Leaves of Ocimum species were collected in the vicinity of Mbuji-Mayi (O. canum) and of the University of Kinshasa (O. basilicum and O. gratissimum) in DRC on May 2013. The collected plants were identified by comparison with reference specimen available at Biology department of University of Kinshasa, DRC. Voucher specimens (Ocimum basilicum: 425/Devred, Ocimum canum: 686/A. Thieband, Ocimum gratissimum: 8016/R. Dechamps) were deposited at the herbarium of the same department. Leaves of plants were dried at the room temperature and finely grounded in a high speed mill (Retsch ZM 100 Model) to 0.02 inches size. The powder leaves were stored in the dark at the room temperature and used for solvent extraction.

The blood samples used for the bioassays in this study were taken from adolescent sickle cell disease patients attending the “Centre de Medicine Mixteet d’Anemie SS” (CMMASS), located in Kinshasa, DRC. A written consent for each patient was approved by the national ethic committee (N˚ BE117). Ethical clearance on the use of SS blood was strictly observed according to international rules [32] .

In order to confirm their SS nature, the above mentioned blood samples were first characterized by hemoglobin electrophoresis on cellulose acetate gel at pH 8.5 and then stored at 4˚C in a refrigerator.

Emmel’s tests were performed as previously reported [3] - [5] [7] [10] [13] . The red blood cells digitized micrographs were treated with a computer assisted image analysis system (Motic Images 2000, version 1.3), Statistical analysis and curves were processed using Microcal Origin 8.6 package Software. All anti-sickling experiments were carried out in triplicate using a sodium citrate suspension of freshly collected blood as previously reported [3] [5] [13] [14] [16] .

All solvents were of analytical grade and purchased from Merck VWR (Leuven, Belgium).

Rosmarinic acid, luteolin, vitexin were HPLC grade from Sigma Aldrich. Water was treated in a Milli-Q water ultra-purification system (Easy Pure Purification System).

The samples were extracted from 1 g plant powder sonicated for 15 min in 10 mL methanol and filtered on a what man filter paper. One milligram of RA of HPLC analytical grade was dissolved in 10 mL methanol and gently shaken.

Analytical thin layer chromatography (TLC) of 10 µL of solution was carried out on Silica gel 60 F₂₅₄ plates (Merck), using ethyl acetate/formic acid/glacial acetic acid/water (100:11:11:26) and methylene Dichloride/ acetone/formic acid (85:25:8.5) as eluents. Luteolin, vitexin, rosmarinic acid were used as standards. The plate was sprayed with Natural Products-PEG reagent and observed at UV-365 nm.

The High performance liquid chromatography (HPLC) analysis was carried out at 25˚C by an “Agilent 1100” HPLC chain connected to a diode array detector (DAD). All samples and standards were filtered through a 0.45 μm pore size syringe-driven filter before 20 μL of each one were injected into the HPLC-UV/DAD system. The separation was carried out using an Hypersil ODS column (4 mm × 250 mm) with a nonlinear gradient of acetonitrile (solvent A) and 0.05% trifluoroacetic acid in ultra-pure water (solvent B) in the following composition: T0: 0% A, 100% B; T1: 3% A, 97% B; T45: 40% A, 60% B; T46: 0% A, 100% B and T60 stop. The time (T) is expressed in minutes. The compounds were eluted at a flow rate of 1 mL/min and detected with UV-DAD. The UV spectra of elution peaks were recorded in the range from 250 to 340 nm and the chromatograms were monitored at 280 and 340 nm. The identification was based on the retention time and the absorption spectra in comparison to the references and the data available in the database of the system [33] .

Figure 1(a) and Figure 1(b) show the TLC fingerprint of methanolic extracts of the three species of Ocimum firstlyalone and then with phenols controls, respectively.

As above noticed, TLC analysis show that the different extracts of Ocimum contain polyphenols, including: Chromatogram (a): flavonoids of Kampférolkind (yellow-orange fluorescent spots) and phenolic acids (blue

fluorescent spots); chromatogram (b): phenols acids RA.

The presence of RA was confirmed by HPLC fingerprint of control pure RA in superimposing this last to the fingerprint of each methanolic extract of the three Ocimum.

Figure 2 shows the structure and the HPLC fingerprint of pure RA used as control by reason of comparison to the three HPLC fingerprints of the three methanolic extracts of three Ocimum.

As it can be clearly seen, the pure RA (a) appears as a single peak to the retention time equal to 32.408 minutes. As above noticed the profile of methanolic extract of O. canum (b) shows two intense peaks. The most intense and consequently major compound appears to the retention time equal to 58.719 minutes, followed by the one at 32.494 minutes. This last has approximately equal RA retention time. These results are near to what found by Berhow et al. [35] who has found a retention time of 30.07 min.

The presence of RA in the methanolic extract of O. canum is confirmed by the overlay of both the pure RA peak and the characterized fingerprint of methanolic extract, at the same retention time. The available data in the database used as references of the system did not allow us to identify the compound to the retention time of 58.719 minutes.

It can also be seen in the fingerprint of O. basilicum (b) the most intense peak to the retention time of 32.300 min. Approximately this last value is equal to the one above of RA. These results are alike to that of Vlase et al. [36] in methanolic extract of O. basilicum from Romania.

The presence of RA is demonstrated by the superimposition of the peak of the HPLC fingerprint of pure rosmarinic acid with the corresponding peak of the HPLC fingerprint of the methanolic extract of O. basilicum, to the identical retention time. By comparing the two intensities identified above peaks between O. canum and O. basilicum, the peak of this last (b) appears to be more intense. Thus, that result confirms that O. basilicum has more important amount of RA than O. canum.

Ultimately, it is noticed in Figure 2 in the fingerprint of O. gratissimum (d) the most intense peak to the retention time of 32.310 min. This seems to be approximately to RA founded value. This result is comparable to that found by Costa et al. [37] in methanolic extract of O. gratissimum from Brazil.

As for the two first methanolic extracts, the presence of RA is shown by the overlay of the two peaks of pure RA and the one of the methanolic extract of O. gratissimum to the almost equal time of retention. Once more

again comparing the three fingerprints of the methanolic extracts of the three species according to their intensity, O. basilicum contains the highest amount of RA, followed by O. gratissimum and then O. canum.

Figures 3(a)-(f) show respectively micrographies of SS blood alone in a NaCl 0.9% solution (negative control), betulinic acid (positive control) and the sickle cells blood incubated with the methanolic extract of O. canum, O. gratissimum and O. basilicum, and RA finally.

As it can be seen from these micrographies, the negative control (Figure 3(a)) contains the majority of sickle- shaped erythrocytes, confirming the Sickle cell nature of the blood. When betulinic acid, well known for its antisickling activity [17] [18] [38] , is added as positive control (Figure 3(b)), Red blood cells show their normal circular (biconcave) shape. In the presence methanolic extracts of the three species of Ocimum (Figures 3(c)-(e))

and RA (Figure 3(f)), the majority of erythrocytes are reversed normal-shape. This fact indicates that the methanolic extracts have an antisickling activity and show also a good normalization effect of drepanocytes in hypoxic condition. The same behavior was already observed form active extracts from some plants used in the management of sickle disease in Congolese traditional medicine, justifying their traditional use. Some molecules contained in these extracts would interact with HbS and compete with its polymerization reaction. This would prevent the sickling of red blood cells. In fact, according to Russu et al. [39] when investigating on molecular basis for the antisickling activity of aromatic amino acids and related compounds by NMR, phenolic acids would interact with HbS on a binding site located at or near the beta 6 position (the site containing the mutation in beta 6Glu to Val). This binding would induce conformational changes in the amino-terminal domains of the beta chains.

To compare the antisickling activity of different drugs, it is necessary to determine the minimal concentration of normalization (MCN) or the concentration that normalizes 50% of drepanocytes (ED50) [5] [9] [10] [15] [16] .

Figure 4 gives the dose dependent antisickling activity of methanolic extracts of three Ocimum leaves and RA.

Figure 4 shows that the sickle cells normalization rate increases with the methanolic extracts doses until reaching the maximum threshold of which the normalization rate remains constant despite the increase of the extracts concentration. The determined minimal concentrations of normalization of each extract and RA i.e. the weakest concentration of extracts or RA for which the normalization rate is maximum are listed in the table.

As it can be noticed from Table 1, RA shows higher antisickling effect than all methanolic extracts. This indicates that the main secondary metabolites responsible of the antisikling effect in the three methanolic extracts have to be found as RA. This is confirmed by the fact that O. basilicum has highest amount of RA and exhibited

the highest antisickling activity, follow by O. gratissimum and then O. canum. In fact, most isolated antisickling active molecules are phenolic acids [18] [38] - [41] . These results indicate that rosmarinic acid is among these antisickling active organic acids. The kwon antioxidant activity of RA [29] [35] [37] may also contribute to its antisickling activity, because sickle disease is in relation with the oxidative stress [42] [43] .