Anti-Inflammatory, Antioxidant Activities, and Phytochemical Characterization of Edible Plants Exerting Synergistic Effects in Human Gastric Epithelial Cells

Dietary bioactive compounds from natural sources (e.g., herbal medicines, foods) are known to potentially suppress acute or chronic inflammation and promote the effectiveness of treatment to reduce the harmful effects of gastritis alone or in combination. In this regard, we have characterized four Cameroonian spice extracts, namely Aframomum citratum, Dichrostachys glomerata, Tetrapleura tetraptera, and Xylopia parviflora through reverse phase-high-performance liquid chromatography (RP-HPLC), ultra-performance liquid chromatography-electrospray ionization high-resolution mass spectrometry (UPLC-ESI-HRMS/MS), and Fourier transform infrared spectroscopic (FTIR) analyses and investigated their antioxidant and synergistic anti-inflammatory activities in human gastric adenocarcinoma (AGS) and gastric epithelial (GES-1) cells. The extracts showed a high amount of total phenolic (TPC: 150–290 mg gallic acid equivalents (GAE)/g of extract) and flavonoid content (TFC: 35–115 mg catechin equivalents (CE)/g of extract) with antioxidant properties in a cell-free system (1,1-Diphenyl-2-picryl-hydrazyl (DPPH) half maximal inhibitory concentration (IC50s) ≤ 45 µg/mL; 2,2′-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) half maximal inhibitory concentration (IC50s) ≤ 29 µg/mL. The extracts in combination (MIX) exert a synergistic beneficial effect (combination index (CIs) < 1 and dose reduction index (DRIs) > 1) on inflammatory markers (interleukin (IL)-8 and -6 release, and nuclear factor kappa B (NF-κB) driven transcription) in human gastric epithelial cells, which may result from the presence of phenolic compounds (phenolic acids, flavonoids) or other compounds (protein, lipid, aromatic, and polysaccharide compounds) tentatively identified in the extracts. The general findings of the present study provide supporting evidence on the chemical composition of four Cameroonian dietary plants and their significant synergistic inhibitory activities on inflammatory markers of gastric epithelial cells.


Introduction
Gastric ulcers and gastritis are among the most common gastrointestinal disorders, with a strong relationship with Helicobacter pylori (H. pylori) infection and nonsteroidal . Fetal bovine serum (FBS) and disposable materials for cell culture were purchased by Euroclone (Euroclone S.p.A., Pero-Milan, Italy). All reagents were at the highest grade available.

Sample Preparation
Xylopia parviflora Spruce, Tetrapleura tetraptera (Schum. and Thonn.) Taub, Dichrostachys glomerata (Forssk.) Chiov., and Aframomum citratum (C.Pereira) K.Schum were harvested as previously described by Nwakiban et al. [14,18]. The plant material was cleaned, taken to dryness, and stored at room temperature. The air-dried and powdered samples (100 g) were subjected to magnetic stirring with 100 mL of a hydroalcoholic (ethanol: water, 70:30) mixture for 4 and 6 hrs at room temperature, in dark conditions. The solvent was removed through a rotary evaporator (Laborota 4000 efficient, Heidolph Instruments, Schwabach, Germany), and subjected to lyophilization. The lyophilized extracts were maintained at 4 • C, and 10 mg were freshly dissolved in 1 mL of HPLC grade or MS grade methanol for chemical characterization. Each extract and its combination (MIX) (mixed in the same proportion) were dissolved in pure dimethyl sulfoxide (DMSO), aliquoted, and stored at −80 • C. For cell treatment and anti-inflammatory assays under sterile conditions, the final concentration of DMSO added to the cells was not above 0.1%.

Quantification of Phenolic Compounds through HPLC-PDA
The phenolic compounds content within each extract was analyzed as described above [17] by HPLC (Nexera X-2 LC-30A, Shimadzu, Japan), and chromatography separations were carried out on a Chromasol™ RP-C18 column (250 mm × 4.6 mm, 5 µm). The mobile phase was made up of water set at a pH of 2.65 with acetic acid (solvent A) and solvent B (20% solvent A and 80% acetonitrile). The extracts were dissolved in HPLC-grade methanol and filtered through a syringe filter (0.45 µm PVDF, Millipore, MA, USA). A 20 µL volume for each standard or sample was injected into the HPLC system, and a linear elution gradient was applied in the following manner: 0-20% B in 0-35 min, 20-50% B in 35-40 min, 50-100% B in 40-45 min, and 100-0% B in 45-60 min. The temperature of the column was kept at 20 • C. The flow rate was 1.2 mL/min, and the PDA detector was adjusted to λ 1 = 280 and λ 2 = 320 simultaneously using LabSolutions software (version 6.50, Shimadzu, Japan). The identified phenolic compounds were quantified using the peak area obtained from the standards, comparing their retention times with those of corresponding standards and by spiking samples with appropriate standards.

UPLC-ESI-HRMS/MS Analysis
The extracts were diluted with MS-grade methanol and injected directly into the UPLC-ESI-HRMS/MS system with an ekspert 110 binary pump, an ekspert 110-XL autosampler, an ekspert PDA detector, and ekspert 110 column compartment (AB Sciex Instruments, Netherlands) via a syringe pump. The Kinetex C18 100A (30 × 2.1 mm 1.7 µm) Column (Phenomenex) was used for separation at a flow rate of 0.4 mL/min. The mobile phase was made up of a combination of two solvents: A (0.1% acetic acid in water) and B (0.1% acetic acid in acetonitrile and methanol in the ratio 8:2) [17]. The mass spectra analyses were performed in negative ion mode in the m/z range from 100 to 2000 m/z at an IRDx resolution of 15,000 using an LC/MS/MS Quadrupole-TOF Hybrid Mass Spectrometer (Sciex Triple ToF 5600, Singapore) and under the following conditions: gases were GS1-45, GS2-60, and Curtain GAS (CUR)-40, source voltage was 4.0 kV, Duospray ion source was set at 400 • C, and the collision energy was 10 V. CID-MS/MS experiments were carried out on mass-selected precursor ions using standard isolation and excitation configuration. All data acquisition and analysis were performed using the Peak View 2.1 Software (AB SCIEX Triple TOF 5600, Singapore), which has MasterView™ (Version 1.0, AB SCIEX). The Master View's XIC management tool was used to detect quasi-molecular weights, mass errors, and isotope patterns of both targeted and non-targeted compounds. ChemSpider, elemental composition analysis, as well as literature review, were used to define consistent tentative structures of the identified compounds.

Fourier Transform Infrared Spectroscopic (FTIR) Analysis
Dried powder from various extracts of each plant material was used for FTIR analysis. Approximately 10 mg of each extract were loaded into the FTIR spectroscope (Bruker, Germany), which is connected to a computer operating on a Windows system and to an OPUS software (Version 7.0 Bruker Optic). The spectra were recorded at a resolution of 4 cm −1 from 4000 to 400 cm −1 using 64 co-added scans. All spectra were subtracted against a background of air spectra. The attenuated total reflectance (ATR) plate was carefully cleaned by scrubbing with isopropyl 70% twice, then drying with soft tissue before being filled with the following sample. The spectra were recorded as absorbance values at each data point in triplicate and two times for spectrum confirmation.

Estimation of Polyphenols and Antioxidant Activity Assays
The polyphenol estimation and antioxidant capacity assays were modified and translated into 96-well plates based on the methods used in previous literature reports [19][20][21]. The stock solution (100 µg/mL) of each extract was prepared for experimental analysis and the data was measured using a UV-VIS spectrophotometer (Shimadzu, Model: UV 2100, Kyoto, Japan). All tests were run in triplicate at least three times.

Total Phenolic Content Assay
Total phenolic content (TPC) was determined using a Folin-Ciocalteu assay with slight modifications [22]. Briefly, a 25 µL sample was mixed with 25 µL Folin's reagent (1:3 diluted with water), and 200 µL deionized water was added, then the mixture was incubated at room temperature for 5 min. The reaction mixture was then alkalized by adding 25 µL of sodium carbonate (10%, w/v) and allowed to stand in the dark at 25 • C for 60 min. The absorbance was read at 765 nm against the reagent blank. Gallic acid was used as a standard in a concentration range between 0 and 50 µg/L. TPC was expressed in mg gallic acid equivalents (GAE) per g of dried extracts based on the calibration curve.

Total Flavonoid Content Assay
Total flavonoid content (TFC) was determined by the colorimetric method as described by Zhishen et al. [23] and slightly modified by Moukette et al. [24]. To a 15 µL sample or standard, 45 µL of distilled water, followed by 4.5 µL of sodium acetate (5% solution) were added. The mixture was left in the darkness at 25 • C for 5 min and 4.5 µL of aluminum chloride (10% of Al 2 Cl 3 ) was added. Afterward, 30 µL of 1 mM NaOH and 150 µL distilled water were added to the reaction mixture. The absorbance of the resulting solution was measured at 765 nm wavelength against the reagent blank. Catechin was used as a standard in a concentration range between 0 and 60 µg/L. TFC was expressed in mg catechin equivalents (CE) per g of dried extracts based on the calibration curve. The total flavonol assay (FC) was adapted from [25,26] with quercetin as the standard compound. Briefly, 80 µL of the samples were mixed with 80 µL of aluminum chloride (2% of Al 2 Cl 3 ) diluted in ethanol and 120 µL of 50 g/L sodium acetate solution. The mixture was incubated at 25 • C for 2.5 hrs and the absorbance was measured at 440 nm. Concentrations (0-60 µg/mL) of quercetin dissolved in ethanol were used to draw the standard curve. The results were expressed as mg quercetin equivalents (QE) per g of dried extracts based on the calibration curve.
The DPPH scavenging activity of extracts was determined based on the method of Sogi et al. [27] and modified from the method reported by Peng et al. [20]. Samples were dissolved in methanol and tested at concentrations between 1 and 100 µg/mL. An aliquot of a 40 µL sample was mixed with 260 µL of 0.1 mM DPPH radical methanolic solution in a 96-well plate and incubated for 30 min at 25 • C. Afterward, the absorbance was measured at 517 nm against the reagent blank. Concentrations ranging from 0 to 50 µg/mL ascorbic acid dissolved in water were used to draw the standard curve. The results were expressed as mg ascorbic acid equivalents (AAE) per g of dried extracts and the IC 50 concentration showing 50% radical scavenging activity was determined.

2,2 -Azinobis-(3-ethylbenzothiazoline-6-sulfonic Acid) (ABTS) Assay
The ABTS antioxidant activity of spice extracts was carried out using the ABTS + radical cation decolorization assay with some modifications [28]. To 5 mL of 7 mmol/L of ABTS solution, 88 µL of a 140 mM potassium persulfate solution was added to produce ABTS + . The mixture was allowed to stand in the dark at 25 • C for 16 hrs, then 0.5 mL of the ABTS + solution was diluted by adding 45 mL analytical grade ethanol to obtain an initial absorbance of 0.70 ± 0.02 at 734 nm. A 10 µL sample extract and 290 µL prepared ABTS + solution were mixed in a 96-well plate and incubated at 25 • C for 6 min in the darkness. The absorbance was measured at 734 nm against the reagent blank. Concentrations from 0 to 15 µg/mL Trolox were used to draw the standard curve. The results were expressed as mg Trolox equivalents (TE) per g of dried extracts and the IC 50 concentration showing 50% radical scavenging activity was determined.

Cytotoxicity Assay and Cell Treatment
Cell viability was measured as previously described by Nwakiban et al. [5], after 6 h of co-treatment with the stimulus (TNFα, 10 ng/mL) and the combination (MIX) of extracts (assessed in the range 0.1-10 µg/mL), by the 3,4,5-dimethylthiazol-2-yl-2,5diphenyltetrazolium bromide (MTT) method. Briefly, the medium was removed from each well at the end of the treatment and 200 µL of MTT solution (0.1 mg/mL) was added for 45 min at 37 • C in dark conditions. The formazan salt was extracted from the cells with 200 µL of a mixture of DMSO: isopropanol (10:90), and the absorbance was measured at 570 nm (Envision, PerkinElmer, Walthman, MA, 02451, USA). To study the release of pro-inflammatory mediators (IL-8 and IL-6) and the NF-κB driven transcription, cells were seeded in 24-well plates at a density of 30,000 cells/well. Seventy-two hours later, cells were co-treated with the pro-inflammatory stimulus (TNFα, 10 ng/mL) and the MIX extracts for 6 h using serum-free medium: DMEM/F12 or RPMI 1640 medium, supplemented with penicillin 100 units/mL, L-glutamine 2 mM, and streptomycin 100 mg/mL. At the end of the treatment, the media were collected for the biological assays. All tests were run in triplicate at least three times.

Transient Transfection Assay
Gastric epithelial cells were transiently transfected with the reporter plasmid NF-κB-Luc [29], which contains three κB responsive elements controlling the luciferase gene. AGS cells were transfected by the calcium phosphate method, whereas GES-1 cells were transfected using Lipofectamine ® (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). The day after, the cells were treated for 6 hrs with the MIX extracts and TNFα (10 ng/mL), as previously described. Then, cell lysis and the luciferase enzymatic reaction were conducted through Britelite TM Plus reagent (PerkinElmer Inc., Milan, Italy), according to the manufacturer's instructions. The results (mean ± SD of at least three experiments) were expressed as relative% with respect to the luminescence of the pro-inflammatory conditions (100%).

Measurement of IL-8 and IL-6 Secretion
IL-8 and IL-6 were quantified using two different ELISA kits, as previously reported [5], according to the manufacturer's instructions. Briefly, Corning 96 well EIA/RIA plates (Sigma-Aldrich, St. Louis, MO, USA) were coated with the corresponding antibody and incubated overnight at room temperature. Then, cell media and biotinylated antibodies were added to construct a sandwich ELISA. The cytokines were measured in the samples at 450 nm through the colorimetric reaction between horseradish peroxidase enzyme and 3,3 ,5,5 -tetramethylbenzidine substrate (Sigma-Aldrich, St. Louis, MO, USA) using a spectrophotometer (Victor X3, PerkinElmer, Walthman, MA, 02451, USA). Epigallocatechin-3-O-gallate (EGCG, 20 µM) was used as a reference inhibitor, for its ability to decrease both TNFα-induced IL-8 and IL-6 secretion [4,5,9].

Synergistic Effect Analysis of Extracts in Combination
The Chou Talalay equations [30,31] and the CompuSyn software (version 1.0; Com-boSyn, Paramus, NJ, USA) were used to determine the combination index (CI) and the dose reduction index (DRI). The half-maximal inhibitory concentration (IC 50 ) (µg/mL) of every single extract on IL-6, IL-8, and NF-κB has already been obtained during our latest research work [5]. The CI was used to determine the types of drug interactions in which CI < 1 indicates a synergistic effect, CI = 1 indicates an additive effect, and CI > 1 represents an antagonistic effect. The Equation (1) The dose reduction index (DRI) was calculated using the following Formula (2), by measuring how many times each extract could be reduced in the combination compared to monotherapy:

Statistical Analysis
All results are expressed as mean ± SD. Statistical data were determined with a one-way analysis of variance (ANOVA) followed by the multiple comparison analysis performed with the Bonferroni post hoc test. To compare data obtained from RP-HPLC, the Waller-Duncan test of SPSS software (IBM Corporation, NY, USA, Version 25) was used to test for differences in means. Pearson's test was applied to understand the correlation between the antioxidant variables through the XLSAT software (Version 2022). Statistical analyses were calculated, and graphs were prepared using GraphPad Prism 9.0 software (GraphPad Software Inc., San Diego, CA, USA).

Characterization of Compounds in Extracts by UPLC-ESI-HRMS/MS Analysis
The UPLC-ESI-HRMS/MS profiles of X. parviflora, D. glomerata, T. tetraptera, and A. citratum hydro-ethanolic extracts are shown in Figure S1 and the chromatographic, MS, and MS/MS data are reported in Table 1. The structure of 33 compounds present in the four spice extracts from Cameroon was tentatively identified using the combined interpretation of the fragmentation patterns and the retention time obtained from the UPLC-ESI-HRMS/MS analysis.

Phenolic Acids
Nine compounds belonging to two different classes of phenolic acids have been identified in the extracts (two hydroxybenzoic acids and six hydroxycinnamic acids) ( Table 1). Caftaric acid derivatives were identified in D. glomerata (compound 9b, t R = 16.05 min) and T.  (Table 1) [21,34]. Compound 8b (t R = 15.10 min) and compound 5a (t R = 15.75 min) (Figure S1B,C) revealed molecular ions at m/z 325.2 and m/z 295.2, respectively (Table 1). Based on these data and comparing retention times and MS 2 fragments, they were tentatively assigned to hydroxycinnamic acid compounds [17,21]. Furthermore, compound 6d (t R = 12.58 min) and compound 10d (t R = 15.65 min) ( Figure S1D) in X. parviflora spectra exhibited [M−H] − at m/z 335.2 and m/z 339.2, respectively (Table 1) The MS 2 fragmentation patterns of compound 6d are typical of caffeoylshikimic acid [35]. Compound 10d showed fragment ions at m/z 183 [M−H−156] − characterized for methylgallate residue, and at m/z 197 [M−H−146] − corresponding to the loss of one deoxyhexose, which indicated the presence of caffeoyl tricarboxylic acid isomers [33]. Compound 11d (t R = 16.96 min) was conditionally identified as an isomer of ellagic acid with a deprotonated molecule at m/z 301.2, based on its fragmentation pattern, including the characteristic aglycone fragment ( Figure S1D) [36].

Other Compounds
In addition to the phenolic compounds described above, an alkylmethoxyphenol, compound 6b ([M−H] − at m/z 221.1, t R = 12.63 min), and a phytolaccagenic acid isomer, compound 3a ([M−H] − at m/z 809.3, t R = 11.94 min) were also detected in D. glomerata and A. citratum extracts, respectively ( Figure S1A,B). The above compounds 6b and 3a respectively produced MS 2 spectra with losses of water molecules and aliphatic residues (Table 1), which complied with the published literature [21,44].

Quantitative Determination of Some Phenolic Compounds in Extracts
Phenolic compounds have always been an integral part of the secondary metabolites found in most plant extracts. During phenolic profiling by RP-HPLC, the presence of seven phenolic compounds was observed in the extracts: the most represented in each plant were underlined by statistical analysis ( Table 2). The qualitative composition, with regards to the selected phenols, was only partially similar among plants. One phenolic acid, i.e., caffeic acid (7.07 ± 1.45 µg/100 mg of extract) was determined in X. parviflora extract. In addition, two flavanols, i.e., catechin (2.95 ± 0.70 µg/100 mg of extract) and epicatechin (106.91 ± 0.67 µg/100 mg of extract) were identified, with epicatechin significantly (p ≤ 0.05) higher than the two other compounds. According to the present study, it was reported in other recent papers [14] the presence of catechin compounds (amount of 3.31%) in the X. parviflora extract using GC-MS analysis. In A. citratum, three phenolic acids, i.e., protocatechuic acid (31.12 ± 0.96 µg/100 mg of extract), p-coumaric acid (8.49 ± 1.66 µg/100 mg of extract), rosmarinic acid (4.62 ± 0.15 µg/100 mg of extract) and kaempferol (111.79 ± 0.22 µg/100 mg of extract), were significantly (p ≤ 0.05) higher than the other compounds. [17] reported that the p-coumaric acid identified in this study was exactly 4-O glucoside p-coumaric acid. As well as in A. citratum, T. tetraptera contained kaempferol (13.96 ± 0.90 µg/100 mg of extract). Only two phenolic acids i.e., p-coumaric acid (22.39 ± 1.23 µg/100 mg of extract) and rosmarinic acid (22.03 ± 0.13 µg/100 mg of extract) were identified in D. glomerata. In comparison with the UPLC-ESI-HRMS/MS analysis of D. glomerata extract, no catechin molecules were found. It was therefore interesting to note that even other bioactive compounds were detectable and quantified.

FTIR Analysis of Extracts
The wavenumber ranges of the FTIR peaks and functional groups of each extract were determined by comparing them to the previous reports [46][47][48] (Table 3). The FTIR spectrum of A. citratum, D. glomerata, T. tetraptera and X. parviflora extracts ( Figure  Among these absorption signals, the FTIR spectra of T. tetraptera and X. parviflora extracts ( Figure S2A,D) showed specific wavenumber ranges, namely aromatic secondary amines at 1259-1371 cm −1 (C-N) and mono-, oligo-carbohydrates, oligosaccharides, and glycoproteins at 926 cm −1 , 931-979 cm −1 (trans C-H, Csp 3 -O (C-OH)). However, there were four signals, in which their function groups were still unknown, i.e., 1918-2350 cm −1 , 1917-2350 cm −1 , 1918-2350 cm −1, and 1768-2326 cm −1 . In corresponding previous reports, it has been reported that hydro-ethanolic extracts from the plants used in this study contained phenols, lipids, polysaccharides (generally in the form of glycosides), and aromatic compounds which are also consistent with the results obtained by the UPLC-ESI-HRMS/MS analysis [14,17,18]. Furthermore, these results may be due to the different solvation properties of ethanol and water.

Polyphenol Estimation (TPC, TFC, and FC Content) and Antioxidant Activities (DPPH and ABTS) of Extracts
The polyphenol content in each extract was measured as TPC, TFC, and FC (Table 4) and the results are expressed as gallic acid, catechin, and quercetin equivalents, respectively. Among the extracts, D. glomerata and X. parviflora possessed the highest TPC, with 282.62 ± 3.88 and 271.18 ± 7.10 mg gallic acid equivalents (GAE)/g respectively, followed by T. tetraptera (150.33 ± 0.036 mg gallic acid equivalent (GAE)/g) and A. citratum (129.36 ± 2.13 mg gallic acid equivalent (GAE)/g) [18,49] reported similarities and discrepancies between their findings on the same plants compared to this study. This may arise from different experimental protocols (i.e., organic solvent extraction vs. aqueous extraction) or the harvest period of plant material [16]. A similar trend, but with lower values in TFC was observed in each extract with 35.43 ± 1.33, 114 ± 1.32, 48.73 ± 4.38, 96.61 ± 0.86 mg catechin equivalents (CE)/g for A. citratum, D. glomerata, T. tetraptera, and X. parviflora respectively, indicating that the majority of phenolics present in extracts are flavonoids. Regarding the FC estimation, extracts showed values with a range of 0.58 to 0.21 mg quercetin equivalent (QE)/g, D. glomerata values (0.58 ± 0.14 mg quercetin equivalent (QE)/g) significantly (p ≤ 0.05) higher than the others. This study was similar and comparable with several studies on cultivars, including mango, blueberry, strawberry, raspberry, grapes, garlic, and ginger previously reported by [19,21,50]. The antioxidant capacities of each extract were evaluated through DPPH and ABTS methods. As shown in Table 4, D. glomerata showed the highest ABTS free radical scavenging activity (IC 50 = 5.28 µg/mL) with 51.57 ± 0.74 mg Trolox equivalent (TE)/g of extract. D. glomerata was respectively followed by X. parviflora (IC 50 = 14.01 µg/mL, 47.06 ± 1.05 mg TE/g of extract), T. tetraptera (IC 50 = 28.15 µg/mL, 60.47 ± 1.05 mg TE/g of extract) and A. citratum (IC 50 = 29.21 µg/mL, 52.66 ± 2.66 mg TE/g of extract). As previously noted, D. glomerata showed the highest DPPH antioxidant potential (IC 50 = 15.06 µg/mL) with 254.30 ± 0.15 mg ascorbic acid equivalent (AAE)/g of extract, followed by X. parviflora (IC 50 = 20.38 µg/mL, 253.54 ± 1.88 mg AAE/g of extract), A. citratum (IC 50 = 41.04 µg/mL, 184.88 ± 0.14 mg AAE/g of extract) and T. tetraptera (IC 50 = 45.67 µg/mL, 218.08 ± 1.20 mg AAE/g of extract). In both ABTS and DPPH assays, D. glomerata and X. parviflora extracts showed higher antioxidant potential when compared to the others. This could be due to a greater amount of antioxidant compounds (i.e., phenolic and flavonoid) contained in those extracts that likely could contribute to a strong antioxidant potential [51]. In addition, it is known that the level of antioxidant activity depends on the type and concentration of the phenolic compounds present because their structures greatly affect their bioactivity. Similar experimental procedures and findings were also reported by [49,[52][53][54] when investigating the antioxidant activities of some dietary plants.

Correlations of Antioxidant Assays with Phenolic Compounds
The correlations between the phenolic estimation and antioxidant assays were carried out using the principal component analysis (PCA) and Pearson's correlation test as shown in the supplementary data ( Figure S3 and Table S1). The total phenolic acids and flavonoids were calculated by summing the proposed compounds from the HPLC-PDA quantitation to provide an idea of the overall correlations between all phenolic compounds and antioxidant tests. Polyphenols not found in the HPLC-PDA were not considered. Here, 90.16% variability of the initial data was kept by the first two main components (PC1 = 51.61% and PC2 = 38.52%, respectively) ( Figure S3). DPPH and ABTS were highly correlated (p ≤ 0.05) to both TPC and TFC with positive correlation coefficients of R ≥ 0.96 (Table S1). These strong correlations suggested that polyphenols were the major contributors to the scavenging activity of the plant extracts. Moreover, the correlation between TPC and TFC parameters (R > 0.9, p ≤ 0.05) indicated that flavonoids were the main antioxidant polyphenols. Similar findings were also reported by [18,19,21,55] who studied the polyphenolic composition of some spice extracts. Furthermore, it is observed that DPPH and ABTS tests were positively correlated since Pearson's correlation coefficient was R = 0.97 (p ≤ 0.05) between the assays (Table S1), in line with previous studies [56][57][58]. Since DPPH is only applicable to hydrophobic systems due to the use of a radical dissolved in organic media, the strong correlation with ABTS indicated that additional and less hydrophilic compounds may also contribute to the scavenging effect [59]. On the contrary, the HPLC-detected phenolic acids were negatively correlated (p ≤ 0.05) with the antioxidant assays (Table S1), suggesting that within the selected samples in this study, the phenolic acids do not significantly contribute to the antioxidant activities, maybe because of vitamins (carotenoids, vitamin C, vitamin E) and antioxidant heteropolysaccharides and polypeptides.

Cytotoxicity and Effect of the MIX on the TNFα-Induced NF-κB Driven Transcription in AGS and GES-1 Cells
These studies were performed using two models of gastric epithelial cells (AGS and GES-1 cells) ( Figure S4) stimulated with TNFα (10 ng/mL), a cytokine that contributes to the inflammatory process during gastric epithelium infection. The cytotoxicity of extracts in combination (namely MIX) was assessed in the concentration range of 0.1-10 µg/mL through the MTT assay ( Figure 1A). Similar to our earlier study [5], in which extracts were used as a single treatment, no cytotoxic effects of extracts in combination were observed in the cell cultures after 6 h treatment ( Figure 1A). Extracts in combination (MIX) were investigated for their ability to inhibit the TNFα-induced NF-κB driven transcription in a concentration-dependent manner ( Figure 1B). The MIX was tested in the range of 0.1-10 µg/mL and the corresponding IC 50 values were calculated (Table 5). In AGS cells, the MIX significantly impaired the activation of NF-κB driven transcription with more than 95% reduction (p ≤ 0.001) observed at 10 µg/mL and IC 50 (0.7 µg/mL) below the previously reported IC 50 s (2.14-9.67 µg/mL) for each plant [5]. In GES-1 cells, the MIX at concentrations ranging between 5 and 10 µg/mL, significantly (p ≤ 0.01) impaired the activation of NF-κB driven transcription (32.14-62.39% reduction). The IC 50 (4.9 µg/mL) value was higher compared to the one obtained in AGS cells, but below IC 50 s (5.22-12.17 µg/mL) for each plant summarized in Table 5. The treatment with the reference compound (20 µM EGCG) yielded a significant (p ≤ 0.001) inhibition of NF-κB driven transcription in AGS (>70% reduction) and GES-1 (>50% reduction) cells, which was higher than the MIX at concentrations below 1 µg/mL). Moreover, the basal concentration of the NF-κB driven transcription was not affected by the extracts used as single or in combination (all tested at 10 µg/mL) ( Figure S5). The chemical analysis of each extract reported the presence of many groups of compounds (polyphenols, carbohydrates, lipids, and proteins). Based on this characterization, these metabolites can act as NF-κB inhibitors, such as epigallocatechin, quercetin, cyanidin, chlorogenic acid, and catechins, identified in some extracts. Several studies reported that phenolic acids stimulate the inhibition of NF-κB activation and macrophage infiltration, resulting in the reduction of inflammation in vitro and in animal models [60][61][62][63]. In addition to their role in food intake regulation and nutrition absorption, a growing body of evidence supports that flavonoids counteract NF-κB and inducible nitric oxide synthase (iNOS) signaling pathways, resulting in reduced oxidative damage and inflammation [64,65].
eral studies reported that phenolic acids stimulate the inhibition of NF-κB activation and macrophage infiltration, resulting in the reduction of inflammation in vitro and in animal models [60][61][62][63]. In addition to their role in food intake regulation and nutrition absorption, a growing body of evidence supports that flavonoids counteract NF-κB and inducible nitric oxide synthase (iNOS) signaling pathways, resulting in reduced oxidative damage and inflammation [64,65].   Table 5. Half-maximal inhibitory concentrations (IC 50 ) (µg/mL) as individual extracts [5] or a combination of the TNFα-induced cytokines release and NF-κB driven transcription in human gastric adenocarcinoma (AGS) and gastric epithelial (GES−1) cells.

The Combination of Extracts (MIX) Inhibits TNFα-Induced IL-8 and IL-6 Release in AGS and GES-1 Cells
IL-8 and IL-6 are well-known NF-κB-dependent cytokines that are implicated in the gastric inflammatory process [66,67]. The MIX at concentrations ranging between 0.1 and 10 µg/mL, inhibited IL-8 release in AGS and GES-1 cells with different IC 50 s (Figure 2A, Table 5). In AGS cells, the MIX was more active at 5 and 10 µg/mL (the highest concentrations tested) compared to the reference compound (20 µM EGCG, >58% reduction) and significantly (p ≤ 0.001) inhibited the release of IL-8 (90% reduction) with an IC 50 (0.27 µg/mL) below the one of A. citratum (0.35 µg/mL), T. tetraptera (1.37 µg/mL), X. parviflora (0.30 µg/mL), but over the IC 50 of D. glomerata (0.29 µg/mL) used as single. This value was lower in AGS cells compared to GES-1 cells (IC 50 > 10 µg/mL), as noted earlier in the NF-κB driven transcription results. However, the IC 50 s of the extracts used as single were lower in GES-1 cells (2.30-8.37 µg/mL) in comparison to the IC 50 of the MIX. In this study, no detectable IL-6 secretion was induced in the AGS cells (data not shown); thus, only GES-1 cells were considered. The MIX also inhibited the IL-6 release in GES-1 cells with an IC 50 of 1.8 µg/mL ( Figure 2B, Table 5), a value which was below the IC 50 s (3.47-5.11 µg/mL) of the extracts used as single. In addition, as noted in AGS cells, the MIX was more active at 5 and 10 µg/mL (the highest concentrations tested) compared to the reference compound (20 µM EGCG, around 40% inhibition). This suggests that it can reduce the release and gene expression of NF-κB-dependent pro-inflammatory mediators, which contribute to the amplification of the gastric inflammatory process.

Synergistic Effect of Extracts in Combination on AGS and GES-1 Cells
Combined therapy has shown a variety of advantages over monotherapy, including decreasing the concentration and toxicity of drugs, improving efficiency, targeting multiple molecular pathways, and sensitizing cells to treatment [68]. As the above statements make clear, the definition of synergy can overlap with potentiation. However, a concrete definition derives only from a mathematical approach, shown and proven by different methodologies, such as Berenbaum's pioneering work, the isobologram method of Loewe, the fractional product method of Webb, the combination index method of Chou and Talalay, which provided the foundation for its use in pharmacology and phytopharmacology [69]. Therefore, in the current study, we tested the synergistic effects of the extracts on inflammatory markers involved in gastric inflammation. The effects of each extract (used

Synergistic Effect of Extracts in Combination on AGS and GES-1 Cells
Combined therapy has shown a variety of advantages over monotherapy, including decreasing the concentration and toxicity of drugs, improving efficiency, targeting multiple molecular pathways, and sensitizing cells to treatment [68]. As the above statements make clear, the definition of synergy can overlap with potentiation. However, a concrete definition derives only from a mathematical approach, shown and proven by different methodologies, such as Berenbaum's pioneering work, the isobologram method of Loewe, the fractional product method of Webb, the combination index method of Chou and Talalay, which provided the foundation for its use in pharmacology and phytopharmacology [69]. Therefore, in the current study, we tested the synergistic effects of the extracts on inflammatory markers involved in gastric inflammation. The effects of each extract (used as a single) on the TNFα-induced cytokines (IL-8, IL-6) release and NF-κB driven transcription of AGS and GES-1 cells, previously reported by Nwakiban et al. [5] and their combination are shown in Table 5. The CI values of IL-8, IL-6 release, and NFκB-driven transcription were more than one (6.18) in GES-1 cells and 0.82 (slight synergism) in AGS cells; 0.46 (synergism) in GES-1 cells; 0.61 (synergism) in GES-1 cells and 0.18 (strong synergism) in AGS cells respectively, which in overall indicates a synergistic effect of the MIX on cells against the gastric inflammatory process. Moreover, the means DRI of extracts in combination were 0.80 ± 0.40 in GES-1 and 8.18 ± 8.13 in AGS cells for IL-8 release; 9.22 ± 1.91 in GES-1 cells for IL-6 release; 14.28 ± 4.6 in GES-1 and 28.74 ± 11.58 in AGS cells for NFκB driven transcription, which suggests a very high-fold dosage reduction compared to each extract in monotherapy, i.e., eight-fold dosage reduction for IL-8 release in AGS cells and fourteen-fold dosage reduction for the NF-κB driven transcription in GES-1 cells. the MIX has been noted to act antagonistically in GES-1 cells on NF-κB transcription (CI > 1), however, it was found to exert an NF-κB inhibition activity in a concentration-dependent fashion ( Figure 1B). Overall, the biological activities of the MIX were lower in AGS cells compared to GES-1 cells as was previously reported [5]. This could be attributed to the differences in mutation and differentiation of GES-1 and AGS cells, but also due to the differences in their genetic profiles [31]. Furthermore, the natural bioactive compounds found in each extract have a more structural diversity that could more effectively inhibit certain targets of gastric disorders (gene expression of proinflammatory cytokines). They also inherently target other biologically relevant NF-κB pathways, i.e., enzymes such as prostaglandin-endoperoxide synthase 2 (PTGS2) (COX-2), because many natural bioactive compounds are secondary metabolites or signaling molecules [5,70]. Based on the above findings, it is suggested that the MIX contains various types of bioactive compounds that act in synergy in human gastric epithelial cells by two or more mechanisms (synergistic multi-target effects and elimination or neutralization potential).

Conclusions
In this study, a comparative analysis of the chemical composition of each hydroalcoholic extract through RP-HPLC, UPLC-ESI-HRMS/MS, and FTIR indicates that they mostly contain a great number of phenolic compounds, but also proteins, lipids, polysaccharides (generally in the form of glycosides) and aromatic compounds. The extracts showed a high amount of total phenolic (TPC: 150-290 mg GAE/g of extract) and flavonoid content (TFC: 35-115 mg CE/g of extract) with antioxidant properties in a cell-free system (DPPH IC 50 s ≤ 45 µg/mL; ABTS IC 50 s ≤ 29 µg/mL). The extracts in combination (MIX) exert a synergistic beneficial effect (CIs < 1 and DRIs > 1) on inflammatory markers (IL-8, IL-6 release, and NF-κB driven transcription) in human gastric epithelial cells which may be due to the presence of phenolic compounds (mostly phenolic acids and flavonoids). Among the phenolic compounds, phenolic acids (hydroxybenzoic and hydroxycinnamic acids) and flavonoids (anthocyanins, flavanols, flavanonols, flavanones, flavones, flavonols, and isoflavonoids) which promoted an antioxidant potential, have been reported in the extracts. The MIX enhances the efficacy of extract used in monotherapy and reduces the potential adverse effects related to high concentrations, as indicated by high DRI. Beyond the plausible pharmacodynamic interaction among phenolic compounds, further studies are required to assess other possible mechanisms of synergistic interaction, and to in vivo confirm the synergistic effect herein reported. To the best of our knowledge, this study provides a scientific basis for the traditional practice of using a combination of extracts useful to alleviate gastric inflammation.