Entinostat-Bortezomib Hybrids against Multiple Myeloma

Although proteasome inhibitors have emerged as the therapeutic backbone of multiple myeloma treatment, patients often relapse and become drug refractory. The combination between proteasome and histone deacetylase inhibitors has shown to be more efficient compared to monotherapy by enhancing the anti-myeloma activity and improving the patient’s lifetime expectancy. Hybrid molecules, combining two drugs/pharmacophores in a single molecular entity, offer improved effectiveness by modulating more than one target and circumventing differences in the pharmacokinetic and pharmacodynamic profiles, which are the main disadvantages of combination therapy. Therefore, eleven histone deacetylase-proteasome inhibitor hybrids were synthesized, combining pharmacophores of entinostat and bortezomib. Compound 3 displayed the strongest antiproliferative activity with an IC50 value of 9.5 nM in the multiple myeloma cells RPMI 8226, 157.7 nM in the same cell line resistant to bortezomib, and 13.1 nM in a 3D spheroid model containing multiple myeloma and mesenchymal stem cells. Moreover, the compound inhibited 33% of histone deacetylase activity when RPMI 8226 cells were treated for 8 h at 10 µM. It also inhibited the proteasome activity with an IC50 value of 23.6 nM.


Introduction
Multiple myeloma (MM) is a hematological malignancy that occurs with an uncontrolled growth of malignant plasma cells, predominantly residing in the bone marrow [1,2]. It is the second most important blood disease following non-Hodgkin lymphoma, representing 10% among the hematological syndromes and accounting for 1% of all types of cancer [2]. It is generally preceded by asymptomatic phases termed monoclonal gammopathy of undetermined significance (MGUS) and smoldering multiple myeloma (SMM) [1,2]. SMM evolves to MM, which then presents well defined clinical manifestations such as hypercalcemia, kidney injury, anemia, bone lesions and an increased risk of infections [1,2]. Even though significant advances were made in the treatment options, MM patients are characterized by a history of relapses and drug resistance [3]. One of the main classes of anti-MM agents among the therapies of choice are proteasome inhibitors (PIs) [4] with the first-in-class bortezomib (BTZ) [5]. BTZ is a potent, selective and slowly reversible inhibitor of the β1 and β5 subunits of the 26S proteasome, the enzyme complex known as the intracellular machinery for protein degradation [5]. As a consequence, the accumulation of misfolded and unfolded proteins triggers various types of cellular stress responses, followed by the induction of apoptosis, inhibition of cell proliferation and cytotoxicity [5,6]. This also leads to the formation of aggregates, which activate the aggresome-autophagy pathway leading to protein clearing as a mechanism of resistance to treatment [5]. Histone deacetylase 3 (HDAC3) is another important target for the inhibition of MM proliferation.
HDAC3 knockdown significantly induced apoptosis, and inhibitors showed synergistic effects with BTZ [7,8]. Therefore, combinations between PIs [9] and HDAC inhibitors (HDACi) [10,11], became a promising therapeutic strategy to address drug resistance in MM. Among the HDACi, entinostat (Ent), a class I selective HDACi, displayed synergistic antiproliferative activity when combined with BTZ in MM cells, including those resistant to BTZ [8,12]. Furthermore, the combination of the two inhibitors led to cell cycle arrest in G2/M phase and enhanced apoptosis [8,12].
The strategy of "hybrid molecules" has recently drawn increased attention in drug design and development, especially for the treatment of multifactorial diseases such as cancer [13]. This approach combines in one molecular entity the key structural features of various drugs, which are essential for their biological activity. This may result in a potent and selective hybrid molecule, able to activate two or more mechanisms of action toward cancer cell growth inhibition [14]. Compared to combination therapies, which involve the co-administration of multiple drugs, a hybrid molecule may provide certain advantages, such as improved pharmacokinetic properties, reduced side effects and a decrease in the development of drug resistance. Additionally, they may provide reduced costs and improved patient compliance, which is also of high importance [15].
As a result of our current interest in the development of novel therapeutic strategies to overcome resistance in MM, a library of 11 HDACi-BTZ hybrids, suitable for structureactivity relationship studies were synthesized ( Figure 1). Based on the promising results of Zhou et al. [11] regarding similar dual inhibitors, the key pharmacophore structure of BTZ was kept. Thus, the dipeptide L-phenylalanine-L-boroleucine, including the pinanediol chiral auxiliary (as prodrug) was combined with various spacers and a zinc-binding group HDACi pharmacophore, such as the one of Ent. Moreover, an interesting set of compounds (8,9,11) was prepared, replacing the L-phenylalanine residue of the BTZ counterpart with melphalan, which combines the phenylalanine structure with a mustard type cytotoxic warhead.

Antiproliferative Activity of Compounds 1-11 in Various MM Models
The antiproliferative potential of the compounds was assessed by measuring their metabolic activity in various MM models, and it was compared to BTZ and Ent. All of the compounds inhibited RPMI 8226 cell proliferation in a dose-dependent manner, except for compound 10, which did not show any activity below 2000 nM (Table 1). This loss of activity can be attributed to the absence of phenylalanine residue in the BTZ part of the molecule.
Compound 3, which incorporates the full structure of BTZ in its skeleton, was the most active with an IC50 value of 9.5 ± 1.4 nM which is only slightly inferior to BTZ (IC50 = 2.1 ± 0.4 nM). On the other hand, compound 2, where the pyrazine-2,5-dicarboxylic acid moiety of the BTZ part was replaced by a terephthalic acid, was ca. 4 times less active with an IC50 value of 37.1 ± 2.9 nM. A similar decrease in activity was observed for the hybrid 5 (IC50 = 25.2 ± 2.2 nM), which, compared to compound 3, has one p-(aminomethyl)benzoyl group inserted between the pyrazine-2,5-dicarboxylic acid moiety and the 1,2-dianiline ring. The same structural change was applied to compound 2 to provide compound 4. This led to a loss of activity (IC50 = 802.4 ± 72.7 nM). A similar activity was observed for compound 1 (IC50 = 677.7 ± 42.5 nM), in which the terephthalyl moiety of hybrid 4 was replaced by a succinyl one. Moreover, compound 11, which, compared to compound 3, has the phenylalanine moiety replaced by melphalan, showed ca 2.4 times less activity than 3. In addition, compounds 8 and 9, bearing melphalan's warhead and having a struc- The synthesis of the hybrid compounds (1-11) is described in Scheme 2 and was achieved by fragment coupling of the BTZ building block part (E5 or E6) and the HDACispacer part (A9 or A10 or B6 or B7 or D4 or D5), using typical amide coupling conditions (TBTU/NMM or PyBOP/DIPEA).

Antiproliferative Activity of Compounds 1-11 in Various MM Models
The antiproliferative potential of the compounds was assessed by measuring their metabolic activity in various MM models, and it was compared to BTZ and Ent. All of the compounds inhibited RPMI 8226 cell proliferation in a dose-dependent manner, except for compound 10, which did not show any activity below 2000 nM (Table 1). This loss of activity can be attributed to the absence of phenylalanine residue in the BTZ part of the molecule.
Compound 3, which incorporates the full structure of BTZ in its skeleton, was the most active with an IC 50 value of 9.5 ± 1.4 nM which is only slightly inferior to BTZ (IC 50 = 2.1 ± 0.4 nM). On the other hand, compound 2, where the pyrazine-2,5-dicarboxylic acid moiety of the BTZ part was replaced by a terephthalic acid, was ca. 4 times less active with an IC 50 value of 37.1 ± 2.9 nM. A similar decrease in activity was observed for the hybrid 5 (IC 50 = 25.2 ± 2.2 nM), which, compared to compound 3, has one p-(aminomethyl)benzoyl group inserted between the pyrazine-2,5-dicarboxylic acid moiety and the 1,2-dianiline ring. The same structural change was applied to compound 2 to provide compound 4. This led to a loss of activity (IC 50 = 802.4 ± 72.7 nM). A similar activity was observed for compound 1 (IC 50 = 677.7 ± 42.5 nM), in which the terephthalyl moiety of hybrid 4 was replaced by a succinyl one. Moreover, compound 11, which, compared to compound 3, has the phenylalanine moiety replaced by melphalan, showed ca 2.4 times less activity than 3. In addition, compounds 8 and 9, bearing melphalan's warhead and having a structural similarity with BTZ, exhibited lower activity than BTZ itself. The above results indicate that the insertion of a cytotoxic warhead to BTZ or hybrid 3 structures did not offer any improvement of activity compared to BTZ. The frequent occurrence of drug resistance in MM patients is one of the reasons for the need to identify new active molecules. Therefore, compounds that showed an IC 50 value below 50 nM in RPMI 8226 cells were tested in RPMI 8226 cells resistant to BTZ (RPMI 8226/BTZ100) to evaluate if the hybrids could overcome PI resistance ( Table 1). The results demonstrated that compounds 2, 3, and 5 were more potent than BTZ against RPMI 8226/BTZ100 cells. Moreover, the increased resistance was different depending on the compounds and corresponded to 3.0, 16.6, and 8.0-fold for compounds 2, 3, and 5, respectively, compared to 116.8-fold for BTZ. Furthermore, some compounds that had similar IC 50 values in RPMI 8226 cells showed considerable differences in their activity in the resistant cells (compounds 2 and 7).
The interactions between MM cells and the bone marrow microenvironment play an important role in MM progression. The culture of a single cell line as a monolayer attached to a plastic surface does not properly represent the tumor complexity and interactions with the extracellular compartment. Therefore, a 3D co-culture spheroid model containing RPMI 8226 cells and mesenchymal stem cells (MSC) in a 5:1 ratio was used to measure the antiproliferative activity of compounds that showed an IC 50 value inferior to 50 nM in RPMI 8226 cells (Table 1). Compound 3 showed the strongest activity with an IC 50 value of 13.1 ± 6.9 nM. For the most active compounds, the activity remained the same or was slightly lower in the spheroids than in RPMI 8226 cells. This showed that those compounds kept their activity in a more complex model.

26S Proteasome Inhibitory Activity of Compounds 1-11
To evaluate whether the proteasome inhibitory activity of BTZ was maintained in the complex molecular structure of the hybrids, each compound was tested for its capacity to inhibit the 26S proteasome in RPMI 8226 cell lysate and the IC 50 was measured when possible ( Table 1). The inhibitory activity of compounds 1, 2, 7, and 8 was similar to BTZ, while compounds 3 and 5 were slightly less active. In general, all of the above-mentioned compounds maintained a very good PI activity in the low nM range, although their antiproliferative activity against RPMI 8226 cells may differ and follow a different order.

HDAC Inhibitory Activity in RPMI 8226 Cells
To evaluate whether the HDAC inhibitory activity of Ent was maintained in the complex molecular structure of the hybrids, the activity of compounds inhibiting RPMI 8226 cell proliferation with an IC 50 value lower than 50 nM was tested in the same cell line using a UHPLC-MS method. Compounds 2, 3, 7, and 11 inhibited 50, 33, 37, and 25%, respectively, of HDAC activity at 10 µM. Ent, on the other hand, inhibited 51% of HDAC activity at 10 µM and showed an IC 50 value of 9.3 ± 0.7 µM. Data indicated that compound 2 and Ent had a similar activity at 10 µM. However, it was not possible to measure the IC 50 value of the hybrids due to poor water solubility.

Drug Combination
The antiproliferative effect of Ent and BTZ in combination at various concentrations was evaluated to determine if there was any synergy between both compounds. Combinations were evaluated using Combenefit data analysis. Ent (20-1600 nM) and BTZ (0.3-20 nM) did not show any synergistic effects when tested around their IC 50 values nor when tested in a 1:1 ratio (0.3-20 nM). However, when the drug combination in a 1:1 ratio was analyzed using the CompuSyn software, Ent and BTZ displayed synergistic effects when combined at 5 nM (CI = 0.59) with 91% cell proliferation inhibition. Despite having a CI < 1 indicating synergy, the IC 50 values between BTZ and the drug combination did not differ significantly (3.3 vs. 3.2 nM), suggesting that the addition of Ent at low doses did not improve the biological activity of BTZ.

General Methods
All solvents were dried and purified according to standard procedures prior to use. When required, reactions were performed under inert atmosphere (Ar) in pre-flamed glassware. Anhydrous Na 2 SO 4 was used for drying solutions, and the solvents were then routinely removed at ca. 40 • C under reduced pressure using a rotary vacuum evaporator. All reagents employed in the present work were commercially available and used without further purification. Flash column chromatography (FCC) was performed on silica gel (70-230 and 230-400 mesh, Merck, Darmstadt, Germany) and analytical thin layer chromatography (TLC) on silica gel 60-F254 precoated aluminum foils (0.2 mm film, Merck, Germany). Spots on the TLC plates were visualized with UV light at 254 nm and using ninhydrin solution. 1 H NMR spectra were recorded in CDCl 3 at 600.13 MHz and 13 C spectra at 150.9 MHz on a Bruker AVANCEIII HD spectrometer. Chemical shifts (δ) are indicated in parts per million downfield from TMS and coupling constants (J) are reported in Hz. Copies of selected 1 H and 13 C NMR spectra can be found at Supplementary Materials. ESI mass spectra were recorded at 30 V, on a Micromass-Platform LC spectrometer using MeOH as solvent. HR mass spectra were performed using a Bruker AUTOFLEX SPEED MALDI-TOF/TOF or a Bruker Maxis Impact QTOF Spectrometer. FT-IR spectra (4000-400 cm −1 ) were recorded using a Thermo Scientific Nicolet iS20 spectrometer with samples prepared as KBr pellets.
5-(methoxycarbonyl)pyrazine-2-carboxylic acid (A6): Into a suspension of dimethyl pyrazine-2,5-dicarboxylate) (0.3 g, 1.53 mmol) in MeOH (38.2 mL), 1 M NaOH in water (1.53 mmol) was dripped and the solution was stirred at room temperature for 24 h [16]. The resulting mixture was concentrated to dryness under vacuum, and the residue was dissolved in water (10 mL). Then HCl (conc.) (~5 mL) was dripped into the solution until a pale-yellow solid was precipitated. The suspension was filtrated under vacuum and air dried to give the acid as a pale-yellow solid (0.250 g, 98% yield), without further purification. Dimethyl terephthalate (A3): Suspension of terephthalic acid (4 g, 24.07 mmol) in MeOH (200 mL) was heated under reflux for 2 h [17]. Then the suspension was cooled at 0 • C and SOCl 2 (50 mL, 485 mmol) was dripped carefully into the system. The reaction mixture was stirred at 80 • C for 24 h. The resulting solution was evaporated to dryness and the residue was subjected to FCC purification using toluene/ethyl acetate 8:2 as eluent, to give the desired diester as white fluffy crystals (4.1 g, 88% yield). Rf (PhMe/EtOAc 4-((tritylamino)methyl)benzoic acid (B1): To a stirring suspension of 4-(aminomethyl) benzoic acid (4.53 g, 30 mmol) in a mixture of anhydrous DCM (52.5 mL) and ACN (7.5 mL), trimethylchlorosilane (4.19 mL, 33 mmol) was added, and the resulting mixture was heated under reflux for 30 min. After cooling at room temperature, the solution was cooled at 0 • C and triethylamine (17.5 mL, 120 mmol) was added dropwise into the system, followed by a portion-wise addition of TrtCl (8.78 g, 31.5 mmol) within 30 min. The resulting white emulsion was stirred vigorously for 1 h at 0 • C, and additionally for 3 h at room temperature. Then, the mixture was cooled at 0 • C, MeOH (3 mL) was added and after 30 min the mixture was concentrated to dryness. The resulting residue was diluted in 4 M NaOH (aq) (30 mL), placed in a separatory funnel, and twice extracted with diethyl ether. The aqueous layer was acidified with ice-cooled 5% citric acid (aq) up to pH 4-5 and twice extracted with ethyl acetate. The combined organic layers were, thereupon, washed with water and brine, dried over anhydrous Na 2 SO 4 , filtered and concentrated to dryness to afford the product as white foam (8 g and HATU (988 mg, 2.6 mmol) were added and the mixture was cooled at 0 • C. Then, 1,2-phenylenediamine (325 mg, 3 mmol) was added and the mixture was stirred at room temperature for 24 h. Upon completion, the mixture was placed in a separatory funnel and washed with ice-cooled 5% citric acid (aq) , water and brine. The organic layer was dried over anhydrous Na 2 SO 4, filtered and concentrated to dryness. The residue was subjected to FCC purification using toluene/ethyl acetate 9:1 as eluent and the desired amide was afforded as white crystalline solid (812 mg, 66% yield). Rf (PhMe/EtOAc  The reaction mixture was initially stirred at 0 • C and then at room temperature for 10 h. Then, the solution was diluted in DCM and twice successively washed with water, 5% NaHCO 3(aq) solution, and brine. The organic layer was dried over anhydrous Na 2 SO 4 , filtered, and concentrated to dryness to afford an oily residue. The desired amide was afforded as a pale-white solid (40 mg, 44% yield) after FCC purification of the residue using toluene/ethyl acetate 9:1 as eluent. Rf (PhMe/EtOAc) 9:1 = 0.    [1,3,2]dioxaborol-2-yl)butyl)amino)-1-oxo-3-phenylpropan-2-aminium chloride salt (35 mg, 0.08 mmol) was added and the reaction mixture was left under stirring at 0 • C for 30 min and subsequently at room temperature for 20 h. Upon completion, the mixture was subjected to FCC purification using toluene/ethyl acetate 8:2 as eluent, and the hybrid was afforded as a yellow oil (24 mg, 40% yield). Rf (PhMe/EtOAc) 8:2 = 0.29; IR (KBr): 2962, 1729, 1616, 1120, 1020 cm −1 ; HR-MALDI (m/z) [M+Na + ] calculated for 11 by replacing the phenylalanine moiety with melphalan failed to offer the presumed improvement of activity. Based on this work, compounds with improved activity against bortezomib-resistant cells may be designed.