Antioxidant activities and inhibitory effects of free and bound polyphenols from date ( Phoenix dactylifera L . ) seeds on starch digestive enzymes

Date (Phoenix dactylifera L.) seeds have gained interest as a valuable by-product of the date fruit industry and have been identified as a rich source of functional and bioactive ingredients. In this study, date seeds from five varieties (Medjool, Deglet Nour, Barhee, Bou Sthammi and Dayrie) cultivated in Australia were analysed for their total phenolic and flavonoid contents, ferric reducing antioxidant power (FRAP), Trolox equivalent antioxidant capacity (TEAC), and in-vitro inhibition against α-amylase and α-glucosidase. The date seed powders (DSP) were extracted with acetone-methanol-water (2:2:1, v:v:v) to obtain free polyphenols. The bound polyphenols were extracted from the residual solids with butanolHCl (97.5:2.5; v/v) treatment at 100 °C. The greatest quantities of total phenols (4166±227 mg Gallic Acid Equivalents 100g−1 DSP), total flavonoids (52.1±9.6 mg Quercetin Equivalents 100g−1 DSP), FRAP (1589±47 μmol Iron (II) Equivalents g−1 DSP) and TEAC (45.2±1.4 μmol Trolox Equivalents g−1 DSP) were detected in seeds from Deglet Nour variety. Both free and bound polyphenol fractions showed significant (p<0.05) inhibition against rat intestinal α-glucosidase with little or no inhibition against pancreatic α-amylase. Bound polyphenols showed stronger α-glucosidase inhibition compared with free polyphenols. The IC50 for α-glucosidase were reported as mg GAE mL−1 ranged from 0.39±0.02 (Deglet Nour) to 0.68±0.02 (Medjool) for bound polyphenols, and from 0.907±0.08 (Medjool) to 1.75±0.15 (Barhee-khalal) for free polyphenols. Bound polyphenol fraction, with strong α-glucosidase inhibition and weak α-amylase inhibition, was suggested as a significant source of functional food ingredients with anti-hyperglycaemic properties.


Introduction
The Date crop has been identified as a suitable crop for the arid regions of Australia, with a very good marketing potential for both date fruit industry and its by-products (Reilly, Reilly, & Lewis, 2010) The total world production of date fruits was 7.5 million tons in 2012 (FAO-STAT, 2015), and date seeds (approximately 10% of fruit weight) are an abundant by-product from this well-established industry (Besbes et al., 2004).Date fruit processing plants produce large quantities of date seeds as a result of value addition processing operations which involve pitting (deseeding) dates to manufacture dates filled with edible seeds, date paste and syrup.It has been suggested that utilizing this by-product in profitable ways will benefit the date industry (Al-Farsi & Lee, 2008a).
It has been reported that date seeds contain large amounts of polyphenols (Al-Farsi & Lee, 2008b).Plant polyphenols are considered to be the key components responsible for many of the health benefits attributed to the consumption of fruits and seeds (Macedo et al., 2013;Mahbub et al., 2012;Gondoin, Grussu, Stewart, & Mc-Dougall, 2010).Health benefits of consuming date seed products have been known in middleeastern folk medicine.Date seed infusions have been consumed in Turkey and Arab countries as herbal coffees with memory enhancing properties (Habib & Ibrahim, 2009).A recent in-vitro study has indicated that date seed extracts can inhibit enzymes related to neurodegeneration (Sekeroglu et al., 2012).In-vivo studies on rats have shown that date seed extract could reduce DNA binding and methylation with N-Nitroso-N-methyl urea, which is a well known carcinogen (Diab & Aboul-Ela, 2012).Anti-hyperglycaemic agents such as acarbose, which act by inhibiting starch digestion enzymes, such as α-amylase and α-glucosidase, are widely used in combination with drug treatment for Type 2 diabetes mellitus (Tundis, Loizzo, & Menichini, 2010).Inhibition of α-amylase and/or intestinal α-glucosidase delays the rise in blood glucose levels, thus alleviating the complications arising from postprandial hyperglycaemia for diabetics.However, the generally prescribed antihyperglycaemic agents are strong α-amylase inhibitors and are known to exert unpleasant side effects (flatulence, bloating, abdominal distention) due to suppression of starch digestion in the small intestine and bacterial fermentation in the colon (Yee & Fong, 1996).Consequently, searching for more specific polyphenolic compounds from plant sources that could inhibit α-amylase and α-glucosidase might provide a solution to these side effects, and be cheaper and more tolerable alternatives.Plant polyphenols (free and bound) is expected to delay postprandial hyperglycaemia by specifically inhibiting intestinal α-glucosidase (Phan, Wang, Tang, Lee, & Ng, 2013), and provide additional benefits against oxidative diseases such as diabetics due to their antioxidant properties (Maritim, Sanders, & Watkins, 2003).
Recent studies have demonstrated that the health benefits attributed to the plant polyphenols should include both free (also referred to as extractable) polyphenols (FPP) and bound (also referred to as non-extractable) polyphenols (BPP) (Saura-Calixto, 2012;Tow, Premier, Jing, & Ajlouni, 2011).Most of the reported polyphenol contents considered the fraction extracted into aqueous-organic solvent system, whereas the residual solids are not analysed.However these residual solids have been found to contain a large amount of polyphenols including polymeric proanthocyanidins (PC).Some of the PCs are strongly bound to cell wall components (Hamauzu & Mizuno, 2011), and Perez-Jimenez, Elena Diaz-Rubio, and Saura-Calixto (2013) indicated that BPPs contribute 60-90% to the total polyphenol content in most fruits.These studies have emphasised the importance of including the bound (non-extractable) polyphenols in plant polyphenol analysis.
The aim of this study was to determine the effect of date seeds' free and bound polyphenols IJFS October 2016 Volume 5 pages 212-223 on the activity of starch the digesting enzymes, α-amylase and α-glucosidase.The phenolic contents and antioxidant capacities of seeds from those Australian grown date varieties were also reported.
2 Materials and Methods .Seeds were separated manually from the fruits (Figure 1) washed, then left to air dry for 5 days.Seeds were then ground using a cutting mill (Retsch SM100, Germany) to pass a 1mm sieve.

Preparation of bound polyphenols (BPP) extract
The generally used method for depolymerisation of the proanthcyanidins is HCl-butanol reaction with Fe 3+ catalyst (Schofield, Mbugua, & Pell, 2001).Following this reaction, the presence of Fe 2+ /Fe 3+ ions in the final BPP extract was inevitable, thus rendering extract not suitable for subsequent analyses.Several trials showed that Fe 2+ /Fe 3+ ions interfered with total flavonoid assay and antioxidant activity measurements using FRAP and TEAC assays.Therefore the depolymerisation was performed without the ferric chloride catalyst.Optimization of the extraction was monitored via measurements of TEAC and TPC, and photometric scans showed absorption maxima at λ 550 nm characteristic to anthocyanidins which were significantly enhanced after 3 h of extraction (data not shown).
The residues from the FPP preparation were used for BPP extraction following the method of Perez-Jimenez and Saura-Calixto ( 2005), with modifications.From the FPP residue, 2 g was mixed with 25 ml butanol/HCl (97.5:2.5 v/v) and heated at 100 °C for 3 hours in a water bath.
The mixture was centrifuged at 10,000 rpm for 10 minutes, supernatant was collected and residue was subjected to two washings with 10 mL butanol.Combined supernatants were rotary evaporated at 60 °C, and pH of the concentrated extract was adjusted to 4 using 1 M sodium hydroxide.pH above 5 caused precipitation of the extract, therefore pH 4 was selected to maintain both an intact extract and the compatibility with buffers.Extract was freeze-dried and stored at -20 °C until analysis.

Total polyphenol content (TPC)
Assay for total phenol content was based on reduction of the molybdenum-tungstate complex in F-C Reagent by phenols (Herald, Gadgil, & Tilley, 2012).The procedure was adapted to 96-well microplate platform (250 µL assay volume per well) with absorbance acquired at λ 765 nm using a microplate reader (Multiskan GO, Thermo Scientific, Australia) at 25 °C.Gallic acid (0-60 µg) was used to generate the standard plot, and TPC was expressed as mg Gallic Acid Equivalent (GAE) 100 −1 g date seed powder (DSP).

Total flavonoid content (TFC)
Assay for total flavonoid content was based on chelation with Al 3+ ion and absorbance measurement of the flavonoid-Al 3+ complex as described by Xu, Wang, Hu, Lee, and Wang (2010), and was adopted to the microplate platform.Absorbance was acquired at λ 430 nm instead of 415 nm based on our optimisation study, which showed that quercetin-aluminium complex exhibit maximum absorbance at 430 nm (data not shown).Quercetin (0-30 µg) was used to generate the standard plot and TFC was expressed as mg Quercetin Equivalents (QE) 100 −1 g DSP.

Trolox equivalent antioxidant activity (TEAC)
Free radical scavenging antioxidant activity was measured as TEAC using ABTS radical as the organic radical (Re et al., 1999).The assay was adapted to 96 well microplate platform and absorbance was measured at λ 734 nm after 10 min at room temperature (RT).ABTS radical scavenging standard curve was obtained using Trolox (0 -400 µmol) and the TEAC expressed as µmole Trolox Equivalents g −1 DSP.

Ferric reducing antioxidant power (FRAP)
Reducing antioxidant activity was measured as FRAP based on the reduction of Fe 3+ -TPTZ complex (Benzie & Strain, 1996).The absorbance was measured at λ 593 nm at a fixed incubation time of 10 min at RT. Reducing standard curve was obtained using Fe 2+ -TPTZ (0-500 µmol Fe 2+ ), and the FRAP was expressed as µmol Fe 2+ equivalents g −1 DSP.
IJFS October 2016 Volume 5 pages 212-223 2.9 α-Glucosidase assay α-Glucosidase extract from rat intestine was prepared using rat intestinal acetone powder and activity assayed following the method of Kee, Koh, Oong, and Ng (2013).The protein content of the enzyme preparation was determined to be 12.65±0.81mg mL −1 using the Lowry protein determination method.Date seed freeze dried extract was reconstituted in DMSO and 10-75 µL was mixed with 700 µL of buffer (0.1 M potassium phosphate, pH 6.8 + 1 % (w:w) NaCl) and 100 µL of α-Glucosidase (in the same buffer), and pre-incubated for 5 minutes at RT.A control (100% enzyme activity) was prepared the same way with DMSO replacing the sample extract.
Enzyme activity was compatible with 10% (v:v) DMSO with no inhibition of activity observed (data not shown).A colour correction solution was prepared for each extract with buffer replacing the enzyme to correct for background colour of extract.100 µmol L −1 of p-nitrophenyl β-D-glucopyranoside (PNPG) substrate was added to initiate the reaction, and the mixture was incubated at 37 °C for 1 h.Reaction was terminated by chilling microplate on ice and the absorbance promptly measured at λ 410 nm with spectrophotometer (Novaspec II, USA).The % inhibition of α-glucosidase activity was calculated using following equation: Where I is the % Inhibition, Abs ctrl is the Absorbance of control incubation corrected for enzyme color and Abs extract is the Absorbance of extract incubation corrected for extract and enzyme color 2.10 α-Amylase assay I (starch-DNS method) α-Amylase activity was assayed by the starch-DNS method according to Wan, Yuan, Cirello, and Seeram (2012) with some modifications.Porcine pancreas α-amylase was diluted to 1 unit mL −1 in 20 mM sodium phosphate buffer containing 7 mmol L −1 NaCl and 1 mmol L −1 CaCl 2 .Soluble potato starch (0.5%; w:w) was prepared in the same buffer.50 µL Sample (in DMSO), 100 µL α-amylase and 400 µL buffer were mixed in 10 mL screw capped plastic tube and incubated for 5 minutes at RT. Enzyme activity was compatible with 10% (v:v) DMSO and no inhibition of activity was observed (data not shown).500 µL 0.5 % starch solution (w:w) was added and mixture was incubated at 37 °C for 20 min.then 500 µL dinitrosalicylic acid (DNS) was added.Tube was placed in boiling water bath for 15 minutes.It was then cooled and absorbance was measured at λ 540 nm.A colour correction solution was prepared for each extract.Percentage inhibition of the α-amylase activity was calculated using equation 1.
Porcine pancreas α-amylase was diluted to 50 units mL −1 in 20 mM sodium phosphate buffer (pH 6.8), containing 7 mM NaCl and 1 mM CaCl 2 .PNPG5 was made up to 25 mM in the same buffer.The assay was performed in 96 well microplates.50 µL of PNPG5, 10 µL of extract in DMSO, and 190 µL buffer were added to each well, and 50 µL of enzyme was added to initiate the reaction.Absorbance was measured at λ 405 nm using microplate reader at 3 minute intervals for 90 minutes.Absorbance of control (without inhibitor) and for extract background (without added enzyme) was also acquired.The percentage inhibition of amylase activity was calculated by the ratio of the linear gradients of control and extract incubation, using the linear range of the kinetic curve.

IC 50 determination
Active extracts against rat intestinal αglucosidase were assessed for their potency by their IC 50 values.The data points (mean ± SD, n = 3) were fitted into a nonlinear sigmoid plot to take into account non-linear concentration dependent of enzyme inhibitor interaction at IJFS October 2016 Volume 5 pages 212-223 low and high concentrations (Burlingham & Widlanski, 2003).

Statistical analysis
One-way analysis of variance (ANOVA) was used to analyse the data followed by Least Significant Difference (LSD) test (p<0.05)for significant differences using SAS software (Version 9.2).

Results and Discussion
As previously mentioned in the introduction (section 1), it is evident that analysis of antioxidant capacity and other health benefits derived from fruit polyphenolics must include both FPPs and BPPs.In this study we have taken into account FPPs and BPPs when discussing the polyphenol contents, antioxidant activity, and starchdigestive enzyme inhibition.BPP fraction has been subjected to acid-catalysed depolymerisation to study the above properties of the resulting degraded end-products.

Total phenol and flavonoid content
We found significant differences (p<0.05) between the total phenol contents (TPC) in FPP fractions of the date seed varieties (Table 1).Deglet Nour and Dayrie seeds showed significantly (p<0.05)high phenolic contents (4166±227 and 3777±309 mg 100g −1 , respectively) compared with other varieties, whereas Medjool showed the lowest (2088±124 mg 100g −1 ).Considerable levels of TPC were found in the BPP fractions in all varieties (Table 1).The TPC quantities in BPP fraction (reported as mg 100g −1 DSP) were the greatest in Medjool seeds (514±13), followed by Barhee (khalal) (405.9±9.2),Barhee (rutab) (378.9±8.9),Deglet Nour (373±4.6)and Dayrie (345.2±5.0).The reported TPCs of date seeds in the literature varied due to different methodologies used in extraction and quantification, as well as varietal and geographic factors (Sirisena, Ng, & Ajlouni, 2015).That was also evident in the current study where varieties differed significantly (p<0.05) in terms of phenolic contents.Most authors have reported very high TPC for the FPP of date seeds, ranging from 2000-4500 mg 100g Since not all phenolic compounds are flavonoids, the flavonoids content are expected to be smaller than the phenolic contents.Results confirmed that analyses showed significantly (p<0.05)smaller quantities of TFCs in comparison with TPCs in both FPPs and BPPs (Table 1).The TFCs in Medjool seeds were the lowest (9.4±0.58 mg QE 100g −1 DSP) in the FPPs, and the highest in the BPPs (30.4±1.7 mg QE 100g −1 DSP), while other varieties showed the opposite trend.
Measuring flavonoids contents can provide an indirect assessment of its proportional contribution to the total phenols and also its contribution to antioxidant activity.

Antioxidant properties of FPP and BPPs
Antioxidant activity of food components can provide functionality contributing to disease prevention by reducing oxidative stress in the gastrointestinal tract and, if absorbed, in body tissues.The antioxidant capacity in the FPPs and BPPs of date seeds were evaluated using TEAC and FRAP assays.These analyses were based on the ability of antioxidants to scavenge (by electron or hydrogen atom donation) an organic free radical or to reduce iron, respectively.Significant differences in antioxidant activity were observed between the seed varieties as measured by TEAC and FRAP (

Inhibition of carbohydrate hydrolysing enzymes
Inhibition of intestinal α-amylase and αglucosidase (membrane bound sucraseisomaltase and maltase-glucoamylase complexes) can prevent starch digestion and delay or reduce postprandial hyperglycaemia due to inhibition of starch digestion (Funke & Melzig, 2006).Since control of hyperglycaemia is an essential treatment strategy for diabetes, α-amylase and α-glucosidase inhibitors play a role in the management of post-prandial hyperglycaemia (Tundis et al., 2010).Ideally, the drugs should have mild α-amylase inhibition and moderate α-glucosidase inhibition, which would minimize the side effects such as flatulence and abdominal pain caused by colonic fermentation of intact starch (Yee & Fong, 1996).

Inhibition of α-glucosidase
To our knowledge, no previous studies have reported the inhibitory effect of date seed extract against α-glucosidase.However, El-Fouhil, Ahmed, and Darwish (2010) reported that diabetic rats fed on boiled aqueous date seed extract resulted in significant hypoglycaemia.In addition, Takaeidi et al. (2014) reported that methanolic date seed extract was able to promote the activities of serum paraoxonase (PON1) enzyme, which is known to be protective against oxidative stress related to diabetic complications.These studies pointed to the possible role of date seeds polyphenol in diabetic control.We showed that BPPs from DSP inhibited rat intestinal α-glucosidase (Table 2), ranging from 58-96%, with BPPs from Deglet Nour seeds showing the strongest percentage inhibition (96.1±1.9) at 1 mg GAE mL −1 .Inhibition from FPPs was moderate to low, ranging from 39-65 % at 1 mg GAE mL −1 , with FPPs from Medjool seeds showing the strongest percentage inhibition (65.3±1.5).Acarbose, which was used as a positive inhibitor of α-glucosidase, inhibited activity completely at a 1 mg mL −1 .The IC 50 values gave a clearer indication of the inhibitory potential of DSP against the α-glucosidase (Table 3).The IC 50 of FPPs from the seeds ranged from 0.907-1.75mg GAE mL −1 , and again showing Medjool seeds with the strongest effect (0.91±0.07 mg GAE mL −1 ).BPPs from DSP showed a much stronger inhibition (smaller IC 50 ) than FPP, with values ranging from 0.39-0.68mg GAE mL −1 .Deglet Nour BPP showed the strongest inhibition with IC 50 value of 0.399±0.018mg GAE mL −1 .The difference in α-glucosidase inhibitory effect between FPPs and BPPs from the seeds reflects on the different phenolic composition between the two phenolic fractions.Detailed characterisation of the FPP and BPP fractions of date seeds has been performed using UHPLC-MS in a separate study, which showed that FPPs constitute mainly simple and glycosylated phenolic acids, flavonols and flavanols, while the BPPs contained partially depolymerised proanthocyanidins and IJFS October 2016 Volume 5 pages 212-223 simple phenolic acids including several dihydroxybenzoic acid isomers (unpublished data).

Inhibition of α-amylase
We initially measured the inhibition of α-amylase using the starch-DNS method according to literature (α-amylase assay I), but observed significant absorbance interference due reduction of the DNS reagent by polyphenols (data not shown).The reducing effect of polyphenols with DNS have been documented before (Xu, Leng, Wang, & Zhang, 2012), thus this method is not suitable for the assay of α-amylase activity with polyphenols.Therefore an alternative method was adapted using a chromogenic maltopentaoside substrate (α-amylase assay II), which is similar to the α-glucosidase assay in that liberated p-nitrophenol is measured by absorbance to indicate enzyme activity.
Figure 2 shows the linear inhibitory kinetics of porcine pancreatic α-amylase exhibited by FPPs and BPPs from Dayrie seeds as measured by αamylase assay II.Acarbose, used as a positive inhibitor of α-amylase, almost completely inhibited activity at 1 mg mL −1 .The percentage inhibition can be obtained by the ratio of the linear gradient of extract incubation over that of control incubation, which were 18.8% and 40.8% for FPP and BPPs, respectively at 1 mg GAE mL −1 of Dayrie seed extract.Table 2 summarised the % α-amylase inhibition by FPPs and BPPs for the other seed varieties.It showed that inhibition by BPPs (ranged from 9-57%) were generally stronger than inhibition by FPPs (ranged from 7-18%) for the same seed variety.Furthermore, the inhibition of α-glucosidase was generally stronger than the inhibition of α-amylase for the same seed variety.IC50 was not determined for α-amylase due to the low inhibitory activity shown by both fractions.The data did not show a clear correlation between α-glucosidase and α-amylase inhibition with phenolic content and antioxidant capacity of the FPP and BPP fractions.This could be due to different polyphenolic 'acids profile' of the 2 fractions.How polyphenols mediate enzyme inhibition still require more thorough investigation.At the molecular level, the number and position of free hydroxyl groups in the polyphenol structure play an important role in both αamylase and α-glucosidase inhibition (Funke & Melzig, 2006).Simple phenolic acids such as hydroxybenzoic acids and hydroxycinnamic acids were not inhibitory (Phan et al., 2013).Tannins showed mild inhibitory activity towards αamylase, but, ellagitannins with β-galloyl groups at the glucose C-1 position showed a higher inhibition than those with α-galloyl and desgalloyl compounds (Li, Tanaka, Zhang, Yang, & Kouno, 2007).In a plant extract however, the IJFS October 2016 Volume 5 pages 212-223

Conclusions
Our study showed that Australian date seeds have high polyphenol contents and in-vitro antioxidant activities.Bound polyphenol fraction from date seeds showed larger inhibitory capacity towards both starch digestive enzymes; α-amylase and α-glucosidase, compared to the free polyphenol fraction.Also, the bound polyphenol fraction showed strong inhibition towards α-glucosidase but weak inhibition towards α-amylase, indicating they are good source of functional ingredients with potent antihyperglycaemic properties.

Figure 1 :
Figure 1: Date seeds from the studied varieties, including seeds from two maturity stages of Barhee (khalal -early and rutab-later stages)

Figure 2 :
Figure 2: Inhibition of porcine pancreatic α-amylase using PNPG5 as substrate (α-amylase assay II).The absorbance at different time points was plotted as mean±SD (n≥3).FPP (long dash) and BPP (long dash and dot) of Dayrie variety at 1 mg GAE mL −1 reaction was compared to control (solid line) without added inhibitor.Acarbose (dots) at 1 mg mL −1 reaction was used as a positive inhibition control

Table 2 :
Inhibition of rat intestinal α-glucosidase and porcine pancreatic α-amylase by date seed free polyphenol (FPP) and bound polyphenol (BPP) extracts Medjool 65.3 ± 1.5 d 15.56 ± 0.85 a 68.58 ± 0.60 b 57.84 ± 0.42 d Percentage inhibition is relative to control without added inhibitor, and values listed are mean % inhibition ± SD (n≥3) a,b,c,d : Values sharing different superscript letters in a column are significantly different (p<0.05)

Table 3 :
IC 50 values for the inhibition of rat intestinal α-glucosidase by FPP and BPP of date seed extract