ORIGINAL RESEARCH published: 30 April 2020 doi: 10.3389/fonc.2020.00657 Sirtuin 1 Inhibiting Thiocyanates (S1th)—A New Class of Isotype Selective Inhibitors of NAD+ Dependent Lysine Deacetylases Nathalie Wössner 1 , Zayan Alhalabi 2 , Jessica González 3 , Sören Swyter 1 , Jin Gan 4 , Karin Schmidtkunz 1 , Lin Zhang 5 , Alejandro Vaquero 3 , Huib Ovaa 4 , Oliver Einsle 5 , Wolfgang Sippl 2 and Manfred Jung 1* 1 Department of Pharmaceutical and Medicinal Chemistry, Institute of Pharmaceutical Sciences, University of Freiburg, Freiburg im Breisgau, Germany, 2 Department of Medicinal Chemistry, Institute of Pharmacy, University of Halle-Wittenberg, Halle, Germany, 3 Chromatin Biology Laboratory, Josep Carreras Leukaemia Research Institute (IJC), Badalona, Barcelona, Spain, 4 Department of Cell and Chemical Biology, Oncode Institute, Leiden University Medical Center, Leiden, Netherlands, 5 Department of Protein Crystallography, Institute of Biochemistry, University of Freiburg, Freiburg im Breisgau, Germany Edited by: Sirtuin 1 (Sirt1) is a NAD+ dependent lysine deacetylase associated with the Cristian Scatena, University of Pisa, Italy pathogenesis of various diseases including cancer. In many cancer types Sirt1 expression Reviewed by: is increased and higher levels have been associated with metastasis and poor prognosis. Sumit Sahni, However, it was also shown, that Sirt1 can have tumor suppressing properties and Kolling Institute of Medical Research, Australia in some instances even a dual role for the same cancer type has been reported. Sujit Kumar Bhutia, Increased Sirt1 activity has been linked to extension of the life span of cells, respectively, National Institute of Technology organisms by promoting DNA repair processes and downregulation of tumor suppressor Rourkela, India proteins. This may have the downside of enhancing tumor growth and metastasis. *Correspondence: Manfred Jung In mice embryonic fibroblasts depletion of Sirt1 was shown to decrease levels of the manfred.jung@ DNA damage sensor histone H2AX. Impairment of DNA repair mechanisms by Sirt1 pharmazie.uni-freiburg.de can promote tumorigenesis but also lower chemoresistance toward DNA targeting Specialty section: therapies. Despite many biological studies, there is currently just one small molecule Sirt1 This article was submitted to inhibitor in clinical trials. Selisistat (EX-527) reached phase III clinical trials for treatment Cancer Metabolism, of Huntington’s Disease. New small molecule Sirt1 modulators are crucial for further a section of the journal Frontiers in Oncology investigation of the contradicting roles of Sirt1 in cancer. We tested a small library of Received: 09 December 2019 commercially available compounds that were proposed by virtual screening and docking Accepted: 08 April 2020 studies against Sirt1, 2 and 3. A thienopyrimidone featuring a phenyl thiocyanate moiety Published: 30 April 2020 was found to selectively inhibit Sirt1 with an IC50 of 13 µM. Structural analogs lacking the Citation: Wössner N, Alhalabi Z, González J, thiocyanate function did not show inhibition of Sirt1 revealing this group as key for the Swyter S, Gan J, Schmidtkunz K, selectivity and affinity toward Sirt1. Further analogs with higher solubility were identified Zhang L, Vaquero A, Ovaa H, Einsle O, through iterative docking studies and in vitro testing. The most active compounds (down Sippl W and Jung M (2020) Sirtuin 1 Inhibiting Thiocyanates (S1th)—A New to 5 µM IC50 ) were further studied in cells. The ratio of phosphorylated γH2AX to Class of Isotype Selective Inhibitors of unmodified H2AX is lower when Sirt1 is depleted or inhibited. Our new Sirtuin 1 inhibiting NAD+ Dependent Lysine Deacetylases. Front. Oncol. 10:657. thiocyanates (S1th) lead to similarly lowered γH2AX/H2AX ratios in mouse embryonic doi: 10.3389/fonc.2020.00657 fibroblasts as Sirt1 knockout and treatment with the reference inhibitor EX-527. In Frontiers in Oncology | www.frontiersin.org 1 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) addition to that we were able to show antiproliferative activity, inhibition of migration and colony forming as well as hyperacetylation of Sirt1 targets p53 and H3 by the S1th in cervical cancer cells (HeLa). These results reveal thiocyanates as a promising new class of selective Sirt1 inhibitors. Keywords: sirtuins, lysine deacetylase, thiocyanate, DNA damage, histone, H2AX, p53 INTRODUCTION types both tumor promoting and suppressing actions have been described, for example in prostate cancer. In Sirt1 knockout mice Sirtuins are deacylases able to cleave off acetyl and also increased cell proliferation of prostatic intraepithelial neoplasia longer chain acyl groups from the ε-amino residue of lysines was observed, implicating Sirt1 as a tumor suppressor (24). in histones and non-histone proteins in a NAD+ -dependent However, via global transcriptome analysis increased levels of manner. For their activity on histones they have been designated Sirt1 were identified as a key biomarker for prostate cancer to form class III of the histone deacetylases (HDACs). Generally, suggesting a tumor promoting influence of Sirt1 (25). Several HDACs are known to deacetylate resp. deacylate various protein important tumor suppressors like p53, FOXO3a, or E2F1 that substrates and for sirtuins many findings actually focus on induce apoptosis in malignant cells (e.g., in breast cancer) are non-histone substrates (1). The human genome contains seven deacetylated by Sirt1, and thereby inactivated, promoting cell sirtuins isotypes (Sirt1-7) that differ in substrate spectrum and survival (26–29). In breast cancer tissue elevated Sirt1 expression localization (2). With regard to drug discovery the majority of correlates with tumor size, high histological grades and lymph efforts has focused on Sirt2 where many small molecule inhibitors node metastasis (30). Nevertheless, Sirt1 can still act as a tumor are now present. Other isotypes have been addressed as well but suppressor in breast cancer cells as well. It is crucial for DNA only for Sirt1 a drug is currently in clinical trials (see below) damage response, regulates several DNA repair enzymes like (3, 4). Due to conflicting evidence of roles of sirtuin isotypes Ku70 and thereby enables stable, efficient DNA repair (31, 32). in different diseases both activators and inhibitors have been Some known oncogenes like NF-κB are directly deacetylated by investigated (5–7). Sirt1 promoting downregulation of the NF-κB-dependent cell Sirt2 has been discussed as an anticancer target but also tumor survival pathway (33). suppressive activities of Sirt2 have been mentioned (8–15). These In some cancer types (e.g., glioma, bladder, or ovarian cancer) effects might be tissue or organ specific which complicates drug lower expression levels of Sirt1 have been detected, although in development and no Sirt2 inhibitor has reached clinical trials most cancer types an increased expression was observed (34). yet. Sirt6 has been identified as a tumor suppressor and hence A meta-analysis showed that Sirt1 overexpression significantly Sirt6 activators are the focus of drug discovery efforts regarding correlates with poor prognosis in solid tumors (35). Anticancer this phenotype (16, 17). But also tumor promoting effects of effects of sirtuin inhibitors have been described on a cellular Sirt6 have been described (18). In addition, Sirt6 inhibition level. In cervical cancer cells EX-527 induced cell death while might be a way to increase the efficacy of cytostatic drugs with inhibition of the isotype Sirt2 led to cell cycle arrest. In the DNA-damaging mode of action (19). breast carcinoma cell line MCF-7 though, the opposite effect Regarding Sirt1, it has been studied strongly in the context of was observed, Sirt1 inhibition by EX-527 led to cell cycle arrest neurodegenerative diseases and Selisistat [EX-527 (1), Figure 1] while treatment with Sirtinol or Salermide (Sirt1/2 inhibitors is a potent and selective Sirt1 inhibitor that was undergoing with a stronger effect on Sirt2) resulted in cell death (36, clinical testing in Huntington’s disease (20). A strong focus 37). In melanoma, chronic lymphocytic leukemia as well as attention was dedicated to potentially lifespan extending Sirt1 hepatocellular carcinoma cell lines both Sirt1 inhibitors (EX-527) activators. The initial studies mostly dealt with resveratrol, and Sirt2 inhibitors impaired cell growth and viability (38–40). a natural product with pleiotropic activities which makes As already mentioned, Sirt1 plays a pivotal role in DNA mechanistic studies difficult. Later, drug-like sirtuin activating damage response (DDR). For example, phosphorylation of compounds (STACs) were identified and went into clinical the DNA damage sensor H2AX which gets phosphorylated trials. While the discussion on the relevance and robustness of to γH2AX upon double strand breaks (DSBs) in healthy activation of deacetylation by Sirt1 has been very controversial, cells is significantly downregulated when Sirt1 is inhibited some effects of resveratrol have indeed been linked to Sirt1 in or depleted (41, 42). DDR can be regulated through Sirt1 animal models and depending on the deacetylation substrate, either by direct histone deacetylation which changes chromatin activation has been proven in reliable biochemical assays (21). compaction or by activation and inactivation of non-histone With regard to cancer, Sirt1 was postulated as an anticancer proteins that are involved in the major DNA repair mechanisms: target, e.g., promoting epithelial-to-mesenchymal transition homologous recombination (HR), non-homologs end joining (EMT) in many cancer types (22) but it was also reported (NHEJ), nucleotide excision repair (NER), mismatch repair to suppress EMT in types like ovarian cancer. Yang et al. (MMR), and base excision repair (BER) (43–47). Due to postulated that the conflicting behavior of Sirt1 in cancer cells Sirt1s implications in DNA damage sensing and recruitment may depend on its subcellular localization (23). In some cancer of repair proteins it can alter resistance toward some cancer Frontiers in Oncology | www.frontiersin.org 2 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 1 | Structures of S1th 1, 7, S1th 10–13, the pan-DUB inhibitor PR-619 and the reference Sirt1 inhibitor EX-527. therapies targeting DNA stability. Application of EX-527 reduced the role of Sirt1 in different cancer types and to better examine chemoresistance in endometrial carcinoma cells (48). its therapeutic potential in cancer further Sirt1 inhibitors are EX-527 is widely used for studies on the effects of Sirt1 needed. Here we present a new class of sirtuin 1 inhibitors inhibition in cells and living organisms, partially due to a lack based on a specific interaction with a thiocyanate moiety of other selective Sirt1 inhibitors. The only two compounds that that lead to altered γH2AX/H2AX ratios in mouse embryonic show similar potency toward Sirt1 are its analog CHIC-35 and fibroblast cells. the Suramin analog NF675 (49, 50). However, both of them show slightly less selectivity over Sirt2 (51) and Suramins are MATERIALS AND METHODS usually poorly cell permeable. Other than that, only few specific Sirt1 inhibitors have been identified. Several examples with low All test compounds are commercially available and were micro molar affinity exist, like certain Splitomicin derivatives purchased from Princeton BioMolecular Research, Sigma or the so-called spiro series (52, 53). To get more insights into Aldrich, Enamine, or Chembridge and used as received. Frontiers in Oncology | www.frontiersin.org 3 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) Enzyme Expression and Purification fluorescence intensity was measured in a microplate reader Recombinant human Sirt537−310 was purchased readily in Tris (λEx = 390 nm, λEm = 460 nm, BMG POLARstar Optima, Buffer [25 mM Tris, 100 mM NaCl, 5 mM DTT, 10% glycerol BMG Labtech, Germany). An enzyme-free blank control and (v/v)] from Enzo Life Sciences (NY, USA). Human Sirt3118−395 a 100% conversion control containing AMC instead of ZMAL was expressed and purified as described previously (54). For were measured in addition. Inhibition was calculated in % in expression of human Sirt1134−747 and Sirt256−356 chemically relation to a DMSO control after blank signal subtraction and competent cells of E. coli BL21 Star (DE3) were transformed IC50 values were determined using a non-linear regression to fit with the expression vectors pET30S-hSirt1134−747 or pET30S- the dose-response curve with OriginPro 9G (OriginLab, USA). hSirt256−356 . Bacteria were grown at 37◦ C in 2×YT medium Pre-tests as well as IC50 determination was carried out at least supplemented with 50 µg·mL−1 of kanamycin to an OD600 of 0.6. twice in duplicates. Then isopropyl-β-D-1-thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM to induce gene expression. Homogeneous ZKsA-Based Fluorescence Assay for After further cultivation at 20◦ C for 12 h, the cells were Sirt5 harvested by centrifugation for 15 min at 5,000 g. The cells were Inhibition of Sirt5 was measured using a general procedure resuspended in lysis buffer [100 mM Tris/HCl buffer at pH described before with small modifications (58). Sirt5 was 8.0, 150 mM NaCl and 10% (v/v) glycerol] and disrupted by purchased from Enzo Life Science (NY, USA) and used as ultrasonication (Branson Digital Sonifier 250) at 70% amplitude received. Z-Lys(succ)-aminomethyl coumarin (ZKsA) was used for 10 min (3 s working, 10 s pause). The crude extract was as a substrate for Sirt5 mediated desuccinylation. In black cleared by centrifugation at 100,000 g for 1 h, and the supernatant 384-well non-binding plates (Greiner Bio-One, Monroe, NC) was loaded onto a Strep-Tactin Superflow cartridge (5 ml Sirt5 was mixed with 2 µL of ZKsA (1 mM stock solution in bed volume, IBA Lifescience, Germany). Target proteins were assay buffer [(50 mM Tris·HCl, 137 mM NaCl, 2.7 mM KCl, eluted with lysis buffer containing 5 mM D-Desthiobiotin (IBA 1 mM MgCl2, pH 8.0, and 0.1% PEG8000), 100 µM final assay Lifescience, Germany) and further separated by size-exclusion concentration], 1 µL of Inhibitor dissolved in DMSO or DMSO chromatography (Superdex S200 26/60, GE Healthcare, IL, USA) as a control [final DMSO concentration 5% (v/v)], 2 µL NAD+ equilibrated in Tris/HCl buffer (20 mM, 150 mM NaCl, pH (5 mM stock solution in assay buffer, final assay concentration 8.0). Pure protein was concentrated by ultrafiltration, flash- 500 µM) and filled up to a total volume of 20 µL with assay frozen in liquid nitrogen and stored at −80◦ C until further buffer. The mixture was incubated for 1 h at 37◦ C and 140 use. Identity as well as purity were verified by SDS-PAGE (55) rpm before 4 µL of trypsin solution (50 mM Tris, 100 mM and protein concentration was determined by the bicinchoninic NaCl, pH 8.0, 6 mg/mL trypsin buffer, 1 mg/mL final assay acid (BCA) method, using bovine serum albumin (BSA) as a concentration) was added to stop the enzymatic reaction. After standard (56). Deacetylase activity was confirmed to be NAD+ - 2 min of incubation at 37◦ C and 140 rpm fluorescence intensity dependent and could be inhibited with the physiological sirtuin was detected as described above. Sirt5 concentration was adjusted inhibitor nicotinamide. to 15–30% substrate conversion. A negative control without enzyme and a positive control containing AMC instead of ZKsA In vitro Characterization were performed as well. Inhibition was calculated in % in relation Homogeneous ZMAL-Based Fluorescence Assay for to a DMSO control and was determined in triplicates. Class I Sirtuins All compounds were tested in the trypsin-coupled Homogeneous ZMAL-Based Fluorescence Assay for high-throughput ZMAL-assay in black 96-well plates HDAC1 and 6 (OptiPlateTM−96F, black, 96 well, Pinch bar design, The inhibition of HDAC1 and 6 by the S1th and PR-619 was PerkinElmer, USA), using ZMAL (Z-Lys(acetyl)-AMC) as a determined via the ZMAL-assay according to the same general substrate (57). Sirt1134−747 , Sirt256−356 , and Sirt3118−395 were procedure described for sirtuins (2.2.1). 5 µL substrate (10.5 µM mixed with 5 µL substrate (10.5 µM final assay concentration, final assay concentration) were mixed with 3 µL of inhibitor in diluted from a 12.6 mM stock in DMSO) and 3 µL Inhibitor in DMSO or DMSO and 10 µL of HDAC solution (concentration DMSO at various concentrations or DMSO as a control [final adjusted to 15–30% conversion) and filled up to 60 µL with assay DMSO concentration 5% (v/v)]. The mixture was supplemented buffer (50 mM Tris, 137 mM NaCl, 1 mM MgCl2, 2,7 mM KCl, 1 with assay buffer (50 mM Tris/HCl, 137 mM NaCl, 2.7 mM KCl, mg/ml BSA, pH = 8.0). The mixture was incubated at 37◦ C and 1 mM MgCL2 , pH 8.0) to a total volume of 55 µL. Enzyme 140 rpm for 1.5 h. For the stop solution Trichostatin A (3.3 µM) concentration was adjusted to get a final conversion of 20–30%. was used instead of nicotinamide. Fluorescence intensity was Addition of 5 µL NAD+ (6 mM in assay buffer, final assay measured as described above. An enzyme free blank control and concentration of 500 µM) initiated the catalytic reaction and the an AMC containing 100% control were performed as well and plates were incubated at 37◦ C for 4 h with agitation at 140 rpm. inhibition was calculated in % in relation to a DMSO control. The catalytic reaction was stopped by addition of 60 µL stop solution [50 mM Tris, 100 mM NaCl, 6.7% (V/V) DMSO, 5.5 Jump Dilution Assay U/µL trypsin, 8 µM nicotinamide, pH 8.0]. The plate was again To test the compounds for reversibility a jump dilution assay incubated at 37◦ C and 140 rpm for 20 min to release free AMC based on the trypsin-coupled ZMAL assay described above was from the deacetylated substrate by trypsin digestion. Afterwards, used. In black 96 well plates (OptiPlateTM−96F, black, 96 well, Frontiers in Oncology | www.frontiersin.org 4 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) Pinch bar design, PerkinElmer, USA) 1 µL Sirt1134−747 in assay Sirt1134−747 (3 mg/mL, final assay conversion 20–30%) was buffer (50 mM Tris/HCl, 137 mM NaCl, 2.7 mM KCl, 1 mM mixed with 5 µL of the diluted peptide and 3 µL of inhibitor MgCl2 , Sirt1 concentration 100-fold higher than normally used) dissolved in DMSO in various concentrations or DMSO as with 1 µL inhibitor in a concentration 10-fold higher than the a control [final DMSO concentration 5% (v/v)] and filled up IC50 were preincubated at room temperature for 10 min before to 55 µL with assay buffer. After addition of 5 µL NAD+ the mixture was rapidly diluted 100-fold with assay mix [assay (6 mM, final assay concentration 500 µM) to start the enzymatic buffer, ZMAL (final assay concentration 500 µM), DMSO (final reaction the plate was incubated for 1 h at 37◦ C and 140 rpm. assay concentration 5% (v/v), NAD+ (final assay concentration Afterwards pH was adjusted with 5 µL of 0.1 M NaOH and 500 µM), pH 8.0] to a total volume of 200 µL. The reaction the enzymatic reaction was stopped with stopping solution was stopped after 2.5, 5, 7.5, 10, 12.5, 15, 20, 25, and 30 min [8.73 mM nicotinamide (final assay concentration 4 mM), by addition of 60 µL stop solution [50 mM Tris, 100 mM NaCl, 0.455 mM fluorescamine (final assay concentration 62.5 µM) 6.7% (V/V) DMSO, trypsin 5.5 U/µL, 8 µM nicotinamide, pH in acetone]. The fluorescence signal was measured using 8.0]. The plate was again incubated at 37◦ C and 140 rpm for a microplate reader (λEx = 390 nm, λEm = 485 nm, BMG 20 min to get the free AMC from the deacetylated substrate by POLARstar Optima, BMG Labtech, Germany). Additionally, trypsin digest. Afterwards fluorescence intensity was measured a 100% inhibition control containing the physiological sirtuin in a microplate reader (λEx = 390 nm, λEm = 460 nm, BMG inhibitor nicotinamide (6 mM final assay concentration) and a POLARstar Optima, BMG Labtech, Germany). An enzyme-free control simulating 100% conversion containing the deacetylated blank control and a 100% conversion control containing AMC FOXO3a peptide (Ac-DSPSQLSKWPPGTSS-NH2, custom instead of ZMAL were performed additionally. Conversion was synthesized by PSL, Heidelberg, Germany) in equivalent calculated in % in relation to the 100% control after subtraction concentration as the substrate were performed. Inhibition was of the blank fluorescence signal. Conversion in % was plotted calculated in % in relation to DMSO control after subtraction against time with OriginPro 9G (OriginLab, USA) and an of the 100% inhibition fluorescence signal. IC50 values were exponential fit was performed to fit the curves. determined using a non-linear regression to fit the dose-response curve with OriginPro 9G (OriginLab, USA). NAD+ Competition Assay NAD+ competition was determined using the trypsin coupled Fluorescent Thermal Shift Assay (FTSA) ZMAL assay. Sirt1134−747 , the ZMAL substrate and the inhibitors In a 96-well plate (Hard-Shell R PCR-plates, 96-well, thin- or DMSO were mixed and filled up to 55 µL in black 96-well wall, BioRAD, USA) 14 µL Sirt1134−747 in assay buffer (0.3 plates as described above. The reaction was initiated by the mg/mL final assay concentration, 50 mM Tris/HCl, 137 mM addition of 5 µL of NAD+ in concentrations ranging from 62.5 NaCl, 2.7 mM KCl, 1 mM MgCL2 , pH 8.0) were mixed with to 2,000 µM final assay concentration. After 4 h of incubation 1 µL inhibitor in DMSO (100 µM) or DMSO [final assay at 37◦ C and 140 rpm the reaction was stopped by addition of concentration 5% (v/v)] as a control and 5 µL SyproOrange the trypsin containing 60 µL stop solution, incubated again for (1:100 in assay buffer). The fluorescence intensity was monitored 20 min at 37◦ C and 140 rpm and fluorescence was detected during a constant increase of temperature from 20 to 95◦ C, with a BMG POLARstar (λEx = 390 nm, λEm = 460 nm, BMG 1 K per 20 s, using a real-time-PCR-machine (C1000 TouchTM POLARstar Optima, BMG Labtech, Germany). An enzyme-free Thermal Cycler, CFX96TM Real-Time System, BioRAD, USA). blank control and a control containing AMC instead of ZMAL Melting temperatures were calculated via GraphPad Prism were measured additionally. Conversion was calculated in % in according to a published procedure (60). relation to the 100% control (AMC) after subtraction of the blank fluorescence signal. Conversion in % was plotted against Cellular Assays NAD+ concentration with OriginPro 9G (OriginLab, USA) and DUB Labeling on HEK293T Cell Lysate an exponential fit was performed to fit the curves. Pan-inhibition of deubiquitinases (DUBs) was tested using HEK293T cell lysate (human embryonic kidney cells, human). FOXO3a Substrate Competition Assay Cells were harvested and resuspended in HR buffer (50 mM Peptide substrate competition was measured using a Tris, 5 mM MgCl2 , 250 mM sucrose, 1 mM DTT, pH 7.4). homogeneous fluorescence-based fluorescamine assay similar Cell lysis was achieved by sonication (Bioruptor, Diagenode, to that previously reported as an activity assay for Sirt2 and high intensity for 10 min with an ON/OFF cycle of 30 s) at Sirt3 with Ac-α-tubulin as a substrate (59). For Sirt1 a partial 4◦ C and the cell debris was removed by centrifugation at sequence of the physiological Sirt1 substrate FOXO3a [Ac- 13,500 rpm for 15 min. Cell lysate protein concentration was DSPSQLSK(Ac)WPPGTSS-NH2, custom synthesized by PSL, determined with a NanoDrop spectrophotometer (NanoDrop Heidelberg, Germany], hereafter called FOXO3a-ac, was used OneTM Spektrophotometer, Thermo Fisher, MA, USA) by as substrate. Substrates were stored as 1 mM stocks in assay measuring absorbance at 280 nm. Nineteen microliter of the buffer [HEPES 25 mM, NaCl 137 mM, KCl 2.7 mM, MgCl2 lysate (2 µg/µL) were incubated with inhibitors dissolved in 1 mM, Triton-X 100 0.015% (v/v), pH 8.0] and diluted with DMSO at various concentrations or DMSO as a control for assay buffer to 240 µM FOXO3a-ac peptide (20 µM final assay 30 min at 37◦ C. Afterwards 1 µL of Rhodamine-Ub-PA (1 µM concentration). In black 96-well plates (OptiPlateTM−96F, final assay concentration) was added to each sample, followed by black, 96 well, Pinch bar design, PerkinElmer, USA) 0.05 µL further incubation for 40 min at 37◦ C (61). The labeling reaction Frontiers in Oncology | www.frontiersin.org 5 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) was stopped by addition of NuPAGE 4× LDS sample buffer washed with PBS, and cells were fixed with cold methanol for (Invitrogen Life Technologies, Carlsbad, CA, USA) containing β- 5 min at RT. Cells were stained with crystal violet for 10 min mercaptoethanol and boiling for 7 min at 95◦ C. The proteins of at RT and washed with H2 O. Images were acquired using the samples as well as protein marker (PageRulerTM Pre-stained iBright (Thermo Fisher Scientific, MA, USA). For quantification, Protein Ladder, 10–250 kDa, Thermo Fisher, MA, USA) were cells of each well were resuspended in 10% glacial acetic resolved by a 4–12% SDS-PAGE using the NuPAGE system with acid to dissolve crystal violet whose levels were monitored by MOPS running buffer (Invitrogen Life Technologies, Carlsbad, measuring absorbance at 590 nm. The experiment was performed CA, USA). The resulting gel was scanned with a Typhoon imager in duplicates. (GE Healthcare Life Sciences, USA) to visualize the Rhodamine- Ub-PA probe and Cy5 (marker). Subsequently the gel was stained Analysis of Acetylation Status in HeLa Cells by with InstantBlueTM Protein Stain (Expedeon, UK) and scanned Western Blotting on a Amersham scanner (Amersham Typhoon gel and blot 5 × 104 cells were plated in 6-well plates and grown imaging system, GE Healthcare Life Sciences, USA). with the conditions and drug concentrations as described for the proliferation assay. Cells were harvested at 48 h and γH2AX Level Determination resuspendend in protein loading buffer. After sonication the Sirt1 WT and KO MEFS cells were treated with 10 µM EX- samples were centrifuged at 14,000 rpm in an Eppendorf 527, 12 µM S1th 13 (6), 20 µM S1th 12 (5), 30 µM S1th 10 microcentrifuge (Eppendorf, DE) and the supernatants were (7) and 30 µM SirReal2 [Sirt2 inhibitor, synthesized according analyzed by western-blot. The following antibodies were used: to (62)] for 48 h. Oxidative stress was induced using 10 µM histone H3 (Cell Signaling #9715), histone H3K9ac (Cell Camptothecin 2 h. Whole-cell extracts were performed according Signaling # C5B11), GAPDH (Cell Signaling #D16H11), p53 to the Dignam protocol (63). Primary antibodies used for the (ThermoFisher #PA527822), and p53K382ac (Abcam 75754). western blot were anti-H2AX and anti-γH2AX (ab11175 and Quantification was carried out using imageJ. ab2893 resp., Abcam, UK). Densiometric analysis of the western blots was performed with Quantity One software (Bio-Rad Computational Methods Laboratories, Inc., USA). 3D structures of all compounds under study were generated from SMILES strings, and a subsequent energy minimization Cell Proliferation Assay With HeLa or MCF7 Cells was carried out using the MMFF94x force field implemented in 1.5 × 104 cells/well were seeded in 24-well plates and grown Molecular Operating Environment System (MOE) 2014.10 in DMEM 10% in presence of DMSO, 20 µM EX-527, 12 µM (Chemical Computing Group, Montreal, Canada). All S1th 13, 20 µM S1th 12, or 30 µM S1th 10. The growth media compounds were used in their neutral form. A maximum containing the drugs was replaced every 48 h. Cells were collected of 100 conformations were generated for each ligand using the and counted at the indicated times starting 24 h after seeding (0, Conformational Search module implemented in MOE. 24, 48, 72, 96, and 120 h). The experiment was performed twice The structure of Sirt1 protein in complex with NAD+ and in duplicates. the small molecule inhibitor EX-527 was downloaded from the Wound Healing Assay With HeLa Cells Protein Data Bank (PDB ID 4I5I) (64). In addition, crystal For wound healing assay 6 × 105 cells per well were seeded structures of Sirt2 in complex with the EX-527 analog CHIC35 in 6-well plates and grown for 24 h in DMEM containing 10% and ADPR (PDB ID 5D7Q) and Sirt3 in complex with EX- FBS. After 24 h, a lineal gap between cells was created in the 527 and NAD+ (PDB ID 4BV3) were investigated. The protein middle of the plate by scratching the cell monolayer with a structures were prepared by using the Structure Preparation 1 mL pipette tip in the same position in each well. Once the module in MOE. Hydrogen atoms were added, for titratable scratching is done, plates were washed twice with PBS and then amino acids the protonation state was calculated using the grown in DMEM containing 1% FBS and the drugs in the Protonate 3D module in MOE. Protein structures were energy same concentrations as for the proliferation assay. As described minimized using the AMBER99 force field with a tethering force previously, the growth media 1% FBS containing the drugs constant of (3/2) kT/2 (σ = 0.5 Å) for all atoms during the was replaced every 48 h. Images were acquired in an Optimal minimization (65). AM1-BCC charges were used for the studied microscope (Leica microscopes, DE) using Leica Application ligands (66). All molecules except the zinc ion were removed Suite X (LAS X) every 24 h (0, 24, 48, and 72 h) until the gap from the structures. closed. The quantification of the area of the gap at the indicated Protein-ligand docking was performed using GOLD5.6 (67). times was performed with Image J-MRI Wound Healing Tool. For Sirt1, Ser442 was used to define the size of the grid box The experiment was performed only once. (15 Å radius). In case of Sirt2 and Sirt3, the corresponding Ser263 and Ser321 were used, respectively. Ten docking poses Colony Forming Assay With HeLa or MCF7 Cells were calculated for each inhibitor. All other options were left at For colony assays, 50 cells per well were seeded in 6-well their default values. Using the docking setup, the cocrystallized plates and grown in the same conditions with the same drug inhibitors EX-527 and CHIC35 could be correctly docked with concentrations as for the proliferation assays. In this case, the RMSD values below 0.6 Å. Virtual screening was carried out growth media containing the drugs was replaced every 48 h. using program GOLD5.6 and the settings described above. To After 7–10 days, when isolated colonies were formed, wells were decrease calculation time, the Virtual Screening setup was used Frontiers in Oncology | www.frontiersin.org 6 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) within GOLD5.6 and only the top-ranked pose was stored virtually (considering only phenyl thiocyanates) and docked the for further evaluation. In total 16 compounds were purchased resulting 113 thiocyanates to Sirt1. Eight promising hits were and submitted to biochemical testing (eleven compounds from cherry picked, purchased and submitted to a third round of Princeton Biomolecular Research, one from Sigma-Aldrich and biochemical testing (Table 2). Among the eight compounds, four from Enamine, Supplementary Table 1). All docking scores S1th 7 (3) showed increased inhibition compared to S1th 1 displayed in Supplementary Table 1. with an IC50 of 6.34 µM (Table 2) which is also supported by the best docking score. The predicted binding pose of S1th 7 shows two hydrogen bonds to Asp272 and Asn465 (Figure 2B). RESULTS Since we encountered solubility problems of the active hits Docking in cellular testing at higher concentrations, we purchased six Based on a previously collected library of putative sirtuin more polar compounds and submitted them to a fourth round inhibitors we carried out a virtual screening using the GOLD5.6 of biochemical testing. S1th 11 (4), 12 (5), and 13 (6) (from docking software and the available crystal structure of Sirt1 in Enamine) were found to be better soluble and equally active as complex with the inhibitor EX-527 and NAD+ (64). A first S1th 7. The binding mode of the active hit S1th 13 (IC50 of in vitro screening (for primary in vitro screening data see 5.2 µM, Table 3) is similar to that observed for the previous hits Supplementary Table 2) on the three class I sirtuins indicated showing interactions with Cys482, Asn465 and in addition to a thienopyrimidone carrying a thiocyanate moiety which we Asp272 (Figure 2C). termed S1th 1 (2) (structure Figure 1) as a promising hit for All active inhibitors, both from the 3rd and the 4th round, Sirt1 inhibition. Two structural analogs of S1th 1 with the retained the extremely high selectivity of the initial hit for Sirt1 same heterocyclic system but different functional groups were over the isotypes Sirt2 and 3. Docking to Sirt2 and Sirt3 was docked to the NAD+ binding site of Sirt1 and were submitted subsequently carried out for the active hits in order to rationalize to a second round of biochemical testing (Table 1). Among the observed selectivity. In case of Sirt2, there are three different the three selected thienopyrimidones only S1th 1 was able to amino acid residues in the putative thiocyanate binding pocket inhibit Sirt1 in the micro molar range (IC50 13 µM). The docking that affect the docking results (Figure 2D): Val266, Arg466, and pose of the active analog showed that the phenyl thiocyanate Asp481 of Sirt1 are substituted by Thr89, Lys287, and Glu323 in moiety is located in the adenine pocket, engaging in a hydrogen the Sirt2 structure, respectively. The inhibitors could be docked bond to Cys482 (backbone NH, Figure 2A). The pyrimidine in a similar orientation to the Sirt2 binding pocket, however ring is located in the polar phosphate pocket of Sirt1 and with less favorable docking scores. Especially the interaction shows a hydrogen bond to Ser442. A third hydrogen bond was of the thiocyanate phenyl ring with Arg466 is lost in case of observed between the amide group of the inhibitor and Gln445. Sirt2, which might explain the lower docking scores. In case The two inactive analogs (OSSK_338451 and OSSK_531963, of Sirt3, there are four amino acid residues substituted in the structures Supplementary Figure 1) showed a similar binding putative binding pocket. Val266, Ser267, Cys482, and Asp483 of mode lacking an interaction in the adenine pocket. Due to the Sirt1 are substituted by Thr150, Pro151, Val366, and Val367 in close proximity of the potentially reactive thiocyanate group of the Sirt3 structure (Figure 2E). Val366, Val367, and Pro151 are S1th 1 and Cys482 we speculated that an irreversible binding restricting the size of the putative thiocyanate binding pocket and might occur. Therefore, we tested the reversibility of enzyme consequently the thiocyanate moiety is not able to interact with inhibition as well as NAD+ competition and competition toward the protein as observed for Sirt1. a peptide analog of the physiological Sirt1 substrate FOXO3a to confirm the proposed binding mode. In vitro results showed In vitro Characterization of S1th that the inhibitors are reversible binders and NAD+ competitive We wanted to further elucidate the binding mode and selectivity but not substrate competitive which is in agreement with the of the S1th. Apart from the closely related class I sirtuins (Sirt2 predicted binding mode of the thiocyanates (Figure 3). and 3) also inhibition of Sirt5 as a representative of other sirtuin To confirm the importance of the thiocyanate group we classes was tested. Although Sirt5 sequence and structure in screened the whole Princeton BioMol. Res. Compound collection general shows less overlap with Sirt1 than the class I sirtuins, the active side residue Cys482 which is explicitly important for S1th binding is conserved in Sirt5. Yet, no inhibition of Sirt5 by S1th 10–13 at 10 µM was observed. In general, thiocyanates TABLE 1 | Inhibition of class I sirtuins by S1th 1 and its two structural analogs are known to act as chelating groups making them candidates OSSK_338451 and OSSK_531963 (2nd round of in vitro screening). for inhibition of ion-dependent enzymes like classical HDACs Compound IC50 [µM] or % inhibition @ 50 µM which feature a zinc ion in their active site. It has been shown that cruciferous vegetable isothiocyanates like sulforaphane can Sirt1 Sirt2 Sirt3 act as potent pan-HDAC inhibitors (68). Consequently, to further S1th 1 13 ± 0.6 23% n.i.* investigate the selectivity of our new thiocyanate Sirt1 inhibitors, OSSK_33845 n.i. n.i. n.i. they were tested against two representative zinc dependent OSSK_531963 n.i. n.i. n.i. HDACs, HDAC1 (Class I) and HDAC6 (Class II). HDAC1 was not inhibited by the tested thiocyanates at 100 µM at all. For *n.i., no inhibition (<10%). HDAC6 a very weak inhibition by S1th 10 (7), 12 and 13 Frontiers in Oncology | www.frontiersin.org 7 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 2 | Sirt1 ribbon is colored purple, hydrogen bonds are shown as dashed lines. (A) Docking pose obtained for S1th 1 (colored beige) at the Sirt1 NAD+ binding pocket. (B) Docking pose obtained for S1th 7 (colored cyan) at the Sirt1 NAD+ binding pocket, NAD+ (colored orange) is shown for comparison. (C) Docking pose obtained for S1th 13 (colored green) at the Sirt1 NAD+ binding pocket. The molecular surface of the binding pocket is colored according to the hydrophobicity (hydrophobic = green, hydrophilic = magenta). (D) Superimposition of the crystal structure of Sirt1 (PDB ID 4I5I, magenta colored ribbon, white colored residues) with docked S1th 13 (colored green) and the crystal structure of Sirt2 (PDB ID 5D7Q, brown colored ribbon, pink colored residues). The amino acid residues of Sirt2 that are different are colored cyan. (E) Superimposition of the crystal structure of Sirt1 (PDB ID 4I5I, magenta colored ribbon, white colored residues) with docked S1th 13 (colored green) and the crystal structure of Sirt3 (PDB ID 4BV3, orange colored ribbon, salmon colored residues). The amino acid residues of Sirt3 that are different are colored cyan. (19, 36, and 29% at 100 µM) was observed (Table 4). Another performed. Binding of a ligand usually results in a stabilization likely off-target effect of thiocyanates could be the inhibition of the protein which can be observed as an increase of the of deubiquitinases (DUBs). Deubiquitination activity of these melting temperature (Tm ) in FTSA. For example, binding of enzymes relies on a cysteine residue in the catalytic core which the NAD+ metabolite adenosine diphosphate ribose (ADPR), acts as a nucleophile (69). Most known DUB inhibitors therefore which is formed during the catalytic reaction of the sirtuins, feature a functional group that can form specific interactions with leads to a stabilization up to 4 K. The known Sirt1 binders this cysteine. PR-619 (8), a compound featuring two aromatic EX-527 and SRT1720 (72) only cause a very small shift or no thiocyanate moieties, was shown to inhibit more than twenty shift at all (Supplementary Figure 3a). For the S1th however DUBs with IC50 values ranging from 5 to 20 µM and, as we could we observed a strong left shift of the Tm from 5.9 up to 9.6 K show, is also able to inhibit Sirt1 with an IC50 of 2.7 ± 0.2 µM for the most potent inhibitors of the fourth round at 100 µM (70, 71). HDAC1 and 6 are inhibited to ∼50% residual activity (Figure 3). The decrease of Tm is concentration dependent for by PR-619 as well. The S1th however did not show inhibition all tested compounds (S1th 10–13), resulting in a smaller 1T at of DUBs which we could show in a fluorescence based activity lower concentrations (Supplementary Figure 3c). Shifts of Tm to assay using HEK-293 cell lysate and a Rhodamine-Ub-PA probe decreased temperatures have been associated with an apparent (Supplementary Figure 2). These results support the excellent destabilization of the protein by covalent binding compounds, selectivity of the S1th not only amongst sirtuins but also over detergent-like compounds or compounds that extract stabilizing other possible off-targets. ions from the structure in several cases (73). This could be a To better understand the binding of the S1th a series hint toward a covalent interaction of the thiocyanate moiety of fluorescent thermal shift assay (FTSA) experiments was with Cys482. However, another widely recognized explanation Frontiers in Oncology | www.frontiersin.org 8 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 3 | In vitro characterization of the thiocyanates. Evaluation of reversibility, competition with NAD+ or the peptide substrate FOXO3a and enzyme stabilization; (A) Sirt1 and the respective thiocyanate inhibitor (50 µM) were preincubated for 10 min and then rapidly diluted 1:100 with assay solution (buffer, ZMAL, NAD+ ), activity of Sirt1 was measured at different time points. Increasing activity can be observed for non-covalent inhibitors; (B) thiocyanate inhibitors (100 µM) were incubated with increasing concentrations of NAD+ , showing that the inhibitors can be replaced by NAD+ ; (C) binding of S1th 10–13 (100 µM) to Sirt1 leads to a strong decrease of Tm ; (D) S1th (100 µM) were incubated with increasing concentrations of the peptide substrate FOXO3a, the inhibitors are not competitive toward FOXO3a. All biochemical assays were performed at least twice in duplicates. for left shifts is that these compounds bind more strongly to a metabolite ADPR. Simultaneous application of ADPR and the conformation different from the native one (74, 75). In regard to S1th resulted in a significantly smaller decrease of Tm (Sirt1) this, one has to take into account that Tm is considerably more than treatment with S1th alone (Supplementary Figure 3b). To affected by entropy than by enthalpy. Consequently, enthalpy- ensure that the observation of a left shift is specific for Sirt1 the driven binding to the native state can be outnumbered by compounds were also tested in an FTSA using Sirt2. There was weaker entropy-driven binding to a different conformation or no decrease of Tm observed, in fact binding of the thiocyanates even the denatured state resulting in a left shift (76). The leads to a very small positive shift (0.2–0.5 K) of Tm for Sirt2 shift of Tm induced by S1th 10–13 can be reversed through (Supplementary Figure 3d). As the Cys482 residue is conserved addition of ADPR. As the thiocyanates are binding to the NAD+ in five out of the seven human sirtuin isotypes including Sirt2 pocket they are also competitive toward the physiological NAD+ and 5, this data shows a unique and specific binding mode of the Frontiers in Oncology | www.frontiersin.org 9 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) TABLE 2 | Inhibition of class I sirtuins by thiocyanate analogs of 3rd round of of Sirt1 inhibitors. To demonstrate the maximum change in virtual screening (S1th 2–9). phosphorylation possible, the γH2AX/H2AX levels of wild type Compound IC50 [µM] or % inhibition @ 50 µM MEFs and Sirt1 KO MEFs were determined. The effect of S1th 10, 12, and 13 (30, 20, and 12 µM, respectively) was compared to that Sirt1 Sirt2 Sirt3 of EX-527 (10 µM) as a positive control and to the specific Sirt2 inhibitor SirReal2 (30 µM) as a negative control (Figure 4, for S1th 2 6.4 ± 0.7 n.i.* n.i. western blots see Supplementary Figure 4). All three inhibitors S1th 3 2.8 ± 0.5 n.i. n.i. showed a reduction of γH2AX/H2AX levels and the observed S1th 4 5.3 ± 0.5 n.i. n.i. effect is similar to that of EX-527. The effect of the sirtuin 1 S1th 5 9.4 ± 1.4 n.i. n.i. inhibitors becomes even more apparent when the topoisomerase S1th 6 5.1 ± 0.6 n.i. n.i. I inhibitor Camptothecin is added to the cells (78). Treatment S1th 7 6.3 ± 0.5 n.i. n.i. with Camptothecin induces DNA damage which in functional S1th 8 40% n.i. n.i. cells leads to elevated phosphorylation levels of H2AX. When S1th 9 n.i. n.i. n.i. EX-527 or the S1th were administered to these cells again a strong *n.i., no inhibition (<10%). decrease of γH2AX/H2AX was detected. This shows that in cells with higher stress levels through DNA damage, the S1th can significantly alter DDR. TABLE 3 | Inhibition of sirtuins 1, 2 3 and 5 by thiocyanate analgues of 4th round of virtual screening (S1th 10-13) and PR-619. Impact of S1th on Cervical Cancer Cells (HeLa) Compound IC50 [µM] or % inhibition @ 50 µM After showing Sirt1 dependent effects in non-cancerous mouse cells we wanted to investigate whether our new inhibitors Sirt1 Sirt2 Sirt3 Sirt5** have an impact on proliferation, migration and colony forming S1th 10 23 ± 6.0 n.i.* n.i. n.i. properties of human cancer cells. As already mentioned Sirt1 S1th 11 6.3 ± 0.8 n.i. n.i. n.i. can play very contradictory roles in different tissues and even S1th 12 5.9 ± 1.4 n.i. n.i. n.i. within one cancer type. Still, cases have been reported where Sirt1 inhibition impairs cell growth. For cervical cancer cells S1th 13 5.2 ± 1.0 n.i. n.i. n.i. (HeLa) effects on cell proliferation upon administration of EX- PR-619 2.7 ± 0.2 36% n.i. 28% 527 have been reported (79). Based on these findings we chose *n.i., no inhibition (<10%), **inhibition tested @ 10 µM. HeLa cells for further examination of the S1th. The three best characterized inhibitors S1th 10, 12, and 13 as well as EX-527 were administered to HeLa cells and the effects on TABLE 4 | Selectivity over histone deacetylases and deubiquitinases compared to pan-DUB inhibitor PR-619. cell proliferation, migration and colony forming were observed (Figure 5). EX-527, S1th 10 and S1th 13 significantly decreased Compound % inhibition @ 100 µM Qualitative pan-DUB inhibition cell proliferation, with EX-527 being slightly more effective than HDAC1 HDAC6 @ 10 µM @ 50 µM the S1th. S1th12 only showed a mild effect on proliferation. Interestingly, even though having the most impact on cell S1th 10 n.i.* 29% n.i. n.i. proliferation, EX-527 failed to impair migration of HeLa cells S1th 11 n.i. n.i. n.i. n.i. in a wound healing assay within the first 24 h. A monolayer S1th 12 n.i. 36% n.i. n.i. of HeLa cells was plated and a “wound” was introduced by S1th 13 n.i. 19% n.i. n.i. scratching. Through cell migration the cells grow back together PR-619 54% 49% + +++ to heal the wound in the monolayer. Other than EX-527, the S1th all showed good inhibition of cell migration already after *n.i., no inhibition (<10%), + low inhibition, + + + full inhibition. 24 h. S1th 13 appeared to be most effective showing an ∼50% lower reduction of total wound area compared to a control where no drug was applied. After 72 h also EX-527 showed a thiocyanates to the Sirt1 structure that was already proposed by mild effect on wound healing and the inhibition by the S1th the docking to Sirt1, 2 and 3 (77). that could already be observed after 24 h became more apparent. In consistency with the results on proliferation and migration Effects of S1th in Cellular Systems also a colony forming assay proved S1th 13 to be the most Sirt1 Dependent Effects in Mice Embryonic potent drug of our new class. EX-527 completely suppressed Fibroblasts (MEFs) colony forming of HeLa cells. S1th 13 proved to be almost as As described previously by Wang et al., the phosphorylation level potent as EX-527 while for S1th 10 and 12 only mild effects of H2AX in mice embryonic fibroblasts (MEFs) is decreased were observed. Effects of the S1th on proliferation and colony if Sirt1 is depleted or inhibited (42). Phosphorylation of forming were also confirmed in initial studies in the breast H2AX is a key step in DNA damage sensing. Determination cancer cell line MCF7 (Supplementary Figure 5). In addition, we of γH2AX/H2AX levels in MEFs via western blotting was performed western blot analysis of the Sirt1 substrates H3K9ac employed to show target engagement in cells of our new class and p53K382ac (Figure 6). To determine the acetylation status Frontiers in Oncology | www.frontiersin.org 10 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 4 | Quantification of γH2AX/H2AX levels in MEF cells after Sirt1 depletion or incubation with various inhibitors by western blot analysis (original western blots: Supplementary Figure 4) with anti-γH2AX and anit-H2AX. γH2AX/H2AX levels are decreased in Sirt1 KO cells and upon addition of EX-527 (10 µM), S1th 10 (30 µM), S1th 12 (20 µM), or S1th 13 (12 µM). The selective Sirt2 inhibitor SirReal2 (30 µM) does not affect γH2AX/H2AX levels. Addition of Camptothecin increases DNA damage in cells (A) γH2AX/H2AX levels without Camptothecin; (B) γH2AX/H2AX levels after addition of Camptothecin (10 µM). The experiments were performed at least three times each. *p < 0.05, **p < 0.01, ***p < 0.001. binding of α-H3K9ac and α-p53K382ac were compared to α- Interestingly, the Sirt2 melting temperature was not affected H3 and α-p53 binding, respectively. Thereby stable expression at all by the S1th, although the respective cysteine residue of H3 and p53 in all samples was ensured and additionally α- is conserved in Sirt2 (Cys324). These results indicate that no GAPDH was used as a loading control. All three S1th tested were unspecific binding of the S1th’s thiocyanate moiety occurs and able to significantly increase H3K9 and p53K382 acetylation and the interaction between the thiocyanate and Cys482 in Sirt1 is thereby affirm selective Sirt1 inhibition in HeLa cells. For EX-527 highly selective. Further we could demonstrate that S1th are no hyperacetylation of H3K9 but increased acetylation levels of selective over sirtuin isotypes 3 and 5, representatives of HDAC p53K382 were observed. class I and II (HDAC 1 and 6) as well as a set of deubiquitinases (DUBs). Selectivity over DUBs is especially remarkable since DISCUSSION they are known to be inhibited by thiocyanates through binding of a catalytically relevant cysteine residue in the active site To further investigate the role of Sirt1 in cancer, new selective of DUBs. The thiocyanate PR-619 for example is a pan-DUB inhibitors for this isotype will be of great value. However, so far inhibitor and as we showed also inhibits Sirt1 with a low micro only few such inhibitors have been reported. In this study we molar IC50 . After ensuring high selectivity and potency of our identified a new class of selective and potent Sirt1 inhibitors, the new inhibitor class we wanted to prove target engagement in Sirtuin 1 inhibiting Thiocyanates (S1th) by an iterative process cells. S1th 10, 12, and 13 were applied to MEF cells and the of virtual screening and biochemical testing. Molecular docking effect on H2AX phosphorylation was detected. H2AX is a DNA of the S1th to the crystal structure of Sirt1 in complex with damage sensor that gets phosphorylated upon DNA damage. its cofactor NAD+ revealed their putative binding mode. In Phosphorylation of H2AX was significantly decreased by S1th general, the inhibitors are proposed to bind to the NAD+ binding 10, 12, and 13 as well as by the positive control EX-527 while pocket of Sirt1. This could be confirmed in a competition assay, a selective Sirt2 inhibitor (SirReal2) did not show any effect. The showing competition between the inhibitors and the cofactor reduction of γH2AX/H2AX levels observed after application of NAD+ but not toward a peptide substrate analog. The most Sirt1 inhibitors was similar as in Sirt1 KO MEFs. Additional potent inhibitor of this class S1th 13 is thought to engage in treatment with camptothecin, a drug that induces DNA damage two hydrogen bonds with Asn465 and Asp272 and a potentially through inhibition of topoisomerase I and thereby increases covalent interaction with Cys482. As indicated by competition γH2AX/H2AX levels, didn’t surpass the inhibitory effect of the assays binding of the S1th is reversible leading to the conclusion S1th or the positive control but even seemed to increase the that even though the interaction with Cys482 could be covalent efficacy especially for S1th 13. Finally, we tested the effects of it is also fully reversible. Fast reversible covalent inhibitors have S1th on proliferation, migration and colony forming capabilities been reported before e.g., for kinases (80). In thermal shift assays of human cervical cancer cells. The cervical cancer cell line HeLa covalent inhibitors often show a characteristic left shift, as they was treated with S1th 10, 12, 13 and EX-527 as a positive control. can destabilize the thermodynamically most stable conformation S1th 13 showed significant inhibition of proliferation, migration of an enzyme or can stabilize a different confirmation. S1th and colony forming while S1th 10 and 12 only had moderate binding resulted in a strong left shift of Sirt1 melting temperature. effects. EX-527 also showed robust inhibition of cell proliferation Frontiers in Oncology | www.frontiersin.org 11 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 5 | (A) Impact of EX-527 and S1th 10, 12, and 13 on cell migration of HeLa cells in a wound healing assay. All compounds were able to impair cell migration. The S1th showed a clear effect already after 24 h, while a decrease of migration only came apparent after 72 h for the positive control EX-527. This experiment was only carried out once. (B) Impact of S1th 10, 12, and 13 and EX-527 on cell proliferation of HeLa cells. All compounds are able to slow down cell proliferation of HeLa cells, with S1th 12 only showing a mild effect. The experiments were performed at least three times each. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (C) Impact of EX-527 and S1th 10, 12, and 13 on colony forming capabilities of HeLa cells. Colony forming is significantly decreased by EX-527 and S1th 13, but barely affected by S1th 12 and 10. Experiments were carried out in duplicates and quantification was done by staining the colonies with crystal violet and measuring absorbance at 590 nm. and colony forming, however for migration a clear effect became potencies in cells as in vitro. This demonstrates their high apparent only 72 h after treatment but not already after 24 h, as potential to further study the role of Sirt1 in cellular model observed for the S1th. Finally, western blot analysis confirmed systems for cancer research but also in other diseases. The that the effect of S1th in HeLa cells is associated with concomitant structural knowledge obtained from available crystal structures hyperacetylation of H3K9 and p53K382. of Sirt1 and our docking studies can be utilized for future Although the in vitro potency of the S1th is yet lower than inhibitor optimization. Knowing that thiocyanates are able to that of the reference EX-527, their discovery, especially their engage in a specific interaction right in the catalytic core of very high selectivity, still opens up new possibilities. Remarkably, the enzyme, new structures can be designed and synthesized. unlike EX-527 that has strongly decreased potency in the cellular Already now the S1th present a valid alternative to EX-527 for setting as compared to the biochemical assay, they show similar cellular studies. Frontiers in Oncology | www.frontiersin.org 12 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) FIGURE 6 | Western blot analysis of acetylation status of p53K382 and H3K9 in HeLa cells. A α-GAPDH antibody was used as a loading control. Histone H3 and p53 expression was monitored by α-H3 and α-p53 antibodies and was not altered by application of the inhibitors, while α-H3K9ac and α-p53K382ac indicate the impact on acetylation status upon inhibitor treatment. 1 (EX-527) was used as a positive control, however it did not show clear hyperacetylation of H3K9. In contrast, a clear effect on acetylation of these Sirt1 targets is observed when 5, 6, and 7 (S1th 12, 13, 10 resp.) are applied. Quantification of the shown blot is displayed as the ratio between acetylated and deacetylated protein in relation to a control where no inhibitor was applied. DATA AVAILABILITY STATEMENT FUNDING All datasets generated for this study are included in the MJ, WS, and NW thank the DFG (Ju295/14-1, Si868/15- article/Supplementary Material. 1; MJ and NW additionally thank 235777276/GRK1976 and CRC992) for funding. AV thanks the Spanish Ministry AUTHOR CONTRIBUTIONS of Economy and Competitiveness - MINECO SAF2017- 88975R (AV) cofunded by FEDER funds/European Regional MJ and WS contributed conception and design of the study. ZA Development Fund (ERDF) - A Way to Build Europe and the performed the computational studies. HO and JGa performed Catalan government agency AGAUR 2017-SGR148 (AV) and selectivity testing on DUBs. KS performed selectivity testing on the CERCA Programme/Generalitat de Catalunya and the Josep HDACs. SS performed initial in vitro testing. Cellular tests were Carreras Foundation for institutional support. planned by AV and JGo and performed by JGo. OE and LZ designed and executed expression and purification of enzymes SUPPLEMENTARY MATERIAL for in vitro testing. NW performed in vitro studies and wrote major parts of the manuscript. MJ, ZA, and WS wrote sections of The Supplementary Material for this article can be found the manuscript. All authors contributed to manuscript revision, online at: https://www.frontiersin.org/articles/10.3389/fonc. read and approved the submitted version. 2020.00657/full#supplementary-material REFERENCES 7. Lee S-H, Lee J-H, Lee H-Y, Min K-J. Sirtuin signaling in cellular senescence and aging. BMB Rep. (2019) 52:24–34. doi: 10.5483/BMBRep.2019.52.1.290 1. Houtkooper RH, Pirinen E, Auwerx J. Sirtuins as regulators of 8. Park S-H, Zhu Y, Ozden O, Kim H-S, Jiang H, Deng C-X, et al. SIRT2 is metabolism and healthspan. Nat Rev Mol Cell Biol. (2012) 13:225–38. a tumor suppressor that connects aging, acetylome, cell cycle signaling, and doi: 10.1038/nrm3293 carcinogenesis. Transl Cancer Res. (2012) 1:15–21. 2. Schiedel M, Robaa D, Rumpf T, Sippl W, Jung M. The current state of NAD+ 9. Serrano L, Martínez-Redondo P, Marazuela-Duque A, Vazquez BN, Dooley SJ, -dependent histone deacetylases (sirtuins) as novel therapeutic targets. Med Voigt P, et al. The tumor suppressor SirT2 regulates cell cycle progression and Res Rev. (2018) 38:147–200. doi: 10.1002/med.21436 genome stability by modulating the mitotic deposition of H4K20 methylation. 3. Carafa V, Rotili D, Forgione M, Cuomo F, Serretiello E, Hailu GS, et al. Sirtuin Genes Dev. (2013) 27:639–53. doi: 10.1101/gad.211342.112 functions and modulation: from chemistry to the clinic. Clin Epigenetics. 10. Jing H, Hu J, He B, Negrón Abril YL, Stupinski J, Weiser K, et al. (2016) 8:61. doi: 10.1186/s13148-016-0224-3 A SIRT2-selective inhibitor promotes c-Myc oncoprotein degradation 4. Robaa D, Monaldi D, Wössner N, Norio K, Rumpf T, Schiedel M, and exhibits broad anticancer activity. Cancer Cell. (2016) 29:297–310. et al. Opening the selectivity pocket in the human lysine deacetylase doi: 10.1016/j.ccell.2016.02.007 Sirtuin2 - new opportunities, new questions. Chem Rec. (2018) 18:1701–7. 11. Schiedel M, Rumpf T, Karaman B, Lehotzky A, Oláh J, Gerhardt doi: 10.1002/tcr.201800044 S, et al. Aminothiazoles as potent and selective Sirt2 inhibitors: a 5. Dang W. The controversial world of sirtuins. Drug Discov Today Technol. structure-activity relationship study. J Med Chem. (2016) 59:1599–612. (2014) 12:e9–e17. doi: 10.1016/j.ddtec.2012.08.003 doi: 10.1021/acs.jmedchem.5b01517 6. Dai H, Sinclair DA, Ellis JL, Steegborn C. Sirtuin activators and inhibitors: 12. Shah AA, Ito A, Nakata A, Yoshida M. Identification of a selective sIRT2 promises, achievements, and challenges. Pharmacol Ther. (2018) 188:140–54. inhibitor and its anti-breast cancer activity. Biol Pharm Bull. (2016) 39:1739– doi: 10.1016/j.pharmthera.2018.03.004 42. doi: 10.1248/bpb.b16-00520 Frontiers in Oncology | www.frontiersin.org 13 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) 13. Moniot S, Forgione M, Lucidi A, Hailu GS, Nebbioso A, Carafa V, but alters gene expression during aging. Cell. (2008) 135:907–18. et al. Development of 1,2,4-oxadiazoles as potent and selective inhibitors doi: 10.1016/j.cell.2008.10.025 of the human deacetylase Sirtuin 2: structure-activity relationship, X-ray 33. Yeung F, Hoberg JE, Ramsey CS, Keller MD, Jones DR, Frye RA, et al. crystal structure, and anticancer activity. J Med Chem. (2017) 60:2344–60. Modulation of NF-κB-dependent transcription and cell survival by the SIRT1 doi: 10.1021/acs.jmedchem.6b01609 deacetylase. EMBO J. (2004) 23:2369–80. doi: 10.1038/sj.emboj.7600244 14. Kozako T, Mellini P, Ohsugi T, Aikawa A, Uchida Y-I, Honda S-I, et al. Novel 34. Wilking MJ, Ahmad N. The role of SIRT1 in cancer: the saga continues. Am J small molecule SIRT2 inhibitors induce cell death in leukemic cell lines. BMC Pathol. (2015) 185:26–8. doi: 10.1016/j.ajpath.2014.10.002 Cancer. (2018) 18:791. doi: 10.1186/s12885-018-4710-1 35. Wang C, Yang W, Dong F, Guo Y, Tan J, Ruan S, et al. The prognostic role 15. Spiegelman NA, Price IR, Jing H, Wang M, Yang M, Cao J, et al. Direct of Sirt1 expression in solid malignancies: a meta-analysis. Oncotarget. (2017) comparison of SIRT2 inhibitors: potency, specificity, activity-dependent 8:66343–51. doi: 10.18632/oncotarget.18494 inhibition, and on-target anticancer activities. ChemMedChem. (2018) 36. Peck B, Chen CY, Ho KK, Di Fruscia P, Myatt SS, Coombes RC, 13:1890–4. doi: 10.1002/cmdc.201800391 et al. SIRT inhibitors induce cell death and p53 acetylation through 16. Desantis V, Lamanuzzi A, Vacca A. The role of SIRT6 in tumors. targeting both SIRT1 and SIRT2. Mol Cancer Ther. (2010) 9:844–55. Haematologica. (2018) 103:1–4. doi: 10.3324/haematol.2017.182675 doi: 10.1158/1535-7163.MCT-09-0971 17. Huang Z, Zhao J, Deng W, Chen Y, Shang J, Song K, et al. Identification of a 37. Singh S, Kumar PU, Thakur S, Kiran S, Sen B, Sharma S, et al. cellularly active SIRT6 allosteric activator. Nat Chem Biol. (2018) 14:1118–26. Expression/localization patterns of sirtuins (SIRT1, SIRT2, and SIRT7) doi: 10.1038/s41589-018-0150-0 during progression of cervical cancer and effects of sirtuin inhibitors 18. Ran L-K, Chen Y, Zhang Z-Z, Tao N-N, Ren J-H, Zhou L, et al. SIRT6 on growth of cervical cancer cells. Tumor Biol. (2015) 36:6159–71. overexpression potentiates apoptosis evasion in hepatocellular carcinoma via doi: 10.1007/s13277-015-3300-y BCL2-associated X protein-dependent apoptotic pathway. Clin Cancer Res. 38. Wilking MJ, Singh CK, Nihal M, Ndiaye MA, Ahmad N. Sirtuin deacetylases: (2016) 22:3372–82. doi: 10.1158/1078-0432.CCR-15-1638 a new target for melanoma management. Cell Cycle. (2014) 13:2821–6. 19. Cagnetta A, Soncini D, Orecchioni S, Talarico G, Minetto P, Guolo F, et al. doi: 10.4161/15384101.2014.949085 Depletion of SIRT6 enzymatic activity increases acute myeloid leukemia cells’ 39. Bhalla S, Gordon LI. Functional characterization of NAD dependent de- vulnerability to DNA-damaging agents. Haematologica. (2018) 103:80–90. acetylases SIRT1 and SIRT2 in B-cell chronic lymphocytic leukemia (CLL). doi: 10.3324/haematol.2017.176248 Cancer Biol Ther. (2016) 17:300–9. doi: 10.1080/15384047.2016.1139246 20. Süssmuth SD, Haider S, Landwehrmeyer GB, Farmer R, Frost C, Tripepi G, 40. Ceballos MP, Decándido G, Quiroga AD, Comanzo CG, Livore VI, et al. An exploratory double-blind, randomized clinical trial with selisistat, a Lorenzetti F, et al. Inhibition of sirtuins 1 and 2 impairs cell survival SirT1 inhibitor, in patients with Huntington’s disease. Br J Clin Pharmacol. and migration and modulates the expression of P-glycoprotein and MRP3 (2015) 79:465–76. doi: 10.1111/bcp.12512 in hepatocellular carcinoma cell lines. Toxicol Lett. (2018) 289:63–74. 21. Borra MT, Smith BC, Denu JM. Mechanism of human SIRT1 activation by doi: 10.1016/j.toxlet.2018.03.011 resveratrol. J Biol Chem. (2005) 280:17187–95. doi: 10.1074/jbc.M501250200 41. Kuo LJ, Yang L-X. γ-H2AX - a novel biomarker for DNA double-strand 22. Choupani J, Mansoori Derakhshan S, Bayat S, Alivand MR, Shekari Khaniani breaks. In Vivo. (2008). (2008) 22:305–9. M. Narrower insight to SIRT1 role in cancer: a potential therapeutic target to 42. Wang R-H, Sengupta K, Li C, Kim H-S, Cao L, Xiao C, et al. Impaired DNA control epithelial-mesenchymal transition in cancer cells. J Cell Physiol. (2018) damage response, genome instability, and tumorigenesis in SIRT1 mutant 233:4443–57. doi: 10.1002/jcp.26302 mice. Cancer Cell. (2008) 14:312–23. doi: 10.1016/j.ccr.2008.09.001 23. Yang T, Zhou R, Yu S, Yu S, Cui Z, Hu P, et al. Cytoplasmic SIRT1 43. Madabushi A, Hwang B-J, Jin J, Lu A-L. Histone deacetylase SIRT1 modulates inhibits cell migration and invasion by impeding epithelial-mesenchymal and deacetylates DNA base excision repair enzyme thymine DNA glycosylase. transition in ovarian carcinoma. Mol Cell Biochem. (2019) 459:157–69. Biochem J. (2013) 456:89–98. doi: 10.1042/BJ20130670 doi: 10.1007/s11010-019-03559-y 44. Jang J, Huh YJ, Cho H-J, Lee B, Park J, Hwang D-Y, et al. SIRT1 enhances the 24. Di Sante G, Pestell TG, Casimiro MC, Bisetto S, Powell MJ, Lisanti MP, survival of human embryonic stem cells by promoting DNA repair. Stem Cell et al. Loss of Sirt1 promotes prostatic intraepithelial neoplasia, reduces Rep. (2017) 9:629–41. doi: 10.1016/j.stemcr.2017.06.001 mitophagy, and delays PARK2 translocation to mitochondria. Am J Pathol. 45. Jarrett SG, Carter KM, Bautista RM, He D, Wang C, D’Orazio JA. Sirtuin 1- (2015) 185:266–79. doi: 10.1016/j.ajpath.2014.09.014 mediated deacetylation of XPA DNA repair protein enhances its interaction 25. Long Q, Xu J, Osunkoya AO, Sannigrahi S, Johnson BA, Zhou W, et al. with ATR protein and promotes cAMP-induced DNA repair of UV damage. J Global transcriptome analysis of formalin-fixed prostate cancer specimens Biol Chem. (2018) 293:19025–37. doi: 10.1074/jbc.RA118.003940 identifies biomarkers of disease recurrence. Cancer Res. (2014) 74:3228–37. 46. Yu W, Li L, Wang G, Zhang W, Xu J, Liang A. KU70 Inhibition impairs doi: 10.1158/0008-5472.CAN-13-2699 both non-homologous end joining and homologous recombination DNA 26. Vaziri H, Dessain SK, Ng Eaton E, Imai SI, Frye RA, Pandita TK, et al. damage repair through SHP-1 induced dephosphorylation of SIRT1 in adult hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase. Cell. (2001) T-cell leukemia-lymphoma cells. Cell Physiol Biochem. (2018) 49:2111–23. 107:149–59. doi: 10.1016/S0092-8674(01)00527-X doi: 10.1159/000493815 27. Wang C, Chen L, Hou X, Li Z, Kabra N, Ma Y, et al. Interactions between E2F1 47. Alves-Fernandes DK, Jasiulionis MG. The role of SIRT1 on DNA damage and SirT1 regulate apoptotic response to DNA damage. Nat Cell Biol. (2006) response and epigenetic alterations in cancer. Int J Mol Sci. (2019) 20:3153. 8:1025–31. doi: 10.1038/ncb1468 doi: 10.3390/ijms20133153 28. Abdelmawgoud H, El Awady RR. Effect of Sirtuin 1 inhibition on matrix 48. Asaka R, Miyamoto T, Yamada Y, Ando H, Mvunta DH, Kobara H, et al. metalloproteinase 2 and Forkhead box O3a expression in breast cancer cells. Sirtuin 1 promotes the growth and cisplatin resistance of endometrial Genes Dis. (2017) 4:240–6. doi: 10.1016/j.gendis.2017.11.001 carcinoma cells: a novel therapeutic target. Lab Investig. (2015) 95:1363–73. 29. Rifaï K, Idrissou M, Penault-Llorca F, Bignon Y-J, Bernard-Gallon D. Breaking doi: 10.1038/labinvest.2015.119 down the contradictory roles of histone deacetylase SIRT1 in human breast 49. Napper AD, Hixon J, McDonagh T, Keavey K, Pons J-F, Barker J, et al. cancer. Cancers. (2018) 10:409. doi: 10.3390/cancers10110409 Discovery of indoles as potent and selective inhibitors of the deacetylase 30. Jin X, Wei Y, Xu F, Zhao M, Dai K, Shen R, et al. SIRT1 promotes formation SIRT1. J Med Chem. (2005) 48:8045–54. doi: 10.1021/jm050522v of breast cancer through modulating Akt activity. J Cancer. (2018) 9:2012–23. 50. Trapp J, Meier R, Hongwiset D, Kassack MU, Sippl W, Jung M. doi: 10.7150/jca.24275 Structure-activity studies on suramin analogues as inhibitors of NAD + - 31. Jeong J, Juhn K, Lee H, Kim S-H, Min B-H, Lee K-M, et al. SIRT1 promotes dependent histone deacetylases (sirtuins). ChemMedChem. (2007) 2:1419–31. DNA repair activity and deacetylation of Ku70. Exp Mol Med. (2007) 39:8–13. doi: 10.1002/cmdc.200700003 doi: 10.1038/emm.2007.2 51. Zhou Z, Ma T, Zhu Q, Xu Y, Zha X. Recent advances in inhibitors of 32. Oberdoerffer P, Michan S, McVay M, Mostoslavsky R, Vann J, Park S- sirtuin1/2: an update and perspective. Future Med Chem. (2018) 10:907–34. K, et al. SIRT1 redistribution on chromatin promotes genomic stability doi: 10.4155/fmc-2017-0207 Frontiers in Oncology | www.frontiersin.org 14 April 2020 | Volume 10 | Article 657 Wössner et al. Sirtuin 1 Inhibiting Thiocyanates (S1th) 52. Pagans S, Pedal A, North BJ, Kaehlcke K, Marshall BL, Dorr A, et al. SIRT1 69. Amerik AY, Hochstrasser M. Mechanism and function of deubiquitinating regulates HIV transcription via tat deacetylation. PLoS Biol. (2005) 3:0210–20. enzymes. Biochim Biophys Acta Mol Cell Res. (2004) 1695:189–207. doi: 10.1371/journal.pbio.0030041 doi: 10.1016/j.bbamcr.2004.10.003 53. Ismail, Kuthati B, Thalari G, Bommarapu V, Mulakayala C, Chitta 70. Altun M, Kramer HB, Willems LI, McDermott JL, Leach CA, Goldenberg SK, et al. Synthesis of novel spiro[pyrazolo[4,3-d]pyrimidinones and SJ, et al. Activity-based chemical proteomics accelerates inhibitor spiro[benzo[4,5]thieno[2,3-d]pyrimidine-2,3′ -indoline]-2′ ,4(3H)-diones and development for deubiquitylating enzymes. Chem Biol. (2011) 18:1401–12. their evaluation for anticancer activity. Bioorg Med Chem Lett. (2017) doi: 10.1016/j.chembiol.2011.08.018 27:1446–50. doi: 10.1016/j.bmcl.2017.01.088 71. Seiberlich V, Goldbaum O, Zhukareva V, Richter-Landsberg C. The 54. Schiedel M, Herp D, Hammelmann S, Swyter S, Lehotzky A, Robaa D, et al. small molecule inhibitor PR-619 of deubiquitinating enzymes affects the Chemically induced degradation of sirtuin 2 (Sirt2) by a proteolysis targeting microtubule network and causes protein aggregate formation in neural chimera (PROTAC) based on sirtuin rearranging ligands (SirReals). J Med cells: implications for neurodegenerative diseases. Biochim Biophys Chem. (2018) 61:482–91. doi: 10.1021/acs.jmedchem.6b01872 Acta Mol Cell Res. (2012) 1823:2057–68. doi: 10.1016/j.bbamcr.201 55. Laemmli UK. Cleavage of structural proteins during the assembly of the head 2.04.011 of bacteriophage T4. Nature. (1970) 227:680–5. doi: 10.1038/227680a0 72. Milne JC, Denu JM. The Sirtuin family: therapeutic targets to treat diseases of 56. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano aging. Curr Opin Chem Biol. (2008) 12:11–7. doi: 10.1016/j.cbpa.2008.01.019 MD, et al. Measurement of protein using bicinchoninic acid. Anal Biochem. 73. Weber PC, Salemme FR. Applications of calorimetric methods (1985) 150:76–85. doi: 10.1016/0003-2697(85)90442-7 to drug discovery and the study of protein interactions. Curr 57. Heltweg B, Trapp J, Jung M. In vitro assays for the determination Opin Struct Biol. (2003) 13:115–21. doi: 10.1016/S0959-440X(03)0 of histone deacetylase activity. Methods. (2005) 36:332–7. 0003-4 doi: 10.1016/j.ymeth.2005.03.003 74. Cimmperman P, Baranauskiene L, Jachimovičiute S, Jachno J, Torresan 58. Maurer B, Rumpf T, Scharfe M, Stolfa DA, Schmitt ML, He W, et al. Inhibitors J, Michailoviene V, et al. A quantitative model of thermal stabilization of the NAD+-dependent protein desuccinylase and demalonylase sirt5. ACS and destabilization of proteins by ligands. Biophys J. (2008) 95:3222–31. Med Chem Lett. (2012) 3:1050–3. doi: 10.1021/ml3002709 doi: 10.1529/biophysj.108.134973 59. Swyter S, Schiedel M, Monaldi D, Sippl W, Lehotzky A, Rumpf T, et al. 75. Simeonov A. Recent developments in the use of differential New chemical tools for probing activity and inhibition of the NAD+ - scanning fluorometry in protein and small molecule discovery dependent lysine deacylase sirtuin 2. Philos Trans R Soc. (2018) 373:20170083. and characterization. Expert Opin Drug Discov. (2013) 8:1071–82. doi: 10.1098/rstb.2017.0083 doi: 10.1517/17460441.2013.806479 60. Niesen FH, Berglund H, Vedadi M. The use of differential scanning 76. Garbett NC, Chaires JB. Thermodynamic studies for drug design fluorimetry to detect ligand interactions that promote protein stability. Nat and screening. Expert Opin Drug Discov. (2012) 7:299–314. Protoc. (2007) 2:2212–21. doi: 10.1038/nprot.2007.321 doi: 10.1517/17460441.2012.666235 61. Ekkebus R, van Kasteren SI, Kulathu Y, Scholten A, Berlin I, Geurink PP, et al. 77. Parenti MD, Bruzzone S, Nencioni A, Del Rio A. Selectivity hot- On terminal alkynes that can react with active-site cysteine nucleophiles in spots of sirtuin catalytic cores. Mol Biosyst. (2015) 11:2263–72. proteases. J Am Chem Soc. (2013) 135:2867–70. doi: 10.1021/ja309802n doi: 10.1039/C5MB00205B 62. Rumpf T, Schiedel M, Karaman B, Roessler C, North BJ, Lehotzky A, et al. 78. Slichenmyer WJ, Rowinsky EK, Donehower RC, Kaufmann SH. The current Selective Sirt2 inhibition by ligand-induced rearrangement of the active site. status of camptothecin analogues as antitumor agents. J Natl Cancer Inst. Nat Commun. (2015) 6:6263. doi: 10.1038/ncomms7263 (1993) 85:271–91. doi: 10.1093/jnci/85.4.271 63. Dignam JD, Lebovitz RM, Roeder RG. Accurate transcription initiation by 79. Ryckewaert L, Sacconnay L, Carrupt P-A, Nurisso A, Simões-Pires RNA polymerase II in a soluble extract from isolated mammalian nuclei. C. Non-specific SIRT inhibition as a mechanism for the cytotoxicity Nucleic Acids Res. (1983) 11:1475–89. doi: 10.1093/nar/11.5.1475 of ginkgolic acids and urushiols. Toxicol Lett. (2014) 229:374–80. 64. Zhao X, Allison D, Condon B, Zhang F, Gheyi T, Zhang A, et al. The 2.5 doi: 10.1016/j.toxlet.2014.07.002 Å crystal structure of the SIRT1 catalytic domain bound to nicotinamide 80. Kim H, Jo A, Choi S-S, Nam H, Byun WG, Bae H, et al. Rational adenine dinucleotide (NAD+) and an indole (EX527 analogue) reveals a novel design and synthesis of reversible covalent PPARγ antagonistic ligands mechanism of histone deacetylase inhibition. J Med Chem. (2013) 56:963–9. inhibiting Ser273 phosphorylation. Asian J Org Chem. (2019) 8:1698–706. doi: 10.1021/jm301431y doi: 10.1002/ajoc.201800668 65. Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA. Development and testing of a general amber force field. J Comput Chem. (2004) 25:1157–74. Conflict of Interest: The authors declare that the research was conducted in the doi: 10.1002/jcc.20035 absence of any commercial or financial relationships that could be construed as a 66. Jakalian A, Jack DB, Bayly CI. Fast, efficient generation of high-quality atomic potential conflict of interest. charges. AM1-BCC model: II. Parameterization and validation. J Comput Chem. (2002) 23:1623–41. doi: 10.1002/jcc.10128 Copyright © 2020 Wössner, Alhalabi, González, Swyter, Gan, Schmidtkunz, Zhang, 67. Jones G, Willett P, Glen RC, Leach AR, Taylor R. Development and validation Vaquero, Ovaa, Einsle, Sippl and Jung. This is an open-access article distributed of a genetic algorithm for flexible docking11 edited by F. E Cohen J Mol Biol. under the terms of the Creative Commons Attribution License (CC BY). The use, (1997) 267:727–48. doi: 10.1006/jmbi.1996.0897 distribution or reproduction in other forums is permitted, provided the original 68. Ho E, Clarke JD, Dashwood RH. Dietary sulforaphane, a histone author(s) and the copyright owner(s) are credited and that the original publication deacetylase inhibitor for cancer prevention. J Nutr. (2009) 139:2393–6. in this journal is cited, in accordance with accepted academic practice. No use, doi: 10.3945/jn.109.113332 distribution or reproduction is permitted which does not comply with these terms. Frontiers in Oncology | www.frontiersin.org 15 April 2020 | Volume 10 | Article 657