2.1. Isolation and identification:
The total Gls of both plants were isolated and further purified to afford two compounds (CA1 and CA2) from CA and five compounds (FA1- FA5)from FA which were identified as follow:
CA1- 4-methylthio-3-butenyl Gls: The compound was isolated as an off-white amorphous powder and appear as a brown spot (Rb = 0.52) after spraying with modified AgNO3 reagent. The UV absorption spectrum of the compound in methanol showed λmax .at 208 nm which shifted to 217 nm by addition of NaOH. When sliver nitrate solution (1% ) was added a new maximum appeared at 260 nm in addition to the first peak at 215 nm with increase in intensity, indicates the glucosinolate nature of the compound [18].
The 1H-NMR spectrum (DMSO) showed signals at δin ppm 2.13(3H, s, CH3-S), 2.35(2H, t, H-1), 2.7(2H, q, H-2), 4.75(1H, d, J = 8.2 Hz with a small t, H-3) ,5.8(1H, d, J = 8.2 Hz, H-4), the anomeric proton of glucose displayed as a doublet at 5.1 ppm. The complex group of signals between 3.1 and 3.85are characteristic for the rest proton of the glucose moiety. The 13C-NMR spectrum gave the anomeric carbon of the glucose unit at 82.16, the two carbon atoms(C-3 and C-4) of the double bond appeared at 137 and 115.6 ppm respectively and the central carbon of the Gls was assigned at 156.21 ppm. The other data were summarized in table-1. These findings were coincided with that reported by Manuela et al in 1992 [11]. The GC/MS of the isolated aglucone of this compound after the enzymatic hydrolysis gave only one peak for the corresponding isothiocyanate which identified through it's fragmentation pattern as follow: the mass spectrum displayed a mass to charge ratio (m/z) of 159 due to the molecular ion peak M+ which correspond to the molecular formula C5H7S2N, in addition to the fragment ion peaks at m/z 144(M+ - CH3), 101(M+ - NCS) and 73(M+ - CH2NCS). By comparing these data with the reported data, it was found that, the fragmentation pattern is the same that reported for 4-methylthio-3-butenyl ITC, so, the chromatographic and spectroscopic measurements could be identified compound CA1 as 4-methylthio-3-butenyl glucosinolate.
Table 1:13C-nmr data of compound CA1:
Carbon no.
|
δin ppm
|
C1
|
31.75
|
C2
|
30.44
|
C3
|
137.01
|
C4
|
115.6
|
S-C5
|
24.35
|
C=N
|
156.21
|
C1'
|
82.16
|
C2'
|
72.94
|
C3'
|
78.15
|
C4'
|
69.79
|
C5'
|
81.15
|
C6'
|
60.93
|
CA2- 6-methylsulfonylhexyl Gls: this compound was isolated in a small amount about 3 mg as a white yellowish powder which is not enough for nmr measurements. so, it subjected to enzymatic hydrolysis and identification the resultant ITC as follow: the mass spectrum exhibited M+ at m/z of 221(20.5%) which fit to the molecular formula C8H15O2NS2. The other fragments at m/z = 220(M+-1, 23%), 205(M+- CH3, 100%), 163(M+- NCS, 40%) and 142 (M+- CH3SO2, 35%).This fragmentation pattern confirm the identification of the aglucone as 6-methylsulfonylhexyl ITC [CH3SO2-CH2-(CH2)4-CH2-NCS] and accordingly the compound was identified as: 6-methylsulfonylhexyl Gls. This is the first report about the isolation of these compounds from C. annua growing in Egypt.
The GC/MS data of the Gls. Hydrolysis products from FA was matched with the previously published data [18, 19], it was found that, the Gls, hydrolysis products included only isothiocyanates which were identified in an increasing of their retention manner as follow:
FA1 - 6-methylsulfonyl6-hydroxy-hexyl ITC: Rt. 15.3 min., the mass spectrum displayed a molecular ion peak M+ at m/z (% relative abundance),=237(5%) correspond to the molecular formula C8H15O3NS2. Other fragments at 236(M+ -1, 12.3%), 221(M+ - CH4, 28.6%), 204(M+- SH, 17.5%), 180(M+- NCS, 74%), 165(M+- CH2NCS, 100%), 137(M+- (CH2)3NCS, 32%), 123(M+- (CH2)4NCS, 12.5%), 109(M+- (CH2)5NCS, 18%) and 91(M+- [H2O +(CH2)5NCS], 11.8%) or (M+-[CH3SO2-CH-1]). The latter two fragments (m/z 91 and 109) showed the presence of the OH group at C-6 and the fragmentation pattern proved the identification of the aglucone as 6-methylsulfonyl-6-hydroxy hexyl ITC [CH3SO2-CH(OH)-(CH2)4-CH2- NCS] (Chart 2).
FA2- 4-pentenyl ITC: Rt. 14.4 min., the molecular ion peak M+ was present at m/z = 127(14.3%) which correspond to the molecular formula C6H9NS. Another important fragments at m/z 100(M+- (CH2 = CH+, 18%), 85(M+-(CH2)3, 76%) and 57(M+- (CH2 = CH-(CH2)3, 100%). These data substantiate the identification of the compound as 4-pentenyl ITC CH2 = CH-(CH2)3-NCS(Chart 2).
FA3- 3-Methylthio propyl ITC, Rt. 16.9 min., the mass spectrum gave M + at m/z of 147(12%) which fit with the molecular formula C5H9NS2. The peaks at m/z 101 (M+- CH3S, 100%), 86 (M+- CH3S-CH2, 2%), 72 (M+- CH3S-CH2-CH2, 50%) and 61 (M+- -CH2-CH2-NCS, 44%) confirmed the presence of a methylthio unit attached to a propyl group and the compound identified as 3-Methylthio propyl ITC CH3S-CH2-CH2-CH2-NCS (Chart 2).
FA4- 5-hydroxypentyl ITC: Rt. 19.2 min., the mass spectrum of this compound showed M+ at m/z = 145(3%) correspond to the molecular formula C6H11ONS. The fragments at 101(M+ -(CH2-OH), 22.5%), 87(M+ - NCS, 20.6%), 73(M+- CH2-NCS, 67.8%) and 59(M+ - [CH2-(OH)CH2CH2, 100%) confirm the identification of the aglucone as 5-hydroxypentyl ITC [CH2 (OH)-(CH2)3-CH2- NCS] (Chart 2).
FA5- sulphoraphane (4-methylsulphinyl butyl ITC): Rt. 28.7 min., the MS of the compound revealed the presence of M+ at m/z = 177 corresponding to C6H11ONS2 as a molecular formula. The fragmentation giving rise to 114 (M+ - CH3SO-, 6%), 86 (M+ - CH3SO-CH2, 5%) and 72 (M+ - CH2NCS, 100%). So the compound was identified as 4-methylsulphinyl butyl ITC, CH3SO-(CH2)4NCS(Chart 2).
The variation of glucosinolate hydrolysis products between previous studies and the present work may be due to geographical and seasonal differences, environment, soil type, stress and plant part examined, or due to the differences in the experimental conditions.
2.3. Docking studies:
The docking study have targeted a α-glucosidase and β-amylase, to examine a mode of action of the 4-methylthio-3-butenylglucosinolate. The ligand–protein interaction behavior were estimated based gold score function as implemented in MOE 2015.10[22]. All calculations of docking experiment were represented (Table 2). The crystal structures for α-glucosidase(PDB: 4yvx[23]) and α-amylase(PDB: 4gqr[24]) have been obtained from protein databank, which their complexed with Glimipride and myricetin as reference drugs, respectively. The tested ligand docked into active site for two enzymes.
Figure 2 :
The active site of these enzymes were defined to include residues within a 3.5 Å radius around reference drugs atoms. The molecular dynamic (MD) using (MMFF94) force field to 0.05 Kcal/mol gradient convergence range, which their was applied to minimize energy for obtained ligand-enzyme complexes. The binding affinity for the tested ligand was determined by highest MOE scoring function (Table 3). The premier ligands revealed MOE score (-7.511 and − 6.157 Kcal/mol), respectively, for α-glucosidase and β-amylase. In α-glucosidase binding site, Glimipride make important strong hydrogen bond with His117 and Tyr55, at the time this ligand forms π-π bonds with Tyr24 and Trp22. The analysis binding site of β-amylase-Myricetin pose exhibited that, the ligand interacted with important amino acid backbone as Asp197, which contacted with 2-H-bonds. The tested compound exhibited a higher binding affinity with MOE scores (-6.611 Kcal/mol.) than reference drug (-6.115 Kcal/mol.). This ligand was formed an important hydrogen bond with important Asp amino acid (Fig. 3). Our compound arranged in binding pock at parallel mode with Asp ( Fig. 3).
Table 3
Docking energy scores (Kcal/mol) derived from the MOE for 4-methythio-3-butenyl gluconate.
Enzyme
|
4yvx
|
4gqr
|
|
Glimipride
|
Tested Ligand
|
myricetin
|
Tested Ligand
|
E_score1
|
-7.75118
|
-7.11842
|
-101.157
|
-131.364
|
E.dGE
|
1.62884
|
1.122716
|
-5.19
|
-7.209
|
E_conf
|
-263.798
|
85.79013
|
-101.987
|
-117.44
|
E_place
|
-49.6922
|
-104.664
|
-367.2
|
-566.262
|
E.Int.
|
-13.9234
|
-14.7619
|
-101.987
|
-117.44
|
E.H.B.
|
-42.0326
|
-39.5142
|
-37.74
|
-56.14
|
Eele
|
-7.75118
|
-7.11842
|
-3.07
|
-3.6
|
Evdw
|
-7.49118
|
-7.28842
|
2.59
|
3.24
|
RMS
|
1.62884
|
1.122716
|
0.9652
|
0.865
|
E_score1; Initial free binding energy of the ligand from a given pose. Ed.G.: Final free binding energy of the ligand from a given pose, E_conf; Free binding energy of the ligand from a given conformer. E_place. Free binding energy of the ligand from a receptor. E.Int.: Affinity binding energy of ligand with receptor, H.B.: Hydrogen bonding energy between protein and ligand. Eele: Electrostatic interaction with the receptor, Evdw: Van der Waals energies between the ligand and the receptor. RMSD; The root mean square deviation of the pose of the docking pose compared to the co-crystal ligand position. |
The hydrophilicity of binding pocket is postulated that, the hydrophobicity and membrane permeability are an important pharmacokinetic character for absorption molecule in biological system. The results clearly revealed that, the amino acid residues close to the reference molecules are mostly the same as observed in the tested compound.
Figure 3 :
2.4. In silico Toxicological study
2.4.1. In silico pharmacokinetic Profile:
Cytotoxicity screening in silico through ADMET parameters acting a fundamental route in therapeutic bioactive molecule. The calculated descriptors for 4-methylthio-3-butenylglucosinolate and Glimipride were calculated by MOE, SwissADME [25] and admet-SAR model[26], which these disclosed in (Table 4). The physicochemical and ADME parameters for 4-methythio-3-butenyl glucosinolate Glimipride (reference drug), which stated that this ligand suitable for Lipinski's rule with one violation related to molecular weight. Furthermore, when applied Ghose Veber, Egan and Muegge rules, the tested compounds don’t passed these rules, except Glimipride drug is qualified for Muegge rule. Consequently, the tested compound exhibited bioavailability Scores (0.11 and 0.55) with biodegradation values (0.6362 and 0.4523). These data confirmed that the investigated compound have good oral bioavailability. The lead-likeness profile plotted the Bioavailability Radar planner for tested compound (Fig. 4). The reader figured relation between polarity, size, lipophilicity, solubility, saturation and flexibility[20], the optimal range for each parameters represented in pink color. From the first glance for the reader, 4-methythio-3-butenyl glucosinolate and Glimipride exhibited deviation for polarity and flexibility, respectively, (Fig. 4).
Table 4
ADMET(absorption, distribution, metabolism, and excretion - toxicity ) for 4-methythio-3-butenyl gluconate and Glimipride.
|
Ligand
|
Glimipride
|
|
|
Ligand
|
Glimipride
|
|
Molecular weight
|
434.48
|
490.62
|
Physicochemical
Properties
|
Log Po/w (iLOGP)
|
2.8
|
2.42
|
Lipophilicity parameters
|
Num. heavy atoms
|
26
|
34
|
Log Po/w (XLOGP3)
|
1.38
|
3.85
|
Num. arom. heavy atoms
|
0
|
6
|
Log Po/w (WLOGP)
|
0.42
|
3.77
|
Fraction Csp3
|
0.75
|
0.54
|
Log Po/w (MLOGP)
|
2.37
|
1.88
|
Num. rotatable bonds
|
9
|
11
|
a Consensus Log Po/w
|
1.84
|
2.76
|
Num. H-bond acceptors, donner
|
10,5
|
5,3
|
Lipinski
|
Yes; violation = 1; No rO < 10
|
Yes; 0 violation
|
Drug
likeness
|
TPSA (Topological surface area)
|
186.32A2
|
133.06A2
|
Ghose
|
No; violation = 1: WLOGP>-0.4
|
No; 2 violations: MW > 480, MR > 130
|
Absorption percentage
|
44.48
|
63.09
|
Veber
|
No; violation = 1: TPSA140<
|
No; 1 violation: Rotors > 10
|
Mutagenic
|
none
|
none
|
Pharmacokinetics
parameters
|
Egan
|
No; violation = 1: TPSA131.6<
|
No; 1 violation: TPSA > 131.6
|
Tumorigenic
|
none
|
none
|
Muegge
|
No violation = 2: TPSA < 150 H, acc < 10
|
Yes
|
Carcinogens
|
None
|
None
|
Bioavailability Score
|
0.11
|
0.55
|
Reproductive Effective
|
high
|
high
|
PAINS
|
0 alert
|
0 alert
|
Medicinal chemistry
|
Irritant
|
low
|
low
|
bBrenk
|
0 alerts
|
0 alert
|
HIA (Human Intestinal Absorption)
|
0.9893
|
0.6412
|
Leadlikeness
|
2 violations: MW > 350, Rotors > 7
|
3 violations: MW > 350, Rotors > 7, XLOGP3 > 3.5
|
Caco2- Permeability
|
0.5734
|
0.68523
|
Synthetic accessibility
|
5.66
|
4.71
|
hERG_inhibition
|
ambiguous
|
ambiguous
|
Log S (ESOL)
|
-1.07
|
-4.71
|
water
solubility
|
BBB(blood brain barrier Permeability)
|
No
|
No
|
Solubility
|
0.36 mg/ml; 0.85 × 10− 2 mol/l
|
9.52 × 10 − 3 mg/ml ; 1.94 × 10 − 5 mol/l
|
Biodegradation
|
0.6362
|
0.4523
|
Qualitative solubility
|
Very soluble
|
Poorly soluble
|
GI absorption
|
Low
|
Low
|
cAMES Toxicity
|
None
|
None
|
toxicity
prediction
|
P-gp substrate
|
Yes
|
Yes
|
Acute Oral Toxicity
|
None
|
None
|
CYP1A2 inhibitor
|
No
|
No
|
LD50
|
16 mg/kg
|
3250 mg/mol
|
CYP2C19 inhibitor
|
No
|
No
|
dPredicted Toxicity Class
|
2
|
5
|
CYP2C9 inhibitor
|
No
|
Yes
|
Average similarity%
|
34.27
|
60.83
|
CYP2D6 inhibitor
|
No
|
No
|
Prediction accuracy%
|
23
|
68.07
|
CYP3A4 inhibitor
|
No
|
Yes
|
Binding to Toxicity
|
No binding targets
|
No binding targets
|
Log Kp (skin permeation)
|
-9.93 cm/s
|
-6.56 cm/s
|
Toxic fragments
|
Nil.
|
Nil.
|
a Consensus Log Po/w: average four prediction of partition coefficient between n-octanol- water. |
bhet-C-het_not_in_ring, imine_1, imine_2, oxygen-nitrogen_single_bond |
c In vitro Ames test result for each TA100 strain (Metabolic activation by rat liver homogenate), TA100 strain (No metabolic activation), TA1535 strain (Metabolic activation by rat liver homogenate) and TA1535 strain (No metabolic activation)[9]. |
d Toxicity Class ranging from 1 to 6 according to the Global Harmony System (GHS) [8]. |
Both compounds showed high saturation degree(Fraction Csp3 = 0.75,0.45, respectively), which pass the filter of Fraction Csp3 < 0.25[27]. The solubility parameters are important key for absorption, the 4-methythio-3-butenyl glucosinolate exhibited strong soluble in H2O when applied ESOL topological model with absorption% (%ABS = 44.48)[28], in contrary Glimipride showed poor solubility in this model (Table 4). Also, lipophilicity parameters in medicinal chemistry incorporated as lead-likeness filter, which identified problematic fragments in bioactive molecules based on two Structural alerts as PAINS and Brenk’s filters [20]. The tested compounds exhibited no structural alerts against PAINS and Brenk’s filters.
Figure 4
Mutagenic, Tumorigenic, Reproductive Effective, Irritant, Human Intestinal absorption, were investigated in silico using admet-sar pharmacokinetic parameters[29]. Furthermore, the other (SVM) algorithm was applied to identify substrate or non-substrate of the permeability for skin permeation (Log Kp), Caco-2, blood brain barrier (BBB) and p-glycoprotein (P-gp), as well as, to recognize inhibition effect against the chief cytochromes P450 isoenzymes (CYP1A2, CYP2C19, CYP2C9, CYP2D6, CYP3A4). The results obtained from (Table 3) exhibited that, Glimipride drug inhibited both CYP2C9 and CYP3A4, but our tested ligand has no inhibition action for all types of P450. The BOILED-Egg model is represented relation for WLOGP vs TPSA in (Fig. 4), which demonstrated that, High BBB permeability, high GI absorption, high brain penetration, ambiguous inhibition action for against human Ether-à-go-go-Related Gene (hERG). In addition, the Glimipride showed activation action against P-gp (multidrug resistance protein 1), but our compound act as P-gp inhibitor. Furthermore, skin permeability (Kp) was predicted for our compounds, which described that, the higher negative value combined with low skin permeability for compounds. The data in (Table 4) showed that, the Glimipride drug is higher skin permeability than our ligand. The carcinogen behavior was investigated by admet-sar for the investigated compounds, through compared with 981 various carcinogenic chemical structures, that obtained from “Carcinogenic Potency Database (CPDB [30])”, the results exhibited no carcinogenicity, mutagenicity and tumorigenicity effects for tested compounds.
2.4.2. Oral toxicity prediction:
The possible toxicity was predicted[31], through calculate estimate rodent oral toxicity based on extracted from the Chemical European Biology Laboratory (ChEMBL) database[26]. Also, median lethal doses (LD50) values was estimated in rodents, which our ligand (LD50 = 16 mg/Kg)showed lower value than reference drug (LD50 = 3250 mg/Kg). The calculated toxicity using this database depend on highest endpoints including of 33 models (Table 4). The calculated schematic is devided into several stages of toxicity as “toxicity, toxicological endpoints (mutagenicity, carcinotoxicity, organ toxicity (hepatotoxicity), cytotoxicity and immunotoxicity), toxicological pathways (AOPs) and toxicity targets thereby”[25], which provided a suitable visions into the possible molecular mechanism as well as toxic response (Table 5). The Glimipride located in lower toxicity classes 5 than 4-methythio-3-butenyl glucosinolate, which sited in Class (No. 4), while both compounds do not exhibited any toxic fragments without non-binding to any toxicity targets(Table 5). In general, the tested compounds have been a good oral bioavailability, high ability BBB transport, and no marked health effects observed for rodent toxicity profiles.
Table 5: Toxicity Model Report for ligand which extracted from admet-SAR.