Proteasome inhibitors

Abstract

The present invention relates to a compound of formula (I), wherein X is C═O, C═S or B—OH; Y is an electrophile and Z is a leaving group, or Y═Z is an electrophile; R.sup.1 comprises or consists of (a) (i) a first group binding to a proteolytic site of a proteasome, the first group being bound to X; and (ii) optionally a second group enhancing delivery; or (b) a group binding between subunits β1 and β2 of a proteasome; R.sup.2 and R.sup.3 are independently selected from H, methyl, methoxy, ethyl, ethenyl, ethynyl and cyano, wherein methyl and ethyl may be substituted with OH or halogen. ##STR00001##

Claims

1. A compound of formula (I) ##STR00020## wherein X is C═O, C═S or B—OH; Y is an electrophile and Z is a leaving group, or Y═Z is an electrophile; R.sup.1 is a peptidic group, wherein said peptidic group consists of three α-amino acids and wherein (a) the N-terminal amino acid is selected from Ser(OMe), Leu, Phe and Ala; the middle amino acid is selected from Ser(OMe), Leu, Phe and Ala; and/or the C-terminal amino acid is attached to X and is a truncated amino acid residue that lacks a carbonyl group, wherein the truncated amino acid residue is based upon an amino acid selected from Phe, Tyr, Leu, Ser(OMe) and Ala; or (b) said peptidic group consists of Ser(OMe)-Ser(OMe)-Phe, Leu-Leu-Tyr or Ala-Ala-Ala R.sup.2 and R.sup.3 are independently selected from H, methyl, methoxy, ethyl, ethenyl, ethynyl and cyano, wherein methyl and ethyl may be substituted with OH or halogen.

2. The compound of claim 1, wherein Y═Z is (a) CH═O, CH.sub.2—I, CH.sub.2—Br, CH.sub.2—Cl, CH.sub.2—OPO(OH).sub.2, CH.sub.2-OTs, CO—NHS or CH═CH.sub.2, wherein OTs is p-toluene sulfonyloxy and NHS is N-oxy-succinimide; or (b) O—I, O—Br, O—Cl, S—I, S—Br or S—I.

3. The compound of claim 1, wherein R.sup.2 and R.sup.3 are identical.

4. The compound of claim 1, wherein X is C═O and Y═Z is CH═O or CO—NHS.

5. A method of inhibiting a proteasome, said method comprising bringing into contact a proteasome and a compound as defined in claim 1.

6. A method of treating, ameliorating or preventing cancer, an autoimmune disease, muscular dystrophy, emphysema, or cachexia accompanying cancer or AIDS, comprising administering a compound as defined in claim 1 to a subject in need thereof.

7. The method of claim 6, wherein (a) said cancer is a lymphoid malignancy, wherein the lymphoid malignancy is multiple myeloma (MM) or non-Hodgkin lymphoma, wherein the non-Hodgkin lymphoma is a B-cell lymphoma selected from mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), and Waldenström macroglobulinaemia; or (b) said autoimmune disease is rheumatoid arthritis, systemic lupus erythematosus, Sjörgen's syndrome or scleroderma.

8. The compound of claim 3, wherein R.sup.2 and R.sup.3 are selected from methyl, methoxy, and —CH.sub.2OH.

Description

(1) The figures show:

(2) FIG. 1: Chemical structures of representative members of the epoxyketone class of proteasome inhibitors. Shown are the natural products epoxomicin (top right) and dihydroeponemycin (bottom right), which were isolated from bacterial strains. Also depicted are synthetic derivative of the parent natural product molecules. Carfilzomib (top left) was recently approved for the treatment of multiple myeloma and is currently in clinical trials for the treatment of other cancers. Oprozomib is currently in clinical trials as an orally available multi-potent inhibitor for the treatment of several cancers.

(3) FIG. 2: Crystal structures of co-crystal structures of human 20S proteasomes in complex with epoxyketone inhibitors. 2Fo-Fc electron density maps are shown contoured at 1.5σ for the proteasome active site inhibited with carfilzomib (top left), oprozomib (bottom left), epoxomicin (top right) and dihydroeponemycin (bottom right). Note that the electron density map reveals density for an additional carbon atom. In all cases, the modeled cyclic ring structure is consistent with a 1,4-oxazepane structure formed by the reaction of epoxyketone inhibitors with the active site catalytic threonine side chain.

(4) FIG. 3: Confirmation that a 7-membered 1,4-oxazepane ring structure is formed upon inhibition of proteasomes with epoxyketone inhibitors. Shown are the chemical structures of Oprozomib (top left) and dihydroeponemycin (top right). The oval highlights that Oprozomib contains a methyl group and dihydroeponemycin a methanolic group in the carbon atom in α-position to the ketone, respectively. The bottom panels illustrate the β5 active site inhibited with Oprozomib (bottom left) and Dihydroeponemycin (bottom right), along with an omit map contoured at 46 for the inhibitor, the cyclic linkage and β5Thr2. The main chain segments of β5 residues 2, 19-21, 33, 45-50, 129-131, 169,170 and β6 125,126 are indicated along with the β5 side chains of Thr2, Thr23, Lys33, Ser130 and the side chains of β6 Asp125, Pro126 as sticks. Dashed lines signify hydrogen bonds (≤3.2 Å distance). In the respective bottom right panels close-up views of the inhibitor-Thr1 linkages are shown along with an omit map contoured at 66. Note that the electron density map does not support a chiral center in the case of epoxomicin, which contains a methyl and a methanolic group, which would be a consequence of the proposed mechanism for 1,4-morpholine ring formation (bottom left) (Groll et al., loc. cit.). The dihydroeponemycin structure does not reveal electron density consistent with the presence of two methanolic groups, yet again disproving 1,4-morpholine ring formation (bottom right).

(5) FIG. 4: Attempts to model a 1,4-morpholine ring structure in the linkage occurring between epoxyketone inhibitor and the proteasome active site Threonine1. Shown is an overlay of the proteasome active site (main chain) inhibited by a 1,4-oxazepane linkage and the attempted 1,4-morpholine linkage. The attempted 1,4-morpholine refinement reveals that a van-der-Waals clash occurs with the methyl group of the inhibitor and the main chain carbonyl atoms of R19 and Y169. Additionally, strong negative difference density at 5 sigma is visible at the C5 methanol atom of the 1,4-morpholine linkage. Positive densities are visible at positions 4 and 5 of the 1,4-oxazepane linkage. These findings entirely exclude formation of 1,4-morpholine ring formation in the epoxyketone inhibited proteasome active site.

(6) FIG. 5: Enzymatic characterization of proteasome catalytic activity. Measurements were performed as described in methods. The left panel depicts a typical kinetic experiment, where the increase in fluorescence signal of the AMC released by proteolytic cleavage is plotted against time. The left window signifies an activation phase in enzymatic peptide cleavage, whereas the right window represents the steady-state phase. In the right panel the specific activities for the human 20S proteasome purified by our method are indicated for the pre-steady-state (prior to activation) or and the steady-state phases, respectively. The equation shown in the right panel was used to perform fits against the experimental data. F.sub.0 designates the initial fluorescence, ΔF.sub.ss: fluorescence increase in the steady-state part of the measurement, k.sub.act: the rate constant. The exponential term with the rate constant k.sub.at is used to describe the activation phase of the reaction.

(7) The examples illustrate the invention.

EXAMPLE 1

(8) Crystallography

(9) Initial phases for human 20S proteasomes were determined by molecular replacement using the murine 20S structure (PDB ID: 3UNE). The model was then optimized by several rounds of interactive manual model building in Coot and refinement in Refmac5. The obtained structures display excellent stereochemistry with typical values for R.sub.work=18% and R.sub.free=21%, reveal superb electron densities for all 6724 residues and reveal several ligands as present in buffers used for purification and crystallization.

(10) Using a dataset to 1.8 Å resolution, we created a reference model. With the availability of this excellent model for the human 20S proteasome, now structure determination takes minutes by automated refinement of the reference model against integrated and scaled X-ray data from related crystals. Bound ligands can then be rapidly identified in difference density maps and modeled interactively in Coot.

(11) Native crystals were soaked with the epoxyketone inhibitors shown in FIG. 1. Co-crystal structures of human 20S proteasomes in complex with these inhibitors at resolutions between 1.9-2.1 Å (0.3-0.5 Å better resolution than presently available structures) were solved. Surprisingly, after refinement of these co-crystal structures, which allow the modelling of all parts of human 20S proteasomes at atomic resolution, it was not possible to visualize the presumed 1,4-morpholine ring structure in the inhibited state. The electron density maps in the new epoxyketone/human 20S proteasome co-crystal structures in all cases did not agree with the formation of a 1,4-morpholine 6-ring. Instead, density for an additional atom became clearly visible, which is consistent with a 7-ring structure. In fact, modelling the cyclic molecule visible in the inhibited state formed by the reaction of the epoxyketone inhibitors with the active site catalytic threonine amino acid revealed that it represents a 1,4-oxazepane 7-ring structure (FIG. 2).

(12) The observation of a 1,4-oxazepane structure contrasts with previously presumed chemical inhibition mechanisms of 20S proteasomes by epoxyketones. Therefore, a control experiment was performed to ensure that the observation of this 7-membered ring structure is true and the modelling of the 1,4-oxazepane structure in the inhibited proteasome active site is justified. Additionally, this control experiment should provide insight into the chemical inhibition mechanism by which the 7-ring structure is formed upon proteasome inhibition. For this control experiment, the co-crystal structures determined using the epoxyketone inhibitors epoxomicin and dihydroeponemycin were compared. Epoxomicin contains an epoxide group with a methyl ligand at the carbon atom, where the nucleophilic attack is presumed to occur in order to form the presumed 1,4-morpholine inhibited ring structure (FIG. 3, top left). Additionally, after formation of the presumed 1,4-morpholine structure ring opening of the epoxide at the carbon atom in α-position to the ketone should yield a stereo center, which contains both a methyl and a methanolic group (FIG. 3, right panel) (Groll et al., loc. cit.). Dihydroeponemycin in contrast already contains a methanolic ligand at the carbon atom, where the nucleophilic attack is presumed to occur in order to form the proposed 1,4-morpholine inhibited ring structure (FIG. 3, top right). As a consequence, after formation of the proposed 1,4-morpholine structure ring opening of the epoxide at the carbon atom in α-position to the ketone should yield a non-chiral center, which contains two methanolic groups. At the resolutions at which both co-crystal structures were determined (1.9 and 2.0 Å), it is possible to verify if this is the case. Surprisingly however, the electron density maps of both structures revealed that the inhibited state is formed by the covalent attack of the electrophilic carbon atom in β-position to the ketone (FIG. 3, bottom panels), excluding the possibility that a 1,4-morpholine ring structure is formed in the inhibited state by the absence of electron density compatible with a methanolic group in the case of epoxomicin and two methanolic groups in dihydroeponemycin.

(13) Attempts to model the electron density in the active site of the dihydroeponemycin complex by a 1,4-morpholine ring structure linkage were unsuccessful (FIG. 4). 1,4-morpholine linkage refinement resulted in a severely distorted molecular geometry characterized by the elongation of the N4-carbon bonds by 0.1-0.2 Å, shortening of the C5-alcohol carbon bond by 0.1 Å and deviation of the methyl-C5-alcohol bond angle by −20 degrees off the expected value. Additionally, the methyl carbon exhibited a van-der-Waals distance of 3.0 Å to the Arginine 19 and Tyrosine 169 main chain oxygen atoms, which is too close. Moreover, strong negative difference density peaks in difference maps contoured at 5 sigma levels at the C5 methanol oxygen of the 1,4-morpholine ring model, as well as positive density peaks contoured at 4.5 sigma levels close to positions 4 and 5 of our 1,4-oxazepane ring model remained after this 1,4-morpholine ring refinement. In contrast, no residual difference (neither negative nor positive) density was present in the refined 1,4-oxazepane linkage in density maps contoured above 2.3 sigma.

EXAMPLE 2

(14) Synthesis

(15) ##STR00013##

(16) ##STR00014##

(17) ##STR00015##

(18) A) To a solution of N-Boc-L-Phenylalanine (1) in CH.sub.2Cl.sub.2 is added a EDC (1 eq.), DMAP (1 eq.) and Meldrum's acid (1 eq.). The reaction is stirred at room temperature for 17 hours, then poured into 1 M HCl. The layers are separated and the aqueous layer is extracted three times with CH.sub.2Cl.sub.2. The organic layers are combined, washed with brine, dried over MgSO.sub.4 and concentrated in vacuum. The residue is heated to 80° C. in toluene and after the addition of benzyl alcohol (1 eq.) for 4 hours. The solvent is the removed in vacuum and the residue purified by flash chromatography and elution with ethyl acetate to yield (2).

(19) B) A solution of (2), methyl iodide (3 eq.) and potassium carbonate (2 eq.) in acetone is heated under reflux for 17 hours. 2 volume equivalents of water are added and the resulting mixture extracted three times with 2 volume equivalents ethyl acetate. The organics are combined, dried over MgSO.sub.4 and concentrated in vacuum. The residue is purified by preparative HPLC in 0.1% formic acid in water, with a gradient to 0.1% formic acid in acetonitrile to give (3).

(20) C) A solution of (3) is stirred in a 10% TFA solution in CH.sub.2Cl.sub.2 for 17 hours at room temperature. The solvent is subsequently removed in vacuum to yield (4).

(21) D) To (4) (1 eq.) in CH.sub.2Cl.sub.2 and triethylamine (0.001 eq.) is added (19) (1 eq.), HOBT (0.2 eq.) and EDC (2 eq.). The solution is stirred at 25° C. for 24 hours and then washed three times with saturated sodium hydrogencarbonate solution, once with deionized water and brine each, and the organic layer is dried over MgSO.sub.4. The solvent is removed in vacuum and the residue purified by flash chromatography eluting with 1:1 ethyl acetate:n-hexane to yield (5).

(22) E) (5) is stirred in methanol containing 10% Pd/C under a hydrogen atmosphere (1 atm). After 2 hours the mixture is filtered through celite and the solvent removed in vacuo to the product (6).

(23) F) To (6) in CH.sub.2Cl.sub.2 is added, EDC (2 eq.) and N-Hydroxysuccinimide (2 eq.) and the mixture stirred for 2 hours. The solvent is then removed in vacuo and the residue purified by flash chromatography, eluting with 1:5 ethyl acetate:n-hexane to yield (7).

(24) G) To a stirred solution of Boc-L-Phenylalanine in CH.sub.2Cl.sub.2 is added O,N-dimethylhydroxylamine hydrochloride (1 eq.), triethylamien (2 eq.) and BOP (1 eq.). After 3.5 hours the solution is diluted 4-fold with CH.sub.2Cl.sub.2 and washed three times with 3 M HCL, three times with saturated sodium hydrogencarbonate and three times with brine. The organic layer is then dried over MgSO.sub.4, the solvent removed in vacuo and the residue purified by flash chromatography with 1:3 ethyl acetate:n-hexane to yield (8).

(25) H) The Weinreb amide (8) in THE under an Argon atmosphere is cooled to 0° C. and 1,3-Dioxacyclopentyl-2-MgBromide (5 eq.) in THE added dropwise. The reaction is allowed to reach 25° C. and after 4 hours of stirring, is quenched with 1 M HCl forming a precipitate. The precipitate is removed by filtration and washed three times with ethyl acetate. The combined organics are then washed with brine, dried over MgSO.sub.4 and the solvent removed in vacuum. The residue is then purified by flash chromatography using 1:4 ethyl acetate: n-hexane as an eluent to yield (9).

(26) I) A solution of (9) is stirred in a 10% TFA solution in CH.sub.2Cl.sub.2 for 17 hours at room temperature. The solvent is subsequently removed in vacuum to yield (10).

(27) J) To (10) (1 eq.) in CH.sub.2Cl.sub.2 and triethylamine (0.001 eq.) is added (19) (1 eq.), HOBT (0.2 eq.) and EDC (2 eq.). The solution is stirred at 25° C. for 24 hours and then washed three times with saturated sodium hydrogencarbonate solution, once with deionized water and brine each, and the organic layer is dried over MgSO.sub.4. The solvent is removed in vacuo and the residue purified by flash chromatography eluting with 1:1 ethyl acetate:n-hexane to yield (11).

(28) K) A solution of (11) and p-TsOH (0.1 eq.) in acetaldehyde (0.5 eq.) is stirred under an argon atmosphere at 15° C. for 23 hours. The solvent is then removed in vacuum and the residue purified by flash chromatography, eluting with 1:5 ethyl acetate:n-hexane to yield (12).

(29) L) To a solution of Boc-methylserine (13) in DCM (Dichlormethane) TEA (Triethylamine) and DMAP (4-Dimethylaminopyridine) are added. The resulting solution is cooled to −5° C., and benzyl chloroformate is then slowly added via an addition funnel under an atmosphere of argon. The reaction is kept at the same temperature for 3 h and then diluted with brine. The layers are separated, and the aqueous layer is extracted with DCM. The organic layers are combined and dried over Na.sub.2SO.sub.4. The Na.sub.2SO.sub.4 is removed by filtration, and the volatiles are removed under reduced pressure. The resulting residue is purified by flash chromatography using a mixture of hexane and ethyl acetate to provide intermediate (14) as white solid.

(30) M) To a 0° C. solution of intermediate (14) in DCM TFA (Trifluoroacetic acid) is added slowly via a funnel. The reaction is kept at the same temperature for 1 h, concentrated, and dried under high vacuum overnight. The resulting residual TFA salt (15) is used in the next step without further purification.

(31) N) To a −5° C. mixture of aforementioned TFA salt (15), Boc-methylserine (13), HOBt (Hydroxybenzotriazole), and HBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate) in THE (Tetrahydrofuran) (600 mL) is added DIEA slowly via an addition funnel. The reaction is kept at the same temperature for 4 h, followed by dilution with EtOAc (Ethylacetate) and brine. The layers are separated, and the aqueous layer is extracted with EtOAc (2×300 mL). The organic layers are combined and dried over Na.sub.2SO.sub.4. The Na.sub.2SO.sub.4 is removed by filtration, and the volatiles are removed under reduced pressure. The resulting residue is purified by flash chromatography using a mixture of hexane and ethyl acetate to provide dipeptide (16) as white solid.

(32) O) To a 0° C. solution of aforementioned intermediate (16) in DCM was added TFA slowly via an addition funnel. The reaction is kept at the same temperature for 2 h, concentrated, and dried under high vacuum overnight. The resulting residual TFA salt (17) is used in the next step without further purification.

(33) P) To a −5° C. mixture of TFA salt (17), 2-methylthiazole-5-carboxylic acid, HOBt, and HBTU in THE is added DIEA slowly. The reaction is kept at the same temperature for 4 h and then diluted with EtOAc and brine. The layers are separated, and the aqueous layer is extracted with EtOAc. The organic layers are combined and dried over Na.sub.2SO.sub.4. The Na.sub.2SO.sub.4 is removed by filtration, and the volatiles are removed under reduced pressure. The resulting residue is purified by flash chromatography using a mixture of hexane and ethyl acetate to provide benzyl ester (18) as white solid.

(34) Q) (18) is stirred in methanol containing 10% Pd/C under a hydrogen atmosphere (1 atm). After 2 hours the mixture is filtered through celite and the solvent removed in vacuum to the product (19).

EXAMPLE 3

(35) Proteasome Assays

(36) Activity Measurement

(37) Preferred In Vitro Assay.

(38) All kinetic measurements were performed using a FluoroMax®-4 fluorescence spectrophotometer (Horiba Scientific). Succinyl-Leucine-Leucine-Valine-Tyrosine-7-amido-4-methylcoumarin (Suc-LLVY-AMC (SEQ ID NO: 1), Bachem) was used as substrate to determine the chymotryptic-like activity of the β5 catalytic active site of the human 20S proteasome (R. L. Stein, F. Melandri, L. Dick, Kinetic characterization of the chymotryptic activity of the 20S proteasome. Biochemistry 35, 3899-3908 (1996)). The fluorescence emission of hydrolyzed AMC was continuously monitored at 460 nm (λex=380 nM). The reaction temperature was kept at 37° C. for all measurements and the reaction buffer for enzymatic assays specified in Table S1 was used. Suc-LLVY-AMC (SEQ ID NO: 1) and inhibitors (such as Oprozomib, Dihydroeponemycin, Z-LLY-Ketoaldehyde; “Suc” designating Succinyl and “Z” designating Benzyloxycarbonyl) were dissolved in DMSO and stored at −80° C. until usage. The DMSO concentration did not exceed 2% (v/v) in any measurement. For kinetic characterization of Suc-LLVY-AMC (SEQ ID NO: 1) conversion, 0.035 mg/mL (50 nM) human 20S proteasome in reaction buffer was pre-incubated for 3 minutes at 37° C. The reaction was started by the addition of substrate and the fluorescence signal was measured continuously. For determination of the first-order rate constant of inhibition of the respective inhibitors, the reaction mixture containing reaction buffer, 150 μM substrate and either Oprozomib (50 μM), Dihydroeponemycin (50 μM) or Z-LLY-Ketoaldehyde (15 μM) were pre-incubated at 37° C. for 3 minutes. The reaction was then started by the addition of human 20S proteasome to a final concentration 50 nM. The fluorescence signal was measured continuously.

(39) Data were analyzed and fitted with OriginPro 9.1 and KaleidaGraph 4.03. The equation shown in FIG. 5 was used to analyze the chymotryptic-like catalytic activity and catalytic activation of the 20S proteasome. For the determination of the first-order inactivation rate constants, equations were used that contained either two exponential terms in case of Z-LLY-Ketoaldehyde, or two exponential terms plus a linear term for epoxyketones (Oprozomib and Dihydroeponemycin). The first of the two exponential terms accounts for the catalytic activation, whereas the second exponential term represents the catalytic inactivation by inhibitory action. The linear term in case of the epoxyketones was used to account for the residual activity of the proteasome after inactivation.

(40) Alternative Kinetic Assay.

(41) Time point measurements of the activity assays are performed to acquire an initial tendency of inhibitor binding. Different concentrations of 20S proteasome are used for each active site: 0.05 mg/ml for CL (Chymotryptic-like activity) and PGPH (Peptidyl-glutamyl peptide hydrolyzing activity) and 0.075 mg/ml for TL (Tryptic-like activity). The final reaction volume is 30 μl/well. A total number of five repetitions are performed to obtain root mean square deviation (RMSD), including a blank and a 100% initial activity reaction. The steps are as follows.

(42) (1) A master mix is prepared: Respective amount proteasome (according to PGPH, TL, or CL activity determination).

(43) (2) Eppendorf tubes are prepared with the amount of the respective inhibitor to be analysed; e.g. 500 μM concentration of the ligand in 30 μl is 1.5 μl per assay of a 10 mM inhibitor stock solution.

(44) (3) 28.5 μl of the master mix is added to each Eppendorf tube. This solution is incubated for 15 min at room temperature and transferred to the respective wells of the 96 well plates.

(45) (4) Following incubation, 1 μl of a 7.5 mM stock solution of substrate for the caspase, chymotryptic or tryptic site is added, giving a final substrate concentration of 250 μM. The plate is centrifuged and incubated at RT for 1 h.

(46) (5) 300 μl of buffer is added to the reaction and the remaining proteasome activity is subsequently recorded by fluorescence at Ex (Excitation wavelength) 360 nm-Em (Emission wavelength) 460 nm.

(47) (6) The remaining activity is calculated using the blank and 100% initial activity.

(48) Once proteasome inhibition is observed through the time point measurements, the half maximal inhibitory concentration (IC.sub.50) measurements can be performed. The percentage of the remaining activities is then plotted against the log concentration of the respective inhibitor. The obtained data are fitted with a conventional statistical program, for example as described in: Groll M, Gallastegui N, Maréchal X, et al (2010) 20S Proteasome Inhibition: Designing Non-Covalent Linear Peptide Mimics of the Natural Product TMC-95A. ChemMedChem 5:1701-1705.

(49) Reference for Assay: A description of an activity assay can be found in Gallastegui, N., & Groll, M. (2012). Analysing Properties of Proteasome Inhibitors Using Kinetic and X-Ray Crystallographic Studies. In Methods in Molecular Biology (Vol. 832, pp. 373-390). doi:10.1007/978-1-61779-474-2_26.

(50) Cellular Assays.

(51) A further example is the Z-Sensor Proteasome assay from Takara/Clontech. ZsProSensor-1 is a proteasome-sensitive fluorescent reporter. It is a fusion of a bright green fluorescent protein (Exmax=496 nm, Emmax=506 nm) with a degradation domain which targets the protein for rapid degradation by the proteasome. The cells emit green fluorescence when there is a drop in proteasome activity.

(52) Alternative Methods:

(53) Proteasome-Glo™ Chymotrypsin-Like, Trypsin-Like and Caspase-Like Cell-Based Assays (Promega).

(54) GFP-Assay: Bence N, Bennett E, Kopito R, Deshaies R. Application and analysis of the GFP(u) family of ubiquitin-proteasome system reporters. Ubiquitin and Protein Degradation, Pt B. 2005; 399:481-490.

(55) A cellular assay for the immunoproteasome is the following:

(56) Fluorogenic in vitro assay (Immunoproteasome): (Basler, M., & Groettrup, M. (2012).

(57) Immunoproteasome-Specific Inhibitors and Their Application. In Methods in Molecular Biology (Vol. 832, pp. 391-401). doi:10.1007/978-1-61779-474-2_27)

(58) In order to test whether your IP inhibitor is cell permeable, the following method based on proteasome immuno-precipitation and in vitro activity assay can be used. The steps are as follows.

(59) (1) Incubate cells for 2 h with desired concentration of IP inhibitors in cell culture media at 37° C. We normally use mouse splenocytes (one spleen per sample). As control, use an equal number of cells without inhibitor.

(60) (2) Wash cells three times with PBS to remove unbound inhibitor.

(61) (3) Lyse cells in 500 μl lysis buffer and incubate for 20 min on ice.

(62) (4) Centrifuge the lysates for 10 min at 20,800×g to remove debris.

(63) (5) Discard pellet and add 3 μl of polyclonal rabbit-anti-mouse proteasome antibody and 50 μl protein A microbeads to the supernatant and incubate for 30 min on ice.

(64) (6) Insert μ column into magnet.

(65) (7) Equilibrate μ column with 1 ml NET-TON buffer.

(66) (8) Load lysate on μ column and discard flow through.

(67) (9) Wash column twice with 1 ml NET-TON buffer and three times with NET-T buffer.

(68) (10) Add 50 μl of a fluorogenic substrate and incubate column for 30 min at 37° C.

(69) (11) Add 200 μl lysis buffer and collect eluate.

(70) (12) Measure the fluorescence in 100 μl of the eluate (96-well plate, flat bottom, black). The fluorescence in the eluate corresponds to the activity of the retained proteasome in the column.

(71) LacZ assay (Immunoproteasome); see Basler, M., & Groettrup, M. (2012). Immunoproteasome-Specific Inhibitors and Their Application. In Methods in Molecular Biology (Vol. 832, pp. 391-401). doi:10.1007/978-1-61779-474-2_27.

(72) Numerous MHC-I restricted CD8+ T-cell epitopes have been described to be dependent on IP subunits. Investigating the processing of such T-cell epitopes can test specificity of IP inhibitors. In order to analyse the LMP7-selective inhibitor PR-957, we investigated the male HY-derived CTL-epitope UTY 246-254, which was reported to be LMP7 dependent. Therefore, we treated male splenocytes with PR-957 and detected MHC-I presented UTY 246-254 peptides with the help of UTY 246-254-specific T-cell hybridomas in lacZ assays.

(73) (1) Remove spleen of one male and one female mouse and take up spleen in 5 ml RPMI 10% FCS.

(74) (2) Make a single-cell suspension by pressing spleen through a grid.

(75) (3) Centrifuge cells for 5 min at 347×g and discard supernatant.

(76) (4) Lyse the erythrocytes by resuspending cells in 5 ml pre-warmed 1.66% (w/v) NH 4Cl solution (in 15-ml tubes).

(77) (5) Incubate for 2 min at room temperature.

(78) (6) Fill up to 15 ml with RPMI 10% FCS and centrifuge cells for 5 min at 347×g and discard supernatant.

(79) (7) Wash cells with 15 ml PBS, centrifuge cells for 5 min at 347×g, and discard supernatant.

(80) (8) Take up cells in 5 ml RPMI 10% FCS and count cells using a Neubauer chamber.

(81) (9) Incubate 10 7 splenocytes in 3 ml RPMI 10% FCS per well (6-well tissue culture plate).

(82) (10) Add desired amounts of inhibitor. You need one well of male splenocytes without inhibitor for comparison of untreated and treated samples. For female splenocytes, you only need one well without inhibitor.

(83) (11) Incubate overnight at 37° C.

(84) (12) Harvest splenocytes, wash cells twice with 15 ml PBS, and count splenocytes.

(85) (13) Resuspend cells in RPMI 10% FCS at 10.sup.7/ml.

(86) (14) Use 96-well round-bottom tissue culture plate and add 150 μl per well to wells A1-D1. Make four serial threefold dilutions of splenocytes (100 μl/per well).

(87) (15) Harvest T-cell hybridomas, count, and resuspend in RPMI 10% FCS at 10{circumflex over ( )}6/ml. (We use the UTY 246-254-specific T-cell hybridoma (5).)

(88) (16) Add 100 μl of T-cell hybridomas per well (A1-A4; B1-B4). Add to half of your samples (C1-C4; D1-D4) 100 μl RPMI 10% FCS as background control.

(89) (17) Female splenocytes are used as negative control and untreated male splenocytes as positive control and for comparison. You can make an additional positive control adding synthetic peptide (we use UTY 246-254 peptide at a concentration of 10{circumflex over ( )}-7 M) to female splenocytes.

(90) (18) Incubate o/n at 37° C.

(91) (19) Centrifuge plate at 541×g for 90 s and discard supernatant.

(92) (20) Add 100 μl lacZ buffer and incubate at 37° C.

(93) (21) Measure absorbance at 570/620 nm when colour change is visible (approx. after 1-3 h).

EXAMPLE 4

(94) Alternative Synthesis

(95) Synthesis of Tripeptide 28584

(96) ##STR00016##

(97) The synthesis of the tripeptide 28584 was successfully carried out according to the literature procedure. The formation of the Benzylester 28579 was performed in 4 g scale and 4.6 g product was obtained (>95% purity by .sup.1H-NMR, 98% purity by LC/MS, 82% yield). The following Boc-deprotection of 28579 led to 4.8 g of 28580 in quant. yield and 95% purity by 1H-NMR and 86% purity by LC/MS.

(98) The peptide coupling was performed on a 2 g scale and 1.8 g of 28581 was obtained (95% purity by LC/MS, 90% purity by .sup.1H-NMR, 73% yield).

(99) Synthesis of 28582 was carried out in 1.7 g scale and 1.7 g product (28582) was obtained as TFA salt (quant. yield, 90% purity by .sup.1H-NMR).

(100) After amide formation of 2-Methyl-5-thiazolecarboxylic acid with 28582, 1.5 g of 28583 was obtained (80% yield, 93% purity by .sup.1H-NMR; >95% by LC/MS).

(101) The benzyl ester deprotection of 28583 with 10% Pd/C was performed on 1.5 g scale and 1.0 g 28584 was obtained (79% yield, 95% purity by .sup.1H-NMR and LC/MS). It was found out that the benzyl ester deprotection requires more than catalytic amount 10% Pd/C. For fast deprotection at least equal amount of Pd/C is needed compared to the used amount of 28583.

(102) Synthesis of NHS-Ketoester 28880 and Thioester 29502, 29865

(103) ##STR00017## ##STR00018##

(104) The formation of the Meidrum's acid intermediate 28864 was performed by using N-Boc-L-Leu-OH and EDC/DMAP in DCM as reported in the literature. It was observed that under those reaction conditions product with complex mixture of side products was formed. One major side product was identified by LC/MS as cyclized Leucine derivative. The reaction crude has a purity of approx. 30-40% determined by .sup.1H-NMR.

(105) ##STR00019##

(106) When 28864 was treated with benzyl alcohol for Meldrum's acid building block opening the desired product 28865 was obtained. From 6.7 g reaction 1.7 g of 28865 was obtained (26% yield, 85% purity by .sup.1H-NMR).

(107) The double alkylation of 28865 was carried out by using excess of methyl iodide in the presence of K2CO3 in acetone which led to the formation of 28895 in 30-38% yield. In the end, out of a 3.7 g scale reaction 1.3 g of 28895 was obtained (38% yield, 95% purity by .sup.1H-NMR).

(108) The N-Boc deprotection of 28877 was performed using 10% TFA in DCM and was done as reported in the literature. After work up .sup.1H-NMR clearly showed the mixture of compounds.

(109) Another reaction was carried out and it was found out that the reaction was completed in two hours instead of 17 h which was reported in the literature.

(110) The reaction mixture was concentrated at room temperature.

(111) The deprotection was performed shortly before the amide coupling of 28878. 400 mg reaction of 28877 was carried out and the reaction crude (purity 85% by LC/MS) was used immediately for the next reaction step.

(112) HATU was used as a coupling reagent for the synthesis of 28878 successfully. After work up the .sup.1H-NMR indicated that epimerization was occurred (15% by .sup.1H-NMR). This is corresponding to the reported literature. The synthesis of 28878 was successfully performed in a 100 mg and 300 mg scale by using HATU as coupling reagent. After purification 65 mg (36% yield, purity 95% by LC/MS) and 298 mg (55% yield, purity 95% by LC/MS) product 28878 was obtained. In both cases the presence of epimer was decreased from 15 mol % to 8 mol %.

(113) The synthesis of 28879 was performed using N,N-dicyclohexylcarbodiimide and N-hydroxysuccinimide in THF/DMF as solvent system. By LC/MS the formation of a peak with the right mass was observed.

(114) After precipitation of dicyclohexylurea in DCM and removal of excess of NHS by washing with water, 25 mg were obtained (48% purity by LC/MS). The crude compound was tried to purify by preparative HPLC on reverse phase column. The fractions were extracted by DCM in order to avoid hydrolysis/decarboxylation. 0.7 mg with 93% purity and a peak with the right product mass in LC/MS was obtained. .sup.1H-NMR in CDCl3 provided no clear indication due to the small amount of sample.

(115) For structure elucidation, reaction of 28880 was performed on a 140 mg scale by using N,N-dicyclohexylcarbodiimide and N-hydroxysuccinimide (NHS) in EtOAc/DMF as solvent system. LC/MS analysis showed the formation of a peak with the right product mass. After precipitation of dicyclohexylurea in DCM and removal of excess of NHS by washing with water, 120 mg were obtained (48% purity by LC/MS). The crude was tried to purify by preparative HPLC.

(116) To enhance the coupling efficiency PyBOP was used as a coupling reagent. By using 1.3 eq. of PyBOP, 2.0 eq. DIPEA and 3.0 eq. NHS a very prominent peak with right product mass was observed also in much higher intensity than before.

(117) In order to prove whether the peak with the right product mass observed in LC/MS is an artefact or belongs to the desired compound 28880, it was treated with MeOH to form the methylester of 28880. LC/MS analysis showed that the peak with the mass of the expected NHS-ester was disappeared and the formation of a peak with the right mass of the methylester was observed.

(118) Purification was done by doing two times short plug filter column chromatography. In the first case the reaction mixture was directly filtered through a pad of silica without any solvent removal after the reaction. THE was used as eluent. After concentration of this above mentioned fraction, compound was purified again by short column chromatography using DCM/THF mixture as eluent. 9.8 mg of product was obtained (62% purity by LC/MS). Due to very small amount the purity of product cannot be improved further. However, .sup.1H-NMR is corresponding to the product. For larger scale the purification was improved by optimizing the column chromatography using c-hexane/THF as solvent mixture.

(119) Another 100 mg reaction was performed by using 1.3 eq. of PyBOP, 2.0 eq. DIPEA and 3.0 eq. NHS as coupling conditions. Purification was done by doing short plug filter column chromatography. In the first case the reaction mixture was filtered through a pad of silica without any concentration. THF was used as eluent. After concentration of the above mentioned fraction, compound was purified by column chromatography using c-hexane/THF mixture as eluent. 47 mg of desired title compound 28880 was obtained with purity of 88% by .sup.1H-NMR, containing 13.5 mol % of decarboxylated 28879.

(120) In conclusion, 47 mg of 28880 was delivered successfully in 88% purity by .sup.1H-NMR.