Bioactivities of Alchemilla mollis , Alchemilla persica and Their Active Constituents

The aim of this study was to determine antimicrobial activities of Alchemilla mollis, Alchemilla persica as well as ellagic acid and miquelianin against Staphylococcus aureus, Enterococcus faecalis, Bacillus subtilis, Escherichia coli, Pseudomonas aeruginosa, Candida albicans by using microbroth dilution method and anti-inflammatory activity by using human red blood cell (HRBC) membrane stabilization method. Microbroth dilution method was used to determine the antimicrobial activities. Extracts possessed activity having MIC values of 2.5-5-10mg/ mL, compounds possessed activity having MIC values of 1.25-2.5-4-5mg/mL. A.mollis aerial parts displayed the highest anti-inflammatory activity (IC 50 =1.22±0.07mg/mL). Ellagic acid and miquelianin were also determined as anti-inflammatory agents with 0.57±0.01mg/mL and 1.23±0.02mg/mL IC 50 values, respectively. Total phenolic content and tannin content of the A.mollis and A.persica were determined as 357.00±75.80mg, 282.50±28.70mg PGE/g plant material and 18.02%, 18.63% respectively according to the method described by European Pharmacopoeia. Ellagic acid, miquelianin and catechin were analyzed by HPLC. The highest catechin content was detected in A.persica roots (6.69±0.05g/100g plant material). A.mollis aerial parts contain higher miquelianin (0.39±0.02g/plant material) and ellagic acid (1.56±0.01g/ plant material) than A.persica . activities of Alchemilla species.

Alchemilla vulgaris L. known as Lady's mantle, Bear's foot, Lion's foot in Europe is the most studied species of Alchemilla genus. This plant is used in folk medicine internally for menopausal complaints, gynaecological disorders such as dysmenorrhea and menorrhagia, menstrual pain, gastrointestinal disorders, inflammation, weight loss, and externally as a gargle for mouth and throat inflammation as well as for wounds and skin disorders such as eczema and rashes (Said et al., 2011;Ozbek et al., 2017). European Pharmacopoeia described A. vulgaris aerial parts and Commission E approved usage of this plant for treatment of non-specific diarrhea (Blumenthal et al., 1998). On the basis of clinical studies and long-term usage, A. vulgaris is suggested against non-specific diarrhea, gastrointestinal disorders and dysmenorrhea by ESCOP Monographs (Mills, Hutchins, 2013).
In order to reveal wound healing activities of A.mollis and A.persica, which were established previously, due to inhibition of inflammation and prevention of microbial infections, we aimed to evaluate antimicrobial and antiinflammatory activities in current study.
Chemical composition of Alchemilla mollis (Buser) Rothm. and Alchemilla persica Rothm. aerial parts and roots were investigated in this study in order to define responsible compounds for mentioned activities. Miquelianin (1) and ellagic acid (2) were isolated from the ethylacetate part of the methanol: water extracts of A.mollis aerial parts by using chromatographic techniques. Catechin, ellagic acid and miquelianin were analyzed by HPLC qualitatively and quantitatively. Antimicrobial activities of the A.mollis and A.persica as well as miquelianin and ellagic acid were evaluated against Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212, Bacillus subtilis ATCC 6633,Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853 and Candida albicans ATCC 10231. Anti-inflammatory activity of the extracts and isolated compounds were also tested by measuring their HRBC membrane stabilizing effects.

Plant Material
A.mollis was collected from Sivas, Sarıyar in 2013, while A.persica was collected from Erzincan, Kop Passage in 2015. A.mollis was identified by Mehmet Tekin and deposited in Cumhuriyet University, Faculty of Science Herbarium (CUFH 1344). A.persica was identified by Hayri Duman and deposited in Ankara University Faculty of Pharmacy Herbarium (AEF 26371). Aerial parts and roots of the plants were dried at room temperature separately and then ground.

Extraction and Isolation
Dried and powdered aerial parts of A.mollis (995.96 g) were macerated with 1500 mL; roots of A.mollis (54.71 g), aerial parts of A.persica (51.71 g) and roots of A.persica (51.44 g) were macerated with 150 mL methanol:water (80:20) solvent system during 24 hours at room temperature separately, stirred in ultrasonic bath for 30 minutes and then filtered. Extracts were evaporated under vacuum at 35-45 °C and then lyophilized. Crude extract of A.mollis aerial parts (284.4801 g) was dissolved in water and subjected to liquid-liquid extraction by petroleum ether, chloroform and ethylacetate, to yield 5.4229 g; 15.0525 g; 64.7894 g of corresponding extracts, respectively.
Fractions 14-16 obtained from silica gel column were applied to Sephadex LH-20 (25-100 μm, Sigma-Aldrich) column and eluted with methanol. Quercetin glucuronide, namely miquelianin (1) and antifungal activity test respectively. Microbroth dilution method was used for determination of the minimum inhibitory concentrations (MIC). The cultures were obtained in Mueller Hinton Broth (Difco, Difco Laboratories, Detroit, MI, USA). Serial two-fold dilutions ranging from 10 to 0.078 mg/mL were prepared in medium. A set of wells containing only inoculated broth, 20 % DMSO, ampicillin, ofloxacin and fluconazole were used as control. After incubation for 18-24 h at 35±1°C for bacteria-48 h for fungi, the last well with no microbial growth was recorded to represent MIC value (mg/mL).

Anti-inflammatory Activity
The study protocol was approved by the ethics committees of the Faculty of Medicine of Ankara University, Ankara-Turkey (26.10.2015/16-695-15).
Preparation of Human Red Blood Cells (HRBC) Suspension: Fresh whole human blood was collected from healthy human volunteer who had not taken any anti-inflammatory or steroidal drug for 2 weeks prior the experiment and transferred to the centrifuge tubes. The tubes were centrifuged at 3000 rpm for 10 min. The packed cells were washed at least three times with equal volume of isosaline (0.85 %, pH7.2). The volume of the blood was measured and reconstituted as 10 % v/v suspension with isosaline.
Heat Induced Hemolysis: Membrane stabilizing activity of the extracts was assessed using heat-induced human erythrocyte hemolysis. A stock solution of each sample was prepared by dissolving 20 mg of extract/ acetylsalicylic acid with 1 ml of their solvents. The reaction mixture consisted of test sample (extracts) and 10 % HRBC suspension. Instead of test sample, only solvent was used as a control. Acetylsalicylic acid was used as a standard drug. All centrifuge tubes containing reaction mixture were incubated in water bath at 56 ºC for 30 min. Then, the incubation tubes were cooled under running tap water. The reaction mixture was centrifuged at 2500 rpm for 5 min. The absorbance of the supernatant was measured at 560 nm (Anosike, Obidoa, Ezeanyika, 2012). Then, the efficient concentrations of the samples were determined by diluting the solution and the half maximal inhibitory concentration (IC 50 ) values were calculated.

Statistical analyses
All the experiments were performed at least in triplicates for all the test samples. SPSS 23.0 was used to examine the results. Results were expressed as mean and standard deviation (SD). One-way analysis of variance was performed by ANOVA procedures. Significant differences between means were determined by the Fisher's least significant difference (LSD) test. P values less than 0.05 were reported as "statistically significant".

Total Phenolic Content
Total phenol content of the aerial parts and roots of A.mollis and A.persica was carried out using the method described in European Pharmacopeia 6.0. Aerial parts and roots of A.mollis and A.persica were powdered. 0.5 g of each material was heated with 150 mL water at water-bath for 30 minutes, then diluted to 250 mL with water and filtered. First 50 mL of the solution was discarded. 5 mL of each extract was diluted to 25 mL with water. Mix 2 mL of each solution with 1 mL phosphomolybdotungstic reagent (Sigma-Aldrich) and 10 mL water, then diluted to 25 mL with 29 % sodium carbonate (Atabay) solution. After 30 minutes final mixture's absorbance was measured at 760 nm as A 1 against the water as a blank solution.
Total phenolic content of the extracts were expressed as gallic acid and pyrogallol equation. For pyrogallol equation, 50 mg pyrogallol was dissolved in water and diluted to 100 mL with water. 5 mL of the solution was diluted to 100 mL with the same solvent. 2 mL of this solution was mixed with 1 mL phosphomolybdotungstic reagent (Sigma-Aldrich) and 10 mL water, then diluated to 25 mL with 29 % sodium carbonate (Atabay) solution. After 30 minutes, absorbance was read at 760 nm as A 3 , against the water. Following equation was used for calculation of total phenolic content.
Total phenolic content = 62.5 x A 1 x m 2 A 3 x m 1 m 1 = weight of test material (g), m 2 = weight of pirogallol (g) For gallic acid equation, standard compound, gallic acid was weighed as 50 mg and dissolved in 10 mL of water. Different concentrations of standard compound were prepared (0.5, 0.1, 0.05, 0.025 mg/ mL). 2 mL of each concentration was mixed with 1 mL phosphomolybdotungstic reagent (Sigma-Aldrich) and 10 mL water, then diluted to 25 mL with 29 % sodium carbonate (Atabay) solution. After 30 minutes, final mixture's absorbance was measured at 760 nm against the water as blank solution. Calibration curve was drawn and the equation was calculated. Linear correlation between concentration and absorbance was expressed as y = mx + n equation. Total phenolic content of the tested extracts was calculated by the following equation y=1.417x+0.067. y-absorbance, x-total phenolic content as gallic acid mg/g plant material equivalent.

Tannin Content
Tannin content of the aerial parts and roots of A.mollis and A.persica was carried out using the method described in European Pharmacopeia 6.0. Aerial parts and roots of A.mollis and A.persica were powdered. 0.5 g of each material was heated with 150 mL water at water-bath for 30 minutes, then diluted to 250 mL with water and filtered. First 50 mL of the solution was discarded. 5 mL of each extract was diluted to 25 mL with water. 2 mL of each solution was mixed with 1 mL phosphomolybdotungstic reagent (Sigma-Aldrich) and 10 mL water, then diluted to 25 mL with 29 % sodium carbonate (Atabay) solution. After 30 minutes, final mixture's absorbance was measured at 760 nm as A 1 , against the water as blank solution. 10 mL of each extract was stirred with 0,10 g hide powder for 60 minutes and filtered. 5 mL of each extract was made up to 25 mL with water. 2 mL of each solution was mixed with 1 mL phosphomolybdotungstic reagent (Sigma-Aldrich) and 10 mL water, then diluted to 25 mL with 29 % sodium carbonate (Atabay) solution. After 30 minutes, final mixture's absorbance was read at 760 nm as A 2 , against the water as blank. Tannin content of the extracts was calculated from following equation and expressed as pyrogallol equation.
Tannin content = 62.5 ˣ (A 1 -A 2 ) ˣ m 2 A 3 ˣ m 1 For determining tannin content as gallic acid, equivalent calibration curve of gallic acid was used and each extract was stirred with hide powder using the method described in European Pharmacopeia (2008) and final mixture's absorbance was read at 760 nm as phenolic compounds were not absorbed by hide powder against the water as blank. Tannin content of the extracts was calculated from the same equation by subtracting absorbance of phenolic compounds not absorbed by hide powder from absorbance of total phenolic content.

HPLC Analyses
HPLC analyses were carried out using Agilent LC 1260 chromatograph (Agilent Technologies, Darmstadt, Germany). The diode array detector (DAD) was set at wavelengths of 210 and 254 nm and peak areas were integrated automatically using Agilent Chem Station Software. ACE 5 C18 (250 mm; 4.6 mm; 5 µm) column ,v/v) was applied from 55th min to 60th min. The flow rate was 0.5 mL/min, column temperature was maintained at 25 ºC. The sample injection volume was 10 µL. Both aerial parts and roots of A. mollis and A. persica (0.5 g) were extracted with methanol:water (80:20) mixture in ultrasonic bath for 30 min. After filtration, all extracts were adjusted to 10 mL in volumetric flask. Each extract was filtered from 0.45 μm filters before injection to HPLC system. Catechin, gallic acid, chlorogenic acid and caffeic acid purchased from Sigma; ellagic acid and miquelianin obtained from A. mollis were used as standard compounds. To prepare calibration curves, peak area of each solution was plotted against the concentration. Tested compounds were prepared at six different concentration levels (0.01, 0.02, 0.05, 0.1, 0.2, 0.25 mg/mL). Triplicate injections of 10 µL of each standard solution were performed. The compounds were determined at 210 and 254 nm absorbances. LOD (Limit of detection) and LOQ (Limit of quantification) of each compound were determined as signal to noise ratio of 3 and 10 respectively and verified by six injections at 10 µL.

RESULTS AND DISCUSSION
The MIC values (mg/mL) of the methanol:water extracts of A. mollis and A. persica as well as miquelianin and ellagic acid were shown in Table I. Neither extracts nor compounds showed antibacterial activity against E.coli. Both of the aerial part extracts of the tested plants and also miquelianin didn't show antibacterial activity against B. subtilis. Root extract of A. mollis, ellagic acid and miquelianin didn't show antibacterial activity against E. faecalis. No antibacterial activity was observed against P. aeruginosa for ellagic acid and no antifungal activity was observed against C. albicans for root extract of A. mollis. Except these, extracts possessed activity having MIC values of 2.5-5-10 mg/mL, miquelianin and ellagic acid possesed activity having MIC values of 1.25-2.5-4-5 mg/mL against the tested microorganisms. Anti-inflammatory activity of the samples was determined by measuring stabilization capacity against heat induced hemolysis of HRBC membrane. The extracts showed significant protection of the HRBC membrane against hemolysis induced by heat compared to acetylsalicylic acid used as a reference drug. A.mollis aerial part extract (IC 50 =1.22±0.07 mg/mL) exhibited the highest HRBC membrane stabilizing effect almost 4 times lower than acetylsalicylic acid (IC 50 =0.29 mg/mL) (Table II). Total phenol and tannin content of the A. mollis and A. persica was tested according to the method described in European Pharmacopoeia 6.0 in current study. Gallic acid and pyrogallol are both used as standards. Results of our analysis revealed that the total polyphenol content of A. persica and A. mollis varied from 244±72.7 to 331.7± 117.7 mg GAE/ g and tannin content varied from 8.3±4.42 to 12.95±3.00 g GAE/100 g plant material. As pyrogallol equivalent, total phenolic content varied from 237.8±25.9 to 357±75.8 mg PGE/ g and tannin content between 9.35± 0.29 and 18.63 ± 0.29 g PGE/100 g plant material. The highest tannin and total phenolic content as pyrogallol equivalent were determined for both aerial parts of A.mollis and A. persica extracts (Table III and Table IV). Our research reveals that aerial parts of A.mollis exhibited the highest red blood cell membrane stabilization activity (IC 50 : 1.22±0.07 mg/mL) with a good protection against heat-induced lysis and followed by A.mollis root extract (IC 50 : 1.34±0.08 mg/mL). These results can be attributed to the higher phenolic and tannin content.  Total phenol contents of the A. mollis and A. persica were also determined and the highest content was observed in A. persica roots (331.69±117.27 mg GAE/ g plant material; 282.50±28.70 mg PGE/g plant material). Ellagic acid and miquelianin as well as catechin were also analysed by HPLC qualitatively and quantitatively (Table V) LOD and LOQ values of catechin, miquelianin and ellagic acid were also determined and are given in Table VI. Table V  persica, which were found to contain catechin while aerial parts did not. The highest content of the catechin was determined in A. persica roots (6.69±0.05 g/100 g plant material). On the other hand, ellagic acid and miquelianin were only detected in aerial parts of both species. Ellagic acid and miquelianin were detected as major compounds in HPLC chromatogram of the A. mollis and A. persica aerial part extracts. Miquelianin percentages of the both species were higher than A. vulgaris. However, ellagic acid content of the A. persica was lower than A. vulgaris while A. mollis contained a higher amount. Both miquelianin and ellagic acid were isolated from ethylacetate part of the methanol-water extract of A. mollis aerial parts using different chromatographic techniques in the present study. Structures of the isolated compounds were determined according to the NMR 1 H and 13 C NMR results and comparing with literature data (Li et al, 1999;Nugroho, Lee, 2014;Vijayalakshmi, Madhira, 2014). Previous studies have revealed that A. mollis contains miquelianin, hyperoside, isoquercetin, cis-and trans-tiliroside, sinocrassoside D2, rhodiolgin and gossypetin-3-O-β-D-galactopyranosyl-7-O-α-L-rhamnopyranoside (Trendafilova et al., 2011). In acetone/water extracts of A. mollis, pedunculagin, agrimoniin, sanguiin H-10, castalagin/vescalagin as monomeric and oligomeric ellagitannins as well as galloyl-bis-hexahydroxydiphenoyl (HHDP) hexose, gallic acid, chlorogenic acid, catechin, a procyanidin trimer, quercetin glucuronide (miquelianin), several other quercetin glycosides, kaempferol glucuronide and methylated quercetin glucuronide were determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Duckstein et al., 2013). Sesquiterpenes and diterpenes as well as phenolic compounds including catechin, gallic acid, chlorogenic acid, gallic acid glycoside, gallic acid methoxy glycoside, pedunculatin isomers, galloyl-HHDP-hexoside isomers, epicatechin, quercetin-3-O-glucuronide, sanguiin isomers, casuarictin, agrimoniin, procyanidin B1, kaempferol-3-O-rutinoside, methyl gallate, ellagic acid, digalloylgalloylgalloside, aromadendrin glucoside derivative were determined in A. persica using GC-MS and HPLC-MS techniques (Afshar et al., 2015). Ellagic acid was isolated from A. mollis aerial parts for the first time in the current study. European Pharmacopoeia describes A. vulgaris dried leaves as "Alchemillae herba" an official pharmaceutical drug. According to the European Pharmacopoeia and ESCOP monographs, Alchemillae herba should contain minimum 6.0 per cent of tannins expressed as pyrogallol, approximately 2.2-2.5 % flavonoid glycosides and the miquelianin namely as main flavonoid (European Pharmacopoeia, 2008;Mills, Hutchins, 2013). Additionally, according to Fraisse et al. (2000), levels of the total flavonoids were detected as 2.22 %, miquelianin 1.18 %, tannins 16.0 % and ellagic acid 0.36 % for A. vulgaris.  According to the literature survey, antibacterial activities of the A. mollis were tested against S. aureus, S. S. epidermidis, S. pyogenes, P. aeruginosa, K. pneumonia and E. coli previously. Water, ethanol and methanol extracts of the A. mollis displayed antibacterial activity against E. coli and P. aeruginosa in disc diffusion and well diffusion assays. Inhibition zones were determined as 9.33 ± 0.66 mm, 15.33 ± 0.66 mm and 9.33 ± 0.66 mm for water, ethanol and methanol extracts against E. coli and 22.67 ± 0.66 mm, 22.67 ± 1.33 mm for ethanol and methanol extracts against P. aeruginosa in disc diffusion test. Ethanolic and methanolic extracts of A. mollis antimicrobial activity were similar or greater than tested antibiotics; ovobiocin (23 mm), carbenicillin (22 mm), tetracycline (13 mm), bacitracin (11 mm), ampicillin (8 mm) and lincomycin (7 mm) in disc diffusion methods against P. aeruginosa. Same results were obtained in well diffusion method (Usta, Birinci Yıldırım, Uçar Türker, 2014 (Shahzad et al., 2015). Ellagic acid was determined as one of the active constituents of Syzygium aromaticum for its antimicrobial activity against oral pathogens such as Porphyromonas gingivalis, Prevotella intermedia, Streptococcus mutans, Actinomyces viscosus with 1.25 mg/mL MIC value (Cai, Wu, 1996). It has been reported that ellagic acid exhibited antibacterial activity against S.aureus ATCC 9538, S.epidermidis NCTC 11047, Micrococcus luteus NCTC 9341, E.coli NCTC 8196, B.subtilis ATCC 6633 with 0.63 mg/mL, 0.31 mg/mL, 0.31 mg/mL, 2.5 mg/mL, 0.63 mg/mL and antifungal activity against C.albicans ATCC 10231 with 1.25 mg/ mL MIC values (Thiem, Goslinska, 2004). Present study results revealed that similar results were obtained for ellagic acid. This compound displayed antibacterial activity against B. subtilis and S. aureus with 5 mg/mL, 1.25 mg/mL MIC values respectively and antifungal activity against C. albicans 2.5 mg/mL MIC value as shown in Table I. Martini et al. (2009) have reported that antimicrobial activity of ellagic acid on H.pylori G21 and H.pylori 10K was also observed at 0.002 mg/mL and 0.01 mg/mL concentrations, respectively. Ellagic acid inhibited growth of oral pathogens such as Streptococcus mutans, Streptococcus sanguis, Streptococcus salivarius, Actinomyces naeslundii, Actinomyces viscosus, Lactobacillus rhamnosus, Porphyromonas gingivalis and Bacteroides forsythus (Loo et al., 2010). Antibacterial activity of ellagic acid against Staphylococcus epidermatis, Bacillus cereus, Klebsiella pneumonia and Salmonella typhi was also tested using plate-agar method. Activity of the compound was determined as more effective than gentamycin and streptomycin. Minimum effective dose of ellagic acid was determined as 0.15 mg/mL while gentamycin and streptomycin were 0.3 mg/mL and 2 mg/ mL respectively (Ghudhaib, Hanna, Jawad, 2010).
According to current study results, A.mollis aerial parts contain tannin, miquelianin and ellagic acid in a proper percentage when compared to A.vulgaris as an official pharmaceutical drug. A. persica also contains a higher amount of tannin and miquelianin than A. vulgaris and a lower amount of ellagic acid. These results may explain the traditional usage of A. mollis in European traditional medicine for same purposes with A. vulgaris as a commercial drug "Herba Alchemillae" for its astringent, diuretic, antispasmodic, wound healing activities and against excessive menstruation (Trendafilova et al., 2011;Özbek et al., 2017).
In conclusion, antimicrobial activity was observed with aerial parts and roots of A. mollis and A. persica as well as miquelianin and ellagic acid as isolated compounds against tested microorganisms. All the plant extracts tested as well as ellagic acid and miquelianin stabilized the HRBC membrane against heat significantly. Miquelianin, ellagic acid and tannin content of the Alchemilla species could be responsible for their antimicrobial and anti-inflammatory activities. Additionally, tannin and miquelianin content of A. mollis was determined as high as A. vulgaris. These results may suggest the usage of A. mollis instead of A. vulgaris in countries where A. vulgaris does not grow naturally.