Study of Antioxidant Activity of Some Medicinal Plants Having High Content of Caffeic Acid Derivatives

Recently, there has been increasing interest in medicinal plants, due to their content of health-promoting compounds, e.g., caffeic acids derivatives. Hence, the aim of this work was to study the antioxidant activity of extracts obtained from the following medicinal plants: caraway (Carum carvi L.), coltsfoot (Tussilago farfara L.), dandelion (Taraxacum officinale F.H.Wigg.), lovage (Levisticum officinale L.), tarragon (Artemisia dracunculus L.) and white mulberry (Morus alba L.), characterized by their high content of caffeic acid derivatives. The water-ethanolic extracts were characterized on average by about 9 times higher contents of caffeic acid derivatives level than water extracts. Both in water and water-ethanolic extracts, the dominant phenolic acid was 5-CQA (5-O-caffeoylquinic acid) and 3,4-diCQA (3,4-dicaffeoylquinic acid), then CCA-1 (chicoric acid isomer 1), which appeared only in water-ethanolic extracts. Extracts from dandelion contained compounds such as CTA (caftaric acid), CCA-1 (chicoric acid isomer 1) and CCA-2 (chicoric acid isomer 2), which were not detected in other plant extracts examined in this work. The water-ethanolic extracts from coltsfoot and tarragon were characterized by a high content of di-caffeoylquinic acids, especially 3,4-diCQA and 3,5-diCQA, respectively. It has been stated that there is a positive correlation between caffeic acid derivatives and antioxidant activity (radical cation scavenging activity (ABTS) and radical scavenging activity (DPPH)), especially in water-ethanolic extract of medicinal plants.


Introduction
In recent times, interest in the application of plant extracts in the food industry has continued to grow, mainly because of their antioxidant properties and associated health-promoting effects for humans. These properties of extracts are connected to the presence of vitamins, minerals and different phenolic compounds in plants [1][2][3]. In the group of phenolic compounds, special attention should be paid to phenolic acids, which can counteract the development of coronary heart disease, inflammation, diabetes and cancer [3][4][5][6].
In plant tissues, there are a number of phenolic acids, including chlorogenic acids (CGA). Chlorogenic acids are a family of esters formed between certain phenolic acids (trans-cinnamic acids) Extracts of the above-mentioned medicinal plants could be used in the food industry to improve the quality and healthiness of products. The cheapest way to obtain plant extracts is preparation with water to extraction; although, better results are observed when using ethanol or methanol, however methanol, because of toxicity, cannot be used in food production.
Despite extensive research into the chemical composition of various plants, detailed information on the content of caffeic acid derivatives in herb and spice plant extracts is not collected in the scientific literature. Therefore, the authors undertook the task of thoroughly analysing a number of these better and less well-known plants in terms of the content of phenolic acids that are valuable for the human body and determining their amount depending on the extractant used in the work.

Material
The material comprised dried seeds of caraway (Carum carvi L.), as well as leaves of coltsfoot (Tussilago farfara L.), dandelion (Taraxacum officinale L.), lovage (Levisticum officinale L.), tarragon (Artemisia dracunculus L.) and white mulberry (Morus alba L.), and were purchased from a local producer and grower situated in the Lower Silesia. The plants came from the 2018 growing season. In order to prepare extracts, raw plant material was freeze-dried in lyophilizer (type 5411, BOC Ltd., Edwards, England) and crushed in an electric mill (FA-5485 SP-742, TZS First, Austria). Ground dry material was packed in plastic closable packaging, after which the samples were used directly for analyses (at room temperature).

Extraction
Extraction was carried out using water and 50% (v/v) ethanol in water. Extracts from medicinal plants were prepared according to the method of Bucić-Kojić et al. [31], with some modifications. The extracts were prepared as follows: 2 grams of ground plant material were mixed with 100 mL of distilled water (water extracts) or 100 mL of a mixture (1:1) of distilled water and ethyl alcohol (water-ethanolic extracts) with the addition of 0.1% NaHSO 3 . The obtained mixture was sonicated in an ultrasonic shaker (UM-2, Unitra-Unima Olsztyn, Poland) for 15 min and heated at 80 • C for 5 min. Afterwards, the mixture was cooled down to room temperature and stored at 4 • C for 12 h. Next, the mixtures were re-sonicated and centrifuged at 5000 rpm (MPW-351R, Mpw Med. Instruments, Poland) for 10 min (Figure 1). Finally, extracts were used for analyses of phenolic compounds and antioxidant activity (AA).
There is a lack of detailed information in the literature about quantitative analysis of other caffeic acids (especially of caffeic acid derivatives), so information described below are valuable in works on medicinal plants. It was reported that plant extracts were characterised by various contents of caffeic acid derivatives depending on the types of plant and solvent used (Table 1).
Among water extracts, coltsfoot extract contained the highest total amount of the analysed compounds (16.18 mg CQA·g −1 ), whereas very small amounts were noted in total white mulberry extract (0.18 mg CQA·g −1 ). Water-ethanolic extracts were characterised by higher amounts of caffeic acids compounds than water extracts (on average by about 9 times). The highest content of the total investigated compounds was found in the coltsfoot water-ethanolic extract (101.15 mg CQA·g −1 ) and then in the dandelion water-ethanolic extract (94.93 mg CQA·g −1 ). In the tarragon water-ethanolic extract, the amount of analysed compounds was also high and equal to 60.81 mg CQA·g −1 , while caffeic acid derivatives were not found in the water extract of tarragon. The lowest content of total analysed acids was found in water extract of white mulberry (0.18 mg CQA·g −1 ).
It has been found that among all other caffeic acid derivatives, the highest quantity was of CCA-1 acid in water-ethanolic extract of dandelion. However, generally, both in water and water-ethanolic extracts, the dominant phenolic acid was 5-CQA (5-O-caffeoylquinic acid). The content of that compound in water extracts ranged from 0.14 mg CQA·g −1 (white mulberry) to 9.41 mg CQA·g −1 (coltsfoot), while from 2.23 mg CQA·g −1 (caraway) to 28.92 mg CQA·g −1 (tarragon) in water-ethanolic extracts. 5-O-caffeoylquinic acid represented 40% to 73% of all of the extracted caffeic acid derivatives. It was stated that both extracts were characterised by the significantly smaller amounts of 3-O-caffeoylquinic acid and 4-O-caffeoylquinic acid among the group of caffeyolqunic acids (average about 7-times lower than compared to the amount of acid 5-CQA). It has been stated that the highest content of 3-CQA (3-O-caffeoylquinic) was in water-ethanolic extract of tarragon (3.86 mg CQA·g −1 ) and 4-CQA (4-O-caffeoylquinic) in water-ethanolic extract of coltsfoot (2.53 mg CQA·g −1 ). However, the content of both compounds was not found in water and water-ethanolic extract of dandelion and water-ethanolic extract of tarragon.
Extracts from dandelion contained compounds such as CTA (caftaric acid), CCA-1 (chicoric acid isomer 1) and CCA-2 (chicoric acid isomer 2), which were not detected in other plant extracts examined in this work. It was stated that the highest content of CCA-1 (chicoric acid isomer 1) was found in water-ethanol of dandelion (73.83 mg CQA·g −1 ). The water-ethanolic extract of this plant also contained CTA (11.97 mg CQA·g −1 ) and CCA-2 (4.73 mg CQA·g −1 ), while in the water extract, there was a small amount of CTA (0.39 mg CQA·g −1 ). Generally, it was stated that among the water extracts, the highest content of the total studied compounds was characterised by coltsfoot.
Investigation of phenolic compound contents in coltsfoot was carried out by Dobravalskytė et al. [37]. They conducted analyses of coltsfoot extracts and detected higher caffeoylquinic acid using the HPLC-UV method. Research conducted by this author confirmed the occurrence of phenolic acid in coltsfoot extracts, as well as other phenolic acids such as dicaffeoylquinic and quinic acid. Ivanov [38] found phenolic (e.g., chlorogenic acid and chicoric acid) in dandelion extracts; depending on the extraction method, the content of that compound ranged from 18 mg·100 g −1 d.m. to 37 mg·100 g −1 d.m and from 484 mg·100 g −1 d.m. to 3148 mg·100 g −1 d.m., respectively. It is noteworthy that better extraction results were obtained by using a 50% water-ethanolic mixture as the extractant than 95% ethanol, while part of the investigated compounds were not detected in water extracts. Tsai et al. [39] stated that 50% ethanol is most favourable for extracting individual caffeic acid derivatives from plant Echinacea purpurea. Mirjalili et al. [40] analysed extracts from the dried leaves of lovage, and confirmed that caffeoyoloquinic acids are the main phenolic acids in those extracts. The authors stated that the content of these compounds in dried lovage leaves was 2-60-fold lower in comparison to caffeoylquinic acids.
Lin and Harnly [41] confirmed, by using the LC-PDA-ESI/MS method, that tarragon contained caffeoylquinic acids (3-CQA, 4-CQA and 5-CQA). On the other hand, Khezrilu et al. [42] did not find any derivatives of those acids in tarragon extract, but they identified other phenolic compounds, i.e., gallic acid, p-hydroxy benzoic acid, vanillic acid, p-coumaric acid, syringic acid, ferulic acid and sinapic acid. However, it should be noted that the authors used methanol as a solvent in their research. Chu et al. [43] reported that phenolic acids in white mulberry included caffeic acid, gallic acid and chlorogenic acid. Memon et al. [44] reported that caffeoyoloquinic acids were the dominant acids in white mulberry leaf extracts and the amount of these compounds varied from 47.9 mg·100 g −1 to 64.73 mg·100 g −1 depending on extraction method. In the mentioned works, none used the water and water-ethanol methods to extract compounds from the six selected medicinal plants investigated in this research.
Based on the results presented in Figures 3 and 4, differences in the percentage concentration of all caffeic acid derivatives can be observed. In water extracts of caraway, coltsfoot and dandelion, caffeyolquinic acids represented 68% of the extracted caffeic acid derivatives.
In contrast, in water-ethanolic extracts of these plants, the content of other caffeic acid derivatives was dominant and represented about 78% of all extracted compounds. The content of other caffeic acid derivatives was not found in both the water and water-ethanolic extract from lovage and white mulberry, and the share of caffeyolquinic acids in both types of extracts was comparable. Differences between these two types of eluents can also be observed in tarragon. Caffeic acid derivatives (caffeyolquinic acids and other caffeic acid derivatives) were not detected in water extracts of tarragon, yet they represented about 58% and 42% of all phenolic compounds in water-ethanolic extract, respectively. It was stated that regardless of the eluent used, the dominant constituent in most of the analysed samples was 5-CQA (5-O-caffeoylquinic acid). Turkmen et al. [45], who examined the influence of extraction solvents on polyphenol concentrations, reported that 50% ethanol solvent is about 2.5-fold more effective than water. Also, Tan et al. [46] confirmed that the water-ethanol mixture is better for polyphenol extraction than distilled water, although they used a 60% ethanol-water mixture. Pudziuvelyte et al. [47] stated that the amount of compounds extracted from the plant is influenced by the type of solvent used to prepare the extract, the extraction process parameters and the anatomical parts of the plants used.
It was stated that the highest antioxidant activity in the analysed water extracts assayed with the ABTS method was determined in the coltsfoot extract (ca. 321 µmol TE·g −1 d.m.), whereas the lowest activity was detected in white mulberry extract (ca. 59 µmol TE·g −1 d.m.) ( Table 2).  The antioxidant activity (AA) of the water-ethanolic extracts from the investigated plants obtained by the ABTS method was about 1.6-fold higher on average compared to extracts obtained using water as an extractant. The biggest difference was observed in white mulberry extracts, where antioxidant activity was 3-fold higher. With regard to the content of caffeic acids derivatives, the highest AA among water-ethanolic extracts was also observed in tarragon extract (ca. 406 µmol TE·g −1 d.m.), which was characterised by the highest amount of caffeyolqunic acids (3-CQA, 4-CQA and 5-CQA) and dicaffeyolqunic acid (3,5-diCQA) ( Table 1). Both water and water-ethanolic extracts of coltsfoot were characterised by the highest AA measured by the DPPH method (ca. 52 µmol TE·g −1 d.m. and ca. 269 µmol TE·g −1 d.m., respectively) ( Table 2), which could be related to the fact that coltsfoot was characterised by the highest amount of dicaffeyolqunic acid (3,4-diCQA and 4,5-diCQA) and higher amounts of 3,5-diCQA and 5-CQA in water-ethanolic extract, as well as the highest content of these compounds in water extract (Table 1). Similar to the ABTS assay results, the lowest DPPH antioxidant activity among water-ethanolic extracts was determined in extracts from white mulberry, which also had the lowest AA among water extracts, probably because the white mulberry contained only a small amount of caffeyolqiunic acids (3-CQA, 4-CQA and 5-CQA) both in water and water-ethanolic extract ( Table 1). The AA of water-ethanolic extracts obtained by the DPPH assay was about 4-fold higher on average, compared to water extracts from the investigated plants.
Vallveedú-Queralt et al. [17] reported that the AA of caraway water-ethanolic extract obtained by ABTS was 2.32 mmol TE·g −1 and for the DPPH method 1.63 mmol TE·g −1 . Moreover, this extract was characterised by the strongest AA compared to extracts from turmeric, dill, marjoram and nutmeg. Dobravalskytė et al. [37] confirmed high antioxidant properties of coltsfoot extracts in their investigation. Biel et al. [48] demonstrated that dandelion has a low AA (25.9 µmol TE·100 g −1 d.m.), but they used other eluents in their work. However, Ivanov [38] reported that the AA of water-ethanolic extract from dandelion measured by the DPPH method was about 3-fold higher compared to water extract. Sepahpour et al. [49] stated that generally, the antioxidant activity (DPPH) of all the analysed medicinal plant extracts showed the lowest values when water was used for extraction in comparison to organic solvents (methanol, ethanol and aceton). The results presented by Mariutti et al. [50] showed that tarragon exhibits moderate antioxidant properties in comparison to herbs such as rosemary and thyme. However, they also showed that this plant was a stronger antioxidant than caraway, the tendency of which was also noticeable in this study. Mumivand et al. [24] reports that the DPPH activity of tarragon extracts ranged from 0.034 to 0.101 (IC50 mg-mL −1 ) depending on the place of harvest. Memon et al. [44] studied the antioxidant activity of white mulberry extract using the DPPH method and showed that the plant has an average antioxidant activity; however, using 80% methanol as an eluent, the activity of white mulberry extract significantly increased. Sotiropoulou et al. [51] stated that other parameters, like extraction temperature, time and amount of solvent, also play important roles in the evaluation of properties of herbal plants. Table 3 shows the correlations between the content of caffeic acid derivatives and antioxidant activity in water and water-ethanolic extracts from medicinal plants. It was stated that among the phenolic acids present in water-ethanolic extracts of investigated medicinal plants, the following acids: 3-CQA, 4-CQA, 5-CQA, 3,4-diCQA, 3,5-diCQA and 4,5-diCQA acid, were correlated with antioxidant activity (ABTS and DPPH). Correlation coefficients indicated that correlation in the majority was very high or relatively high. The highest correlation between phenolic acids and antioxidant activity (ABTS and DPPH) was found for 5-CQA acid (r = 0.995 ** and r = 0.858 **, respectively). Also, high correlation was found for 3,4-diCQA (r = 0.823 ** and r = 0.776 **) respectively, for both assays of antioxidant activity, ABTS and DPPH (Table 3). Regarding the water extracts, the 5-CQA acid content had the highest correlation with ABTS and DPPH (r = 0.623 ** and r = 0.601 *, respectively). Also, 4-CQA acid showed significant correlation for assays of antioxidant activity (ABTS and DPPH), while 3-CQA only for ABTS. However, the 3,4-diCQA (3,4-dicaffeoylquninic acid), 3,5-diCQA (3,5-dicaffeoylquninic acid) and 4,5-diCQA (4.5-dicaffeoylquninic acid) had no significant correlation for antioxidant activity in water extract (Table 3).
Many authors [13,27,52,53] reported that there is a significant correlation between the content of phenolic compounds (especially phenolic acids) in different plants (e.g., medicinal plants, potatoes) and antioxidant activity. Cai et al. [27] found a positive, linear relationship between antioxidant activity (ABTS) and total phenolic content (all R 2 values ≥ 0.95) in several medicinal herbs.
[53] reported a positive correlation (r = 0.69) between the content of phenolic compounds and antioxidant activity in Rosmarinus officinalis L., while Zhang et al.
[52] reported a positive correlation in methanolic extracts from S. miltiorrhiza. Ru et al. [13] demonstrated that in the free fraction, the content of each individual phenolic acid (in yellow, white, red and purple fleshed potatoes) was positively correlated with antioxidant activity. Cai et al. [27] stated that all analysed medicinal herbs exhibited far stronger antioxidant activity and contained significantly higher levels of phenolics than common vegetables and fruits.
Generally, the application of an appropriate extraction method in plants, including selection of the appropriate eluent to use these extracts in food production, is of great importance, both in terms of obtaining valuable compounds (including phenolic acids) from plants, increasing the antioxidant activity of these compounds and indicating the particular health-promoting plants to food producers and consumers.

Conclusions
It was found that the plant extracts obtained were characterised by different contents of caffeic acid derivatives, other caffeic acid derivatives and diversified antioxidant activities, which was influenced by both raw material and the type of eluent used. The water-ethanol extracts were characterised on average by about 9-times higher contents of caffeic acids derivatives level than water extract. Both in water and water-ethanolic extracts, the dominant phenolic acid was 5-CQA (5-O-caffeoylquinic acid) and 3,4-diCQA (3,4-dicaffeoylquinic acid), then CCA-1 (chicoric acid isomer 1), which appeared only in water-ethanol extract. Extracts from dandelion contained compounds such as CTA (caftaric acid), CCA-1 (chicoric acid isomer 1) and CCA-2 (chicoric acid isomer 2), which were not detected in other plant extracts examined in this work.
The water-ethanolic extract from coltsfoot and tarragon was characterised by a high content of di-caffeoylquinic acids, especially 3,4-diCQA and 3,5-diCQA, respectively. Among the water extracts, the highest content of the total studied compounds was characterised by coltsfoot. None of the caffeic acid derivatives group was found in the water extract from tarragon, although the amounts of the tested acids in the water-ethanolic extract from this plant were significant. The lowest content of total analysed acids was stated in water extract of white mulberry. It was stated that 5-O-caffeoylquinic acid was represented in 40% to 74% of all of the extracted phenolic acids. Antioxidant activity measured by ABTS and DPPH was higher (on average by almost 1.5-fold and 6-fold, respectively) in extracts obtained with eluent (ethanol) in comparison to water extracts. It was stated that there is a positive correlation between caffeic acid derivatives and antioxidant activity (ABTS and DPPH), especially in water-ethanolic extracts of medicinal plants. The highest correlation between 5-CQA and the antioxidant activity of ABTS and DPPH was found for both water-ethanolic and water extracts. In conclusion, extracts from coltsfoot, dandelion and tarragon appear to be more highly recommended for use in food production and human consumption.