9050417 (2020).
Moreno-Córdova, E. N. et al. Gallotannins are uncompetitive inhibitors of pancreatic lipase activity. Biophys. Chem. 264, 106409, https://doi.org/10.1016/j.bpc.2020.106409 (2020).
Khongrum, J., Yingthongchai, P. & Tateing, S. Exploring the synergistic effect of celery and Chinese kale on antioxidant and anti-pancreatic lipase and anti-acetylcholinesterase activities. Open Access J. Biomed. Sci. 3, 899-904, https://doi.org/10.38125/OAJBS.000264 (2021).
Jiang, H. w. et al. The evaluation of antioxidant interactions among 4 common vegetables using isobolographic analysis. J. Food Sci. 80, C1162-C1169, https://doi.org/10.1111/1750-3841.12896 (2015).
Zhang, L., Virgous, C. & Si, H. Synergistic anti-inflammatory effects and mechanisms of combined phytochemicals. J. Nutr. Biochem. 69, 19-30, https://doi.org/10.1016/j.jnutbio.2019.03.009 (2019).
Yang, Y. et al. Synergy effects of herb extracts: pharmacokinetics and pharmacodynamic basis. Fitoterapia 92, 133-147, https://doi.org/10.1016/j.fitote.2013.10.010 (2014).
Wu, S. et al. Ethanol extract of Sargarsum fusiforme alleviates HFD/STZ-induced hyperglycemia in association with modulation of gut microbiota and intestinal metabolites in type 2 diabetic mice. Food Res. Int. 147, 110550, https://doi.org/10.1016/j.foodres.2021.110550 (2021).
Yoon, S.-Y., Ahn, D., Hwang, J. Y., Kang, M. J. & Chung, S. J. Linoleic acid exerts antidiabetic effects by inhibiting protein tyrosine phosphatases associated with insulin resistance. J. Funct. Foods 83, 104532, https://doi.org/10.1016/j.jff.2021.104532 (2021).
Liu, D.-K. et al. Evaluation of bioactive components and antioxidant capacity of four celery (Apium graveolens L.) leaves and petioles. Int. J. Food Prop. 23, 1097-1109, https://doi.org/10.1080/10942912.2020.1778027 (2020).
Lau, H., Laserna, A. K. C. & Li, S. F. Y. 1H NMR-based metabolomics for the discrimination of celery (Apium graveolens L. var. dulce) from different geographical origins. Food Chem. 332, 127424, https://doi.org/10.1016/j.foodchem.2020.127424 (2020).
Jin, D. X., He, J. F., Zhang, K. Q., Luo, X. G. & Zhang, T. C. α‐Glucosidase inhibition action of major flavonoids identified from Hypericum attenuatum Choisy and their synergistic effects. Chem. Biodivers. 18, e2100244, https://doi.org/10.1002/cbdv.202100244 (2021).
Al-Malki, A. L. Inhibition of α-glucosidase by thiosulfinate as a target for glucose modulation in diabetic rats. Evid.-based Complement. Altern. Med. https://doi.org/10.1155/2016/7687915 (2016).
Son, H.-U. et al. Effects of synergistic inhibition on α-glucosidase by phytoalexins in soybeans. Biomolecules 9, 828, https://doi.org/10.3390/biom9120828 (2019).
Proença, C. et al. α-Glucosidase inhibition by flavonoids: an in vitro and in silico structure–activity relationship study. J. Enzyme Inhib. Med. Chem. 32, 1216-1228, https://doi.org/10.1080/14756366.2017.1368503 (2017).
Malapermal, V., Botha, I., Krishna, S. B. N. & Mbatha, J. N. Enhancing antidiabetic and antimicrobial performance of Ocimum basilicum, and Ocimum sanctum (L.) using silver nanoparticles. Saudi J. Biol. Sci. 24, 1294-1305, https://doi.org/10.1016/j.sjbs.2015.06.026 (2017).
Poureini, F., Mohammadi, M., Najafpour, G. D. & Nikzad, M. Comparative study on the extraction of apigenin from parsley leaves (Petroselinum crispum L.) by ultrasonic and microwave methods. Chem. Pap. 74, 3857-3871, https://doi.org/10.1007/s11696-020-01208-z (2020).
Li, C. et al. Two-dimensional visualization of nitrogen distribution in leaves of Chinese cabbage (Brassica rapa subsp. chinensis) by the Fourier transform infrared photoacoustic spectroscopy technique. J. Agric. Food Chem. 64, 7696-7701, https://doi.org/10.1021/acs.jafc.6b03234 (2016).
Nair, S. & Mukne, A. Assessment of chemical stability of constituents in thiosulfinate derivative-rich extract of garlic by a validated HPTLC method. Indian J. Pharm. Sci. 79, 438-450, https://doi.org/10.4172/pharmaceutical-sciences.1000247 (2017).
Sharma, K., Mahato, N., Nile, S. H., Lee, E. T. & Lee, Y. R. Economical and environmentally-friendly approaches for usage of onion (Allium cepa L.) waste. Food Funct. 7, 3354-3369, https://doi.org/10.1039/C6FO00251J (2016).
Romo-Pérez, M. et al. Metabolite profiling of onion landraces and the cold storage effect. Plant Physiol. Biochem. 146, 428-437, https://doi.org/10.1016/j.plaphy.2019.11.007 (2020).
Chowdhary, P. et al. Detection and identification of hazardous organic pollutants from distillery wastewater by GC-MS analysis and its phytotoxicity and genotoxicity evaluation by using Allium cepa and Cicer arietinum L. Chemosphere 297, 134123, https://doi.org/10.1016/j.chemosphere.2022.134123 (2022).
Woodyer, R. D., Wymer, N. J., Racine, F. M., Khan, S. N. & Saha, B. C. Efficient production of L-ribose with a recombinant Escherichia coli biocatalyst. Appl. Environ. Microbiol. 74, 2967-2975, https://doi.org/10.1128/aem.02768-07 (2008).
AOAC. Official methods of analysis (17 th). (Association of Official Analytical Chemists, 2000).
Bavisetty, S. C. B. & Venkatachalam, K. Physicochemical qualities and antioxidant properties of juice extracted from ripe and overripe wax apple as affected by pasteurization and sonication. J. Food Process. Preserv. 45, e15524, https://doi.org/10.1111/jfpp.15524 (2021).
Ali, A. M. M., Gullo, M., Rai, A. K. & Bavisetty, S. C. B. Bioconservation of iron and enhancement of antioxidant and antibacterial properties of chicken gizzard protein hydrolysate fermented by Pediococcus acidilactici ATTC 8042. J. Sci. Food Agric. 101, 2718-2726, https://doi.org/10.1002/jsfa.10898 (2021).
Chatsumpun, N., Sritularak, B. & Likhitwitayawuid, K. New biflavonoids with α-glucosidase and pancreatic lipase inhibitory activities from Boesenbergia rotunda. Molecules 22, https://doi.org/doi:10.3390/molecules22111862 (2017).
Mittal, A., Singh, A. & Benjakul, S. α-amylase inhibitory activity of chitooligosaccharide from shrimp shell chitosan and its epigallocatechin gallate conjugate: kinetics, fluorescence quenching and structure–activity relationship. Food Chem. 403, 134456, https://doi.org/10.1016/j.foodchem.2022.134456 (2023).
Ali, A. M. M., Prodpran, T. & Benjakul, S. Effect of squalene as a glycerol substitute on morphological and barrier properties of golden carp (Probarbus Jullieni) skin gelatin film. Food Hydrocoll. 97, 105201, https://doi.org/10.1016/j.foodhyd.2019.105201 (2019).
Murugesu, S. et al. Characterization of α-glucosidase inhibitors from Clinacanthus nutans Lindau leaves by gas chromatography-mass spectrometry-based metabolomics and molecular docking simulation. Molecules 23, 2402, https://doi.org/10.3390/molecules23092402 (2018).

Table 1. The scientific name, family, and common name of the individual of green leafy vegetables and samples code of green leafy vegetable combinations used.

Individual of green leafy vegetables
No.	Scientific name	Family	Common name
1.	Allium cepa	Amaryllidaceae	Spring onion
2.	Allium fistulosum	Amaryllidaceae	Bunching onion
3.	Apium graveolens L.	Apiaceae	Celery
4.	Brassica perviridis	Brassicaceae	Chinese spinach
5.	Brassica chinensis L.	Brassicaceae	Bok choy
6.	Coriandrum sativum	Apiaceae	Coriander
7.	Ipomoea aquatica Forsk	Convolvulaceae	Water spinach
8.	Ocimum basilicum	Lamiaceae	Basil
9.	Petroselinum crispum	Apiaceae	Parsley
10.	Spinacia oleracea L.	Amaranthaceae	Spinach
Combination of green leafy vegetables
Samples code	Combination of green leafy vegetables (1:1:1; w:w:w)
1.	Allium cepa + Allium fistulosum + Apium graveolens L.
2.	Allium fistulosum + Brassica perviridis + Apium graveolens L.
3.	Allium cepa + Brassica perviridis + Apium graveolens L.
4.	Allium cepa + Apium graveolens L. + Coriandrum sativum
5.	Allium cepa + Coriandrum sativum + Ipomoea aquatica Forsk
6.	Allium cepa + Apium graveolens L. + Ipomoea aquatica Forsk
7.	Petroselinum crispum + Spinacia oleracea L. + Brassica chinensis L.

Table 2. Total phenolic content, antioxidant, and enzyme inhibitory activities crude extracts from green leafy vegetables.

No.	Scientific name	Total phenolic content¹	DPPH radical scavenging activity²	Metal chelating activity³	Ferric reducing antioxidant power⁴	α-Amylase inhibitory activity⁵	α-Glucosidase inhibitory activity⁶	Lipase inhibitory activity⁷
1.	Allium cepa	11.76 ± 0.16	12.67 ± 1.03	11.07 ± 1.02	21.62 ± 1.58	10.16 ± 0.95	595.28 ± 48.97^a	17.74 ± 1.20
2.	Allium fistulosum	6.61 ± 0.43^f	8.99 ± 0.65^h	14.70 ± 1.22	7.30 ± 0.71^g	36.73 ± 2.79^a	282.97 ± 24.64	4.85 ± 0.47^e
3.	Apium graveolens L.	23.78 ± 1.46^a	125.57 ± 3.09^a	70.33 ± 5.65	34.78 ± 3.42	7.27 ± 0.21^e	130.50 ± 7.94	27.45 ± 1.17
4.	Brassica perviridis	9.21 ± 0.08	46.77 ± 3.47	53.07 ± 3.09	38.95 ± 1.95	8.85 ± 0.32	210.42 ± 3.15	26.65 ± 2.04
5.	Brassica chinensis L.	7.18 ± 0.37	28.74 ± 1.18	2.63 ± 0.22^f	27.05 ± 2.05	26.96 ± 1.21	136.81 ± 11.94	16.38 ± 1.23
6.	Coriandrum sativum	10.34 ± 0.44	119.42 ± 5.33	92.85 ± 5.23^a	46.98 ± 4.64^a	26.06 ± 1.56	131.55 ± 10.93	13.33 ± 1.21
7.	Ipomoea aquatica Forsk	11.92 ± 0.31	93.35 ± 3.19	8.93 ± 0.78	33.96 ± 2.28	17.32 ± 1.55	23.24 ± 1.82^g	16.33 ± 0.81
8.	Ocimum basilicum	6.20 ± 0.12	14.80 ± 1.29	11.89 ± 1.14	39.05 ± 3.90	34.46 ± 2.28	153.63 ± 8.35	6.64 ± 0.34
9.	Petroselinum crispum	19.94 ± 1.98	21.04 ± 1.46	77.96 ± 5.53	19.25 ± 0.97	30.61 ± 2.38	111.57 ± 7.94	31.07 ± 0.89^a
10.	Spinacia oleracea L.	10.14 ± 0.89	26.77 ± 1.36	8.78 ± 0.64	15.59 ± 1.47	14.50 ± 1.26	55.84 ± 5.46	14.13 ± 1.05
¹Results are expressed as gallic acid equivalents in mg GAE/g extract ²Results are expressed as mg of L-ascorbic acid equivalent radical scavenging capacity per g of extract ³Results are expressed as mg of EDTA equivalent chelating capacity per g of extract ⁴Results are expressed as mg of L-ascorbic acid equivalent reducing capacity per g of extract ⁵Results are expressed as mmol of acarbose equivalent inhibition capacity per g of extract ⁶Results are expressed as mmol of acarbose equivalent inhibition capacity per g of extract ⁷Results are expressed as percentage (%) inhibition of lipase enzyme activity Mean ± SD from triplicate. Different lowercase superscripts in the same column indicate significant difference (p < 0.05).

Table 3. Total phenolic content, antioxidant, and enzyme inhibitory activities of crude extracts from green leafy vegetables, in synergistic combinations.

Samples code	Total phenolic content¹	DPPH radical scavenging activity²	α-Amylase inhibitory activity³	α-Glucosidase inhibitory activity⁴	Lipase inhibitory activity⁵
1	20.64 ± 0.65	40.56 ± 3.47	30.61 ± 2.48	974.46 ± 81.57^a	36.80 ± 1.06
2	19.79 ± 0.85	43.33 ± 1.67	29.78 ± 1.64	502.59 ± 39.28	36.60 ± 0.79
3	20.51 ± 0.72	151.67 ± 9.28	17.05 ± 1.52^c	220.25 ± 14.97	42.79 ± 0.98
4	19.86 ± 0.67	671.67 ± 6.01^a	29.44 ± 2.66	217.64 ± 17.43^e	28.92 ± 2.53
5	20.31 ± 0.64	318.89 ± 15.75	36.39 ± 3.25	224.82 ± 19.74	14.91 ± 1.40^e
6	20.18 ± 0.46	137.78 ± 13.37	18.56 ± 1.38	427.43 ± 41.08	23.42 ± 1.07
7	19.06 ± 0.89	14.44 ± 0.96^e	39.69 ± 3.32^a	302.60 ± 27.89	43.06 ± 2.54^a
¹Results are expressed as gallic acid equivalents in mg GAE/g extract ²Results are expressed as mg of L-ascorbic acid equivalent radical scavenging capacity per g of extract ³Results are expressed as mmol of acarbose equivalent inhibition capacity per g of extract ⁴Results are expressed as mmol of acarbose equivalent inhibition capacity per g of extract ⁵Results are expressed as percentage (%) inhibition of lipase enzyme activity Mean ± SD from triplicate. Different lowercase superscripts in the same column indicate significant difference (p < 0.05).

Table 4. Enzyme inhibitory activity and detailed kinetics of α-glucosidase inhibition by Allium cepa, combination-1, and acarbose at different concentrations.

Samples		Concentration (mg/mL)					Inhibition type
		0	5		10
Allium cepa	v_max^*	0.0072	0.0044		0.0029		Non-competitive
	K_m^**	3.9701	4.2757		4.6219
Combination-1	v_max^*	0.0072	0.0069		0.0068		Competitive
	K_m^**	3.9701	7.1734		10.9002
Positive control		Concentration (mg/mL)
		0		10		20
Acarbose	v_max^*	0.0072		0.0225		0.0828	Competitive
	K_m^**	3.9701		89.7802		517.4381

*v_max and **K_m were expressed in mM p-Nitrophenol/min and mg/mL, respectively.

Table 5. The metabolites of Allium cepa and combination-1 extracts by GC-MS.

Compounds	*Allium cepa*		Combination-1
Compounds	Abundance¹	Similarity (%)	Abundance¹	Similarity (%)
Sugars
Arabinofuranose	13.93	52	8.34	58
Fructose	63.34	91	68.97	93
Fructose oxime	52.47	90	50.39	91
Galactose	0.61	80	38.51	94
Galactose oxime	7.49	90	6.65	87
Glucose	57.26	95	62.82	95
Myo-inositol	11.32	86	5.62	58
Ribofuranose	0.75	87	-	-
Sorbopyranose	8.10	87	6.02	91
Sugar alcohols
2,3-Butanediol	-	-	0.62	80
Galactitol	-	-	5.74	76
Mannitol	-	-	34.12	86
Amino acids
Alanine	-	-	1.56	90
Asparagine	-	-	1.65	98
Isoleucine	0.62	86	-	-
Phenylalanine	0.64	68	-	-
Proline	3.32	90	3.00	96
Threonine	0.78	91	0.59	91
Valine	1.43	87	1.13	91
Fatty acids
Hexadecanoic acid	0.87	99	0.48	72
9,12-Octadecadienoic acid	-	-	0.53	86
Propanoic acid	13.70	99	3.70	99
Organic acids
Gluconic acid	1.45	90	-	-
Succinic acid	9.00	98	3.64	99
2,3,4-Trihydroxybutyric acid	4.40	86	-	-

¹Values are expressed as abundance (x 10⁵).

-, not detected.

No competing interests reported.

Supplementarydata.docx

Selective green leafy vegetables and their synergistic combination approach as natural anti-diabetic agents: therapeutic potential

Status:

Version 1

Abstract

Figures

1. Introduction

2. Results and Discussion

3. Materials and methods

4. Conclusion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1