New Synthetic 1H-1,2,3-Triazole Derivatives of 3-O-Acetyl-β-boswellic Acid and 3-O-acetyl-11-keto-β-boswellic Acid from Boswellia Sacra Inhibits Carbonic Anhydrase II in Vitro

Boswellic acids are genus specic to Boswellia; they are the principal biologically active compounds holding exceptionally potent anti-inammatory activity. A series of new 1H-1,2,3-triazole tethered of 3-O-acetyl-β-boswellic acid (ABA, 1) and 3-O-acetyl-11-keto-β-boswellic acid (AKBA, 2) derivatives (10a-d and 11a-d) were synthesized and their carbonic anhydrase II (CA II) inhibitory activity was evaluated in vitro. All compounds were characterized by 1 H NMR, 13 C NMR, 2D NMR (HMBC, HSQC, COSY and NOESY) experiments, ESI-MS, and when applicable by 19 F NMR spectroscopy (10b, 10c and 11b, 11c). This series has displayed a moderate to strong inhibition against CA II with IC 50 values of 13.2–60.1 µM. All the active compounds were reported for the rst time for their CA II inhibition potential. Kinetics studies on the most active inhibitors (5 and 10b) were carried out to investigate their mode of inhibition and to determine their inhibition constants K i . Both compounds (5 and 10b) were found to be non-competitive inhibitors with K i values of 10.40 ± 0.013 and 14.25 ± 0.017 µM, respectively. Molecular docking studies showed that all compounds were well accommodated in the allosteric site of CA II. The current study has demonstrated the usefulness of incorporating a 1H-1,2,3-triazole moiety into the boswellic acids skeleton.

We were interested to explore the potential and biological activity of the natural products having this scaffold. Previously, several novel analogues of these molecules have been prepared in our laboratory as well as by other groups (Figure 3) 4, [13][14][15][16][17][18] . Furthermore, several 1H-1,2,3-triazole substituted compounds were synthesized 19 and found to be excellent inhibitors against carbonic anhydrase II (CA II) enzyme 20,21 . Given the interesting biological activities of both triazole derivatives and boswellic acids, we aimed to synthesize triazole tethered boswellic acids ( Figure 3).
Carbonic anhydrases (CAs, EC 4.2.1.1) are class of metallo-enzymes using zinc as a cofactor for the reversible inter-conversion of carbon dioxide and bicarbonate 26,27 . Among the 16 CA isoforms reported so far, only 12 isoforms are catalytically active and varies with respect to their location, kinetic properties, and inhibitor pro les 28,29 . CAs are involved in different physiological and pathological processes [30][31][32] , as a consequence, they seem to be interesting therapeutic targets to treat pathological disorders [33][34][35] . CA II is mainly involved in the regulation of the bicarbonate concentration in the eyes. CA II inhibitors can be used to reduce the intraocular pressure usually associated with glaucoma [36][37][38] . Other than that CA II is also expressed in malignant brain tumors 39 , renal, gastritic and pancreatic carcinomas [40][41][42] . These inhibitors have also been considered as an adjunct in the chemotherapy of cancer. There are number of sulfonamides derivatives reported for CA II inhibitions [43][44][45][46][47][48][49] .
Herein, we report for the rst time the synthesis and in-vitro CA II inhibitory activity of a novel series of 1H-1,2,3-triazole analogues of ABA and AKBA. Furthermore, some structure-activity relationships, molecular docking, and kinetic studies of the active analogues were discussed.
Thereby, in the initial step, compounds 1 and 2 were treated in DMF with 1-bromo propanol in the presence of potassium carbonate at room temperature to afford the esters 3 (yield 76%) and 4 (yield 74%), respectively 51 . Compound 3 holds a free -OH group, which was converted into tosylated 5 (yield 80%) by its treatment with p-toluenesulfonyl chloride in the presence of triethylamine (Et 3 N) and 4dimethylaminopyridine (DMAP) in dry DCM 52 . Under similar conditions, the AKBA derivative 6 was obtained from compound 4 in 83% yield. In the next step, compounds 5 and 6 were treated with sodium azide (NaN 3 ) in DMF at 70 °C to afford the corresponding azides 7 and 8 in good yields of 74% and 71%, respectively 53 . The nal step was carried out using "Click" chemistry. Thereby, a 1,3-dipolar cycloaddition reaction between β-ABA-azide 7 and different alkyne derivatives 9a-d in the presence of copper iodide (CuI) and Hünig's base in MeCN furnished the desired products, 1H-1,2,3-triazole analogues of β-ABA 10a-

In Vitro Carbonic Anhydrase Ii Inhibition
Different 1H-1,2,3-triazole derivatives of 1 and 2 were evaluated for their ability to act as inhibitor of CA II. All the assays were carried out at micro molar level using acetazolamide as a standard inhibitor showing an IC 50 value of 18.2 ± 1.23 µM. After preliminary screening, the parent compounds 1 and 2 held weak activity with IC 50 values of 55.5 ± 1.32 and 95.5 ± 1.74 µM as compared to the reference compound (Table 1). The addition of bulky groups on different position enhanced their inhibitory activity, and as a result, all compounds (3-11d) showed signi cant CA II inhibition with IC 50 values in the range of 13.2-60.1 µM. Compounds 5 and 10b were found to be the best inhibitors in this series, while compounds 3, 7, 10a, 10c, 10d, 8 and 11c showed only moderate inhibition with respect to the standard. However, compounds 6 and 11d were weak inhibitors of this series (Table 1).  From the kinetics studies, it was deduced that 5 and 10b are non-competitive inhibitors. The type of inhibition was determined from Lineweaver-Burk plots. (Figures 4 and 5).
The K i values were determined from secondary replots of the Lineweaver-Burk plots by plotting the slope of each line in the Lineweaver-Burk plots against different concentrations of compounds 5 and 10b ( Figure 4B and 5B). The K i values were con rmed from Dixon plots after plotting the reciprocal of the rate of reaction against different concentrations of compounds 5 and 10b ( Figure 4B and 5B).

Molecular docking and predicted structure-activity relationship
Derivatives of ABA and AKBA (3, 5-8, 10a-10d, 11c and 11d) were targeted at the allosteric site (AS) of human CA II 56 . This revealed that these compounds are tted neatly at the entrance of the active site between AS1 and AS2. The entrance of the active site of CA II is lined with several hydrophobic residues.
The allosteric site 1 (AS1) is located 3-5Å away from the active site and mainly constituted by Val121, Val143, Val207, Trp209, Leu198 and Thr199. AS2 is found near 5-7Å of active site and composed of Tyr7, His64, Asn62, Asn67 and Thr200. Moreover, there is another site present behind the active site where carbon dioxide can bind; this site was termed as AS3 in this study. This site is located 10-12Å far behind the active site and mainly formed by hydrophobic residues like Trp97, Phe226, Val223, Gln222, Val218, Leu157, Leu148, Ala116 and Phe95 . All the binding sites are shown in Figure 6.
The most active compound, 5 (IC 50 = 13.2 ± 0.51µM), demonstrated a 5-fold better activity than the standard 'acetazolamide'. The docked view of 5 revealed that the substituted toluene sulfonic acid tted deep inside AS1 and held hydrophobic interaction with Trp209, while sulfonic acid interacted with Thr200 of AS2 and Gln92 (lining the active site gorge) through H-bonding. The propyl ester and the acetate group formed H-bonds with the side chain of Asn67 and His64 of AS2, respectively. The triterpene moiety of the compound remained at the entrance of the active site, thus blocking the access to the active gorge. The binding mode of 5 is shown in Figure 7. Both AS1 and AS2 stabilize the compound at the surface of the enzyme where residues provided strong bonding to the compound and made the compound highest active within this series if compounds. The docking results are tabulated in Table 1.
Additionally, we predicted the solubility potential of compounds through logP, logS, and TPSA values. The partition coe cient (logP) of a drug is the ratio of its concentrations in a mixture of two immiscible solvents at equilibrium. In medicinal chemistry, one of the solvents is water (which represent blood serum) and the second is n-octanol which is hydrophobic (indicates lipid bilayer). Hence logP measures hydrophilicity and hydrophobicity of a compound, therefore, estimates the distribution of drugs within the body. Hydrophobic drugs with high octanol-water logP are distributed to hydrophobic areas (lipid bilayers of cells) while hydrophilic drugs (low octanol/water logP) are circulated in aqueous regions (blood serum). Low hydrophilicities and high logP values cause poor absorption or permeation. It has been shown for compounds to have a reasonable propability of being well absorbt their logP value must not be ≥ 5.0. The calculated logP of compounds depict that the compounds are highly lipophilic, thus they can easily penetrate the lipid bilayer, however their absorption through the intestine is very low due to their poor solubility or insolubility in the water.
The polar or topological polar surface area (PSA or TPSA) is also a commonly used metric in the medicinal chemistry for the optimization of a compound or a drug's ability to permeate cells. TPSA is de ned as the surface sum over all polar atoms (primarily oxygen and nitrogen with their attached hydrogen atoms) of a compound. Molecules with a PSA of ≥140Å 2 are usually have poor permeability in the cell membranes, however, for molecules which act on CNS, PSA should be ≤90Å 2 in order to effectively penetrate the blood-brain barrier. The TPSA values of compounds 3, 5-8, 10a-10d, 11c and 11d were in range of 72.83 to 126.68 Å 2 , indicating that these molecules cannot pass blood brain barrier, however, possess moderate permeability to the cell membrane. The results are shown in Table 2.

Conclusions
In summary, novel 1H-1,2,3-triazole analogues of 1 and 2 were synthesized (10a-d and 11a-d), and evaluated for their CA II inhibitory potential in vitro. The C-4 acid group of ABA and AKBA was transformed into an ester containing a hydroxyl group which was modi ed to an azide through tosylation. The azide compounds were subsequently converted to 1H-1,2,3-triazoles having aromatic and ester substituents. Except three compounds (4, 11a and 11b), all compounds exhibited good inhibitory potential against this enzyme. Kinetic assays demonstrated that these derivatives are non-competitive inhibitors. Additionally, molecular docking indicated that the active compounds have perfectly tted into the allosteric site of CA II.

General
Reagents were obtained from Sigma-Aldrich, Germany. Silica gel for column chromatography were of 100-200 mesh. Solvents were puri ed by following standard procedures. Thin layer chromatography (TLC) was carried using silica gel F 254 pre-coated plates. UV-light and I 2 stain were used to visualize the spots. The 1 H and 13 C NMR spectra were recorded on NMR spectrometer (Bruker: 600 MHz for 1 H, 150 MHz for 13 C and 564 MHz for 19 F) using CDCl 3 as a solvent. The high-resolution electrospray ionization mass spectra (HR-ESI-MS) were recorded on Agilent 6530 LC Q-TOF instrument. Organic extracts and solutions of pure compounds were dried over anhydrous MgSO 4 .

Molecular Docking
The X-ray crystallographic structure of CA II in complex with carbon dioxide and bicarbonate ion (PDB code: 2VVB, resolution: 1.66Å) was used in docking. The structures of all the compounds were prepared on ChemDraw software and saved in mol format, then imported into MOE database where each molecule was minimized with MMFF94x force eld until an RMSD gradient of 0.1 kcal•mol −1 Å −1 was achieved, and the partial charges were automatically applied on each molecule during the minimization process.
Hydrogen atoms and partial charges were added on the enzyme structure with the default settings of the Protonate 3D protocol in MOE. The binding site was de ned by selecting the residues of the allosteric site. Triangle Matcher placement method and London dG scoring function were used for docking. Later, thirty docked conformations of each compound were saved for conformational sampling. The best docked pose of each compound with respect to the docking score and binding interactions, was selected for SAR analysis.
The logP, logS, and TPSA values of active compounds were calculated by MOE using the 3D-structures of compounds, while gastro-intestinal absorption and blood brain barrier permeability was calculated through SwissADME webserver (http://www.swissadme.ch/index.php).

Declarations
Con ict of interest All authors con rm that this article content has no con ict of interest.