Design, Modeling, Synthesis and in Vitro Evaluation of New DPP-IV Inhibitors

Dipeptidyl peptidase-IV enzyme is one of the targets for drug design to compact type-2 diabetes. The current research addresses the design, modeling, synthesis and in vitro testing of potential new DPP-IV inhibitors. X-ray structure of Sitagliptin with DPP-IV indicated hydrophobic interaction of the triazolo-piperazine ring system with the S2 pocket through the amino acid Phe357. Using molecular modeling; we designed new compounds with better hydrophobic properties facing Phe357. Six compounds were designed, docked, synthesized and tested for DPP-IV inhibitory activity. All designed compounds showed comparable affinity to the catalytic site of the enzyme to that of the standard drug, however, the activity as inhibitors, as reflected by the IC50 range of 0.3 μM – 1.3 μM, was lower than that of Sitagliptin (IC50 22 nM). The most active compound in the new series with, IC50 of 0.3 μM, showed interesting a flipped over docking mode that is different from all other compounds including Sitagliptin with the lipophilic area directed away from the amino acid Phe357. The forces of interaction of the most active compound with the enzyme catalytic site were determined from modeling studies, binding can be optimized by further structural modifications.


Introduction
DPPIV inhibitors are relatively new class of drugs introduced to the market for the treatment of type2 diabetes. Inhibition of DPP-IV results in prolonging the half-life of the glucagon-like peptide (GLP-1). The latter, is a polypeptide hormone secreted by L-cells of the gastric mucosa in response t

Rationale of design
X-ray crystal structure of Sitagliptin bound to DPP-IV, (Figure-2), illustrates that the triazolo-piperazine fragment of the drug is oriented towards the Phe357 residue in the catalytic site, and the trifluromethyl group towards the Ser209 residue. Both residues are part of what is known as the S2 pocket of the catalytic site.  The current research aimed to design molecules maintaining the β-amino amide pharmacophore of Sitagliptin, but with the triazolo-piperazine ring system replaced by more lipophilic entities for better hydrophobic interactions with Phe 357 in the S2 pocket. (Figure 3), depicts the chemical structures of the designed molecules. Guided by molecular modeling, the six designed compounds were docked into DPP-IV catalytic sites to identify their binding poses and compare it with the standard drug pose in the crystal structure. Sitagliptin was also subjected to the same process and energy minimization and superimposition over the crystal structure form to validate the docking process. Compounds 8 and 9 differ in the number of carbons in the linker between the aromatic ring and the β-amino-amide functionality. Same applies for compounds 10 and 12 with the two methoxy groups. Compound 11, on the other hand, with fluro-substituent, designed to test the effect of electron withdrawing group on binding mode and activity, if any. Compound 2 New DPP-IV Inhibitors 13 (the phenylalanine derivative) differs from other five compounds shown in figure 3 in having a carboxylic acid group in the side chain. The latter, was designed to reach out to Arg358 observed in the active site of the S2 pocket for binding enhancement.

Assay for PP-IV inhibition
To determine the DPP-IV inhibition activity of new compounds, we followed previously established standard protocol described in the literature [11][12][13] . Sitagliptin IC 50 was determined as positive control to validate the assay validation. The IC 50 values of the new compounds and Sitagliptin are listed in (table 1).

Material and Methods
Reagents and solvents were generally used as received from the commercial suppliers. Tetrahydrofuran (THF) was distilled using sodium and benzophenone in the still under argon. All the reactions for the synthesis of amides were run under anhydrous condition using nitrogen balloon. Melting Points were determined using a MEL-TEMP apparatus and are uncorrected. 1 H and 13 C NMR spectra were recorded at 90 MHz on an Anasazi instrument Pulse Probe using tetramethylsilane (TMS) as an internal Standard; the values of chemical shift (o) are given in parts per million (ppm) and coupling constants (J) in hertz (Hz). IR spectra were recorded on a Nicolet 10 FT spectrometer with samples placed directly in the path of the light beam. LC/ MS (Mass Spectrum) was used to characterize the final test compounds; the value of m/z was reported as M+H peak. Progress of the reaction was monitored by TLC on silica gel plate (Whatman; PE SIL GIUV). Extracts were dried over calcium chloride or magnesium sulfate, and solvent was removed under reduced pressure. Yields refer to the unpurified products which were further purified by recrystallization technique. Aldrich Silica Gel, 70 -230 mesh, 60A0 was used for column chromatography. Compounds 1 -7 have been previously reported as referenced under experimental (references (41-48), and were characterized by melting points and spectral data. Intermediate (8a-13a) are new compounds and were characterized by spectral data and melting points, and were used directly to prepare the desired target products (8-13) that were fully characterized by melting points, spectral data and elemental analyses. Elemental analyses were performed by Galbraith Laboratories Inc, Knoxville TN.

3-amino-4-(2,4,5,-triflourophenyl) but-2-enoate (4)
In a 200 ml Round bottom flask was dissolved 3-oxo-4-(2, 4, 5-triflourophenyl) butanoate (0.0377 moles) in 65 ml of methanol. In a separate beaker was dissolved in (0.169 moles) of ammonium acetate in 60 ml of methanol. Transfer the ammonium acetate solution in round bottom flask containing starting material and set to Reflux for 7 -8 hrs. The reaction was then evaporated to get a crude oil, which was then extracted with weakly alkaline water (pH-8) and chloroform (2 x 25 ml) to get rid of excess acetic acid. The organic layer then was dried using calcium chloride, filtered and evaporated to get product (0.0359) 95% yield, m.p-65 -66°C. The crude product was recrystallized 4 New DPP-IV Inhibitors using hexane to get fine needle shape crystals 10.56 g m.p-70-71ºC. 1  Methyl 3-amino-4-(2,4,5-triflourophenyl) butanoate (5) In a 250 ml was added in portions sodium borohydride (0.0788 moles) in 70 ml of acetic acid for about 30 min at 15 -20 ° C in Room temperature water bath. The solution was kept stirring until all H2 evolution ceased, after which in a single portion was added enamine (0.0263 moles) to the solution and the reaction mixture was stirred for an hour. Reaction was monitored with TLC at an interval of 30 min in (20% EtOAc/Hexane). The reaction was completed after an hour as indicated by TLC spot at an Rf value 0 compared with sm Rf-0.4, indicating the amine sticking at the base line. Acetic acid was evaporated under vacuum below 50°C. The residue obtained after evaporation was dissolved in methylene chloride and washed with saturated Sodium carbonate solution (4 x 25 ml). The organic layer was then dried using anhydrous calcium chloride. The organic layer was then evaporated to give β-amino ester (oil) 2.2 g (0.00890 moles) 87.32% yield. 1

Methyl-3-((tert-butoxycarbonyl) amino)-4-(2,4,5-triflourophenyl) butanoate (6)
To a stirred solution of amine (0.0220 moles) in 30 ml of dichloromethane was added Boc anhydride (0.0230 moles) dissolved in 30 ml of dichloromethane. The reaction mixture was stirred and left overnight at room temperature. TLC was spotted and found to completion. The reaction mixture was diluted with brine and washed. The organic layer was separated and dried using Calcium chloride. The organic layer was evaporated to get a crude product, m.p-85 -92°C. Column chromatography was performed to purify the product to get (0.0190 moles) 86.47% yield, m.p. 97 -100°C. The crude product was washed with hot petroleum ether to get pure product of 2.500 g (0.00720 moles) m.p-110 -111ºC.

Tert-butyl 4-oxo-4-(phenethylamino)-1-(2,4,5-trifluorophenyl) butan-2-ylcarbamate (8a)
In a 25 ml Round Bottom Flask (r.b.f) 1 was taken 0.300 g (0.000846 moles) of Boc protected β-amino acid and 0.276 g (0.00092) of DEPBT [3-(Diethoxyphosphoryloxy)-(1,2,3)-benzotriazin-4(3H)-one] in 10 ml of anhydrous THF. To this solution was added initially in equivalent molar quantity 0.092 g (0.00092) of triethylamine and the pH of the solution was adjusted to 8 ~ 9 by adding triethylamine. The solution was set to stir for 30 minutes under inert condition using nitrogen gas. In another round bottom flask was taken amine (phenyl ethylamine) 0.1024 g (0.000846 moles) and dissolved in 5 ml of THF. The amine solution was transferred under anhydrous condition to the Rbf 1 using the dry cannula. The amine containing r.b.f was rinsed with another 5 ml of anhydrous THF and transferred. The reaction was set to stir for 2 hrs. A little white solid started appearing after an hour. After 2 hrs reaction was completed as indicated by TLC (90% CHCl 3 / MeOH). The reaction solution was evaporated after and oily residue was obtained. The oily residue obtained was dissolved in EtOAc and was successively washed with 1N HCl (cold 2 x 20 ml), Sat NaHCO 3 (2 x 20 ml) and Brine (2 x 20 ml). The organic layer was dried using sodium sulfate and evaporated to give 0.220 g (0.000522 mol) of white solid product. Yield 61.76 % m.p -182 -185ºC. The white solid product was recrystallized using acetonitrile gave a pure compound with sharp melting point of 185 -187ºC. 1

4-oxo-4-(phenethylamino)-1-(2,4,5-trifluorophenyl)butan-2-aminium chloride (8)
In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added compound (8) 0.200 g (0.000459 moles) and stirred under inert atmosphere. After 30 min a white precipitate started appearing into the solution. The solution was kept for another 30 min and monitored with TLC (90% etlylacetate/hexane). The reaction was completed within an hour and the reaction solution was filtered to collect the precipitate 0.101 g (0.000269 mol) yield 58.57%. The precipitate was washed with diethyl ether to get rid of excess of dioxane and dried. m.p -164 -166 ºC. The 40 mg HCl salt (8) was further recrystallize using MeOH/Diethyl ether to get 25 mg of pure salt after 24hrs m.p-167 -168 ºC which was used for elemental analysis and biological testing. 1

Tert-butyl 4-(benzylamino)-4-oxo-1-(2,4,5-trifluorophenyl) butan-2-ylcarbamate (9a)
In a 25 ml round bottom flask, 1 was taken 0.396 g (0.00112 moles) of Boc protected β-amino acid and 0.389 g (0.0013) of DEPBT [3-(Diethoxyphosphoryloxy)-(1,2,3)-benzotriazin-4(3H)-one] in 7 ml of anhydrous THF. To this solution was added initially in equivalent molar quantity 0.1315 g (0.0013) of triethylamine and the pH of the solution was adjusted to 8 ~ 9 by adding triethylamine. The solution was set to stir for 30 minutes under inert condition using nitrogen gas. In another round bottom flask was taken amine (Benzylamine) 0.0683 g (0.000563 moles) and dissolved in 5 ml of THF. The amine solution was transferred under anhydrous condition to the Rbf 1 using the dry canula. The amine containing r.b.f was rinsed with another 5 ml of anhydrous THF and transferred. The reaction was set to stir for 2 hrs. After 2 hrs reaction was completed as indicated by TLC (90% CHCl 3 / MeOH). The reaction solution was evaporated after and oily residue was obtained. The oily residue obtained was dissolved in EtOAc and was successively washed with 1N HCl (cold 2 x 20 ml), Sat NaHCO 3 (2 x 20 ml) and Brine (2 x 20 ml). The organic layer was dried using sodium sulfate and evaporated to give 0.326 g (0.00077 mol) of white solid product. Yield 68.97% m.p -163 -165ºC. The white solid product was recrystallized using 1-pentanol gave a pure compound with sharp melting point of 163 -164 ºC. 1

4-(benzylamino)-4-oxo-1-(2,4,5-trifluorophenyl)butan-2-aminium chloride (9)
In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added compound (9a) 0.200 g (0.000474 moles) and stirred under inert atmosphere. After 30 min a white precipitate started appearing into the solution. The solution was kept for another 30 min and monitored with TLC (90% EtOAc/hexane). The reaction was completed within an hour and the reaction solution was filtered to collect the precipitate 0.112 g (0.000313 moles) yield 66.00 %. The precipitate was washed with diethyl ether to get rid of excess of dioxane and dried m.p -197 -200ºC. The 40 mg HCl salt (9) was further recrystallize using MeOH/Diethyl ether to get 25 mg of pure after 24hrs salt m.p -202 -203ºC which was used for elemental analysis and biological testing. 1 4-(2,4-dimethoxybenzylamino)-4-oxo-1-(2,4,5- The amine solution was transferred under anhydrous condition to the Rbf 1 using the dry canula. The amine containing r.b.f was rinsed with another 5 ml of anhydrous THF and transferred. The reaction was set to stir for 2 hrs. After 2 hrs reaction was completed as indicated by TLC (90% CHCl 3 / MeOH). The reaction solution was evaporated after and oily residue was obtained. The oily residue obtained was dissolved in EtOAc and was successively washed with 1N HCl (cold 2 x 20 ml), Sat NaHCO 3 (2 x 20 ml) and Brine (2 x 20 ml). The organic layer was dried using sodium sulfate and evaporated to give 0.609 g (0.00126 moles) of yellowish white solid product. Yield 89.61 % m.p -155 -159 ºC. The white solid product was recrystallized using acetonitrile gave a pure compound with sharp melting point of 165. 5

4-(2,4-dimethoxybenzylamino)-4-oxo-1-(2,4,5-trifluorophenyl) butan-2-aminium chloride (10)
In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added com-pound (10a) 0.300 g (0.000622 moles) and stirred under inert atmosphere. After 30 min a white precipitate started appearing into the solution. The solution was kept for another 30 min and monitored with TLC (90% EtOAc/hexane). The reaction was completed within an hour and the reaction solution was filtered to collect the precipitate 0.200 g (0.000478 moles) yield 76.92 %. The precipitate was washed with diethyl ether to get rid of excess of dioxane and dried m.p -154 -160ºC. The 60 mg HCl salt (10) was further recrystallize using MeOH/Diethyl ether to get 40 mg of pure salt after 6 hrs m.p -163 -165.5ºC which was used for elemental analysis and biological testing. 1

butan-2-aminium chloride (11)
In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added compound (11) 0.200 g (0.000441 moles) and stirred under inert atmosphere. After 30 min a white precipitate started appearing into the solution. The solution was kept for another 30 min and monitored with TLC (90% EtOAc/hexane). The reaction was completed within an hour and the reaction solution was filtered to collect the precipitate 0.120 g (0.000308 moles) yield 69.77 %. The precipitate was washed with diethyl ether to get rid of excess dioxane, and dried, m.p -147 -152ºC. The 50 mg HCl salt (10) was further recrystallize using MeOH/Diethyl ether and after 3 days the salt came out of the solution to get 34 mg of pure salt m.p -152 -153ºC which was used for elemental analysis and biological testing. 1

Tert-butyl 4-(3,5-dimethoxyphenethylamino)-4-oxo-1-(2,4,5 -trifluorophenyl)butan-2-ylcarbamate (12a)
In a 25 ml Round Bottom Flask (r.b.f) 1 was taken 0.500 g (0.00141 moles) of Boc protected β-amino acid and 0.509 g (0.0017) of DEPBT [3-(Diethoxyphosphoryloxy)-1,2,3-benzotriazin-4(3H)-one] in 7 ml of anhydrous THF. To this solution was added initially in equivalent molar quantity 0.243 g (0.0017) of triethylamine and the pH of the solution was adjusted to 8~9 by adding triethylamine. The solution was set to stir for 30 minutes under inert condition using nitrogen gas. In another round bottom flask was taken amine (3,5 dimethoxy phenyl ethyl-amine) 0.255g (0.00141 moles) and dissolved in 5 ml of THF. The amine solution was transferred under anhydrous condition to the Rbf 1 using the dry canula. The amine containing r.b.f was rinsed with another 5 ml of anhydrous THF and transferred. The reaction was set to stir for 2 hrs. After 2 hrs reaction was completed as indicated by TLC (90% CHCl 3 / MeOH). The reaction solution was evaporated after and oily residue was obtained. The oily residue obtained was dissolved in EtOAc and was successively washed with 1N HCl (cold 2 x 20 ml), Sat NaHCO 3 (2 x 20 ml) and Brine (2 x 20 ml). The organic layer was dried using sodium sulfate and evaporated to give 0.660 g (0.00133 moles) of yellowish white solid product. Yield 94.37% m.p -147 -151 ºC. The yellowish white solid product was recrystallized using acetonitrile gave a pure compound with sharp melting point of 160 -162ºC. 1

4-(3,5-dimethoxyphenethylamino)-4-oxo-1-(2,4,5-trifluorophenyl)butan-2-aminium chloride (12)
5 In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added compound 0.250 g (0.000504 moles) and stirred under inert atmosphere. The solution was kept for an hour and monitored with TLC (90% EtOAc/hexane). The reaction was completed after an hour and the reaction mixture was evaporated. The residue was dissolved in water in a round bottom flask, kept in an ice bath, and the pH of the solution was adjusted to 8 ~ 9 using sodium bicarbonate and the aqueous layer were extracted with ethyl acetate. The organic layer was dried using calcium chloride and evaporated to get 0.120 g (0.000303 moles) yield 60.13%. The precipitate was washed with diethyl ether to get rid of excess of dioxane and dried m.p -150 -154ºC. The 40 mg HCl salt (12) was further recrystallize using MeOH/Diethyl ether/HCl in dioxane and after 24 hrs the salt came out of the solution to get 24 mg of pure salt m.p -154 -155ºC which was used for elemental analysis and biological testing.

2-(3-(tert-butoxycarbonylamino)-4-(2,4,5-trifluorophenyl) butanamido)-3-phenylpropanoic acid (13a)
In a 25 ml of round bottom flask was taken 1.00 g (0.00202 moles) of compound 13aa. The compound was dissolved in 12/2 ml of anhydrous THF/Methanol. The solution was stirred for 30 minutes in ice bath. To this cold solution was added 2 ml of 2N Methanolic KOH and the solution was stirred for an hour. The reaction was monitored using TLC (90% EtOAc/hexane). The reaction was completed after 2 hrs and white solid salt precipitating out of the reaction mixture was observed. The reaction mixture was completely evaporated after the completion of reaction to dryness. The dry white solid salt was dissolved in 30 ml of cold distilled water and stirred until all the residue was dissolved, after this the pH of the solution was adjusted to pH 3 ~ 4 using 1N HCl. The precipitate was observed after adjusting the pH which was extracted with diethyl ether/ethyl acetate (4:3, 4 x 25 ml). The organic layer was separated and dried using anhydrous sodium sulfate and evaporated to dryness under vacuum to get 0.891 g (0.00186 moles) of white solid product yield-92 %. m.p -183. 5

4-((1-carboxy-2-phenylethyl) amino)-4-oxo-1-(2,4,5-trifluorophenyl) butan-2-aminium chloride (13)
In a dry 5 ml round bottom flask was added 2 ml of 4M HCl/Dioxane solution and stirred in an ice bath for 30 minutes under inert atmosphere. To this ice cold solution was added compound (13a) 0.250 g (0.000520 moles) and stirred under inert atmosphere. The solution was kept for an hour and monitored with TLC (90% EtOAc/hexane). The reaction was completed after an hour and the reaction mixture was evaporated. The residue was washed with (2 x 25 ml) diethyl ether and evaporated, after an initial wash the residue was again dissolved in diethyl ether and decanted to get rid of impurities. The residue was dried in a vacuum at 40°C for 3 hrs to get 0.120 g (0.000288 moles) yellowish white solid yield 60.13%. The precipitate was washed with diethyl ether to get rid of excess of dioxane and dried m.p -82 -89ºC. The 40 mg HCl salt (10a) was further recrystallize using MeOH/Diethyl ether/HCl in dioxane and after 24 hrs the salt came out of the solution to get 24 mg of pure salt m.p -112 -114.5ºC which was used for elemental analysis and biological testing. 1

Molecular Modeling
All compounds were subjected to energy minimization before docking into the crystal structure catalytic site. The energy minimization and docking were accomplished using "Vina" software and "Pyrx'' (V.08) program which provides a Microsoft Windows interface for Vina [14,15] .
Docking results were visualized through discovery studio software [16] . The validity of the docking process was assessed by comparing the docking mode of Sitagliptin-lowest energy pose, generated by Pyrx, with the crystal structure mode of binding as shown in figure 4. The modeling process starts with downloading the crystal structure of sitaglipting docked with DPP-IV in a pdb format (PDB ID 1 x 70) from the protein databank (www.rcsb.org). To prepare the protein for docking studies; the .pdb file was imported to Autodock Tools (v.1.5.6). Water molecules and the ligand (sitagliptin) were removed from the structure, the polar hydrogen's were added to the residues and the structure was converted to a .pdb qt file format in which the partial charges on the residues were calculated. To prepare the ligands for docking, the structure of the ligands was drawn using Marvin Sketch (V.15.11.9) then saved as a .sdf file format [17] . The protein as well as the ligands files were then imported to Pyrx software where the energy of ligands were minimized using Open Babel extension in the program, then the files converted to .pd bqt format and save [18] . The grid box for the docking was set to be 25 A° X 25 A° X and was centered on the active site which is located at the following coordinates: X = 40.1242 , Y = 50.9588 , Z = 36.1963. The docking studies were performed three times and the potential energies were reported as average of the three runs as summarized in (Table-1  Interestingly, the energy minimization studies of the six compounds and Sitagliptin revealed very comparable energy minimum values as listed in (table-1). Those values are reflection of the binding affinities to the active site. (Table-1) data indicate that the designed lipophilic areas of the new compounds resulted in comparable affinities to that of Sitagliptin to the catalytic site. (Figure 5 a-d) with new compounds docked into the active site, also confirms that the lipophilic entities introduced in the new compounds are oriented facing the lipophilic amino acid residue Phe 357 in the S2 pocket. (Figure-5 a-d), also depicts super imposable binding mode with sitagliptin in the catalytic site. However, although the hydrophobic interaction with Phe 357 were observed, and the compounds are general super imposable on sitagliptin pose, (figure-5 a-d), the activity of the designed compounds were approximately an order of magnitude less active than sitagliptin, (table-1). The IC 50 values of compounds 8 -12 were generally unaffected by the substitutions on the aromatic ring, nor by the distance the aromatic ring from the β-amino amide. This may signify that other interactions of the triazolo-piperazine ring system of Sitagliptin such as electrostatic and hydrogen bonding may play more important role in the binding process than just the hydrophobic interactions. Another interesting observation from the docking studies; that all designed compounds, except compound 13 which is a carboxylic acid derivative of compound 8, were found to be with almost full superimposition poses over the standard drug in the active site (5a-d). However, interestingly, when compound 13 when docked into the active site it was found to bind in a flipped over mode. The carboxylic group instead of heading toward Arg358 the S-2 pocket, to form ionic interaction, was found to be orient the entire molecule into a flipped over mode with new and different set of interactions with the active site. The trifluro phenyl ring is now facing Phe357, and two of its fluro groups forming new bonds with Glu205, Glu206, Tyr666. The amide carbonyl has interaction with Tyr547 hydroxyl group, and the carboxylic group headed down away from the S2 pocket to interact with Tyr666 hydroxyl group ( figure 5c-13). The new set of interactions, possibly, contributes to the 5-fold increase in activity of compound 13 over the rest of the series and to its closer activity to that of Sitagliptin. Further structural modifications are possible to enhance the binding of compound 13, for possible increase in affinity, hence, higher activity.