Flavonoids from Chionanthus retusus (Oleaceae) Flowers and Their Protective Effects against Glutamate-Induced Cell Toxicity in HT22 Cells

The dried flowers of Chionanthus retusus were extracted with 80% MeOH, and the concentrate was divided into EtOAc, n-BuOH, and H2O fractions. Repeated SiO2, octadecyl SiO2 (ODS), and Sephadex LH-20 column chromatography of the EtOAc fraction led to the isolation of four flavonols (1–4), three flavones (5–7), four flavanonols (8–11), and one flavanone (12), which were identified based on extensive analysis of various spectroscopic data. Flavonoids 4–6 and 8–11 were isolated from the flowers of C. retusus for the first time in this study. Flavonoids 1, 2, 5, 6, 8, and 10–12 significantly inhibited NO production in RAW 264.7 cells stimulated by lipopolysaccharide (LPS) and glutamate-induced cell toxicity and effectively increased HO-1 protein expression in mouse hippocampal HT22 cells. Flavonoids with significant neuroprotective activity were also found to recover oxidative-stress-induced cell damage by increasing HO-1 protein expression. This article demonstrates that flavonoids from C. retusus flowers have significant potential as therapeutic materials in inflammation and neurodisease.


Introduction
With the rapid growth of the aging population, the treatment of age-related diseases has become an important global issue, including in Korea [1]. Neurodisease is among the various illnesses induced by aging [2]. Previous studies have revealed the neuroprotective activities of bioactive compounds such as alkaloids, sterols, and flavonoids [3,4]. Flavonoids perform various neuroprotective actions, such as suppressing neuroinflammation; protecting neurons; and promoting memory, cognitive function, and learning [5,6]. Given the many experiments demonstrating their neuroprotective effects, these compounds may have therapeutic potential in neurodisease [3,[6][7][8][9].
Flavonoids have a phenylchromane (C6-C3-C6) structure and are synthesized from l-phenylalanine and l-tyrosine via the shikimic acid pathway [10]. They comprise one of the most widespread and diverse groups of compounds in nature [11][12][13]. Among various natural resources, flowers (the reproductive organs of plants) contain diverse secondary metabolites, including volatiles, pigments, and flavonoids, which lure pollinating insects and facilitate pollination [14][15][16]. Sun et al. previously determined the total flavonoid content of Chionanthus retusus flowers to be 10.7% [17]. Thus, in this

Contents of Total Phenols and Total Flavonoids in C. retusus Flowers
The contents of total phenols and flavonoids in the extract and fractions were determined as gallic acid and catechin equivalent values, respectively. As shown in Table 1, MeOH extract and EtOAc fraction (fr.) showed the highest contents compared to other fr.s. MeOH extract and EtOAc fr. showed a yellowish color on a thin-layer-chromatography (TLC) plate by spraying 10% H 2 SO 4 and baking (data not shown), suggesting the extract and EtOAc fr. to include high amounts of flavonoids.

Inhibition Effects of Flavonoids 1-12 on NO Production in Lipopolysaccharide (LPS)-Induced RAW 264.7 Cells
Oxidative stress is not only an important feature of several neurodegenerative processes, but also actively triggers intracellular signaling pathways that lead to cell death [28]. We first examined the viability of RAW264.7 cells treated with compounds 1-12 using a 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay. It did not show cytotoxicity or cellular proliferation when treated with compounds 1-4 and 6-12 at concentrations of 40 or 80 µM in RAW264.7 cells. However, compound 5 exhibited cytotoxic effects at 80 µM ( Figure 2a). To investigate the anti-inflammatory effects of compounds 1-12, we appreciated their inhibitory effects on NO production in LPS-induced RAW 264.7 cells. These cells were pretreated with flavonoids 1-12 and butein, a positive control, before one day LPS treatment. As shown in Table 2, compounds 1, 2, 5, and 6 highly inhibited NO production, while compounds 8 and 10-12 showed moderate inhibition effect. Flavonoids with a catechol structure in the B ring (1, 5, 6, 8, 11, and 12) exerted stronger anti-inflammatory effects than those with a phenol structure (3, 4, 7, 9, and 10). In addition, as the number of glucose moieties increased in compounds 1-6, the NO inhibitory effects of these compounds in RAW 264.7 cells decreased. However, compounds with glucopyranosyl moieties at C-7 (10 and 11) exhibited higher activity than aglycones (8 and 9). These results indicate that the presence of a catechol structure in the B ring and a glucopyranosyl moiety in the flavonoid structure were key factors of the antiinflammatory effects of these flavonoids.

Inhibition Effects of Flavonoids 1-12 on NO Production in Lipopolysaccharide (LPS)-Induced RAW 264.7 Cells
Oxidative stress is not only an important feature of several neurodegenerative processes, but also actively triggers intracellular signaling pathways that lead to cell death [28]. We first examined the viability of RAW264.7 cells treated with compounds 1-12 using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. It did not show cytotoxicity or cellular proliferation when treated with compounds 1-4 and 6-12 at concentrations of 40 or 80 µM in RAW264.7 cells. However, compound 5 exhibited cytotoxic effects at 80 µM ( Figure 2a). To investigate the anti-inflammatory effects of compounds 1-12, we appreciated their inhibitory effects on NO production in LPS-induced RAW 264.7 cells. These cells were pretreated with flavonoids 1-12 and butein, a positive control, before one day LPS treatment. As shown in Table 2, compounds 1, 2, 5, and 6 highly inhibited NO production, while compounds 8 and 10-12 showed moderate inhibition effect. Flavonoids with a catechol structure in the B ring (1, 5, 6, 8, 11, and 12) exerted stronger anti-inflammatory effects than those with a phenol structure (3, 4, 7, 9, and 10). In addition, as the number of glucose moieties increased in compounds 1-6, the NO inhibitory effects of these compounds in RAW 264.7 cells decreased. However, compounds with glucopyranosyl moieties at C-7 (10 and 11) exhibited higher activity than aglycones (8 and 9). These results indicate that the presence of a catechol structure in the B ring and a glucopyranosyl moiety in the flavonoid structure were key factors of the anti-inflammatory effects of these flavonoids. Table 2. IC 50 values of flavonoids 1-12 from C. retusus flowers on NO production in lipopolysaccharide (LPS)-induced RAW264.7 cells. The cells were pre-treated with each compound for 12 h, and then stimulated with LPS (1 µg/mL) for 18 h. The production of NO was determined as described in Section 3. Data shown represent the mean ± SD of three experiments. phenol structure (3, 4, 7, 9, and 10). In addition, as the number of glucose moieties increased in compounds 1-6, the cytoprotective effects of these compounds in HT22 cells decreased. However, compounds with glucopyranosyl moieties at C-7 (10 and 11) exhibited higher activity than aglycones (8 and 9). These results indicate that the presence of a catechol structure in the B ring and a glucopyranosyl moiety in the flavonoid structure were key determinants of the effects of these flavonoids on mouse hippocampal HT22 cells.

Effects of Flavonoids 1-12 on Glutamate-Induced Cell Toxicity in Mouse Hippocampal HT22 Cells
To investigate the protective effects of compounds 1-12 against glutamate-induced oxidative neuronal cell death, we also examined their effects on the viability of mouse hippocampal HT22 cells. To investigate the potential for cellular proliferation or cytotoxic effects of compounds 1-12, we first examined the viability of mouse hippocampal HT22 cells treated with compounds 1-12 using an MTT assay. No cytotoxic effects or cellular proliferation by compounds 1-12 were observed at concentrations <40 µM (Figure 2b). These cells were pretreated with compounds 1-12 at concentrations of 20 or 40 µM for 3 h and then were treated with glutamate and reacted for 12 h. Thereafter, cell viability was assessed with an MTT assay. None of the compounds exhibited toxicity at the highest concentration (40 µM). Compounds 1, 2, 5, 6, 8, 10, 11, and 12 significantly increased cell viability following glutamate treatment ( Figure 3). Butein derived from Rhus verniciflua, which is known to protect mouse hippocampal HT22 cells from glutamate-induced death [29], was used as a positive control and indeed exhibited cytoprotective effects ( Figure 3). Flavonoids with a catechol structure in the B ring (1, 5, 6, 8, 11, and 12) exerted stronger cytoprotective effects than those with a phenol structure (3, 4, 7, 9, and 10). In addition, as the number of glucose moieties increased in compounds 1-6, the cytoprotective effects of these compounds in HT22 cells decreased. However, compounds with glucopyranosyl moieties at C-7 (10 and 11) exhibited higher activity than aglycones (8 and 9). These results indicate that the presence of a catechol structure in the B ring and a glucopyranosyl moiety in the flavonoid structure were key determinants of the effects of these flavonoids on mouse hippocampal HT22 cells.

Effects of Compounds 1, 2, 5, 6, 8, and 10-12 on HO-1 Expression in Mouse Hippocampal HT22 Cells
Heme oxygenase (HO) is an important enzyme in the antioxidant cell system. HO-1, one of the HO derivatives, decomposes heme in the cell to produce carbon monoxide, iron, and biliverdin [30]. HO-1 expression has been reported to inhibit brain cell damage resulting from oxidative stress [31]. We examined whether compounds 1, 2, 5, 6, 8, and 10-12 affected the protein expression of HO-1, given their protection against glutamate-induced toxicity in mouse hippocampal HT22 cells. Mouse hippocampal HT22 cells were treated with compounds 1, 2, 5, 6, 8, and 10-12 at three concentrations (10, 20, and 40 µM) and then cultured for 12 h. Cobalt protoporphyrin (CoPP), a well-known HO-1 inducer, was used as a positive control. As shown in Figure 4, compounds 1, 2, 5, 6, 8, and 10-12 all increased HO-1 protein expression in a dose-dependent manner in mouse hippocampal HT22 cells. Flavonoid aglycones (1, 2, 5, and 8) exhibited higher activity than the glycosides (10-12). The flavonol and flavanonol with a catechol structure in the B ring (1 and 11) displayed stronger HO-1 expression than those with a phenol structure (2 and 10). Flavonoids with a hydroxy group at C-3 (8 and 11) exhibited weaker HO-1 expression than those without (5 and 12). In addition, a flavonoid with a double bond between C2 and C3 (1) was a weaker inhibitor of oxidative-stress-induced brain-cell damage than one with a single bond (8). These results indicate that the presence of a hydroxy group at C-3, the structure of the B ring and the type of C2-C3 bond are key determinants of the extent to which these flavonoids protect brain cells from damage due to oxidative stress. Heme oxygenase (HO) is an important enzyme in the antioxidant cell system. HO-1, one of the HO derivatives, decomposes heme in the cell to produce carbon monoxide, iron, and biliverdin [30]. HO-1 expression has been reported to inhibit brain cell damage resulting from oxidative stress [31]. We examined whether compounds 1, 2, 5, 6, 8, and 10-12 affected the protein expression of HO-1, given their protection against glutamate-induced toxicity in mouse hippocampal HT22 cells. Mouse hippocampal HT22 cells were treated with compounds 1, 2, 5, 6, 8, and 10-12 at three concentrations (10, 20, and 40 µM) and then cultured for 12 h. Cobalt protoporphyrin (CoPP), a well-known HO-1 inducer, was used as a positive control. As shown in Figure 4, compounds 1, 2, 5, 6, 8, and 10-12 all increased HO-1 protein expression in a dose-dependent manner in mouse hippocampal HT22 cells. Flavonoid aglycones (1, 2, 5, and 8) exhibited higher activity than the glycosides (10)(11)(12). The flavonol and flavanonol with a catechol structure in the B ring (1 and 11) displayed stronger HO-1 expression than those with a phenol structure (2 and 10). Flavonoids with a hydroxy group at C-3 (8 and 11) exhibited weaker HO-1 expression than those without (5 and 12). In addition, a flavonoid with a double bond between C2 and C3 (1) was a weaker inhibitor of oxidative-stress-induced brain-cell damage than one with a single bond (8). These results indicate that the presence of a hydroxy group at C-3, the structure of the B ring and the type of C2-C3 bond are key determinants of the extent to which these flavonoids protect brain cells from damage due to oxidative stress.

Plant Materials
The flowers of C. retusus Lindl. And Paxton were gathered near Kyung Hee University, Yong-In, South Korea, in August 2014, and were identified by Prof. Dae-Keun Kim, College of Pharmacy, Woosuk University, Jeonju, South Korea. A voucher specimen (KHU-NPCL-201408) has been deposited at the Natural Products Chemistry Laboratory, Kyung Hee University.

General Experimental Procedures
The equipment and chemicals used to isolate and identify flavonoids from C. retusus flowers and evaluate their neuroprotective activity were obtained from the literature [32][33][34][35].

Plant Materials
The flowers of C. retusus Lindl. And Paxton were gathered near Kyung Hee University, Yong-In, South Korea, in August 2014, and were identified by Prof. Dae-Keun Kim, College of Pharmacy, Woosuk University, Jeonju, South Korea. A voucher specimen (KHU-NPCl-201408) has been deposited at the Natural Products Chemistry Laboratory, Kyung Hee University.

General Experimental Procedures
The equipment and chemicals used to isolate and identify flavonoids from C. retusus flowers and evaluate their neuroprotective activity were obtained from the literature [32][33][34][35].

Cell Culture and MTT Assay
Mouse hippocampal HT22 cells were donated by Wonkwang University, Iksan, Korea (Prof. Youn-Chul Kim). Cytoprotective activity assay was performed, as per the previously described method [35]. Cell viability was evaluated using the MTT assay reported in the literature [36].
3.5. Macrophage RAW 264.7 Culture, Viability Assay, and NO Measurement Macrophage RAW 264.7 culture, viability assay, and NO measurement were carried out as per the previously described method [35].

Determination of Total Phenols and Flavonoids Contents in C. retusus Flower
Determination of the total phenolic and flavonoid contents of C. retusus flower was carried out as per the previously described method [37].

Western Blot Analysis
Pelleted HT22 cells were washed with PBS and lysed with an RIPA buffer from Sigma Chemical Co. The same amount of protein from each sample was mixed into a sample loading buffer, subjected to SDS-PAGE, and transferred to a membrane.

Statistical Analysis
Statistical analysis was performed with GraphPad Prism 5 software (ver. 3.03, San Diego, CA, USA). Data are presented as the mean ± standard deviation of 3 independent experiments. The mean differences were derived using one-way ANOVA and Tukey's multiple comparison test, and statistical significance was defined as p < 0.05, p < 0.01, and p < 0.001.

Conclusions
In conclusion, four flavonols (1-4), three flavones (5-7), four flavanonols (8)(9)(10)(11), and one flavanone (12) were isolated from C. retusus flowers. Flavonoids 4-6 and 8-11 were isolated from the flowers of C. retusus for the first time in this study. Flavonoids 1, 2, 5, 6, 8, and 10-12 exhibited significant anti-inflammatory and neurocytoprotective activity, and effectively increased HO-1 protein expression. The flavonoids that displayed significant neuroprotective activity were found to recover oxidative stress-induced cell damage by increasing HO-1 protein expression. The relationships between the structural characteristics of these flavonoids and their anti-inflammatory and neuroprotective activity were revealed. Further studies are needed to investigate the potential therapeutic effects of flavonoids in innovative anti-inflammatory and neuroprotective strategies.

Conflicts of Interest:
The authors declare no conflict of interest.