Antioxidant and Anti-Inflammatory Activity of Citrus Flavanones Mix and Its Stability after In Vitro Simulated Digestion

Recently, several studies have highlighted the role of Citrus flavanones in counteracting oxidative stress and inflammatory response in bowel diseases. The aim of study was to identify the most promising Citrus flavanones by a preliminary antioxidant and anti-inflammatory screening by in vitro cell-free assays, and then to mix the most powerful ones in equimolar ratio in order to investigate a potential synergistic activity. The obtained flavanones mix (FM) was then subjected to in vitro simulated digestion to evaluate the availability of the parent compounds at the intestinal level. Finally, the anti-inflammatory activity was investigated on a Caco-2 cell-based model stimulated with interleukin (IL)-1β. FM showed stronger antioxidant and anti-inflammatory activity with respect to the single flavanones, demonstrating the occurrence of synergistic activity. The LC-DAD-ESI-MS/MS analysis of gastric and duodenal digested FM (DFM) showed that all compounds remained unchanged at the end of digestion. As proof, a superimposable behavior was observed between FM and DFM in the anti-inflammatory assay carried out on Caco-2 cells. Indeed, it was observed that both FM and DFM decreased the IL-6, IL-8, and nitric oxide (NO) release similarly to the reference anti-inflammatory drug dexamethasone.


Introduction
Flavonoids represent one of the most representative classes of plant secondary metabolites, exceeding 8000 compounds, including aglycons, glycosides, and polymers [1] with flavanones, which represents one of the most diffused sub-classes in the Citrus genus [2].

In Vitro Cell-Free Antioxidant and Anti-Inflammatory Screening
In order to select the most promising flavanones, a preliminary antioxidant and antiinflammatory screening by colorimetric in vitro cell-free assays was carried out on the most representative flavanones in the Citrus genus (see Section 2.1). Unless otherwise specified, the absorbance data were recorded by an UV-VIS Spectrophotometer (Shimadzu UV-1601, Kyoto, Japan). Results were expressed as half-maximal inhibitory concentration (IC 50 , µg/mL) with confident limits (C.L.) at 95% calculated by Litchfield and Wilcoxon test, using the PHARM/PCS software version 4 (MCS Consulting, Wynnewood, PA, USA). Data represent the mean ± standard deviation (S.D.) of three independent experiments in triplicate (n = 3). The concentration ranges reported below refer to the final concentrations of the samples or reference compounds in the reaction mixture.

Trolox Equivalent Antioxidant Capacity (TEAC) Assay
TEAC assay was carried out according to Bazzicalupo et al. [25]. The reaction mixture, consisting of 4.3 mM potassium persulfate and 1.7 mM 2,2 -azino-bis (3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) 1:5 v/v, was incubated for 12-16 h in the dark at RT and diluted in deionized water until an absorbance value of 0.7 ± 0.02 at 734 nm. Fifty microliters of each flavanone (2.0-64 µM) or trolox as reference compound (2.50-20 µM) were added to 1 mL of the TEAC reagent and incubated at RT in the dark for 6 min. The absorbance was recorded at 734 nm.
The radical scavenging activity of samples against DPPH was evaluated according to Smeriglio et al. [26]. Briefly, 37.5 µL of each flavanone (3.13-6400 µM) or trolox as reference compound (2.5-20 µM) were added to 1.5 mL of fresh 10 −4 M DPPH methanol solution, mixed for 10 s and incubated for 20 min in the dark at RT. The absorbance was recorded at 517 nm.

Anti-Inflammatory Assays Albumin Denaturation Assay (ADA)
This assay evaluates the inhibitory activity of sample on heat-induced denaturation of bovine serum albumin (BSA) [27]. Briefly, 80 µL of each flavanone (100-6400 µM) or diclofenac sodium as reference compound (4.0-32 µM) were added to 100 µL of 0.4% fatty free BSA solution and 20 µL of phosphate buffer saline (PBS, pH 5.3) into a 96 well-plate. The absorbance was recorded at 595 nm at the starting time (t = 0) and after 30 min (t = 30) at 70 • C by a microplate reader (Multiskan GO, Thermo Scientific, Waltham, MA, USA). A blank consisting of PBS instead of sample was used as negative control. The inhibition percentage of denaturation (ID %) was calculated as follows: where A = sample absorbance (t = 30); B = blank absorbance at (t = 0); C = blank absorbance at (t = 30).
Anti-Protease Activity (APA) The anti-tryptic activity of sample was evaluated according to Smeriglio et al. [27]. Briefly, 200 µL of each flavanone (6.25-140 µM) or diclofenac sodium as reference compound (10-80 µM) were added to 12 µL of trypsin (10 µg/mL) and 188 µL of 25 mM Tris-HCl buffer (pH 7.5). After that, 200 µL of 0.8% casein solution were added and samples incubated for 20 min at 37 • C in a water bath. After that, 400 µL of perchloric acid were added to stop the reaction. The cloudy suspension, due to the protein precipitation, was centrifuged at 3500× g for 10 min and the absorbance of the supernatant was recorded at 280 nm against a blank consisting of deionized water instead of sample.

Flavanones Mix Preparation
Based on the results of the preliminary antioxidant and anti-inflammatory screening (see Section 2.2), the most powerful flavanones (neohesperidin, hesperidin, neoeriocitrin, eriocitrin, and hesperetin) were selected in order to prepare a flavanones mix (FM), which was, at first instance, investigated by the same in vitro colorimetric assays (see Section 2.2) in order to evaluate its potency with respect to the single flavanones and to elucidate potential synergistic mechanisms. The following final concentration range of FM were used: FRAP (2.50-20 µM), ORAC (0.08-0.60 µM), TEAC (0.31-2.50 µM), DPPH (1.25-10 µM), ADA (5.0-40 µM), and APA (2.50-20 µM). After that, being the results of the preliminary antioxidant and anti-inflammatory screening of FM very promising, an appropriate concentration was chosen to proceed with the subsequent analyses. In particular, the FM was prepared on the assumption that digested sample, which would then be applied to the Caco-2 cells for the anti-inflammatory experiments, had to contain an equimolar concentration of the five most powerful flavanones (neohesperidin, hesperidin, neoeriocitrin, eriocitrin, and hesperetin) equal to 10 µM, which represents the mean efficacy concentration taking into account the IC 50 values obtained in the preliminary antioxidant and anti-inflammatory screening of the FM. At this purpose, stock solutions (14 mM) of flavanones in DMSO were mixed and diluted 10-fold with Milli-Q water to obtain the 1.4 mM FM, which was used to carried out the in vitro simulated human digestion.

In Vitro Simulated Human Digestion
In vitro gastric and duodenal digestion of FM was carried out according to Trombetta et al. [28] with some modifications. FM solution (1.5 mL) (see Section 2.3), has been dissolved in 7.5 mL of simulated gastric solution containing 0.127 mM egg-phosphatidylcholine and adjusting the pH to 2.5, by 1 M HCl. Gastric pepsin and lipase (9000 U/mL and 60 U/mL, respectively) were added to the mixture starting the gastric digestion phase, which was protracted for 2 h at 37 • C, incubating under continuous stirring (170 rpm) by an Innova 4000 Benchtop Incubator Shaker (New Brunswick Scientific, Edison, NJ, USA). The gastric phase was stopped by the addition of 1M NaOH in order to reach a pH of 7.5, and 6 mL of the previous solution passed to the duodenal digestion phase. Duodenal solution was prepared by adding 2.10 mL of simulated bile solution (6.5 mM phosphatidylcholine, 4 mM cholesterol, 12.5 mM sodium taurocholate, and 12.5 mM sodium glycodeoxycholate) to 5.9 mL of simulated pancreatic juice containing pancreatic lipase (590 U/mL), colipase (3.2 µg/mL), trypsin (11 U/mL), α-chymotrypsin (24 U/mL), and α-amylase (300 U/mL). The duodenal mixture was incubated for 4 h under continuous stirring at 37 • C as described above. At the end, both gastric and duodenal digesta were immediately stored at −80 • C until the subsequent analyses.
For MS analysis, an ion trap fitted with an electrospray ionization (ESI) source operating in negative ionization mode, was used. Nitrogen (dry gas) flow rate and pressure were set at 9 L/min and 40 psi, respectively, whereas the nebulizer temperature was set at 350 • C. Helium was used as collision gas (1.46 × 10 −5 bar) and collision-induced dissociation spectra were obtained with a fragmentation amplitude of 1.0 V (MS/MS). Data acquisition was performed using the ChemStation A.10.01 software (Agilent technologies, Santa Clara, CA, USA). The identification of flavanones, as well as of any metabolites, was carried out by comparison of retention time, UV-Vis and mass spectra of the analytes with reference compounds where possible or by comparison of UV-VIS and mass spectra with data reported in literature and mass spectra libraries. Quantification was carried out by built external standard calibration curves of each flavanone (0.612-10 µg/mL) and expressing the results as average ± standard deviation (S.D.) of three independent experiments in triplicate (n = 3).
The HPLC method was validated according to the current international guidelines [29], regarding selectivity, linearity, precision, robustness, limit of detection (LOD), limit of quantitation (LOQ) and recovery. LOD and LOQ values were calculated following the approach based on the standard deviation of the response and the slope of the calibration curves [29].
2.6. Anti-Inflammatory Activity on In Vitro Cell-Based Model 2.6.1. Cell Model Experiments were carried out on transwell models (CacoReady™, Readycell, Barcelona, Spain) consisting of Caco-2 cells (8.5 × 104 cells/cm 2 , passage number 41-58) seeded on polyester microporous filters in 24-well HTS plates (6.5 mm diameter, 0.33 cm 2 area and 0.4 µm of pore size) (Corning Incorporated, Corning, NY, USA). Completed Dulbecco's Modified Eagle Medium (DMEM), prepared according to Denaro et al. [30], was added on the apical (0.3 mL) and basolateral side (0.9 mL) of the Transwell ® insert. After 21 days of culture, Caco-2 cells were completely differentiated and polarized, such that their phenotype resembled the morphological and functional features of mature enterocytes lining the small intestine.
Samples were collected and stored at −80 • C until the subsequent analyses. Postquality control assays, such as TEER, as well as the apparent permeability coefficient (Papp) and paracellular flux (Pf) of Lucifer yellow (LY) were assessed in order to evaluate the Caco-2 monolayer integrity [30].

Statistical Analysis
Three independent experiments in triplicate (n = 3) were carried out for each in vitro cell-free and cell-based assay. Results were expressed as mean ± standard deviation (S.D.). Data were analyzed by one-way analysis of variance (ANOVA) followed by Tukey's test by SigmaPlot12 (Systat Software, Inc., San Jose, CA, USA). Results were statistically significant for p ≤ 0.05.

Antioxidant and Anti-Inflammatory Screening of Flavanones and Flavanones Mix
With the aim of identifying the most promising flavanones from an antioxidant and anti-inflammatory point of view, nine of the most representative compounds of the Citrus genus were chosen: neoeriocitrin, eriocitrin, hesperetin, hesperidin, neohesperidin, diosmin, neodiosmin, naringin, and tangeretin. Among these, after a preliminary antioxidant and anti-inflammatory screening by in vitro cell-free assays, the five most promising ones were selected ( Figure 1) and suitably mixed in equimolar ratio (flavanones mix) in order to elucidate possible synergistic effects. xidants 2021, 10, x FOR PEER REVIEW 7 of 18 The antioxidant screening was carried out using four different colorimetric tests (ORAC, FRAP, TEAC, and DPPH) characterized by different environments and reaction mechanisms, as well as by different charged radicals. Results were depicted in Figure 2, which reports the IC50 of the five strongest flavanones (neohesperidin, hesperidin, neoeriocitrin, eriocitrin, and hesperetin) in comparison with the reference standard trolox. The antioxidant screening was carried out using four different colorimetric tests (ORAC, FRAP, TEAC, and DPPH) characterized by different environments and reaction mechanisms, as well as by different charged radicals. Results were depicted in Figure 2, which reports the IC 50 of the five strongest flavanones (neohesperidin, hesperidin, neoeriocitrin, eriocitrin, and hesperetin) in comparison with the reference standard trolox. All investigated flavanones showed the same behavior, showing a greater antioxidant activity in the ORAC test followed by TEAC, FRAP, and DPPH, with the exception of eriocitrin and neoeriocitrin, which showed instead the following order of potency: ORAC > DPPH > TEAC > FRAP. These results shed light on the importance of studying the antioxidant activity with different tests, because the response of a molecule can vary substantially depending on its structural features. Among the strongest molecules, surely stood out hesperidin, hesperetin, and neohesperidin in the TEAC and ORAC tests, and eriocitrin and neoeriocitrin in the FRAP and DPPH tests, which showed different behavior, depending on the radical charge and test reaction mechanism (based on transfer of All investigated flavanones showed the same behavior, showing a greater antioxidant activity in the ORAC test followed by TEAC, FRAP, and DPPH, with the exception of eriocitrin and neoeriocitrin, which showed instead the following order of potency: ORAC > DPPH > TEAC > FRAP. These results shed light on the importance of studying the antioxidant activity with different tests, because the response of a molecule can vary substantially depending on its structural features. Among the strongest molecules, surely stood out hesperidin, hesperetin, and neohesperidin in the TEAC and ORAC tests, and eriocitrin and neoeriocitrin in the FRAP and DPPH tests, which showed different behavior, depending on the radical charge and test reaction mechanism (based on transfer of hydrogen atoms for TEAC and ORAC, and on electron transfer for FRAP and DPPH). The other flavanones taking into account at the beginning of study, were discarded because they were less active with respect to the five selected flavanones, with IC 50 values ≥2 times in ORAC and TEAC assays and IC 50 ≥ 4 times in FRAP and DPPH assays (data not shown). The antioxidant activity of the five selected flavanones resulted statistically significant (p < 0.001) with respect to the trolox in all of the tests carried out, with the exception of TEAC, in which hesperidin, hesperetin and neohesperidin showed a behavior comparable with the reference standard ( Figure 2). Furthermore, it is interesting to note that all five tested flavanones were significantly (p < 0.001) stronger with respect to the trolox in the ORAC assay.
In the last panel of Figure 2, it is possible to observe the behavior of the flavanones mix (FM). By averaging the IC 50 value of the five flavanones in each test, it is easy to see how the FM always showed the greater antioxidant activity (17 times stronger in the DPPH, 8.6 times stronger in the FRAP, 7.8 times stronger in the TEAC, and 1.4 times stronger in the ORAC) compared to the single flavanones, demonstrating that there is effectively a synergistic activity between these molecules, which enhances their intrinsic antioxidant activity. The IC 50 of the FM is, in fact, always statistically significant (p < 0.001) in comparison with that of flavanones in all of the tests carried out (p < 0.001, data not shown). Moreover, an even more interesting aspect is that the FM IC 50 was always significantly lower (p < 0.001) in comparison with the reference standard (trolox).
These preliminary results were corroborated by the evaluation of the anti-inflammatory activity, investigated by the protease inhibition assay and the heat-induced denaturation of BSA. In Figure 3a, it is possible to observe and compare the IC 50 of each flavanone and FM, with respect to the reference anti-inflammatory drug diclofenac sodium in the protease inhibition test.
idants 2021, 10, x FOR PEER REVIEW 9 of 18 hydrogen atoms for TEAC and ORAC, and on electron transfer for FRAP and DPPH). The other flavanones taking into account at the beginning of study, were discarded because they were less active with respect to the five selected flavanones, with IC50 values ≥ 2 times in ORAC and TEAC assays and IC50 ≥ 4 times in FRAP and DPPH assays (data not shown). The antioxidant activity of the five selected flavanones resulted statistically significant (p < 0.001) with respect to the trolox in all of the tests carried out, with the exception of TEAC, in which hesperidin, hesperetin and neohesperidin showed a behavior comparable with the reference standard ( Figure 2). Furthermore, it is interesting to note that all five tested flavanones were significantly (p < 0.001) stronger with respect to the trolox in the ORAC assay.
In the last panel of Figure 2, it is possible to observe the behavior of the flavanones mix (FM). By averaging the IC50 value of the five flavanones in each test, it is easy to see how the FM always showed the greater antioxidant activity (17 times stronger in the DPPH, 8.6 times stronger in the FRAP, 7.8 times stronger in the TEAC, and 1.4 times stronger in the ORAC) compared to the single flavanones, demonstrating that there is effectively a synergistic activity between these molecules, which enhances their intrinsic antioxidant activity. The IC50 of the FM is, in fact, always statistically significant (p < 0.001) in comparison with that of flavanones in all of the tests carried out (p < 0.001, data not shown). Moreover, an even more interesting aspect is that the FM IC50 was always significantly lower (p < 0.001) in comparison with the reference standard (trolox).
These preliminary results were corroborated by the evaluation of the anti-inflammatory activity, investigated by the protease inhibition assay and the heat-induced denaturation of BSA. In Figure 3a, it is possible to observe and compare the IC50 of each flavanone and FM, with respect to the reference anti-inflammatory drug diclofenac sodium in the protease inhibition test. FM and hesperetin showed a statistically significant activity (p < 0.001) in comparison with diclofenac sodium. In particular, FM showed the strongest anti-inflammatory activity (about 2 times greater with respect to the reference compound), whereas hesperetin showed the lowest anti-inflammatory activity (about 4 times less) than the other tested molecules, which instead showed a comparable behavior, with respect to the reference FM and hesperetin showed a statistically significant activity (p < 0.001) in comparison with diclofenac sodium. In particular, FM showed the strongest anti-inflammatory activity (about 2 times greater with respect to the reference compound), whereas hesperetin showed the lowest anti-inflammatory activity (about 4 times less) than the other tested molecules, which instead showed a comparable behavior, with respect to the reference standard. In regards to the inhibition of the BSA denaturation, FM showed a significantly higher activity (p < 0.001) in comparison with the single flavanones (about 38 times stronger, considering an average of the IC 50 of the single compounds) and a significantly lower activity with respect to the diclofenac sodium (approximately 1.3 times).
Therefore, in accordance with the previous results, it is possible to postulate a clear synergistic mechanism of FM, capable of significantly enhancing the intrinsic anti-inflammatory activity of the individual molecules.

Pre-and Post-Digestion Analysis of Flavanones Mix
The quali-quantitative determination of flavanones in FM and gastric and duodenal DFM, was carried out by a LC-DAD-ESI-MS/MS analysis. The analytical method was validated and, as observed from the data, reported in Table 1, the method resulted as sensitive, precise, accurate, and repeatable, with recovery values ≥90.36%. Moreover, according to the current international guidelines [29], the robustness of the method was evaluated, taking into account the following parameters: pH variation of mobile phase, variations in mobile phase composition, different lots, and/or suppliers of column, temperature, and flow rate. Maintaining the conditions reported in Section 2.5, the analytical method results robust, because it is reliable, with respect to deliberate variations in the method parameters mentioned above (data not shown). Representative chromatograms acquired at 292 nm of the FM (black), as well as of the gastric (blue) and duodenal (red) DFM, are reported in Figure 4.
No interferences, such as degradation products, metabolites, or co-eluting compounds, were recorded. Moreover, the chromatographic separation of the constituents of the FM did not show any overlap or interferences from matrix constituents in the digested samples at the retention time of the five flavanones, which appeared well-separated and easy identifiable. No metabolites were identified in the gastric digested samples or in the duodenal digested samples, as observed in Figure 4 and Table 2, in which the five flavanones are identified and quantified, with respect to the starting FM (10 µM).
Indeed, only a mild decrease ≤10.8% in gastric digested samples followed by a further decrease ≤1.6% in duodenal digested samples was observed, recording a total flavanones loss ≤12.4%. This loss is certainly attributable to the sample extraction process, as it is possible to observe from the recovery data reported in Table 1. nts 2021, 10, x FOR PEER REVIEW 11 of 18

Anti-Inflammatory Activity of Flavanones Mix on Caco-2 Cells
Once FM was identified as the most promising formulation from an antioxidant and anti-inflammatory point of view, an anti-inflammatory activity study was conducted on a Caco-2 cell-based model, to evaluate the potential applicability of this FM as an anti-inflammatory agent, and plan further (and more in-depth) in vivo studies on animal models.
As can be seen from Figure 5, the inflammation triggered by 25 ng/mL IL-1β allowed a statistically significant (p < 0.001) increase in IL-6 (~7 folds vs. CTR − ), IL-8 (~27 folds vs. CTR − ) as well as in NO release (~2.5 folds vs. CTR − ), well-known inflammatory mediators. Indeed, only a mild decrease ≤ 10.8% in gastric digested samples followed by a further decrease ≤ 1.6% in duodenal digested samples was observed, recording a total flavanones loss ≤ 12.4%. This loss is certainly attributable to the sample extraction process, as it is possible to observe from the recovery data reported in Table 1.

Anti-Inflammatory Activity of Flavanones Mix on Caco-2 Cells
Once FM was identified as the most promising formulation from an antioxidant and anti-inflammatory point of view, an anti-inflammatory activity study was conducted on a Caco-2 cell-based model, to evaluate the potential applicability of this FM as an antiinflammatory agent, and plan further (and more in-depth) in vivo studies on animal models.
As can be seen from Figure 5, the inflammation triggered by 25 ng/mL IL-1β allowed a statistically significant (p < 0.001) increase in IL-6 (~7 folds vs. CTR − ), IL-8 (~27 folds vs. CTR − ) as well as in NO release (~2.5 folds vs. CTR − ), well-known inflammatory mediators. Figure 5. Evaluation of the anti-inflammatory activity of flavanones mix (FM, 10 µM) and gastric plus duodenal digested flavanones mix (DFM, 10 µM) in comparison with the reference anti-inflammatory drug dexamethasone (Dex, 10 µM) by a cell-based assay carried out on Caco-2 transwell model. IL-1β (25 ng/mL) was used to trigger the inflammation (CTR + ), whereas DMEM with 0.1% DMSO was used as negative control (CTR − ). Results were expressed as the concentration (pg/mL for IL-6 and IL-8, and µM for NO) of inflammatory markers released after cell treatments ± S.D. of Figure 5. Evaluation of the anti-inflammatory activity of flavanones mix (FM, 10 µM) and gastric plus duodenal digested flavanones mix (DFM, 10 µM) in comparison with the reference anti-inflammatory drug dexamethasone (Dex, 10 µM) by a cell-based assay carried out on Caco-2 transwell model. IL-1β (25 ng/mL) was used to trigger the inflammation (CTR + ), whereas DMEM with 0.1% DMSO was used as negative control (CTR − ). Results were expressed as the concentration (pg/mL for IL-6 and IL-8, and µM for NO) of inflammatory markers released after cell treatments ± S.D. of three independent experiments in triplicate (n = 3); * p < 0.001 vs. CTR + ; § p < 0.001 vs. Dex 10 µM/IL-1β 25 ng/mL. FM, as well as the digested flavanones mix (DFM) and dexamethasone (Dex), all at the same concentration (10 µM), did not show any pro-inflammatory activity, so much so that they did not show any statistically significant difference between them, and between them and CTR − , for all the three inflammatory markers taking into account.

Discussion
The bioavailability of flavanones has not been fully clarified, yet several studies have focused on the metabolic patterns of these compounds, trying to understand if the observed biological activities are due, at least in part, to the parent compounds. Evidence suggests that they may be metabolized by the intestinal microbiome, resulting in the formation of aglycones and smaller phenolics, which affect the bioavailability of the parent compounds [10]. However, what emerges from the literature is the lack of studies that evaluate the bioaccessibility of these compounds, which is the fraction of compounds available for the absorption in the small intestine. This aspect becomes even more important when it is necessary to evaluate the anti-inflammatory activity of a compound or formulation at intestinal level. Indeed, since these compounds must exert their anti-inflammatory activity locally, it is important to know what is the real availability of these compounds in the intestinal compartment after the digestive processes. Despite several studies investigating the antioxidant and anti-inflammatory activity of some Citrus flavanones, such as hesperidin, hesperetin, and naringenin [35][36][37][38] are available, this is the first one that carried out an antioxidant and anti-inflammatory screening of the most representative flavanones of the Citrus genus. Moreover, for the first time, the most promising molecules have been selected and mixed in equimolar ratio to investigate the interactions between these compounds and the potential occurrence of synergistic effects, which are often observed with Citrus extracts or juices containing different molecules belonging to the same or different classes of flavonoids [26,27], as well as for other medicinal plants, such as Opuntia ficus indica L., Cannabis sativa L. and Echinacea purpurea (L.) Moench [23,39,40].
The investigated FM (hesperetin, hesperidin, neohesperidin, eriocitrin, and neoeriocitrin) was then subjected to simulated digestion in vitro to mimic what happens in vivo, and to verify what occurs to these compounds once ingested along the gastrointestinal tract. The LC-DAD-ESI-MS/MS analysis of gastric and duodenal DFM showed that any compound has been altered during gastric digestion, and this condition remains unchanged in duodenal digestion too. This stability was previously described in vitro and in vivo for some of the flavanones described in the present study (neohesperidin, hesperidin, and hesperetin) as most representative compounds of a bitter orange extract [41]. Results showed that these flavanones were stable after simulated in vitro digestion. Moreover, they were detected both in the small intestine and in the colon of rats, even 12 and 24 h after oral administration of the bitter orange extract, with a maximum recovery percentage of 72% with respect to the ingested dose [41]. These results corroborate that dietary flavanones can remain unaltered for several hours in the colon where they may be absorbed and/or interact with the intestinal mucosa exerting the local anti-inflammatory activity.
A possible explanation for this particular behavior lies in the peculiar metabolisms of rutinoside and neohesperidoside derivatives, which usually arrive unchanged in the distal part of the intestine, where they are hydrolyzed by colonic microflora enzymes and then absorbed [42]. Considering this, the results obtained in the present study are mainly attributable to the structure of the five selected flavanones: two rutinosides (hesperidin and eriocitrin), two neohesperidosides (neohesperidin and neoeriocitrin), and one aglycon (hesperetin), which not only showed the strongest antioxidant and anti-inflammatory activities, with respect to other Citrus flavanones tested, but they have maintained their structure. Consequently, their activities were unchanged, even after being subjected to simulated digestion. In fact, no chemical or enzymatic degradation was observed after gastric and duodenal digestion, indirectly demonstrating how the presence of these sugars shields the molecule from possible enzymatic attacks or hydrolytic phenomena due to pH changes, which happens, on the contrary for glucose conjugates. It has been demonstrated that glucose conjugates, but not conjugates with other sugars, such as rhamnose and rutinose, are rapidly hydrolyzed in the oral cavity for the presence of β-glucosidases, enzymes expressed abundantly at the intestinal level [43].
Several anti-inflammatory studies have been carried out to understand how flavanones modulate inflammatory response. Results obtained from the investigation on IL-1β stimulated Caco-2 cells showed that FM mix and DFM were able to decrease the release of the inflammatory markers IL-6, IL-8, and NO already at "pharma-nutritional" doses. Indeed, the FM and DFM concentration used (10 µM) is a concentration easily reachable even by simply consuming Citrus fruits or juices [3]. DFM showed, at times, a stronger anti-inflammatory activity than the FM mix, although the LC-DAD-ESI-MS/MS analysis did not show any presence of metabolites or degradation products. This finding could be related to the presence of several enzymes in DFM, which exert their own anti-inflammatory activity. It has been demonstrated that proteolytic enzymes, such as chymotrypsin and trypsin, possess anti-inflammatory activities, and exhibit synergistic effects with anti-inflammatory drugs [44].
The flavanones ability to modulate inflammation depends on different factors. One of these is a consequence of the antioxidant potency, which is strongly related to their structure, such as presence of sugar moieties, number of hydroxyl groups, and their spatial arrangement [3]. Equally important in inflammatory modulation is the inhibitory activity of these compounds on key enzymes responsible for activation and transduction of inflammatory stimuli; for example, hesperidin was able to inhibit mitogen-activated protein kinases (MAPKs) and phosphodiesterases [45]. The MAPKs role in intracellular signaling pathway during pro-inflammatory response is closely related to the NF-kb signaling pathway, which exerts the pivotal role in the expression of iNOS, COX-2, IL-6, and TNF-α [46]. It was demonstrated that flavanones are able to downregulate NF-Kb [47], but the differences in the experimental models can affect the outcomes, which can appear discordant, and needs to be separately interpreted [46]. Recently, Guazelli et al. [48] corroborated the previous observations by investigation of the antioxidant and anti-inflammatory effects of hesperidin methyl chalcone on an animal model of ulcerative colitis. The authors highlighted, in particular, as this methylated derivative of hesperidin can play a pivotal role in increasing the antioxidant defense system, counteracting the inflammatory state by decreasing the pro-inflammatory cytokines release (TNF-α, IL-6, IL-1β, and IL-33) in the colon via NF-Kb, by improving, significantly, the macro and microscopic damages induced by intracolonic administration of acetic acid in rats.
However, despite the abundance of studies on flavanones, their anti-inflammatory mechanisms of action remain unclear, and further studies are needed to better clarify their roles in modulating the inflammatory cascade.
The results of the present study open new perspectives for the possible clinical use of this FM for inflammatory intestinal diseases, from the mild forms, to the more severe, such as ulcerative colitis and Crohn's disease, which are becoming increasingly widespread.
The therapeutic strategies available today mainly focus on the use of corticosteroids and immunosuppressive agents, aim to achieve and maintain remission, and prevent disease progression. However, these drugs have several side effects, and some patients quickly become refractory to the treatment (in some cases requiring surgery) [49]. From this point of view, new nutraceutical formulations, which have efficacy equal to synthetic drugs, but with fewer or no side effects, represent the therapeutic strategy of the future, both alone and in co-treatment with conventional drugs. Moreover, it has been demonstrated that naringin, a flavanone widespread in the Citrus genus, could be a useful agent to counteract bone loss due to steroid treatment of inflammatory intestinal diseases [50], opening new perspectives in the treatment of adverse effects of conventional drugs.
Despite this being an innovative study, because it investigates for the first time, to our knowledge, the synergistic antioxidant and anti-inflammatory activity of a Citrus flavanones mix on in vitro cell-free and cell-based models, and its stability after simulated gastrointestinal digestion, this is a preliminary study, and, furthermore, it shows several limitations. First, only the IL-1β-induced inflammation was used. Second, the in vitro intestinal model used reproduces only partially the intestinal environment lacking the microbiome. Third, we did not investigate the molecular mechanism underlying intestinal inflammatory activity. Considering this, further in vitro cell-based co-culture studies, as well as investigation on specific IBD animal models, are needed to verify the transability of these results on humans, to confirm the postulated clinical applications.

Conclusions
In conclusion, the present study elucidates (for the first time) that eriocitrin, neoeriocitrin, hesperidin, neohesperidin, and hesperetin are the most powerful antioxidant and anti-inflammatory compounds among the investigated Citrus flavanones, and that the particular structural features of these molecules shield them from possible enzymatic attacks or hydrolytic phenomena during simulated digestion, making them available at the intestinal level, where they can exert their remarkable anti-inflammatory activities. Moreover, this study demonstrates (for the first time) that mixing different flavanones enhances the antioxidant and anti-inflammatory activities of individual compounds, experimentally demonstrating a synergistic effect. The flavanones mix exerts already at a pharma-nutritional dose, an anti-inflammatory activity comparable or higher with respect to the reference drugs diclofenac sodium and dexamethasone.
However, this is a preliminary study based on in vitro cell-free and cell-based models; therefore, the translation of results to the complex in vivo scenario is quite difficult. Considering this, further in vitro cell-based studies, as well as animal and human studies, should be performed in order to investigate, in depth, the anti-inflammatory properties of this flavanones mix, as well as the molecular mechanisms and cellular targets involved, which could justify the potential role of this formulation as a possible alternative strategy to counteract the inflammatory intestinal diseases.