Crystalline chloride, crystalline thiobarbituric acid (TBA) 99.5–101% of pharmacopoeial purity in the original packaging produced by Sigma, Germany were used for research. Trichloroacetic acid with a purity of 98% (Erba Lachema, Germany) was used. Malonaldehyde bis (dimethyl acetate) - purity for analysis of Sigma-Aldrich production, Germany. 98% crystalline sodium hydroxide, chemically pure produced by Brenntag CEE Austria. N-butanol for analysis was used 99.9% pure for analysis produced by Merck, Germany.

Plants Trifolium pratense L, which were prepared for research, were grown from the collection seeds of the Department of Botany of the Botany Department of Karazin National University, Kharkiv, Ukraine. Swards (grass) were collected in the suburbs of Zaporizhzhia (urban settlement Primorske [47°37'28"N, 35°17'39"E] during the 2019–2021 seasons. Plants were cultivated under the same conditions and harvested in the same period to exclude the dependence of changes in the structure of raw materials and leaf epidermis on external living conditions, seasonal weather conditions and age in the phenological period of full flowering (May-June). Pre-dried raw materials were dried under cover or in a well-ventilated place, making sure that the inflorescence did not dry out and crumble. We received 20% of dry raw materials.

The identification of the species was confirmed at the Botany Department of V. N. Karazin Kharkiv National University, Ukraine (Head of the Department, Herbarium Keeper, Assoc. Prof. Gamula Yu.G.).

For microscopic analysis used temporary preparations of leaves, stems and flowers, which were pre-boiled in a solution of sodium hydroxide (5%, aq.) without the use of dyes; cross-sections were made with a blade (Grechana et al. 2020). On temporary preparations we determined the shape of epidermal cells, type of respiratory system, vascular tissue, structure of trichomes, glandular (secretory) trichomes and more. Studies of the drugs were performed on a microscope “Axioskop 40” (Zeiss, Germany) at a magnification of 15 × 40. Photographs were taken with a Canon PS G7 (Japan), measurements were performed in 20 replicates using the licensed program AxioVision Rel. 4.7, statistical processing (average value, error, coefficient of variation) was performed using Microsoft Excel. Anatomical parameters of the epidermis were considered low-variable if the coefficient of variation of Cv is less than 20%, as moderately variable - at Cv > 20%, and highly variable - at Cv > 40% (Zakharevich 1964). The type of stomata was determined according to the classification of Baranova (1985), the description of epidermal cells was performed according to the method of SF Zakharevich (1964); Butnik and Timchenko (1987), the respiratory index was calculated by the formula of A. Kästner (Kästner 1972).

Determination of antioxidant activity was performed according to the readily available protocol for non-enzymatic lipid peroxidation (Kohn and Liversedge 1944; Guichardant et al. 2011) free radical (iron ion, II), described by Bayir et al. (2020) and Durojaiye and Adewale (2013). The results were stated using a spectrophotometer.

The grass of the plants was placed in the oven of the universal Memmert UF55 (Memmert, Germany) at 80 °C for 24 hours. The dried plant raw materials were ground to a powdery state with a pestle and mortar. Samples (100 mg) were extracted by extraction with aqueous-ethanol mixture (70%, 10.0 ml) and distilled water (5.0 ml and 10.0 ml), using a Soxhlet-type extractor Behr R 304 (Germany). Our previous studies have tested the optimal residence time of the mixture in the extractor with the highest yield of components to the extractant - 60 minutes - 24 hours. Based on this, we performed extraction during the day without heating - at room temperature 20 ± 2–3 °C. Three batches were prepared for each extractant. Accordingly, three extracts were obtained (alcohol, water 1:5 and water 1:10). The extracts were dried using a Freeze Dryer FD-10 (China). Stored at a temperature of 4 OS until the study. Immediately before measuring the level of TBARS, the extracts were diluted by adding 2 ml of distilled water.

As a substrate was used a suspension of egg lipoproteins (homogenized chicken egg yolk with 1 liter of phosphate buffer, pH = 7.4).

Two ml of tested extracts were added to a suspension of egg lipoproteins and 10% aqueous FeSO₄ × 7H₂O, mixed thoroughly and incubated with a Memmert UF55 oven (Memmert, Germany) for 60 min at 37 °C. To stop the reaction and prevent further oxidation of the medium, 2 ml of a 20% mixture of trichloroacetic acid with Trilon B was used. The samples were shaken for 1 h with a Wisebath bath shaker WSB-18-1257-00010 (DAIHAN Scientific, South Korean) and filtered laboratory deashed blue tape (LLC “VDP Aquachim”, Ukraine) using laboratory small recirculating pump water SHZ-DIII (Biobase, China). The filtrate was centrifuged for 30 min on a SM-6M centrifuge (ELMI SIA, Latvia), the supernatant was drained and used for research.

The supernatant (1 ml) was added to a 10 ml tube and mixed with 1 ml of 0.8% aqueous thiobarbituric acid (TВА). TBА solution, according to Ohkawa et al. (1979) and Wang et al. (2018), was prepare daily fresh.

The mixture was heated on a boiling (95 °C) water thermostatic bath VB-20 (MICROmed, Ukraine) for 60 minutes. The tubes were rapidly cooled to room temperature under running water. The formation of a pink chromogenic trimethyl complex was observed. The stained complex of TBA-active products (TBARS) was extracted with n-butanol. The concentration of TBARS was determined spectrophotometrically on a digital spectrophotometer UV-VIS PD-30000UV (Apel, Japan) in comparison with n-butanol at wavelength (λ = 532 nm). Determination of indicators for each sample was repeated (n = 6) according to the above procedure.

We selected the known antioxidant ascorbic acid (gamma-lactone 2,3-dehydro-L-gulonic acid, solution for injection, 50 mg / ml, 2 ml, PF Darnitsa, Ukraine). With two ml of ascorbic acid solution instead of plant extract did the same steps as described above with repeating the procedure, n = 6.

A “zero” study was performed, where instead of extracts 2 ml of distilled water were taken and in the repeatability n = 6 and all the above experimental actions were performed.

The calculation of antioxidant activity (AOA,%) was performed according to the formula 1:

$\mathrm{AOA}=\frac{\left(E_c-E_e\right)}{E_c}\times 100\%$ (1)

where: E c - optical density of the test extract, E e - optical density of the extractant TВARS (n-butanol).

Data were processed using statistical methods and evaluated both parametrically (by Student’s coefficient, t-test) and non-parametrically (by Mann-Whitney test - u-test). We presented a comparative evaluation of the data obtained using Pearson’s pairwise correlations and scale diagrams described by Vehkalahti and Everitt (2018).

The anatomical structure of the flower, the leaves of bract, petals and generative organs has been described in the literature (Pinar et al. 2001; Halbritter and Auer 2021) and in the State Pharmacopoeia of Ukraine (2019). Therefore, we did not focus on the widely described elements of the studied plant.

The leaves are trifoliate, tough. Leaflets have 2–3 cm in length and 1–1,5 cm in width, in obovate shape, pointed at the end, finely or unevenly dentate along the edges with a dense network of lateral veins that are thickened towards the edges.

The leaf is amphistomatous (stomata are present on the upper and lower leaf surfaces). Among the epidermal cells are pairs of sausage-shaped guard cells (pore or stoma). The morphometric characteristics of the leaf epidermal structures are given in Table 1.

Leaf adaxial epidermis (Fig. 1A) The main epidermal cells are polygonal or irregular, with rectilinear-rounded outlines, from 51.70 to 84.77 microns in length and from 38.64 to 61.97 microns in width. The anticlinal walls are arcuate or wavy. Epidermal cell projection is polygonal, rounded or flattened, compactly arranged. The major epidermal cells are thick-walled, their number varies from 347 to 425 per 1 mm². Intercellular spaces are absent. Adjacent boundaries angles are obtuse, rounded, pointed, and straight.

The stomata are solitary, chaotically arranged. The stomata apparatus have anomocytic type. Stomata have surrounded by 3–5, more often 4 cells. Peri-stomata cells that neighbor with guard cells do not differ in shape and size from the main epidermal cells. A definite orientation of the peri-stomata cells has not been observed, however, in some cases, ones are located in a certain way: two cells are border with the guard cells lateral sides (parallel into stomata long axis) and two - with the stomata poles. Nevertheless, since the peri-stomata cells do not differ in structure, shape, and size from the rest of epidermis cells, in this case, we also referred to this type of stomatal apparatus as the anomocytic one. The stomata are elongated-rounded shape, their number per mm² varies from 25 to 43 pcs. The stomata index is 7.42%. The upper epidermis has no pubescence.

Leaf abaxial epidermis (Fig. 1B, Table 1) The major epidermal cells are irregular in shape with irregular coarsely wavy anticlinal walls and coarsely tortuous outlines, light green, covered with thin cuticle and is interrupted by stomata. Intercellular spaces are absent. The wave’s frequency and amplitude are not constant. The sizes of epidermal cells vary over a very wide range from 39.79 to 80.34 microns in length and 21.46 to 45.55 microns in width. The epidermal cells are thick-walled, in various number (from 329 to 511 per 1 mm²), with their plan projection is stretched out or spread out. Angles are obtuse, pointed, rounded in adjacent boundaries.

The stomata are solitary, in chaotic locations. The stomata apparatus is an anomocytic type: the stomata are surrounded by 3–5, more often 4, cells. As at the upper epidermis, peri-stomata cells that guard cells adjacent, do not differ from the main epidermal cells along with shape and size. Peri-stomata cells’ certain orientation is not noted, however, some cases are the same and have characterized to upper epidermis too. The stomata are rounded, in number per 1 mm² varies from 67 to 183. Stomata index 17.87%. There are single-row, unicellular simple long hairs with a warty cuticle. The trichomes are formed by cuboid cells with an elongated apical cell. There is a base hair site in a rounded shape, which is surrounded by 12 polygonal epidermal cells, diverging in radial directions (Fig. 2).

PS: AW - anticlinal walls of epidermal cells; ELC - epidermal leaf cells; SA - stomatal apparatus; SI - stomatal index; Cv - variations coefficient.

The central vein on a cross-section has a rounded shape (Fig. 3A). Epidermis cells above the vein are parenchymal. Veins represent vascular bundles. They are founded regularly in the mesophyll. The largest and the oldest vein is founded in the center (midrib vein).

Each vein has a bundle sheath composed of a single layer of compactly arranged barrel-shaped parenchyma cells. The bundle sheath encloses both xylem and phloem. In the xylem, many protoxylem and metaxylem vessels are found. Protoxylem orients towards the upper epidermis. Hence, the vascular bundles are described as conjoint and collateral with the end- arch xylem. The bundle sheath of the midrib vein is connected to the upper and the lower epidermal layers by many layers of collenchyma cells, representing bundle sheath extensions or hypodermal collenchymas. A characteristic feature of the central leaf vein is the crystalline coating, which is formed by single calcium oxalate crystals (Fig. 3B). The vascular bundle is covered by a bundle sheath of parenchyma cells. Fibers are absent in both xylem and phloem. The xylem and phloem elements are conspicuous only in large vascular bundles.

The stem is rounded with slightly protruding ribs. Central axial cylinder is fascicular type: the fascicules are open and collateral, arranged in a circle. There is a sclerenchyma lining in the phloem side fascicules (Fig. 4).

Epidermal cells are elongated parenchymal with thickened rectangular walls, permeated by straight pores; without any pubescence (Fig. 5A). There is a two-layered lamellar collenchyma under the stem rib epidermis (Fig. 5B).

Antioxidant activity: Studies of the antioxidant activity of meadow clover grass were performed with plant excerpts (extracts) available for production in the pharmacy network and everyday life of ordinary consumers (ethanol and water). Previously, we tested the optimal terms, the temperature of obtaining the highest content of active substances of grass extracts of the studied plant, and we studied different ratios of plant raw materials: water (1: 5 and 1:10). Brief descriptive statistics of the experimental results are shown in Table 2.

The aqueous extract of clover grass (1: 5) had the lowest TBARS values among the three subjects under investigation (respectively: 0.541 ± 0.0291 μM / g; 0.638 ± 0.0218 μM / g and 0.570 ± 0.0341 μM / g, mean ± SD).

According to all studies, the degree of belief in the results was within normal limits: from p < 0.001 (t-test in aqueous extract 1: 5 and a solution of the reference substance of ascorbic acid) to p < 0.05 (t-test in aqueous extract 1:10).

The mean and median values in the tested extracts exceeded the modal values, which characterized the obtained data with a positive skew (skew was in the range from 0.5 to 1). The value of skew for ascorbic acid solution (exceeded 1) characterized the presence of large data skew.

The level of excess, calculated for the studied extracts, indicated the absence of miscarriages in the data set (kurt = -0.8193; 0.575; 0.211965, respectively). An excess of more than 3 in ascorbic acid indicated a miscarriage in the dataset (kurt = 4.468151).

In the case of the “zero” study, where the equivalent volume of water was analyzed instead of extracts, the antioxidant activity index had the highest level (0.734 ± 0.0241 μM / g, mean ± SD). But the degree of belief in the values obtained in the “zero” study does not allow to take into account these figures (t-test p = 1; u-test p > 0.05).

According to descriptive statistics, TBARS was most pronounced (reduced effect of active products) when using aqueous extract (1: 5). Here, the indicator% to mean contr was equal to -26.28% (p < 0.001 *, p < 0.01 **) under the influence of the reference drug -40.51 (p < 0.001 *, p < 0.01 **) relative to the control group.

Us the measure of correlation we were used in this study classical Pearson correlation, which represents the linear relation between two variables. Pearson correlation was computed using the stats R package (Table 3). Overall, there were moderately positive and negative pairwise correlations between scavenging assays, ranging from r = 0.76, P < 0.01 (water extract 1:5 vs. water extract 1:10) to r = - 0.70, P < 0.01 (water extract 1:5 vs. ethanol extract).

We have been built boxplots that provide a useful way to visualize the range and other characteristics of responses for data groups and to find about the optimal solvent and its ratio with plant material in an expression of maximal antioxidant action (Fig. 6). The boxplot shows that the difference between the medians of the two groups is approximately 1, we can conclude, with 95% confidence, that the true medians do differ. Our investigation had not any data points beyond the whiskers (no outliers), so the whiskers extend to the most extreme data points. The center of distribution water 1:5 is the lowest of the three distributions (median is 0.5113). Q1 was for it far minimum than all extracting data. Moreover, the line marking the median crosses the boxes toward Q1 indicating a distribution with asymmetrical shape (positively skewed) in this distribution. Distribution water 1:10 is the most concentrated, with mine interquartile range of the three distributions while with the highest center of distribution of the three distributions (median is 0.63875). The centers of distribution in the boxes varied, showing asymmetric shifts, where the smallest with a value of 0.55 was observed in the aqueous extract (1:5) of clover. The data of ethanol extract have the longest asymmetric fences.

Thus, based on the positive correlation coefficient and Fig. 6, we concluded the possibility of further recommendations for in-depth pharmacological studies of the phytocomposition Trifolium pratense herb (water_1: 5) as an antioxidant.

Describe T. pratense extract inhibited the growth and the proliferation of cancerous cells but its use in clinical setting is limited due to the therapeutic and toxicity doses. Today, complementary compounds can be used to reduce the side effects and enhance the anticancer potential of chemotherapeutic drugs. Тherefore, those signs in the anatomical structure of the plant are important for the correct identification and belonging of the plant.

Nevertheless, with such a wide phytochemical and pharmacological interest in studying a large number of clover species, we found few articles about anatomical microscopic identification of Trifolii pratense herbs as medicine (Booth et al. 2006; Zoric et al. 2012) and about coumarins, especially simple coumarins, with anticoagulant action because plants can be regarded as “living factories” producing a variety of chemical compounds, including primary important metabolites for the plants’ growth and secondary metabolites. All these components may work together to deliver a synergistic effect in the finished product. Finally, the author team want to call attention to the advantages of the application of boxplots in data analysis of pharmacognostical interest and, most important, can be very useful to uncover information hidden.