Polyphenols in cassava leaves (Manihot esculenta Crantz) and their stability in antioxidant potential after in vitro gastrointestinal digestion.

The study was carried out to assess the effect of variety on polyphenols in cassava leaves and their stability in antioxidant activity before and after in vitro gastrointestinal digestion. The results showed that individual and total polyphenols content (TPC) and antioxidant activity of bound, free and bioaccessible polyphenols were significantly (p < 0.05) influenced by variety at harvesting maturity. The bound polyphenols had lower TPC (5.00-19.16 mg GAE/g) than free (39.16-89.61 mg GAE/g) throughout harvesting maturity. The polyphenols were strongly affected after in vitro digestion, however, salicylic, syringic and benzoic acids are the most bioaccessible. The free polyphenols of variety IRAD4115 had the highest value of FRAP (35.17 μg TE/g) at 12 months after planting (MAP), while, bound polyphenols showed the lowest DPPH (6.59 μg TE/g, variety EN at 12MAP). The antioxidant activity value evaluated by DPPH method was decreased significantly after in vitro gastrointestinal digestion. However, there was no significant difference between antioxidant activity of bioaccessible polyphenols (77.71 μg TE/g) and methanolic polyphenols (79.17 μg TE/g) assessed by FRAP method. These findings showed the stability of antioxidant potential of polyphenols in cassava leaves harvested at different periods after in vitro digestion. Thus cassava leaves harvested at appropriate maturity can be used as ingredient of functional food for nutraceutical benefits.

Introduction

A great portion of the World's population has malnutrition problems, which can be defined as a state of insufficient supply of quantitative and qualitative nutrients. Consequently, non-communicable diseases such as obesity, diabetes, cancer and cardiovascular diseases are becoming important diseases affecting the world population in the different regions. The vulgarisation of dietary diversities rich in nutrients, pigments and polyphenols which are excellent source of natural antioxidants may perhaps help populations to improve their life style. Many studies have shown that vegetables contain health improving compounds, and their consumption may thus help in the prevention of diseases (Gatto et al., 2016; Zhang et al., 2017). Despite the toxic cyanogen levels in cassava leaves, they are consumed as vegetable in various forms in the different parts of the World. The toxic potential of cassava leaf depending on the variety and plant age, hence, proper processing methods should be addressed to cassava leaf before its human consumption (Latif and Müller, 2015). The leaves can be harvested at any growth stages of plant, and many studies have proven that apart from being rich in nutritional composition, they have good concentrations of phytochemicals with high antioxidant activity (Koubala et al., 2015; Oresegun et al., 2016). It is reported that phenolics composition and antioxidant activity of vegetables are greatly affected by seasonal variation and environmental stress (Garcia-Diaz et al., 2018). The polyphenols in vegetables exist both as bound and free fraction (Sosulski et al., 1982; Chandrika et al., 2006; Zhang et al., 2017), and the bound fraction is reported to be more active than free fraction in grains (Kotaskova et al., 2016). The bound polyphenols may be released continuously in a slow manner through the gastrointestinal tract. Their abundant release in the gut during bacterial fermentation, can improve their bioaccessibility and bioavailability (Shahidi and Yeo, 2016). However, in the case of cassava leaves, there is no information about the differences in polyphenols composition in the bound, free fractions and bioaccessibility at different stages of maturity, and seasonal variation are not reported. The levels of bound and free polyphenols as well as their levels after in vitro gastrointestinal digestion of cassava leaves may be affected by the plant maturity, and varietal differences may also influence their antioxidant activity.

Therefore, the present study aimed to assess total polyphenol contents and in vitro antioxidant proprieties of bound and free fraction, and bioaccessible polyphenols of cassava leaves as affected by variety and harvesting maturity.

Materials and methods

Materials

Cassava (Manihot esculenta Crantz) leaves from five varieties such as EN and AD (local varieties), TMS92/0326 and TMS96/1414 from IITA (International Institute of Tropical Agriculture, Nigeria) (improved varieties) and IRAD4115 from IRAD (Institute of Agricultural Research for Development) in Adamawa region of Cameroon (improved variety) were used. Plants were grown in natural conditions in Tokombéré subdivision, the Far North Region of Cameroon, and their tender leaves were harvested in the morning during different plant maturity stages (months after planting, MAP), such as 6 MAP, 9 MAP, 12 MAP and 15 MAP. The cassava leaves were washed with tap water, cut into pieces and dried in an air oven (UN75 Memmert, 30–750, Germany) at 60 °C for 24 h. The dried samples were reduced into a powder, sieved and stored in opaque black polyethylene bag (Handgards®, Texas, USA) prior to analysis.

Chemicals and reagents

All standards were obtained from Sigma-Aldrich (Bengaluru, India). These standards were p-hydroxybenzoic, benzoic, gallic, vanillic, salicylic, protocatechuic, gentisic, syringic and Trolox. Other chemicals and reagents used were of analytical grade purchased from Sisco Research Laboratory (Bengaluru, India).

Free and bound polyphenolics extraction

Free and bound phenolic compounds were extracted following the method of Chen et al. (2017) with slight modifications. 20 ml of methanol was added into the fine powder (300 mg dry leaf sample) in triplicate for each analysis and mixed overnight by placing the test tube on shakers at room temperature (25 °C). The mixture was centrifuged at 10000 rpm for 20 min at 4 °C. The extraction solution was filtered with Whatman paper no.1. Then, the residue was washed two times using 5 ml of methanol and then centrifuged. The supernatants were combined and centrifuged in the same conditions as described above and filtered with Whatman paper no.1 to obtain the free polyphenol fraction. The solvent was evaporated at 45 °C to dryness and dissolved in methanol (HPLC grade) before being stored at -20 °C for subsequent uses. Concerning bound phenolics, the residue resulting from free phenolic extraction was mixed with 10 ml NaOH (2N) solution and boiled in shaking water bath at 50 rpm for 30 min before adding 5 ml HCl (2M) and then incubated for 60 min at 60 °C. The extraction mixture was cooled and 10 ml of ethyl acetate was added into the extraction solution. After 1 h shaking, the extraction solution was filtered and the ethyl acetate fraction was centrifuged under the same conditions as described above. The residue obtained after filtration was further mixed with 10 ml of ethyl acetate and centrifuged in the same conditions as described above before filtering through Whatman paper no.1. Finally, the combined supernatant was evaporated to dryness at 45 °C. The residue obtained was dissolved in methanol and centrifuged in the same conditions described previously to obtain the free polyphenol fraction. The methanol extract was kept at -20 °C until analysis.

In vitro gastrointestinal digestion and extraction of polyphenols

The method described by Ydjedd et al. (2017) with some changes was used. In short, the simulated solution was prepared in phosphate buffer (1M, pH 6.0). The digestion was started with the introduction of 1.5 mL of simulated salivary solution to both fraction (500 mg) containing 5 mg of porcine α-amylase (1000 U/mg). The pH of the mixture was adjusted to 7.0 before stirring for 10 min in a shaking water bath at 37 °C for 90 rpm. The next step consisted in adding 1.5 mL of the gastric digestion stimulating solution (10 mg/mL) consisting of pepsin (1:3000) initially diluted in HCl (0.1N). The pH was adjusted to 2 with HCl (6M) and the mixture was incubated at 37 °C in a shaking water bath (90 rpm) for 2 h. To simulate intestinal digestion, 2 mL of the intestinal solution (10 mg/mL) of pancreatin (100 U/mg) and 4 mg/mL of bile salts were added, respectively. The pH was adjusted to 7.0 with NaOH (1M), and the mixture was incubated for 3 h in a shaking water bath at a speed of 90 rpm. The enzymatic reaction was stopped by adding 1 mL of methanol. The supernatant was collected and after that 1 mL of methanol (HPLC grade) was added to the residue fraction then centrifuged at 4 °C at 10,000 rpm for 10 min. After filtration through Whatman paper no.1, the solvent of combined supernatant was evaporated. The residue resulting was dissolved in methanol (HPLC grade) and filtered once with the 0.22 μm membrane (polypropylene, USA) and kept at -20 °C until the analysis.

Determination of total polyphenols

Bound, free and bioaccesible phenolics were determined using Folin Ciocalteu's reagent using microplate reader according to the protocol as described by Rocchetti et al. (2017) with some modifications. Briefly, 25 μL of standard or sample was mixed with 125 μL 0.2 M Folin-Ciocalteu reagent in a 96-well microplates. The microplate was covered with aluminium foil, and incubated for 10 min at room temperature. After incubation, 125 μL of 7.5 % sodium carbonate solution was added to each well. The microplate was shaken for 40 s and incubated for 40 min at room temperature before taking the absorbance at 765 nm using multimode microplate reader (Tecan SPARK 10M, V1.2.20, Austria). The standard curve plotted with gallic acid was used to evaluate the concentration of polyphenols.

Condition for HPLC-DAD analysis of individual phenolics

HPLC system (Hitachi Elite LaChrom, Hitachi High Technologies Japan, Tokyo, Japan) equipped with a Shodex C18-120-5 4E column, DAD detector and auto sampler, was used for quantification of phenolics. The C18 column (4.6 × 250 mm, 5 μm) (Supelco. Co) was used for separation of phenolics. System control and data acquisition were performed by Empower 3 Software (2010) (Corporation, Kyoto, Japan). Fractions were filtered with RC 0.45 μm (Phenex, USA) and filled in HPLC vials (Waters1.5 ml, USA). The mobile phases consisted of: (A) Milli-Q water containing 1 % acetic acid (v/v) and (B) methanol containing 1 % acetic acid (v/v). The column temperature was maintained at 30 °C and 20 μl of each sample was automatically injected. Elution was carried out at a flow rate of 1 ml/min. The linear gradient was formed as follows: 0–7 min (15 % B), 7–30 min (50 % B), 30–50 min (100 % B), 50–60 min (0 % B), 60–70 min (0 % B). Phenolics were monitored at 280, 320 and 360 nm. Peak of each phenolic was identified with authentic standard by comparing the retention time, and their quantification was done using standard curves.

Antioxidant activity assays

Ferric reducing antioxidant power (FRAP)

The FRAP was carried out according to the method described by Feregrino-Perez et al. (2008) with minor changes. Using a micropipette, 20 μL of the sample, blank or standard was added to 120 μL of the FRAP reagent in a 96 well microplates and the mixture was shaken for 50 s. The mixture was incubated at room temperature for 40 min. The absorbance was recorded at 595 nm with multimode reader (Tecan SPARK 10M, V1.2.20) and the results were determined using the Trolox standard and represented as Trolox Equivalent (TE).

2.2-diphenyl-1-picyhydrazyl (DPPH)

The DPPH radical scavenging activity of the cassava leaves was carried out according to the method described by Feregrino-Perez et al. (2008) modified by Rocchetti et al. (2017) with some modifications. To the sample solution (40 μl), 150 μl of DPPH (1 mM) was added in a 96 well microplates. The solution was mixed for 40 s using microplate reader, and incubated for 40 min in the dark at 35 °C. After incubation, the absorbance was recorded at 517 nm with multimode reader (Tecan SPARK 10M, V1.2.20). The results were evaluated from the Trolox standard.

Statistical analysis

All data were analyzed by ANOVA using Statgraphics software (version. 16). Tukey's test was used to determine any significant difference between different varieties in the same month and in all harvest age for each variety and significance was accepted at level p < 0.05. The results were expressed as means ± standard deviation. All experiments were done in four replicates except individual phenolics.

Results and discussion

Effect of variety on total polyphenol content (TPC) of bound and free polyphenols of dry leaf without digestion

The TPC in free and bound polyphenol fractions of tender cassava leaves at different harvest maturity differed significantly (p < 0.05) for each variety (Table 1). For bound polyphenols, the TPC of variety EN (19.16 mg GAE/g) was significantly higher at 15 MAP compared to the others through the various harvests. However, the variety IRAD4115 (89.61 mg GAE/g) showed the highest concentration of TPC in free polyphenols at 12 MAP (Table 1). The lowest value of TPC in bound polyphenols was shown by variety TMS92/0326 (5.00 mg GAE/g) at 15 MAP which coincided with the rainy season. However, the results suggest that the accumulation of bound and free polyphenols may not only depend on varietal differences and season but also may be impacted by plant maturity. Similarly, some observations reported in literature indicated that the amounts of polyphenol in plants varied with respect to varieties and age maturity (Perez-Lopez et al., 2007). Zhang et al. (2017) also reported that the bound polyphenols contained lower polyphenols than free fraction in sweet corn, and the results of present study are consistent with the findings of Garcia-Diaz et al. (2018) who reported the significant effect of season variation on bioactive accumulation in beans.

Table 1

Effect of variety on total polyphenol contents (mg GAE/gdw) of bound and free polyphenols in cassava leaves of undigested sample (methanolic extract of dry leaf without digestion).

Harvest maturity	Varieties	Total polyphenol contents
		Bound polyphenols	Free polyphenols
6MAP	TMS92/0326	15.14 ± 0.18aA	40.57 ± 0.28dD
	TMS96/1414	12.32 ± 0.23bB	42.63 ± 0.82cD
	IRAD4115	7.00 ± 0.27cD	39.16 ± 1.51dD
	EN	6.12 ± 0.46dD	84.20 ± 0.87aA
	AD	5.19 ± 0.13eD	52.54 ± 0.07bB
9MAP	TMS92/0326	10.52 ± 0.34aB	41.91 ± 0.52dC
	TMS96/1414	8.15 ± 0.56dcC	77.23 ± 0.53bA
	IRAD4115	8.35 ± 0.09dcB	81.38 ± 2.86aB
	EN	10.04 ± 0.28bB	60.52 ± 1.52cB
	AD	8.79 ± 0.39cC	75.07 ± 0.46bA
12MAP	TMS92/0326	8.96 ± 0.52dC	47.17 ± 1.02cB
	TMS96/1414	15.04 ± 0.38aA	53.24 ± 1.14bC
	IRAD4115	11.63 ± 0.29bA	89.61 ± 0.87aA
	EN	8.43 ± 0.35dC	45.90 ± 1.67dD
	AD	9.72 ± 0.26cAB	44.09 ± 0.72dD
15MAP	TMS92/0326	5.00 ± 0.27eD	69.76 ± 0.79aA
	TMS96/1414	7.58 ± 0.04dCD	59.49 ± 0.79bB
	IRAD4115	8.51 ± 0.29cBC	70.64 ± 0.67aC
	EN	19.16 ± 0.36aA	57.19 ± 1.48cC
	AD	10.13 ± 0.19bA	51.27 ± 0.68dC

MAP: month after planting; Cassava leaves harvested at 6MAP (onset dry season), 9MAP (main dry season), 12MAP (onset rainy season) and 15MAP (main rainy season); values represent means ± standard deviation of four replications (n = 4); Values are expressed in milligram (mg) gallic acid equivalent (GAE) per gram dry weight (gdw) of cassava leaves. Values followed by different lowercase letters in different variety in each column for the same month are significantly different (p < 0.05); values followed by different capital letters for each variety in each column for all month are significantly different (p < 0.05).

Total polyphenol contents (TPC) after in vitro gastrointestinal digestion influenced by variety

Figure 1 shows the TPC after in vitro gastrointestinal digestion of cassava leaves. The TPC values varied significantly between 39.42 and 298.32 mg/100g. These variations could be related to genetic characteristics of the varieties and seasons that may influence the concentration of bioaccessible polyphenols. However, bioaccessible polyphenols of the leaves of the variety IRAD4115 at 15 MAP is higher (Figure 1) as compared to the others. This result suggests that bioaccessibility could be influenced by phenol structure and solubilization (Williamson and Clifford, 2017; Blancas-Benitez et al., 2018). In addition, interactions of polyphenols with lipids, proteins and sugars could also affect the bioaccessibility (Karakaya, 2004; Jakobek, 2015). Whereas, methanolic TPC (Table 1) are higher than bioaccessible polyphenols: this can be explained by the presence of the insoluble fraction which is only excreted by microbial flora and fermentation in the large intestine for their high bioactivity (Zhao et al., 2018). However, there was a decrease of polyphenols after in vitro gastrointestinal digestion which was consistent with the results reported by Gunathilake et al. (2018) on six types of leaves.

Figure 1

Effect of variety on bioaccessible polyphenolic contents (digested samples) of cassava leaves. MAP: month after planting; cassava leaves harvested at 6MAP (onset dry season), 9MAP (main dry season), 12MAP (onset rainy season) and 15MAP (main rainy season); Values represent means ± standard deviation of four replications (n = 4); Values are expressed in milligram (mg) per 100 g dry weight (dw). GAE: gallic acid equivalent; Bars followed by different lowercase letters in different variety for the same month are significantly different (p < 0.05); bars followed by different capital letters for each variety for all month are significantly different (p < 0.05).

Effect of variety on hydroxybenzoic acids of bound and free polyphenols of dry leaf without digestion

The highest value (660.67 μg/g) of p-hydroxybenzoic acid of bound fraction was exhibited by variety EN at 9 MAP, whereas, the free fraction of variety TMS96/1414 showed the highest concentration of p-hydroxybenzoic acid (1094.9 μg/g) at the same harvesting maturity (Table 2A). These results indicate that the p-hydroxybenzoic acid accumulated during dry season when plants may develop various mechanisms to fight against stress. The results obtained are similar to that observed by Yang et al. (2018) who reported that variation of p-hydroxybenzoic acid contents in bound and free polyphenols from barley varieties. The bound phenolics of variety TMS92/0326 showed the highest concentration (16.62 μg/g) at 6 MAP, while the highest value of p-catechuic acid in free fraction was found in variety EN (10.62 μg/g) at 12 MAP. The results are in agreement with Yang et al. (2018) who found a significant variation of p-catechuic acid in free and bound polyphenols of barley varieties. The variety EN (1.10 μg/g) showed the highest value of syringic acid at 12 MAP in bound fraction, while, the highest concentration of syringic acid from free phenolics was gotten in the AD variety (1.20 μg/g) at the same harvest maturity. The onset of rainy season was found to induce high syringic acid in both bound and free polyphenols as reported by Yang et al. (2018) who obtained a significant effect of variety on syringic acid of bound and free polyphenols from barley. Gallic acid was highly abundant in variety TMS92/0326 with the highest value for bound (1.72 μg/g) at 6 MAP and free phenolics (58.50 μg/g) at 15 MAP (Table 2A). The bound phenolics of TMS96/1414 variety (0.92 μg/g) and the free polyphenols of AD variety (0.13 μg/g) showed the lowest amount of gallic acid at 12 MAP. These results suggest that gallic acid concentration varied depending on genetic differences of variety. Vanillic acid of bound and free phenolics varied significantly (p < 0.05), however, vanillic acid was not found in bound fractions of EN and AD varieties at 12 MAP. The bound phenolics of TMS92/0326 variety (1.27 μg/g) showed the highest concentration at 12 MAP and the free fraction of TMS92/0326 variety had the best concentration (585.91 μg/g) at 6 MAP. This result is in agreement with that reported by Suriano et al. (2018). The bound polyphenols of EN variety (0.21 μg/g) at 6 MAP and the free fraction at 12 MAP (20.25 μg/g) showed the highest concentrations of gentisic acid. The differences could be related to varietal differences mainly but the variation of environmental conditions also affect the phenolic accumulation. The bound polyphenols of AD variety (34.90 mg/g) and the free polyphenols of TMS96/1414 variety (265.93 mg/g) had the highest concentration of benzoic acid at 15 MAP, and benzoic acid was not detected in bound polyphenols of TMS92/0326, TMS96/1414, AD and EN varieties (Table 2B). This findings are in accordance with the results of Chen et al. (2014) who observed a great variation of benzoic acid in six ramie leaves. The bound fraction of AD variety (59.48 mg/g) had the highest concentration of salicylic acid at 6 MAP, while the IRAD4115 variety had the highest concentration (237.94 mg/g) at 15 MAP (Table 2B). These results suggest that salicylic acid was more accumulated during dry season and decreased during rainy season with respect to varietal differences. The results are similar with those reported by Suriano et al. (2018) on 20 barley genotypes.

Table 2

AEffect of cassava leaf variety of on hydroxybenzoic acids of bound and free phenolics fractions (methanolic extracts of dry leaf without digestion). Values are expressed in microgram per gram dry weight (μg/gdw) of leaves.

Harvest maturity	Var	p-Hydroxybenzoic acid (μg/gdw)		Protocatechiuc acid (μg/gdw)		Syringic acid (μg/gdw)		Gallic acid (μg/gdw)
		Bound fraction	Free fraction	Bound fraction	Free fraction	Bound fraction	Free fraction	Bound fraction	Free fraction
6MAP	92	6.77 ± 2.15cC	90.89 ± 9.06cbB	16.62 ± 0.65aA	3.47 ± 1.24aB	0.18 ± 0.04bB	0.60 ± 0.05cdBC	1.72 ± 0.016aA	4.25 ± 0.25dBCE
	96	15.94 ± 1.84aC	146.66 ± 3.67bCD	0.15 ± 0.12cBC	1.00 ± 0.59bD	0.25 ± 0.01aB	0.75 ± 0.07cC	0.5 ± 0.05cB	9.01 ± 0.66cC
	IRA	6.19 ± 1.09cB	47.48 ± 8.44cdD	0.14 ± 0.58cD	1.01 ± 0.67bC	0.07 ± 0.01cC	3.79 ± 0.77bC	0.13 ± 0.01bA	14.99 ± 1.30bD
	EN	4.99 ± 0.83cdCD	704.80 ± 89.69aA	0.23 ± 0.11bC	0.66 ± 0.39cD	0.24 ± 0.01aC	3.01 ± 0.10cdeC	0.24 ± 0.01dC	9.23 ± 0.38cBC
	AD	53.53 ± 7.37bB	27.82 ± 2.99deD	0.15 ± 0.79cB	0.55 ± 0.30cB	0.12 ± 0.00cC	5.65 ± 0.25aB	0.10 ± 0.00deC	16.74 ± 0.74aA
9MAP	92	121.88 ± 12.15bcA	52.28 ± 7.14eD	0.19 ± 0.22deD	7.76 ± 3.52aA	0.08 ± 0.01cC	0.37 ± 0.06deCD	0.2 ± 0.02deD	26.71 ± 0.96dCB
	96	180.55 ± 4.95bA	1094.99 ± 59.95aA	0.27 ± 0.17dB	5.44 ± 3.91cA	0.05 ± 0.00cdC	6.63 ± 0.82bA	0.03 ± 0.02dC	78.80 ± 1559cB
	IRA	42.50 ± 14.92dA	368.51 ± 7.27cA	0.54 ± 0.68cB	0.94 ± 0.42eCD	0.34 ± 0.12bB	9.50 ± 0.67aA	0.06 ± 0.00cC	181.62 ± 7.43aA
	EN	660.67 ± 91.29aA	502.15 ± 87.84bB	1.12 ± 0.89bA	7.00 ± 1.12bB	0.08 ± 0.02cD	0.38 ± 0.06dC	0.12 ± 0.06bA	13.01 ± 0.58dcA
	AD	31.88 ± 3.43deC	346.94 ± 24.37dcA	1.35 ± 1.00aA	2.00 ± 5.60dA	0.46 ± 0.03aA	4.36 ± 0.64cC	0.15 ± 0.03aA	140.51 ± 6.12bB
12MAP	92	10.12 ± 2.94cB	113.53 ± 5.81cdA	1.34 ± 0.53aB	2.99 ± 0.21dCD	0.60 ± 0.03bA	0.97 ± 0.14dB	0.14 ± 0.06aB	51.55 ± 1.65bB
	96	7.66 ± 0.93dcCD	169.22 ± 20.81abC	0.88 ± 0.11bC	2.68 ± 0.98cB	0.35 ± 0.07cA	0.36 ± 0.02deCD	0.92 ± 0.01deD	7.72 ± 0.97cdC
	IRA	3.51 ± 1.12deBC	130.62 ± 7.67cB	0.26 ± 0.15cC	4.08 ± 2.01bA	0.45 ± 0.04cA	6.15 ± 0.49cB	0.27 ± 0.02cD	96.99 ± 19.80aB
	EN	12.90 ± 2.66bcC	52.68 ± 13.15eC	0.61 ± 0.39bB	10.62 ± 6.85aA	1.10 ± 0.07aA	12.01 ± 0.99aA	0.58 ± 0.05bB	8.69 ± 0.28cCD
	AD	65.82 ± 3.05aA	185.83 ± 12.69aB	0.12 ± 0.06 dB	2.15 ± 0.80deBC	0.8 ± 0.09dC	7.99 ± 0.18bA	0.13 ± 0.00dCD	1.43 ± 0.06cdeD
15MAP	92	6.82 ± 0.64cdC	69.30 ± 6.09bC	0.86 ± 0.07bC	1.8 ± 0.23cdC	0.05 ± 0.02dCD	4.56 ± 0.50Abc	0.89 ± 0.07abcC	585.01 ± 71.59aB
	96	104.22 ± 14.47aB	930.11 ± 13.09aB	1.01 ± 1.24aA	2.14 ± 1.34bC	0.28 ± 0.02cB	5.17 ± 1.35bAB	0.12 ± 0.09aA	159.77 ± 7.05bA
	IRA	6.06 ± 1.73cdeB	86.23 ± 11.72bC	0.72 ± 0.18bA	3.83 ± 3.00aAB	0.05 ± 0.02dCD	1.31 ± 0.25dD	0.75 ± 0.02cdB	41.46 ± 1.45cC
	EN	522.05 ± 66.24bB	61.96 ± 8.28bC	0.11 ± 0.08cD	0.78 ± 0.57cC	0.49 ± 0.09aB	8.72 ± 1.09aB	0.11 ± 0.013eB	10.31 ± 1.34cdB
	AD	15.92 ± 1.00cD	126.06 ± 3.92bcC	1.00 ± 0.02cC	1.61 ± 0.16eBC	0.39 ± 0.00bAB	1.20 ± 0.11dD	0.08 ± 0.01cB	7.54 ± 0.97cdeC

Var: varieties; MAP: month after planting; 92: TMS92/0326; 96: TMS96/1414; IRA: IRAD4115; cassava leaves varieties harvested at 6MAP (onset of dry season), 9MAP (main dry season), 12 (onset of rainy season) and 15MAP (main rainy season); values represent means ± standard deviation (n = 3). Values followed by different lowercase letters in different variety in each column for the same month are significantly different (p < 0.05); values followed by different capital letters for each variety in each column for all month are significantly different (p < 0.05).

Table 2B

Effect of cassava leaf variety on hydroxybenzoic acids of bound and free phenolics of undigested sample (methanolic extract of dry leaf without digestion) (next). Values are expressed in microgram per gram dry weight (μg/gdw), or milligram per gram dry weight (mg/gdw) of leaves.

Harvest maturity	Var	Vanillic acid (μg/gdw)		Gentisic acid (μg/gdw)		Benzoic acid (mg/gdw)		Salicylic acid (mg/gdw)
		Bound fraction	Free fraction	Bound fraction	Free fraction	Bound fraction	Free fraction	Bound fraction	Free fraction
6MAP	92	0.04 ± 0.00eCD	48.27 ± 1.30cdB	0.06 ± 0.03cB	0.83 ± 0.34cC	10.48 ± 0.88aA	3.53 ± 0.64bCD	12.79 ± 2.17dA	56.79 ± 1.59aA
	96	0.37 ± 0.02bC	585.91 ± 32.61aA	0.04 ± 0.01cA	1.12 ± 0.17cdCD	7.46 ± 0.68bA	3.69 ± 0.64bC	23.55 ± 5.8cA	69.64 ± 5.52bB
	IRA	0.10 ± 0.00dC	65.90 ± 1.73bcC	0.03 ± 0.00dB	7.01 ± 1.03bC	0.35 ± 0.05dC	89.07 ± 1.88aB	3.87 ± 0.89eB	54.16 ± 3.74bcD
	EN	0.75 ± 0.05aA	9.20 ± 0.33eB	0.21 ± 0.02aA	0.62 ± 0.26cC	4.09 ± 0.39cB	4.33 ± 0.12bD	47.34 ± 6.94bA	55.46 ± 2.83dD
	AD	0.22 ± 0.02cC	84.87 ± 2.10bB	0.17 ± 0.02bA	11.58 ± 3.51aA	0.21 ± 0.05deC	3.83 ± 0.18bA	59.48 ± 0.25aA	37.89 ± 1.46eD
9MAP	92	0.98 ± 0.02aB	21.34 ± 0.68dC	0.12 ± 0.03aA	1.26 ± 0.28dC	0.63 ± 0.29dC	12.44 ± 1.20cdB	4.55 ± 0.33cC	51.94 ± 6.30aB
	96	0.56 ± 0.03cB	89.83 ± 3.70bB	0.05 ± 0.01bA	9.84 ± 1.32bA	0.69 ± 0.08dC	77.78 ± 6.25bB	8.57 ± 0.12eBCE	24.10 ± 8.47C
	IRA	0.70 ± 0.02bB	137.48 ± 3.69aA	0.06 ± 0.02bB	13.19 ± 0.97aA	9.22 ± 0.25aA	120.69 ± 6.82aA	6.50 ± 0.42bA	33.39 ± 1.5dcB
	EN	0.38 ± 0.03cB	5.64 ± 0.34eC	0.06 ± 0.03bC	0.85 ± 0.27deC	3.53 ± 0.21cA	14.15 ± 0.84cC	7.52 ± 0.81aB	18.21 ± 1.41bB
	AD	0.72 ± 0.01bA	64.89 ± 3.00cC	0.08 ± 0.02bcB	7.06 ± 0.81cB	3.83 ± 0.52bB	3.90 ± 0.11eC	2.55 ± 0.34dB	23.78 ± 4.40bB
12MAP	92	1.27 ± 0.03aA	15.74 ± 0.29cD	0.11 ± 0.0'bcAB	2.55 ± 0.34dcB	4.47 ± 0.20bB	5.44 ± 0.60dC	0.40 ± 0.05eD	22.13 ± 8.98dD
	96	0.60 ± 0.02cA	12.71 ± 0.18dC	0.06 ± 0.04bdA	0.50 ± 0.07deC	1.92 ± 0.25cB	1.99 ± 0.27eCD	1.09 ± 0.11dBC	93.37 ± 3.80aA
	IRA	0.73 ± 0.02bA	88.86 ± 1.53bB	0.09 ± 0.03bA	11.37 ± 5.32bAB	1.74 ± 0.18cdB	82.55 ± 2.81aBC	3.69 ± 0.22aB	35.62 ± 8.98bA
	EN	ND	9.51 ± 0.27cB	0.17 ± 0.02aAB	20.25 ± 1.71aA	11.01 ± 0.25eA	55.13 ± 1.94bB	2.98 ± 0.17bBCE	41.82 ± 4.51bB
	AD	ND	120.99 ± 1.70aA	0.02 ± 0.01dC	7.16 ± 1.60bcB	8.21 ± 0.55aA	8.68 ± 0.83cB	1.19 ± 0.06cC	14.98 ± 1.50deD
15MAP	92	0.07 ± 0.00dC	72.04 ± 2.52cA	0.09 ± 0.02dcAB	7.72 ± 0.73bA	ND	101.95 ± 3.31bA	9.21 ± 0.43aB	28.37 ± 0.48cbC
	96	0.38 ± 0.02bC	101.14 ± 1.13bB	0.05 ± 0.03aA	8.37 ± 1.47bB	ND	265.93 ± 7.53aA	1.32 ± 0.08dB	46.28 ± 5.37bB
	IRA	0.05 ± 0.00dD	24.06 ± 0.34dD	0.04 ± 0.00abB	2.04 ± 0.38cCD	ND	17.92 ± 2.19dD	3.88 ± 0.26bcB	237.94 ± 19.92aB
	EN	0.14 ± 0.02cC	174.87 ± 2.37aA	0.08 ± 0.034aC	14.15 ± 1.19aB	ND	87.77 ± 0.45cA	4.23 ± 0.21bBC	37.39 ± 14.77cbBC
	AD	0.56 ± 0.02aB	7.77 ± 1.13eD	0.06 ± 0.02aB	1.83 ± 0.32dC	34.90 ± 0.57aB	11.49 ± 1.28deA	0.89 ± 0.04deCD	5.34 ± 7.57bA

ND: not detected; Var: varieties; MAP: month after planting; 92: TMS92/0326; 96: TMS96/1414; IRA: IRAD4115; cassava leaves varieties harvested at 6MAP (onset of dry season), 9MAP (main dry season), 12 (onset of rainy season) and 15MAP (main rainy season); values represent means ± standard deviation (n = 3); Values followed by different lowercase letters in different variety in each column for the same month are significantly different (p < 0.05); values followed by different capital letters for each variety in each column for all month are significantly different (p < 0.05).

Stability of hydroxybenzoic acids after in vitro gastrointestinal digestion (digested sample) impacted by the varieties

Table 3 shows a significant (p < 0.05) variation of bioaccessible hydroxybenzoic acids in leaves. The leaves of EN variety harvested at 9 MAP showed the highest value of biaoccessible gallic acid and salicylic acid, the leaves of the same variety and the same harvest maturity are more significantly bioaccessible (2798.49 μg/g). Similarly, the same leaves showed the highest value (6429 μg/g) of syringic acid (Table 3). These results indicate that varietal factors and harvest maturity influence the bioaccessibilty of hydroxybenzoic acids. The values of benzoic acid also varied significantly between 346.98 and 8142.42 μg/g for the leaves at 9 MAP and 6 MAP, respectively, whereas, for p-Hydroxybenzoic acid, the low bioaccessible content (9.63 μg/g) was observed at 9 MAP in leaves of TMS96/1414 variety and the significantly highest value (230.86 μg/g) was found in the leaves of the AD variety at 15 MAP. The highest bioaccessible gentisic acid was found in EN variety at 6 MAP as well as the highest value of 91.07 μg/g in bioaccessible vanillic acid (Table 3). These variations of the contents of bioaccessible hydroxybenzoic acids are attributed to the stability and instability of phenolics during the transition of gastrointestinal digestion and the affinity of digestive enzymes with phenolics (Gonzales et al., 2015; De Santiago et al., 2018), but also the initial concentration of these acids which have been reported by Gunathilake et al. (2018) on the bioaccessibility of phenolics of six leaves.

Table 3

Effect of cassava leaf on bioaccessibility of polyphenols of digested sample. Values are expressed in microgram per gram dry weight (μg/gdw) of leaves.

Harvest maturity	Variety	Gallic	Salicylic	Syringic	Benzoic	P-catechuic	p-Hydroxybenzoic	Gentisic	Vanillic
6MAP	92	7.01 ± 0.29dC	118.42 ± 0.43cD	295.89 ± 0.66eB	2805.34 ± 2.14b	352.22 ± 0.94cA	33.08 ± 0.54dC	4.46 ± 0.17cB	7.45 ± 0.14dC
	96	39.41 ± 0.41bA	115.36 ± 0.66cD	503.56 ± 0.74dA	1661.62 ± 1.54c	625.29 ± 2.17bA	25.04 ± 0.68eC	2.49 ± 0.29d	17.83 ± 0.17cA
	IRA	12.59 ± 0.64cC	132.73 ± 0.30bC	609.61 ± 0.24cA	737.68 ± 2.96d	185.92 ± 0.61eA	183.71 ± 0.89bA	0.88 ± 0.04eC	21.68 ± 0.22bA
	EN	48.14 ± 1.30aB	37.42 ± 0.14dD	2611.79 ± 0.81aB	8142.41 ± 3.14a	5021.89 ± 0.76aA	53.29 ± 1.18cD	66.58 ± 1.40aA	91.07 ± 0.32aA
	AD	11.83 ± 0.24cC	148.05 ± 0.47aC	648.57 ± 3.04bC	693.93 ± 0.57e	218.44 ± 0.62dB	188.98 ± 0.91aA	6.90 ± 0.21bA	21.64 ± 0.27bB
9MAP	92	13.80 ± 0.27dA	1265.02 ± 1.01cA	346.48 ± 0.37cA	591.13 ± 1.39b	130.98 ± 1.34bC	119.52 ± 0.95cB	3.44 ± 0.23aC	5.30 ± 0.12cC
	96	23.49 ± 0.39cB	1558.22 ± 2.78bA	278.04 ± 0.47dB	458.73 ± 1.67c	91.93 ± 0.59dD	9.63 ± 0.16eD	1.67 ± 0.07dB	5.49 ± 0.20cC
	IRA	30.94 ± 0.86bA	493.42 ± 0.32eA	219.13 ± 0.28eA	348.24 ± 0.60d	66.05 ± 0.62eC	11.81 ± 0.11dB	2.94 ± 0.02cC	4.78 ± 0.29eC
	EN	50.05 ± 1.27aA	2798.49 ± 0.20aA	6429.11 ± 0.43aA	718.73 ± 3.64a	141.45 ± 0.89aD	141.59 ± 0.61bB	1.43 ± 0.13eC	16.19 ± 0.22aC
	AD	9.67 ± 0.14eD	510.73 ± 0.19dA	783.63 ± 0.33bB	346.98 ± 2.06d	114.99 ± 0.53cD	165.04 ± 0.35aB	3.34 ± 0.04bB	8.64 ± 0.11bC
12MAP	92	11.20 ± 0.25dA	203.46 ± 0.61bB	163.68 ± 0.60dD	615.27 ± 2.18c	138.01 ± 1.06bC	116.47 ± 0.32bB	2.82 ± 0.02dC	10.47 ± 0.18bB
	96	30.12 ± 1.72bA	201.46 ± 0.15bB	188.30 ± 0.59bC	833.13 ± 2.22a	212.26 ± 0.21cB	95.21 ± 0.25dA	18.88 ± 0.26aA	6.88 ± 0.11dC
	IRA	19.49 ± 0.42cB	217.89 ± 0.39aB	146.05 ± 0.35eB	526.36 ± 2.02e	111.18 ± 1.07dB	18.89 ± 0.12eB	8.04 ± 0.61bB	4.12 ± 0.07eC
	EN	32.05 ± 2.61bD	167.70 ± 0.42dB	207.74 ± 0.30aD	728.97 ± 1.63b	211.32 ± 1.85cC	124.38 ± 0.45aC	2.81 ± 0.11dC	13.17 ± 0.18aC
	AD	49.66 ± 1.54aA	192.16 ± 0.51cD	178.82 ± 0.72cD	600.89 ± 3.83d	150.14 ± 0.71aC	100.92 ± 0.47cC	4.14 ± 0.12cA	8.21 ± 0.19cC+
15MAP	92	9.62 ± 0.24eB	193.11 ± 0.23bC	213.05 ± 0.68cC	1452.07 ± 1.62a	219.59 ± 0.79cB	165.13 ± 0.58bA	55.64 ± 1.39aA	16.21 ± 0.19cA
	96	11.49 ± 0.12dC	175.51 ± 0.45cC	167.53 ± 0.25dD	921.62 ± 1.59b	162.93 ± 0.67eC	81.05 ± 0.45cB	3.41 ± 0.15dB	11.50 ± 0.13dB
	IRA	19.39 ± 0.42bB	97.07 ± 0.18dD	139.13 ± 0.45eC	798.53 ± 1.50d	186.09 ± 0.61dA	12.61 ± 0.11dB	4.67 ± 0.21cC	6.67 ± 0.19eB
	EN	41.17 ± 1.60aC	50.58 ± 0.36eC	861.39 ± 0.84bC	801.83 ± 0.57c	614.50 ± 0.65aB	230.86 ± 0.91aA	42.32 ± 0.59bB	45.63 ± 0.28bB
	AD	14.76 ± 0.44cB	349.36 ± 0.09aB	2670.83 ± 0.43aA	421.81 ± 1.14e	595.37 ± 0.88bA	10.24 ± 0.01eD	2.18 ± 0.04eB	72.27 ± 0.40aA

MAP: month after planting; 92: TMS92/0326; 96: TMS96/1414; IRA: IRAD4115; Cassava leaves harvested at 6MAP (onset dry season), 9MAP (main dry season), 12MAP (onset rainy season) and 15MAP (main rainy season); values represent means ± standard deviation (n = 3); gdw: gram dry weight. Values followed by different lowercase letters in different variety in each column for the same month are significantly different (p < 0.05); values followed by different capital letters for each variety in each column for all month are significantly different (p < 0.05).

Effect of variety on antioxidant activities of undigested sample (methanolic extract of dry leaf without digestion)

FRAP of bound and free polyphenols varied significantly (p < 0.05) across the harvest maturity in different varieties (Table 4). The highest antioxidant activity throughout the harvest was shown by the IRAD4115variety at 12 MAP (35.17 μgTE/g in bound fraction and 79.17 μgTE/g in free fraction). This variation may be linked to the genetic factors of varieties, and the result is in accordance with Quartey et al. (2016) who observed variation of the total antioxidant among cassava leaves varieties, however, they did not report the bound and free polyphenols and effect of harvest maturity in their study.

Table 4

Effect of cassava leaf variety on FRAP and DPPH activity of bound and free fractions (undigested sample or leaves without digestion) and bioaccessible (digested sample) polyphenols. Values are expressed in microgram trolox equivalent per gram dry weight (μgTE/gdw) of leaves.

Harvest maturity	Varieties	FRAP (μgTE/gdw)			DPPH (μgTE/gdw)
		Bound fraction	Free fraction	Bioaccessible	Bound fraction	Free fraction	Bioaccessible
6MAP	TMS92/0326	22.39 ± 0.77cA	36.27 ± 0.38cD	50.04 ± 0.13aD	114.47 ± 0.52bB	176.67 ± 2.8 3aA	9.87 ± 0.84bB
	TMS96/1414	15.09 ± 0.30dC	36.32 ± 0.65bD	30.51 ± 0.22dD	74.86 ± 1.51cB	20.30 ± 0.12dCD	5.99 ± 0.15cB
	IRAD4115	33.48 ± 0.34aBC	34.12 ± 1.68dD	41.72 ± 1.70bB	47.14 ± 1.33dA	9.99 ± 0.21eCD	10.79 ± 1.44bC
	EN	24.76 ± 0.79bB	51.11 ± 0.43aB	41.74 ± 2.32bA	147.25 ± 2.25aA	86.15 ± 0.95bC	11.49 ± 0.95aA
	AD	21.59 ± 0.45cC	31.41 ± 0.61eD	35.09 ± 1.49cB	14.16 ± 0.56eD	61.63 ± 1.61cB	9.18 ± 0.39bB
9MAP	TMS92/0326	19.33 ± 0.35bB	71.60 ± 0.17aA	64.64 ± 2.09bB	117.69 ± 1.78aB	131.19 ± 1.03bB	19.47 ± 3.38aA
	TMS96/1414	19.45 ± 0.12bB	69.76 ± 0.76cA	68.56 ± 2.90aA	21.34 ± 0.71dC	132.46 ± 0.44bA	3.88 ± 1.26cC
	IRAD4115	18.68 ± 0.18cD	72.78 ± 0.73aB	39.24 ± 2.07cC	14.19 ± 0.38eC	70.69 ± 0.63cB	4.40 ± 0.60cD
	EN	34.68 ± 0.26aA	60.25 ± 0.35cA	36.77 ± 1.62cB	64.70 ± 1.94bC	169.59 ± 2.96aA	7.83 ± 0.85bC
	AD	18.84 ± 0.13dD	70.00 ± 0.98bA	38.36 ± 1.31cA	35.31 ± 1.86cB	70.79 ± 1.64cA	1.61 ± 0.82dD
12MAP	TMS92/0326	18.28 ± 0.44cD	41.73 ± 0.89dC	76.85 ± 2.19aA	15.33 ± 0.41dC	172.58 ± 2.31aA	8.77 ± 2.52cC
	TMS96/1414	32.93 ± 0.85bA	47.83 ± 0.17bB	51.89 ± 0.50bB	78.07 ± 2.05aA	20.92 ± 0.47dC	9.87 ± 0.67cA
	IRAD4115	35.17 ± 0.76aA	79.17 ± 1.05aA	77.71 ± 1.67aA	42.92 ± 2.11bAB	10.54 ± 0.69eC	15.40 ± 3.14aB
	EN	14.41 ± 0.366dD	35.89 ± 0.08eD	34.95 ± 0.89cC	6.59 ± 0.29eD	69.41 ± 0.76bD	11.53 ± 0.58abA
	AD	32.91 ± 0.84bA	44.39 ± 0.46cC	32.54 ± 1.53cB	20.83 ± 0.40cC	54.21 ± 1.42cC	14.34 ± 1.85bA
15MAP	TMS92/0326	19.25 ± 0.10dCB	70.97 ± 0.51aAB	58.09 ± 1.00aC	154.56 ± 0.11aA	84.40 ± 0.25bC	20.91 ± 2.31aA
	TMS96/1414	14.52 ± 0.13eCE	45.85 ± 0.44dC	41.92 ± 2.07bC	16.73 ± 0.82dD	24.23 ± 0.33dB	5.98 ± 2.01cB
	IRAD4115	33.61 ± 0.63aB	65.42 ± 1.19bC	39.24 ± 2.07bcC	9.69 ± 0.52eD	90.24 ± 0.25aA	17.32 ± 2.15aA
	EN	24.20 ± 0.13cBC	39.55 ± 0.41eC	36.77 ± 1.62cB	69.16 ± 0.76bB	90.28 ± 0.22aB	9.21 ± 1.39bB
	AD	29.45 ± 0.33bB	48.21 ± 0.18cB	38.37 ± 1.41cA	51. 13 ± 1.34cA	45.22 ± 1.26cD	8. 76 ± 2.87bC

FRAP: ferric reducing antioxidant power; DPPH: 2.2-diphenyl-1-picyhydrazyl; g dw: gram dry weight; μgTE: microgram trolox equivalent; MAP: month after planting; Cassava leaves harvested at 6MAP (onset dry season). 9MAP (main dry season). 12MAP (onset rainy season) and 15MAP (main rainy season); values represent means ± standard deviation of four replications (n = 4); Values followed by different lowercase letters in different variety in each column for the same month are significantly different (p < 0.05); values followed by different capital letters for each variety in each column for all month are significantly different (p < 0.05).

The DPPH radical scavenging activity of bound polyphenols varied significantly during various harvest maturity according to the variety (Table 4). However, it was found that the bound polyphenols of EN variety (6.59 μgTE/g) had the best antiradical activity at 12 MAP, whereas free polyphenols of IRAD4115 variety at 6 MAP (9.99 μgTE/g) was found to possess the highest antioxidant activity as compared to the others. The variation in bound and free polyphenol contents (Table 1) during various harvest maturity may be responsible for these variations of antiradical activity, and these are in agreement with the results observed by Barkat et al. (2018) who reported the significant effect of harvesting time on DPPH radical scavenging activity in spinach. Furthermore, the variation could be due to the influence of plant maturity on antioxidant activity reported by Simao et al. (2013) for cassava varieties.

Stability of antioxidant activities of polyphenols after in vitro gastrointestinal digestion (digested sample) affected by the harvest maturity

The bioaccessibile polyphenols of the IRAD4115 variety at 12 MAP showed the significantly highest FRAP value (77.71 μgTE/g), while, the TMS96/1414 variety harvested at 6 MAP had the lowest value (30.51 μgTE/g) of FRAP (Table 4). These observations are similar to those reported by Bouayed et al. (2011). The results may be linked to the change in pH which could alter the structure of polyphenols, thus affecting antioxidant activity (Arenas and Trinidad, 2017). There was a slight decrease in antioxidant activity after in vitro gastrointestinal digestion of polyphenols that was consistent with observations reported by Taglazucchi et al. (2010).

Cassava leaves of AD variety at 9 MAP (1.61 μgTE/g) had the strongest radical scavenging activity after in vitro gastrointestinal digestion (Table 4). The results may suggest the presence of multitudes of phenolics that act synergistically after in vitro gastrointestinal digestion to exhibit a strong antiradical activity, however, the variation in antioxidant activity could be linked to the maturity of the plant, whose sample matrix could influence the bioaccessibility of polyphenols with structures having a strong affinity for free radicals (Simao et al., 2013). The antiradical activity evaluated could be related only to the bioaccessible polyphenols, therefore the insoluble fraction may be more bioactive after fermentative action of the microbial flora (Williamson and Clifford, 2017). This suggests that the present results are very important because bioaccessible polyphenols showed the strongest radical scavenging activity as compared to those of methanolic polyphenols from the native cassava leaf samples. These observations have also been reported by several authors (Bouayed et al., 2011; Gunathilake et al., 2018) in their studies on six types of leaves.

Conclusion

This finding clearly shows that variety and harvesting maturity significantly affect the polyphenol contents and antioxidant activity of cassava leaves before and after in vitro gastrointestinal digestion. The antioxidant activity of bound polyphenols showed significantly higher activity than the free fraction, however, there were some discrepancies throughout the harvest maturity among the varieties. The antioxidant activity was increased significantly after in vitro gastrointestinal digestion with some variability. The bioaccessible phenolics are stable and showed the highest activity compared to the methanolic extract of leaves with unexpected variability among varieties. These results suggest that cassava leaves can be considered as a good reliable source of natural polyphenols and bioaccessible antioxidant compounds throughout the harvest maturity that could be useful as functional food ingredients.

Declarations

Author contribution statement

Alphonse Laya: Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Benoît B. Koubala: Conceived and designed the experiments.

Funding statement

Alphonse Laya was supported by a CSIR-TWAS fellowship.

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.