Effect of Non‐Volatile Constituents of Elsholtzia ciliata (Thunb.) Hyl. from Southern Vietnam on Reactive Oxygen Species and Nitric Oxide Release in Macrophages

The extract of Elsholtzia ciliata aerial parts was subjected to bio‐guided isolation using the intercellular ROS reduction in J774A.1 macrophages to monitor the anti‐oxidative activity. Fifteen compounds were isolated from the active fractions including eleven flavonoids (vitexin, pedalin, luteolin‐7‐O‐β‐d‐glucopyranoside, apigenin‐5‐O‐β‐d‐glucopyranoside, apigenin‐7‐O‐β‐d‐glucopyranoside, chrysoeriol‐7‐O‐β‐d‐glucopyranoside, 7,3′‐dimethoxyluteolin‐6‐O‐β‐d‐glucopyranoside, luteolin, 5,6,4′‐trihydroxy‐7,3′‐dimethoxyflavone, 5‐hydroxy‐6,7‐dimethoxyflavone (compound 13), 5‐hydroxy‐7,8‐dimethoxyflavone); three hydroxycinnamic acid derivatives (caffeic acid, 4‐(E)‐caffeoyl‐l‐threonic acid, 4‐O‐(E)‐p‐coumaroyl‐l‐threonic acid) and one fatty acid (α‐linolenic acid). The biological evaluation of these compounds (10–2.5 μm) indicated that all of them exerted good antioxidant and anti‐inflammatory activities, in particular compound 13.


Introduction
Reactive oxygen species (ROS), inevitable side products of normal cellular metabolism, are necessary for signaling and regulation of biological functions at physiological levels. [1,2] Especially they are generated in response to pathogens or foreign invasion, playing an imperative role in immune system. [3] However, uncontrolled ROS levels, not adequately balanced by endogenous and exogenous antioxidant defenses, can lead to an oxidative stress status giving rise to cell damage and various diseases. [4] Oxidative stress is known to hasten the aging process, significantly contribute to pathophysiology of a broad variety of disorders or diseases such as cardiovascular diseases, inflammatory bowel disease, rheumatoid arthritis, and others. [5] Moreover, a considerable body of evidence has shown that ROS trigger NF-кB which activates and/or further amplifies the pro-inflammatory response regulating pro-inflammatory mediators, such as nitric oxide (NO). [6,7] Therefore, regulation of redox balance is crucial for healthy living organisms. Besides natural defense mechanisms based on endogenous antioxidants such as antioxidant enzymes (e. g., glutathione peroxidase, catalase and superoxide dismutase), metal-binding proteins (like ferritin, transferrin, metallothionein, lactoferrin, ceruloplasmin) or non-enzymatic scavengers (e. g., uric acid, bilirubin), [8] recent studies proved that exogenous antioxidants originating from a diet rich in spices, culinary herbs, fruits and vegetables providing compounds such as vitamin C, E, carotenoids and phenolics also play an important antioxidant effects in both oxidative stress and inflammatory conditions in J774A.1 murine macrophages. In order to elucidate the contribution of the non-volatile constituents to the observed activities, we performed an activity-guided isolation and evaluated the antioxidant and anti-inflammatory activities of subfractions and pure compounds on J774A.1 macrophages.

Activity Guided Isolation
Bioactive compounds present in food could play a role in physical dysfunction through their effects on inflammatory mediators and pathways, barrier integrity, and/or intestinal microbiota composition. Evaluation of the antioxidant potential was performed in both inflammatory and oxidative stress conditions in J774A.1 macrophages. The impact on ROS production induced by LPS was of special interest due to its similarity to a microbial induced intestinal inflammation with an interaction of diet derived secondary plant metabolites. In order to assess the activity of the constituents of E. ciliata, the air-dried aerial plant parts were extracted with methanol. A part of the obtained dry extract was dissolved in water and fractionated by liquid-liquid extraction with organic solvents of different polarity to yield a petroleum ether, diethyl ether, ethyl acetate, n-butanol and the respective water subfraction. In the performed assays, all the tested extract/subfractions exerted antioxidant effects at concentrations � 5 μg/mL (P < 0.001 vs. LPS or H 2 O 2 ; Table 1). At lower tested concentration, only the petroleum ether, diethyl ether and water fractions were active in inhibiting ROS release during oxidative stress conditions in macrophages (2.5 μg/mL P < 0.001 vs. H 2 O 2 ; Table 1). All the tested extracts and subfractions did not exert cytotoxic activity except for the water fraction, which showed marginal anti-proliferative effects (20.67 � 1.2 vs. untreated macrophages) at the highest tested concentration.
HPLC Analysis of the methanolic extract and its subfractions ( Figure 1) showed a good enrichment of individual compounds in the different subfractions e. g. compound 13, 14, and 15 in the petroleum ether subfraction. On the other hand, the analysis revealed that the activity of the extract and its subfractions could not be explained by the presence of a single active principal but due to the presence of several compounds. Aiming at the identification of those, the three most potent extract fractions (petroleum ether, diethyl ether and ethyl acetate fraction) were selected for further phytochemical investigation.

Anti-Oxidative and Anti-Inflammatory Activity of Pure Compounds
Prior to the evaluation of the antioxidant and antiinflammatory activities, the anti-proliferative activity of the pure compounds at all tested concentrations (10-2.5 μM) were analyzed. With the exception of compound 14, which showed an anti-proliferative activity (38.0 � 9.1 vs. untreated macrophages) and was therefore excluded from the further experiments, none of the tested compounds displayed at the tested concentrations a potential cytotoxic activity. Evaluation of the antioxidant activity indicated that all tested compounds exerted significant effects on ROS release during oxidative stress conditions at the highest tested concentration (10 μM). In particular, compounds 6, 11     (15) was not unexpected but helped to verify the experimental setup. [52] More surprising was the finding that flavonoid glucosides (5 and 6) as well as the flavonoid aglycons (11 and 13) showed activities in the cell-based assay that suggests a sufficient cellular up-take of the substances or more unlikely an interaction with the extracellular LPS-signaling. In order to evaluate the contribution of the tested compounds to the inflammatory response, their effects on NO release in LPS-stimulated macrophages were evaluated. With the exception of compounds 10 and 12, all tested compounds significantly inhibited NO release in macrophages during LPS-induced inflamma-tion (P < 0.001 vs. LPS alone treated macrophages) at all tested concentrations (Figure 3). A possible explanation for the lack of activity of 10 and 12 might be the presence of two methoxy groups at 3' and 7 in these compounds. The most pronounced effect could be observed for 13, which was also found to be one of the most dominant compounds according to HPLC-DAD analysis (Figure 1).

Conclusions
Since the essential oil of the investigated species was already shown to be active in vivo in a rat paw swelling assay induced by serotonin or carrageenan, and in a model for chronic arthritis induced by formaldehyde, [35] we were interested in the contribution of the non-volatile constituents of E. ciliata to the antioxidant effect in inflammatory and in oxidative stress conditions. The impact on ROS production induced by LPS was of special interest, since this  Table S5.
Chem. Biodiversity 2021, 18, e2000577 model mimics the ROS level of a microbial induced intestinal inflammation, which is in agreement with the traditional use of E. ciliata. [13,15] From a phytochemical point of view, several compounds, which have not been reported as components of E. ciliata before, could be identified. Among those, two unusual hydroxycinnamic acid derivatives 4-(E)-caffeoyl-L-threonic acid (2) and 4-O-(E)-p-coumaroyl-L-threonic acid (3), were described for the first time as constituents of a member of the Lamiaceae family. In summary, the performed study supports the traditional use of E. ciliata against ROS/inflammation associated disorders by the found in vitro activity and identifies the aerial plant parts as a valuable source of flavonoids and phenolic acids with anti-inflammatory and anti-oxidative activities.

Intracellular ROS Release Measurement
ROS intracellular production in J774A.1 macrophages was evaluated by the probe of 2',7'-dichlorofluorescein-diacetate (H 2 DCF-DA). J774A.1 cells were seeded in 24-well plates (3 × 10 5 cells/well) and allowed to adhere. After the cellular treatment with the tested compounds in combination with LPS or H 2 O 2 as described above, J774A.1 cells were collected, washed with PBS, and then incubated in PBS containing H 2 DCF-DA (10 μM). Cell fluorescence was evaluated after 15 min at 37°C, using a fluorescence-activated cell sorter (FACSscan, Becton Dickinson, Franklin Lakes, NJ, USA), and was analyzed by CellQuest software version 4 (Becton Dickinson, Milan, Italy), as formerly reported. [54] Measurement of NO Release NO levels were measured as nitrite NO 2 À , index of NO released by cells, in the culture medium of J774A.1 macrophages 24 h after LPS stimulation by Griess reaction, as previously reported. [55] Briefly, after the cell treatment with the tested compounds and LPS, as previously described, 100 μL of cell culture medium were mixed with 100 μL of Griess reagent (equal volumes of 1 % (w/v) sulfanilamide in 5 % (v/v) phosphoric acid and 0.1 % (w/v) N- (1-napthyl) ethylenediamine dihydrochloride in water) and incubated at room temperature for 10 min. Subsequently, the absorbance was measured at 550 nm in a microplate reader Titertek (Dasit, Cornaredo, Milan, Italy). The amount of NO 2 À measured in the samples is expressed in μM concentration, which was calculated via a sodium nitrite standard curve.

Data Analysis
Results of the pharmacological assays are reported as mean values in % inhibition � standard error of the mean (SEM) referred to the LPS-stimulated control group in Table 1 and Figure 3. At least three independent experiments, each in triplicate, were performed for each extract and concentration. Statistical analysis was performed by analysis of variance test, and multiple comparisons were made by Bonferroni's test using Prism 5 (GraphPad Software, San Diego, CA, USA). A P-value less than 0.05 was considered significant.