The leaves and stems of D.ramossisimum were collected in March 2018 in the town of Porto-Novo. At the laboratory, the leaves and stems were washed with distilled water, then dried at room temperature 25˚C ± 3˚C for 16 days. They were then ground into a uniform powder, using a Retsch mill type SM 2000/1430/Upm/Smfet. The obtained powder was used for the preparation of different extracts.

2.2.1. Methanol Extract and Water-Ethanol Extract

The extraction method described by Dah-Nouvlessounon etal. [6] was used. Briefly, 50 g of powder were macerated in 500 ml of solvent: methanol and water-ethanol (30:70, v/v) under continuous stirring for 48 hours. The mixture was then filtered thrice through cotton wool and once with the filter paper Whatman N˚1. The extracts were concentrated to dryness using a rotary evaporator (RE 300). The resulting residues were dried at 40˚C in oven. The resulting powders constitute the methanol dry extract and water-ethanol dry extract ready for use. Dry extracts were stored at 4˚C until analysis.

2.2.2. Successive Extraction with Hexane, Ethanol, and Ethyl Acetate

The extraction technique is an adaptation of the protocol used by N’Guessan etal. [7]. Briefly, 100 g of powder was macerated in 1 L of ethanol 96˚ under continuous stirring for 72 hours. The mixture was filtered thrice. Then, 50 ml of distilled water and 50 ml of hexane were added to it. The mixture obtained was then poured in a separatory funnel. After decantation, two phases are observed: a lower aqueous phase and an upper organic phase. The organic layer was collected and then dried to obtain the hexane dry extract. The first half of the lower layer was dried at 40˚C to obtain the ethanol dry extract. For the second half of aqueous phase, 200 ml of distilled water and 100 ml of ethyl acetate were added to it. The resulting mixture was left decanted to obtain two phases. The upper layer was recovered and dried to obtain the dry extract of ethyl acetate. The lower phase was dried to obtain the residual ethanol dry extract. The extracts obtained were stored in sterile bottles for further use. The extraction yield was determined by the ratio between the mass of dry extract obtained after evaporation and the mass of plant material.

The phenolic compounds were quantified using high-performance liquid chromatography (HPLC) in reverse mode equipped with 600 E pump associated with Waters 486 UV tunable absorbance detector and 20 µl injection and a reversed-phase column (Inertsil ODS RP C18 Column × 150 mm; size, 5 µm) with a flow rate of 1 ml/min. The compound content was determined using a five-point calibration curve of each standard (R² = 0.99) in the linearity range of 10 - 100 μg/ml. The analyses were carried out at a wavelength of 320 nm. The identification and quantification of the phenolic compounds contained in the extracts was done by comparing the retention time of the surfaces of the standards used with those obtained by analyzing the extracts. The gradient used is constituted of three mobile phases:

- Solvent A: 50 mM of phosphate ammonium at 2.6 pH adjusted (NH₄H₂PO₄) with orthophosphoric acid (H₃PO₄).

-Solvent B: composition 80:20 (v/v) of acetonitrile/solvent A.

-Solvent C: composition 200 mM orthophosphoric acid (H₃PO₄) at pH 1.5 adjusted with NaOH).

The solvents were filtered through filter paper, Whatman N˚1 and disposed in ultrasound apparatus for 25 min. The profile of the gradient used is resumed in Table 1.

Table 1. Profile of the gradient for HPLC analyses.

2.4.1. DPPH Radical Scavenging Test

The antioxidant activity of D.ramosissimum extracts, on the basis of the scavenging activity of the stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radical, was determined by the method described by Lamien-Médard etal. [8]. A methanol solution of 100 μg/ml of DPPH was mixed with extracts of Desmodiumramosissimum prepared from a stock solution of 400 μg/ml (400; 200; 100; 50; 25; 12.5; 6.25; 3.125; 1.56; 0.78 μg/ml). Briefly, 1 ml of the DPPH methanol solution was added to 1 ml of each extract, the mixture was vigorously stirred, and then the tubes were incubated at room temperature and in the dark for 30 minutes. Absorbance was read at 517 nm against a blank consisting of DPPH and methanol. Ascorbic acid, quercetin, BHA and BHT were used as positive control. Each test was carried out in triplicates. The antiradical activity is estimated according to the equation below:

The IC₅₀ (Concentration providing 50% inhibition) was calculated graphically using a calibration curve in the linear range by plotting the extract concentration and the corresponding scavenging effect. Antioxidant activity index (AAI) was calculated according to the formula used by Scherer and Godoy [9].

2.4.2. Ferric Reducing/Antioxidant Power (FRAP)

FRAP essay was conducted according to the protocol described by Dieng etal. [10]. Thus, 0.5 ml of sample at different concentrations was mixed with 1 ml of phosphate buffer (0.2 M; pH = 6.6) and 1 ml of 1% potassium hexacyanoferrate [K₃Fe(CN)₆]. After incubating the mixture at 50˚C for 30 minutes, 1 ml of 10% trichloroacetic acid was added to stop the reaction, and then the tubes were centrifuged at 3000 rpm for 10 minutes. Then, 1 ml of the supernatant from each tube was mixed with 0.2 ml of a 0.1% FeCl₃ solution and left to stand in the dark for 30 minutes before measuring optical densities at 700 nm. The antioxidant activity related to the reducing power of the extracts is expressed in reducing power (RP) using the following formula:

IC₅₀ (Concentration that inhibits the 50% of DPPH or reduces the 50% of Fe³⁺) was determined by the probit method.

The invitro anti-inflammatory activity has been evaluated using the method of inhibition of albumin denaturation described by Ghumre [11]. Briefly, 5 ml of the reaction mixture consisted of 0.2 ml of egg albumin, 2.8 ml of phosphate-buffered saline (PBS, pH 6.4) and 2 ml of variable concentration of extracts (1000; 500; 250; 125; 62.5; 31.5 µg/ml). A similar volume of distilled water was used as a control. The mixture was then incubated at 37˚C for about 15 minutes and then heated to 70˚C for 5 minutes. After cooling, absorbances were measured at 660 nm. Diclofenac sodium inj.Pb./Drugs/1804-B was used as standard at a concentration of 6.25 to 25 mg/ml under the same conditions as the extracts. The inhibition’s percentage of protein denaturation was calculated by the formula mentioned below:

The cytotoxic effect of the extracts was evaluated based on the method described by Kawsar etal. [12]. Larvae were obtained by hatching 10 mg of freeze-dried eggs of Artemiasalina (ARTEMIO JBL GmbH D-67141 Neuhofem) under continuous agitation in 1 L of seawater for 72 h. From a concentration of 20 mg/ml, a range of 10 concentrations is achieved with sterile distilled water following dilutions of 1/2 in test tubes numbered T1 to T10. Then, 1 ml of sea water containing 16 alive larvae was added to all tubes. The number of surviving larvae was counted after 24 h incubation. For each extract, the lethal concentration 50 (LC₅₀) was determined from the regression line obtained from the curve representing the number of surviving larvae as a function of extract concentration (the curves are appended to this manuscript). The results were interpreted using the correlation grid associating the degree of toxicity to LC₅₀ proposed by Moshi etal. [13].

Table 3 presents the different phenolic compounds identified in the hydro-ethanol and methanol extracts of D.ramossissimum. The analysis of Table 3 shows that 63% of the compounds were identified in the hydro-ethanol extract, while 50% were found in the methanol extract. Among the quantified compounds, the highest levels of phenolic compounds were obtained with glucose quercetin (0.45 ± 0.03 mg/ml) and quercetin dihydrate (2.12 ± 0.16 mg/ml) for the hydro-ethanol and methanol extracts, respectively. Furthermore, among the quantified compounds, quercetin had the lowest level (0.05 ± 0.04 mg/ml) with the hydro-ethanol extract, while cinnamic acid had the lowest level (0.78 ± 0.07 mg/ml) with the methanol extract (Figures 1-3).

Table 3. Concentration of phenolic compounds contained in hydro-ethanol and methanol extracts.

H-Eth: hydro-ethanol; MEth: methanol extract; Ref.: reference; Rt: retention time; nd: not detected.

(a)

(b)

(c)

1: gallic acid; 2: protocatechuic acid; 3: catechin; 4: chlorogenic acid; 5: rutin; 6: glucose quercetin; 7: quercetin dihydrate; 8: cinnamic acid.

Figure 1. Chromatographic profile of the standard mixture (a), hydro-ethanol (b) and methanol (c) extract.

Table 4 presents the different phenolic compounds identified in the ethanol and residual ethanol extracts of D.ramossissimum. The analysis of Table 4 reveals that 87.5% of the compounds were identified in ethanol extract against 75% in residual ethanol extract. The minority compounds extracted with the ethanol were catechin (40.10 ± 0.04 mg/ml) and gallic acid (50.11 ± 0.01 mg/ml), chlorogenic acid was not determined. However, in the residual ethanol extract, quercetin dihydrate (50.20 ± 0.02 mg/ml) and glucose quercetin (39.39 ± 0.03 mg/ml) are predominantly represented. Rutin and cinnamic acid were not determined in the residual ethanol extract.

Table 4. Concentration of phenolic compounds contained in residual ethanol and ethanol extracts.

Eth: ethanol extract; Ethr: residual ethanol extract; Ref.: reference; Rt: retention time; nd: not detected.

(a) (b)

1: gallic acid; 2: protocatechuic acid; 3: catechin; 4: chlorogenic acid; 5: rutin; 6: glucose quercetin; 7: quercetin dihydrate; 8: cinnamic acid.

Figure 2. Chromatographic profile of the residual ethanol (a) and ethanol extract (b).

Table 5 presents the different phenolic compounds identified in the hexane and ethyl acetate extracts of D.ramossissimum. Table 5 reveals that the hexane extract and the ethyl acetate extract have the same proportion of compound (50%). Among the quantified compounds, chlorogenic acid is predominantly represented (17.10 ± 0.01 mg/ml) in the hexane extract, followed by hydrated quercetin (13.07 ± 0.02 mg/ml) in the ethyl acetate extract. The lowest level of compound recorded in the quantified compounds was obtained with protocatechuic acid (10.05 ± 0.02 mg/ml) for the hexane extract, while catechin showed the lowest level (8.08 ± 0.04 mg/ml) with the ethyl acetate extract.

Table 5. Concentration of phenolic compounds contained in the hexane and ethyl acetate extracts.

Hex: hexane extract; Eth-ace: ethyl acetate extract, Ref.: reference, Rt: retention time, nd: not detected.

(a) (b)

1: gallic acid; 2: protocatechuic acid; 3: catechin; 4: chlorogenic acid; 6: glucose quercetin; 7: quercetin dihydrate.

Figure 3. Chromatographic profile of the hexane (a) and ethyl acetate (b) extract.

The results of the antiradical activity of our extracts were expressed in millimole equivalent of ascorbic acid per gram of extract (Table 6) in reference to a standard curve (y = 0.0522x + 0.056, R² = 0.9934) and in inhibition’s percentage of the free radical DPPH-. Therefore, the concentrations that inhibit 50% of the free radicals (IC₅₀) as well as the antioxidant activity index (AAI) are respectively consigned to Table 6.

The ethanolic extract presented the highest concentration (37 ± 0.66 mmol EAA/g of sample) in ascorbic acid equivalent, with a low IC₅₀ (2.5 ± 0.7 µg/ml), thus showing its strongest antioxidant power. Furthermore, although the hexanic extract presented the lowest concentration (7 ± 0.03 mmol EAA/g of sample), it is the hydro-ethanolic extract that recorded the highest IC₅₀ (37.5 ± 0.06 µg/ml), thus showing its weak antioxidant potential.

Table 6. Equivalent content of ascorbic acid and IC₅₀ of D.ramosissimum extracts.

3.2.1. DPPH Radical Scavenging Test

The inhibition percentages vary depending on the extracts. Indeed, the highest percentage was obtained with the hydroethanolic extract, followed by the residual ethanolic extract, the methanolic extract and finally the ethanolic extract (Figure 4).

Figure 4. Variation in the percentage of inhibition in function of extract concentration.

3.2.2. Ferric Reducing/Antioxidant Power (FRAP)

The ferric ion reduction power was shown in Table 7. The results show that four out of six extracts tested show antioxidant activity using the FRAP method. The IC₅₀ ranges from 1.1 mg/ml (ethanolic extract) to 7 mg/ml (methanolic extract), which shows that the ethanolic extract has a higher ferric ion reduction power than the other extracts. The analysis of variance shows that the interaction between the inhibitory power of extracts varies according to their concentrations. Indeed, the percentage of inhibition of extracts varies significantly (p = 0.0001).

Table 7. Percentage of inhibition of D.ramosissimum extracts and IC₅₀.

The invitro anti-inflammatory effect of D.ramosissimum extracts is summarised in Table 8. Indeed, the highest inhibition percentages obtained were for hydro-ethanolic 92.99% ± 0.04%, ethanolic 92.55% ± 0% and residual ethanolic 92.85% ± 0.04% extracts. In addition, the analysis of variance showed a significant difference (p < 0.05) between the concentrations at 15.75 µg/ml. However, the hydro-ethanol extract was more active than the ethanol extract. This was confirmed by comparing their IC₅₀ values. Indeed, the hydro-ethanol extract shows the strongest inhibition of protein denaturation (egg albumin). Diclofenac sodium (in the concentration range of 6.25 to 25 mg/ml) was used as the reference drug, which also showed a concentration-dependent inhibition of protein denaturation. The results showed that the high activity (94.55% ± 0.12%) of Diclofenac is obtained at 25 mg/ml.

Table 8. Inhibition percentage of Diclofenac and Desmodiumramosissimum extracts.

From this table, it can be seen that the greatest inhibition percentages (92.99% ± 0.04%) obtained are the one of hydro-ethanol extract, ethanol extract (92.55% ± 0%) and residual ethanol extract (92.85% ± 0.04%).

The cytotoxicity of D.ramosissimum extracts has been evaluated on Artemiasalina. After 24-hour of exposure to the tested extracts, larval mortality was investigated using a regression curve obtained from the number of surviving larvae in each tube according to the extract’s concentrations. Figure 5 shows the number of surviving larvae with each extract after 24 hours of incubation.

Figure 5. Variation in larval mortality of Artemiasalina according to D.ramosissimum extracts.

The yields of extracts obtained from D.ramosissimum are in decreasing order: 16% for the ethanol extract, 14.78% for the methanol extract, 13.4% for the hydro-ethanol extract, 5.2% for the acetic extract followed by 4.22% for the residual ethanol extract and 2.24% for the hexane extract. These results reveal that they would contain bioactive molecules that were extracted at various degrees depending on the type of solvent. These results also show that ethanol is the most suitable solvent for the extraction. Alcoholic solvents are capable of increasing the permeability of cell walls by facilitating the extraction of a greater number of polar’s molecules of medium and low polarity [14]. Moreover, the maceration process, under stirring for a spread-out time (72 hours) and at room temperature, allows, respectively, the exhaustion of the solvent in active compounds and the prevention of their alteration or modification probably due to the high temperature. According to Wu etal. [15], the extraction yield could be explained by several factors: the chemical structure of the compounds, the maceration time, the temperature, the nature of the solvent and the storage conditions. For example, Dah-Nouvlessounon etal. [6] showed that the variability in extraction yields could be explained by the extraction capacity of each solvent, which depends on the affinity of the solvent for the phytomolecules on one hand and the polarity of the solvent on the other hand. The drying of the plant was carried out in an open air in a shady environment in order to preserve the active ingredients from degradation by the hydrolases present in the fresh plant material. Drying the plant in the dark prevents chemical transformations such as isomerisation and degradation caused by ultraviolet radiation from sunlight [16]. Grinding and sieving have been carried out to recover as much fine powder as possible with a diameter smaller than 63 μm, because according to the work of Chen etal. [17], this diameter increases the contact surface of the solvent with the powder for better yield. Also, we used several solvents during extraction to ensure separation between polar and non-polar compounds [18]. In addition, extraction with several solvents allows better separation of compounds with antioxidant activity [19]. These extraction yields that we obtained are not consistent with the results of Maddi etal. [20], who found 26.37% for the ethanol extract and 33.32% for the acetate extract. These discrepancies could be due to several reasons, including the stage of maturity of the leaves collected [21], the extraction process, the geographical origin of the plant drug [22], the drying conditions of the leaves [23], the harvesting season and the quality of the solvents used. However, our results corroborate with those of Onzo etal. [24] who reported that the hexane extract has the lowest extraction yield, as in the present study.

The antioxidant activity of the extracts is evaluated by two techniques, iron reduction (Ferric reducing antioxidant power or FRAP) and the scavenging of the free radical 2,2-diphenyl-1-picrylhydrazyl (DPPH). The results of the DPPH test show that D.ramosissimum extracts have interesting antiradical properties that vary from one type of extract to another. The results showed that the equivalent ascorbic acid content of the ethanol (37 ± 0.66 mmol EAA/g extract) and methanol (26 ± 0.3 mmol EAA/g extract) extracts is quite high. This result could be explained by the fact that these extracts have a high reducing power and therefore contain molecules with a stronger electron-donating reducing potential than the other extracts. As a result, the reducing electron-donating power is dependent on the bioactive molecules. These results corroborate with those of Nounagnon etal. [25] who found the best ascorbic acid equivalent contents with the ethanol and methanol extracts of Combretumadenogonium. The results of the DPPH test show that at a concentration of 200 µg/ml, the ethanol, methanol and residual ethanol extracts of D.ramosissimum have respectively a scavenging power of the DPPH free radical at 84.79%; 78.95% and 80.05%, while at the same concentration, ascorbic acid, quercetin, BHA and BHT respectively inhibit 91.73%;74.8%; 65.95% and 72.54% of this free radical. These results show that D.ramosissimum extracts have low IC₅₀ values, which means that they would be excellent antiradical agents. Comparing these values to those of standard antioxidants shows that these extracts are more active than BHT and BHA. This antioxidant property of D.ramosissimum would be linked to the presence of polyphenols such as flavonoids, tannins, and coumarins. These compounds are identified by Hooper etal. [26] in the leaves and roots of D.ramosissimum. The results of the qualitative and quantitative analyses carried out by HPLC have revealed that our extracts are rich in phenolic compounds, including gallic acid, protocatechic acid, catechin, chlorogenic acid, rutin, glucosed quercetin, quercetin dihydrate, and cinnamic acid. These results would explain the strong antioxidant activity observed in the plant.

The high levels of quercetindihydrate and chlorogenicacid identified by HPLC provide a clear biochemical basis for the strong antioxidant and anti-inflammatory activities of D.ramosissimum. Quercetin is a potent flavonoid known for its free-radical-scavenging capacity and its ability to inhibit key inflammatory pathways such as NF-κB and COX-2 (Boots etal., 2008; Li etal., 2016) [27][28]. Likewise, chlorogenicacid exhibits strong antioxidant effects through ROS neutralization and reduces inflammation by suppressing oxidative-stress-mediated signaling (Naveed etal., 2018) [29]. The presence of these well-characterized compounds therefore supports the biological activities observed and strengthens the therapeutic relevance of this species.

Furthermore, considering the results of the antioxidant activity indexes (AAI) according to Scherer and Goddoy [9], we can conclude that apart from the hydro-ethanol extract, all the other extracts of D.ramosissimum have a very high DPPH radical scavenging power (AAI > 2). The action of these antioxidants is assumed to be due to their ability to donate hydrogen atoms or electrons derived mainly from the A-ring hydroxyl of flavonoids [30]. In addition, Albano and Miguel [31] reported that flavonoids and phenolic acids are more effective in scavenging free radicals after extraction by moderate or hydrophilic solvents. The effectiveness of the antioxidant (AH) increases if the strength of the A-H bond is low and the resulting (A) radical must be as stable as possible, which is the case for phenolic and flavonoid compounds, which are better electron or hydrogen donors [32]. Our results regarding inhibition percentages are comparable to those of Kabran etal. [33], who obtained 95% with D.adscendens extracts. D.ramosissimum extracts can protect the organism against the oxidative stress. Stress is a major contributor to the appearance of many pathologies such as cancers, cardiovascular diseases and degenerative pathologies of the nervous system [34]. Faced with this situation, scientific work has shown that natural antioxidants are more protective than synthetic antioxidants [35].

With regard to the determination of antioxidant activity by the FRAP method, the increase in iron reduction is proportional to the concentrations used. Except for the ethanol extract, all our extracts have higher IC₅₀ than BHT. FRAP is based on the capacity of the extracts to reduce ferric iron Fe³⁺ to ferrous iron Fe²⁺. The high activity of D.ramosissimum extracts might be due to their high content of polyphenolic compounds. The reducing power of these extracts would be due to the hydroxyl group in the phenolic compounds that can serve as electron donors. Therefore, antioxidants can be considered as reducing and inactivating oxidants [36]. Some previous studies have also shown that the reducing power of a compound can serve as a significant indicator of its potential antioxidant activity [37]. Alcohol-based extracts are the most active, this is surely due to the composition of these extracts in polyphenolic substances and the synergy between them for a better antioxidant activity [38]. Several studies have shown that the hydroxyl groups in phenolic and flavonoid compounds are responsible for their antioxidant power [39]. These results are consistent with the results obtained in terms of HPLC, which revealed that our extracts are rich in phenolic compounds.

The extracts showed concentration-dependent inhibition of protein denaturation (egg albumin) over the entire concentration range from 15.75 to 1000 µg/ml. These results showed that the extracts have a good inhibition of protein denaturation compared to the reference molecule (Diclofenac sodium).

Referring to the toxicity scale established by Moshi etal. [13], the LD₅₀ of extracts are above 0.1 mg/ml. At this value, the extract is considered to be non-toxic. Indeed, the extracts tested on Artemiasalina were found to be non-toxic at the tested doses. However, the mortality of brine shrimp (A.salina) would increase by increasing the concentration of our extracts. Mayer etal. [40] came to the same conclusion in their study of several species of the genus Euphorbia with the Artemiasalina larvae. The same results were found by Bapna etal. (2014) [41] with different extracts of Citrusaurantifolia.

This toxicity test showed that at a dose of 20 mg/ml, the extracts showed no toxicity on the cells of shrimp larvae. By definition, a substance is toxic if it causes alterations or disturbances in the functions of the living organism, leading to harmful effects. The most serious harm is the death of this organism [42] (Ramade, 1979). Also, according to Cowan etal. [43], the mechanism of toxicity could be linked to the inhibition of hydrolytic enzymes (proteases and carbohydrolase) or other interactions to inactivate microbial adhesins, transport and cell envelope proteins. The Artemiasalina model has been suggested for the evaluation of cytotoxicity in pharmacology because cytotoxicity in shrimp larvae is correlated with 9 PS and 9 KB cells [44]; A-549 cells in lung carcinoma and HT-29 cells in colon carcinoma [45]. It is a test that is much more widely used in underdeveloped countries where traditional medicine plays a key role. The evaluation of this activity with D.ramosissimum can therefore serve as an indication in the choice of therapeutic doses.