Anti-Inflammatory and Antioxidant Effects Induced by Allium sativum L. Extracts on an Ex Vivo Experimental Model of Ulcerative Colitis

Inflammatory bowel diseases (IBDs) are chronic and multifactorial inflammatory conditions of the colonic mucosa (ulcerative colitis), characterized by increased and unbalanced immune response to external stimuli. Garlic and its bioactive constituents were reported to exert various biological effects, including anti-inflammatory, antioxidant and immunomodulatory activities. We aimed to evaluate the protective effects of a hydroalcoholic (GHE) and a water (GWE) extract from a Sicilian variety of garlic, known as Nubia red garlic, on an ex vivo experimental model of ulcerative colitis, involving isolated LPS-treated mouse colon specimens. Both extracts were able to counteract LPS-induced cyclooxygenase (COX)-2, tumor necrosis factor (TNF)-α, nuclear factor-kB (NF-kB), and interleukin (IL)-6 gene expression in mouse colon. Moreover, the same extracts inhibited prostaglandin (PG)E2, 8-iso-PGF2α, and increased the 5-hydroxyindoleacetic acid/serotonin ratio following treatment with LPS. In particular, GHE showed a better anti-inflammatory profile. The anti-inflammatory and antioxidant effects induced by both extracts could be related, at least partially, to their polyphenolic composition, with particular regards to catechin. Concluding, our results showed that GHE and GWE exhibited protective effects in colon, thus suggesting their potential use in the prevention and management of ulcerative colitis.


Introduction
Garlic (Allium sativum L.) is an herbaceous plant, belonging to the Amarillidaceae family, that is used all over the world as traditional medicine and spice [1]. Garlic contains a number of biologically active compounds, including phenolic compounds [2], saponins [3], polysaccharides [4], as well as organosulfur compounds [1] that contribute to its countless pharmacological properties. In particular, the health-promoting properties induced by garlic were suggested to be related to its bioactive compounds, including phenolic compounds,

Preparation of Garlic Extracts
Garlic cloves were supplied as dried powder by il Grappolo S.r.l. (Soliera, Modena, Italy). Plant sample (1 g) was mixed with either a solution of ethanol-water (20:80, v/v) or water (final concentration = 1 g/mL), as previously reported [20][21][22]. The supernatant was filtered and then dried (freeze-drying). The dry residue, a yellow sugary solid, was stored at 4 • C, until chemical analyses were performed.

High Performance Liquid Chromatography (HPLC)-Diode Array (DAD)-Mass Spectrometry (MS) Analysis
Selected phenolic compounds contained in the extracts were identified and quantified by HPLC-DAD-MS analysis. The HPLC apparatus consisted of a two PU-2080 PLUS chromatographic pump, a DG-2080-54 line degasser, a mix-2080-32 mixer, UV, diode array (DAD) and detectors, a mass spectrometer (MS) detector (expression compact mass spectrometer (CMS), Advion, Ithaca, NY, USA), an AS-2057 PLUS autosampler, and a CO-2060 PLUS column thermostat (all from Jasco, Tokyo, Japan). Integration was conducted through ChromNAV2 Chromatography software. The separation was performed on an Infinity lab Poroshell 120-SB reverse phase column (C18, 150 × 4.6 mm i.d., 2.7 µm) (Agilent, Santa Clara, CA, USA). Column temperature was set at 30 • C. The separation was conducted within 60 min of the chromatographic run, starting from the following separation conditions: 97% water with 0.1% formic acid, 3% methanol with 0.1% formic acid, as previously described [23]. Quantitative determination of phenolic compounds was conducted through a DAD detector. Qualitative analysis of GHE and GWE was performed by an MS detector in the positive and negative ion modes. MS signal identification was performed by comparison with a standard solution and MS spectra available in the MassBank Europe database (https://massbank.eu/MassBank/ (accessed on 11 November 2021). All HPLC grade solvents were purchased from Merck Science Life S.r.l. (Milan, Italy). Each analysis was performed in triplicate. The detailed protocol is enclosed as supplementary materials.

Headspace Solid-Phase Microextraction-Gas Chromatography-Mass Spectrometry (HS-SPME-GC-MS) Analysis
The 4 mL vials were loaded with 50 mg of the garlic dry residue. According to the extraction parameters optimized in our previous screening study performed on the garlic powder [15], the sample was allowed to equilibrate at 80 • C for 20 min and then the SPME fiber DVB-CAR-PDMS (Sigma Aldrich, now Merck KGaA, Darmstadt, Germany) was exposed to the head space of the vial for 20 min at 80 • C.
After sampling, fiber was withdrawn in the needle and exposed into the GC inlet at 260 • C for 0.5 min. The desorbed analytes were introduced in a gas chromatograph (6850, Agilent Technologies, Santa Clara, CA, USA) coupled with a mass spectrometer (5975, Agilent Technologies, Santa Clara, CA, USA), equipped with the non-polar capillary column HP-5MS (30 m × 0.25 mm inner diameter, and film thickness 0.25 µm). The gaschromatographic parameters were set as follows: inlet temperature, 260 • C; injection mode, splitless (splitter valve was opened after 0.2 min and split ratio = 20/1); flow rate of the helium carrier gas (99.995% purity), 1.0 mL/min; oven temperature starting from 40 • C, after 5 min raised to 200 • C at 5 • C/min, and kept at this final temperature for 60 min. Mass spectrometry parameters were set as follows: EI energy, 70 eV; source temperature, 230 • C; quadrupole temperature, 150 • C; and mass scan was carried out over the 50-350 m/z range. The analyses were performed in triplicate.
Two analytical criteria were used to allow the identification of the eluted compounds, namely the comparison between the EI experimental spectra and those collected in the commercial (FFNSC 3) and free access databases (NIST 11, Flavor2) and the Kovats index (KI) measured using a mixture of n-alkanes (C7-C40) with the same chromatographic conditions, and then compared with values reported in the FFNSC 3 and NIST 11 databases. A manual integration of chromatographic peaks with a S/N ratio above 3 was performed without any further modification.

Positive-Ion Direct Infusion-Electrospray Ionization-Mass Spectrometry (DI-ESI-MS) Analysis
The dry residue of garlic was dissolved in H 2 O:MeOH (7:3) to a final concentration of 20 µg/mL and directly infused at 10 µL min −1 in the ESI source of a LTQ XL linear ion trap (Thermo Fisher Scientific, Waltman, MA, USA). The source parameters were set as follows: source voltage = 4.5 kV; capillary voltage = 15 V; capillary temperature = 300 • C; tube lens voltage 89.9 V, sheath gas flow rate 10 (arbitrary units). Each spectrum, acquired over the 170-2000 m/z range, was from averaging 10 full scans, each one consisting of 5 micro scans. The major peaks were isolated and submitted to MS tandem experiments to allow the compound identification by comparing their relevant MS/MS spectra with those reported in literature or collected in a free access database (https://massbank.eu/ MassBank/Search (accessed on 19 May 2022). The precursor ion isolation width was 1-2 Da and the normalized collision energy was set to the value needed to reduce the intensity of the precursor ion to approximately 10%.

Colorimetric Analysis
Colorimetric analysis of garlic powder (GP) sample and GWE was performed by X-Rite MetaVueTM ® (Prato, Italy) as previously described [15]. GP sample and GWE analysis was conducted at the time of delivery (t • ) and after 12 months (t 12m ) of storage in the darkness at room temperature (25 ± 2 • C). The detailed description of colorimetric analysis is reported in the Supplementary Materials section.

HPLC-DAD Analysis
GWE was weighed, dissolved in water and filtered before injection into a HPLC Perkin Elmer apparatus (Series 200 LC pump, Series 200 DAD and Series 200 autosampler, Milan, Italy). Chromatography was conducted on RP-18 column (3 µ) using a linear gradient consisting of acetonitrile and acidified water (5% formic acid), from 100% aqueous phase to 85% in 15 min, 85 to 55% in 30 and 55 to 40% in 20 , at a flow rate of 0.8 mL/min, at 254 nm. Alliin was quantified as previously described [22] (y = 6.35x + 50.34; R2 = 0.9987, in the range between 2 and 400 µg/mL, LOD 0.6 µg/g and LOQ 2.0 µg/g extract in dry weight). Each analysis was performed in quadruplicate. The detailed protocol related to HPLC-DAD analysis is described in the Supplementary Materials section.
Extraction of total RNA from colon specimens was performed using TRI reagent (Sigma-Aldrich, St. Louis, MO, USA), according to the manufacturer's instructions. Reverse transcription was performed using High Capacity cDNA Reverse Transcription Kit (ThermoFischer Scientific, Waltman, MA, USA). Gene expression of cyclooxygenase (COX)-2, TNF-α, nuclear factor-kB (NF-kB), IL-6, and nuclear factor erythroid 2-related factor 2 (Nrf2) was measured by quantitative real-time PCR using TaqMan probe-based chemistry [30,31]. The real-time PCR was performed in triplicate. Relative quantification of gene expression was conducted through the comparative 2 −∆∆Ct method [32]. The detailed description of real-time PCR is reported in the Supplementary Materials section.
Extraction of serotonin (5-HT), and 5-hydroxyindolacetic acid (5HIIA) was performed from individual colon specimens homogenized in perchloric acid solution (50 mM). Analysis of colon 5-HT, and 5HIIA levels was performed through high performance liquid chromatography coupled to electrochemical detection consisting of ESA Coulochem III detector equipped with ESA 5014B analytical cell [33,34]. Each analysis was performed in triplicate.

Statistical Analysis
Analysis of the data was performed by using the software GraphPad Prism version 6.0 (Graphpad Software Inc., San Diego, CA, USA). Means ± SEM were assessed for each experimental group and analyzed by one-way analysis of variance (ANOVA), followed by the Newman-Keuls multiple comparison post hoc test. As for quantification of the investigated phenolic compounds detected in GHE and GWE, analysis of the data was conducted by unpaired t test (two-tailed p value). The limit of statistically significant differences between mean values was set at p-value < 0.05. Calculation of the number of animals randomized for each experimental group was performed by using the "Resource Equation" n = (E + T)/T (10 ≤ E ≤ 20) [35].
The DI-ESI-MS full spectrum was dominated by the regular repetition of two monocharged species differing by 16 mass units (Figure 2), that suggests the presence of species complexing Na + and K + cations. The 162 Da intervals correspond to the difference of one -C 6 H 10 O 5 -unit, typical of a homologous series of polysaccharides. The MS/MS spectrum of the 1029 m/z ion, reported in Figure 3 as representative ion of the series, showed the sequential loss of 162 Da (=C 6 H 10 O 5 ) and 18 Da (=H 2 O) according to the MS2 fragmentation of a polysaccharide. The detected oligosaccharides, most probably fructans [41], are characterized by a wide degree of polymerization (DP), namely in the 2-11 DP range ( Table 3). The Na + and K + ions distribution emerging from the fructans series indicates a larger content of the latter, despite what one should expect considering the ubiquitous presence of Na + cation in the ESI-MS experiments. In addition to the fructans, the full scan showed two intense signals at m/z 175 and 214, corresponding to the protonated L-arginine and to N-butylbenzene sulfonamide (a plasticizer contaminant), respectively (Table 3).

Colorimetric Analysis
The color parameters and relative reflectance curves of the analyzed samples are shown in Table 4 and Figure 4.

Colorimetric Analysis
The color parameters and relative reflectance curves of the analyzed samples are shown in Table 4 and Figure 4.   As reported in our previous work [22], the garlic powder analyzed at t • has a very bright color tending to pale yellow (L* = 90.15 and b* = 16.02) that, after eight months of storage, seemed to fade slightly (L* ≈ 94 vs. 90 at t 0 , b* ≈ 14 vs. 16 at t 0 , with a ∆E ≈ 2, data previously reported). In this study, the analyses performed at t 12m , showed a substantial change in color (∆E = 12.34), characterized by a strong darkening of the powder (L* ≈ 78; b* ≈ 20), as also shown by the lowering of the reflectance curve reported in Figure 4. The initial light bleaching and subsequent browning were already described in milk powder samples in Cesa and collaborators (2015) [42], and interpreted as carotenoid degradation followed by Maillard reaction. In the case of garlic, bleaching was not previously described. Nevertheless, the presence of carotenoids in Allium spp. phytocomplex was reported in the literature [43] and the carotenoid bleaching seems the simplest hypothesis, despite all our attempts to extract a carotenoid fraction by organic solvents failed. Therefore, the slight bleaching should be attributed to other yellow, hydrosoluble, components. On the other hand, the powder darkening could be mainly due to the different sulfur components (the presence of many different sulfur compounds was confirmed by HS-SPME-GC-MS analysis (see Section 3.1.2)) which, modifying the pH and the activity of polyphenol oxidases, with a not completely known mechanism depending on the state of preservation and on the temperature, cause darkening of the matrix [44,45].
This darkening is also confirmed in the aqueous extract newly prepared and analyzed after 12 months (see Figure 4). In fact, the L* parameter of GWE t • decreases significantly from 76 to 65, as well as the b* parameter increases even more, from 17 to 28, denoting altogether a strong color change characterized by a substantial browning with respect to the starting points (∆E = 28 respect to G P t • and ∆E = 16 respect to GWE t • ).

HPLC-DAD Analysis
The HPLC-DAD analysis was conducted at 254 nm for the identifying benzoic and hydroxycinnamic acids, flavanols and organosulfur compounds. The molecular profile is confirmed by literature even if no chromatograms of aqueous extracts are reported [46,47].
The chromatogram showed the presence of alliin and an its diastereoisomer (361.1 ± 17.3 µg/g dry extract) as reported by Dethier et al., 2012 [48]. Furthermore, chlorogenic acid, caffeic acid, and epicatechin (see Section 3.1.1 for the relevant quantification) were also shown.
These results are only partially comparable to those present in the literature, because the garlic phytocomplex and relative aqueous extracts are very variable according to the different cultivar, geographic area, and storage conditions. There are no references in the literature on quantitative data related to garlic aqueous extracts. Existing data refer to methanolic or hydroalcoholic extracts reporting alliin content in range of 0.5-33 expressed in mg/g dry weight [47,49]. Compared to the hydroalcoholic extract, reported in our previous work [22], alliin values found in GWE are significantly lower (approximately 360 vs. 1200 µg/g dry extract).

Toxicological and Pharmacological Studies
In a previous study of ours, GHE (1-100 µg/mL) was not able to modify cell viability of H9c2 cells (rat cardiomyoblasts), in basal conditions, confirming its good biocompatibility [22]. Moreover, the same extract (10-100 µg/mL) was effective in protecting cells from cytotoxicity induced by H 2 O 2 (200 µM) [22]. In the present study, we investigated the effects of GHE and GWE, in the dose range 1-100 µg/mL, on colon cancer SW480 cell line viability, in basal conditions. Compared to the control group, GHE was not able to affect SW480 cell viability ( Figure 5). However, GWE (100 µg/mL) significantly suppressed SW480 cell viability, even if the cell viability was not under the biocompatibility limits (70% viability compared to control, respectively) ( Figure 5). Accordingly, both GHE and GWE were not able to modify apoptosis of SW480 cell lines following 48 h of treatment in basal conditions ( Figure 6). Accordingly, both GHE and GWE were not able to modify apoptosis of SW480 cell lines following 48 h of treatment in basal conditions ( Figure 6).

Figure 5.
MTT assay of SW480 cell line treated with garlic hydroalcoholic extract (GHE) (1, 10, and 100 µg/mL), garlic water extract (GWE) (1, 10, and 100 µg/mL), and vehicle (RPMI) for 48 h, in basal conditions. Data are displayed as the means ± SEM. ANOVA, p < 0.001; ** p < 0.01 vs. vehicle. Accordingly, both GHE and GWE were not able to modify apoptosis of SW480 cell lines following 48 h of treatment in basal conditions ( Figure 6). GHE and GWE, in the dose range 1-100 µg/mL, were then tested to evaluate their potential protective activities on oxidative and inflammation pathways in mouse colon specimens treated with LPS. In particular, GHE, in the tested dose range, was effective in suppressing LPS-induced gene expression of pro-inflammatory markers strongly GHE and GWE, in the dose range 1-100 µg/mL, were then tested to evaluate their potential protective activities on oxidative and inflammation pathways in mouse colon specimens treated with LPS. In particular, GHE, in the tested dose range, was effective in suppressing LPS-induced gene expression of pro-inflammatory markers strongly involved in colon inflammation, including COX-2, TNF-α, NF-κB, and IL-6 ( Figure 7A-E) [50][51][52][53]. In the same experimental paradigm, GWE (1-100 µg/mL) inhibited LPS-induced TNF-α, NF-κB, and IL-6 ( Figure 7A-E) gene expression. Moreover, the higher concentrations of GWE suppressed COX-2 ( Figure 7A) gene expression induced by LPS treatment. On the other hand, GHE and GWE did not modify LPS-induced Nrf2 gene expression ( Figure 7E), ruling out a possible role of this mediator in mediating the protective effects exerted by GWE and GHE in mouse colon.
However, GHE and GWE (100 µg/mL) prevented LPS-induced reduction in 5-HT turnover. The decreased levels of 5-HT, measured as the 5HIIA/5-HT ratio, could further account for the anti-inflammatory effects exerted by garlic extracts. To this regard, we have previously found that anti-inflammatory herbal extracts suppressed 5-HT levels in isolated rat colon treated with LPS [82,83]. In particular, GHE was found more effective than GWE in counteracting the LPS-induced decrease in the 5HIIA/5-HT ratio (Figure 9). Actually, the higher activity of GHE compared to GWE could be related to its higher content in benzoic acid and flavonoids, such as quercetin [84,85].
A wide body of evidence showed that imbalance between the oxidative reactions and antioxidant defense mechanisms played a key role in the initiation and progression of IBD. This imbalance generates oxidative stress resulting from either reactive oxygen species (ROS) overproduction or a reduction in antioxidant activity [86,87]. In particular, ROS overproduction was suggested to be involved in functional disruption of the enteric mucosa [52]. Increased production of ROS is known to damage cellular lipids, proteins, as well as nucleic acids, and finally disrupt gastrointestinal barrier integrity [88]. 8-iso-PGF 2α is an isomer of prostaglandins produced from membrane arachidonic acid by free radical-catalyzed peroxidation, which is regarded as a stable marker of lipid peroxidation and oxidative stress [89]. As shown in Figure 10, both GHE and GWE (1-100 µg/mL) were able to counteract 8-iso-PGF 2α levels induced by LPS treatment in isolated mouse colon specimens. In particular, GHE (1 µg/mL) showed higher efficacy in inhibiting 8-iso-PGF 2α levels ( Figure 10). These results agreed with the antioxidant effects induced by GHE, tested in the same concentration range, in isolated mouse heart [22]. Accordingly, aqueous garlic extract was also shown to possess antioxidant properties by scavenging ROS and increasing cellular antioxidant enzymes including superoxide dismutase, catalase, and glutathione peroxidase [36]. The antioxidant effects induced by GHE and GWE are consistent with their polyphenol content [90], with particular regards to catechin [58]. In this context, catechins were found able to reduce the colonic oxidative damages to the colon, by suppressing oxidative stress, via exerting direct or indirect antioxidant effects, and by enhancing the activity of various antioxidant enzymes including glutathione peroxidases [58]. Figure 10. Effects of garlic hydroalcoholic extract (GHE) (1, 10, and 100 µg/mL), garlic water extract (GWE) (1, 10, and 100 µg/mL) and vehicle (RPMI) on LPS-induced 8-iso-prostaglandin (PG)F 2α levels in mouse colon specimens. Data are shown as the means ± SEM. ANOVA, p < 0.0001; *** p < 0.001 vs. LPS; ### p < 0.001 vs. co-respective treatment with GWE.
These results agreed with the antioxidant effects induced by GHE, tested in the same concentration range, in isolated mouse heart [22]. Accordingly, aqueous garlic extract was also shown to possess antioxidant properties by scavenging ROS and increasing cellular antioxidant enzymes including superoxide dismutase, catalase, and glutathione peroxidase [36]. The antioxidant effects induced by GHE and GWE are consistent with their polyphenol content [90], with particular regards to catechin [58]. In this context, catechins were found able to reduce the colonic oxidative damages to the colon, by suppressing oxidative stress, via exerting direct or indirect antioxidant effects, and by enhancing the activity of various antioxidant enzymes including glutathione peroxidases [58].

Conclusions
Concluding, GHE and GWE, particularly GHE, showed protective effects, as confirmed by the inhibitory effects on selected pro-inflammatory and pro-oxidant markers, in LPS-stimulated colon, suggesting a potential role in the prevention and management of ulcerative colitis. The phytochemical analyses suggested these effects could be related, albeit partially, to their phenol and flavonoid content, with particular regards to catechin. Moreover, other components of nutraceutical and pharmaceutical interest were detected in these extracts. On the other hand, further studies are necessary to accurately evaluate the in vivo activity.