An ethnobotanical survey was conducted by Tipaalaga (an association working for sustainable management of natural resources) members to determine plants used in the diet of people with diabetes. This study was carried out from November 22 to December 6, 2019, in Ouagadougou, Pabré, Gampela, and surrounding villages. The research was focused on people without distinction of age, religion, and sex. The interviews were based on a pre-tested questionnaire with specific questions about the informant, the vernacular name of the plant species, the edible parts, and the preparation [15]. The interviews were recorded using a dictaphone. A total of 145 people were interviewed, including 69 people with diabetes, with an average age of 45. Plant samples were collected and identified by the botanical team of Joseph KI-ZERBO University. A herbarium was made.

The plant material was composed of three herbaceous plants: Apium graveolens L. (Ag), Cleome gynandra L. (Cg), and Hibiscus sabdariffa L. (Hs). Plants samples were collected in December 2020 from the botanical garden of Gampela, located 15 km from Ouagadougou. Samples were dried separately at room temperature under ventilation. Samples were powdered into fine powder by an electrical grinder.

All the solvents used without acetonitrile and n-hexane (HPLC grade) were of analytical grade and purchased from Sigma-Aldrich (Taufkirchen, Germany). Water was purified by a Millipore instrument (MOLSHEIM France). Gallic acid, 2,2-diphenyl-1-picrylhydrazyl, Folin-Ciocalteu phenol reagent were obtained from Sigma Chemical Co. (St. Louis, MO). β-carotene (HPLC grade) and Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) standards were purchased from Sigma-Aldrich (St. Louis, MO, USA). Buffer salts and all other chemicals were of analytical grade. Samples and standards solutions were filtered before chromatographic analysis using through Millipore membrane of 0.2 µm.

Twenty-five grams of powder of each herbaceous were extracted by maceration at low temperature (4˚C) for 24 hours with 200 mL of n-hexane. The experiment was repeated much time until the sample was colorless. The different filtrates were collected and concentrated at 40˚C using a rotary evaporator (Buchi). The n-hexane extracts were utilized for non-polar compounds screening by the HPTLC method and quantification of carotenoid contents and lipophilic antioxidant capacity array. The dry residue was taken up with 200 mL of methanol at low temperature (4˚C) for 24 hours. The process was repeated until a colorless sample. After filtration with filter paper, methanol extracts were collected and dried under vacuum on the rotatory evaporator at low temperature (<40˚C). The dried extracts were re-dissolved to a minimum volume of methanol for polar compounds screening and hydrophilic antioxidant activity and phenolic compounds evaluation by spectrophotometer analysis.

The total phenolic contents (TPC) in the herbaceous extracts were determined by the Folin-Ciocalteu colorimetric method [8] [18] with slight modifications. In brief, 0.1 mL of plant extract and gallic acid solution was mixed with 0.1 mL of Folin-Ciocalteu reagent previously diluted ten times with distilled water. After vortexing, the mixture was incubated for 8 min at room temperature, and 1.0 mL of 7.5% saturated sodium carbonate solution was added. The combinations were set at 37˚C for 30 min in the dark. The absorbances of the resulting blue color were read at 760 nm with a SHIMADZU UV-Vis spectrophotometer. The phenolic contents of plant extracts were determined using the equation of the calibration curve (y = 7.3194x + 0.1309, R² = 0.998), and the results are expressed in mg of gallic acid equivalents (GAE) per gram of dry weight. All determinations were performed in triplicate (n = 3).

The determination of total flavonoid contents (TFC) in the herbaceous extracts was performed as reported previously [18] [19] using AlCl₃ colorimetric method. Rutin was used as standard, and the quantification was expressed by reporting the absorbance in the calibration curve of the rutin (2.5608x + 0.0034, R² = 0.9995). The total flavonoid contents were expressed as mg rutin equivalents/g dry weight. All determinations were performed in triplicate (n = 3).

HPLC-DAD analysis was carried on a Prominence UFCL system (Shimadzu, Kyoto, Japan) equipment with two Nexera X2 LC-30AD high-pressure gradient pumps, a Prominence DGU-20A5R degasser, a Nexera SIL-30AC Prominence autosampler (15˚C, injection volume 10 µL), a CTO-20AC Prominence column over at 25˚C and SPD-M20A Prominence diode array detector. Data acquisition and processing were performed by the LabSolutions software (Shimadzu, Kyoto, Japan).

Chromatograms were recorded at 450 nm, and carotenoids were quantified as β-carotene equivalents using an external calibration curve. The β-carotene solution was obtained by dissolving the standard in hexane containing 0.1% BHT. The β-carotene solution’s concentration was measured at 450 nm by a spectrophotometer (Shimadzu, Kyoto, Japan) and calculated with the molar absorption coefficient of β-carotene in light hexane (138,900 L∙mol⁻¹∙cm⁻¹) [20]. The total carotenoid contents were expressed as mg β-carotene equivalents/g dry weight. All the samples were analyzed in triplicate.

Hydrophilic and lipophilic antioxidant activities were measured using the DPPH bleaching method [8] [21] [22]. The process is based on the capacity of plant extracts to scavenge compared to the DPPH radical (DPPH^•) to the Trolox in a dose-response curve. DPPH radical absorbs in the visible at λ_max = 515 nm and disappears with reduction by an antioxidant compound [8] [23]. Briefly 1.5 mL of the DPPH^• solution (0.04 mg/mL in methanol) was mixed with 500 µL of plant extract. After 8 min of incubation at 37˚C in the dark, the absorbance decrease at 515 nm was measured spectrophotometrically (SHIMADZU). Briefly, new standard solutions (Trolox) of different concentrations from 1 to 5 µg/mL in methanol were firstly prepared. 2 mL of each standard and sample were mixed with 2 mL of 0.01 mM DPPH solution, respectively. The mixtures were incubated in the dark at 37˚C for 30 min. The radical scavenging rate of DPPH was calculated as follow:

R Radical scavenging rate = 1 − ( A s − A s 0 ) / A 0

A_s: absorbance of the sample; A_s0: absorbance of the samples with 2 mL methanol; A₀: absorbance of DPPH (0.01 Mm) without the sample.

The importance of each plant is determined by calculating its Use Value species (UVs) according to the simplified formula of Cotton and Wilkie:

U V s = U N [24].

where U denotes the number of uses of the plant and N the number of informants. The data were processed and analyzed with the software SPSS version 15. Mean use values were compared using the One Way ANOVA (Analysis of Variance). Differences are considered statistically significant for a “p-value less than 0.05”.

Leafy vegetables cited in the diet of diabetics are Spinacia oleracea (Chénopodiaceae), Ipomoea batatas (L.) Lam. (Convolvulaceae), Ocimum gratissimum L. (Lamiaceae), Thymus vulgaris (Lamiaceae), Petroselinum crispum L. (Apiaceae), Corchorus olitorius L. (Malvaceae), Cleome gynandra L. (Brassicaceae), Hibiscus sabdariffa L. (Malvaceae), Apium graveolens L. (Apiaceae).

Three plants were distinguished from the others by their high use value (UVs ≥ 0.60): Apium graveolens, Cleome gynandra, and Hibiscus sabdariffa. The leaves of these plants are usually boiled, wrung out, and eaten or mixed with flour in the form of couscous. Some leaves can be eaten as a salad.

Phytochemical screening of different plant extracts was performed on HPTLC plates (glass) silica gel F₅₄ (Merck), and the results were presented in Figures 1-3. The different colors of spots (orange, yellow, blue, green, pink, purple) observed on the chromatogram under UV/366 nm, and white light may correspond to several secondary metabolite groups. To specify the nature of the compounds revealed at UV/366 nm, Natural Products reagent (flavonoids), Fast Blue Salt B reagent (tannins), and Liebermann Burchard reagent (sterols) were used.

The total phenolic, flavonoid, and carotenoid contents of Apium graveolens, Cleome gynandra, and Hibiscus sabdariffa extracts measured in methanol (TPC and TFC) and n-hexane (TCC) were presented in Table 1. Phenolic contents varied from 86.5 ± 0.7 for Apium graveolens (Ag) to 138.4 ± 0.5 mg Gallic Acid Equivalents/g dry weight for Hibiscus sabdariffa (Hs). The methanol extract of Hibiscus sabdariffa showed the highest total phenolic compound than Cleome gynandra and Apium graveolens respectively. The three species’ total phenolic contents were increased in the following order: Apium graveolens < Cleome gynandra < Hibiscus sabdariffa. Significant differences between the results were liked to genotypic and environmental differences within species, time of taking samples, and determination methods [31]. Typical phenolics that possess antioxidant activity are known to be mainly phenolic acids and flavonoids. The total phenolic content measured by the Folin-Ciocalteu method does not give a full picture of the quality or quantity of phenolic constituents. The total flavonoid contents (TFC) showed a similar trend, varying from 29.2 ± 0.6 for Apium graveolens to 52.8 ± 0.6 mg Rutin Equivalents/g dry weight for Hibiscus sabdariffa. As presented in Table 1, the total phenolic and flavonoid contents are ranked in the following order: Apium graveolens < Cleome gynandra < Hibiscus sabdariffa. The total flavonoid contents showed a linear correlation with the total phenolic contents (r = 0.91) (Table 2).

Data are expressed as mean ± SE of triplicate experiments. The means in each column, followed by a different letter, are significantly different (P < 0.05).

In addition to phenolic compounds, carotenoids play an essential role in preventing human diseases and maintaining good health as part of a balanced diet [32]. The total carotenoid contents (TCC) were measured in n-hexane by the high-performance liquid chromatography method. TCC values ranged from 5.2 ± 0.004 for Apium graveolens to 24.4 ± 0.09 mg β-carotene Equivalents/g dry weight for Cleome gynandra (Table 1). As shown in Table 1, the hexane crude extract of Cleome gynandra presents the high content of total carotenoid, followed by Hibiscus sabdariffa, and Apium graveolens, respectively.

Plant samples’ antioxidant properties have been attributed to their bioactive compounds, being the most important carotenoid compounds and phenolic compounds [21]. The antioxidant activity of three herbaceous samples (Apium graveolens, Cleome gynandra, and Hibiscus sabdariffa) was determined using the established assay that uses the radical DPPH^• as chromogen. This method permits the assay of hydrophilic and lipophilic antioxidant activities in the same sample. Table 1 showed the hydrophilic and lipophilic antioxidant activities in the three herbaceous plant studies. As illustrated in Table 1, the three herbaceous samples’ hydrophilic antioxidant activities represented a significant contribution in all cases. In opposite, the lipophilic antioxidant capacities were less important. From the results, it could be concluded that the antioxidant activity level depends on the herbaceous variety. The correlations of hydrophilic antioxidant activity with total phenolic (r = 0.86) and flavonoid (r = 0.99) contents were significant. Among herbaceous plant studies, Hibiscus sabdariffa has exceptionally high antioxidant activity, as demonstrated by DPPH assay. Hibiscus sabdariffa also presented high phenolic and flavonoid contents (Table 1), two primary natural hydrophilic antioxidants [33]. Some epidemiological reports showed that antioxidant consumption could significantly influence health [33]. It was reported in the literature that natural plant products could prevent and management of non-transmissible diseases with minimal side effects and toxicity [29] [34]. Most often, the diabetic person is currently stressed. This situation disrupts hormonal function and causes a rise in cortisol levels, which leads to an increase in blood sugar levels.

However, there is no correlation between carotenoid contents and lipophilic antioxidant activities, which can be due to the low carotenoid contents observed in the herbaceous plants. It was reported that the β-carotene could not be the significant lipophilic antioxidant in certain plant samples explaining the no correlation with the lipophilic antioxidant activity. The significant correlation of hydrophilic antioxidant capacity with the phenolic compounds and flavonoid contents indicated that the Folin-Ciocalteu could be used to determine the hydrophilic antioxidant activity of herbaceous samples.