Influence of Extraction Solvent on the Phenolic Profile and Bioactivity of Two Achillea Species

The phenolic composition, as well as the antioxidant and antimicrobial activities of two poorly investigated Achillea species, Achillea lingulata Waldst. and the endemic Achillea abrotanoides Vis., were studied. To obtain a more detailed phytochemical profile, four solvents with different polarities were used for the preparation of the plant extracts whose phenolic composition was analyzed using UHPLC-MS/MS (ultra-high performance liquid chromatography-tandem mass spectrometry). The results indicate that both of the investigated Achillea species are very rich in both phenolic acids and flavonoids, but that their profiles differ significantly. Chloroform extracts from both species had the highest yields and were the most chemically versatile. The majority of the examined extracts showed antimicrobial activity, while ethanolic extracts from both species were potent against all tested microorganisms. Furthermore, the antioxidant activity of the extracts was evaluated. It was found that the ethanolic extracts possessed the strongest antioxidant activities, although these extracts did not contain the highest amounts of detected phenolic compounds. In addition, several representatives of phenolic compounds were also assayed for these biological activities. Results suggest that ethanol is a sufficient solvent for the isolation of biologically active compounds from both Achillea species. Moreover, it was shown that the flavonoids naringenin and morin are mainly responsible for these antimicrobial activities, while caffeic, salicylic, chlorogenic, p-coumaric, p-hydroxybenzoic, and rosmarinic acid are responsible for the antioxidant activities of the Achillea extracts.


Introduction
Selection of an appropriate solvent and extraction protocol is the key for successful isolation of biologically active compounds from medicinal plants. The extraction solvents are chosen according to their polarity, and therefore, the ability to isolate specific types of compounds with different structures and physicochemical properties. The solvents accepted for use in pharmaceutical formulations are water, ethanol, and glycerol [1]. The polarity of extraction solvents influences the extraction efficiency of phenolic compounds. Less polar solvents extract smaller amounts of phenolic compounds, and therefore, these extracts possess a lesser potential for scavenging free radicals [2]. Generally, highly hydroxylated with the main distribution in the Balkan Peninsula, Carpathians, and Belarus, while A. abrotanoides is endemic to the Balkan Peninsula. To isolate diverse compounds from these plants, four solvents with different polarities were used and all extracts were characterized by their phenolic composition and antioxidant and antimicrobial activities. To the best of our knowledge, there is little data about the phenolic composition and bioactive compounds in the roots [32] and aerial parts [33] of A. lingulata, and no data for A. abrotanoides.

Extraction
The objective of the extraction process is to maximize the amount of target compounds and to obtain the highest biological activity of these extracts. The extraction yield and biological activity of the resulting extract were affected by both the extraction technique and the extraction solvent. Ultrasound-assisted extraction, or sonication, uses cavitation energy in the solvent that accelerates the dissolution and diffusion of the solute as well as the heat transfer, which improves the extraction efficiency. This extraction technique requires low solvent and energy consumption and allows for the reduction of extraction temperature and extraction time. Therefore, it is applicable for the extraction of thermolabile and unstable compounds [34].
Several factors should be considered in the selection of solvents, mainly selectivity, solubility, cost, and safety [35]. In general, alcohols, acetone, and water are used for the extraction of bioactive compounds from the plant material, but the selection is based on the properties of the compound of interest, as well as the plant material used [34].
Plant extracts were prepared through the sonication of dry plant material (vegetative part and inflorescence) in petrol ether, chloroform, ethanol, and water. The yields of the extracts are presented in Table 1, from which it can be seen that the chloroform extracts had the highest yields for both plant species and their investigated organs. Usually, this halogenated solvent with medium polarity can isolate a wide range of compounds from plant material and is commonly used for the non-targeted isolation and analysis of natural products [23]. However, due to its toxicity in long exposures, chloroform is not a suitable solvent for use in pharmaceutical preparations. Furthermore, water extracts also showed relatively high yields for both plants and their organs. On the contrary, petrol ether and ethanol revealed the lowest yields. Table 1. Yields (%) of obtained extracts of two Achillea species.

Species
Plant organ Petrol ether Chloroform Ethanol Water

Extraction
The objective of the extraction process is to maximize the amount of target compounds and to obtain the highest biological activity of these extracts. The extraction yield and biological activity of the resulting extract were affected by both the extraction technique and the extraction solvent. Ultrasound-assisted extraction, or sonication, uses cavitation energy in the solvent that accelerates the dissolution and diffusion of the solute as well as the heat transfer, which improves the extraction efficiency. This extraction technique requires low solvent and energy consumption and allows for the reduction of extraction temperature and extraction time. Therefore, it is applicable for the extraction of thermolabile and unstable compounds [34].
Several factors should be considered in the selection of solvents, mainly selectivity, solubility, cost, and safety [35]. In general, alcohols, acetone, and water are used for the extraction of bioactive compounds from the plant material, but the selection is based on the properties of the compound of interest, as well as the plant material used [34].
Plant extracts were prepared through the sonication of dry plant material (vegetative part and inflorescence) in petrol ether, chloroform, ethanol, and water. The yields of the extracts are presented in Table 1, from which it can be seen that the chloroform extracts had the highest yields for both plant species and their investigated organs. Usually, this halogenated solvent with medium polarity can isolate a wide range of compounds from plant material and is commonly used for the non-targeted isolation and analysis of natural products [23]. However, due to its toxicity in long exposures, chloroform is not a suitable solvent for use in pharmaceutical preparations. Furthermore, water extracts also showed relatively high yields for both plants and their organs. On the contrary, petrol ether and ethanol revealed the lowest yields.

UHPLC-MS/MS Analysis of Phenolic Compounds
A UHPLC-MS/MS (ultra-high performance liquid chromatography-tandem mass spectrometry) analysis was used for the separation and identification of phenolic compounds in the extracts of these two Achillea species. This is the first report of the phenolic composition of aerial parts of both A. lingulata and A. abrotanoides. To the best of our knowledge, only reports on the composition of the roots [32], lignans [33,36], and volatile compounds [37][38][39][40] for A. lingulata are recorded.
Among the 32 phenolic compounds analyzed, 21 phenolic acids and flavonoids were detected and quantified in the extracts of these two Achillea species (Table 2). In general, A. abrotanoides was found to be richer in phenolic compounds than A. lingulata. As an example, the (multiple reaction monitoring (MRM) chromatograms of ethanolic extracts of inflorescences of both investigated Achillea species are presented in Figure 2.   Chloroform as a solvent revealed the greatest diversity in phenolic profiles of both investigated plants. There were 7 phenolic acids and 9 flavonoids found in the vegetative part of A. abrotanoides, while 12 phenolic compounds in total were found in the extracts of A. lingulata. The highest concentration of phenolic compounds was detected in the chloroform extract of the inflorescence of A. abrotanoides (915.3 µmol/g), for which the MRM chromatograms are presented in Figure 3. As expected, the lowest content of phenolics was found in petrol ether extract of the vegetative part of A. lingulata (total concentration 2.1 µmol/g).  Chloroform as a solvent revealed the greatest diversity in phenolic profiles of both investigated plants. There were 7 phenolic acids and 9 flavonoids found in the vegetative part of A. abrotanoides, while 12 phenolic compounds in total were found in the extracts of A. lingulata. The highest concentration of phenolic compounds was detected in the chloroform extract of the inflorescence of A. abrotanoides (915.3 µmol/g), for which the MRM chromatograms are presented in Figure 3. As expected, the lowest content of phenolics was found in petrol ether extract of the vegetative part of A. lingulata (total concentration 2.1 µmol/g). Chloroform as a solvent revealed the greatest diversity in phenolic profiles of both investigated plants. There were 7 phenolic acids and 9 flavonoids found in the vegetative part of A. abrotanoides, while 12 phenolic compounds in total were found in the extracts of A. lingulata. The highest concentration of phenolic compounds was detected in the chloroform extract of the inflorescence of A. abrotanoides (915.3 µmol/g), for which the MRM chromatograms are presented in Figure 3. As expected, the lowest content of phenolics was found in petrol ether extract of the vegetative part of A. lingulata (total concentration 2.1 µmol/g).  Among hydroxybenzoic acids investigated, p-hydroxybenzoic acid was the most abundant in the extracts of A. abrotanoides (up to 205.073 ± 8.626 µmol/g), while the levels of salicylic acid were quite similar in all examined extracts of both species. Other hydroxybenzoic acids are mostly found in polar extracts, i.e., ethanolic and aqueous. In addition, the levels of chlorogenic and caffeic acids, cinnamic acids commonly found in plants, were relatively low in comparison with the levels of hydroxybenzoic acids ( Table 2). On the contrary, significant levels of rosmarinic acid were found in the inflorescence of the endemic A. abrotanoides (up to 207.473 ± 17.557 µmol/g). Unexpectedly, p-coumaric acid was not a dominant phenolic acid in A. abrotanoides, and it was not detected in A. lingulata. However, this hydroxycinnamic acid was identified in many other yarrows, such as A. millefolium, A. distans, A. biserratae, and A. beibrestinii [41][42][43][44].
Furthermore, apigenin was found to be the most abundant flavonoid found in the vegetative parts of both Achillea species, with levels of up to 112.010 ± 6.564 µmol/g and 165.688 ± 9.680 µmol/g for A. abrotanoides and A. lingulata, respectively. Apigenin was suggested as the main flavonoid in other Achillea species, such as A. distans, A. ligustica, A. collina, A. millefolium, etc. [41,[44][45][46][47]. Moreover, isoquercitrin and rutin were the main flavonoids in A. schurii in the study of Benedec et al. [48]. On the contrary, the inflorescences of the investigated Achillea species differ in their flavonoid profiles, i.e., A. abrotanoides was rich in naringenin (up to 362.662 ± 4.922 µmol/g), while the flowers of A. lingulata contained notable amounts of hesperetin and rutin ( Table 2).

Antimicrobial Activity
All isolated extracts, together with the representatives of the phenolic compounds detected, were tested for antimicrobial activity using the diffusion method. The results are summarized in Table 3. This is the first report of antimicrobial activity for the aerial plant parts of A. abrotanoides, while the antimicrobial activity of A. lingulata aerial parts has been studied before [49]. However, this is the first attempt to compare different extraction solvents that isolate the phenolic compounds that might be responsible for the antimicrobial activities of A. lingulata and A. abrotanoides.
All four water extracts of both Achillea species showed no antimicrobial potential, whereas they contained significant amounts of phenolic compounds (Table 2). This result was also reported previously when aqueous extracts exhibited no antimicrobial activity [50], which might be attributed to the inefficient diffusability of aqueous solutions in the agar medium [51], as well as to the lesser ability of aqueous extracts to damage microbe cell walls [52]. However, ethanol extracts of both species showed strong antimicrobial effects against all tested microorganisms except for Staphylococcus aureus with an equal or even higher efficiency than the antibiotic ampicillin or the antimycotic nystatin that were used as positive controls (Table 3). In general, all inflorescence extracts of both species showed more potent antimicrobial activities than the same extracts from the vegetative parts. The inflorescence ethanolic extract of A. abrotanoides was significantly more effective against Enterococcus faecalis compared to the ampicillin. All extracts (except for the chloroform extract of the vegetative part of A. abrotanoides) exhibited the same or even more potent activity against Candida albicans than the antimycotic. In this case, the chloroform extract of A. abrotanoides inflorescence had two times greater an antimicrobial effect than nystatin. Moreover, E. coli was equally as susceptible to nearly all extracts as to the reference antibiotic.
According to the available literature, other Achillea species did not demonstrate such potent antifungal activity against Candida albicans. The inhibition zone of ethanol flower extracts of A. schurri was recorded to be only 6 mm [49], while significantly higher values were recorded for the A. lingulata ethanol extract and the A. abrotanoides chloroform extract (Table 3). A similar observation was recorded for A. millefolium by Maz et al. [53]. No activity of certain A. millefolium plants growing at an average altitude against E. coli was recorded and the activity was limited to Gram-positive bacteria, while the A. abrotanoides and A. lingulata extracts were as effective as the reference antibiotic. In addition, several representatives of each class of phenolic compounds at a concentration of 0.1 mg/mL were also assayed for their antimicrobial activities (Table 3). Among them all, the flavonoid morin was the only compound that successfully inhibited the growth of all four microorganisms. This compound was found in the ethanolic extract of the inflorescences and vegetative parts of both investigated Achillea species. Moreover, the flavonon naringenin showed very high activity against E. faecalis, S. aureus, and C. albicans, with the inhibition zones being bigger than those of ampicillin and nystatin. High concentrations of naringenin were found in the chloroform and ethanolic extracts of A. abrotanoides. Therefore, the activity of these extracts might be explained by the presence of this phenolic compound. Representatives of both hydroxybenzoic and hydroxycinnamic acids revealed similar activities, lower than the activities of flavonoids (Table 3), which is in agreement with the literature data [54].
Antimicrobial activity of phenolic compounds might be explained by the modification of the permeability of cell membranes, the changes in various intracellular functions induced by the hydrogen binding of the phenolic compounds to enzymes, or by the modification of the cell wall rigidity [55][56][57]. Phenolic acids have been shown to disrupt membrane integrity, as they cause consequent leakage of essential intracellular constituents [58], while flavonoids may link to soluble proteins located outside the cells and with bacteria cell walls, thus promoting the formation of complexes [57,59].

Antioxidant Activity
Antioxidant activity of the aerial plant parts of A. lingulata and A. abrotanoides has not been described previously. In this study, the DPPH (2,2-diphenyl-1-picrylhydrazyl) radical scavenging activity of all the extracts was tested and expressed as an IC 50 value, which represents the concentration of the extracts that scavenge 50% of the radicals. In addition, selected phenolic compounds were also assayed for their abilities to scavenge stable DPPH radicals. All results are presented in Figure 4. In general, all extracts of A. lingulata showed a stronger antioxidant potential compared to A. abrotanoides.
represents the concentration of the extracts that scavenge 50% of the radicals. In addition, selected phenolic compounds were also assayed for their abilities to scavenge stable DPPH radicals. All results are presented in Figure 4. In general, all extracts of A. lingulata showed a stronger antioxidant potential compared to A. abrotanoides. The polar aqueous and ethanolic extracts of both investigated species had the lowest IC50 values. The most potent extract was the water inflorescence extract of Achillea lingulata, with an IC50 value of 1.52 µg/mL. This extract contained significant levels of rosmarinic acid, which also possesses strong antioxidant activity against stable DPPH radicals (IC50 5.47 µg/mL). Moreover, the activity of this extract could be attributed to other phenolic compounds that are not quantified due to the lack of standards. Ethanol extracts of the inflorescence, as well as the vegetative part of Achillea lingulata, also showed a high antioxidant capacity, with IC50 values of 18.14 and 12.73 µg/mL, respectively. These extracts were rich in p-hydroxybenzoic acid, but also the flavonoid apigenin. The antioxidant potential is decreased via a decrease in the polarity solvent used [60,61], and the chloroform extracts of A. abrotanoides had the lowest antioxidant potential and the highest IC50 values (Figure 4). On the contrary, chloroform extracts of the vegetative parts and the inflorescence were rich in the phenolic compounds detected. Therefore, its low ability to scavenge stable radicals could be explained by the fact that chloroform isolates a high amount of chlorophylls and other pigments from leaves. These compounds do not possess significant antiradical activities [62].
The antioxidant activity of the extracts was correlated with the concentrations of the selected phenolic compounds, and it was found that caffeic (R = 0.6256), salicylic (R = 0.6296), chlorogenic (R = 0.6394), p-coumaric (R = 0.6402), p-hydroxybenzoic (R = 0.7493), and rosmarinic acid (R = 0.9807) have very high positive correlations with the antioxidant activity. Therefore, it might be concluded that these acids are responsible for the antioxidant activity of the extracts that contain them. The polar aqueous and ethanolic extracts of both investigated species had the lowest IC 50 values. The most potent extract was the water inflorescence extract of Achillea lingulata, with an IC 50 value of 1.52 µg/mL. This extract contained significant levels of rosmarinic acid, which also possesses strong antioxidant activity against stable DPPH radicals (IC 50 5.47 µg/mL). Moreover, the activity of this extract could be attributed to other phenolic compounds that are not quantified due to the lack of standards. Ethanol extracts of the inflorescence, as well as the vegetative part of Achillea lingulata, also showed a high antioxidant capacity, with IC 50 values of 18.14 and 12.73 µg/mL, respectively. These extracts were rich in p-hydroxybenzoic acid, but also the flavonoid apigenin. The antioxidant potential is decreased via a decrease in the polarity solvent used [60,61], and the chloroform extracts of A. abrotanoides had the lowest antioxidant potential and the highest IC 50 values (Figure 4). On the contrary, chloroform extracts of the vegetative parts and the inflorescence were rich in the phenolic compounds detected. Therefore, its low ability to scavenge stable radicals could be explained by the fact that chloroform isolates a high amount of chlorophylls and other pigments from leaves. These compounds do not possess significant antiradical activities [62].
The antioxidant activity of the extracts was correlated with the concentrations of the selected phenolic compounds, and it was found that caffeic (R = 0.6256), salicylic (R = 0.6296), chlorogenic (R = 0.6394), p-coumaric (R = 0.6402), p-hydroxybenzoic (R = 0.7493), and rosmarinic acid (R = 0.9807) have very high positive correlations with the antioxidant activity. Therefore, it might be concluded that these acids are responsible for the antioxidant activity of the extracts that contain them.
DPPH radical scavenging potential has been previously recorded for other Achillea species [48,63], ranging from the low IC 50 of 0.52 µg/mL, as recorded for the essential oils of A. pannonica [64], up to 1.172 µg/mL for extracts of A. eriophora [65]. For some extracts, such as water and ethanol extracts of A. lingulata and A. abrotanoides, reported IC 50 values are even lower than those of some synthetic or natural antioxidants [50,[66][67][68].

Plant Material
Samples of A. lingulata and A. abratonoides were collected during the flowering stage at Mt. Jahorina (June 2017) and Mt. Bjelašnica (July 2017), respectively (Figure 1). Specimens were deposited at the herbarium of the Department of Biology, Faculty of Science, the University of Sarajevo under the voucher no. LERP 355 and LERP 356. The inflorescence and the vegetative parts were separated and air-dried for 7 days at room temperature (23 • C) in a shaded, well-ventilated laboratory. Dried samples were finely powdered in the mill and stored at 4 • C until use.

Preparation of the Extracts
For the inflorescence and vegetative aerial parts, 500 mg each of the powdered plant material was soaked in 12.5 mL of solvent (petrol ether, chloroform, ethanol, and water) and sonicated for 30 min at 23 • C. The supernatant was transferred into a new vial and the sediment was soaked again in the same solvent. Due to the high evaporation rate of petrol ether and chloroform, these extracts were evaporated to dryness and resuspended in DMSO.

UHPLC-MS/MS Analysis of the Extracts
Standard solutions of the target phenolic compounds were first prepared in methanol at 1 mM concentrations, and the solutions were gradually diluted in the mobile phase to the working concentrations that ranged from 0.01 to 50 µM. Quantification was performed by the isotope diluting method using p-coumaric acid-d 6 and salicylic acid-d 4 .
UHPLC-MS/MS was performed on an UltiMate™ 3000 liquid chromatographic system consisting of binary pumps, an autosampler, and a column thermostat coupled to a TSQ Quantum Access Max triple quadrupole mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Chromatographic separation was performed on an Acquity BEHC18 (150 × 3.0 mm; 1.7 µm particle size) UHPLC column (Waters Corp., Milford, MA, USA) kept at 40 • C. The mobile phase consisted of 10 mM formic acid in water (component A) and acetonitrile (component B). Compounds were separated using a binary gradient starting at 5% B for 0.8 min, increasing to 10% B for 0.4 min with an isocratic run for 0.7 min, then at 15% B for 0.5 min with an isocratic run for 1.3 min, at 20% B for 0.3 min with an isocratic run for 1.2 min, at 25% B for 0.5 min, at 35% B for 2.3 min, at 70% B for 2.5 min, then at 100% B for 1 min with an isocratic run for 1 min, and then back to 5% B for 0.5 min. Finally, the equilibration to initial conditions took 3.3 min, with a total chromatographic run of 16 min. The flow rate was 0.4 mL/min and the injection volume was 10 µL.
All analytes were detected in negative electrospray ionization mode (ESI-). Multiple reaction monitoring (MRM) mode was used for their quantification. The spray voltage was 3 kV, the temperature of the ion transfer tube vaporizer was 320 • C, the sheath gas pressure was 45 psi, and the auxiliary gas pressure was 15 psi.

Determination of Antioxidant Activity
DPPH antioxidant activity was evaluated for all extracts and standards according to Meda et al. [69]. Briefly, 200 µL of sample solution was mixed with 50 µM of DPPH solution in ethanol, and the absorbance of the tested mixtures was measured at 517 nm after 30 min. Absolute ethanol was used to zero the spectrophotometer, DPPH solution was used as a blank sample, and different phenolic compounds ( Figure 4) were used as a positive probe. Antioxidant activity was calculated as IC 50 , which is referred to as the concentration of extract that scavenges 50% of free DPPH radicals.
The radical-scavenging activities of the tested samples, expressed as a percentage of the inhibition of DPPH, were calculated according to the formula IC (%) = [(A 0 − A t )/A 0 ] % 100, where A 0 and A t are the absorbance values of the blank sample and the test sample at 0 and 30 min, respectively. Four different concentrations of each sample were assayed. Percent inhibition after 30 min was plotted against concentration, and the equation for the line was used to obtain the IC 50 value. A lower IC 50 value indicates greater antioxidant activity.

Determination of Antimicrobial Activity
Agar well diffusion method was used to evaluate the antimicrobial activity of the plant extracts and standards according to National Committee for Clinical Laboratory Standards (NCCLS) [70,71]. Each well contained 100 µL of extract or standard. Bacterial strains used in the analysis included the Gram-positive bacteria E. faecalis ATCC ® 19433TM and S. aureus subsp. aureus ATCC ® 6538TM and the Gram-negative bacteria S. abony NCTC ® 6017TM, E. coli ATCC ® 8739TM, and the yeast C. albicans ATCC ® 10231TM. Bacterial strains were used as a standardized inoculum of 5×10 5 CFU/mL using a McFarland standard [72].
Müller-Hinton and Sabouraud medium were used for the cultivation of the bacterial strains and yeast, respectively. Ampicillin was used as a positive standard for bacterial strains and nystatin for Candida albicans. Ethanol and DMSO were used as negative controls. The antimicrobial effect was expressed as a diameter of inhibition zone in mm reduced by the inhibition zone of the negative controls, if appropriate.

Statistical Analysis
All data were analyzed using the STATISTICA 10.0 software (Statsoft Inc.). Experimental results were presented in tables as the mean ± standard deviation of three independent replications. Data were subjected to variance analysis (ANOVA) and the Newman-Keuls post hoc test was carried out to identify significant differences between the extract types. Mean values with p < 0.05 were considered statistically significant. Pearson correlations were performed to observe the possible correlation between the phenolic profile, antioxidant capacity, and detected antimicrobial activity.

Conclusions
The phytochemical analysis of the aerial parts of two yarrow species, Achillea lingulata and the endemic Achillea abratonoides, was performed for the first time. The phenolic composition of four extracts with different polarities suggested that yields of chloroform extracts were the highest, and that they had the biggest diversity of phenolic compounds detected. In addition, ethanolic extracts revealed the strongest antioxidant activities and the ability to suppress the growth of selected microorganisms. Presented results suggest that both investigated Achillea species have strong potential for use in different pharmaceutical preparations.