In vitro studies on the effect of watercress juice on digestive enzymes relevant to type 2 diabetes and obesity and antioxidant activity

Inhibition of enzymes involved in carbohydrate and lipids metabolism is an important strategy against type 2 diabetes and obesity, by suppressing dietary sugar and fat absorption. This work reports, for the first time, the in vitro inhibition of α-glucosidase, α-amylase and lipase by watercress juice (WJ). Juice was analyzed for qualitative and quantitative composition and in vitro antioxidant activities. Several components were identified, namely hydroxycinnamic acids, flavonols, and other minor water-soluble phytochemicals. Quantitative data revealed a dimer of caffeoylmalic acid (0.73 mg mL−1 of juice), disinapoylgentibiose (0.64 mg mL−1), ferulic acid (0.56 mg mL−1), and isorhamnetin-O-sophoroside-O-malonyl(hexoside) (0.38 mg mL−1) as the predominant polyphenols. The results showed that WJ had dose-dependent inhibitory potential against targeted enzymes, displaying a more potent inhibitory effect against α-glucosidase relative to α-amylase and lipase. WJ can be considered a potential complementary dietary approach to control hyperglycaemia and hyperlipidaemia, through inhibition of digestive enzymes. 
 
Practical applications 
Currently, raw watercress is widely consumed in liquid form, alone or together with other fresh vegetables, as an ingredient of the “detox juices.” Studies on watercress phenolic composition and antidiabetic properties have been performed on extracts, but not on juice, which is closer to the edible form. The present study supports consumption of watercress juice as source of phytochemicals potentially capable of inhibiting digestive enzymes linked to diabetes and obesity prevention/control.

. Vegetable juices are a natural dietary source of polyphenols, often used as a low-sugar alternative to fruit juice. Watercress (Nasturtium officinale W. T. Aiton; syn: Rorippa nasturtium-aquaticum, Brassicaceae) is an aquatic leafy vegetable with great economic importance in Portugal: about 5.000 tons per year are exported to countries in Northern Europe.
Usually, the raw leaves are consumed fresh as salad greens or steamed and consumed as other vegetables (Mousa-Al-Reza Hadjzadeh, Moradi, & Ghorbani, 2015). Currently, it is widely consumed fresh in liquid form, alone or together with other fresh vegetables, as a trendy ingredient of the so called "detox juices" (Escard o & Cuadra, 2014).
Previously, the vitamin C content in watercress juice (WJ) and its stability with time and temperature was determined (Spínola, Mendes, Câmara, & Castilho, 2013). The aim of the present study was to further characterize its phytochemical profile and assess its in vitro inhibitory effect against key enzymes linked to T2DM and obesity (a-glucoside, a-amylase, and lipase) as well as its antioxidant activity.

| Sample preparation
Watercress (var. Hampshire) was grown in Madeira Island (Portugal) by different producers and supplied to Organic Chemistry and Natural Products Laboratory (NatLab, CQM) two days after harvest, in bunches of about 500 g each (total 3 kg). Portions of watercress leaves and tender stalks were collected from each bunch, mixed, and homogenized in a prechilled blender and the homogenates were centrifuged for 30 min (10,000 rpm; 2-48C) (Spínola et al., 2013). Supernatants were filtered, and the resulting liquid (raw extract, from now on designated as "juice") was stored at 2808C. Three batches were obtained by this procedure.
Total soluble solids were determined using an Atago RX-1000 digital refractometer and the result (2.5 Brix) was similar to that reported previously for watercress (Vinha et al., 2015).

| Chromatographic conditions
HPLC analysis was carried out on a Dionex ultimate 3000 series instrument coupled to a binary pump, a diode-array detector (DAD), an autosampler and a column compartment (kept at 208C). Separation was performed on a Phenomenex Gemini C 18 column (5 mm, 250 3 3.0 mm i.d.) using a mobile phase composed of CH 3 CN (A) and formic acid/ water (0.1%, v/v) at a flow rate of 0.4 mL min 21 . The following gradient program was used: 20% A (10 min), 25% A (20 min), 50% A (40 min), 100% A (42-47 min), and 20% A (49-55 min). Spectral data for all peaks were accumulated in the range of 210-400 nm. Juice was diluted with the initial eluent gradient (dilution 1/10), filtered through (0.45 mm) and 10 lL were injected directly.
For HPLC-DAD/ESI-MS n analysis, a Bruker Esquire model 6000 ion trap mass spectrometer (Bremen, Germany) with an ESI source was used. MS n analysis was performed in negative mode and the scan range was set at m/z 100-1,000 with a speed of 13,000 Da s 21 . The conditions of ESI were as follows: drying and nebulizer gas (N 2 ) flow rate and pressure, 10 mL min 21 and 50 psi; capillary temperature, 3258C; capillary voltage, 4.5 keV; collision gas (He) pressure and energy, 1 3 10 25 mbar and 40 eV; and fragmentor, 1.0 eV. Esquire control software was used for data acquisition and Data Analysis software for processing.

| Quantitative determination of polyphenols
Quantification of main polyphenols was carried out by HPLC-DAD, based on a previous method (Díaz-García, Obon, Castellar, Collado, & Alacid, 2013). Caffeic acid, protocatechuic acid, and quercetin were used for hydroxycinnamic, hydroxybenzoic acids, and flavonols quantification, respectively. Calibration curves (5-100 mg L 21 ) were prepared by diluting the stock solutions (1,000 mg L 21 in methanol) with initial mobile phase. Quantification was carried out by plotting peak area versus concentration (R 2 ! 0.967 in all cases).

| a-Glucosidase inhibition assay
This assay was adapted from a previous report (Podse R dek et al., 2014).
In a 96-well plate, 50 lL of juice (serial dilutions) was combined with 50 lL of enzyme solution (0.1 mg mL 21 in 0.1 mol L 21 phosphate buffer; pH 6.9) and incubated for 20 min. The reaction was initiated by adding 50 lL of 5 mmol L 21 pNPG solution in the above buffer. The mixture was incubated at 378C for 20 min. Finally, 100 lL of 0.1 mol L 21 Na 2 CO 3 solution was added and the absorbance was read at 405 nm (Victor 3 microtiter reader; Perkin-Elmer, Ueberlingen, Germany). Inhibition (IC 50 value) of juices was expressed in terms of acarbose equivalent (AE) concentration (mg AE mL 21 solution) (Tiwari et al., 2013), and acarbose was used as positive control.

| a-Amylase inhibition assay
The assay was performed as described previously (Podse R dek et al., 2014), with slight modifications: 20 lL of juice (serial dilutions) and 40 lL of 2 g L 21 starch solution were mixed with 20 lL of a-amylase (0.1 mg mL 21 in 0.1 mol L 21 phosphate buffer; pH 6.9). After incubation at 378C for 20 min, the reaction was stopped by addition of 80 lL of 0.4 mol L 21 HCl followed by 100 lL of 5 mmol L 21 I 2 (in 5 mmol L 21 KI), and the absorbance was read at 620 nm (Victor 3 microtiter reader; Perkin-Elmer). Acarbose was used as positive control and inhibition (IC 50 value) was expressed as described previously.

| Lipase inhibition assay
The method for measuring lipase activity was based on a previous protocol (Kim et al., 2010): 40 mL of juice was mixed with 20 mL of 10 mM of pNPB solution and 40 mL of the enzyme (2.5 mg mL 21 prepared in 0.1 M phosphate buffer; pH 8.0). After incubation (20 min; 378C), absorbance was read at 405 nm (Victor 3 microtiter reader; Perkin-Elmer). Orlistat was used as positive control and inhibitory activity (IC 50 value) was expressed in terms of orlistat equivalent (OE) concentration (mg OE mL 21 solution).

| Total phenolic content (TPC)
The method was adapted from previous reports (Isabelle et al., 2010): 50 lL of sample was mixed with 1.25 mL of FCR (diluted 1:10) and 1 mL of 75 g L 21 Na 2 CO 3 , added to a 5 mL test tube and mixed. After 30 min, absorbance of the reaction mixture was measured at 765 nm (UV2Vis Lambda 2 spectrophotometer, Perkin Elmer, Oberlingen, Germany). The amount of total phenolics was expressed as mg of gallic acid equivalents (GAE) mL 21 of juice.

| ABTS radical scavenging activity
The ABTS ·1 assay used was a modified version (Tiveron et al., 2012): 40 lL of sample was added to 1.96 mL of the ABTS ·1 solution (diluted in PBS, pH 7.4; absorbance 0.700 6 0.021). The reduction of absorbance at 734 nm (Lambda 2 spectrophotometer, Perkin-Elmer) was measured for 6 min. Results were expressed as mmol Trolox equivalent (TE) mL 21 juice, based on the Trolox calibration curve.

| Oxygen radical absorbance capacity (ORAC) assay
The ORAC assay followed a reported method with some modifications were taken every minute for 60 min (Victor 3 microtiter reader; Perkin-Elmer) and results were expressed as mmol TE mL 21 juice, based on the Trolox calibration curve.

| Statistical analysis
All samples were assayed in triplicate (n 5 3) and results were given as the means 6 standard deviations. Data were analyzed by means of a oneway ANOVA using SPSS for Windows, IBM SPSS Statistics 20 (SPSS, Inc., USA). A value of p < 0.05 was considered statistically significant.

| Phytochemical composition
In this study, identification of phytochemicals was assigned by combining the DAD and mass spectrometry data obtained under negative electron spray ionization (ESI -) together with scientific information available in literature (Table 1).
Compounds were numbered by their elution order and a total of 40 compounds were detected (Fig. S1). These were mainly polyphenols (in particular hydroxycinnamic acids [HCAs]) but other phytochemicals were detected as minor constituents (organic acids, terpenoid, lignan and oligosaccharides). This is the first report of some of these compounds in   (Cartea et al., 2011;Harbaum et al., 2007;Lin & Harnly, 2010). Malic acid derivatives were identified according to previous studies in other Brassica vegetables (Harbaum et al., 2007;Lin & Harnly, 2010;Santos et al., 2014). Except for caffeoylmalic acid, these compounds have not been documented in watercress.
Isorhamnetin glycosides characterized based on previous studies (Harbaum et al., 2007;Lin & Harnly, 2010) were also present. Free kaempferol was identified in association with a formate adduct (probably from mobile phase). An L-AA derivative was tentatively characterized, by comparison with L-AA commercial standard.

| Quantitative analysis
Seventeen main polyphenols from WJ were quantified (Table 1) In another study (Boligon et al., 2013), higher content of HCAs Oxalic acid (OA) content in watercress (12.16 g kg 21 fresh weight, FW) was at the lower end of the range found in literature (7.54-60.12 g kg 21 FW) (Khan et al., 2016;Pinela et al., 2016), although higher than that of leaves of Swiss chard, spinach, or beetroot analyzed under   | 5 of 8 same conditions (results not shown) and generally considered as "high OA containing vegetables" (Getting et al., 2013). OA has been reported as the main organic acid in watercress (Khan et al., 2016) and it is considered an antinutrient due to the reduction of dietary calcium bioavailability and formation of oxalate kidney calculus (Getting et al., 2013).

| In vitro enzyme inhibition assays
WJ was able to inhibit digestive enzymes in a dose-dependent manner, with a stronger inhibitor effect on a-glucosidase than a-amylase and lipase activity (Table 2). Significant differences (p < 0.05) were found between WJ and positive controls (acarbose and orlistat) in tested assays.
The inhibitory effects of some of the main components of WJ on porcine pancreas a-amylase activity had been previously reported.
Ferulic, caffeic and sinapic acids showed higher IC 50 values (>0.86 mg L 21 ) than acarbose (0.015 mg L 21 ) (Funke & Melzig, 2005). Other bioactive compounds in WJ may contribute to these inhibitory activities. Roseoside, pinoresinol and their glycosides are known to be a-glucosidase inhibitory agents (Kwon et al., 2014;Yang, Liang, Xie, & Wei, 2016), while malic acid has been characterized as the active principle for inhibition of a-glucosidase, a-amylase and lipase from Fla- The increased free-radical production by oxidative stress is associated with the initiation and progression of diabetes and related complications (Podse R dek et al., 2014;Tiwari et al., 2013). Therefore, supplementation with radical scavenging antioxidants can be useful for prevention and/or reduction of oxidative stress involved in this metabolic disorder. High antioxidant activities for watercress extracts have been previously reported (Boligon et al., 2013;Isabelle et al., 2010;Martínez-S anchez et al., 2008;Payne et al., 2013;Pereira et al., 2011;Tiveron et al., 2012). Generally, antioxidant measurements obtained from water extracts were lower than those obtained with organic solvents (Oroian & Escriche, 2015;Sulaiman, Sajak, Ooi, & Seow, 2011).
This is in agreement with the present results (Table 3). Antioxidant activities of WJ (Table 3) were at the lower end of the range documented for other vegetable juices (5-2,260 mmol TE mL 21 ) (Simsek et al., 2014;Tabart et al., 2009).

| C O NC LU S I O N S
Thirty-nine compounds (phenolic and non-phenolic) were characterized in WJ. HCAs were the main class of polyphenols, followed by flavonols (isorhamentin and quercetin derivatives), which is a different trend from that reported in the literature.Inhibitiory activities towards digestive enzymes may have been due to polyphenols, however other minor compounds present in the juice could also be involved. This is the first report on digestive enzyme inhibitory activity by WJ, suggesting its potential as a viable source of phytochemicals for the management of