Inhibitory Effects of Selected Medicinal Plants on Bacterial Growth of Methicillin-Resistant Staphylococcus aureus

Methicillin-resistant Staphylococcus aureus (MRSA) is a serious threat to global public health due to its capacity of tolerate conventional antibiotics. Medicinal plants are traditionally used to treat infectious diseases caused by bacterial pathogens. In the present study, 16 medicinal plants were screened for antibacterial activities to preselect more effective species. Ethanol extracts of selected medicinal plants (Caesalpinia sappan L., Glycyrrhiza uralensis Fisch., Sanguisorba officinalis L., and Uncaria gambir Roxb) were partitioned successively with different solvents (n-hexane, chloroform, ethyl acetate, 1-butanol, and water). Disc diffusion assay and broth microdilution were performed to evaluate the antibacterial activities of plant extracts and fractions against Staphylococcus aureus strains. Furthermore, the cytotoxicity of the extracts and fractions was determined against the human hepatoma (HepG2) and human lung carcinoma (A549) cell lines using a trypan blue exclusion method. A few extracts and fractions showed significant inhibitory effects on the bacterial growth of all tested strains, including multidrug-resistance (MDR) clinical isolates. The ethyl acetate fraction of C. sappan had the most potent effects with minimum inhibitory/bactericidal concentrations (MIC/MBC) of 31.2/62.5 μg/mL and showed low cytotoxicity with over 90% cell viability in both cells. Our results suggest that medicinal plants have considerable potential as alternatives to conventional antibiotics.


Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is a prevalent pathogen that causes severe infections in both the healthcare and community settings [1]. MRSA infections have emerged as a global public health burden, substantially increasing morbidity and mortality rates. In hospital settings, Asian countries have reported more MRSA prevalence than countries in Europe [2,3]. The prevalence of MRSA in some Asian countries is particularly high. For example, 73% of clinical isolates in South Korea were identified as MRSA according to reports of a regional resistance surveillance program in 2011 [4,5].
MRSA shows resistance to beta-lactam antibiotics such as penicillin, cephalosporin, and carbapenem and, over time, has become resistant to a few other antibiotic classes including fluoroquinolone, macrolide, aminoglycoside, and clindamycin [6,7]. MRSA infections are difficult to treat due to the capacity of tolerate conventional antibiotics [3]. Thus, new antibiotic agents with different mechanisms of action are urgently needed to control MRSA infections. In fact, the World Health Organization (WHO) has classified MRSA as a high-priority pathogen for the development of novel antibiotics [8].
To search for effective alternatives of antibiotics to treat bacterial infections, plant resources have shown great potential. The use of medicinal plants is still being considered a conventional treatment for diseases or syndromes worldwide [9,10]. Medicinal values

Results and Discussion
In the present study, we confirmed the antibacterial activities of plant extracts and fractions from medicinal plants. To preselect the more effective species, antibacterial activities of ethanol extracts from 16 medicinal plants were evaluated with a broth microdilution method ( Figure S1). Four medicinal plants were selected based on their inhibitory effects on bacterial growth of S. aureus strains. Information on these 16 species of medicinal plants is presented in Table 1. Ethanol extracts of medicinal plants were prepared and partitioned according to the above process ( Figure 1). Extraction yields using ethanol for 16 medicinal plants are shown in Table 1. The highest extraction yield was obtained from S. baicalensis (62.37%), followed by that from U. gambir (61.84%). Water and butanol fractions showed higher average yields at 15.40% and 9.95%, respectively, than ethyl acetate fractions (6.80%) and chloroform fractions (3.39%). Hexane fractions had the lowest average yield, at 0.56%. As the polarity of fraction solvents increased, more compounds or polar compounds with a higher mass were extracted from plants. These results are in line with previous studies showing that compounds in plant extracts are predominantly of intermediate and high polarity [16,17]. erage yields at 15.40% and 9.95%, respectively, than ethyl acetate fractions (6.80%) and chloroform fractions (3.39%). Hexane fractions had the lowest average yield, at 0.56%. As the polarity of fraction solvents increased, more compounds or polar compounds with a higher mass were extracted from plants. These results are in line with previous studies showing that compounds in plant extracts are predominantly of intermediate and high polarity [16,17]. The inhibitory effects of selected medicinal plants were tested by antibacterial bioassays. Two reference strains (MSSA and MRSA) and two clinical isolates (MRSA and MDRSA) of S. aureus were used in these assays, showing different antibiotic resistance patterns (Table 2 and Figure S2). Thus, bioactive extracts with a novel mechanism of action could be confirmed. MDR was defined as being resistant to at least three different antibiotic classes [18].  The inhibitory effects of selected medicinal plants were tested by antibacterial bioassays. Two reference strains (MSSA and MRSA) and two clinical isolates (MRSA and MDRSA) of S. aureus were used in these assays, showing different antibiotic resistance patterns (Table 2 and Figure S2). Thus, bioactive extracts with a novel mechanism of action could be confirmed. MDR was defined as being resistant to at least three different antibiotic classes [18]. Antibacterial activities of extracts and fractions were assessed by disc diffusion method. The diameter of a clear zone was measured and recorded. The results are shown in Tables 3 and 4. All S. aureus strains were extremely susceptible to rifampicin. Distilled water exhibited no antibacterial activity. Antibacterial activities showed the following order from high to low: C. sappan, U. gambir, G. uralensis, and S. officinalis extracts ( Figure S3). C. sappan extract showed a remarkable antibacterial activity with much wider inhibition zones (23.27-25.53 mm) at a concentration of 40 mg/mL and even at a lower concentration of 2.5 mg/mL. In contrast, S. officinalis extracts showed no antibacterial activities at concentrations below 20 mg/mL. We confirmed that the susceptibilities of each bacterial strain to different extracts of selected medicinal plants showed no significant differences ( Figure 2). In other words, plant extracts might possess different mechanisms of action compared to conventional antibiotics. Thus, they could be active against bacterial strains regardless of their resistance [19,20].   Disc diffusion assays were also performed using fractions of selected medicinal plants at a concentration of 10 mg/mL. According to results presented in Figure S4, hexane and chloroform fractions only showed inhibition zones for G. uralensis. Ethyl acetate fractions showed more potent activities than other fractions. Butanol fractions showed antibacterial activities against C. sappan and U. gambir. In contrast, water fractions of all plant species and S. officinalis fractions showed no antibacterial activities. The ethyl acetate fraction of C. sappan resulted in the largest clear zones with diameters of 23.10-23.83 mm. The butanol fraction of C. sappan also showed significant activities with diameters of 16.40-17.77 mm for clear zones. These results are considered to be due to different classes of phytochemicals extracted by different solvents and plant materials. Compounds extracted from medicinal plants are mainly dependent on the type of solvent used in the extraction process. For example, polar solvents are used for extracting phenolic compounds, their glycosides, and saponins, whereas non-polar solvents are used to extract fatty acids and Disc diffusion assays were also performed using fractions of selected medicinal plants at a concentration of 10 mg/mL. According to results presented in Figure S4, hexane and chloroform fractions only showed inhibition zones for G. uralensis. Ethyl acetate fractions showed more potent activities than other fractions. Butanol fractions showed antibacterial activities against C. sappan and U. gambir. In contrast, water fractions of all plant species and S. officinalis fractions showed no antibacterial activities. The ethyl acetate fraction of C. sappan resulted in the largest clear zones with diameters of 23.10-23.83 mm. The butanol fraction of C. sappan also showed significant activities with diameters of 16.40-17.77 mm for clear zones. These results are considered to be due to different classes of phytochemicals extracted by different solvents and plant materials. Compounds extracted from medicinal plants are mainly dependent on the type of solvent used in the extraction process. For example, polar solvents are used for extracting phenolic compounds, their glycosides, and saponins, whereas non-polar solvents are used to extract fatty acids and steroids [21,22]. On the other hand, although water fraction simultaneously had the highest yields, they showed poor inhibitory effects. These results suggest that massive amounts of compounds might be present in more polar fractions. However, not all compounds show antibacterial activities [23].
Antimicrobial mechanisms of plant extracts are not clearly understood yet. Meanwhile, we confirmed that inhibition zones of plant extracts were dependent on their concentrations and fraction solvents. Our results showed that ethanol extracts of G. uralensis, S. officinalis, and U. gambir had an equal MIC value of 250 µg/mL by the broth microdilution method (Table 5). However, differences in inhibition zones of extracts were observed (Table 3). This could be caused by difficulty in diffusion through agar of G. uralensis and S. officinalis extracts because the major active compounds of these extracts possessed high MWs. The diameter of the clear zone measured by the disc diffusion method is dependent on the degree of diffusion of the active substance. Previous studies have reported that the degree of diffusion is affected by the MW and polarity of compounds [23,24]. The above limitation of the disc diffusion method might have caused differences in the antimicrobial activity between different types of extracts. Disc diffusion method could be inappropriate for testing antimicrobial activities of plant extracts due to factors that might adversely affect the accuracy of results [24,25]. Thus, we determined MIC values to accurately compare the effectiveness of different plant extracts. It is widely considered that extracts having MIC values below 8 mg/mL possess some antimicrobial activities, while extracts with MIC values below 1 mg/mL exhibit remarkable antimicrobial activities [7]. According to results presented in Figure S1, half of tested extracts showed remarkable antibacterial activities. Especially, four ethanol extracts (C. sappan, G. uralensis, S. officinalis, and U. gambir) exhibited more significant activities with lower MIC values ( Table 5). The ethanol extract of C. sappan showed the best activity with MIC values of 62.5 µg/mL for all tested strains. Among various fractions, hexane, chloroform, and water fractions showed little activity. However, hexane and chloroform fractions of G. uralensis inhibited the bacterial growth of S. aureus, with MIC values of 62.5 and 31.2 µg/mL, respectively. MIC values of ethyl acetate fractions were determined as follows: C. sappan (31.2 µg/mL), G. uralensis (125 µg/mL), S. officinalis (500 µg/mL), and U. gambir (62.5 µg/mL). The MIC values of butanol fractions were C. sappan (62.5-125 µg/mL), S. officinalis (250 µg/mL), and U. gambir (125 µg/mL). The butanol and water fractions of G. uralensis showed little activity. A previous study has reported that the chloroform fraction of G. uralensis has 2.5-fold higher antimicrobial activity than the hexane fraction, with an MIC value of 0.1 mg/mL against MRSA strains, whereas the butanol and aqueous fractions exhibit no antibacterial activity [26]. Furthermore, prominent antibacterial activity was observed for extracts and fractions from C. sappan. Brazilin, one of the major active compounds in C. sappan, has been reported to show noteworthy antibacterial activity against antibiotic-resistant bacteria, including MRSA [27].
The MBC/MIC ratio determines whether extracts and fractions are bactericidal or bacteriostatic. The results are interpreted as bactericidal for MBC/MIC ≤ 4 or bacteriostatic for MBC/MIC > 4 [28]. The MBC/MIC ratio is shown in Table 5 and Figure 3. Ethanol extracts and water fractions of S. officinalis revealed bacteriostatic activities. Other extracts and fractions showed bactericidal activities. These results indicated that extracts and fractions of selected medicinal plants might act as potent bactericidal agents. Bactericidal agents are preferred over bacteriostatic agents clinically because killing bacteria is considered to be more effective for controlling infections than inhibiting bacterial growth [29].  To elucidate the inhibitory effects of extracts and fractions, a bacterial growth curve was determined toward MRSA 33,591 and CI-2. As shown in Figure 4, ethanol extracts of selected medicinal plants had significant bacterial growth inhibition with a wide range of concentrations. All tested bacteria showed dose-dependent growth inhibition. The time-lag with 31.2 µg/mL of C. sappan extract reaching to the exponential phase was changed from 8 h to 10 h compared to BHI broth (control). However, our results demonstrated that fractions showed much stronger inhibitory effects than ethanol extracts. The ethyl acetate fraction of C. sappan and chlorform fraction of G. uralensis extended the lag phase by 2 h and 4 h at a concentration of 15.6 µg/mL, respectively ( Figure 5).   The cytotoxicity of extracts and fractions was determined using the trypan blue exclusion method. Trypan blue is a negatively charged dye and only stains non-viable cells in blue [30]. HepG2 and A549 cells were treated with MIC concentrations of extracts and fractions for 24 h. The number of viable cells was counted, and cell viability was calculated with the percentage of living cells and dead cells. As shown in Table 6, ethanol extract and ethyl acetate fraction of C. sappan had low cytotoxicity (>90% cell viability) at concentrations of 62.5 and 31.2 μg/mL, whereas other extracts and fractions showed mild cytotoxicity (60-90% cell viability) [31]. Cell viability increased at lower doses, and both cell lines showed a slight difference in the viability between different fraction solvents and plant materials but with no significant differences. These results suggest that plant extracts may be relative efficient and safe for therapeutic use. However, normal cell lines were not used in the present study and further studies are needed to demonstrate the safety of plant extracts. It is critically important to note that in vivo efficacy and toxicity of plant extracts may not reflect in vitro properties due to other pharmacokinetic and pharmacodynamic factors [32].  The cytotoxicity of extracts and fractions was determined using the trypan blue exclusion method. Trypan blue is a negatively charged dye and only stains non-viable cells in blue [30]. HepG2 and A549 cells were treated with MIC concentrations of extracts and fractions for 24 h. The number of viable cells was counted, and cell viability was calculated with the percentage of living cells and dead cells. As shown in Table 6, ethanol extract and ethyl acetate fraction of C. sappan had low cytotoxicity (>90% cell viability) at concentrations of 62.5 and 31.2 µg/mL, whereas other extracts and fractions showed mild cytotoxicity (60-90% cell viability) [31]. Cell viability increased at lower doses, and both cell lines showed a slight difference in the viability between different fraction solvents and plant materials but with no significant differences. These results suggest that plant extracts may be relative efficient and safe for therapeutic use. However, normal cell lines were not used in the present study and further studies are needed to demonstrate the safety of plant extracts. It is critically important to note that in vivo efficacy and toxicity of plant extracts may not reflect in vitro properties due to other pharmacokinetic and pharmacodynamic factors [32].

Preparation and Fractionation of Plant Extracts
Plant materials were blended to powder using a home grinder and soaked in 70% ethanol with shaking (110 rpm) for 24 h. The ratio of plant materials to solvent was 1:10 (w/v). Crude extracts were centrifuged with 3000× g rpm for 30 min. Supernatants were collected and concentrated using a rotary evaporator WEV-1001V (Daihan Scientific Co., Wonju, Korea) in vacuum at 50 • C. The concentrated residue was dissolved in 10% dimethyl sulfoxide (DMSO; Sigma Chemical Co., St. Louis, MO, USA) and finally filtered through a Whatman filter paper No. 2 (Whatman, Kent, UK) to obtain ethanol extracts.
For the preparation of fractions, ethanol extracts were suspended in distilled water and partitioned successively with different solvents (n-hexane, chloroform, ethyl acetate, and 1-butanol) (Sigma Chemical Co., St. Louis, MO, USA) in the order of increasing polarity. Each layer was prepared following the above process. All prepared samples were collected into conical tubes and stored in a refrigerator at 4 • C until further use. The percentage of extraction yield was calculated with the following equation: extraction yield (%) = (dry weight of extract/dry weight of plant material) × 100.

Bacteria Culture
Antibacterial activities of samples were evaluated against standard strains of S. aureus together with MRSA clinical isolates. Standard reference strains (S. aureus ATCC 29,213; MRSA ATCC 33,591) were purchased from American Type Culture Collection (ATCC; Manassas, VA, USA). MRSA clinical isolates were originally isolated from clinical specimens and identified in Gachon University Gil Medical Center (Incheon, Korea) [1]. These isolates were preserved in a −80 • C freezer in 20% glycerol (v/v) until further use. Each bacterium was initially cultivated on brain heart infusion (BHI; Kisan Bio, Seoul, Korea) agar medium in the plate. Before assays, a single colony was picked from each plate and incubated in BHI broth with shaking (110 rpm) at 37 • C for 24 h.

Disc Diffusion Assay
Disc diffusion assay was performed to evaluate antibacterial activities following a previous study [34]. Briefly, each bacterial suspension (1 × 10 8 CFU/mL) was mixed well with 100 mL of fresh BHI medium containing 1.5% agar (w/v). The mixture consisting of bacterial suspension and BHI medium was poured onto Petri dishes. While solidifying the medium, 100 µL of each sample was loaded onto each paper-disc (8 mm/diameter). Rifampicin (30 µg) (Sigma Chemical Co., St. Louis, MO, USA) was used as a positive control. Loaded paper discs were gently placed onto plates. These plates were then incubated at 37 • C for 24 h. Diameters of clear zones were measured after incubation. The experiments were performed in triplicate.

Determination of Minimum Inhibitory Concentration (MIC)
Minimum inhibitory concentration (MIC) was determined using the broth microdilution method according to Bostanci et al. [35] with slight modifications. Briefly, 200 µL of the samples was inoculated to first wells of 96-well microplates and serially diluted two-fold (ranging from 2000 to 3.9 µg/mL). Then, 100 µL of each bacterial suspension (1 × 10 6 CFU/mL) was inoculated to wells. The total volume was 200 µL per well. Rifampicin was used as a positive control. The optical density (OD) was measured at 595 nm with a spectrophotometer (Multiskan FC; Thermo Fisher Scientific, Waltham, MA, USA). Microplates were incubated continuously for up to 18 h for monitoring inhibitory effects of samples on bacterial growth. MIC was defined as the lowest concentration that inhibited visible growth of bacteria.

Determination of Minimum Bactericidal Concentration (MBC)
To determine minimum bactericidal concentration (MBC), the broth microdilution assay was used and 50 µL of the samples was taken from the well that inhibited visible growth of bacteria. Samples were transferred to a new microplate, and 150 µL of BHI broth was added to each well. Plates were incubated for 18 h. MBC was defined as the lowest concentration that killed 99.9% of bacteria.

Cell Culture
Human hepatoma, HepG2 (KCLB 88,065) and human lung carcinoma, A549 (KCLB 10,185) cell lines were purchased from Korean Cell Line Bank (KCLB, Seoul, Korea). These cells were grown in monolayers in 25 cm 2 flasks with 3 mL of Roswell Park Memorial Institute Medium (RPMI 1640; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA) and 1% penicillin and streptomycin (Sigma Chemical Co., St. Louis, MO, USA). They were cultured at 37 • C in a CO 2 cell chamber (MMM Group, Planegg, Germany) with 5% CO 2 . Upon reaching 80-90% confluence, sub-cultured cells were dissociated with a trypsin-ethylenediaminetetraacetic acid (Trypsin-EDTA; Gibco, Grand Island, NY, USA) solution. After centrifugation, the cells were re-suspended with culture medium to 1 × 10 6 cells/mL and plated on 25 cm 2 flasks. After 24 h incubation, logarithmic growth phase cells were used in the cell viability assay.

Cell Viability Assay
The cytotoxicity of samples was determined using the trypan blue exclusion method according to Ndlovu et al. [36] with slight modifications. The cells were seeded at a density of 5 × 10 4 cells/well in 24-well microplates and grown to attachment overnight. Subsequently, the cells were exposed to samples at their MIC concentrations. Each sample was diluted with RPMI 1640 and added to each well. Microplates were incubated at 37 • C with 5% CO 2 for 24 h. After incubation, trypan blue solution (Sigma Chemical Co., St. Louis, MO, USA) was mixed with trypsinized cell suspensions in a ratio of 1: 1. Ten microliters of the mixture were deposited on a hemocytometer. After counting cells, the percentage of living cells and dead cells was calculated. The experiments were performed in duplicate.

Statistical Analysis
Statistical analysis was performed using Microsoft Excel 2016 (Microsoft, Redmond, WA, USA) and SigmaPlot version 12.0 (Systat Software, San Jose, CA, USA). Average values were calculated as means and standard deviation (±SD). Statistical differences were assessed by analysis of variance (ANOVA). A p-value of less than 0.05 was considered statistically significant.

Conclusions
In the present study, we confirmed that ethanol extracts and fractions of medicinal plants had potent antibacterial activities against S. aureus strains, including MRSA clinical isolates. Among the tested samples selected with abilities to inhibit of S. aureus strains, the ethyl acetate fraction of C. sappan was the most effective. The fraction showed remarkable inhibitory effects on bacterial growth with MIC/MBC values of 31.2/62.5 µg/mL and exhibited low cytotoxicity with more than 90% survival rates on HepG2 and A549 cell lines, respectively. The chloroform fraction of G. uralensis also showed significant bacterial growth inhibition. The present study demonstrated a considerable potential of medicinal plants as alternatives to conventional antibiotics.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/molecules27227780/s1, Figure S1: Screening of ethanol extracts from 16 medicinal plants for antibacterial activities by broth microdilution method. Figure S2: Antibiotic susceptibility testing for S. aureus strains. Figure S3: Antibacterial activities of ethanol extracts from selected medicinal plants by disc diffusion assay. Figure S4: Antibacterial activities of fractions from selected medicinal plants by disc diffusion assay.