Synthesis and Antioxidant/Anti-Inflammatory Activity of 3-Arylphthalides

Phthalides are a group of compounds with relevant biological activities in different areas such as cytotoxicity, anti-stroke activity, neuroprotection, and inflammation, among others. In this study we designed and synthesized a series of 3-arylphthalide derivatives in order to identify their antioxidant and anti-inflammatory activities. The synthetic methodology was established in terms of atom and step economy through a dehydrative coupling reaction between 3-hydroxyphthalide and different properly functionalized arene rings. The evaluation of the antioxidant activity was performed by the ABTS assay and for the anti-inflammatory activity the inhibition of LPS-induced nitric oxide (NO) production in microglial cells Bv.2 and macrophage cells RAW 264.7 was measured. The synthesized compound 3-(2,4-dihydroxyphenyl)phthalide (5a) showed better antioxidant activity than the Trolox standard and caused strong inhibition of NO production in LPS-stimulated Bv.2 and RAW 264.7 cells. In addition, compound 5a reduced the expression of the pro-inflammatory cytokines Il1b and Il6 in RAW 264.7 cells. These results, which are the first account of the anti-inflammatory activity of 3-arylphthalides, suggest that compound 5a could be a promising candidate for more advanced anti-inflammatory studies.


Introduction
Naturally occurring phthalides are a relatively small group of metabolites produced by a variety of plants, including species worldwide used in traditional medicine, and also by fungi from terrestrial and marine environments [1]. Structurally, these natural products are characterized by the presence of a 1(3H)-isobenzofuranone nucleus (phthalide), the substitution patterns of which confer wide structural diversity on this family of compounds ( Figure 1). Thus, the aromatic ring of the phthalide often contains hydroxy or alkoxy groups, although sugars and moieties of terpenoid or alkaloid origin may also be present. With regard to the γ-lactone ring, most natural phthalides are substituted at C-3, exhibiting alkyl chains, spirocycles, and aromatic rings either alone or as part of more complex substructures. In addition, compounds with the aromatic ring partially reduced and dimeric phthalides have also been described as natural products [1,2].
Phthalides have been the subject of numerous pharmacological studies which have shown a wide range of properties, such as cytotoxic, antifungal, antibacterial, anti-inflammatory and anti-stroke activities, among others, in this class of natural products [1][2][3]. Two remarkable compounds are mycophenolic acid and 3-n-butylphthalide (NBP) (Figure 1). Mycophenolic acid, first isolated from Penicillium stoloniferum, [4] is used as an immunosuppressive drug in organ transplantation [5]. NBP, firstly isolated from Angelica sinenxis Radix [6], was Over the last decades, the study of inflammatory processes has become a m of research due to the close relationship between inflammation and the develop serious diseases such as certain types of cancer, inflammatory bowel disease, d and arthritis, among others [8,9]. Moreover, inflammation is related to oxidativ since reactive oxygen species (ROS) are potent inflammatory mediators [10]. In tion is usually treated with non-steroidal anti-inflammatory drugs that, after a pr use, often lead to undesirable side effects [11]. Therefore, novel therapeutic age anti-inflammatory properties and minimal side effects are needed. In this context, studies have described the anti-inflammatory activity of several natural phthalide studies have focused on (Z)-ligustilide ( Figure 1) [12][13][14][15][16][17][18], cnidilide [19] and relat pounds and dimers, as well as some semi-synthetic derivatives [12,18,[20][21][22][23]. It noting that all these compounds are phthalides bearing one or two alkyl residue while no data on 3-aryl derivatives have been found in the literature.
On this basis, in the course of our research on anti-inflammatory natural p we turned our attention towards phthalides bearing at C-3 an aromatic ring, isopestacin ( Figure 1) [24]. As the first approach to exploring the anti-inflamma tential of 3-arylphthalides, we undertook the synthesis of a series of phthalides sub at C-3 with benzene rings bearing oxygenated, halogenated, and/or sulfur functi tion. In this regard, functional groups such as hydroxy or alkyloxy groups can important influence on antioxidant activity [25], and sulfur-containing drugs h exhibited good efficiency in several aspects, including antioxidant, anti-free-radi neuroprotective activities, among others [26]. Then, the synthesized compoun tested in antioxidant and anti-inflammatory assays.

Synthesis of 3-Arylphthalides
The important role of phthalides, both in total synthesis and pharmaceutical try, has attracted attention of chemists and several synthetic strategies have bee oped [1,3,[27][28][29]. The strategies first described were based preferentially on th thetic aspects: the construction of the γ-lactone on a six carbon cycle, the constru the benzene system on a pre-existing γ-lactone, or the concomitant formation Over the last decades, the study of inflammatory processes has become a main area of research due to the close relationship between inflammation and the development of serious diseases such as certain types of cancer, inflammatory bowel disease, diabetes, and arthritis, among others [8,9]. Moreover, inflammation is related to oxidative stress since reactive oxygen species (ROS) are potent inflammatory mediators [10]. Inflammation is usually treated with non-steroidal anti-inflammatory drugs that, after a prolonged use, often lead to undesirable side effects [11]. Therefore, novel therapeutic agents with anti-inflammatory properties and minimal side effects are needed. In this context, various studies have described the anti-inflammatory activity of several natural phthalides. These studies have focused on (Z)-ligustilide ( Figure 1) [12][13][14][15][16][17][18], cnidilide [19] and related compounds and dimers, as well as some semi-synthetic derivatives [12,18,[20][21][22][23]. It is worth noting that all these compounds are phthalides bearing one or two alkyl residues at C-3, while no data on 3-aryl derivatives have been found in the literature.
On this basis, in the course of our research on anti-inflammatory natural products, we turned our attention towards phthalides bearing at C-3 an aromatic ring, such as isopestacin ( Figure 1) [24]. As the first approach to exploring the anti-inflammatory potential of 3-arylphthalides, we undertook the synthesis of a series of phthalides substituted at C-3 with benzene rings bearing oxygenated, halogenated, and/or sulfur functionalization. In this regard, functional groups such as hydroxy or alkyloxy groups can exert an important influence on antioxidant activity [25], and sulfur-containing drugs have also exhibited good efficiency in several aspects, including antioxidant, anti-free-radical, and neuroprotective activities, among others [26]. Then, the synthesized compounds were tested in antioxidant and anti-inflammatory assays.

Synthesis of 3-Arylphthalides
The important role of phthalides, both in total synthesis and pharmaceutical chemistry, has attracted attention of chemists and several synthetic strategies have been developed [1,3,[27][28][29]. The strategies first described were based preferentially on three synthetic aspects: the construction of the γ-lactone on a six carbon cycle, the construction of the benzene system on a pre-existing γ-lactone, or the concomitant formation of both rings, the benzene and the γ-lactone rings [3]. More recently, other strategies have been developed, searching either one-pot multicomponent reaction (MCRs) versions as powerful tools for the rapid assembly of molecules [30] or innovative C-H bond Pharmaceuticals 2022, 15, 588 3 of 15 functionalization using simple and unactivated starting materials with the achievement of atom economy and molecular diversity [29].
The synthetic strategy herein, proposed to obtain 3-substituted phthalide derivatives, is established in terms of atom and step economy through a dehydrative coupling reaction in which a C-H and a C-OH bonds react to form a C-C bond with elimination of a water molecule [31]. In particular, the direct condensation of 3-hydroxyphthalide with different arene rings under acid catalysis is established. The synthetic route for the preparation of 3-arylphthalide derivatives is outlined in Figure 2. developed, searching either one-pot multicomponent reaction (MCRs) versions as powerful tools for the rapid assembly of molecules [30] or innovative C-H bond functionalization using simple and unactivated starting materials with the achievement of atom economy and molecular diversity [29]. The synthetic strategy herein, proposed to obtain 3-substituted phthalide derivatives, is established in terms of atom and step economy through a dehydrative coupling reaction in which a C-H and a C-OH bonds react to form a C-C bond with elimination of a water molecule [31]. In particular, the direct condensation of 3-hydroxyphthalide with different arene rings under acid catalysis is established. The synthetic route for the preparation of 3-arylphthalide derivatives is outlined in Figure 2. Following this, the use of arenes 4a-4f allowed obtaining the 3-arylphthalides 5a-5g ( Figure 3).  Following this, the use of arenes 4a-4f allowed obtaining the 3-arylphthalides 5a-5g ( Figure 3). developed, searching either one-pot multicomponent reaction (MCRs) versions as powerful tools for the rapid assembly of molecules [30] or innovative C-H bond functionalization using simple and unactivated starting materials with the achievement of atom economy and molecular diversity [29]. The synthetic strategy herein, proposed to obtain 3-substituted phthalide derivatives, is established in terms of atom and step economy through a dehydrative coupling reaction in which a C-H and a C-OH bonds react to form a C-C bond with elimination of a water molecule [31]. In particular, the direct condensation of 3-hydroxyphthalide with different arene rings under acid catalysis is established. The synthetic route for the preparation of 3-arylphthalide derivatives is outlined in Figure 2. Following this, the use of arenes 4a-4f allowed obtaining the 3-arylphthalides 5a-5g ( Figure 3).  Synthesis starts with the commercial phthalide (1) that was subjected to benzylic radical bromination with N-bromosuccinimide (NBS), yielding quantitatively 3-bromophthalide (2). Compound 2 was treated with aqueous KOH under reflux, yielding 3-hydroxyphthalide (3). Then, the treatment of 3 with acidic conditions led to the corresponding electrophilic phthalidyl ion that underwent condensation with arenes 4a-4f, giving rise to the 3-arylphthalide derivatives 5a-5f, respectively. The reaction conditions involved the use of H 2 SO 4 /H 2 O (3:7) mixtures, except for the synthesis of 5a, which was accomplished with dioxane/H 2 O (1:4, ca. 5% HCl) ( Figure 3). The condensation reaction for the synthesis of compounds 5a-5f proceeded smoothly and with good efficiency. Thus, although compound 5a was previously synthesized from o-phthalaldehydic acid with a 75% yield [32], our procedure proved to be more efficient, yielding 5a quantitatively. Compounds 5b, 5c, and 5d were also efficiently prepared, with yields of 87%, 96%, and 90%, respectively, while compounds 5e and 5f were obtained in lower yields (60% and 50%, respectively). These results also demonstrated excellent functional group compatibility.
Thus, the condensation of 3 with the commercially available compounds resorcinol (4a) and guaiacol (4b) yielded phthalides 5a and 5b, respectively. For the synthesis of compounds 5c-5f, it was necessary to prepare the arenes 4c-4f, respectively ( Figure 4). Thus, compounds 4c and 4d were obtained from the commercial compounds 6a and 6b by treatment with PPh 3 and CBr 4 under Apple conditions, causing the displacement of the hydroxy group and the installation of a halogen on the side chain. Demethylation of compounds 4c and 4d with EtSNa led to the corresponding hydroxy groups along with the concomitant nucleophilic substitution of the bromine atom by the ethylthio group on the alkyl side chain. On the other hand, the phthalide 5g was synthesized by the treatment of 5d with EtSNa ( Figure 3). The structures of all target compounds were characterized by extensive spectroscopic analyses (NMR and IR) and HRMS-ESI. Synthesis starts with the commercial phthalide (1) that was subjected to benzylic radical bromination with N-bromosuccinimide (NBS), yielding quantitatively 3-bromophthalide (2). Compound 2 was treated with aqueous KOH under reflux, yielding 3-hydroxyphthalide (3). Then, the treatment of 3 with acidic conditions led to the corresponding electrophilic phthalidyl ion that underwent condensation with arenes 4a-4f, giving rise to the 3-arylphthalide derivatives 5a-5f, respectively. The reaction conditions involved the use of H2SO4/H2O (3:7) mixtures, except for the synthesis of 5a, which was accomplished with dioxane/H2O (1:4, ca. 5% HCl) ( Figure 3). The condensation reaction for the synthesis of compounds 5a-5f proceeded smoothly and with good efficiency. Thus, although compound 5a was previously synthesized from o-phthalaldehydic acid with a 75% yield [32], our procedure proved to be more efficient, yielding 5a quantitatively. Compounds 5b, 5c, and 5d were also efficiently prepared, with yields of 87%, 96%, and 90%, respectively, while compounds 5e and 5f were obtained in lower yields (60% and 50%, respectively). These results also demonstrated excellent functional group compatibility.
Thus, the condensation of 3 with the commercially available compounds resorcinol (4a) and guaiacol (4b) yielded phthalides 5a and 5b, respectively. For the synthesis of compounds 5c-5f, it was necessary to prepare the arenes 4c-4f, respectively ( Figure 4). Thus, compounds 4c and 4d were obtained from the commercial compounds 6a and 6b by treatment with PPh3 and CBr4 under Apple conditions, causing the displacement of the hydroxy group and the installation of a halogen on the side chain. Demethylation of compounds 4c and 4d with EtSNa led to the corresponding hydroxy groups along with the concomitant nucleophilic substitution of the bromine atom by the ethylthio group on the alkyl side chain. On the other hand, the phthalide 5g was synthesized by the treatment of 5d with EtSNa ( Figure 3). The structures of all target compounds were characterized by extensive spectroscopic analyses (NMR and IR) and HRMS-ESI. The synthesized phthalides 5a-5g contain a stereogenic center at C-3, and due to the lack of stereochemical control in the synthetic methodology used, these compounds were obtained as racemic mixtures. Biological studies described for various 3-substituted phthalides have shown that the configuration at C-3 does not seem to be determinant for bioactivity [33]. Nonetheless, we were encouraged to evaluate, if possible, phthalides 5a-5g as pure enantiomers. We attempted to obtain optically pure enantiomers of 5a by chiral HPLC, but none of the chromatographic conditions used led to the separation of the enantiomers. In all cases, low resolution chromatograms were obtained, which suggested an inherent racemization process, likely due to the γ-lactone ring opening favored by the ability of the phenolic ring to form an o-or p-quinone methide ( Figure S15). Considering the pitfall associated with the use of chiral HPLC in 5a, we investigated other alternatives, such as the protection of the hydroxy groups by esterification, which could preclude the formation of quinone methide structures and thus racemization at C-3. In fact, a racemic mixture of esters of 5a was successfully separated by chiral HPLC into two enantiomers. However, when each enantiomerically pure ester was subjected to hydrolysis to remove the ester group, only the racemic compound 5a was recovered. These results again suggested a rapid racemization of pure enantiomers of 5a. On the other hand, we also subjected compound 5f to chiral HPLC, which in this case led to the corresponding pure enantiomers. However, the value of the specific optical rotations of these enantiomers [α](ca. The synthesized phthalides 5a-5g contain a stereogenic center at C-3, and due to the lack of stereochemical control in the synthetic methodology used, these compounds were obtained as racemic mixtures. Biological studies described for various 3-substituted phthalides have shown that the configuration at C-3 does not seem to be determinant for bioactivity [33]. Nonetheless, we were encouraged to evaluate, if possible, phthalides 5a-5g as pure enantiomers. We attempted to obtain optically pure enantiomers of 5a by chiral HPLC, but none of the chromatographic conditions used led to the separation of the enantiomers. In all cases, low resolution chromatograms were obtained, which suggested an inherent racemization process, likely due to the γ-lactone ring opening favored by the ability of the phenolic ring to form an oor p-quinone methide ( Figure S15). Considering the pitfall associated with the use of chiral HPLC in 5a, we investigated other alternatives, such as the protection of the hydroxy groups by esterification, which could preclude the formation of quinone methide structures and thus racemization at C-3. In fact, a racemic mixture of esters of 5a was successfully separated by chiral HPLC into two enantiomers. However, when each enantiomerically pure ester was subjected to hydrolysis to remove the ester group, only the racemic compound 5a was recovered. These results again suggested a rapid racemization of pure enantiomers of 5a. On the other hand, we also subjected compound 5f to chiral HPLC, which in this case led to the corresponding pure enantiomers. However, the value of the specific optical rotations of these enantiomers [α](ca. +10.0 • and −9.5 • ) were lower than those of structurally similar compounds described in reference [34]. Moreover, we noticed that the optical rotation decreased with time until turning null in a one-hour time period. Similar results were also obtained with 5e.
Interestingly, most natural 3-arylphthalides which bear hydroxy groups in the aryl ring at C-3 have been described as racemates [24] or with unknown stereochemistry at C-3 [35,36], suggesting the racemization of these compounds. Since enzymes act in stereospecific routes leading to pure enantiomers, it is plausible to conclude that the racemization observed in some natural 3-arylphthalides is produced in a post-biosynthetic process. It is also worth noting that although several asymmetric syntheses of 3-arylphthalides can be found in literature [34,[37][38][39][40][41][42][43][44], none of them include in their scope the synthesis of compounds bearing phenolic rings at C-3. The easy racemization of 3-arylphthalides which bear ortoor para-hydroxy groups in the aryl ring at C-3 has been explained in terms of the formation at C-3 of a cationic intermediate stabilized by the ability to form quinone methide structures under almost neutral conditions [45].
Once established that racemization in solution precluded the evaluation of the biological activity of the pure enantiomers of 5a, 5e, and 5f, we decided to perform the activity assays of all synthesized compounds 5a-5g as racemates.

Antioxidant Activity
The in vitro antioxidant properties of compounds 5a-5g were assessed by using the ABTS assay and Trolox as positive control. Compounds 5c, 5d, and 5g were inactive while compounds 5a, 5b, 5e, and 5f displayed significant activities. Compound 5a exhibited slightly better activity than the Trolox standard, while the activity values of 5b, 5e, and 5f were 57%, 43%, and 56% that of the Trolox, respectively (Table 1). As expected, only compounds bearing hydroxy groups on the aromatic ring at C-3 (5a, 5b, 5e, 5f) displayed significant antioxidant activity. Moreover, the results showed that the number of hydroxy groups and their relative position on the benzene ring are crucial for the activity. The antioxidant properties of compounds exhibiting free phenolic groups is well-known, and in particular, the radical-scavenging activity of the natural 3-arylphthalide isopestacin was already reported [24]. Herein, we have shown the activity of the related analogue 5a. On the other hand, it has been described that the introduction of methoxy groups can increase the antioxidant activity in simple phenolic compounds and in compounds having conjugated systems such as stilbenes and flavonoids [25]. Compound 5b exhibited higher antioxidant activity than 5e, which presents the same relative position of the oxygenated substituents but lacks a methoxy group. Although sulfur-containing compounds have been described to reduce oxidative stress, our results are not conclusive [26].

Anti-Inflammatory Activity
The anti-inflammatory activity of the synthesized compounds 5a-5g was tested in assays aimed to detect the inhibition of nitric oxide (NO) production in both immune cell line Bv.2 (microglia) and RAW 264.7 (macrophages) cells. For the more active compounds, the expression of the pro-inflammatory cytokines Tnfa, Il1b, and Il6 was also tested.
First, to ensure the safety of the compounds and obtain reliable results in the antiinflammatory evaluation, the cytotoxicity of the compounds on Bv.2 and RAW 264.7 cells was checked (Figures S16 and S17). We found that compounds 5a-5g exhibited no cytotoxicity at concentrations equal to or below 10 µM. This concentration was selected to evaluate their anti-inflammatory activity. For the anti-inflammatory assays, Bv.2 and RAW 264.7 cells were stimulated with lipopolysaccharide (LPS) to trigger the release of inflammatory mediators such as NO and pro-inflammatory cytokines, which cause local inflammation upon binding to membrane receptors of cells.
In this study, cells were cotreated with compounds 5a-5g and LPS. The concentration of nitrite (NO 2 − ), which is one of the major metabolites derived from NO, was measured. The level of nitrites did not change in Bv.2 and RAW 264.7 cells treated with the compounds at 10 µM (data not shown). However, the level of NO production was significantly increased in cells stimulated with LPS. When these cells were cotreated with LPS and compounds 5a-5g the production of nitrites was significantly inhibited (Figures 5 and 6).
For the anti-inflammatory assays, Bv.2 and RAW 264.7 cells were stimulated lipopolysaccharide (LPS) to trigger the release of inflammatory mediators such as NO pro-inflammatory cytokines, which cause local inflammation upon binding to memb receptors of cells.
In this study, cells were cotreated with compounds 5a-5g and LPS. The concentr of nitrite (NO2 − ), which is one of the major metabolites derived from NO, was measu The level of nitrites did not change in Bv.2 and RAW 264.7 cells treated with the pounds at 10 μM (data not shown). However, the level of NO production was signific increased in cells stimulated with LPS. When these cells were cotreated with LPS compounds 5a-5g the production of nitrites was significantly inhibited (Figures 5 an In the case of Bv.2 cells (Figure 5), the most active compounds were 5a and 5e, cau 79.84% and 45.82% inhibition, respectively, in nitrite production with respect to LPSulated but non-treated cells. Compounds 5b, 5f, and 5g were less active showing in tion of NO production below 23%, while 5c and 5d showed no inhibition.   Nitrite accumulation in the culture media was measured using the Griess reagent. Results were expressed as a fold change relative to the LPS condition and mean ± SD (n ≥ 3 independent experiments performed in duplicate). Significant differences were determined by one-way ANOVA followed by Bonferroni t test; * p ≤ 0.05 vs. LPS.
For the anti-inflammatory assays, Bv.2 and RAW 264.7 cells were stimulated lipopolysaccharide (LPS) to trigger the release of inflammatory mediators such as NO pro-inflammatory cytokines, which cause local inflammation upon binding to memb receptors of cells.
In this study, cells were cotreated with compounds 5a-5g and LPS. The concentra of nitrite (NO2 − ), which is one of the major metabolites derived from NO, was measu The level of nitrites did not change in Bv.2 and RAW 264.7 cells treated with the c pounds at 10 μM (data not shown). However, the level of NO production was significa increased in cells stimulated with LPS. When these cells were cotreated with LPS compounds 5a-5g the production of nitrites was significantly inhibited (Figures 5 an In the case of Bv.2 cells (Figure 5), the most active compounds were 5a and 5e, cau 79.84% and 45.82% inhibition, respectively, in nitrite production with respect to LPS-s ulated but non-treated cells. Compounds 5b, 5f, and 5g were less active showing in tion of NO production below 23%, while 5c and 5d showed no inhibition. Figure 5. Effects of compounds 5a-5g on NO release in microglial cells. Bv.2 microglial cells cotreated with the compound at 10 μM and 200 ng/mL LPS for 24 h. Nitrite accumulation i culture media was measured using the Griess reagent. Results were expressed as a fold chang ative to the LPS condition and mean ± SD (n ≥ 3 independent experiments performed in dupli Significant differences were determined by one-way ANOVA followed by Bonferroni t test; 0.05 vs. LPS. Figure 6. Effects of compounds 5a-5g on NO release in macrophage cells. RAW 264.7 cells cotreated with the compound at 10 μM and 200 ng/mL LPS for 24 h. Nitrite accumulation i culture media was measured using the Griess reagent. Results were expressed as a fold chang ative to the LPS condition and mean ± SD (n ≥ 3 independent experiments performed in dupli Significant differences were determined by one-way ANOVA followed by Bonferroni t test; 0.05 vs. LPS. Figure 6. Effects of compounds 5a-5g on NO release in macrophage cells. RAW 264.7 cells were cotreated with the compound at 10 µM and 200 ng/mL LPS for 24 h. Nitrite accumulation in the culture media was measured using the Griess reagent. Results were expressed as a fold change relative to the LPS condition and mean ± SD (n ≥ 3 independent experiments performed in duplicate). Significant differences were determined by one-way ANOVA followed by Bonferroni t test; * p ≤ 0.05 vs. LPS. In the case of Bv.2 cells (Figure 5), the most active compounds were 5a and 5e, causing 79.84% and 45.82% inhibition, respectively, in nitrite production with respect to LPSstimulated but non-treated cells. Compounds 5b, 5f, and 5g were less active showing inhibition of NO production below 23%, while 5c and 5d showed no inhibition.
Similar results were obtained in assays with RAW 264.7 cells (Figure 6). After the cotreatment with LPS and compounds 5a and 5e the production of nitrites was significantly inhibited by 76.31% and 86.47%, respectively, with respect to LPS-stimulated but nontreated cells. Compounds 5b, 5d, 5f, and 5g again showed less activity with NO inhibition levels below 30%, while compound 5c did not show NO inhibition.
Although the number of compounds is limited, the results strongly suggest that the presence of two hydroxy groups on the aromatic ring at C-3 of the phthalide correlates with anti-inflammatory activity, while the presence of lateral side-chains with either bromine or sulfur seems to lack of significance for the inhibitory activity. Thus, compounds 5a and 5e, with two OH substituents on the C-3 aryl ring, caused significant inhibition of LPS-induced NO production in both types of cells. There were significant differences in the effects caused by 5a (o-and p-OH) and 5e (m-and p-OH) on Bv.2 cells, suggesting that the position of the hydroxy groups is a limiting factor for determining the NO inhibition. However, this trend was not maintained in RAW 264.7 cells, where compounds 5a and 5e showed similar levels of NO inhibition. Compounds 5b and 5f, which only have one hydroxy group, were less active than 5a and 5e, and the position of the hydroxy group (p-OH for 5b and o-OH for 5f) seems irrelevant. Compounds 5c and 5d lacking of hydroxy groups did not show apparent NO inhibitory activity in Bv.2 and RAW 264.7 cells, while 5g only showed weak NO inhibitory activity.
Our study describes for the first time the anti-inflammatory activity of 3-arylphthalides. Previous studies have reported the capability of (Z)-ligustilide ( Figure 1) [14,16], the 3-alkylphthalide cnidilide [19] and a few related compounds [20,21,23] to inhibit LPSinduced NO production in RAW 264.7 macrophages. However, most of these compounds only at concentrations higher than 50 µM caused inhibitions higher than 50%, and for ligustilide IC 50 values of 32.3 µM [14] or 8.45 µM [16] have been reported. These activity levels are significantly lower than those found herein for compound 5a.
According to these results, compound 5a combined the highest antioxidant activity in the ABTS assay and the highest potency on the inhibition of NO production. Therefore, compound 5a was selected to evaluate its effects in the inhibition of the pro-inflammatory cytokine expression in RAW 264.7 cells. LPS-induced inflammatory response is characterized by releasing of pro-inflammatory cytokines such as TNF-α IL-1β, and IL-6. These cytokines play crucial roles during inflammation and are recognized as important early inflammatory mediators, and their over-expression can lead to strong inflammatory reactions. Quantitative RT-PCR analysis revealed a great increase in the mRNA expression of Il1b and Il6 in RAW 264.7 cells after LPS stimulus (Figure 7). However, when cells were cotreated with compound 5a at 10 µM, the expression of Il1b and Il6 decreased by 71.38% and 82.64%, respectively, while no effect was observed on Tnfa. The different behavior of compound 5a in the downregulation of the cytokines opens the way to forthcoming studies on the mechanism of action by which compound 5a exerts its anti-inflammatory activity.
These data suggest that compound 5a could have therapeutic benefit in a variety of pathological conditions in which there is an excessive increase in ROS and pro-inflammatory cytokines. Small molecules such as 5a do not block the normal actions of cytokines, they only reduce their expression and are less likely to cause the adverse immune effects of the cytokine antagonists [46,47]. Il1b and Il6 in RAW 264.7 cells after LPS stimulus (Figure 7). However, when cells wer cotreated with compound 5a at 10 μM, the expression of Il1b and Il6 decreased by 71.38% and 82.64%, respectively, while no effect was observed on Tnfa. The different behavior o compound 5a in the downregulation of the cytokines opens the way to forthcoming stud ies on the mechanism of action by which compound 5a exerts its anti-inflammatory activ ity. Results were expressed as a fold change relative to the LPS condition and mean ± SD (n ≥ 3 independent experiments performed in duplicate). Significant differences were determined by one-way ANOVA followed by Bonferroni t-test; * p ≤ 0.05 vs. LPS.
Optical rotations were measured on a Jasco P-2000 polarimeter (Jasco, Easton, MD, USA), operating at the sodium D line with a 100 mm path length cell. Infrared spectra (IR) were recorded on a Perkin-Elmer FT/IR spectrometer (Perkin Elmer, Boston, MA, USA). Frequencies are given as wavenumbers in cm -1 . 1 H and 13 C NMR spectra were recorded on Agilent 400 or Agilent 500 spectrometers (Agilent Technologies, Santa Clara, CA, USA) or Bruker Avance NEO 400 (Bruker BioSpin GmbH, Rheinstetten, Germany) using CD 3 OD, CDCl 3, or CD 3 COCD 3 as solvents. Chemical shifts were referenced using the corresponding solvent signals (δ H 3.30 and δ C 49.0 for CD 3 OD; δ H 7.26 and δ C 77.0 for CDCl 3 ; δ H 2.04 and δ C 29.9 for CD 3 COCD 3 ). COSY, HSQC, HMBC, and NOESY experiments were performed using standard Agilent pulse sequences. High resolution mass spectra (HRMS) were obtained on a Waters XEVO G2-S Mass spectrometer (Waters, Milford, MA, USA).

Synthesis of the Aromatic Derivatives 4c-4f
Synthesis of the Bromoderivatives 4c and 4d To a solution of 3.3 mmol of the alcohol 2-(3,4-dimethoxyphenyl)ethanol (6a) or 2-(4methoxyphenyl)ethanol (6b) and 1.17 g of PPh 3 (4.46 mmol) in 20 mL of CH 2 Cl 2 at 0 • C, and 1.22 g of CBr 4 (3.63 mmol) were added. The resulting mixture was stirred at rt for 2 h and then concentrated under reduced pressure to give a residue that was purified by column chromatography (SiO 2 , 2 × 17 cm, hexanes/AcOEt 7:3) to yield the corresponding bromoderivatives 4c and 4d in quantitative yield.

Synthesis of 4e and 4f
A solution of bromoderivative 4c or 4d (0.24 mmol) in 1.5 mL of DMF was added to a round bottom flask containing 207 mg of NaEtS (2.4 mmol) under inert atmosphere. After heating under reflux for 5 h, the reaction was cooled to 0 • C and added to a solution of HCl 5%. Then, the mixture was treated with AcOEt (2 × 5 mL). The organic layers combined, dried over anhydrous MgSO 4 , and the solvent concentrated under reduced pressure, yielding a residue that was purified by column chromatography (SiO 2 , 2 × 16 cm, hexanes/AcOEt (7:3) for compound 4e and hexanes/AcOEt (9:1) for compound 4f), to yield 80-90% of the corresponding demethylated compounds.
3-(4-hydroxy-3-methoxyphenyl)phthalide (5b): 1     and of the tested compounds were prepared in EtOH. For the assay, 100 µL of the Trolox solution or 100 µL of tested compound solution were mixed with 2 mL of the ABTS +• solution. The final concentrations of Trolox or tested compounds were 1, 5, 10, 20, 25, and 30 µM. Controls were prepared by adding 100 µL of EtOH to 2 mL of ABTS +• solution. Six minutes after mixing, the absorbance at 734 nm was measured after 6 min in a UV-Vis spectrophotometer, VWR UV-1600PC (VWR, Radnor, PA, USA). All determinations were carried out in triplicate. The percentage of inhibition of the absorbance was calculated by the following equation: % Inhibition = [(A0 − A1)/A0] × 100, where A0 expresses the absorbance of control and A1 the absorbance of the tested compound. ). An amount of 1.5 × 10 5 cells/well was seeded in a 6-multiwell plate (Sarstedt, Germany). The culture conditions were 37 • C in a humidified atmosphere with 5% CO 2 in RPMI supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS), 1% (v/v) penicillin/streptomycin (Sigma), and 2 mM L-glutamine (Gibco, Carlsbad, CA, USA). All experimental cell approaches were performed in complete medium without FBS.

Analysis of the Cellular Viability by Crystal Violet Staining
Cells were cultured in 24-well plates and grown up to 70% confluence. The cells were treated with solutions of the compounds to reach final concentrations of 0.1, 1.0, 10.0, 25.0, and 50.0 µM and incubated in serum-free medium. After 24 h, the medium was discarded and cells were fixed by adding 0.5 mL of glutaraldehyde 1% (v/v) for 30 min. Then, the plates were rinsed with phosphate buffer saline (PBS) and the remaining viable adherent cells were stained with crystal violet 0.1% (w/v) for 30 min. After rinsing plates with water and drying for 24 h, 0.5 mL of acetic acid 10% (v/v) was added. The absorbance of each plate was read spectrophotometrically at 590 nm in a microplate reader (Power Wave, Bioteck, Torino, Italy).

Analysis of Nitrites (NO 2 − )
Cells were cultured in 6-well plates and grown up to 70% confluence. The cells were co-treated with lipopolysaccharide (LPS, 200 ng/mL) and the compounds at 10 µM in serum-free medium for 24 h. Dexamethasone (Dx) was used as positive control at 2.5 µM. At this concentration of Dx, the NO production in LPS-stimulated cells decreased to the basal values. After cell treatments, levels of NO 2 − were measured by using the Griess reagent [49]. Briefly, cell cultured medium was treated with an acid solution containing 1% sulfanilamide and 0.1% N-(1-naphthyl)ethylenediamine (NEDA) and read spectrophotometrically at 548 nm in a microplate reader.

Quantitative Real-Time PCR (qPCR) Analysis
Total RNA was extracted with TRI ® reagent (Sigma, Madrid, Spain) and reversetranscribed using the iScript gDNA Clear cDNA Synthesis Kit from BioRad (Madrid, Spain). qPCR was performed with the iTaq Universal Probes Supermix from BioRad (Madrid, Spain) in a CFX Connect Real-Time System from BioRad (Madrid, Spain). Analyses of relative gene expression data were performed using the 2 −∆∆al method. Primer-probe sets for mouse Il1b, Il6 and Tnfa, and actin-b were purchased as predesigned TaqMan probe expression assays (Applied Biosystems, Foster City, CA, USA).

Statistical Analysis
Data are presented as mean ± standard deviation (SD) and were compared by using the ANOVA test and Bonferroni post hoc test. All statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software Inc., San Diego, CA, USA). Differences were considered statistically significant at p ≤ 0.05.

Conclusions
In conclusion, the use of a dehydrative coupling reaction between 3-hydroxyphthalide and substituted arenes allowed obtaining a series of 3-arylphthalides with good yields, good levels of site selectivity, and with high functional group tolerance. In antioxidant and anti-inflammatory assays, compounds bearing hydroxy groups on the 3-aryl ring displayed significant activities. In particular, compound 5a has been identified as possessing strong antioxidant activity and causing high inhibition of LPS-induced NO production in Bv.2 and RAW 264.7 cells. Moreover, compound 5a significantly decreases mRNA expression of pro-inflammatory cytokines Il1b and Il6 in LPS-stimulated RAW 264.7 cells. These results have disclosed the anti-inflammatory potential of phthalides bearing a phenolic ring at C-3 and provide compelling evidence that slight structural modifications on the aryl ring derivatives conferred a remarkable impact on their antioxidant and anti-inflammatory activities. The reduction of mRNA levels in classical pro-inflammatory cytokines reveals the potential effects on the signaling pathways involved. Thus, further investigation of the inflammasome complex and/or the kinase-stress pathways will be analyzed in order to elucidate the specific anti-inflammatory pathways promoted by the compound 5a and related 3-arylphthalides.

Patents
Compounds 5a and 5e in this manuscript are the subject of a patent application at the Spanish Patent Office. Application no. P202130739.