Contents lists available at ScienceDirect

European Journal of Pharmacology

journal homepage: www.elsevier.com/locate/ejphar

Molecular and cellular pharmacology

Cephalostatin 1 analogues activate apoptosis via the endoplasmic reticulumstress signaling pathway

Lubna H. Tahtamounia,⁎, Mansour M. Nawasrehb, Zainab A. Al-Mazaydeha, Rema A. Al-Khateeba,Reem N. Abdellatifa, Randa M. Bawadic, James R. Bamburgd, Salem R. Yasina

a Department of Biology and Biotechnology, Faculty of Science, The Hashemite University, Zarqa 13115, JordanbApplied Sciences Department, Faculty of Engineering Technology, Al-Balqa Applied University, Amman 11134, Jordanc Department of Physiology and Biochemistry, The University of Jordan, Amman 11942, Jordand Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO 80523, USA

A R T I C L E I N F O

Keywords:CephalostatinCytotoxicityCaspaseCytochrome cSmac/DIABLO

A B S T R A C T

The current study was conducted to compare the cytotoxicity of two stereospecific cephalostatin 1 analogues(CAs) against several human normal cell types and cancer cell lines and to determine their cytotoxic mechanism.Both CA analogues induced apoptosis and were cytotoxic with 50% growth inhibition (GI50) at ~1 µM or less insix human cancer cell lines but neither analogue at 10 µM killed more than 14% of any of three types of normalhuman cells suggesting their cytotoxicity is cancer-specific. CA treatment inhibited clonogenic tumor growth andactivated caspase 3 and 9 but not caspase 8. CA-induced apoptosis was inhibited by the pan caspase inhibitorindicating the importance of caspase activation. CA treatment released smac/DIABLO but not cytochrome c frommitochondria and induced phosphorylation of eIF-2 and the activation of procaspase 4 in cancer cells, similar tocell treatment with thapsigargin, a known endoplasmic reticulum (ER) stress inducer. Finally, cells pretreatedwith a caspase 4 inhibitor were resistant to CA-induced apoptosis. In conclusion, both CAs induced apoptosis bytriggering ER stress. Because of their ease of synthesis and low GI50, these cephalostatin analogues representpromising anticancer drugs.

1. Introduction

Cancer is a leading cause of death worldwide and the number ofdeaths is projected to rise in the coming years (Siegel et al., 2016).Surgery, chemotherapy and radiotherapy are the most common treat-ment options against cancer (Abdullah and Chow, 2013). However, theresponse to treatment varies substantially in different types of cancer,or even among patients with the same type of cancer (Gatti and Zunino,2005). The variations in responses might indicate that intrinsic or ac-quired therapeutic resistance exists in cancer patients (Hammond et al.,2016). Nowadays, there are great efforts focusing on understanding themolecular mechanisms responsible for drug resistance and identifyingnew drugs which might work through signaling pathways that differfrom what regular drugs use to overcome drug resistance (Housmanet al., 2014).

Genetic and biochemical studies have identified two major path-ways of apoptosis, the death receptor-mediated pathway and the mi-tochondria-dependent pathway (Elmore, 2007). Recently, many studieshave implicated a regulatory role for the endoplasmic reticulum (ER) in

apoptosis (Bravo-Sagua et al., 2013). The ER under stress induces theoverexpression of chaperones and phosphorylation of the eukaryoticinitiation factor-2 (eIF-2), which attenuates translation initiation andprotein synthesis (Bravo-Sagua et al., 2013).

In addition and under severe stress, the ER can initiate its ownapoptotic signals that independently activate caspase 4 by localizing itto the ER membrane. Caspase 4 in turn activates caspase 9 in-dependently of cytochrome c release from the mitochondria(Breckenridge et al., 2003; Iurlaro and Muñoz-Pinedo, 2016). Caspase 9then activates one or more of the effector caspases (caspase 3, 6 and 7)(Groenendyk and Michalak, 2005). In addition, ER stress causes therelease of second mitochondria-derived activator of caspase/direct in-hibitor of apoptosis-binding protein with low isoelectric point (smac/DIABLO) from the mitochondria which inhibits the inhibitors ofapoptosis proteins (IAPs) contributing to the activation of caspases(Parrish et al., 2013).

Cephalostatin 1 (Fig. 1) is a member of strongly-related bis-steroidalcompounds (Pettit et al., 1988, 2011). Cephalostatin 1 has proved to beone of the most powerful experimental anticancer agents tested with a

https://doi.org/10.1016/j.ejphar.2017.11.025Received 9 August 2017; Received in revised form 7 November 2017; Accepted 14 November 2017

⁎ Correspondence to: Dept. of Biology and Biotechnology, Faculty of Science, The Hashemite University, P.O. Box 150459, Zarqa 13115, Jordan.E-mail address: lubnatahtamuni@hu.edu.jo (L.H. Tahtamouni).

European Journal of Pharmacology 818 (2018) 400–409

Available online 15 November 20170014-2999/ © 2017 Elsevier B.V. All rights reserved.

T

GI50 in the nano to pico molar range (Lee et al., 2009). In addition,cephalostatin 1 induces apoptosis through the ER-mediated pathwayand selectively uses smac/DIABLO as a mitochondrial signaling mole-cule (Dirsch et al., 2003; Muller et al., 2005; Rudy et al., 2008).

However, the availability of cephalostatin 1 from its natural sourcesis extremely limited and its synthesis is very complicated (Gryszkiewiczet al., 2003; Li and Dias, 1997; Poza et al., 2010). Difficulties in ob-taining quantities of cephalostatin 1 led to approaches synthesizingcephalostatin 1 analogues (CAs) looking for utilizable alternatives (Guoet al., 1996; Iglesias-Arteaga and Morzycki, 2013; Li et al., 2002; Li andFuchs, 2003; Nawasreh, 2007).

Two 12'α-derivatives of cephalostatin 1, 12'-α hydroxy-12α-hydro-xymethyl-bis-steroidal pyrazine (CAA) and 12'α-hydroxy-12β-hydro-xymethyl-bis-steroidal pyrazine (CAB) (Fig. 1), were synthesized pre-viously in small amounts and when tested as a mixture with the 12'-β-hydroxy derivatives, showed cytotoxicity (Nawasreh, 2007). However,

the cytotoxicity of the isolated 12'-β-hydroxy derivatives was dimin-ished. Thus, we have scaled up synthesis of these compounds to obtainsufficient amounts of the purified 12'α-hydroxy derivatives to char-acterize their biological activity in cell culture and to deduce theirmechanism of cell killing. The amounts of these compounds should besufficient for their eventual use in vivo to analyze their pharmacologicalefficacy against cancers.

2. Materials and methods

2.1. Compounds

Broad spectrum pan-caspase inhibitor (benzyloxycarbonyl-Val-Ala-Asp(OMe)-fluoromethylketone; zVAD-fmk) was purchased fromBachem Americas, Inc. (CA, USA), thapsigargin (TG) was purchasedfrom LC Laboratories (MA, USA). Caspase 4 inhibitor (Ac-LEVD-CHO)

Fig. 1. Structures of molecules important to this study. Chemicalstructure of cephalostatin 1, cephalostatin analogue A (CAA) andCAB.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

401

and etoposide (ET) were purchased from Sigma-Aldrich (MO, USA). Thetwo isomers CAA and CAB (Fig. 1) were prepared as previously de-scribed (Nawasreh, 2000, 2007).

2.2. Cell culture

Human K562 chronic myelogenous leukemic and human MCF-7breast cancer cells were cultured in RPMI-1640 (Euroclone, Italy)supplemented with 10% fetal bovine serum (FBS; Capricorn Scientific,Germany). Human DU-145 prostate cancer cell line was cultured in α-MEM supplemented with 10% FBS. Human HeLa cervical cancer cells,SAOS-2 and U2OS osteosarcoma cancer cells were cultured in highglucose DMEM (HGDMEM) (Gibco, USA) supplemented with 10% FBS.Human normal ARPE-19 retinal pigmented epithelial cell line wascultured in DMEM/F12 (Gibco, USA) supplemented with 10% FBS.Human normal GM2149 skin fibroblast cell line was cultured in α-MEM(Gibco, USA) supplemented with 15% FBS. Human normal white bloodcells were cultured in in RPMI-1640 supplemented with 10% FBS.Trypsin-EDTA (Lonza, Switzerland) was routinely used for subcultures.Cell growth was accomplished at 37 °C in a 5% carbon dioxide/95% airatmosphere.

2.3. In vitro cytotoxicity (3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide) MTT test

Cytotoxicity of CAA and CAB on human normal and cancer cells wasevaluated by means of MTT (tetrazolium salt reduction) test (Fotakisand Timbrell, 2006; Mosmann, 1983). CA stock solutions were preparedin 10% DMSO, final DMSO concentration in media did not exceed 0.1%.For each analogue, six concentrations (0.005, 0.01, 0.05, 0.1, 1, and10 µM) were prepared in growth media. Viable cells (50,000) wereadded to each well of a 96-well tissue culture plate containing growthmedium supplemented with FBS. Cells were kept in a humidified 5%CO2 incubator at 37 °C for 24 h. The next morning, the different con-centrations of CAs were added, and the cells incubated for 24 h, 48 h,and 72 h. Freshly prepared MTT salt (3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide) (5 mg/ml) was then added to each wellto give a final concentration of 0.5 μg/μL. The plates were incubated for4 h and the formation of formazan crystals was checked using an in-verted microscope. Equal volume of 1:1 (200 μL) DMSO and iso-propanol mixture was added to each well and incubated for 30–45 min.The inhibition of cell growth induced by the two analogues was de-tected by measuring the absorbance of each well at 570 nm using a StatFax microplate reader (FL, USA). Percent growth was calculated ac-cording to the following formula: Growth (%) = OD treated/OD ve-hicle-treated control×100%. The concentration-percent growth curvewas used to calculate the concentration which caused 50% growth in-hibition (GI50) by linear interpolation from a semi-log plot of a dose-response curve, while the concentration that caused total growth in-hibition (TGI) was read as the x-axis intercept at the line for 100% non-

viable cells. The experiment was performed three times in triplicates.

2.4. Detection of apoptosis

Apoptosis was detected using a Nikon Eclipse Ti-E microscope byvisualization of apoptotic nuclei after staining with DAPI and by flowcytometry of cell surface phosphatidylserine using the annexin V-FITCapoptosis detection kit (Sigma, USA).

2.5. Clonogenic assay

Cells (2 × 105) cells were seeded in tissue culture dishes containinggrowth media supplemented with FBS. Cells were kept in a humidified5% CO2 incubator at 37 °C for 24 h. Afterwards, the media were re-placed and the cells were incubated for 3 h in the presence of increasingconcentrations (0.005, 0.01, 0.05, 0.1, 1, and 10 µM) of CAA and CAB.Aliquots of 1000 cells were seeded on soft agar for all cell lines exceptfor K562 leukemic cells which were seeded on methylcellulose(Franken, 2006; Rafehi et al., 2011). Plates were incubated for 12 days.The colonies were then stained with 0.01% crystal violet and counted,discarding colonies with less than 50 cells (Rafehi et al., 2011). Thesurviving fraction (SF) was calculated as plating efficiency (PE) oftreated sample/ PE of vehicle-treated control × 100, where PE equalsnumber of colonies counted/ number of cells seeded × 100 (Franken,2006). The experiment was performed three times in duplicates.

2.6. Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

The quantity of caspase 3, caspase 4, caspase 8, caspase 9 and smac/DIABLO mRNA was determined by qRT-PCR (Gesing et al., 2011). TotalRNA from vehicle-treated control and GI50-treated cells was extractedaccording to the manufacture's instructions (Total RNA IsolationSystem, Sigma-Aldrich, USA). After RNA extraction, cDNA was pre-pared using Power cDNA synthesis kit (Intron Biotechnology, SouthKorea). Amplification of target cDNA for apoptosis markers and β-actin(as a normalization gene) was done using KAPA SYBR FAST qPCR KitMaster Mix (KAPA BIOSYSTEMS, USA) on Line Gene 9680 BioGR in-strument. cDNA (5 µl aliquots) was mixed with 1 µl of forward primer(25X), 1 µl reverse primer (25X) (Bian et al., 2009; Van Geelen et al.,2010; Yan et al., 2004; Yaoxian et al., 2013; Zhou et al., 2013), 5.5 µlnuclease free water and 12.5 µl master mix. All experiments wereperformed in triplicate. The relative mRNA quantity was normalized toβ-actin.

2.7. Cytochrome c releasing apoptosis assay

Vehicle-treated control and cells treated with the GI50 amounts ofCAs were collected by centrifugation at 600 g for 5 min, washed withice cold PBS and centrifuged at 600 g for 5 min and re-suspended with1X cytosolic extraction buffer mix containing dithiothreitol (DTT) andprotease inhibitors. The cells were incubated on ice for 10 min,homogenized and centrifuged at 700 g for 10 min. The supernatant wascentrifuged again at 10,000 g for 30 min to get the cytosolic fraction,while the pellet was resuspended in mitochondrial extraction buffer toget mitochondrial fraction (Cytochrome c Releasing Apoptosis AssayKit, Abcam, USA).

2.8. Western blot analysis

Vehicle-treated control and cells treated with GI50 amounts of CAswere lysed in cold lysis buffer (2% SDS, 10 mM Tris pH 7.5, 10 mMNaF, 2 mM EGTA, 10 mM dithiothreitol). Cell extract was heated in aboiling water bath for 5 min and sonicated. Aliquots of lysates werediluted in 4X SDS-PAGE sample buffer (0.5 M Tris-HCl pH 6.8, 2% SDS,20% glycerol, 20% 2-mercaptoethanol and 0.16% bromophenol blue)

Table 1The MTT (3-(4,5-dimethylthiazol-2-yl)−2,5 diphenyltetrazolium bromide) assay results.

CAA CAB

Human Cancer Cell Line GI50 µM TGI µM GI50 µM TGI µM

K562 (Leukemia) 0.012a 8.2a 0.013a 8.7a

MCF−7 (Breast) 1.2a 10.6a 1.2a 12.3c

DU−145 (Prostate) 0.12a 8.8a 0.23a 9.6a

HeLa (Cervical) 1.4a 14.3b 1.9a 16.2c

SAOS−2 (Osteosarcoma) 0.72a 12.7b 0.89a 15.4c

U2OS (Osteosarcoma) 0.45a 9.3a 0.67a 10.2a

a 24 h.b 48 h.c 72 h.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

402

Fig. 2. Cephalostatin analogue A and B induce apoptosis in cancer cells. Cells were treated with CAA or CAB (GI50). A) Fluorescence images of tumor cell lines showing fragmented nucleiafter DAPI staining. Scale bar: 10 µm, B) Results from flow cytometry analysis of annexin V-FITC and PI-stained cells. Live cells (Annexin V negative, PI negative), apoptotic cells (AnnexinV positive, PI negative) and necrotic cells (Annexin V positive, PI positive).

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

403

and proteins were resolved by electrophoresis on SDS-containing 10%or 12.5% polyacrylamide gels. Proteins were transferred onto ni-trocellulose membrane, blocked using 2% (w/v) bovine serum albumin(BSA) in Tris-buffered saline (TBS), and exposed overnight at 4 °C to thefollowing primary antibodies: mouse monoclonal anti-caspase 3 (1:500;Invitrogen, USA), mouse monoclonal anti-caspase 9 (1:1000;Invitrogen, USA), rabbit polyclonal anti-caspase 4 (1:1000; Invitrogen,USA), mouse monoclonal anti-caspase 8 (1:1000; Invitrogen, USA),rabbit polyclonal anti-smac/DIABLO (1:1000; R&D Systems, Germany),mouse GAPDH (1:6000; CHEMICON, USA), rabbit monoclonalphospho-eukaryotic initiation factor-2 (p-EIF-2) (1:2000; Cell Signaling

Technology, USA) and mouse monoclonal β-actin (1:2000; NovusBiologicals, USA), each diluted in 1% BSA in TBS containing 0.05%Tween 20. After washing and incubation with appropriate secondaryantibodies conjugated to IRDye 680 or 800 nm fluorescent dyes, themembranes were washed and the bands were visualized on FluorChemR system (Oxford, UK). Signals were quantified using AlphaView soft-ware (ProteinSimple, USA).

2.9. Statistical analysis

All experiments were performed in triplicates. Results are expressedas mean± S.E.M. Statistical analysis was performed using GraphPadPrism version 5.0 (GraphPad Software, San Diego, CA, USA). To de-termine differences between 3 or more means, one-way ANOVA withFisher's LSD for multiple comparisons post-tests were performed. Pvalue< 0.05 was considered significant.

Fig. 3. Cephalostatin analogue-induced apoptosis is dose-dependent. Cells were treatedfor 24 h with 104 range of concentrations of CAA or CAB and the percentage of non-viablecells was quantified. The percentage of non-viable cells relative to the vehicle-treatedcontrol cells measured by MTT assay. Bars = mean±S.E.M. of three independent ex-periments performed in triplicates.* P< 0.05; ** P< 0.01; *** P<0.001 compared tovehicle-treated control cells.

Fig. 4. Cephalostatin analogue-induced apoptosis is cancer-specific. White blood cells,ARPE-19 and GM2149 human normal cells were treated with 104 range of concentrationsof CAA or CAB for 72 h and the percentage of non-viable cells was quantified. Percentageof non-viable cells relative to the vehicle-treated control cells is measured by MTT assay.Bars = mean±S.E.M. of three independent experiments performed in triplicates. *P<0.05; ** P< 0.01 compared to vehicle-treated control cells.

Fig. 5. CAA and CAB inhibit clonogenic tumor growth. A) Representative images showingcolonies produced by vehicle-treated control and CA treated cells following plating of1000 cells and 12 days incubation, B) Cells were treated for 3 h with 104 range of con-centrations of CAA or CAB and the number of colonies was quantified after 12 days.Cephalostatin analogue treatment caused an inhibition of clonogenic tumor growth.Bars= mean±S.E.M. of three independent experiments performed in duplicates. *P< 0.05; ** P< 0.01; *** P<0.001 compared to vehicle-treated control cells.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

404

3. Results

3.1. Cephalostatin analogues CAA and CAB induce apoptosis in multipletumor cell lines but not in normal cells

The cytotoxic activities of CAA and CAB were evaluated against sixdifferent human cancer cell lines. These lines and the correspondingGI50 (50% Growth Inhibition) and TGI (Total Growth Inhibition) valuesfor each CA analogue are shown in Table 1. Unless otherwise stated, the

amounts of CAA and CAB used in subsequent experiments were the GI50values from this table.

The cytotoxicity of CAA and CAB was attributed to the induction ofapoptosis as indicated by nuclei fragmentation detected by fluorescencemicroscopy after DNA staining with DAPI (Fig. 2A) and by flow cyto-metry analysis of annexin V-FITC and PI staining (Fig. 2B). The cyto-toxicity results at 24 h presented in Fig. 3 show that apoptosis inductionby CAA and CAB occurred in a dose-dependent manner. In addition, thecytotoxicity of CAA and CAB was tested against human normal white

Fig. 6. CAA- and CAB-induced apoptosis is dependent on caspases activation. A) qRT-PCR analysis of caspase 3, 8 and 9 mRNA in CAA- and CAB-treated cells (GI50) as compared tovehicle-treated control cells [set as 1 arbitrary unit (a.u.)]. Note that MCF-7 cells do not express caspase 3. Values were normalized to β-actin. B) Representative Western blots showingcleavage “activation’ of procaspase 3 to the active form p17, procaspase 9 to the active forms p35 and p37 and procaspase 8 to the active forms p42 and p44 in K562-treated cells (GI50).Etoposide (ET) was used (25 µg/ml, 24 h) as positive control (classical caspases-dependent inducer of apoptosis). Similar results were achieved in MCF-7 (except for caspase 3 activation)and DU-145 cancers cells. The experiment was repeated three times and the corresponding quantification is shown in (C). C) Quantification of procaspase 3, 8 and 9 and their cleavedforms in CAA-; CAB- and ET-treated K562 relative to the control values (set as 1.0) obtained from densitometers of immunoblots. D) Inhibition of CAA and CAB-induced apoptosis by thepan caspase inhibitor zVAD-fmk. Cells were treated with 0.1% DMSO (control), treated with CAA or CAB (GI50), or pretreated with zVAD-fmk (25 µM, 1 h) and then treated with CAA orCAB. Percent non-viable cells: percentage of non-viable cells relative to the vehicle-treated control cells measured by MTT assay. Similar results were achieved in HeLa, SAOS-2 and U2OScancers cells. Bars = mean± S.E.M. of three independent experiments performed in triplicates. * P < 0.05; ** P< 0.01; *** P< 0.001 compared to vehicle-treated control cells, ψψψ

P<0.001 compared to etoposide-treated cells.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

405

blood cells (WBC), ARPE-19 retinal pigmented epithelial and GM2149skin fibroblast cells (Fig. 4, but note change in scale from Fig. 3). Ourresults revealed that neither analogue had greater than 14% cytotoxi-city in normal cells even after 72 h of treatment at the highest (10 µM)drug concentration tested (Fig. 4). Thus, the cytotoxic effect of CAA andCAB is directed toward cancer cells.

3.2. Cephalostatin analogues (CAs) inhibit clonal tumor cell growth

In vitro clonogenic assays are used commonly to investigate the longterm survival of cells which were irradiated or treated with a che-motherapeutic agent by measuring the proliferative ability of a singlecell to form a colony (Fig. 5A) (Sumantran, 2011). The results presentedin Fig. 5 show that cancer cell treatment for 3 h with either CAA or CABresulted in a dose-dependent inhibition of clonogenic tumor growth(Fig. 5B).

3.3. Apoptosis induced by both CAA and CAB is dependent on the activationof caspases

To investigate if caspase activation occurs during CAs-inducedapoptosis, the mRNA levels of caspase 3, caspase 8 and caspase 9 werequantified by qRT-PCR (Fig. 6A) and their activation was assessed byWestern blot analysis (Fig. 6B). The quantity of caspase 3 mRNA in-creased in CA-treated K562 and DU-145 cell lines and caspase 9 mRNAincreased in all of the tested cancer cell lines as compared to vehicle-treated cells (Fig. 6A). On the other hand, caspase 8 mRNA levels de-creased in all of the treated cells as compared to vehicle-treated controlcells, although the decrease was not significant (Fig. 6A).

In response to CAA, CAB and etoposide (caspase-dependent drug)treatment, K562 and DU-145 cells showed evidence of cleavage ofprocaspase 3 to the active form p17 and procaspase 9 was cleaved to theactive forms p35 and p37 (Fig. 6B and C). MCF-7 cells showed cleavageof caspase 9 but not caspase 3. Caspase 8 cleavage products (p42 andp44) appeared in cells treated with etoposide but not in cells treatedwith CAA or CAB (Fig. 6B and C). The densitometry values (Fig. 6C)from the blots shown in Fig. 6B were normalized to actin or GAPDH andthen expressed relative to the caspase pro- or cleaved form set at 1.0 invehicle-treated control cells. In addition, cells pretreated with the pancaspase inhibitor (zVAD-fmk) (25 µM, 1 h) prior to treatment withetoposide or CAs at the GI50 concentration were protected from cyto-toxicity (Fig. 6D). These results indicate the importance of caspaseactivation in CAs-induced apoptosis.

3.4. CAA and CAB induce the release of smac/DIABLO but not cytochromec from the mitochondria

The activation of caspase 9 in response to cell treatment with CAAor CAB (Fig. 6B) indicated a possible role for the mitochondria-medi-ated apoptotic pathway. Two of the proapoptotic factors released frommitochondria after various proapoptotic stimuli are cytochrome c andsmac/DIABLO. Treatment of K562, MCF-7 and DU-145 cancer cell lineswith CAA or CAB did not result in release of cytochrome c (CF fraction)from mitochondria (MF fraction), although the same cells treated withetoposide as a positive control showed cytochrome c release (Fig. 7Aand B). On the other hand, smac/DIABLO was released from mi-tochondria in cells treated with CAA, CAB or etoposide (Fig. 7A and B).The amount of smac/DIABLO mRNA increased relative to vehicle-treated controls in all three tested cancer cell lines treated with CAA orCAB (Fig. 7C).

3.5. CAA and CAB induce ER stress in tumor cells

Since CAs did not induce cytochrome c release or caspase 8 acti-vation in cancer cells, we hypothesized that CA treatment led toapoptosis through the ER stress pathway. To test this, we examined thephosphorylation of eIF-2, a target of ER stress that is initiated bythapsigargin, a well-characterized inducer of ER stress (Oslowski andUrano, 2011). Treatment of the six different cancer cell lines with GI50amounts of CAA, CAB or thapsigargin (3 µM, 2 h), induced a strongincrease in phospho-eIF-2 levels (Fig. 8A and B).

3.6. Cephalostatin analogues-induced caspase 4 activation is necessary forapoptosis

During ER stress-mediated apoptosis, caspase 4 is localized to andactivated at the ER membrane where it is the specific caspase involvedin triggering the apoptotic response (Bravo-Sagua et al., 2013; Iurlaroand Muñoz-Pinedo, 2016). CAA, CAB and thapsigargin treatment ofK562, MCF-7 and DU-145 cancer cells results in increased cleavage ofcaspase 4 indicated by the reduction in procaspase 4 after treatmentand the appearance of the cleavage product p20 (Fig. 8C and D). Inaddition, the quantity of caspase 4 mRNA increased in all of the tested

Fig. 7. CAA and CAB induce the release of smac/DIABLO but not cytochrome c from themitochondria into the cytosol. A) Representative Western blot showing the release ofsmac/DIABLO but not cytochrome c from the mitochondria (MF) into the cytosol (CF) oftreated K562 cancer cells (GI50). Similar results were achieved in MCF-7 and DU-145cancer cells. Etoposide (ET) was used as a positive control (induces the translocation ofboth smac/DIABLO and cytochrome c from the mitochondria into the cytosol). Equalprotein loading was controlled by staining membranes with Ponceau S (a representativesection of the stained membrane is shown). The experiment was repeated three times andthe corresponding quantification is shown in (B). B) Quantification of smac/DIABLO andcytochrome C levels in vehicle-treated control K562 cells and cells treated with GI50amounts of CAA and CAB or etoposide as a positive control. Relative levels of smac/DIABLO and cytochrome c in treated cells compared to controls (set as 1.0) obtained fromdensitometry of immunoblots. MF: mitochondria fraction; CF: cytosol fraction. C) qRT-PCR analysis of smac/DIABLO mRNA in CAA- and CAB-treated cells as compared to ve-hicle-treated control cells [set as 1 arbitrary unit (a.u.)]. Values were normalized to β-actin. Bars = mean± S.E.M. of three independent experiments performed in triplicates. *P<0.05; ** P<0.01; *** P< 0.001 compared to vehicle-treated control cells; ψψψ

P<0.001 compared to etoposide-treated cells.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

406

cell lines treated with CAA and CAB compared to vehicle-treated con-trol cells (Fig. 8E). Furthermore, cells pretreated with the caspase 4inhibitor Ac-LEVD-CHO (20 µM, 1 h) were protected from apoptosisinduced by CAA or CAB (Fig. 8F). Together these results strongly sug-gest that the cephalostatin 1 analogues A and B direct death of cancercells through the ER-mediated apoptotic pathway.

4. Discussion

Chemotherapy, a cornerstone in the development of present daycancer therapy, is one of the most effective and potent strategies to treattumors. Several cytotoxic compounds have proved to induce tumorregression. However, most anticancer drugs lack tumor specificity,

cause damage to normal tissues and lead to adverse effects. In somecases, drug-resistant cancer cells remain alive after chemotherapycausing chemoresistance (Liang et al., 2010). There is still a need fordeveloping effective anticancer drugs with new mechanisms of actionsuch as growth factors, cell-cycle proteins, modulators of apoptosis andmolecules that inhibit angiogenesis (Masui et al., 2013).

Cephalostatin 1, a bis-steroidal compound, is a potent anti-tumordrug. Key features in the biological activity of cephalostatin 1 are po-sition-12 functionality, the presence of the Δ14 bond, the spiroketalmoiety, the 17α-hydroxyl group and the polarity difference betweenthe two steroidal halves (Iglesias-Arteaga and Morzycki, 2013). How-ever, isolation of cephalostatin 1 from its natural source and difficultiesin its chemical synthesis (Gryszkiewicz et al., 2003; Lee et al., 2009; Li

Fig. 8. CAA and CAB induce endoplasmic reticulum-mediated apoptosis. A) Representative Western blot showing the induction of phosphorylation of eukaryotic initiation factor-2 (eIF-2)in K562 cells treated with GI50 amounts of CAA and CAB. For comparison purposes, thapsigargin (TG), an ER stress-causing drug was used (3 µM, 2 h) as a positive control. Similar resultswere achieved in the other five cancers cell lines. The experiment was repeated three times and the corresponding quantification is shown in (B), B) Quantification of PeIF-2 levels inCAA-, CAB- and thapsigargin-treated K562 cells. Relative levels of PeIF-2 in K562-treated cells compared to controls (set as 1.0) obtained from densitometry of immunoblots. C)Representative Western blot showing cleavage “activation’ of procaspase 4 to the active form p20 in K562 cancer cells. Similar results were achieved in MCF-7 and DU-145 cancers cells.The experiment was repeated three times and the corresponding quantification is shown in (D). D) Quantification of procaspase 4 and its cleaved form in CAA-; CAB- and TG-treated K562relative to the control values (set as 1.0) obtained from densitometers of immunoblots. E) qRT-PCR analysis of caspase 4 mRNA in CAA- and CAB-treated cells as compared to vehicle-treated control cells [set as 1 arbitrary unit (a.u.)]. Values were normalized to β-actin. F) Inhibition of CAA and CAB-induced apoptosis by the caspase 4-specific inhibitor Ac-LEVD-CHO.Cells were treated with 0.1% DMSO (control), treated with CAA or CAB (GI50), or pretreated with Ac-LEVD-CHO (20 µM, 1 h) and then treated with CAA or CAB. Percent non-viable cells:percentage of non-viable cells relative to the vehicle-treated control cells measured by MTT assay. Bars= mean±S.E.M. of three independent experiments performed in triplicates. **P<0.01; *** P< 0.001 compared to vehicle-treated control cells; ψ P<0.05 compared to thapsigargin-treated cells.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

407

and Dias, 1997; Poza et al., 2010), have led to an interest in thesynthesis of cephalostatin 1 analogues (CAs) as potential antitumoragents.

The selected route to synthesize CAs with potent activity was viaproper functionalization that could partially fulfill these requirementsyielded the four diol isomeric mixture 12'β, 12α and 12'β, 12β (as twomajors) and 12'α, 12α (CAA); 12'α, 12β (CAB) (as two minors)(Nawasreh, 2007, 2008). Previous work showed the more potent bio-logical activity resided with the minor isomers (Nawasreh, 2007), andhere we scaled up the synthesis of the minor isomers, CAA and CAB, inwhich the hydroxyl group at position-12' has α-configuration. Wesuggest that the enhancement in activity of CAA and CAB is based onstructural comparison between these compounds and cephalostatin 1.Although the interacting partners of cephalostatin 1 are unknown, wesuggest that the α-configurated hydroxyl group at the 12'-position inCAA and CAB could act similarly to the hydroxyl group at position-17 incephalostatin1 (Fig. 1). In addition, or alternatively, the hydroxymethylgroup at position-12 of CAA could substitute for the hydroxyl group atposition 17 of cephalostatin 1. The oxygenated environment aroundposition 18’ of cephalostatin 1 might also be mimicked by a similarenvironment around position 18’ of both CAA and CAB.

CAA and CAB treatment caused cytotoxicity in the six tested celllines (Table 1). This cytotoxicity was attributed to apoptosis as evidentby DNA fragmentation (Fig. 2A) and exposure of phosphatidylserine onthe outside of the plasma membrane (Annexin V- FITC staining;Fig. 2B). In addition, CAs-induced apoptosis is cancer cell line in-dependent (Fig. 3) and cancer cell specific (Fig. 4). A higher con-centration of CAs was needed (almost 100-fold higher concentration)for cell death induction in MCF-7 and HeLa cells than other cancer linestreated (1). However, even with these cell lines and using the in vitrolong-term clonogenic assay, CAs treatment was able to induce inhibi-tion of clonal tumor growth (Fig. 5), suggesting that these cephalostatinanalogues might be useful to circumvent chemoresistance (Beier et al.,2011).

CAs-induced apoptosis is caspase-dependent in the cell lines tested(Fig. 6A-C) since CAs-induced cell death was inhibited by the broadspectrum caspase inhibitor zVAD-fmk (Fig. 6D). The same result wasachieved for etoposide, a classical caspases-dependent inducer ofapoptosis (Fig. 6D). Apoptosis in MCF-7 breast cancer cells lackingcaspase 3 due to a deletion mutation was also induced after CA treat-ment (data not shown) indicating that another effector caspase such ascaspase 6 or 7 may have compensated for caspase 3 (Kurokawa et al.,1999; Semenov et al., 2003).

Cephalostatin 1 was shown to use the ER stress-induced apoptoticpathway involving caspase 4 independent of cytochrome c release andcaspase 8 activation, and it selectively uses smac/DIABLO as a mi-tochondrial signaling molecule (Dirsch et al., 2003; Imperatore et al.,2014). The findings presented in the current work suggest that CAA andCAB function to induce apoptosis in a highly similar manner to ce-phalostatin 1 rather than through the extrinsic or intrinsic apoptoticpathways (Hwang and Kim, 2015; Kantari and Walczak, 2011; Lópezet al., 2006). The following results support this: a) CA treatment did notcause caspase 8 mRNA level or cleavage products to increase (Fig. 6Aand B); b) CA treatment did not cause the release of cytochrome c fromthe mitochondria (Fig. 7A), however, smac/DIABLO was released fromthe mitochondria into the cytosol; c) CA treatment caused the phos-phorylation of the eukaryotic initiation factor-2 (eIF-2; Fig. 8A) and d)ER stress caused the activation of caspase 4 the main player in ER-mediated apoptosis (Fig. 8B). CA-induced cell death is dependent oncaspase 4 activation; the use of caspase 4 specific inhibitor z-LEVD-fmkled to a marked inhibition of CA-induced cell death (Fig. 8D). More-over, since the two isomers are very comparable in their in vitro activityand mechanism of action, we anticipate that the separation of eachisomer in pure form might not be necessary, increasing their yield anddecreasing their cost.

5. Conclusion

In summary, the current work showed that the two 12'α-derivativesof cephalostatin 1 induced cell death by activating the atypical ER stressapoptosis signaling pathway indicting their potential applicability tocancer treatment. In addition, CAA and CAB were shown to targetcancer cells over normal cells showing a 102 to 103 concentration biastoward human cancer cell lines over human normal cells. Nevertheless,several questions remain unanswered; how is caspase 9 activated in theabsence of cytochrome c release and apoptosome formation? Is caspase9 activation essential for CAs-induced apoptosis? Does caspase 4 actupstream or downstream of caspase 9? Would the over expression ofBcl-xL block smac/DIABLO release and apoptosis? The promising cy-totoxic activity results obtained, in addition to the availability of suf-ficient amounts of CAA and CAB to carry out further biological studymakes this research of high priority in the future.

Acknowledgments

The authors are grateful to the Deanship of Research and GraduateStudies, Hashemite University for supporting the current work. Wegratefully acknowledge the College of Natural Sciences, Colorado StateUniversity for the Summer International Scholar Award for LHT anddonors to the Colorado State University Foundation Development Fundsupporting research in the Bamburg laboratory.

Conflict of interest

The authors declare that they have no conflict of interest.

References

Abdullah, L.N., Chow, E.K., 2013. Mechanisms of chemoresistance in cancer stem cells.Clin. Transl. Med. 2, 3.

Beier, D., Schulz, J.B., Beier, C.P., 2011. Chemoresistance of glioblastoma cancer stemcells - much more complex than expected. Mol. Cancer 10, 128.

Bian, Z., Elner, S., Elner, V., 2009. Dual involvement of caspase-4 in inflammatory and erstress-induced apoptotic responses in human retinal pigment epithelial cells. Investig.Ophthalmol. Vis. Sci. 50, 6006–6014.

Bravo-Sagua, R., Rodriguez, A.E., Kuzmicic, J., Gutierrez, T., Lopez-Crisosto, C., Quiroga,C., Díaz-Elizondo, J., Chiong, M., Gillette, T.G., Rothermel, B.A., Lavandero, S., 2013.Cell death and survival through the endoplasmic reticulum-mitochondrial axis. Curr.Mol. Med. 13, 317–329.

Breckenridge, D.G., Germain, M., Mathai, J.P., Nguyen, M., Shore, G.C., 2003. Regulationof apoptosis by endoplasmic reticulum pathways. Oncogene 22, 8608–8618.

Dirsch, V., Müller, I., Eichhorst, S., Pettit, G., Kamano, Y., Inoue, M., Xu, J., Ichihara, Y.,Wanner, G., Vollmar, A., 2003. Cephalostatin 1 selectively triggers the release ofsmac/diablo and subsequent apoptosis that is characterized by an increased densityof the mitochondrial matrix. Cancer Res. 63, 8869–8876.

Elmore, S., 2007. Apoptosis: a review of programmed cell death. Toxicol. Pathol. 35,495–516.

Fotakis, G., Timbrell, J.A., 2006. In vitro cytotoxicity assays: comparison of LDH, neutralred, MTT and protein assay in hepatoma cell lines following exposure to cadmiumchloride. Toxicol. Lett. 160, 171–177.

Franken, N.A., Rodermond, H.M., Stap, J., Haveman, J., van Bree, C., 2006. Clonogenicassay of cells in vitro. Nat. Protoc. 1, 2315–2319.

Gatti, L., Zunino, F., 2005. Overview of tumor cell chemoresistance mechanisms. MethodsMol. Med. 111, 127–148.

Gesing, A., Masternak, M.M., Wang, F., Lewinski, A., Karbownik-Lewinsk, M., Bartke, A.,2011. Decreased expression level of apoptosis-related genes and/or proteins in ske-letal muscles, but not in hearts, of growth hormone receptor knockout mice. Exp.Biol. Med. 236, 156–168.

Groenendyk, J., Michalak, M., 2005. Endoplasmic reticulum quality control and apop-tosis. Acta Biochim. Pol. 52, 381–395.

Gryszkiewicz, A., Jastrzebska, I., Morzycki, J., Romanowska, D., 2003. Approaches to-wards the synthesis of cephalostatins, ritterazines and saponins from ornithogalumsaundersiae-new natural products with cytostatic activity. Curr. Org. Chem. 7,1257–1277.

Guo, C., Bhandaru, S., Fuchs, P., Boyd, M., 1996. An efficient protocol for the synthesis ofunsymmetrical pyrazines. Total synthesis of dihydrocephalostatin 11. J. Am. Chem.Soc. 118, 10672–10673.

Hammond, W.A., Swaika, A., Mody, K., 2016. Pharmacologic resistance in colorectalcancer: a review. Ther. Adv. Med. Oncol. 8, 57–84.

Housman, G., Byler, S., Heerboth, S., Lapinska, K., Longacre, M., Snyder, N., Sarkar, S.,2014. Drug resistance in cancer: an overview. Cancers 6, 1769–1792.

Hwang, H., Kim, H., 2015. Chondrocyte apoptosis in the pathogenesis of osteoarthritis.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

408

Int. J. Mol. Sci. 16, 26035–26054.Iglesias-Arteaga, M.A., Morzycki, J.W., 2013. Cephalostatins and ritterazines. Alkaloids

Chem. Biol. 72, 153–279.Imperatore, C., Aiello, A., D'aniello, F., Senese, M., Menna, M., 2014. Alkaloids from

marine invertebrates as important leads for anticancer drugs discovery and devel-opment. Molecules 19, 20391–20423.

Iurlaro, R., Muñoz-Pinedo, C., 2016. Cell death induced by endoplasmic reticulum stress.FEBS J. 283, 2640–2652.

Kantari, C., Walczak, H., 2011. Caspase-8 and bid: caught in the act between death re-ceptors and mitochondria. Biochim. Biophys. Acta 1813, 558–563.

Kurokawa, H., Nishio, K., Fukumoto, H., Tomonari, A., Suzuki, T., Saijo, N., 1999.Alteration of caspase-3 (CPP32/Yama/apopain) in wild-type MCF-7, breast cancercells. Oncol. Rep. 6, 33–37.

Lee, S., LaCour, T.G., Fuchs, P.L., 2009. Chemistry of trisdecacyclic pyrazine anti-neoplastics: the cephalostatins and ritterazines. Chem. Rev. 109, 2275–2314.

Li, W., Fuchs, P., 2003. An efficient synthesis of the c-23 deoxy, 17α-hydroxy south 1hemisphere and its cephalostatin 1 analog1. Org. Lett. 5, 2849–2852.

Li, W., Lacour, T., Fuchs, P., 2002. Dyotropic rearrangement facilitated proximal func-tionalization and oxidative removal of angular methyl groups: efficient syntheses of23'-deoxy cephalostatin 1 analogues 1. J. Am. Chem. Soc. 124, 4548–4549.

Li, Y., Dias, J., 1997. Dimeric and oligomeric steroids. Chem. Rev. 97, 283–304.Liang, X., Chen, C., Zhao, Y., Wang, P., 2010. Circumventing tumor resistance to che-

motherapy by nanotechnology. Methods Mol. Biol. 596, 467–488.López, N., Rudy, A., Barth, N., Schmitz, L., Pettit, G., Schulze, K., Dirsch, V., Vollmar, A.,

2006. The marine product cephalostatin 1 activates an endoplasmic reticulum stress-specific and apoptosome-independent apoptotic signaling pathway. J. Biol. Chem.281, 33078–33086.

Masui, K., Gini, B., Wykosky, J., Zanca, C., Mischel, P., Furnari, F., Cavenee, W., 2013. Atale of two approaches: complementary mechanisms of cytotoxic and targetedtherapy resistance may inform next generation cancer treatments. Carcinogenesis 34,725–738.

Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival: applicationto proliferation and cytotoxicity assays. J. Immunol. Methods 65, 55–63.

Muller, I.M., Dirsch, V.M., Rudy, A., López-Antón, N., Pettit, G.R., Vollmar, A.M., 2005.Cephalostatin 1 inactivates Bcl-2 by hyperphosphorylation independent of M-phasearrest and DNA damage. Mol. Pharmacol. 67, 1684–1689.

Nawasreh, M., 2000. Stereoselective Synthesis of Cephalostatin Analogues as AnticancerAgents. Ph.D. Thesis. Leibniz Hannover University, Germany.

Nawasreh, M., 2007. Stereoselective synthesis of bis-steroidal pyrazine derivatives. Nat.Prod. Res. 21, 91–99.

Nawasreh, M., 2008. Chemo-, regio-, and stereoselectivity of F-ring opening reactions in

the cephalostatin series. Bioorg. Med. Chem. 16, 255–265.Oslowski, C.M., Urano, F., 2011. Measuring ER stress and the unfolded protein response

using mammalian tissue culture system. Methods Enzymol. 490, 71–92.Parrish, A., Freel, C., Kornbluth, S., 2013. Cellular mechanisms controlling caspase acti-

vation and function. Cold Spring Harb. Perspect. Biol. 5, 1–25.Pettit, G.R., Inoue, M., Kamano, Y., Herald, D.L., Arm, C., Dufresne, C., Christie, N.D.,

Schmidt, J.M., Doubek, D.L., Krupa, T.S., 1988. Isolation and structure of the pow-erful cell-growth inhibitor cephalostatin-1. J. Am. Chem. Soc. 110, 2006–2007.

Pettit, G.R., Mendonça, R.F., Knight, J.C., Pettit, R.K., 2011. The cephalostatins.21. Synthesis of bis-steroidal pyrazine rhamnosides (1). J. Nat. Prod. 74, 1922–1930.

Poza, J., Rodríguez, J., Jiménez, C., 2010. Synthesis of a new cytotoxic cephalostatin/ritterazine analogue from hecogenin and 22-epi-hippuristanol. Bioorg. Med. Chem.18, 58–63.

Rafehi, H., Orlowski, C., Georgiadis, G.T., Ververis, K., El-Osta, A., Karagiannis, T.C.,2011. Clonogenic assay: adherent cells. J. Vis. Exp. 49. http://dx.doi.org/10.3791/2573.

Rudy, A., López-Antón, N., Barth, N., Pettit, G.R., Dirsch, V.M., Schulze-Osthoff, K., Rehm,M., Prehn, J.H., Vogler, M., Fulda, S., Vollmar, A.M., 2008. Role of Smac in cepha-lostatin-induced cell death. Cell Death Differ. 15, 1930–1940.

Semenov, D., Aronov, P., Kuligina, E., Potapenko, M., Richter, V., 2003.Oligonucleosomal DNA fragmentation in MCF-7 cells undergoing palmitate-inducedapoptosis. Biochemistry 68, 1335–1341.

Siegel, R.L., Miller, K.D., Jemal, A., 2016. Cancer statistics, 2016. CA Cancer J. Clin. 66,7–30.

Sumantran, V.N., 2011. Cellular chemosensitivity assays: an overview. Methods Mol. Biol.731, 219–236.

Van Geelen, C.M., Pennarun, B., Ek, W.B., Le, P.T., Spierings, D.C., De Vries, E.G., DeJong, S., 2010. Downregulation of active caspase 8 as a mechanism of acquired trailresistance in mismatch repair-proficient colon carcinoma cell lines. Int. J. Oncol. 37,1031–1041.

Yan, Y., Mahotka, C., Heikaus, S., Shibata, T., Wethkamp, N., Liebmann, J., Suschek, C.V.,Guo, Y., Gabbert, H.E., Gerharz, C.D., 2004. Disturbed balance of expression betweenxiap and smac/diablo during tumour progression in renal cell carcinomas. Br. J.Cancer 91, 1349–1357.

Yaoxian, W., Hui, Y., Yunyan, Z., Yanqin, L., Xin, G., Xiaoke, W., 2013. Emodin inducesapoptosis of human cervical cancer hela cells via intrinsic mitochondrial and extrinsicdeath receptor pathway. Cancer Cell Int. 13, 71.

Zhou, Y., Su, J., Shi, L., Liao, Q., Su, Q., 2013. DADS downregulates the Rac1-ROCK1/PAK1-LIMK1-ADF/cofilin signaling pathway, inhibiting cell migration and invasion.Oncol. Rep. 29, 605–612.

L.H. Tahtamouni et al. European Journal of Pharmacology 818 (2018) 400–409

409