Effectiveness of different depuration procedures in removing reagents interference on in vitro digested strawberry extracts for reliable antioxidant determinations

Abstract

BACKGROUND:

Healthy benefits associated with strawberries consumption are mostly related to their antioxidant composition, mainly polyphenols. Quality assessment on fresh fruits is commonly done by spectrophotometric methods, but intake and digestion may alter their composition and healthy properties. To asses antioxidants bioavailability at different gastrointestinal-tract levels, in vitro digestion (IvD) simulations are used but reagents involved in may interfere in antioxidant determinations despite depuration procedures are employed.

OBJECTIVE:

To test the magnitude of reagents interference in IvD approaches and the effectiveness of different depuration procedures for reliable antioxidant quantifications on strawberries.

METHODS:

IvD assays were done with water and strawberry samples to obtain digested fractions (gastric and intestinal). After passing-through hydrophilic cotton, digested extracts were subjected to different depuration procedures: centrifugation, Sep-Pack and 0.45 μm nylon-filter. Antioxidant content and capacity were evaluated spectrophotometrically.

RESULTS:

IvD reagents interfered in all antioxidant determinations, especially in the intestinal fractions. Depuration procedures differed in their effectiveness for reagents removal and in their antioxidant retrieval efficiency, with hydrophilic cotton displaying better recovery efficiency.

CONCLUSIONS:

Reagents interference should be considered for antioxidant content and capacity determinations after IvD but, for reliable estimations of healthy compounds of food matrices, depuration methods should prioritize antioxidant recovery over reagents removal.

Keywords

Bioavailability biliar salts pepsin gastrointestinal digestion phenolic compounds spectrophotometric

1 Introduction

Strawberries (Fragaria×ananassa, Duch.) are among the most valuable fruits for fresh consumption, widely consumed in the world. Besides their pleasant taste, fruits are highly appreciated by consumers because of their healthy benefits [1, 2], which have been related to the antioxidant capacity of the polyphenolic compounds they contain [3, 4], including anthocyanins, which are responsible of the characteristic red colour of the fruits [5, 6].

In this sense, improvement of healthy quality traits of the fruit (i.e. polyphenols and antioxidant capacity) for variety selection has recently become a target of many strawberry breeding programs [7]. In such context, assessment of polyphenol composition (i.e. total phenolic, flavonoid and anthocyanin content) and antioxidant capacity is commonly determined in fresh fruits by spectrophotometric methods [8].

However, intake and digestion process may alter fruit composition and strawberry healthy properties. In fact, it has been reported that phenolic compounds entering in the gastrointestinal tract could be hydrolyzed and/or conjugated to be released from the food matrix (bioaccessibility) and, therefore, absorbed (bioavailability) in such modified chemical form [9]. Also, the rate of phenolic compounds absorbed differ depending on their chemical nature. Thus, it has been shown that ingested anthocyanins that are actually absorbed at the gastrointestinal level is quite low [10].

For determining to what extent strawberry fresh-fruit phenolic composition is related to healthy effects after intake and digestion, in vitro digestion approaches have been developed providing a powerful tool to elucidate the antioxidant modification, release and absorption at the different gastrointestinal tract levels [11 –13]. These approaches involve the use of reagents such as pepsin, pancreatin and bile salts, that may interfere on the quantification of the antioxidant compounds, despite the digested fractions are commonly depurated by centrifugation [14, 15] or by the use of solid phase extraction techniques [16], which are expensive and time-consuming. However, no studies to date have tested how effective these depuration procedures in removing these reagents are, and to what extent the filtered extracts gather all type of antioxidants released, which is of importance to avoid misleading results in the spectrophotometric quantification of antioxidants after digestion. In this sense, this study aims to evaluate the feasibility of different depuration methods for in vitro digestion approaches and to assess their reliability for achieving an accurate determination of bioavailable and bioaccessible antioxidants.

2 Materials and methods

2.1 Experimental design and plant material

In vitro digestion assays were performed to evaluate the effect of in vitro digestion reagents on the antioxidant profile of fruit samples and to test the effectiveness of different depuration procedures in removing these reagents. Therefore, strawberry (Fragaria×ananassa Duch., ‘Charlene’) puree samples or Milli-Q water (used as negative control) were used for the assays, which were carried out per triplicate on each type of sample. Strawberry samples, consisting on 5–10 fully-ripen fruits, were harvested on three plots (50 plants each), at the IFAPA field experimental station ‘El Cebollar’ (Huelva, Spain). Whole fruits were blended immediately after harvesting and puree samples were stored at – 20°C until analysis.

2.2 In vitro digestion process

The in vitro digestion procedure was performed according to the method developed by Gil-Izquierdo et al. [11], with some modifications [13]. Briefly, strawberry puree or Milli-Q water (10 g/100 mL) were immersed in a beaker containing Milli-Q water, 514 units of pepsin g^–1, HCl (pH ∼1.7–2) and a dialysis membrane (molecular weight cut-off of 12,000 Da) containing Milli-Q water. Samples were kept in darkness during 2 h at 37°C for simulating digestion at the gastric level. Afterwards, the potentially absorbed (GF-IN; inside the membrane fraction = bioavailable) and released (GF-Out; outside the membrane fraction = bioaccessible) gastric fractions were collected. Then, the GF-Out was used to continue the simulation at the intestinal level, by adding pancreatin and bile extract (4 g L^–1 and 25 g L^–1, respectively) mixture into the beaker and by adjusting the pH to ∼7.8 (slightly alkaline). An additional dialysis membrane, containing Milli-Q water and the necessary amount of NaHCO₃ to titrate the mixture to pH 7.8, was also immersed into the beaker. After keeping the mixture at 37°C in darkness for 2 h in a water bath, the potentially absorbed (IF-In; bioavailable) and released (IF-Out; non-bioavailable or bioaccessible) intestinal fractions were collected.

2.3 Depuration assay

Four different depuration procedures were tested to compare their effectiveness in removing chemical interferences and reagents noise:signal ratio in the fractions collected after digestion simulations. These procedures are among the most commonly used in in vitro assays [14 –16]. After the depuration, sample fractions were stored at – 20°C until their spectrophotometric analysis.

2.3.1 Hydrophilic Cotton (HC)

One piece of hydrophilic cotton (HC) was used for filtrating solid residues on each digested fraction [17]. HC was used as a reference for comparisons.

2.3.2 Hydrophilic Cotton + Centrifugation (HC + C)

The digested fractions were passed through HC and the filtrates were centrifuged at 4°C and 10000 rpm for 10 min [18]. The pellet was discarded from the assay.

2.3.3 Hydrophilic Cotton + 0.45-μm nylon filter (HC+N)

Coupled to HC, fractions were passed through nylon membrane syringe filters with 0.45 μm pore size (Agilent Technologies, Santa Clara, CA) and hydrophilic properties [19].

2.3.4 Hydrophilic Cotton + solid phase extraction (HC + SPE)

After HC, collected samples were carefully charged into C-18 Sep-Pack Vac 6cc cartridges (Waters Assoc., Milford, MA) for solid phase extraction (SPE) according to the method described in Anthony et al. [20] with minor modifications. This procedure allows removal of sugars, organic acids and other water-soluble compounds to obtain polyphenol-enriched eluted extracts.

2.4 Spectrophotometric determination of total phenols, flavonoids, anthocyanins and antioxidant capacity

Total phenolic, flavonoid and anthocyanin contents were measured according to Singleton et al. [21], Dewanto et al. [22] and Giusti et al. [23], respectively. Results were expressed as milligrams of standard equivalent (Gallic acid: GAE; Catechin: CE; Pelargonidin: PE, respectively) per milliliter of fraction, or milligrams of standard equivalent (GAE, CE or PE) per 100 grams of Fruit fresh weight (F_fw). For the determination of the antioxidant capacity, TEAC (Trolox Equivalent Antioxidant Capacity) method was followed as in Re et al. [24]. TEAC results were expressed as micromols of Trolox equivalents (TE) per milliliter of fraction or micromols of TE per 100 grams of F_fw. In all assays, three technical replicates were measured on each filtrated fraction.

2.5 Calculations and statistical analysis

The relative contribution of reagents to antioxidant amount or capacity of each fraction and protocol in digested fruit samples, i.e. reagents interference index (Ri), was calculated as follows: Ri = [(Conc X)_water/(Conc X)_sample]×100, being (Conc X) = concentration of the X compound in mg X eq/ mL of digested extract.

To allow comparisons on the efficacy of depuration procedures in retrieving antioxidants from fruit samples, an effectiveness index (Ei) was calculated as: Ei = [(AOXs_DM/AOXs_HC)]×100, where AOXs_DM is the amount of antioxidants of each depuration method and AOXs_HC is the corresponding HC antioxidant content, used as a reference for comparisons among methods. Low values of Ei indicate a low antioxidant recovery compared to HC alone (i.e. high magnitude of antioxidant losses due to depuration methods), whereas values above 100 involve higher antioxidant recovery than HC.

All data were subjected to analysis of variance (ANOVA) and differences between mean values were assessed by Tukey honest significant differences (HSD) at a significance level of P≤0.05, with the analytical software STATISTIX 9.0 (Analytical Software, Florida, USA).

3 Results and discussion

This work is the first report highlighting the importance of the reagents interference in the analytical methods commonly used for antioxidant quantification in in vitro digestion approaches. It also evidences the great discrepancy in the effectiveness of antioxidant retrieving among the different depuration procedures. These issues are of special significance for a proper assessment of the healthy properties of food matrices, such as strawberries, after in vitro digestion approaches, and for avoiding misleading attribution of healthy properties to the fruits.

3.1 Interference of reagents on spectrophotometric antioxidant determination after in vitro digestion

The effect of in vitro digestion reagents on spectrophotometric antioxidant determination was assessed by simulating the digestion process on water samples (i.e. reference or control, without strawberry sample). Results revealed that these reagents caused a detectable signal in all the antioxidant protocols used, suggesting certain antioxidant overestimation (Table 1). This was reflected by the reagents interference index (Ri), which reached values up to 25.4% and 100% at the gastric and intestinal level, respectively. The signal observed in most protocols tested in both gastric fractions (i.e. GF-Out and GF-In) might be due to the presence of pepsin, which is capable to pass through the dialysis membrane. However, no signal was detected on the gastric fractions under the conditions of the anthocyanin protocol, suggesting that pepsin interference for anthocyanins determinations are negligible.

Table 1
Total phenolic, flavonoid and anthocyanin content, as well as, antioxidant capacity of digested fractions without strawberry (as a control), employing four different depuration methods (HC: hydrophilic cotton; HC + C: hydrophilic cotton + centrifuge; HC + N: hydrophilic cotton + nylon and HC + SPE: hydrophilic cotton + solid phase extraction). Values within parenthesis indicate the reagents interference index (Ri). All analyses were performed in triplicate. Data (mean±SE) are expressed in mg or μmol standard equivalent /mL fractions

Digestion fractions Depuration methods Total phenolics (mg GAE/mL fraction) Flavonoids (mg CE/mL fraction) Anthocyanins (mg PE/mL fraction) Antioxidant capacity (μmol TE/mL fraction)

GF-In HC 0.0006±3.6E–05^ab 0.0007±3.8E–20^a <LOD 0.0388±4.4E-03^a

(0.8) (6.3) (15.4)

HC + C 0.0005±3.3E–05^b 0.0007±3.8E–20^a <LOD 0.0364±4.9E-03^a

(0.8) (5.0) (15.4)

HC + N 0.0002±4.9E–05^c 0.0007±3.8E–20^a <LOD 0.0329±5.6E-03^a

(0.4) (25.4) (31.4)

HC + SPE 0.0007±2.2E–04^a 0.0008±7.2E–05^a <LOD 0.0309±8.1E-03^a

(1.4) (6.9) (12.4)

GF-Out HC 0.0006±0.0E+00^ab 0.0007±3.8E–20^a <LOD 0.0473±2.1E-03^a

(0.3) (1.8) (5.7)

HC + C 0.0005±2.7E–05b^c 0.0007±3.8E–20^a <LOD 0.0473±2.1E-03^a

(0.3) (2.3) (6.0)

HC + N 0.0005±1.7E–05^c 0.0007±4.1E–05^a <LOD 0.0376±7.3E-03^a

(0.7) (14.1) (14.0)

HC + SPE 0.0006±6.4E–05^a 0.0008±1.4E–04^a <LOD 0.0487±9.6E-03^a

(0.5) (3.5) (6.8)

IF-In HC 0.0087±1.1E–03^ab 0.0013±4.7E–05^a 0.0003±4.5E–05^a 0.1456±8.7E-03^a

(27.2) (21.8) (6.7) (33.5)

HC + C 0.0097±5.3E–04^a 0.0009±1.1E–04^a 0.0001±4.0E-05^b 0.1529±8.4E-03^a

(19.1) (12.5) (2.1) (31.7)

HC + N 0.0076±3.1E–04^ab 0.0012±9.5E–05^a 0.0001±4.4E–05^ab 0.1353±1.0E-02^a

(21.7) (32.1) (8.6) (38.9)

HC + SPE 0.0071±2.1E–04^b 0.0013±2.0E–04^a 0.0000±3.1E-06^b 0.0644±1.1E-02^b

(22.9) (22.5) (0.3) (100.0)

IF-Out HC 0.0343±1.2E–03^a 0.0103±7.1E–04^a 0.0022±1.2E-04^a 0.3984±2.2E-02^a

(18.8) (25.0) (6.1) (25.1)

HC + C 0.0224±1.7E–03^bc 0.0035±7.5E–05^b 0.0021±8.9E–05^a 0.3679±9.1E-03^a

(16.8) (16.1) (7.1) (27.7)

HC + N 0.0253±3.0E–03^b 0.0027±2.1E–04^b 0.0021±8.1E–05^a 0.3704±1.9E-02^a

(20.9) (28.5) (8.2) (37.1)

HC + SPE 0.0178±8.2E–04^c 0.0031±8.4E-05^b 0.0021±6.5E–05^a 0.0927±1.1E-02^b

(22.7) (19.0) (7.4) (21.2)

Digestion fractions	Depuration methods	Total phenolics (mg GAE/mL fraction)	Flavonoids (mg CE/mL fraction)	Anthocyanins (mg PE/mL fraction)	Antioxidant capacity (μmol TE/mL fraction)
GF-In	HC	0.0006±3.6E–05^ab	0.0007±3.8E–20^a	<LOD	0.0388±4.4E-03^a
		(0.8)	(6.3)		(15.4)
	HC + C	0.0005±3.3E–05^b	0.0007±3.8E–20^a	<LOD	0.0364±4.9E-03^a
		(0.8)	(5.0)		(15.4)
	HC + N	0.0002±4.9E–05^c	0.0007±3.8E–20^a	<LOD	0.0329±5.6E-03^a
		(0.4)	(25.4)		(31.4)
	HC + SPE	0.0007±2.2E–04^a	0.0008±7.2E–05^a	<LOD	0.0309±8.1E-03^a
		(1.4)	(6.9)		(12.4)
GF-Out	HC	0.0006±0.0E+00^ab	0.0007±3.8E–20^a	<LOD	0.0473±2.1E-03^a
		(0.3)	(1.8)		(5.7)
	HC + C	0.0005±2.7E–05b^c	0.0007±3.8E–20^a	<LOD	0.0473±2.1E-03^a
		(0.3)	(2.3)		(6.0)
	HC + N	0.0005±1.7E–05^c	0.0007±4.1E–05^a	<LOD	0.0376±7.3E-03^a
		(0.7)	(14.1)		(14.0)
	HC + SPE	0.0006±6.4E–05^a	0.0008±1.4E–04^a	<LOD	0.0487±9.6E-03^a
		(0.5)	(3.5)		(6.8)
IF-In	HC	0.0087±1.1E–03^ab	0.0013±4.7E–05^a	0.0003±4.5E–05^a	0.1456±8.7E-03^a
		(27.2)	(21.8)	(6.7)	(33.5)
	HC + C	0.0097±5.3E–04^a	0.0009±1.1E–04^a	0.0001±4.0E-05^b	0.1529±8.4E-03^a
		(19.1)	(12.5)	(2.1)	(31.7)
	HC + N	0.0076±3.1E–04^ab	0.0012±9.5E–05^a	0.0001±4.4E–05^ab	0.1353±1.0E-02^a
		(21.7)	(32.1)	(8.6)	(38.9)
	HC + SPE	0.0071±2.1E–04^b	0.0013±2.0E–04^a	0.0000±3.1E-06^b	0.0644±1.1E-02^b
		(22.9)	(22.5)	(0.3)	(100.0)
IF-Out	HC	0.0343±1.2E–03^a	0.0103±7.1E–04^a	0.0022±1.2E-04^a	0.3984±2.2E-02^a
		(18.8)	(25.0)	(6.1)	(25.1)
	HC + C	0.0224±1.7E–03^bc	0.0035±7.5E–05^b	0.0021±8.9E–05^a	0.3679±9.1E-03^a
		(16.8)	(16.1)	(7.1)	(27.7)
	HC + N	0.0253±3.0E–03^b	0.0027±2.1E–04^b	0.0021±8.1E–05^a	0.3704±1.9E-02^a
		(20.9)	(28.5)	(8.2)	(37.1)
	HC + SPE	0.0178±8.2E–04^c	0.0031±8.4E-05^b	0.0021±6.5E–05^a	0.0927±1.1E-02^b
		(22.7)	(19.0)	(7.4)	(21.2)

Different letters within each digestion fraction of the same protocol indicate significant differences (P < 0.05).

In the intestinal fractions, the interference of reagents was more remarkable than in the gastric ones. The signal detected in IF-In and IF-Out might be the result of the intrinsic antioxidant properties of bile salts at physiological concentrations described by DeLange et al. [25], since pepsin interferences at this level can be discarded because it is irreversibly denatured at pH 7 or above [26]. However, in the IF-Out the interference was more marked than in the IF-In, probably due to the physical barrier imposed by the dialysis membrane to bile salts passthrough (Table 1).

At each gastrointestinal level, the magnitude of reagents interference differed among antioxidant protocols (Table 1). Thus, in the anthocyanin protocol, reagents showed the lowest interference, while the reverse was true for the antioxidant capacity determinations. This highest interference is consistent with the ability of TEAC protocol to quantify all molecules able to donate protons, including bile salts [25].

These findings are pointing out that reagents used in the in vitro digestion simulations contribute significantly to the determinations of antioxidant content and capacity of digested fractions. This fact should be taken into account on studies aiming to quantify antioxidant capacity after in vitro digestion processes of berry fruits [27 –30].

Additionally, these results also underline the need of using methods for removing these reagents and minimizing their signal (Table 1).

3.2 Effectiveness of depuration procedures in removing reagent signal after in vitro digestion

Four depuration methods were assessed in order to test their effectiveness in removing reagents from digested control samples after in vitro digestion (Table 1).

For most antioxidant protocols and digested fractions, the use of hydrophilic cotton (HC) yielded higher signal values indicating a lower effectiveness in removing reagents interference in comparison with the other depuration procedures evaluated.

In contrast, the effectiveness of the other three depuration methods (i.e. HC + C, HC + N, HC + SPE) varied depending on the protocol of antioxidant determination, being similar for the flavonoids and anthocyanins while the reverse was true for the total phenolics and antioxidant capacity determinations (Table 1). For total phenolic content, spectrophotometric signals obtained on each depuration method differed between gastric and intestinal levels, being HC + N and HC + SPE the most effective methods for removing reagents on each level, respectively. For antioxidant capacity, no differences among methods were found at the gastric level whereas, at the intestinal level, the combination of HC + SPE resulted in significantly lower signal suggesting better depuration effectiveness. These results a priori are pointing out that HC + SPE could be a reliable option for estimating antioxidant content after in vitro digestion, despite being the most time consuming and expensive method tested. However, prior to recommend this procedure, it is necessary to verify how accurate is in discriminating between reagents and antioxidants in fruit samples (i.e. to retain selectively reagents in favour of antioxidant compounds; see section 3.3).

3.3 Effectiveness of depuration procedures for recovering fruit antioxidants after in vitro digestion

To overcome the inaccuracies on antioxidant determinations due to reagent interferences, despite using depuration procedures, values of the digested control samples (i.e. water; Table 1) were subtracted from the spectrophotometric quantifications of antioxidants on strawberry samples after in vitro digestion. These values are shown in Table 2.

Table 2
Total phenolic, flavonoid and anthocyanin content, as well as, antioxidant capacity of digested fractions of ‘Charlene’ strawberry, employing four different depuration methods (HC: hydrophilic cotton; HC + C: hydrophilic cotton + centrifuge; HC + N: hydrophilic cotton + nylon and HC + SPE: hydrophilic cotton + solid phase extraction). Effectiveness index (Ei) is shown within parenthesis. All analyses were performed in triplicate. Data (mean±SE) are expressed in mg or μmol of standard equivalent /100 g fruit fresh weight (F_fw).

Digestion fractions Depuration methods Total phenolics (mg GAE/100 g F_fw) Flavonoids (mg CE/100 g F_fw) Anthocyanins (mg PE/100 g F_fw) Antioxidant capacity (μmol TE/100g F_fw)

GF-In HC 40.24±0.61^a 4.48±0.23^ab 3.00±0.20^a 106.50±6.58^a

HC + C 37.46±0.96^a (93) 4.32±0.36^b (97) 3.05±0.19^a (102) 99.93±4.78^a (94)

HC + N 23.98±0.67^b (60) 1.01±0.12^c (23) 2.00±0.16^b (67) 35.90±3.47^b (34)

HC + SPE 24.71±0.86^b (61) 5.14±0.23^a (115) 3.13±0.23^a (104) 108.95±7.12^a (102)

GF-Out HC 208.15±3.03^a 38.18±1.15^a 9.92±0.73^a 777.91±31.88^a

HC + C 185.24±5.70^b (89) 29.75±1.30^b (78) 10.36±0.77^a (104) 745.08±30.66^ab (96)

HC + N 70.54±1.91^d (34) 4.43±0.27^d (12) 7.68±0.76^b (77) 193.46±21.51^c (25)

HC + SPE 113.94±3.77^c (55) 22.74±0.66^c (60) 9.46±0.64^a (95) 669.81±14.83^b (86)

IF-In HC 23.33±1.31^a 4.77±0.61^a 3.67±0.44^a 288.66±15.00^a

HC + C 27.43±2.19^a (118) 4.87±0.80^a (102) 3.63±0.45^a (99) 328.93±21.22^a (114)

HC + N 18.25±2.59^b (78) 2.46±0.57^c (52) 1.53±0.12^c (42) 182.00±7.34^b (63)

HC + SPE 15.90±2.79^b (68) 4.34±0.57^b (91) 2.51±0.30^b (68) <LOD^c (0)

IF-Out HC 147.94±2.23^a 27.90±0.45^a 26.14±1.78^ab 1190.41±32.70^a

HC + C 111.06±4.41^b (75) 18.49±0.39^b (66) 27.61±1.67^a (106) 960.19±21.09^b (81)

HC + N 74.63±6.60^c (50) 6.85±0.47^d (25) 23.09±1.13^b (88) 522.34±54.73^c (44)

HC + SPE 60.74±2.39^d (41) 13.39±0.13^c (48) 25.81±0.42^ab (99) 344.74±22.94^d (29)

Digestion fractions	Depuration methods	Total phenolics (mg GAE/100 g F_fw)	Flavonoids (mg CE/100 g F_fw)	Anthocyanins (mg PE/100 g F_fw)	Antioxidant capacity (μmol TE/100g F_fw)
GF-In	HC	40.24±0.61^a		4.48±0.23^ab		3.00±0.20^a		106.50±6.58^a
	HC + C	37.46±0.96^a	(93)	4.32±0.36^b	(97)	3.05±0.19^a	(102)	99.93±4.78^a	(94)
	HC + N	23.98±0.67^b	(60)	1.01±0.12^c	(23)	2.00±0.16^b	(67)	35.90±3.47^b	(34)
	HC + SPE	24.71±0.86^b	(61)	5.14±0.23^a	(115)	3.13±0.23^a	(104)	108.95±7.12^a	(102)
GF-Out	HC	208.15±3.03^a		38.18±1.15^a		9.92±0.73^a		777.91±31.88^a
	HC + C	185.24±5.70^b	(89)	29.75±1.30^b	(78)	10.36±0.77^a	(104)	745.08±30.66^ab	(96)
	HC + N	70.54±1.91^d	(34)	4.43±0.27^d	(12)	7.68±0.76^b	(77)	193.46±21.51^c	(25)
	HC + SPE	113.94±3.77^c	(55)	22.74±0.66^c	(60)	9.46±0.64^a	(95)	669.81±14.83^b	(86)
IF-In	HC	23.33±1.31^a		4.77±0.61^a		3.67±0.44^a		288.66±15.00^a
	HC + C	27.43±2.19^a	(118)	4.87±0.80^a	(102)	3.63±0.45^a	(99)	328.93±21.22^a	(114)
	HC + N	18.25±2.59^b	(78)	2.46±0.57^c	(52)	1.53±0.12^c	(42)	182.00±7.34^b	(63)
	HC + SPE	15.90±2.79^b	(68)	4.34±0.57^b	(91)	2.51±0.30^b	(68)	<LOD^c	(0)
IF-Out	HC	147.94±2.23^a		27.90±0.45^a		26.14±1.78^ab		1190.41±32.70^a
	HC + C	111.06±4.41^b	(75)	18.49±0.39^b	(66)	27.61±1.67^a	(106)	960.19±21.09^b	(81)
	HC + N	74.63±6.60^c	(50)	6.85±0.47^d	(25)	23.09±1.13^b	(88)	522.34±54.73^c	(44)
	HC + SPE	60.74±2.39^d	(41)	13.39±0.13^c	(48)	25.81±0.42^ab	(99)	344.74±22.94^d	(29)

Different letters within each digestion fraction of the same protocol indicate significant differences (P < 0.05).

In most of the antioxidant protocols and digested fractions, HC was the method allowing the higher antioxidant recovery (i.e. high Ei) and the reverse was true for HC + SPE (Table 2). Thus, HC showed up to 2.4-fold higher total phenolic content than HC + SPE, and the differences between both methods were even higher for the antioxidant capacity protocol, reaching up to 3.4-fold higher values than in HC + SPE.

It is remarkable that HC + SPE values of antioxidant capacity in the IF-In were lower than the limit of detection (<LOD), despite significant amounts of phenolic, flavonoid and anthocyanin compounds were detectable. These results are suggesting that eluted antioxidants by this method might not be contributing to the antioxidant capacity of this fraction. In this sense, it is known that depuration by SPE removes selectively polar non-phenolic compounds such as sugars, organic acids and other water-soluble fruit constituents [20, 31] but also can differentially retain polymeric against non-polymeric polyphenols [32]. This could be consistent with the low recovery (i.e. low Ei values) of polyphenols observed in our study when this depuration method is used (Table 2), and therefore, it is not a reliable approach to estimate antioxidant composition of digested fruits samples.

The other two methods (HC + C, HC + N) displayed intermediate values between HC and HC + SPE and their effectiveness depended on both the antioxidant protocol and digested fraction. Differences between these two methods are mainly noticeable in the protocol for total phenolic determinations. In this case, HC + C yielded higher amounts than HC + N (Table 2). These differences could result from a higher restriction of high-molecular-weight polyphenols to pass through the nylon membrane compared to centrifugation.

Although the ‘In’ fractions (i.e. bioavailable fractions) are the most interesting for fruit healthiness assessment, it is worthy to note that reagent interferences in ‘In’ and ‘Out’ fractions are of different magnitude, either at the gastric and intestinal levels, pointing out that reagent effects should be considered for proper estimations of the balance between bioavailable and bioaccessible antioxidant compounds after intake, avoiding, in this way, underestimations of total potential bioavailability and health benefits of fruits.

4 Conclusions

The results of the present study clearly demonstrate that in vitro digestion reagents interfere in all the spectrophotometric quantifications of antioxidants, even when depuration methods are employed. These interferences must be considered to avoid overestimations of the antioxidant content of food matrices when using in vitro digestion approaches.

Although reagents effects can be removed by depuration procedures, these differ in their efficiency for antioxidant recovery leading, in some cases, to significant losses (i.e. HC + SPE depuration). Therefore, for reliable antioxidant quantifications on food matrices such as strawberries, an adequate depuration method election is critical and its efficiency for antioxidant recovery must be prioritized over reagents removal. In this sense, among the depuration methods tested in this study, HC or HC + C were the most effective for eluting antioxidants despite their ability for reagents removal were not the greatest. Thus, they are recommended for determining the potential benefits of fruit intake on human health.

Conflict of interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Funding

The authors report no funding.

Footnotes

Acknowledgments

This research was funded by the TRA201600.5, AVA201601.10 and AVA2019.034 projects and co-financed by the European Union (FEDER funds). Dr. Ariza is supported by IFAPA, Junta de Andalucía (20%) and by the “Programa Operativo Fondo Social Europeo (FSE) de Andalucía 2007–2013” (80%) under the topic “Andalucía se mueve con Europa”. Lucía Cervantes is hired by MINECO under the topic “Ayudas para la promoción de empleo Joven e Implantación del Sistema de Garantía Juvenil en I+D+I”.

References

Hannum

. Potential impact of strawberries on human health: A review of the science. Critical Reviews in Food Science and Nutrition. 2004;44:1–17.

Giampieri

, Tulipani

, Alvarez-Suarez

, Quiles

, Mezzetti

, Battino

. The strawberry: Composition, nutritional quality, and impact on human health. Nutrition. 2012;28:9–19.

Meyers

, Watkins

, Pritts

, Liu

. Antioxidant and antiproliferative activities of strawberries. Journal of Agricultural and Food Chemistry. 2003;51:6887–92.

Aaby

, Skrede

, Wrolstad

. Phenolic composition and antioxidant activities in flesh and achenes of strawberries (Fragaria ananassa). Journal of Agricultural and Food Chemistry. 2005;53:4032–40.

Stack

, Wray

. Anthocyanins. In: Harborne, J.B., editor. Methods in Plant Biochemistry, Plant Phenolics. Academic Press, London. 1993;1:325–56.

Carbone

, Preuss

, de Vos

RCH

, D’Amico

, Perrotta

, Bovy

, Martens

, Rosati

. Developmental, genetic and environmental factors affect the expression of flavonoid genes, enzymes and metabolites in strawberry fruits. Plant, Cell & Environment. 2009;32:1117–31.

Mezzetti

, Balducci

, Capocasa

, Zhong

, Cappelletti

, Di Vittori

, Mazzoni

, Giampieri

, Battino

. Breeding strawberry for higher phytochemicals content and claim it: Is it possible? International Journal of Fruit Science 2016;16:194–206.

Andrianjaka-Camps

, Heritier

, Ançay

, Andlauer

, Carlen

. Evolution of the taste-related and bioactive compound profiles of the external and internal tissues of strawberry fruits (Fragaria x ananassa) cv. ‘Clery’ during ripening. Journal of Berry Research. 2017;7:11–22.

Manach

, Scalbert

, Morand

, Rémésy

, Jiménez

. Polyphenols: Food sources and bioavailability. The American Journal of Clinical Nutrition. 2004;79:727–47.

10.

McGhie

, Walton

. The bioavailability and absorption of anthocyanins: Towards a better understanding. Molecular Nutrition & Food Research. 2007;51:702–13.

11.

Gil-Izquierdo

, Zafrilla

, Tomás-Barberán

. An in vitro method to simulate phenolic compound release from the food matrix in the gastrointestinal tract. European Food Research and Technology. 2002;214:155–9.

12.

Minekus

, Alminger

, Alvito

, Ballance

, Bohn

, Bourlieu

, Carrière

, Boutrou

, Corredig

, Dupont

, Dufour

, Egger

, Golding

, Karakaya

, Kirkhus

, Le Feunteun

, Lesmes

, Macierzanka

, Mackie

, Marze

, McClements

, Ménard

, Recio

, Santos

, Singh

, Vegarud

, Wickham

MSJ

, Weitschies

, Brodkorb

. A standardised static in vitro digestion method suitable for food - an international consensus. Food & Function. 2014;5:1113–24.

13.

Ariza

, Reboredo-Rodríguez

, Cervantes

, Soria

, Martínez-Ferri

, Gonzalez-Barreiro

, Cancho-Grande

, Battino

, Simal-Gándara

. Bioaccesibility and potential bioavailability of phenolic compounds as a new target for strawberry breeding programs. Food Chemistry. 2018;248:155–65.

14.

McDougall

, Dobson

, Smith

, Blake

, Stewart

. Assessing potential bioavailability of raspberry anthocyanins using an in vitro digestion system. Journal of Agricultural and Food Chemistry. 2005;53:5896–904.

15.

Sengul

, Surek

, Nilufer-Erdil

. Investigating the effects of food matrix and food components on bioaccessibility of pomegranate (Punicagranatum) phenolics and anthocyanins using an in-vitro gastrointestinal digestion model. Food Research International. 2014;62:1069–79.

16.

Ariza

, Reboredo-Rodríguez

, Mazzoni

, Forbes-Hernández

, Giampieri

, Afrin

, Gasparrini

, Soria

, Martínez-Ferri

, Battino

, Mezzetti

. Strawberry achenes are an important source of bioactive compounds for human health. International Journal of Molecular Sciences. 2016;17:1103.

17.

Marmon

, Undeland

. Effect of alkaline pH-shift processing on in vitro gastrointestinal digestion of herring (Clupea harengus) fillets. Food Chemistry. 2013;138(1), 214–9.

18.

Martini

, Conte

, Tagliazucchi

. Bioaccessibility, bioactivity and cell metabolism of dark chocolate phenolic compounds after in vitro gastro-intestinal digestion. Journal of Functional Foods. 2018;49:424–36.

19.

Marhuenda

, Alemán

, Gironés-Vilaplana

, Pérez

, Caravaca

, Figueroa

, Mulero

, Zafrilla

. Phenolic Composition, Antioxidant Activity, and In Vitro Availability of Four Different Berries Journal of Chemistry 2016–org/10.1155/2016/5194901.

20.

Anthony

, Fyfe

, Stewart

, McDougall

. Differential effectiveness of berry polyphenols as anti-giardial agents. Parasitology. 2011;138:1110–16.

21.

Singleton

, Orthofer

, Lamuela-Raventós

. Analysis of total phenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent. Methods in Enzymology. 1999;299:152–78.

22.

Dewanto

, Wu

, Adom

, Liu

. Thermal processing enhances the nutritional value of tomatoes by increasing total antioxidant activity. Journal of Agricultural and Food Chemistry. 2002;50:3010–4.

23.

Giusti

, Wrolstad

. Characterization and measurement of anthocyanins by UV-Visible spectroscopy. In: Giusti

, Wrolstad

, editors. Current Protocols in Food Analytical Chemistry John Wiley and Sons, Inc., Hoboken. 2001. (F1.2.1.-F1.2.13)

24.

, Pellegrini

, Proteggente

, Pannala

, Yang

, Rice-Evans

. Antioxidant activity applying an improved ABTS radical cation decolorization assay. Free Radical Biology & Medicine. 1999;26:1231–7.

25.

DeLange

, Glazer

. Bile acids: Antioxidants or enhancers of peroxidation depending on lipid concentration. Archives of Biochemistry and Biophysics. 1990;276:19–25.

26.

Hornbuckle

, Simpson

, Tennant

. Gastrointestinal Function. In: Kaneko

, Harvey

, Bruss

, editors.Clinical Biochemistry of Domestic Animals (Sixth Edition), Elsevier, Boston; 2008. pp. 413–458.

27.

Bermúdez-Soto

, Tomás-Barberán

, García-Conesa

. Stability of polyphenols in chokeberry (Aronia melanocarpa) subjected to in vitro gastric and pancreatic digestion. Food Chemistry. 2007;102:865–74.

28.

Kahlon

, Smith

. In vitro binding of bile acids by blueberries (Vaccinium spp.), plums (Prunus spp.), prunes (Prunus spp.), strawberries (Fragaria X ananassa), cherries (Malpighia punicifolia), cranberries (Vaccinium macrocarpon) and apples (Malus sylvestris). Food Chemistry. 2007;100:1182–7.

29.

Ryan

, Prescott

. Stability of the antioxidant capacity of twenty-five commercially available fruit juices subjected to an in vitro digestion. International Journal of Food Science & Technology. 2010;45:1191–7.

30.

Tavares

, Figueira

, Macedo

, McDougall

, Leitão

, Vieira

HLA

, Stewart

, Alves

, Ferreira

, Santos

. Neuroprotective effect of blackberry (Rubus s) polyphenols is potentiated after simulated gastrointestinal digestion. Food Chemistry. 2012;131:1443–52.

31.

Dai

, Mumper

. Plant phenolics: Extraction, analysis and their antioxidant and anticancer properties. Molecules. 2010;15:7313–52.

32.

Pinelo

, Laurie

, Waterhouse

. A simple method to separate red wine nonpolymeric and polymeric phenols by solid-phase extraction. Journal of Agricultural and Food Chemistry. 2006;54:2839–44.