Extracts of medicinal plants with natural deep eutectic solvents: enhanced antimicrobial activity and low genotoxicity

Natural deep eutectic solvents (NADES) are a new alternative to toxic organic solvents. Their constituents are primary metabolites, non-toxic, biocompatible and sustainable. In this study four selected NADES were applied for the extraction of two medicinal plants: Sideritis scardica, and Plantago major as an alternative to water-alcohol mixtures, and the antimicrobial and genotoxic potential of the extracts were studied. The extraction efficiency was evaluated by measuring the extracted total phenolics, and total flavonoids. Best extraction results for total phenolics for the studied plants were obtained with choline chloride-glucose 5:2 plus 30% water; but surprisingly these extracts were inactive against all tested microorganisms. Extracts with citric acid-1,2-propanediol 1:4 and choline chloride-glycerol 1:2 showed good activity against S. pyogenes, E. coli, S. aureus, and C. albicans. Low genotoxicity and cytotoxicity were observed for all four NADES and the extracts with antimicrobial activity. Our results confirm the potential of NADESs for extraction of bioactive constituents of medicinal plants and further suggest that NADES can improve the effects of bioactive extracts. Further studies are needed to clarify the influence of the studied NADES on the bioactivity of dissolved substances, and the possibility to use such extracts in the pharmaceutical and food industry.


Introduction
One of the most important aims of green chemistry has been to find green solvents for extraction of bioactive compounds from natural sources in order to replace the currently used hazardous organic solvents. One of the eco-friendly alternatives are deep eutectic solvents (DES) and particularly the natural deep eutectic solvents (NADES). DES are mixtures of organic compounds that have melting points lower than those of the individual components of the mixture and are liquid at ambient temperature. In the case of NADES, the constituents of the eutectic mixture are natural compounds: primary metabolites, which are easily available, non-toxic, biocompatible and sustainable [1]. NADES have low vapor pressure, an advantage with respect to environmental and human health protection. This property, however, poses a serious problem to the recovery of the active ingredients from the extract. Recently, several studies have demonstrated that NADESs retained or even improved the biological activity of dissolved substances [2,3]. Therefore, the NADES could function as an active ingredient, and the extract could be directly used as part of cosmetic or pharmaceutical formulations, bypassing the difficulties of Open Access BMC Chemistry *Correspondence: bankova@orgchm.bas.bg Grozdanova et al. BMC Chemistry (2020) 14:73 solute recovery. The aim of the present study was to apply selected NADES for extraction of two popular Bulgarian medicinal plants: the mountain tea Sideritis scardica Griesb., and the broadleaf plantain Plantago major L. as an alternative to water-alcohol mixtures, and to evaluate the antimicrobial, cytotoxic and genotoxic potential of both NADES solvents and extracts.

Preparation of NADES
The NADES were prepared by mixing the components and subsequently stirring in water bath (300 rpm) combined with mild heating at 50 °C until a homogeneous liquid was formed [1].

Polarity measurements
The NADESs polarity was measured by the solvatochromic dye Nile red [4]. The dye was dissolved in each NADES in the concentration range 0.01-0.1 mM and absorption spectra were recorded. The λ max was used to calculate the molar transition energy ENR, based on the equation: ENR = hc NA /λ max = 28,591/λ max , (ENR in kcal/ mol, λ max in nm).

Density measurements
The NADESs density was determined as follows: 2 ml of NADES were put in a volumetric flask at 20 °C and the weight of the liquid was measured. The density was calculated using the formula: ρ = m NADES /V NADES , where ρ is density, g/ml at 20 °C, m NADES − weight, g at 20 °C and V NADES − volume in ml at 20 °C (2 ml). For each solvent the procedure was performed in duplicate.

Extraction
Air-dried plant material was ground using a coffee mill, the average particle size was 0.75 mm. The extraction was performed in a 2 ml Eppendorf tube with 50 mg of plant material and 1.5 ml solvent in an ultrasound bath (Elmasonic S 30 H), without heating, for 1 h. The mixture was then centrifuged at 13,000 rpm for 40 min and filtered through cotton in a 1 ml volumetric flask. This extract was further used for antimicrobial tests, and analyzed to determine the main groups of bioactive compounds in the extracts. Each extraction procedure was performed in triplicate.

Quantitative determination of total phenolics and total flavonoids
For measuring those two groups of bioactive compounds, previously reported spectrophotometric methods were used [5]. For blank: solution of respective NADES instead of the test sample was used in analogous procedures. Total phenolics content was estimated using caffeic acid as standard, and total flavonoid content with rutine as standard. Every assay was performed in triplicate.

Minimal inhibitory (MIC) and bactericidal (MBC) concentrations
The antimicrobial activity was studied by the broth microdilution method according to ISO 20776-1:2006. Briefly, bacterial and fungal inoculums with concentration 105 CFU/ml were added to 96-well plates containing BHIB or MHB loaded with twofold serial dilutions of pure solvents or extracts differing in the concentration of total phenolics. Pure solvents were applied in an equivalent concentration as for testing the antimicrobial activity. Plates were incubated overnight at 37 °C, excepting the plates with B. cereus and Y. enterocolitica, which were incubated at 30 °C, respectively 26 °C. Gentamicin, penicillin and tetracycline were used as reference antibiotics for bacteria and amphotericin B-for C. albicans, following the requirements of EUCAST. Experiments were performed in triplicate. MICs and MBCs were determined as described before [6].

Dehydrogenase (DEHA) activity
The DEHA activity of the test microorganisms was assessed by MTT-test (3-(4,5-dimethylthiazolyl-2)-2,5diphenyltetrazolium bromide, M2128-1G, Sigma-Aldrich). The method is based on the reduction of the MTT dye by the membrane located bacterial enzyme NADH: ubiquinone reductase (H+-translocation) to insoluble formazan crystals. Briefly, the treated and untreated bacterial, respectively fungal cells, were incubated for 2 h with MTT dye in a final concentration of 0.05 mg/ml. An equivalent volume of 5% HCOOH in isopropanol dissolved the formed crystals. Absorption was measured using ELISA reader (BioTek Elx800, USA) at 550 nm (reference 690 nm) against a blank solution.

MTT test-calculation of IC 50 and statistics
The MTT test was conducted according to Annex C, ISO 10993-5 [7,8]. Cells with a density of 1 × 10 5 ml −1 were seeded in 96-well plates (flat bottom, 100 µl/well). For cells to start exponential growth (log phase), plates were incubated for 24 h. After entering the log phase cells were exposed to NADES and their extracts at concentrations ranging between 2 and 0.004% volume fraction for 24 and 72 h. PBS was used as a solvent and 4 wells were used for each treatment. MTT (0.5 mg/ml final concentration) was added to each well, followed by a 2-h incubation at 37 °C. The medium above the cells was removed and 100 μl/well 2-propanol supplemented with 5% formic acid were used to dissolve the formed formazan crystals and as a blank solution. Absorption was measured at 540 nm (reference filter 690) on a microplate reader ELx800 (BioTek Instruments, Inc., United States). The IC 50 values (inhibitory concentration 50 which reduces vital cells by half ) were calculated with a non-linear regression analysis (inhibition dose-response model, variable slope) using the GraphPad Prizm software. Untreated cells were considered as negative control and normalized for 100%.

Genotoxic activity The method of neutral Comet Assay
The method of Comet Assay was performed under neutral conditions. CCL1 cells-control and treated with increasing concentrations of the tested NADES solvents and extracts for 24 h were mixed with 1.4% low-melting agarose and spread onto already pre-coated with 0.5% normal agarose microscopic slides. The microgels, covered with coverslips to assure equal distribution, were incubated at 4 °C for 10 min. The coverslips were removed after solidification of the microgels. This was followed by a 20-min incubation in a lysis buffer (

Extraction and evaluation of extraction efficiency
Four NADES were selected based on literature data of their polarity, close or higher than 70% ethanol. They are described in Table 1, with corresponding abbreviations, and data of their density and polarity measured with the solvatochromic dye Nile red (lower E NR values mean higher polarity, [4]. The dry aerial parts of both selected plants were extracted with the four NADES and 70% ethanol as a reference solvent. Ultrasound assisted extraction was applied to accelerate the process, because of the significant viscosity of the NADES. The NADES choline chloride-glucose 5:2 was too viscous and to enable mass transfer, water was added to make it suitable for extraction. It is known that addition of water decreases the viscosity of NADES and weakens the hydrogen bonding interaction between its components, but dilution under 50% usually does not lead to a matrix disruption into individual components [9].
The extraction yield was evaluated by measuring the total phenolics and total flavonoids in the extracts. Extraction with 70% ethanol under the same conditions was used as a reference. The results are presented in Table 2.
The results demonstrated that some of the NADES extracted more bioactive compounds than the classic water-alcohol mixture. The water-containing NADES XXGlH was the most effective one, it extracted more phenolics and more flavonoids from the mountain tea that 70% ethanol. In the case of plantain phenolics XXGlH was again the most effective solvent: it extracted 25% more phenolics, compared to the reference solvent. However, none of the four NADES extracted flavonoids from plantain more efficiently than the 70% ethanol. This difference could be explained considering the specific   [10], while the major flavonoids in P. major are flavone monoglycosides [11], which are less polar than diglycosides. As the chosen NADESs have polarities higher than 70% ethanol, the better extraction of more polar flavonoid diglycosides can be explained by this fact, at least to some extent.

Antimicrobial activity
Recently, it became clear that NADES components can be selected not only to fine-tune solvent physicochemical characteristics but also to improve the biological activity of dissolved active compounds [12,13]. That is why we decided to check the antimicrobial activity of the extracts, and to compare the results with the activity of ethanol extracts, based on total phenolics concentration. The minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC) of total phenolics in the extracts were measured. In addition, the metabolic activity of the microorganisms after the treatment was studied by measuring the dehydrogenase activity (DEHA) with the MTT test. The results of these tests are shown in Table 3. The most potent antimicrobial activity was observed for CAPD/S against S. pyogenes, E. coli, S. aureus and C. albicans with MICs between 0.098 and 0.39 µg/ml total phenolics. CAPD/P was active against S. aureus, E. coli and C. albicans (MICs were in the range 1.99-3.98 µg/ml total phenolics). The plantain extracts had MBC = MIC for the three test microorganisms against which they were active. In this case the DEHA activity cannot be determined because it is equal to zero, as there are no metabolically active cells present. The extracts XXPD/S, XXPD/P, XXGlH/S, and XXGlH/P were inactive against the tested microorganisms. Surprisingly, the extracts with the NADES which was most effective in the extraction of phenolics and flavonoids, XXGlH, showed no antimicrobial activity. This could be due to qualitative differences between the extracts obtained with different solvents, however, qualitative analysis of the extracts was not performed. The present work aimed to find the most effective NADES for extraction of S. scardica and P. major in terms of antimicrobial potential. Optimization of the extraction conditions with the most effective solvent is the subject of future work, when we intend to apply more detailed analysis of individual constituents.
In most cases where antimicrobial activity was observed, it was much higher than the one of the traditional ethanol extracts. Such antimicrobial potentiating effects of NADESs was also observed in the case of propolis NADES extracts [14]. Although the addition of more than 50% of water to NADES during the antimicrobial tests breaks the NADES supramolecular complex, the individual NADES constituents could contribute to the overall effect of the solution [15] and some synergistic effects are also to be considered. Especially in the case of the CAPD extract of S. scardica, the increased effect compared to the other extracts can be partly explained by the effect of the presence of citric acid as an element of the NADES. In the literature there are indications that NADES containing organic acids possess higher antimicrobial activity, since some organic acids present many pharmacological effects [16].

Genotoxicity and cytotoxicity
Any further implication of the tested four NADES in the practice requires proofs for their genotoxicity safety. Many data in the literature show the genotoxicity of a broad range of food, cosmetic and pharmaceutical additives regardless of their nature [17][18][19]. Some data point to the fact that these substances added to the products consumed by or applied on people exert their genotoxicity at allowed concentrations [17,20,21]. This unambiguously entails the application of sensitive tests for fast and accurate evaluation of the potential of any substance that is planned to be implicated in food, pharmaceutical and cosmetic practices to induce general cytotoxicity via inhibiting cell proliferation and/or damage in DNA [22][23][24]. The method of Comet Assay is a brilliant technology for fast and sensitive analysis of genotoxicity [25,26]; it requires single cells, is fast and with high precision evaluates all kinds of DNA damages. Data quantification allows precise estimation of genotoxicity. In this study, the tested NADES solvents and the extracts which demonstrated antimicrobial activity: the extracts with XXGly and CAPD of S. scardica and P. major, were tested for their genotoxicity. The preparation for this test first required the evaluation of the in vitro cytotoxicity of the tested substances. Therefore, the in vitro cytotoxicity of the NADESs and extracts was determined on the normal mouse fibroblast cell line CCL-1 (Table 4) for two exposure times-24 and 72 h. The IC 50 values calculated after 24 h exposure time were used for performing the genotoxicity assay. The calculated IC 50 concentrations of all tested samples are given in Table 4.
As visible from the median inhibitory concentrations in Table 4, the solvent XXGly was less cytotoxic than CAPD. The same trend was observed for the relevant extracts. Interestingly, the cytotoxicity of the solvents XXGly, CAPD and XXGIH diminished with time and was less pronounced after 72 h of exposure than after 24 h. All extracts showed dose and time dependent cytotoxicity. The XXGly/S extract exhibited more potent antiproliferative effect on mouse fibroblasts than XXGly/P and was more cytotoxic to the cells than the solvent itself in both exposure periods. The cytotoxicity of XXGly/P was more pronounced than that of the solvent after the longer exposure period (72 h). CAPD/S showed a twofold higher antiproliferative effect on the cells after 24 h of exposure than CAPD. However, after 72 h the effect of CAPD/S diminished slightly compared to that of the solvent at the first incubation period. The cytotoxicity of the CAPD/P extract was time-dependent and the IC 50 values were significantly lower than those of the pure solvent, especially after long exposure time. Considering the MIC values determined by the BMD assay, it can be seen that CAPD/P exhibited anti-staphylococcal and antifungal activities at concentrations 3.9 and 1.99 µg/ml tP, respectively, that are significantly lower than the median inhibitory concentration cytotoxic for normal mouse fibroblasts (IC 50 /24 h = 4.6 µg/ml tP).
After estimation of the cytotoxicity of the studied NADES solvents and extracts, CCL1 cells were subjected to Comet assay. CCL1 cells were treated with the four tested NADES and extracts of S. scardica and P. major with XXGly and CAPD, for 24 h at optimal conditions. Both the pure NADES and the NADES extracts were applied to the monolayer cells at concentrations of IC 50 , ½IC 50 and ¼IC 50 , estimated by MTT tests (Table 4). Untreated cells were used as negative control while CCL1 cells treated for 30 min at 37 °C with 5 mM H 2 O 2 were used as a positive control for genotoxicity.
The Comet Assay results were quantified with the CometScore software and the values for the Olive Moment are presented in Fig. 1. Results demonstrated that all NADESs at a concentration of IC 50 showed moderate to subtle genotoxicity effect except XXGly that at a concentration of IC 50 demonstrated the highest genotoxic potential in comparison to all other solvents applied at IC 50 . Dilution two and four times of the applied IC 50 of all solvents showed lack of genotoxic potential on the tested CCL1 cells. The less genotoxic were all tested concentrations of CAPD (Fig. 1a). Further, extracts of S. scardica and P. major with XXGly and CAPD were tested for genotoxicity with the method of Comet Assay and results are shown in Fig. 1b. All extracts applied at concentration of IC 50 demonstrated genotoxicity on CCL1 cells, especially CAPD/S. Specifically, XXGly/S and XXGly/P, CAPD/S and CAPD/P showed a decrease in the detected genotoxicity when applied at ½ and ¼ of IC 50 . The most harmless was CAPD/P. These NADES probes demonstrated a lack of genotoxicity at all tested concentrations. Conserning the detected genotoxicity the NADES solvents can be classified in a row showing an increase in genotoxicity with an increase in the tested concentrations of the compounds. The NADES solvents and extracts are listed in the direction left to right, which marks increased genotoxicity. The NADES solvents are arranged as follows: CAPD < XXGIH < XXPD < XXGly. The extracts can be classified like this: CAPD/P < XXGly/P < XXGly/S < CAP D/S. Generally, all tested NADES revealed little changes in the distribution of cells in the different phases of the cell cycle after treatment for 24 h with increasing concentrations. The closer look at the graph in Fig. 2a shows that XXGIH at all tested concentrations, most explicitly at ½ and ¼IC 50 demonstrated decrease in the population of cells in G0/G1, suggesting a slight cytostatic effect. CCL1 cells treated with CAPD/S applied at IC 50 , and CAPD/P at all tested concentrations, especially at ¼IC 50 showed fewer cells in G0/G1. These results correspond with the detected genotoxicity in the samples. The detected genotoxicity led to moderate cytostatic effect on the CCL1 cells.
All tested NADES proved harmless genome integrity of the tested CCL1 cells. The detected little changes in the percentage of cells in different phases of the cell cycle can be due to cell culture asynchronization.

Conclusions
In conclusion, our results confirm the promising potential of NADESs as solvents for extraction of the biologically active constituents of popular medicinal plants and confirm the suggestion that NADES can improve the biological effects of bioactive extracts [12]. Best extraction results for total phenolics for the studied plants were obtained using XXGlH, but surprisingly these extracts were inactive against all tested microorganisms and were not subjected to further studies of in vitro cytotoxic and genotoxic activity. The most effective were the extracts with CAPD. The presence of citric acid and some synergistic effects with Sideritis constituents may play a role, as CAPD extract of S. scardica was much more active compared to the respective P. major extract.