Bortezomib – JQ1 chemical

Synergistic induction of apoptosis in resistant head and neck carcinoma and leukemia by alkoxyamide-based histone deacetylase inhibitors

Leandro A. Alves Avelar a, 1, Christian Schrenk a, 1, Melf So€nnichsen a, 1, Alexandra Hamacher a, Finn K. Hansen b, Julian Schliehe-Diecks c, Arndt Borkhardt c, Sanil Bhatia c, **, 1, Matthias U. Kassack a, ***, 1, Thomas Kurz a, *, 1

A B S T R A C T

Targeting epigenetic dysregulation has emerged as a valuable therapeutic strategy in cancer treatment. Especially epigenetic combination therapy of histone deacetylase inhibitors (HDACi) with established anti-cancer drugs has provided promising results in preclinical and clinical studies. The structural optimization of alkoxyamide-based class I/IIb inhibitors afforded improved analogs with potent efﬁcacy in cisplatin-resistant head and neck carcinoma cells and bortezomib-resistant leukemia cells. The most promising HDACi showed a superior synergistic cytotoxic activity as compared to vorinostat and class I HDACi in combination with cisplatin, leading to a full reversal of the chemoresistant phenotype in head and neck cancer cell lines, as well in combination with the proteasome inhibitors (bortezomib and carﬁlzomib) in a panel of leukemic cell lines. Furthermore, the most valuable alkoxyamide-based HDACi exhibited strong ex vivo anticancer efﬁcacy against primary patient samples obtained from different therapy-resistant leukemic entities.

Keywords:
Histone deacetylases HDAC isozyme Proﬁle Class I
Class IIb inhibitors Leukemia
Solid cancer Cancer resistance
Chemosensitizing effects

1. Introduction

Epigenetic dysregulation is a hallmark of cancer and its impor- tance in cancer therapy has increased considerably in the last de- cades [1]. Among epigenetic modiﬁcations, reversible acetylation of speciﬁc lysine residues of histones along with DNA methylation are the most important regulators of gene transcription and DNA accessibility [2]. Lysine acetylation is a ubiquitous process present in all intracellular compartments. It is essential for gene tran- scription, metabolic processes and homeostasis of bioenergetic processes [3]. The acetylation of lysine residues is performed by lysine acetyltransferases (KATs), while the deacetylase process is catalyzed by lysine deacetylases (KDACs), commonly referred to as histone deacetylases (HDACs), due to their ﬁrst discovered biolog- ical role [3]. HDACs are transcriptional repressors that erase epigenetic marks. Humans express eighteen different HDAC isozymes, which are grouped into four classes: the zinc-dependent class I, II and IV, and the NADþ dependent class III (sirtuins). Class I is formed by HDAC1-3 and HDAC8, while class II is further subdivided into class IIa containing HDAC4, HDAC5, HDAC7, and HDAC9 and class IIb, which includes HDAC6 and HDAC10. HDAC11 is the only repre- sentative of class IV [4]. Pan- and class I-selective histone deacetylase inhibitors (HDACi) are being used in hematological cancer therapy since the approval of vorinostat (Fig. 1A) by the FDA [5].
Epigenetic dysregulation is an important characteristic of practically all cancer cells. Approximately 80e90% of deaths in cancer patients are attributed to cancer resistance. Anticancer drug combinations are one promising tool to combat cancer resistance. Recent preclinical and clinical studies demonstrated that the combination of HDACi with cytotoxic, target-based, and immuno- therapeutic anti-cancer drugs provided improved therapeutic re- sults due to various synergistic effects [6,7]. Most of the currently approved HDACi are pan-inhibitors that cause in part signiﬁcant side effects and off-target toxicity such as diarrhea, nausea, anorexia, cardiac toxicity and thrombocytopenia [5]. Even though preferential/selective class I HDACi are in clinical use for lymphoma treatment (e.g. romidepsin and tucidinostat); it is still under investigation if inhibition of speciﬁc HDAC isozymes will improve these drawbacks. Although it has been observed that as a single treatment, the efﬁcacy of HDACi in solid cancers is limited, various HDACi-based combination therapies have entered clinical trials for the treatment of a broad range of cancers [7]. Especially the use of HDACi as chemosensitizers that improve the efﬁciency of currently used cancer drugs has shown great potential in preclinical and clinical drug research. However, it remains to be clariﬁed which isozyme proﬁles are responsible for the chemosensitizing effects in certain drug combinations depending on the type of cancer. Thus, more preclinical and clinical data are required to prove which tu- mor speciﬁc HDAC isozyme proﬁle is needed for efﬁcient therapy regimes.
Previously we reported, that the epigenetic modulation of ovarian and head and neck carcinomas by either HDAC class I speciﬁc inhibitors [10,11] or HDAC6 selective HDACi [12] led in part to restored cisplatin efﬁcacy. We also observed pronounced che- mosensitizing effects with alkoxyamide- and alkoxyurea-based class I/IIb inhibitors (Fig. 1B). To further explore the potential of class I/IIb inhibitors as chemosensitizing epigenetic anti-cancer agents, a new series of alkoxyamide-based HDACi focusing on cap group optimization were synthesized and tested. Intriguingly, the most valuable HDACi exhibited synergistic cytotoxic activity in combination with cisplatin leading to a full reversal of the che- moresistant phenotype in head and neck cancer cells. Furthermore, the lead HDACi were potent as single-agents against resistant leu- kemia cell lines and primary patient samples obtained from different therapy-resistant leukemia entities and showed strong synergistic interaction with proteasome inhibitors (bortezomib and carﬁlzomib) in a panel of leukemia cell lines.

2. Results and discussion

2.1. Development of new alkoxyamide-based HDACi

In previous studies, we demonstrated the importance of the cap group for the cytotoxic and chemosensitizing effects of alkoxyamide-based HDACi [8,13]. The most potent class I/IIb in- hibitors contain 3,5-disubstituted-phenyl moieties as cap groups (e.g. YAK61, Fig. 1B). The enlargement of the cap-region of alkoxyurea-based HDACi also proved to be a valuable strategy to improve their in vitro anti-cancer effects. Replacement of the phenyl ring with naphthalene and a substituted quinoline moiety (e.g. KSK64, Fig. 1B) signiﬁcantly increased their cytotoxic and chemosensitizing properties [9,13]. Based on these results, novel alkoxyamide-based HDACi with bicyclic (hetero)aromatic cap groups were designed and synthesized. In this study, we evaluated the isozyme proﬁle, cytotoxicity, apoptotic effects, drug combina- tions and the chemosensitizing properties of several newly syn- thesized alkoxyamide-based class I/IIb inhibitors against sensitive and cisplatin-resistant head and neck (Cal27) and ovarian (A2780) carcinomas, as well as in various resistant leukemia cell lines and primary resistant leukemic patient samples.

2.2. Synthesis of novel alkoxymide HDACi

The synthesis of compounds 6a-k and 8a-g was performed in ﬁve steps according to a previously published procedure (Scheme 1) [14]. Starting from 6-bromohexanoic acid (1), the O-trityl pro- tected hydroxamic acid 2 was synthesized via mixed anhydride coupling with O-trityl-hydroxylamine. The hydroxylamine 4 was prepared via the protected intermediate 3 in two steps using Gabriel synthesis conditions. Next, selected cap groups were introduced by coupling reactions of 4 with the respective carboxylic acid using EDCl/DMAP as an amide coupling system. Finally, the free hydroxamic acids (6a-k, 8a-g) were obtained through the acid- catalyzed removal of the trityl group. The synthesis of compound 6a and 6b has been previously reported [14].

2.3. In vitro anticancer activity and cellular HDAC inhibition of compounds 6a-k and 8a-g in head and neck (Cal27) and ovarian (A2780) carcinoma cell lines

Analogs 6a-k (Table 1) and 8a-g (Table 2) were evaluated in MTT antiproliferative and whole-cell HDAC assays in the ovarian epithelial cancer cell line A2780 and the oral squamous cell carci- noma line Cal27. Vorinostat and cisplatin served as controls. The antiproliferative activity in the oral squamous cell carci- noma line Cal27 was stronger than in the ovarian epithelial cancer cell line A2780. Compounds 6b, 6c, 8a, 8d, and 8e displayed sub- micromolar cytotoxic activity against Cal27, while only one com- pound (8a) exhibited submicromolar activity against A2780. Notably, the compounds with the strongest antiproliferative ac- tivity exhibiting IC50 values below 0.5 mM against Cal27 are the 2- indolyl-substituted inhibitor 6c (IC50 0.3 mM), the 2-quinolyl- substituted analog 8a (IC50 0.25 mM) and the 2-naphthyl-derivative 8e (IC50 0.18 mM). Thus, several alkoxy-amide-based HDACi are signiﬁcantly more active than the reference compound vorinostat. Cellular HDAC assays were performed as previously published [13] and the results are summarized in Tables 1 and 2 6c, 8a, 8d, and 8e were the most active compounds in the cellular HDAC assays displaying IC50 values ranging from 0.10 to 0.67 mM against both cancer cell lines. Results for 6c, 8a, 8d, and 8e from cellular HDAC assays are in good agreement with the antiproliferative effects observed in MTT assays except for 6c and 8d in A2780. The good

2.4. Isozyme proﬁling: of inhibitors 6c, 8a, and 8c-e

To analyze the isozyme proﬁle, the most potent compounds in MTT and whole-cell HDAC assay were tested against human re- combinant HDAC2, 4, 6, and 8 (Table 3), which represent HDAC class I (HDAC2 and 8), IIa (HDAC4) and IIb (HDAC6). The pan HDACi vorinostat, compound 2a [10] (class I selective HDACi), CHDI00390576 (class IIa selective HDACi) and tubastatin A (class IIb selective HDACi) served as reference compounds. Except for 8c (IC50: 759 nM), vorinostat and 2a, all selected alkoxyamide-based HDACi inhibited HDAC2 with IC50 values in the double-digit nanomolar range (6c, 8a, 8d, 8e: IC50: 41e92 nM). HDAC6 inhibition was even stronger with IC50 values between 11 and 25 nM (6c, 8a, 8d, 8e) and 173 nM for 8c. Thus, compounds 6c, 8a, 8c, 8d and 8e are highly active HDAC class I/IIb inhibitors, with a moderate preference for HDAC6. The inhibition of HDAC8 was moderate for all tested compounds with IC50 values in the mM range. Furthermore, all tested compounds (except reference com- pound CHDI00390576) displayed weak HDAC4 inhibitory activity.

2.5. Acetylation of a-tubulin and histone H3

Western Blot analysis was used to test for cellular HDAC inhib- itory activity in Cal27 cells in addition to data from whole-cell HDAC assays displayed in Tables 1 and 2 A representative immu- noblot is shown in Fig. 2. Incubation with vorinostat, entinostat, and compounds 6c, 8e, 8a and 8c for 24 h induced acetylation of histone H3 compared to control, indicating inhibition of class I HDACs. The highest level of histone H3 acetylation was observed for 6c, 8a and 8c. Notably, these compounds also induced the expres- sion of histone H3 itself. All compounds, except entinostat, led to an accumulation of acetylated a-tubulin, indicating HDAC6 inhibition.

2.6. Enhancement of cisplatin-induced cytotoxicity

The effects of the most potent compounds 6c, 8a, 8c, 8d and 8e on cisplatin-induced cytotoxicity were then tested in the Cal27 and its cisplatin-resistant sub-cell line Cal27CisR [16]. The results are shown in Table 4. Compounds 6c, 8a and 8c, which induced histone H3 expression and showed the strongest accumulation of acetyl- histone H3 (Fig. 2), increased the potency of cisplatin in both cell lines. 8a and 8c were even more efﬁcient in shifting cisplatin IC50 values than the reference HDACi vorinostat (pan HDACi) or class I- selective compound 2a as can be seen by shift factors (SF) of 13.2 for 8a and 10.0 for 8c compared to 8.91 for vorinostat or 7.51 for 2a in Cal27CisR cells (Table 4). A 48 h preincubation with 0.5 mM of compound 8a reduced the cisplatin IC50 from 48.5 mM to 3.67 mM, thus completely reversing cisplatin resistance in Cal27CisR (Fig. 3). Notably, the ability of these new HDACi to increase the potency of cisplatin was more pronounced in the cisplatin-resistant cell line Cal27CisR compared to the parental cell line Cal27 as evident by the higher shift factors in Cal27CisR. A SF of 7.51 by 2a (class I selective HDACi) compared to 1.5 by Tubastatin A (in Cal27CisR) further demonstrates the importance of class I HDAC inhibition in reversing cisplatin resistance, which has been investigated by our group in previous studies [17,18]. These data demonstrate that our class I/HDAC6 HDACi described in this paper are superior to class I HDACi 2a and to pan HDACi vorinostat in reversing cisplatin resistance. In addition, class I/IIb inhibition could be beneﬁcial in combination with proteasome inhibitors [19,20]. Therefore, the combination of compound 8a with proteasome inhibitors borte- zomib and carﬁlzomib was further evaluated in section 2.11.

2.7. Synergistic interaction between HDACi and cisplatin

To further investigate the interaction type of cisplatin and the new HDACi, the combination index, according to Chou Talalay [21], was estimated for the three compounds 6c, 8a, and 8c showing the highest increase in histone acetylation (Fig. 2) and the highest shift factors for cisplatin (Table 4). 6c, 8a and 8c showed strong synergy as shown by CI < 0.5 for almost all concentrations tested in the combination MTT assays (Table 5). 2.8. Synergistic interaction of 6c, 8a and 8c with cisplatin is associated with caspase 3/7 activation Compounds 6c, 8a and 8c increased cisplatin-induced cytotox- icity signiﬁcantly (Table 4, Fig. 3) prompting us to test caspase 3/7 activation as a marker of apoptosis induction. The results are pre- sented in Fig. 4 5. Compounds 6c, 8a and 8c alone had a little or moderate effect on caspase 3/7 activation in Cal27 or Cal27CisR. Further, a single treatment with cisplatin gave only a slight increase in caspase 3/7 activation. However, in combination treatments, all three HDACi with cisplatin, respectively, resulted in hyperadditive activation of caspase 3/7 (Fig. 4). Quantiﬁcation of the ﬂuorescent intensities conﬁrmed the strong enhancement of caspase 3/7 acti- vation by the combination treatment (Fig. 5). The difference be- tween the measured values from combination treatment (white bars) and the sum of single treatments (cisplatin alone plus HDACi alone, black bars) showed a signiﬁcant difference. This is in line with a synergistic interaction of the HDACi with cisplatin as shown by Chou-Talalay analysis in Table 5. 2.9. Analysis of DNA double-strand breaks (phosphorylation of g- H2AX) Next, the combination of 6c, 8a and 8c with cisplatin was tested for their ability to enhance DNA double-strand breaks estimated by gH2AX assay (Figs. 6 and 7). Compounds 6c, 8a and 8c alone only slightly increased phosphorylation of H2AX, while cisplatin dis- played a concentration-dependent increase in gH2AX foci in both cell lines (Fig. 7). In combination with cisplatin, all three HDACi showed a hyperadditive effect in gH2AX formation in Cal27 and Casl27CisR (Fig. 7). 2.10. Functional speciﬁcity of compound 8a and compound 8e in leukemic cell lines To explore the broad anti-cancer activity of quinoline/naph- thalene series compounds from solid cancer cell line screenings, the functional speciﬁcity and efﬁcacy of the lead compounds (8a and 8e) were later analyzed in resistant cell lines obtained from different leukemia entities. Initially, the speciﬁcity of the com- pounds 8a and 8e against different HDAC classes was measured with a cellular-based (HDAC class I/IIb and class IIa) activity assay, using the HL60 (acute myeloid leukemia or AML) cell line. Ric- olinostat (preferential HDAC6i, Fig. 1A) and CAY10603 (selective HDAC6i, Fig. 9B) served as reference inhibitor controls. As expected, compounds 8a and 8e were able to deacetylate class I/IIb speciﬁc substrates with <1 mM concentrations, however, class IIa speciﬁc substrates remained deacetylated (Fig. 8). The IC50 for HDAC class I/ IIb activity for 8a and 8e (73.2 nM and 85.1 nM, respectively) was comparable to CAY10603 (109 nM), whereas IC50 was signiﬁcantly higher for ricolinostat (902 nM). The cellular HDAC inhibition was further correlated with the cell viability data (IC50), which was conducted by incubating the HL60 cells at the same time point with the respective inhibitors for 72 h, later the readout was taken using ATP-based Celltitre-Glo luminescent cell viability assay (Fig. 8). The lowest IC50 values were found for compound 8a with 270 nM compared to 8e with 561 nM, whereas CAY10603 with 594 nM and ricolinostat with >10000 nM (Fig. 8 and SI Table S1).
Encouraged by the superior activity, the compounds 8a and 8e were further tested in comparison to a panel of commercially available (ranging from selective to preferential) HDAC6 inhibitors in leukemic cell lines (B-cell acute lymphoblastic leukemia or B- ALL; HAL01, AML; HL60, T-cell acute lymphoblastic or T-ALL; Jurkat), a bortezomib resistant cell line HL60 (HL60r) and four primary leukemia patient samples. In this analysis, compound 8a came up again as the most potent hit with the lowest IC50 con- centrations in the nanomolar range (data represented as a clustered heat map), comparable to compound 8e and CAY10603, whereas all the other tested selective HDAC6 inhibitors had IC50 values in the micromolar range (Fig. 9 and SI Table S1). Moreover, the potent anti-leukemic activity of compound 8a prompted us to further explore its ability to induce apoptosis in caspase 3/7 enzyme- dependent assay (Fig. 10). Brieﬂy, HL60 cells were treated with compound 8a at its IC50 and IC75 concentrations for 48 h, whereas ricolinostat (10-fold higher concentration than 8a) served as a positive control. Compound 8a signiﬁcantly induced apoptosis at its IC50 concentration (213.7 nM) and showed a more potent induction at its IC75 (320.5 nM) concentration similar to ricolinostat, however with a tenfold higher IC50 concentration of 8a (Fig. 10).

2.11. Synergistic analysis of compound 8a in combination with proteasome inhibitor bortezomib and carﬁlzomib

8a exhibited potent cytotoxicity activity and signiﬁcant apoptosis induction in leukemic cell lines and was potent against the bortezomib resistant HL60 (HL60r) cell line, which encouraged us to ﬁnd the synergistic combination partners. In recent studies, it was shown that HDACi show promising anti-cancer properties when combined with proteasome inhibitors [6,22e26]. Therefore we screened compound 8a in against increasing concentrations (10 X 10 dose-response matrices) of the proteasome inhibitors carﬁl- zomib or bortezomib in the same set of leukemic cell lines that were used in prior experiments (B-ALL: HAL01, T-ALL: Jurkat and AML: HL60) using a semi-automated drug screening platform. In line with the previous studies, we also detected signiﬁcant syner- gistic interaction of compound 8a and bortezomib or carﬁlzomib in all three tested cell lines, using Bliss synergy analysis [27]. Espe- cially in the HL60 and Jurkat cells, stronger ﬁelds of synergistic interactions were detected, as compared to HAL01(Fig. 11). We further screened vorinostat against increasing concentrations of bortezomib in the HL60 and Jurkat cell lines (10 x 10 dose-response matrices). Vorinostat and bortezomib show no synergistic inter- action in the Jurkat cell line and only minor synergistic interaction in the HL60 cell line, indicating the superiority of 8a over vorinostat for the combination of HDACi and a proteasome inhibitor (Fig. S1).

3. Conclusion

The combination of histone deacetylase inhibitors with different anti-cancer drugs has shown promising results against cancer resistance in different types of hematological and solid tumors [6,28,29]. Based on previously observed chemosensitizing effects during the combination of alkoxyamide-based class I, IIb HDACi and cisplatin, new analogs were synthesized which contain bicyclic (hetero) aromatic cap groups. Compounds 6c, 8a, 8d, and 8e were identiﬁed as the most potent compounds in the MTT and HDACi whole-cell assay exerting IC50 values in the submicromolar range in cell line Cal27. The reference HDACi vorinostat was considerably less active in both assays. The new HDACi are nanomolar HDAC2/ HDAC6 inhibitors, with only moderate activity against HDAC8 and low activity towards HDAC4. The selectivity proﬁle (class I/IIb in- hibitor) was further conﬁrmed by western blot analysis (acetylated histone and tubulin, Fig. 2). A 48 h preincubation of 8a before treatment with cisplatin signiﬁcantly increased the cytotoxic effect of cisplatin in cell lines Cal27 and Cal27CisR, thus leading to a complete resensitization of the cisplatin-resistant phenotype. Of interest, 8a was superior in resensitization of cisplatin resistance as estimated by shift factors over the pan HDACi vorinostat, the class IIb selective inhibitor tubastatin A and the class I selective HDACi 2a.
Overexpression of certain HDAC classes directly plays a pivotal role in the identiﬁcation, prognosis, and treatment of leukemia [30e33]. Therefore, we also examined the promising inhibitors 8a and 8e in cell lines obtained from different leukemia subtypes, a bortezomib resistant form of the HL60 cell line (HL60r), and four primary leukemic patient samples. 8a and 8e showed high cyto- toxic properties as a single agent against the major acute subtypes of leukemia cell line models (AML, B-ALL, and T-ALL) and even in primary leukemic patient samples cells in the nanomolar range, and at much better efﬁciency than the clinical candidate ricolino- stat [34]. HDAC inhibitors have been demonstrated promising anti- cancer properties when combined with proteasome inhibitors [6,22e26]. Furthermore, the new HDACi were also potent against bortezomib resistant cell lines. In this line, we showed that com- pound 8a synergized strongly with proteasome inhibitors in a bliss matrix synergy screen against the HL60 (AML) cell line. In sum- mary, the structural optimization afforded improved alkoxyamide- based HDAC class I/IIb inhibitors exerting unique chemosensitizing effects in resistant head and neck and leukemia cells. Further pre- clinical studies will be addressed along with further SAR and initial MOA studies in a follow-up lead optimization program.

4. Experimental section

4.1. Chemistry

Chemicals and solvents were purchased from commercial sup- pliers (Sigma-Aldrich, Acros Organics, TCI, Fluorochem, Fluka, ABCR, Alfa Aesar, J&K, Carbolution and Merck) and used without further puriﬁcation. The reactions were monitored by thin-layer chromatography (TLC) using Merck precoated silica gel plates (with ﬂuorescence indicator UV254), being visualized by irradia- tion with ultraviolet light (254 nm) or staining in potassium permanganate solution. Column chromatography was performed using Macherey-Nagel Silica 60 (0.040e0.063 mm) and Flash chromatography using prepacked silica cartridge with the solvent mixtures of hexane/ethyl acetate or dichloromethane/methanol according to the separation. Melting points (mp) analyses were performed using a Mettler FP 5 melting-point apparatus and are uncorrected. Proton (1H) and carbon (13C) NMR spectra were recorded on a Bruker Avance 300, 500, or 600 MHz using DMSO‑d6 as the solvent. Chemical shifts are given in parts per million (ppm), relative to residual solvent peak for 1H and 13C. High-resolution mass spectra (HRMS) analysis was performed on a UHR-TOF maXis 4G, Bruker Daltonics, Bremen by electrospray ionization (ESI). Analytical HPLC analyses were carried out on a Varian Prostar system equipped with a Prostar 410 (autosampler), 210 (pumps) and 330 (UV-detector) using a Phenomenex Luna 5u C18 1.8 mm particle (250 mm 4.6 mm) column, supported by Phenomenex Security Guard Cartridge Kit C18 (4.0 mm 3.0 mm) or on in an AZURA P 6.1L (pumps) and Smartline 2600 VIS LWL (UV-detector) using a column type Vertex Plus Column (length 150 4 mm with precolumn). In both pieces of equipment, the UV absorption was detected at 254 nm with a linear gradient of 10% B to 100% B in 30 min using HPLC-grade water 0.1% TFA (solvent A) and HPLC- grade acetonitrile 0.1% TFA (solvent B) for elution at a ﬂow rate of 1 mL/min.
The general procedure for the synthesis and characterization of compounds 6c-k and 8a-g are bellowed. The synthesis and char- acterization of protected hydroxamates 2, 3, 4, 5c-k, and 7a-g are reported in the Electronic Supplementary Information. [M H]: m/z calcd for C18H23N2O4S:331.1652, found: 331.1646. HPLC: retention time 8.45, purity 95.3%.

4.2. Biological Evaluation

4.2.1. Reagents

Cisplatin was purchased from Sigma (Germany) and dissolved in 0.9% sodium chloride solution. Vorinostat was synthesized ac- cording to known procedures [13]. Stock solutions (10 mM) of the respective compounds were prepared with DMSO and diluted to the desired concentrations with the appropriate medium. All other reagents were supplied by PAN Biotech (Germany) unless other- wise stated.

4.2.2. Cell lines and cell culture

(Solid tumor) The human ovarian carcinoma cell line A2780 was obtained from the European Collection of Cell Cultures (ECACC, Salisbury, UK). The human tongue cell line Cal27 was obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ, Germany). The corresponding cisplatin-resistant CisR cell line Cal27CisR was generated by exposing the parental cell line to weekly cycles of cisplatin in an IC50 concentration over 24e30 weeks as described in Gosepath et al. and Eckstein et al. [16]. All cell lines were grown at 37 ◦C under humidiﬁed air supplemented with 5% CO2 in RPM I 1640 (A2780) or DMEM (Cal27) containing 10% fetal calf serum, 120 IU/mL penicillin, and 120 mg/mL streptomycin. The cells were grown to 80% conﬂuency before being used in further assays. (Leukemia): The leukemic cell lines, HAL01 (B-ALL) and HL60 (AML) were cultured in RPMI 1640 GlutaMax supplemented with 10% FCS. Bortezomib resistant HL60 (HL60r) clones were generated by gradually increasing the concentration over three months up to its 80 nM concentration. However HL60r cells were generally grown with 20 nM bortezomib. Jurkat (T-ALL) was cultured in RPMI 1640 GlutaMax supplemented with 20% FCS (DSMZ, Braunschweig, Germany). The primary leukemic patient cells were cultured in RPMI 1640 supplemented with 15% FCS, 0.6% sodium pyruvate, 0.012% mercaptoethanol, 0.1% gentamycin. All cells were cultured in a 37 ◦C humidiﬁed incubator with 5% CO2.

4.2.3. Cell viability assay

(Solid tumors) The rate of cell-survival under the action of test substances was evaluated by an improved MTT assay as previously described [35]. In brief, A2780 and Cal27 cell lines were seeded at a density of 5,000 and 2,500 cells/well in 96 well plates (Corning, Germany). After 24 h, cells were exposed to increased concentra- tions of the test compounds. Incubation was ended after 72 h and cell survival was determined by the addition of MTT solution (Serva, Germany, 5 mg/mL in phosphate-buffered saline). The formazan precipitate was dissolved in DMSO (VWR, Germany). Absorbance was measured at 544 nm and 690 nm in a FLUOstar microplate- reader (BMG LabTech, Offenburg, Germany). To investigate the ef- fect of 6c, 8a, 8c, 8d, 8e on cisplatin-induced cytotoxicity, com- pounds were added 48 h before cisplatin administration. After 72 h, the cytotoxic effect was determined with an MTT cell viability assay and shift factors were calculated by dividing the IC50 value of cisplatin alone by the IC50 value of the drug combinations.

4.2.4. Cell viability assay

(Leukemia) The assay was performed to determine the IC50 values for the leukemic cell lines. The compounds were printed on white 384-well plates (Thermo Fisher Scientiﬁc, Waltham, USA) with increasing concentrations (5 nM – 10 mM) by using a digital dispenser (D300e, Tecan, Ma€nnedorf, Switzerland). Cell viability was monitored after 72 h using ATP based CellTiter-Glo luminescent assay (Promega, Madison, USA) using a microplate reader (Spark, Tecan) [25]. The IC50 values for the compounds were determined by plotting raw data (normalized to controls) using the sigmoid dose curve and nonlinear regression (GraphPad Prism Inc., San Diego, CA). The IC50 data was plotted as a clustered heat map, followed by unsupervised hierarchical clustering.

4.2.5. Whole-cell HDAC inhibition assay
The cellular HDAC assay was based on a procedure published by Ciossek et al. [36] and Bonﬁls et al. [37] with minor modiﬁcations. Brieﬂy, human cancer cell lines Cal27 and A2780 were seeded in 96-well tissue culture plates (Corning, Germany) at a density of 1.5 104 cells/well in a total volume of 90 ml culture medium. After 24 h, cells were incubated for 18 h with increasing concentrations of test compounds. The reaction was started by adding 10 ml of 3 mM Boc-Lys(ε-Ac)-AMC (Bachem, Germany) to reach a ﬁnal concentration of 0.3 mM7 Cells were incubated with Boc-Lys(ε-Ac)- AMC for 3 h under cell culture conditions. After this incubation, 100ml/well stop solution (25 mM Tris-HCl (pH 8), 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 1% NP40, 2.0 mg/mL trypsin, 10 mM vorinostat) was added and the reaction was developed for 3 h under cell cul- ture conditions. Fluorescence intensity was measured at an exci- tation of 355 nm and an emission of 460 nm in a NOVOstar microplate-reader (BMG LabTech, Offenburg, Germany).

4.2.6. Cellular HDAC activity analysis

The compounds were printed on white 384-well plates (Thermo Fisher Scientiﬁc, Waltham, USA) with increasing concentrations (5 nM – 10 mM) by using a digital dispenser (D300e, Tecan, M€annedorf, Switzerland). Cells were added to the plate and incu- bated for 30 min. The HDAC-Glo class I/II and HDAC-Glo class IIa Assay and Screening System (Promega) were used, following manufacturer’s instructions. The HDAC activity for the compounds was determined by plotting raw data (normalized to controls) using a sigmoid dose curve and nonlinear regression (GraphPad Prism Inc., San Diego, CA).

4.2.7. Caspase 3/7 activation assay

(Solid tumors) Compound-induced activation of caspases 3 and 7 were analyzed by using the CellEvent Caspase 3/7 green detection reagent (Thermo Scientiﬁc Germany) according to the manufac- turer’s instructions. Brieﬂy, Cal27 and Cal27 CisR cells were seeded in 96-well-plates (Corning, Germany) at a density of 4000/ 3000 cells per well. Cells were treated with 6c, 8a or 8c for 24 h prior to cisplatin. After a further incubation period of 18 h, the medium was removed and 50 ml of CellEvent Caspase 3/7 green detection reagent (2 mM in PBS supplemented with 5% heat- inactivated FBS) was added. Cells were incubated for 30 min at 37 ◦C in a humidiﬁed incubator before imaging by using the Thermo Fisher ArrayScan XTI high content screening (HCS) system (Thermo Scientiﬁc). Hoechst 33342 was used for nuclei staining. (Leukemia) The Caspase 3/7 assay was conducted after HL60 cells were treated with the vehicle control (0.5% DMSO), 8a (IC50: 213.7 nM, IC75: 320.55 nM) and Ricolinostat (IC50: 2120 nM, IC75: 3180 nM). Caspase-Glo® 3/7 Assay System (Promega) was used measure the caspase-3/7 enzymatic activity (measure of apoptosis induction), following manufacturer’s guidelines.

4.2.8. yH2AX assay

Similar to the Caspase 3/7 activation assay, Cal27 and Cal27 CisR cells were seeded in 96-well plates (Corning, Germany) and pre- treated with the same concentration of 6c, 8a or 8c for 24 h prior to the addition of cisplatin. After an 18 h incubation period, the me- dium was removed and cells were washed with PBS. Cells were then ﬁxed by the addition of 50 ml 3.7% PFA for 20 min, washed again with PBS and permeabilized by 50 ml of 0.1% TritonX in PBS for 5 min. After another washing step with PBS, the cells were blocked with 1% BSA in PBS. Cells were stained with Anti-phospho-Histone H2A.X (Ser139) clone JBW301 (Merck KGaA, Germany) in a con- centration of 2 mg/mL in 30 ml blocking buffer (1% BSA in PBS) for 1 h at room temperature. Cells were washed again with PBS following incubation with Donkey Anti-Mouse IgG Northern- Lights™ NL557-conjugate Antibody (R&D Systems, USA) according to the manufacturer’s recommendations for 1 h. After a ﬁnal washing step with PBS, cells were then stained with Hoechst 33342 and imaging was performed by using the Thermo Fisher ArrayScan XTI high content screening (HCS) system (Thermo Scientiﬁc).

4.2.9. Enzyme assay

All human recombinant enzymes were purchased from Reaction Biology Corp. (Malvern, PA, USA). The HDAC activity assay of HDAC2 (catalog nr. KDA-21-277), 4 (catalog nr. KDA-21-279), 6 (catalog nr. KDA-21-213) and 8 (catalog nr. KDA-21-481) was performed in 96- well plates (Corning, Germany). Brieﬂy, 20 ng of HDAC2 and HDAC8, 17.5 ng of HDAC6 and 2 ng of HDAC4 per reaction were used. Re- combinant enzymes were diluted in assay buffer (50 mM Tris-HCl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, and 1 mg/mL BSA). 80 ml of this dilution was incubated with 10 ml of different con- centrations of inhibitors in assay buffer. After a 5 min incubation step, the reaction was started with 10 ml of 300 mM (HDAC2), 150 mM (HDAC6) Boc-Lys(Ac)-AMC (Bachem, Germany) or 100 mM (HDAC4), 60 mM (HDAC8) Boc-Lys(TFA)-AMC (Bachem, Germany). The reac- tion was stopped after 90 min by adding 100 ml stop solution (16 mg/mL trypsin, 2 mM panobinostat for HDAC2, HDAC6 and HDAC8, 2 mM CHDI0039 (kindly provided by the CHDI Foundation Inc., New York, USA) for HDAC4 in 50 mM Tris-HCl, pH 8.0, and 100 mM NaCl. 15 min after the addition of the stop solution the ﬂuorescence intensity was measured at an excitation of 355 nm and emission of 460 nm in a NOVOstar microplate reader (BMG Lab- Tech, Offenburg, Germany).

4.2.10. Immunoblotting

Cells were treated with 1 mM of vorinostat, entinostat, 6c, 8a, 8, and 5 mM of 8c and 8d or vehicle for 24 h. Cell pellets were dissolved with RIPA buffer (50 mM Tris-HCl pH 8.0, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM sodium chloride, 2 mM EDTA, supplemented with protease and phosphatase inhibitors (Pierce™ protease and phosphatase inhibitor mini tablets, Thermo Scientiﬁc)) and clariﬁed by centrifugation. Equal amounts of total protein (30 mg) were resolved by SDS-PAGE and transferred to polyvinylidene ﬂuoride membranes (Merck Millipore). Blots were incubated with primary antibodies against acetylated a-tubulin, a- tubulin (Santa Cruz Biotechnology, Germany), histone H3, and acetyl histone H3 (Lys24) (biotech, Germany). Immunoreactive proteins were visualized using luminol reagent (Santa Cruz Biotechnology, Germany) with an Intas Imager (Intas, Germany).

4.2.11. Combinatorial drug screening

(Leukemia) The compounds were printed on white 384-well plates (Thermo Fisher Scientiﬁc, Waltham, USA) with increasing concentrations in a dose-response 10×10 matrices by using a digital dispenser (D300e, Tecan, Ma€nnedorf, Switzerland). For bortezomib and carﬁlzomib the concentration range was 0.05 nM up to 4 nM and for 8a from 25 nM up to 4 mM, seeded in a logarithmic distri- bution. The concentration range for vorinostat was from 20 nM up to 200 nM.Cell viability was monitored after 72 h using CellTiter- Glo luminescent assay (as described above), using a microplate reader (Spark, Tecan). The Bliss synergy scores were determined by using the Combeneﬁt synergy analysis software [27].

4.2.12. Data analysis

Concentration-effect curves were constructed with Prism 7.0 (GraphPad, San Diego, CA) by ﬁtting data from at least three ex- periments performed in triplicates to the four parameters logistic equation. Statistical analysis was performed using a t-test.

References

[1] S. Sharma, T.K. Kelly, P.A. Jones, Epigenetics in cancer, Carcinogenesis 31 (2009) 27e36, https://doi.org/10.1093/carcin/bgp220.
[2] V. Pande, Understanding the complexity of epigenetic target space, J. Med. Chem. 59 (2016) 1299e1307, https://doi.org/10.1021/acs.jmedchem.5b01507.
[3] I. Ali, R.J. Conrad, E. Verdin, M. Ott, Lysine acetylation goes global: from epi- genetics to metabolism and therapeutics, Chem. Rev. 118 (2018) 1216e1252, https://doi.org/10.1021/acs.chemrev.7b00181.
[4] I.V. Gregoretti, Y. Lee, H.V. Goodson, N. Dame, N. Dame, Molecular evolution of the histone deacetylase Family : functional implications of phylogenetic analysis, J. Mol. Biol. (2004) 17e31, https://doi.org/10.1016/j.jmb.2004.02.006.
[5] B.S. Mann, J.R. Johnson, M.H. Cohen, R. Justice, R. Pazdur, FDA approval sum- mary: vorinostat for treatment of advanced primary cutaneous T-cell lym- phoma, Oncology 12 (2007) 1247e1252, https://doi.org/10.1634/ theoncologist.12-10-1247.
[6] A. Suraweera, K.J. O’Byrne, D.J. Richard, Combination therapy with histone deacetylase inhibitors (HDACi) for the treatment of cancer: achieving the full therapeutic potential of HDACi, Front. Oncol. 8 (2018) 1e15, https://doi.org/ 10.3389/fonc.2018.00092.
[7] H.P. Mohammad, O. Barbash, C.L. Creasy, Targeting epigenetic modiﬁcations in cancer therapy: erasing the roadmap to cancer, Nat. Med. 25 (2019) 403e418, https://doi.org/10.1038/s41591-019-0376-8.
[8] Y. Asfaha, C. Schrenk, L.A. Alves Avelar, F. Lange, C. Wang, J.J. Bandolik, A. Hamacher, M.U. Kassack, T. Kurz, Novel alkoxyamide-based histone deacetylase inhibitors reverse cisplatin resistance in chemoresistant cancer cells, Bioorg. Med. Chem. 28 (2020), https://doi.org/10.1016/ j.bmc.2019.115108.
[9] K. Stenzel, A. Hamacher, F.K. Hansen, C.G.W. Gertzen, J. Senger, V. Marquardt, L. Marek, M. Marek, C. Romier, M. Remke, M. Jung, H. Gohlke, M.U. Kassack, T. Kurz, Alkoxyurea-based histone deacetylase inhibitors increase cisplatin potency in chemoresistant cancer cell lines, J. Med. Chem. 60 (2017) 5334e5348, https://doi.org/10.1021/acs.jmedchem.6b01538.
[10] V. Krieger, A. Hamacher, F. Cao, K. Stenzel, C.G.W. Gertzen, L. Scha€ker-Hübner, T. Kurz, H. Gohlke, F.J. Dekker, M.U. Kassack, F.K. Hansen, Synthesis of peptoid- based class I-selective histone deacetylase inhibitors with chemosensitizing properties, J. Med. Chem. 62 (2019) 11260e11279, https://doi.org/10.1021/ acs.jmedchem.9b01489.
[11] J.J. Bandolik, A. Hamacher, C. Schrenk, R. Weishaupt, M.U. Kassack, Class i- histone deacetylase (Hdac) inhibition is superior to pan-hdac inhibition in modulating cisplatin potency in high grade serous ovarian cancer cell lines, Int. J. Mol. Sci. 20 (2019), https://doi.org/10.3390/ijms20123052.
[12] D. Diedrich, A. Hamacher, C.G.W. Gertzen, L.A. Alves Avelar, G.J. Reiss, T. Kurz, H. Gohlke, M.U. Kassack, F.K. Hansen, Rational design and diversity-oriented synthesis of peptoid-based selective HDAC6 inhibitors, Chem. Commun. 52 (2016) 3219e3222, https://doi.org/10.1039/c5cc10301k.
[13] L. Marek, A. Hamacher, F.K. Hansen, K. Kuna, H. Gohlke, M.U. Kassack, T. Kurz, Histone deacetylase (HDAC) inhibitors with a novel connecting unit linker region reveal a selectivity proﬁle for HDAC4 and HDAC5 with improved ac- tivity against chemoresistant cancer cells, J. Med. Chem. 56 (2013) 427e436, https://doi.org/10.1021/jm301254q.
[14] L.A. Alves Avelar, J. Held, J.A. Engel, P. Sureechatchaiyan, F.K. Hansen, A. Hamacher, M.U. Kassack, B. Mordmüller, K.T. Andrews, T. Kurz, Design and synthesis of novel anti-plasmodial histone deacetylase inhibitors containing an alkoxyamide connecting unit, Arch. Pharm. (Weinheim) 350 (2017) 1e11, https://doi.org/10.1002/ardp.201600347.
[15] M. Pﬂieger, A. Hamacher, T. O€ z, N. Horstick-Muche, B. Boesen, C. Schrenk, M.U. Kassack, T. Kurz, Novel a,b-unsaturated hydroxamic acid derivatives overcome cisplatin resistance, Bioorg. Med. Chem. 27 (2019) 115036, https:// doi.org/10.1016/j.bmc.2019.07.052.
[16] E.M. Gosepath, N. Eckstein, A. Hamacher, K. Servan, G. von Jonquieres, H. Lage, B. Gyo€rffy, H.D. Royer, M.U. Kassack, Acquired cisplatin resistance in the headeneck cancer cell line Cal27 is associated with decreased DKK1 expres- sion and can partially be reversed by overexpression of DKK1, Int. J. Canc. 123 (2008) 2013e2019, https://doi.org/10.1002/ijc.23721.
[17] V. Krieger, A. Hamacher, F. Cao, K. Stenzel, C.G.W. Gertzen, L. Scha€ker-Hübner, T. Kurz, H. Gohlke, F.J. Dekker, M.U. Kassack, F.K. Hansen, Synthesis of peptoid- based class I-selective histone deacetylase inhibitors with chemosensitizing properties, J. Med. Chem. 62 (2019) 11260e11279, https://doi.org/10.1021/ acs.jmedchem.9b01489.
[18] J.J. Bandolik, A. Hamacher, C. Schrenk, R. Weishaupt, M.U. Kassack, Class i- histone deacetylase (Hdac) inhibition is superior to pan-hdac inhibition in modulating cisplatin potency in high grade serous ovarian cancer cell lines, Int. J. Mol. Sci. 20 (2019), https://doi.org/10.3390/ijms20123052.
[19] F.-I. Huang, Y.-W. Wu, T.-Y. Sung, J.-P. Liou, M.-H. Lin, S.-L. Pan, C.-R. Yang, MPT0G413, A novel HDAC6-selective inhibitor, and bortezomib synergisti- cally exert anti-tumor activity in multiple myeloma cells, Front. Oncol. 9 (2019) 249, https://doi.org/10.3389/fonc.2019.00249.
[20] K.F. Hui, A.K.S. Chiang, Combination of proteasome and class I HDAC inhibitors induces apoptosis of NPC cells through an HDAC6-independent ER stress- induced mechanism, Int. J. Canc. 135 (2014) 2950e2961, https://doi.org/ 10.1002/ijc.28924.
[21] T.C. Chou, Drug combination studies and their synergy quantiﬁcation using the chou-talalay method, Canc. Res. 70 (2010) 440e446, https://doi.org/ 10.1158/0008-5472.CAN-09-1947.
[22] T. Hideshima, P.G. Richardson, K.C. Anderson, Mechanism of action of pro- teasome inhibitors and deacetylase inhibitors and the biological basis of synergy in multiple myeloma, Mol. Canc. Therapeut. 10 (2011) 2034e2042, https://doi.org/10.1158/1535-7163.MCT-11-0433.
[23] Y. Messinger, P. Gaynon, E. Raetz, R. Hutchinson, S. DuBois, J. Glade-Bender, R. Sposto, J. van der Giessen, E. Eckroth, B.C. Bostrom, Phase I study of bor- tezomib combined with chemotherapy in children with relapsed childhood acute lymphoblastic leukemia (ALL): a report from the therapeutic advances in childhood leukemia (TACL) consortium, Pediatr. Blood Canc. 55 (2010) 254e259, https://doi.org/10.1002/pbc.22456.
[24] Y.H. Messinger, P.S. Gaynon, R. Sposto, J. Van Der Giessen, E. Eckroth, J. Malvar, B.C. Bostrom, Bortezomib with chemotherapy is highly active in advanced B- precursor acute lymphoblastic leukemia: therapeutic Advances in Childhood Leukemia & Lymphoma (TACL) Study, Blood 120 (2012) 285e290, https:// doi.org/10.1182/blood-2012-04-418640.
[25] S. Bhatia, V. Krieger, M. Groll, J.D. Osko, N. Reßing, H. Ahlert, A. Borkhardt, T. Kurz, D.W. Christianson, J. Hauer, F.K. Hansen, Discovery of the ﬁrst-in-class dual histone deacetylase-proteasome inhibitor, J. Med. Chem. 61 (2018) 10299e10309, https://doi.org/10.1021/acs.jmedchem.8b01487.
[26] N. Reßing, M. So€nnichsen, J. Osko, A. Scho€ler, J. Schliehe-Diecks, A.J. Skerhut, A. Borkhardt, J. Hauer, M.U. Kassack, D.W. Christianson, S. Bhatia, F.K. Hansen, Multicomponent synthesis, binding mode and structure-activity relationships of selective histone deacetylase 6 (HDAC6) inhibitors with bifurcated capping groups, J. Med. Chem. (2020), https://doi.org/10.1021/acs.jmedchem.9b01888 acs.jmedchem.9b01888.
[27] G.Y. Di Veroli, C. Fornari, D. Wang, S. Mollard, J.L. Bramhall, F.M. Richards, D.I. Jodrell, Combeneﬁt, An interactive platform for the analysis and visuali- zation of drug combinations, Bioinformatics 32 (2016) 2866e2868, https:// doi.org/10.1093/bioinformatics/btw230.
[28] H.M. Hesham, D.S. Lasheen, K.A.M. Abouzid, Chimeric HDAC inhibitors: comprehensive review on the HDAC-based strategies developed to combat cancer, Med. Res. Rev. 38 (2018) 2058e2109, https://doi.org/10.1002/ med.21505.
[29] N. Blagitko-Dorfs, P. Schlosser, G. Greve, D. Pfeifer, R. Meier, A. Baude, D. Brocks, C. Plass, M. Lübbert, Combination treatment of acute myeloid leu- kemia cells with DNMT and HDAC inhibitors: predominant synergistic gene downregulation associated with gene body demethylation, Leukemia 33 (2019) 945e956, https://doi.org/10.1038/s41375-018-0293-8.
[30] A. Ahmadzadeh, E. Khodadi, M. Shahjahani, J. Bertacchini, T. Vosoughi, N. Saki, The role of HDACs as leukemia therapy targets using HDI, Int. J. Hematol. Stem Cell Res. 9 (2015) 203e214.
[31] P. Wang, Z. Wang, J. Liu, Role of HDACs in normal and malignant hemato- poiesis, Mol. Canc. 19 (2020) 5, https://doi.org/10.1186/s12943-019-1127-7.
[32] San Jose´-Ene´riz, Gimenez-Camino, Agirre, Prosper, HDAC inhibitors in acute myeloid leukemia, Cancers 11 (2019) 1794, https://doi.org/10.3390/ cancers11111794.
[33] M.C. Stubbs, W. Kim, M. Bariteau, T. Davis, S. Vempati, J. Minehart, M. Witkin, J. Qi, A.V. Krivtsov, J.E. Bradner, A.L. Kung, S.A. Armstrong, Selective inhibition of HDAC1 and HDAC2 as a potential therapeutic option for B-ALL, Clin. Canc. Res. 21 (2015) 2348e2358, https://doi.org/10.1158/1078-0432.CCR-14-1290.
[34] D.T. Vogl, N. Raje, S. Jagannath, P. Richardson, P. Hari, R. Orlowski, J.G. Supko, D. Tamang, M. Yang, S.S. Jones, C. Wheeler, R.J. Markelewicz, S. Lonial, Ric- olinostat, the ﬁrst selective histone deacetylase 6 inhibitor, in combination with bortezomib and dexamethasone for relapsed or refractory multiple myeloma, Clin. Canc. Res. 23 (2017) 3307e3315, https://doi.org/10.1158/ 1078-0432.CCR-16-2526.
[35] L.H. Engelke, A. Hamacher, P. Proksch, M.U. Kassack, Ellagic acid and resver- atrol prevent the development of cisplatin resistance in the epithelial ovarian cancer cell line A2780, J. Canc. 7 (2016) 353e363, https://doi.org/10.7150/ jca.13754.
[36] T. Ciossek, H. Julius, H. Wieland, T. Maier, T. Beckers, A homogeneous cellular histone deacetylase assay suitable for compound proﬁling and robotic screening, Anal. Biochem. 372 (2008) 72e81, https://doi.org/10.1016/ J.AB.2007.07.024.
[37] C. Bonﬁls, A. Kalita, M. Dubay, L.L. Siu, M.A. Carducci, G. Reid, R.E. Martell, J.M. Besterman, Z. Li, Evaluation of the pharmacodynamic effects of MGCD0103 from preclinical models to human using a novel HDAC enzyme assay, Clin. Canc. Res. 14 (2008) 3441e3449, https://doi.org/10.1158/1078- 0432.CCR-07-4427.