Givinostat

Title: Potent Induction of Apoptosis by Givinostat in BCR-ABL1-positive and BCR-ABL1-negative Precursor B-cell Acute Lymphoblastic Leukemia Lines

Abstract— We have previously shown that givinostat can induce potent apoptosis in the BCR-ABL1-positive, TP53-wild type B-cell acute lymphoblastic leukemia (B-ALL) cell line SUP-B15. We extend our studies here to two additional B-ALL cell lines, BCR-ABL1- negative CCRF-SB and p210 BCR-ABL1-positive NAML1. Givinostat induced significant cell growth inhibition in both cell lines, with an IC50 of 0.65±0.052 M and 0.25±0.028 M in CCRF-SB and NAML1, respectively. The key signal protein of the BCR-ABL1, Crk- L1, was significantly reduced by givinostat treatment in NAML1. As in SUP-B15, givinostat induced apoptosis in both cell lines but showed different levels of cleavage of the procaspase proteins Casp-3, Casp-7 and PARP. Levels of cell cycle-DNA repair regulator p21, CHK1 and FANCD2 levels were markedly affected by givinostat treatment. These data further enrich our understanding of the mechanisms of the antineoplastic effects of givinostat in B-ALL and provide a preclinical rationale for the inclusion of givinostat or similar agent in leukemia therapy.

I. INTRODUCTION

Given the key role epigenetic dysregulation in mediating growth and maturation arrest in hematopoiesis, histone deacetylases (HDACs) are potential targets in the treatment of leukemia. Consequently, inhibitors of HDACs (HDIs) are being studied for therapeutic purposes [1-2]. HDIs promote or enhance several different anticancer mechanisms, such as induction of apoptosis, cell- cycle arrest, and cellular differentiation [3-4]. Recently, three HDACi (vorinostat, belinostat and romidepsin) received FDA approval for treatment of cutaneous or peripheral T-cell lymphoma, and the HDACi panobinostat for the treatment of multiple myeloma in combination with others medications [5-6].
There is the need for improved treatment of acute lymphoblastic leukemia (ALL), which particularly in adults has a low cure rate and high mortality. There are a large number of previous studies on the effects of HDACi on cultured cell lines as well as xenograft tumor models, which have indicated activity of these agents in lymphoid malignancies. Givinostat exerted significant DNA damages in some responders of T-cell ALL xenografts with apoptosis induced by this agent operating at least in part through genotoxic effects and modulation of DNA repair pathway [6-7].

Our recent study showed that givinostat significantly inhibited cell proliferation of p190 BCR-ABL1 positive SUP-B15 cell line through inhibition of BCR-ABL1 signaling pathway. In that line, givinostat also activated caspase cascades and increased expression of p21 protein [8]. This study investigates the effects of givinostat treatment in B-cell ALL cell line with the p210/major variant of BCR-ABL1 or the absence of BCR-ABL1 fusion.

II. MATERIALS AND METHODS

BAX, Bcl-xL, AIF, LC3B, BECN1, PUMA, FANCD2, and ELM4 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Other sources for antibodies, including phospho-CrkL, caspase-3 and -7, PARP1 and GAPDH, were described previously [8]. Sources and preparation of propidium iodide (PI) and FITC Annexin V-FITC detection kit was described previously [8].

Cell culture and drug treatment

CCRF-SB and NAML1 cells (ATCC) were grown in IMDM or RPMI-1640 culture medium supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin (ATCC). The cells were seeded at 2 x105/mL in 5% CO2 at 37°C in a T-50 polypropylene flask (Becton Dickinson Labware, Lincoln Park, N.J.), with the cap loosely placed to allow for gas exchange. The cells were grown with dimethyl sulfoxide vehicle (DMSO), 0 to 1.0 M Givinostat, or imatinib for 12-72 hours (hrs) when intact cells for MTT test (ATCC) or cells were harvested for protein or RNA extraction. BCR-ABL1 fusion genes were detected by quantitative reverse transcription polymerase chain reaction (qRT-PCR) as previously described [9-10].

Cell proliferation assay (MTS assay).

Cell proliferation was assessed using CellTiter96 AQueous One solution cell proliferation assay kit (Promega, Madison, WI). Four replicates per condition were measured using 100μl of cell suspension at a density of 5×104 cells/well. Givinostat at 1.0 M or DMSO was added to the suspension and cultured at 37°C in 5% CO2. Before the conclusion of the experiment at 24, 48, 72 and 96hrs, 20μl of the tetrazolium reagent MTS was added to each well with additional culture for 4 hours. The absorbance was measured at 490 nm using a microplate spectrophotometer (PowerWave, BioTek Instruments, Winooski, VT, USA), and the results were expressed as a percentage (%) of the control (vehicle alone).

Western blot analysis

Briefly, CCRF-SB (1× 106) or NALM1 (1×106) cells were harvested, washed with PBS and resuspended in 1×Laemmli sample buffer containing proteinase and phosphatase inhibitors. Fifty (50) g of protein was separated by sodium dodecyl (lauryl) sulfate- polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretically transferred to nitrocellulose membranes. Membranes were probed with antibodies against corresponding proteins. The blots were visualized with ECL reagent (Amersham, Arlington Heights, IL) and then exposed to Hyperfilm (Amersham) as described previously [8-9].

Flow cytometry analysis

Cells treated or untreated cells for 48 hours were fixed with ice-cold 70% ethanol for 1 hour and/or stored at -20C until use.
The cells were then washed with cold PBS and treated with RNase for 15 minutes at 37C, stained with 50 g/ml PI and analyzed in flow cytometric analysis (FACSCalibur, Becton Dickinson, CA, USA). DNA content was calculated in the untreated cells with intact G1, S, and G2/M phases using Mod-Fit software (Verity Software House). The percentage of sub-G0/G1 were averaged based on triplicate measurements.
Flow cytometric analysis measures necrosis/apoptosis using an Annexin V-FITC/PI detection kit [11]. After treatment with 0.5
M of Givinostat or vehicle for 12 to 48 hours, cells were harvested by centrifugation and resuspended in 0.5% FBS-PBS. The cell suspension was double-stained with FITC annexin-V (50µg/ml) and PI (50µg/ml) and kept in the dark for 15 minutes. The fluorescence of each cell was analyzed by a using a nitrogen argon laser operating at 488/535 nm excitation/emission for annexin-V, with the fluorescence captured on the FL-1H channel, and at 488/620 nm excitation/emission for PI, with the fluorescence captured on FL-2H channel with logarithmic amplification. For each determination, 10,000 cells were counted. Data were analyzed using FlowJo software (Tree Star, Ashland, OR).

PCR and RT-PCR assays

Total RNA was isolated from CCRF-SB and NALM1 cells using Trizol (QIAzol) and miRNeasy Mini Kit (Qiagen) followed by cDNA reverse transcription by using random hexamer, dNTP, and SuperScript™ II Reverse Transcriptase according to the manufacturer’s procedure (Invitrogen, USA). PCR volume was 25 L, including 5 l of cDNA (or DNA), primers, and master mix with Hotstar Taq polymerase. The primers for amplifications of p190 and p210 of BCR-ABL1 were published as previous [10].

Conventional PCR was performed as described previously [8]. The amplicon sizes were analyzed by capillary electrophoresis on ABI 3130 (Life Technology) with appropriate size controls. DNA sequencing of TP53 was performed using the dideoxy chain- termination (Sanger) method.

Statistics

All results were expressed as means ± SD unless stated otherwise. The unpaired Student’s t test was used to evaluate the significance of differences between groups, accepting p<0.05 as level of significance. III. RESULTS Inhibition of proliferation induced by treatment with givinostat but not by imatinib in CCRF-SB and NALM1.In CRF-SB and NAML1, imatinib treatment at a dose range of 0.01-0.5 M showed minimal toxicities or growth inhibition (by MTS) with 24 to 72 hours of treatment (data not shown), which is similar to relative resistance noted for SUP-B15 [8]. In contrast, treatment with 0.1 M Givinostat inhibited 66% of NALM1 growth and 0.5 M givinostat inhibited 39% of CCRF-SB growth at 48hrs (Figure 1A). The IC50 values at 48hrs for a single givinostat treatment were determined from cell survival plots (SigmaPlot) and were 0.65 ± 0.3M and 0.25 ±0.3M in CCRF-SB and NALM1, respectively. The treatment of the cells by Givinostat at 1M resulted in time-dependent decrease of pCrk-L1 by Western blot (Figure 1C). As expected, qRT-PCR showed an absence of BCR- ABL1 fusion gene in CCRF-SB but p210 (b2a2) transcripts detected in NALM1 (Figure 1B). Givinostat treatment induced cell cycle arrest at G1-phase and DNA content changes We analyzed the effect of givinostat on the cell cycle of the treated cells by flow cytometry. As shown in Figure 2 and summarized in Table 1, there were significant increases in the G1 fraction in CCRF-SB and NALM1 with 24 and 48hrs of givinostat treatment. Consistent with MTS result above, there was a significant decrease in the S phase in 24 hour-treated NALM1 cells. Therefore, our data suggests that givinostat can arrest G1-phase on ALL cells, similar to its effect on SUP-B15 [8]. Treated cells at 24hrs also showed slight increases in the sub-G0/G1 fractions, likely representing fragmented or subdiploid DNA content that were significantly increased by 48 hours compared to control (P<0.001) (Table 1). Givinostat induced 3-fold and 6.6-fold increases in sub-G0/G1 signal in CCRF-SB and 7.6- and 22.8-fold increases in NALM1 at 24 and 48 hours, respectively. Givinostat treatment induced apoptosis The increased signal in the sub-G0/G1 fractions could be due to apoptotic or necrotic cells due to direct genotoxic effects of high-dose givinostat. To distinguish these possibilities, we treated the cells with 0.5M givinostat and analyzed fractions for apoptotic cells and necrotic cells by flow cytometric detection of FITC-Annexin V/PI staining. As shown in Figure 3, apoptosis was already apparent after 24hrs of treatment, with more than 50% of the total counts/cells at 48hrs in both cell lines. As summarized in Table 2, the apoptotic fraction in treated CCRF-SB at 24hrs was 18.4% as compared to 9% in the control, with a nearly 5-fold increase at 48hrs (9 vs. 44.3%, P<0.01). In NALM1 the increase in treated cells was 1.9-fold at 24 hours (8 vs. 15.0%) and 7.5-fold at 48 hours (7.9% vs. 60%, P<0.01). The necrotic fractions were minimal in both cell lines at less than 1%. As additional evidence of apoptotic effects, we examined the cleavage (cleaved-) and intact (Pro-) caspase (C)-3 and C-7 and PARP by Western blotting, using control at 72hrs as the reference (Figure 4). NALM1 cells showed appearance of cleaved PARP (85 kilodalton (KDa)), C3 (17 and 20 KDa forms), and C7 (20kD). In contract, no cleavage of C3 and pro-C7 proteins were noted in CCRF-SB with givinostat treatment but only increases in cleaved PARP. These caspase cleavages were generally less prominent, though, than seen previously for SUP-B15 [8]. We also examined BAX/Bcl-XL-dependent mitochondrial apoptosis as well as autophagy pathways by Western blot (Figure 5). Givinostat-treated NALM1 showed loss of AIF, PUMA, LC3B, and BECN1 and mild decreases in BAX over time, with increases in Bcl-xL. In contrast, CCRF-SB showed more mild changes in these protein levels that paralleled the lower levels of apoptosis and caspase cleavage as compared to NALM1. Due to limited selected panel proteins, it is inconclusive to inconsistence between apoptosis induced detected by flow cytometric analysis and protein expression in Givinostat-treated CCRF-SB cells. Apoptosis induced by Givinostat in parallel with suppression of cell checkpoint and DNA repair proteins Given the previous effects of givinostat treatment on p21 in SUP-B15, we also assessed various components of the p53-linked DNA repair pathway. As shown in Figure 6, both p53 and p21 were expressed in both lines at high levels. DNA sequencing analysis revealed no TP53 mutations in either line. Although p53 levels were only modestly decreased in NALM1 after 72 hours of treatment, the levels of p21 dropped dramatically by 48hrs in both cell lines. Interestingly, both cell lines showed striking decreases in FANCD2 and CHEK1, which were apparent within 24hrs of givinostat treatment. IV. DISCUSSION In this study, using a single type II HDAC inhibitor, givinostat, we demonstrated potent apoptotic induction in two genetically different ALL cell lines, namely the p210 BCR-ABL1-positive NALM1 and the BCR-ABL1-negative CCRF-SB. Inhibition of growth/cell death induced by this drug appears to affect several different pathways, including BCR-ABL1 signaling, classical apoptosis, autophagy and the p53-linked DNA repair axis. Both cell lines showed significant G0/G1 arrest prior to timepoints with high levels of apoptosis, with effects on the DNA repair protein levels also noted at the earliest assessed timepoint. Differential effects of givinostat on the two lines were also noted. For example, in NALM1, the BCR-ABL1 target CrkL, caspase cleavage and modulation of autophagy proteins were readily. In contrast, givinostat induced significant apoptosis in CCRF-SB but showed less striking effects on caspase activation or autophagy pathway protein levels. There are two major forms of BCR-ABL1 fusion genes in Ph-positive leukemia, based on the molecular weight of the fusion proteins, i.e. the minor transcript/p190 and the major transcript/p210. p210+ BCR-ABL1 is more commonly found in patients with chronic myelogenous leukemia (CML) but is also found in a subset of adult ALL, whereas p190+ BCR-ABL1 is nearly always associated with ALL [10, 12]. Previous studies have suggested a differential outcome possibly due to differential signaling in ALL containing [13-14]. Therefore, it is rational to check if givinostat would have similar antineoplastic effects on genetically different ALL cell models. Our results showed IC50 of givinostat for proliferation inhibition, as measured by cytotoxicity assay, is higher in p210+ NALM1 than in the P190+ SUP-B15. In contrast, apoptotic induction appears more potent in the latter cell line. However, givinostat’s effects are clearly at least partially independent of BCR-ABL1 given its effects on the BCR-ABL1-negative CCRF-SB. Given the wild-type TP53 in both CCRF-SB and NALM1, the effects of givinostat on the p53 pathway may reflect classical activation leading to the observed increased expression of the cyclin-dependent kinase inhibitor p21 at early post-treatment timepoints. In our previous work on SUP-B15, we observed persistent elevation of p21 protein with givinostat treatment [8]. In contrast, p21 increases were transient in both CCRF-SB and NALM1 and decreased by the 48 hour timepoint when apoptosis was prominent (Figure 2). This may reflect the ability of enhanced p53 expression to overcome p21WAF1/CIP1-mediated G1 arrest and induces apoptosis in human cancer cells [15]. In SUP-B15, apoptosis as measured by flow cytometry using Annexin V/FITC-PI dual staining and caspase cleavage by Western blot were highly correlated. In NALM1 and CCRF-SB, the level of apoptosis measured by flow cytometry was more prominent than the observed caspase cleavage (Figure 4). This likely reflects intrinsics differences in the susceptibility to apoptosis of different cells lines and possibly differences in the differential activation of necrosis and autophagy by givinostat. We did not observe significant levels of ultra-fast (necrotic) cell death in either NALM1 or CCRF-SB, as compared to SUB-B15 [8]. Differences in drug-induced “slow cell death” [16] between these lines may reflect mutations or altered regulations of components of the apoptotic pathway. If such mutations are present, cleavage of caspases might be expected to less efficient but still sufficient to initiate cell death [16]. In summary, the Type II HDAC inhibitor givinostat produces similar anti-proliferative and pro-apoptotic effects in three different cell lines representing distinct subtypes of ALL. It thus represents a promising candidate as a therapeutic agent. The multiple pathway affected by this agents suggest a potential for use in ALL that have genetic alteration distinct from the ones studied here which could be assessed in future clinical trials. Figure Legends Figure 1. The effect of givinostat on the proliferation of CCRF-SB and NALM1 cells. (A) After treatment with the indicated concentrations of givinostat (0~0.5M) for 48 hours, the rate of cell growth was determined by MTS assay. (B) The expression of the BCR-ABL1 fusion gene in CCRF-SB and NALM1 cells was assayed by RT-PCR. (C) After treatment with 1M givinostat for 24~72 hours, pCrk-L1 levels were detected using Western blot. *P<0.05; **P<0.01 when compared to non-treated controls; Mean ± S.D. of at least three experiments. b2a2: p210 BCR-ABL1. Figure 2. The Effect of givinostat treatment on the cell cycle in CCRF-SB (A) and NALM1 (B). After treatment with 1.0 M givinostat for 0~48 hours, the distributions of propidium iodide signal was detected by flow cytometry and analyzed using Mod- Fit software. The first and second red peaks indicate the diploid G1 and G2/M peaks; S-phase is hatched areas between the two red peaks. Sub-G0/G1 peaks are solid purple areas. Figure 3. Givinostat induces apoptosis in CCRF-SB and NALM1. After treatment with 0.5 M givinostat for 24~48 hours, apoptosis in CCRF-SB and NALM1 cells was analyzed using FITC-Annexin V/PI dual staining followed by flow cytometry. Viable cells are negative for Annexin V and PI; early apoptotic cells are Annexin V positive and PI negative; late apoptotic or dead cells are positive for both Annexin V and PI. Fractions were analyzed using software FlowJo software (TreeStar). Figure 4. Givinostat modulates the expression of apoptosis-associated proteins. After treatment with 0.5 M givinostat for 24~72 hours, levels of cleaved and intact (Pro-) caspase-3 and -7 and PARP in CCRF-SB and NALM1 cells were determined by Western blot. Figure 5. Effect pf givinostat on the BAX/Bcl-XL-dependent mitochondrial apoptotic and autophagy pathways. After treatment with 0.5 M givinostat for 24~72 hours, the proteins level of BAX, Bcl-xL, AIF, PUMA, LC3B and BECN1 in CCRF- SB and NALM1 cells were determined by Western blot.. Figure 6. Givinostat treatment alters levels of checkpoint and DNA repair-associated proteins. After treatment with 0.5 M givinostat for 24~72 hours, the proteins level of FANCD2, CHEK1, p53 and p21 in CCRF-SB and NALM1 cells were determined by Western blot. ACKNOWLEDGMENT The work is supported to W.Z. by the NIH Cancer Center Core Grant (CCSG) from James Comprehensive Cancer Center, The Ohio State University, Columbus, OH. REFERENCES [1] Manal M, Chandrasekar MJ, Gomathi Priya J, Nanjan MJ. Inhibitors of histone deacetylase as antitumor agents: A critical review. Bioorg Chem. 2016 Aug;67:18-42. Review. PubMed PMID: 27239721. [2] Johnstone RW. Histone-deacetylase inhibitors: novel drugs for the treatment of cancer. Nat Rev Drug Discov. 2002; 1: 287-299. [3] Marks PA, Richon VM, Rifkind RA. Histone deacetylase inhibitors: inducers of differentiation or apoptosis of transformed cells. J Natl Cancer Inst. 2000;92:1210-1216. [4] Vrana JA, Decker RH, Johnson CR, Wang Z, Jarvis WD, Richon VM, Ehinger M, Fisher PB, Grant S. Induction of apoptosis in U937 human leukemia cells by suberoylanilide hydroxamic acid (SAHA) proceeds through pathways that are regulated by Bcl-2/Bcl-XL, c-Jun, and p21CIP1, but independent of p53. Oncogene. 1999;18(50):7016-25. [5] Stubbs MC, Kim W, Bariteau M, Davis T, Vempati S, Minehart J, Witkin M, Qi J, Krivtsov AV, Bradner JE, Kung AL, Armstrong SA. Selective Inhibition of HDAC1 and HDAC2 as a Potential Therapeutic Option for B-ALL. Clin Cancer Res. 2015 May 15;21(10):2348-58. doi: 10.1158/1078-0432.CCR-14-1290. PubMed PMID: 25688158;PubMed Central PMCID: PMC4433811. [6] Campbell P, Thomas CM. Belinostat for the treatment of relapsed or refractory peripheral T-cell lymphoma. J Oncol Pharm Pract. 2017 Mar;23(2):143-147. PMID: 26921086. [7] Pinazza M, Borga C, Agnusdei V, Minuzzo S, Fossati G, Paganin M, Michielotto B, De Paoli A, Basso G, Amadori A, te Kronnie G, Indraccolo S. An immediate transcriptional signature associated with response to the histone deacetylase inhibitor Givinostat in T acute lymphoblastic leukemia xenografts. Cell Death Dis. 2016 Jan 14;6:e2047. doi: 10.1038/cddis.2015.394. PubMed PMID: 26764573; PubMed Central PMCID: PMC4816177. [8] Li Y, Zhao K, Yao C, Kahwash S, Tang Y, Zhang G, Patterson K, Wang QE, Zhao W. Givinostat, a type II histone deacetylase inhibitor, induces potent caspase-dependent apoptosis in human lymphoblastic leukemia. Genes Cancer. 2016 Sep;7(9-10):292-300. doi: 10.18632/genesandcancer.117. PubMed PMID: 28050230; PubMed Central PMCID: PMC5115170.
[9] Jones D, Thomas D, Yin CC, O’Brien S, Cortes JE, Jabbour E, Breeden M, Giles FJ, Zhao W, Kantarjian HM. Kinase domain point mutations in Philadelphia chromosome positive acute lymphoblastic leukemia emerge after therapy with BCR-ABL kinase inhibitors. Cancer. 2008 Sep 1;113(5):985-94.
[10] Kantarjian HM, Talpaz M, Dhingra K, Estey E, Keating MJ, Ku S, Trujillo J, Huh Y, Stass S, Kurzrock R. Significance of the P210 versus P190 molecular abnormalities in adults with Philadelphia chromosome-positive acute leukemia. Blood. 1991 Nov 1;78(9):2411-8. PubMed PMID: 1932753
[11] Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled Annexin V. J Immunol Methods. 1995 Jul 17;184(1):39-
51. PubMed PMID: 7622868.
[12] DeBoer R, Koval G, Mulkey F, Wetzler M, Devine S, Marcucci G, Stone RM, Larson RA, Bloomfield CD, Geyer S, Mullighan CG, Stock
W. Clinical impact of ABL1 kinase domain mutations and IKZF1 deletion in adults under age 60 with Philadelphia chromosome-positive (Ph+) acute lymphoblastic leukemia (ALL): molecular analysis of CALGB (Alliance) 10001 and 9665. Leuk Lymphoma. 2016 Oct;57(10):2298-306. doi: 10.3109/10428194.2016.1144881. PubMed PMID: 26892479; PubMed Central PMCID: PMC5008253.
[13] Cutler JA, Tahir R, Sreenivasamurthy SK, Mitchell C, Renuse S, Nirujogi RS, Patil AH, Heydarian M, Wong X, Wu X, Huang TC, Kim MS, Reddy K, Pandey A. Differential signaling through p190 and p210 BCR-ABL fusion proteins revealed by interactome and phosphoproteome analysis. Leukemia. 2017 Feb 17. doi: 10.1038/leu.2017.61. [Epub ahead of print] PubMed PMID: 28210003.
[14] Specchia G, Mininni D, Guerrasio A, Palumbo G, Pastore D, Liso V. Ph positive acute lymphoblastic leukemia in adults: molecular and clinical studies. Leuk Lymphoma. 1995;18 Suppl 1:37-42. PubMed PMID: 7496353.
[15] Kagawa S, Fujiwara T, Hizuta A, Yasuda T, Zhang WW, Roth JA, Tanaka N. p53 expression overcomes p21WAF1/CIP1-mediated G1 arrest and induces apoptosis in human cancer cells. Oncogene. 1997 Oct 16;15(16):1903-9. PubMed PMID: 9365236.
[16] Blagosklonny MV. Cell death beyond apoptosis. Leukemia. 2000 Aug;14(8):1502-8. Review. PubMed PMID: 10942250.
[17] Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau MM, Bloomfield CD, Cazzola M, Vardiman JW. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood. 2016 May 19;127(20):2391-405. doi: 10.1182/blood-2016-03-643544. Review. PubMed PMID: 27069254.