Chemical Characterization and Antioxidant, Antibacterial, Antiacetylcholinesterase and Antiproliferation Properties of Salvia fruticosa Miller Extracts

The Salvia fruticosa (Mill.) is the most medicinal plant used in Lebanon. The aim of this study is to investigate the phytochemical composition and the biological activities (in vitro) of its extracts. The plant was extracted by cold maceration with four solvents presenting an increasing polarity: cyclohexane (CHX), dichloromethane (DCM), ethyl acetate (EtOAc) and methanol (MeOH). The extracts were screened for their chemical composition by a HPLC-DAD detector for phenolic compounds identification and quantification and by GC-MS for volatile compounds detection. The antioxidant capacity (DPPH inhibition) was tested. Biological activities, mainly anti-Alzheimer activity (acetylcholinesterase inhibition), the antiproliferation of two human colon cancer cell lines (HCT-116 and Caco-2 cells) and antibacterial activity, were evaluated. Ten aromatic compounds were quantified by HPLC-DAD analysis. A total of 123 compounds were detected by GC-MS analysis. The MeOH extract showed a very interesting antioxidant activity with an inhibition percentage (IP) of 76.1% and an IC50 of 19.4 μg/mL. The EtOAc extract exhibited the strongest inhibition against the acetylcholinesterase activity (IP = 60.6%) at 50 μg/mL. It also strongly inhibited the proliferation of the HCT-116 cells (IP = 87.5%), whereas the DCM extract gave the best result with the Caco-2 cells (IP = 72.3%). The best antibacterial activity was obtained with the MeOH extract against Staphylococcus aureus (MIC = 1.2 μg/mL) and with the EtOAc extract against Escherichia coli (MIC = 2.4 μg/mL). This study highlights the chemical composition and therapeutic potential of S. fruticosa. It is important to mention that the following chemical compounds were identified for the first time in plant extracts: 2,6,11,15-tetramethyl-hexadeca-2,6,8,10,14-pentaene; 4,5,6,7-tetrahydroxy-1,8,8,9-tetramethyl-8,9-dihydrophenaleno [1,2-b]furan-3-one; podocarpa-1,8,11,13-tetraen-3-one,14-isopropyl-1,13-dimethoxy; podocarpa-8,11,13-trien-3-one,12-hydroxy-13-isopropyl-,acetate; 3′,8,8′-trimethoxy-3-piperidin-1-yl-2,2′-binaphthyl-1,1′,4,4′-tetrone; and 2,3-dehydroferruginol, thus underlining the originality of this study.


Introduction
In recent decades, due to the excessive need for healthy medications devoid of harmful synthetic and chemical products, there was a growing interest in finding new efficient, nontoxic and natural bioactive compounds. Aromatic and medicinal plants, including Salvia species, are known to have a potent role in the treatment of various illnesses, such as aches, epilepsy, colds, bronchitis, tuberculosis, hemorrhage and menstrual disorders [1]. Although

Plant Materials and Extraction Yields
The yields of the four extracts of S. fruticosa are shown in Figure 1. The highest one was obtained with the MeOH extract (9.2%), followed by the CHX extract (3.6%), DCM extract (3.4%) and EtOAc extract (1.1%).
Dincer et al. reported that Salvia sp. extraction yields ranged between 17.8 and 20.3% [1]. Bozan et al. showed that the methanolic extracts yields of eight Salvia species including the S. halophila Hedge, S. tomentosa Miller, S. fruticosa Miller, S. chrysophylla Stapf, S. sclarea L, S. clicica Boiss. and Kotschy, S. cryptantha Montbret and Aucher ex Bentham, and S. palaestina Bentham varied between 12.8 and 26.3% [13]. The highest yield (26.3%) was recorded with the S. fruticosa Miller. Variations in the yields and composition could be affected by many factors, such as the plant development stage, extraction and ecological conditions, soil nature and geographical coordinates [4].  [13]. The highest yield (26.3%) was recorded with the S. fruticosa Miller. Variations in the yields and composition could be affected by many factors, such as the plant development stage, extraction and ecological conditions, soil nature and geographical coordinates [4].

Total Phenolic Content
The largest amount of phenolic compounds was obtained with the MeOH extract and was 135.1 mg GAE/g of dw as shown in Figure 2. This indicates that either the majority of the phenolic compounds of the S. fruticosa were polar or the most abundant ones were polar. The TPCs of the other extracts were 54.4, 44.4 and 23.5 mg GAE/g of dw, respectively, for the DCM, EtOAc and CHX extracts.

Total Phenolic Content
The largest amount of phenolic compounds was obtained with the MeOH extract and was 135.1 mg GAE/g of dw as shown in Figure 2. This indicates that either the majority of the phenolic compounds of the S. fruticosa were polar or the most abundant ones were polar. The TPCs of the other extracts were 54.4, 44.4 and 23.5 mg GAE/g of dw, respectively, for the DCM, EtOAc and CHX extracts.  [13]. The highest yield (26.3%) was recorded with the S. fruticosa Miller. Variations in the yields and composition could be affected by many factors, such as the plant development stage, extraction and ecological conditions, soil nature and geographical coordinates [4].

Total Phenolic Content
The largest amount of phenolic compounds was obtained with the MeOH extract and was 135.1 mg GAE/g of dw as shown in Figure 2. This indicates that either the majority of the phenolic compounds of the S. fruticosa were polar or the most abundant ones were polar. The TPCs of the other extracts were 54.4, 44.4 and 23.5 mg GAE/g of dw, respectively, for the DCM, EtOAc and CHX extracts. Depending on the weather conditions through the years, the TPC detected in the aerial part, especially in the leaves of the S. fruticosa, ranged between 63.7 and 144 mg GAE/g dw [12]. Dincer et al. stated that the methanolic extracts of the S. fruticosa collected from Turkey presented a TPC between 41.5 and 44.6 mg GAE/g dw, almost 3.2 times lower than the amount obtained in the current study [2]. Moreover, they mentioned that the TPC of the MeOH extracts of S. fruticosa was more abundant than the one obtained with other Salvia species extracts. Duletić-Laušević et al. obtained a TPC of 132 mg GAE/g dw in the methanolic extract of the S. fruticosa, which is slightly lower than the one obtained in the current study [14]. Salvia with a high phenolic content has several interesting applications. It can be used in oily food because of its significant capacity to reduce undesirable fragrances, extend shelf life, delay the formation of toxic oxidation products, increase nutritional value and prevent microbial growth. It is also widely employed in the cosmetic industry [15].

Identification and Quantification of Phenolic Compounds by HPLC-DAD
Ten compounds, of which nine were phenolic compounds and one a methoxy-phenolic compound, were detected and quantified by HPLC-DAD as reported in Figure 3 and Table 1. ial part, especially in the leaves of the S. fruticosa, ranged between 63.7 and 144 mg GAE/g dw [12]. Dincer et al. stated that the methanolic extracts of the S. fruticosa collected from Turkey presented a TPC between 41.5 and 44.6 mg GAE/g dw, almost 3.2 times lower than the amount obtained in the current study [2]. Moreover, they mentioned that the TPC of the MeOH extracts of S. fruticosa was more abundant than the one obtained with other Salvia species extracts. Duletić-Laušević et al. obtained a TPC of 132 mg GAE/g dw in the methanolic extract of the S. fruticosa, which is slightly lower than the one obtained in the current study [14]. Salvia with a high phenolic content has several interesting applications. It can be used in oily food because of its significant capacity to reduce undesirable fragrances, extend shelf life, delay the formation of toxic oxidation products, increase nutritional value and prevent microbial growth. It is also widely employed in the cosmetic industry [15].
The identification of the compounds by HPLC-DAD was based on the comparison of the HPLC retention times and the DAD spectra to those found in the literature. It is worth mentioning that the detected compounds were found for the first time in S. fruticosa extracts. The amount of compound 10 was the highest in the CHX extract and decreased gradually in the DCM and EtOAc extracts underlining its non-polar character. Polydatin (compound 4), which was found in the EtOAc and MeOH extracts, presented the highest amount in the MeOH extract, which underlines its polarity. These results are in agreement with those obtained with the TPC determination. The MeOH extract presented the highest TPC mainly because of the presence of polydatin, which was the most abundant phenolic compound detected by HPLC-DAD.  (1); 3,4-dihydroxy-5-methoxybenzoic acid (2); rutin (3); polydatin (4); 5 ,3 -dihydroxyflavone (5); 5,7-dihydroxy-4-phenylcoumarine (6); 3-benzyloxy-4,5-dihydroxy-benzoic acid methyl ester (7); 4 ,5-dihydroxy-7-methoxyflavone (8); pinosylvin monomethyl ether (9); 3,6,3 -trimethoxyflavone (10).
The identification of the compounds by HPLC-DAD was based on the comparison of the HPLC retention times and the DAD spectra to those found in the literature. It is worth mentioning that the detected compounds were found for the first time in S. fruticosa extracts. The amount of compound 10 was the highest in the CHX extract and decreased gradually in the DCM and EtOAc extracts underlining its non-polar character. Polydatin (compound 4), which was found in the EtOAc and MeOH extracts, presented the highest amount in the MeOH extract, which underlines its polarity. These results are in agreement with those obtained with the TPC determination. The MeOH extract presented the highest TPC mainly because of the presence of polydatin, which was the most abundant phenolic compound detected by HPLC-DAD.

Polydatin
Molecules 2023, 28, x FOR PEER REVIEW 5 of 18 The results are the means of duplicate experiments (±SD).

GC-MS Analysis of the S. fruticosa Extracts before and after Derivatization (Trimethylsilylation)
A total of 58 compounds were identified by GC-MS before derivatization (trimethylsilylation) and 65 additional ones after derivatization ( Table 2).
where n is the number of carbon of lower alkane.

DPPH Assay for the Determination of the Antioxidant Activity
The antiradical activity of the aerial part of the S. fruticosa organic extracts was assessed and compared with the standard ascorbic acid (IC 50 = 4 µg/mL). The concentration of the extracts was adjusted to 50 µg/mL. As listed in Table 3, it can be stated that the strongest inhibition of DPPH was obtained with the MeOH extract (76.1%) with an IC 50 value of 19.4 µg/mL. It was followed by the EtOAc extract, which showed an inhibition percentage of 20.9%, then the DCM extract with 6.5% and finally no inhibition was registered with the CHX extract.  The strongest DPPH inhibition was obtained with the MeOH extract. Several authors tested the capacity of the MeOH extracts of many Libyan, Turkish and Iranian Salvia species to quench the DPPH free radical and found that the IC 50 values were as follows: S. fruticosa, IC 50 = 36.37 µg/mL; S. multicaulis, IC 50 = 386.9 µg/mL, S. viridis, IC 50 = 570 µg/mL [20]; and S. macrosiphon, IC 50 = 2743.05 µg/mL [21]. Interestingly, the concentrations were, respectively, 9, 97, 142 and 686 times higher than those obtained with the MeOH extract of the current study (IC 50 = 19.4 µg/mL). El Boukhary et al. assessed the ability of S. fruticosa MeOH extracts prepared from the roots and the aerial parts to inhibit DPPH. The inhibition percentages obtained were 32.16 and 41.5%, respectively, and were 2.37 and 1.83 times less effective than the MeOH extract of the current study (76.1%) [3]. Moreover, the capacity of the MeOH extract to inhibit the DPPH was 3.3 times less effective than that of the ascorbic acid (19.4 vs. 4.0 µg/mL). Nevertheless, the antioxidant activity of the methanolic extract was significant and resulted from the presence of several molecules in the extract. The fractioning and separation of these molecules may give more interesting IC 50 values than those obtained with the ascorbic acid. The high correlation coefficient (R 2 = 0.92) between the phenolic content of the extracts and their corresponding antioxidant activity confirmed that the phenolic compounds were the main components responsible for the antioxidant activity of the S. fruticosa extracts. The MeOH extract, which exhibited the highest antioxidant activity, contained, per gram of extract, 0.3 mg of 3-amino-4-hydroxybenzoic acid; 74.3 mg of polydatin; and 0.1 mg of 5,7-dihydroxy-4-phenylcoumarine (Table 1). It also contained salvianolic acid A (22 ); 2,3-dehydroferruginol (33 ); and rosmarinic acid (60 ) ( Table 2). These molecules may have worked synergistically in order to give a good inhibition. In the EtOAc extract, the presence of phenol, 2,2 -methylenebis 6-(1,1-dimethylethyl)-4-methyl-(43); 4-hydroxybenzoic acid (9 ); and 2-palmitoyl glycerol (40 ) that have a labile proton may have contributed to the quenching of the DPPH free radical ( Table 2). The CHX and DCM extracts showed a poor antioxidant activity. However, some of their chemical compounds may have inhibited the DPPH free radical, such as the β-sitosterol (IC 50 = 140 µg/mL) [22], caryophillene oxide (IC 50 = 84.0 µg/mL [23]), carnosol (IC 50 = 0.59 µM [24]) and carnosic acid (IC 50 = 60 µM) [24]. The lupeol (56) was previously shown to exhibit a good antioxidant behavior. In fact, it quenched the DPPH free radical by 50.0% at 70 µg/mL [25].
2.6. Biological Activities 2.6.1. Antiacetylcholinesterase Activity (Anti-AChE) The analysis was carried out with 50 µg/mL of each S. fruticosa extract. The results were compared to that of the standard GaHbr. As shown in Table 3, a similar inhibition percentage was registered for the CHX (59.5%), DCM (60.5%) and EtOAc (60.6%) extracts. Although the MeOH extract presented the lowest inhibition with 52.46%, it was not significantly different from the previous ones (p > 0.05). This behavior against the AChE enzyme could be attributed to many molecules that are present in the samples. Ayaz et al. proved that the β-sitosterol exhibited a considerable AChE inhibition with an IC 50 value of 55 µg/mL. This compound was present in the non-polar CHX and DCM extracts (55-  Table 2) and the (−)-camphor (12- Table 2) from S. lavandulaefolia EO [16]. These compounds were also found in the S. fruticosa CHX and/or DCM extracts. The (−)-β-pinene inhibited the AChE activity with an IC 50 value of 0.2 mg/mL, while the camphor reduced its activity by 39.0% at 0.5 mg/mL. Recently, Karakaya et al. proved that the caryophyllene oxide isolated from Salvia verticillata subsp. Amasiaca EO presented a good anti-AChE potency since it reduced the enzyme activity by 41.4% at 200 µg/mL. It can be concluded from the previous data that the anti-AChE activity of the extracts resulted from complex interactions, both synergistic and antagonistic, between their terpene constituents [23].

MTT Assay for the Measurement of the Antiproliferation Activity
The antiproliferation activity of the S. fruticosa extracts prepared at 50 µg/mL was tested on two lines of cancer cells: Caco-2 and HCT-116. The tamoxifen was used as a positive control. The highest growth inhibition of the Caco-2 cells was registered with the DCM extract (72.3%), followed by the EtOAc extract (62.1%), whereas the highest inhibition of the HCT-116 cells was recorded with the EtOAc extract (87.5%), followed by the DCM extract (70.7%) ( Table 3). The CHX extract was not active at all, while the MeOH extract only reduced the growth of the HCT-116 cells by 7.2% (p ≤ 0.05). Therefore, the S. fruticosa extracts were more active against the proliferation of the HCT-116 cells than the Caco-2 ones. The American National Cancer Institute (NCI) considered that the extracts with an IC 50 < 30 µg/mL had a promising cytotoxic activity (Suffness, 1990), which is the case of the DCM and EtOAc extracts of the current study (Table 3) [25]. Duletić-Laušević et al. tested the antiproliferation activity of the Libyan S. fruticosa-EtOH extract against the HCT-116 cells and found that it was able to reduce their growth with an IC 50 of 375.96 µg/mL [14]. The obtained value was 25.7 times higher than the one recorded with the EtOAc extract (14.6 µg/mL) of the current study. The latter gave the best antiproliferation activity and showed a promising anticancer application. Polydatin was found to inhibit the growth of Caco-2 cells and gave an IC 50 of 74.9 µg/mL [26]. This compound was present in the EtOAc extract that presented an IC 50 of 31.1 µg/mL when tested against the same cancer cell line. Therefore, polydatin may have contributed to the antiproliferation activity of the EtOAc extract along with other compounds leading to a better result than when tested alone.

Antimicrobial Activity Assay
The four S. fruticosa extracts were tested individually for their capacity to inhibit the growth of seven foodborne pathogenic bacterial strains including four Gram-negative and three Gram-positive bacteria. The minimum inhibitory concentrations (MICs) obtained are displayed in Table 4.   [14]. When comparing these values to the best MIC values obtained in the current study against the same bacterial species, we find that the MIC values of this study were, respectively, 625, 19.21, 833 and 208 times lower than those found in the literature, underlining the important antibacterial activity of our extracts. However, we should take into consideration that the extraction solvents used, the geographical origin and the bacterial strains were not the same. Based on the study conducted by Kosová et al., the EtOAc extract's behavior could be explained by the presence of the 4-hydroxybenzoic acid (9 - Table 2) considered a preservative [27]. This acid was shown to inhibit the growth of S. aureus and E. coli at 20 mmol/l. The extracts of the current study exhibited a bacteriostatic activity against some of the bacterial strains tested, while some others, mainly Gram + ones, were not affected. This can be explained by the presence of resistance mechanisms in these strains [28].

Principal Component Analysis (PCA)
The PCA analysis was used in this study in order to establish the relation between the different biological activities and the chemical composition of the extracts. As shown in Table 5, the axes of inertia have been hidden from this analysis. The percentage of total variation was recorded at 98.6% and proven by the structuring accessions in Figure 4.  The axes were retained because they expressed 57.1% (PC1) and 41.5% (PC2). Simultaneously, the loadings in the PCA loading plots expressed how good the correlation was between the major components and the original variables studied. There was a very good correlation between the antioxidant activity and the TPC. PC1 was highly correlated only with the TPC with a loading of 0.93 (Table 6). The second axis was well correlated with HCT-116, Caco-2 and AChE with loadings of 0.73, 0.63 and 0.59, respectively (Table 6). When applying the principal component analysis, it seemed that there was a discriminate structure. The oval forms grouped the different extracts into three classes: C1 (S. fruticosa-DCM and S. fruticosa-EtOAc), C2 (S. fruticosa-CHX) and C3 (S. fruticosa-MeOH). Since the two plots (biplot) were gathered together, it can be noticed that the high TPC and antioxidant activity were related to the S. fruticosa-MeOH extract. In addition, the S. fruticosa-DCM, S. fruticosa-EtOAc and S. fruticosa-CHX (poor in TPC) were located symmetrically in the negative side of the PC1 axis, which suggested that the high antiproliferation and antiacetylcholinesterase activities of these extracts were not only related to the phenolic compounds. The axes were retained because they expressed 57.1% (PC1) and 41.5% (PC2). Simultaneously, the loadings in the PCA loading plots expressed how good the correlation was between the major components and the original variables studied. There was a very good correlation between the antioxidant activity and the TPC. PC1 was highly correlated only with the TPC with a loading of 0.93 (Table 6). The second axis was well correlated with HCT-116, Caco-2 and AChE with loadings of 0.73, 0.63 and 0.59, respectively (Table 6). When applying the principal component analysis, it seemed that there was a discriminate structure. The oval forms grouped the different extracts into three classes: C1 (S. fruticosa-DCM and S. fruticosa-EtOAc), C2 (S. fruticosa-CHX) and C3 (S. fruticosa-MeOH). Since the two plots (biplot) were gathered together, it can be noticed that the high TPC and antioxidant activity were related to the S. fruticosa-MeOH extract. In addition, the S. fruticosa-DCM, S. fruticosa-EtOAc and S. fruticosa-CHX (poor in TPC) were located symmetrically in the negative side of the PC1 axis, which suggested that the high antiproliferation and antiacetylcholinesterase activities of these extracts were not only related to the phenolic compounds.

Preparation of the Extracts
The collected aerial parts of S. fruticosa were dried in shade at an ambient temperature and transformed into powder with a particle size of 0.8 mm [29]. A cold maceration with four solvents presenting an increasing polarity (CHX, DCM, EtOAc and MeOH) was performed to yield the four organic extracts. A total of 100 g of powder was successively extracted with 2 L of each solvent during 2 h with an agitation of 300 rpm. The filtrates were recovered after filtration through whatman filter papers (Fisher, Asiane, France). The extracts were obtained by evaporating the solvent under vacuum at 35 • C. The extraction yield was calculated as follows: Yield (%) = m M × 100, "m" being the dry weight obtained in grams and "M" being the weight of the plant material in grams.

Total Phenolic Content Determination
The total phenolic content (TPC) of each extract was evaluated spectrophotometrically at 765 nm using the Folin-Ciocalteu (FC) method as described by Dawra et al. [29]. Gallic acid (0-115 µg/mL) was used for the calibration curve. The results were expressed as mg of gallic acid equivalents (GAE)/g dw.

HPLC-DAD Fingerprint
The HPLC analysis was performed in an ultimate 3000 pump-Dionex and Thermo separation product detector DAD model (Thermo Fisher Scientific, Waltham, MA, USA). Separation was achieved on an RPC18 reversed-phase column (Phenomenex, Le Pecq, France), 25 cm × 4.6 mm and particle size of 5 µm, thermostated at 25 • C as described by Dawra et al., with modifications [28]. The elution was performed at a flow rate of 1.2 mL/min, using a mobile phase consisting of MilliQ water (pH 2.6) (solvent A) and acidified water/MeCN (20:80 v/v) (solvent B). The samples were eluted by the following linear gradient: from 12% B to 30% B for 35 min, from 30% B to 50% for 5 min, from 50% B to 88% B for 5 min and finally from 88% B to 12% B for 15 min. The extracts were prepared at a concentration of 20 mg/mL using the mixture acidified water/MeCN (80:20 v/v) and then filtered through a Millex-HA 0.45 µm syringe filter (Sigma Aldrich). Then, 20 µL of each sample were injected and the detection was registered at 280 nm. The phenolic compounds were identified by comparison to the retention time of some known standards and then quantified using their corresponding calibration curves.

Gas Chromatography GC-MS Analysis
The identification of the volatile compounds of the organic extracts, before and after derivatization, was conducted using the protocol described by Dawra et al., with some modifications [29]. The analyses were conducted using an Agilent gas chromatograph 6890 coupled to a 5975 Mass Detector. The 7683 B auto sampler injected 1 µL of each extract. A fused silica capillary column DB-5 MS (30 m × 0.25 mm internal diameter, film thickness 0.25 µm) (Supelco, Sigma-Aldrich, Darmastadt, Germany) was employed. The temperature ramp was settled between 35 and 300 • C. The column temperature was initially set to 35 • C before being gradually increased to 85 • C at 15 • C/min, held for 20 min at 85 • C, raised to 300 • C at 10 • C/min and finally held for 5 min at 300 • C. Helium (purity 99.99%) was used as a carrier gas at a flow rate of 0.8 mL/min. Mass spectra were registered at 70 eV with an ion source temperature held at 310 • C and a transfer line heated at 320 • C. The record of each acquisition was made in full-scan mode (50-400 amu). The main target was to find the maximum resemblance in terms of spectra between the compounds found in the extracts and those suggested by the NIST08 database (National Institute of Standards and Technology, https://www.nist.gov/ (accessed on 15 April 2021)), using AMDIS software, and the retention time was used to facilitate many tasks. The derivatization method consisted of dissolving 5 mg of each extract in 1 mL of its own solvent except for the MeOH extract. The latter was dissolved in MeCN. After that, 150 µL of BSTFA and 1.5 µL of TMSC were added to the solution. The mixture was agitated for 30 s in order to increase the solubility. The reaction mixture was maintained at 40 • C for 30 min. A total of 10 µL of each derivatized solution were injected into the GC-MS and analyzed as previously reported.

Free Radical Scavenging Activity: DPPH Test
The antioxidant scavenging activity of the extracts was examined using the DPPH method as described by Dawra et al. [29]. A total of 20 µL of the diluted plant extract (500 µg/mL) was added to 180 µL of a 0.2 mM methanolic DPPH solution in a 96-well microplate (Micro Well, Thermo Fisher Scientific, Bordeaux, France). After an incubation period of 30 min at 25 • C, the absorbance was measured at 515 nm. The antioxidant activity was expressed as the inhibition percentage of DPPH using the following equation: . The extract concentration providing a 50% reduction in the DPPH initial absorbance (IC 50 ) was calculated using the linear relation between the extract concentrations and the corresponding % INB of DPPH. All measurements were performed in quadruplicate.

Antiacetylcholinesterase Activity
The antiacetylcholinesterase (AChE) activity was tested using the Ellman's procedure as previously reported by Dawra et al. [29]. In a 96-well microplate, 50 µL of 0.1 mM sodium phosphate buffer (pH = 7.5), 125 µL of DTNB (5,5 -dithiobis-2-nitrobenzoic acid), 25 µL of the diluted plant extract (500 µg/mL) and 25 µL of the enzyme solution (493.2 U) were mixed. The microplate was incubated at 25 • C for 15 min. Then, 25 µL of ACTHI was added and the final blend was incubated at 25 • C for 25 min. Finally, the absorbance was measured at 421 nm. The A blank was measured without the extract. The inhibition percentage of the enzyme activity was calculated as follows:

Antiproliferation Activity
The antiproliferation activity of the plant extracts was assessed on two different types of human colon cancer cells (HCT-116 and Caco-2). The test was based on the MTT reduction by the mitochondrial dehydrogenases of intact cells to a purple formazan product. MTT is a yellow water-soluble tetrazolium salt. The cell lines were purchased from Sigma-Aldrich (Manassas, VA, USA). A volume of 100 µL of a suitable culture medium containing 3 × 10 4 cells was added to each well of a 96-well microplate. Then, 100 µL of the same culture medium containing the plant extract was added. The final concentration of the extract in each well was 50 µg/mL. The culture media used were, respectively, the RPMI 1640 (Sigma Aldrich, St. Louis, MO, USA) for the HCT-116 cells and the Dulbecco's modified Eagle's medium GlutaMAX (DMEM, Sigma Aldrich, USA) for the Caco-2 cells. The microplate was incubated at 37 • C for 48 h. The supernatant was then removed, and 50 µL of the MTT solution was added followed by an incubation of 40 min. After removing the MTT reagent, 80 µL of DMSO was added to solubilize the formazan crystals. Finally, the absorbance was measured at 605 nm. The tamoxifen was used as a positive control whereas the negative control was composed of the cell suspension without the plant extracts (blank). The inhibition percentage of the cells' proliferation was calculated as

Antimicrobial Activity Assay
The Gram-negative strains used in this assay were Escherichia coli ATCC 8739 and the Kentucky, Infantis and Enteritidis serotypes of Salmonella enterica provided by the Lebanese Agriculture Research Institute (LARI), Lebanon. The Salmonella serotypes were isolated from chicken samples collected from slaughterhouses. The Gram-positive strains were Staphylococcus aureus ATCC 25923, Listeria monocytogenes ATCC 19115 and Listeria monocytogenes isolated from "fish-filet" at the LARI. A bacterial suspension of each bacterial strain was prepared in a Mueller-Hinton Broth (MHB) at a concentration of 2 × 10 8 CFU/mL (0.5 McFarland standard) [30]. The minimum inhibitory concentration (MIC) values of the Salvia fruticosa extracts were determined by serial dilution in a 96-well microplate. Each well first contained 100 µL of MHB. The dried extracts were dissolved in pure DMSO to a concentration of 5 mg/mL. Then, the extract solutions were half-diluted with MHB to obtain a concentration of 2.5 mg/mL. Afterwards, 100 µL of the latter were placed in the first well and a serial dilution was conducted in order to obtain the following concentrations in each row: 1250, 625, 312.5, 156.2, 78.1, 39, 19.5, 9.7, 4.8, 2.4 and 1.2 µg/mL. Next, 100 µL of each bacterial strain tested were added to the extract solutions. Therefore, the initial bacterial concentration was adjusted to 10 8 CFU/mL in each well. The negative control was composed of 100 µL of DMSO and 100 µL of the bacterial strain tested while the positive control contained 100 µL of MHB and 100 µL of the bacterial strain. The absorbance was measured at time 0 min and after an overnight incubation at 37 • C (24 h) using a Multiskan Sky Microplate Spectrophotometer (Thermo Fisher Scientific, Cleveland, OH, USA).

Statistical Analysis
The data represent the mean of four replicates ± standard deviation (SD). The results were subjected to a multiway analysis of variance, and the mean comparisons were performed by a Tukey's multiple range test using SPSS version 20.0 (Statistical Package for the Social Sciences, Inc., Chicago, IL, USA). The differences between means were considered significant at p-value < 0.05. The linear correlation coefficient (R 2 ) was calculated to establish the relationship between the TPC and the antioxidant or any other biological activity. For exploratory data analysis, the results were processed by one of the multivariate analysis techniques, the principal components analysis (PCA). The PCA was conducted using XLSTAT (version 2020.1, Addinsoft, Pearson edition, Waltman, MA, USA) for a better discrimination between the studied parameters.

Conclusions
This novel research shed light on the phytochemistry and biological activities of the Salvia fruticosa Miller extracts native to Lebanon. The HPLC-DAD analysis permitted the identification and quantification of ten aromatic compounds; nine of which were recognized as phenolic compounds. The GC-MS analysis revealed the presence of 123 volatile compounds. S. fruticosa extracts showed interesting biological activities. The MeOH extract showed a high antioxidant activity. The four extracts presented a good antiacetylcholinesterase activity. The DCM and EtOAc extracts revealed a significant antiproliferation activity against the HCT-116 and Caco-2 cancer cells. Interestingly, the four extracts exhibited an excellent antibacterial activity against pathogenic foodborne Gram-negative and Gram-positive bacteria with low MIC values, particularly against Escherichia coli, Staphylococcus aureus and Listeria monocytogenes. The above-mentioned promising pharmacological activities highlight the plant's potential use in the development of new antimicrobial drugs. They encourage us to identify and purify the bioactive compounds by performing a bioguided fractionation of the most active extracts. The extracts that possess interesting antioxidant and antimicrobial activities should also be tested in food preservation. In vivo studies of the most active compounds should also be performed in order to better assess their therapeutic potential.