Synthesis of dehydrodipeptide esters and their evaluation as inhibitors of cathepsin C

The procedures for the synthesis of esters of dehydropeptides containing C-terminal (Z)-dehydrophenylalanine and dehydroalanine have been elaborated. These esters appeared to be moderate or weak inhibitors of cathepsin C, with some of them exhibiting slow-binding behavior. As shown by molecular modeling, they are rather bound at the surface of the enzyme and are not submersed in its binding cavities.

Although from some years we have been engaged in studies on the dependence of three-dimensional structure of dehydropeptides on their inhibitory activity toward cathepsin C, no clear structure-activity relationship could be drawn (Makowski et al., 2001;Latajka et al., 2006Latajka et al., , 2008. In this paper, we present synthesis of esters of glycyl Z dehydrophenylalanine (Gly-Z DPhe), glycyldehydroalanine (Gly-DAla) and L-phenylalanyldehydroalanine (Phe-DAla) and evaluation of their action toward this enzyme.

Materials and methods
General All reagents and solvents were purchased from Sigma-Aldrich, Avantor Performance Materials or Merck. Ethyl acetate (EtOAc), dichloromethane (DCM), diethyl ether (Et 2 O) and tetrahydrofuran (THF) were dried over P 2 O 5 and then distilled. N,N-dimethyformamide (DMF) was distilled under reduced pressure and stored over molecular sieves 4 Å . Other chemicals were used without purification. Reaction progress was monitored by TLC on Merck 60 silica plates. The spots were visualized by placing Electronic supplementary material The online version of this article (doi:10.1007/s00044-015-1366-0) contains supplementary material, which is available to authorized users. chromatogram plate at chlorine vapor followed by spraying with o-tolidine in water/acetic acid mixture. NMR spectra were recorded on Bruker Ultrashield 400 MHz instrument, operating at 400 MHz ( 1 H) and 100 MHz ( 13 C). Samples were prepared in . Chemical shifts are reported in ppm relative to TMS used as internal standard or to the signal of solvent ( 1 H NMR 2.5 ppm; 13 C NMR 39.52 ppm for DMSO-d6), and coupling constant is reported in Hertz. In the description of dipeptide 1 H NMR and 13 C NMR spectra, the tosylate group is omitted for better readability (Tos 1 H NMR (DMSO, 400 MHz): d 7.48 (d, J = 8.0 Hz, 2H, ArH), 7.11 (d, J = 8.0 Hz, 2H, ArH), 2.29 (s, 3H, CH 3 ); 13 C NMR (DMSO, 100 MHz): d 20.84 (CH 3 ), 125.57,128.14,137.80,145.71 (4 9 ArC). The copies of all NMR spectra are available at electronic supplementary material. Melting points were determined on a Stuart SMP30 apparatus and are reported uncorrected. Mass spectra were recorded on Bruker micrOTOF-Q II high-resolution mass spectrometer with electrospray ionization (ESI). IR spectra were recorded on Nicolet 6700 FT-IR spectrophotometer (Thermo Scientific) operating at resolution 2 cm -1 and scanning range 4000-400 cm -1 . Samples were measured as KBr disks.

Synthesis of N-protected dehydrodipeptides
Boc-protected dehydrodipeptides containing C-terminal dehydroalanine (DAla) or (Z)-dehydrophenylalanine (D Z-Phe) were synthesized earlier by condensation of appropriate carboxamides with a-keto acids in benzene in the presence of p-toluenesulfonic acid as catalyst (Makowski et al., 1985).

Synthesis of dehydrodipeptide methyl, ethyl and isopropyl esters
Syntheses were based on procedure of Cossec et al. (2008). Thus, Boc-Gly-DAla or Boc-(S)Phe-DAla was dissolved in methanol (0.2 or 0.4 M, respectively), and 0.5 equivalent of Cs 2 CO 3 was added. The mixture was stirred for 1 h at room temperature followed by evaporation of solvent. The dipeptide cesium salt was dissolved in DMF (0.28 M), and fivefold or fourfold excess (respectively) of methyl, ethyl or isopropyl iodide was added in portions. After completion of the reaction (3-5 h, controlled by TLC), solvent was evaporated under reduced pressure. The obtained residue was dissolved in ethyl acetate and washed subsequently with: 1 M HCl, saturated KHCO 3 , 0.1 M Na 2 S 2 O 3 and brine (each one in triplicate). Organic layer was dried over anhydrous MgSO 4 . Product was crystallized from mixtures of diethyl ether/hexane or ethyl acetate/hexane providing Boc-Gly-DAla-OMe in 91 %, Boc-(S)Phe-DAla-OMe in 94 %, Boc-(S)Phe-DAla-OEt in 94 % and Boc-(S)Phe-DAla-OPr i in 81 % yields. Deprotection of amine group was performed in 20 % solution of TFA in DCM. Deprotection of amine group of dehydrodipeptide esters containing dehydroalanine required the use of anisole (3 % v/v) for protection against oligomerization reactions. Mixture was stirred at room temperature for 30 min, and equivalent of p-toluenesulfonic acid was added. Mixing was continued for 15 min, and solvent was removed under reduced pressure. The residue was dissolved in dichloromethane, and solvent was carefully evaporated to remove the excess of trifluoroacetic acid. Products were crystallized from mixtures of isopropanol/hexane. ), 5.00 (hept, J = 6.2 Hz, 1H, CH(CH 3 ) 2 ), 4.43-4.34 (m, 1H, CH Phe ), 3.10 (dd, J = 13.9, 6.1 Hz, 1H, ABX system CH A H B Phe ), 2.99 (dd, J = 13.9, 7.8 Hz, 1H, ABX system CH A H B Phe ), 1.26 (d, J = 6.2 Hz, 6H, CH(CH 3 ) 2 ). 13

Synthesis of allyl and propargyl esters of dipeptides containing dehydroalanine
A Cs 2 CO 3 0.163 g (0.5 mmol) was added to solution of Boc-Gly-DAla 0.244 g (1 mmol) or Boc-(S)Phe-DAla 0.334 g (1 mmol) in 5 mL of methanol. The mixture was stirred at room temperature for 2 h, and solvent was removed under reduced pressure. Solid residue was dissolved in 5 mL of THF for Boc-Gly-DAla or 5 mL of DMF for Boc-(S)Phe-DAla, and allyl bromide 0.856 mL (10 mmol) or propargyl bromide 1.114 mL (10 mmol) was added dropwise over 15 min. When peptide substrate was consumed (controlled by TLC), the solvent and excess of bromide were removed under reduced pressure. The residue was dissolved in 80 mL of ethyl acetate, filtrated and washed with: 1 M HCl (4 9 5 mL), saturated KHCO 3 (4 9 5 mL) and brine. Organic layer was dried over MgSO 4 and filtered, and 0.2 mL of anisole was added. The solvent was removed under reduced pressure at 35°C. The residue was dissolved in 10 mL DCM, 1.5 mL of TFA was added and the mixture was stirred for 1 h at room temperature followed by addition of 0.190 g (1 mmol) of ptoluenesulfonic acid. Stirring was continued for additional 20 min, and solvent was removed under reduced pressure. The residue was evaporated two times with 20 mL of DCM to remove TFA excess. Products were crystallized from mixtures of isopropanol/hexane

Synthesis of allyl and propargyl esters of dipeptides containing (Z)-dehydrophenylalanine
Boc-Gly-D Z Phe 0.320 g (1.0 mmol) was dissolved in 5 mL DMF, and Cs 2 CO 3 0.163 g (0.5 mmol) was added. Mixture was stirred for 3 h, and allyl bromide 0.856 mL (10 mmol) or propargyl bromide 1.114 mL (10 mmol) was added dropwise over 15 min. The reaction was continued for 12 h stirring at room temperature. Further steps of synthesis were done according to procedure described for allyl and propargyl esters of Boc-Gly-DAla. The deprotection reaction of amine group was performed without addition of anisole and p-toluenesulfonic acid.

Efforts to synthesize dehydrodipeptide glycidyl esters
Method I Boc-Gly-D Z Phe 0.160 g (0.5 mmol), Et 3 N 0.196 mL (1.1 mmol) and (S)-glycidol 0.266 mL (2.0 mmol) were dissolved in 2.0 mL of acetonitrile, and TBTU (Abdelmoty et al., 1994) 0.208 g (0.65 mmol) was then added. Mixture was stirred at room temperature for 2.5 h, and solvent removed under reduced pressure. The residue was dissolved in 70 mL of ethyl acetate and washed subsequently with: 1 M HCl (3 9 5 mL), saturated KHCO 3 (3 9 5 mL) and brine. Organic phase was dried over MgSO 4 and filtered, and solvents were removed. We were unable to purify a mixture of products obtained using column chromatography with silica gel 60H (Merck) as stationary phase and various eluents. Thus, crude mixture was used in deprotection step. HRMS (ESI) indicated the presence of the desired product as a major one: m/z calcd for C 19 H 24 N 2 O 6 (M ? Na) ? 399.1526; found 399.1529.
Method II iso-butyl chloroformate 0.066 mL (0.5 mmol) was added to solution of Boc-Gly-D Z Phe 0.160 g (0.5 mmol) and Et 3 N 0.070 mL (0.5 mmol) in dichloromethane when cooling in ice bath to -15°C. After 1.5 min, glycidol 0.133 mL (1.0 mmol) was added. The mixture was left to warm to room temperature, and stirring was continued for next 24 h. Further steps of synthesis were performed according to the methodology described for Method I and afforded similar mixture of products.

Deprotection of amine group
Method I Trifluoroacetic acid 0.5 mL was added to solution of Boc-Gly-D Z Phe-OGdl 0.098 g (0.25 mmol) in 2 mL of dichloromethane. Mixture was stirred for 20 min at room temperature, and solvent was removed under reduced pressure. The residue was evaporated three times with 20 mL of dichloromethane and 20 mL of diethyl ether to remove the excess of trifluoroacetic acid. Mixture of products was obtained as oily residue. HRMS (ESI) indicated the presence of the two major products-desired glycidol ester (Gly-D Z Phe-OGdl(S)) and the product of oxirane ring opening-Gly-D Z Phe-OCH 2 CH(OH)CH 2 OH: m/z calcd for C 14  Method II HCl in methanol (*3.8 M) solution was prepared by bubbling dry HCl gas through methanol for 1 h at 0°C. Crude Boc-Gly-D Z Phe-OGdl(S) 0.129 g (0.34 mmol) was dissolved in methanol (1.2 mL), and HClmethanol solution was added (1.3 mL). After 1 h at room temperature, solvent was evaporated under reduced pressure. The oil residue was evaporated three times with 5 mL of dichloromethane. Product was crystallized from mixture of isopropanol/diethyl ether/hexane (2:1), filtered and dried in vacuo.

Enzymatic studies
Cathepsin C was isolated from bovine spleen by modified method of McDonald et al. (1972). The K M value of 2.3 mM for the enzyme was measured using synthetic substrate-glycine-L-phenylalanine-p-nitroanilide (Gly-L-Phe-pNA). Purity of the enzyme was confirmed by electrophoresis.

Inhibitory studies
Cathepsin C was activated for 0.5 h in a water bath at 37°C in 1 % NaCl solution containing 1 mM EDTA-Na 2 and 5 mM 2-mercaptoethanol. The enzymatic reaction was carried out at 37°C in 100 mM acetate buffer, pH 5.0, containing 1 mM EDTA-Na 2 , 1 mM DTT and 30 mM NaCl (all final concentrations). The progress of the reaction was monitored spectrophotometrically (UV-Vis spectrophotometer Cintra 303) at a wavelength of 405 nm against a control sample containing no enzyme. Attempting mixture contained: solution of the synthetic substrate Gly-L-Phe-pNA in acetate buffer at pH 5 containing 1 mM EDTA-Na 2 , 1 mM DTT, 30 mM NaCl (substrate concentration: 2.7-0.01 mM-final concentration), the solution of inhibitor in reaction buffer (concentration of compound depended on inhibitory potential), and enzyme.
Kinetic constants K M , V max and K i and type of inhibition were determined by using Lineweaver-Burk, Dixon, Hanes-Woolf and half-inhibitory concentration methods using the computer program provided kindly by dr Józef Hurek (University of Opole). The K i values presented in the Table 1 are the average ones calculated by using all these methods. All measurements were taken in a three repetitions.

Molecular modeling
The structures of studied dehydropeptides were optimized in Gaussian09 program at the B3LYP/6-311 g (d,p) level (Frisch et al., 2004) in gas phase with using Merz-Singh-Kollman scheme (Besler et al., 1990) to the determination of the atomic charges. The calculations of the docking process were performed using AutoDock program (Morris et al., 2009). The starting geometry and charges of the dehydropeptides were taken from the ab initio calculations. The structure of cathepsin C was extracted from the structure of human dipeptidyl peptidase I deposited EC 3.4.14 in Protein Data Bank (Turk et al., 2001). Structure of the enzyme has been protonated on the H?? server (Myers et al., 2006) at pH = 5.7, and also charges of all enzymatic atoms have been assignment on this server. During the docking process, main chain of the dehydropeptide was fixed, whereas side chains and the terminal groups were left as flexible. The coordinates of the SH proton from the Cys234 were taken as a grid center in the docking process. In the simulation, docking process was performed 100 times. Analysis of the obtained results has been performed by using AutoDock Tools (Morris et al., 2009).

Results and discussion
Cathepsin C (EC 3.4.14.1) is a lysosomal cysteine protease expressed in majority of mammalian tissues and is primarily responsible for activation of serine proteases in inflammatory and immune cells (Reiser et al., 2010). It sequentially removes dipeptides from the N-termini of protein and peptide substrates (Lindley, 1972;Poręba et al., 2014). Increasing evidence of the key role of DPPI in various diseases, such as sepsis, asthma, Duchenne muscular dystrophy, rheumatoid arthritis, basal cell carcinomas, chronic obstructive pulmonary disease and other inflammatory disorders (Guay et al., 2010;Laine and Busch-Petersen, 2010), stimulates interest in this enzyme as the possible medicinal target.
Dehydropeptides appear to be weak inhibitors of the enzyme (Latajka et al., 2006(Latajka et al., , 2008. In this paper, we synthesized series of structurally variable esters of glycyl Z dehydrophenylalanine and its analogs. We speculated that the possible binding of the aromatic part of the inhibitor within S2 pocket of the enzyme might result in reaction between active ester (allyl, propargyl or glycidyl) with thiol moiety of the active-site cysteine. Unfortunately, obtained compounds exerted moderate inhibitory activity acting as competitive inhibitors. More likely this results from different than expected binding mode of these compounds.

Synthesis of inhibitors
Esters of dehydropeptides have been synthesized using classical methods of peptide chemistry. The synthetic schemes are outlined in Figs. 1 and 2. As seen from the figures for each group of esters, specific method of their preparation should be elaborated. Direct esterification of Boc-Gly-DAla with DMTMM (Kunishima et al., 1999) as coupling agent gave non-satisfactory results (30 % of yield). Far better results for esterification of Boc-Gly-DAla were obtained via nucleophilic substitution of alkyl halides with dipeptide cesium salts (Fig. 1). This method gives product with yield 91 %. Glycidyl esters seem to be more interesting as inhibitors of cathepsin since they posses oxirane ring, which is known to react preferably with the enzyme active-site cysteine. In order to prepare these esters, two standard methods, both basing on the activation of carboxylic moiety, have been elaborated (Fig. 2). Unfortunately, the reaction afforded inseparable mixture of glycidyl ester and some products of oxirane ring opening. Efforts to remove Boc protection by trifluoroacetic acid were unsuccessful and gave even more complex mixture of products, whereas using hydrogen chloride in methanol we were able to isolate 3-chloro-2-hydroxypropyl ester of Boc-Gly-D Z Phe.

Inhibitory studies
Inhibitory activities of the synthesized esters are collected in Table 1 and compared to action of methyl L-phenylalanyl-L- Gly-DAlaOPrgÁTos 320 ± 20 Gly-Z DPheOPrgÁTFA 33 ± 2 (S)Phe-DAlaOPrgÁTos 86 ± 4 Gly-D Z PheO-CH 2 CH(OH)CH 2 ClÁHCl 5.5 ± 0.5 NI-no inhibition up to 1245 mM Fig. 1 Synthesis of dehydrodipeptide methyl, ethyl, isopropyl, allyl and propargyl esters alaninate (Phe-AlaOMe). All the compounds appeared to be competitive inhibitors, as shown in Fig. 3 for Gly-Z DPheOAll trifluoroacetate as a representative example. The most active appeared to be Gly-D Z Phe-OCH 2 CH(OH)CH 2 ClÁHCl, Gly-Z DPheOAllÁTFA and (S)Phe-DAlaOAllÁTos, which inhibitory constants were in micromolar range. Quite  interesting, six of the peptides-Gly-Z DPheOPrgÁTFA, Phe-DAlaOMeÁTos, Phe-DAlaOEtÁTos, (S)Phe-DAlaOPr i ÁTos, (S)Phe-DAlaOAllÁTos and (S)Phe-DAlaOPrgÁTos-inhibit cathepsin C according to slow-binding mechanism. This mechanism is of B type and considers conformational rearrangement of inhibitor after binding to the enzyme (Pawełczak and Hurek, 2014). From the data shown in Table 1, it is also not possible to derive clear-cut structureactivity relationship. Contrary to recent studies on the structural requirements for the specific substrates for cathepsin C (Poręba et al., 2014), introduction of N-terminal phenylalanine into peptide chain results in elevation of affinity of Phe-DAlaOMeÁTos if compared with Gly-DAlaOMeÁTos. This suggests that both dipeptide and dehydrodipeptide esters are bound differently than synthetic substrate of this enzyme. Therefore, simple studies on their presumable binding using AutoDock program had been undertaken.

Molecular modeling
Simple molecular modeling using AutoDock has shown that dehydrodipeptide esters are bound at the surface of the enzyme in a non-typical manner. Their phenyl rings are not, as expected, submerged in the cathepsin C cavity responsible for binding aromatic fragments of the substrates and inhibitors but are rather placed at the surface of the enzyme. The most probable binding mode of Gly-Z DPheOAll is shown in Fig. 4. As seen from this figure, allylic double bond of the inhibitor, albeit directed toward cathepsin C active-site cysteine 234, is too far away from thiol moiety (7-9 Å ) to form a covalent adduct. This non-typical pattern of binding of dehydrodipeptide esters found from calculations well explains moderate inhibitory activity of these compounds.

Conclusions
Synthesis of esters of dehydropeptides is not an easy task and requires the choice of specific method tailored to each case. Esters of dehydrodipeptides containing C-terminal dehydroalanine or (Z)-dehydrophenylalanine appeared to be moderate or weak inhibitors of cathepsin C. As suggested by molecular modeling, they are bound rather on the surface of the enzyme than inside of the binding cavities of the enzyme.