Naja Atra Cardiotoxin 3 Elicits Autophagy and Apoptosis in U937 Human Leukemia Cells Through the Ca2+/PP2A/AMPK Axis

Article Naja atra Cardiotoxin 3 Elicits Autophagy and Apoptosis in U937 Human Leukemia Cells through the Ca2+/PP2A/AMPK Axis

Jing-Ting Chiou 1, Yi-Jun Shi 1, Liang-Jun Wang 1, Chia-Hui Huang 1, Yuan-Chin Lee 1 and Long-Sen Chang 1,2,*

1 Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan 2 Department of Biotechnology, Kaohsiung Medical University, Kaohsiung 807, Taiwan * Correspondence: [email protected]; Tel.: +886-7-5252-000-5813

Received: 23 August 2019; Accepted: 10 September 2019; Published: 12 September 2019

Abstract: Cardiotoxins (CTXs) are suggested to exert their cytotoxicity through cell membrane damage. Other studies show that penetration of CTXs into cells elicits mitochondrial fragmentation or lysosome disruption, leading to cell death. Considering the role of AMPK-activated protein kinase (AMPK) in mitochondrial biogenesis and lysosomal biogenesis, we aimed to investigate whether the AMPK-mediated pathway modulated Naja atra (Taiwan cobra) CTX3 cytotoxicity in U937 human leukemia cells. Our results showed that CTX3 induced autophagy and apoptosis in U937 cells, whereas autophagic inhibitors suppressed CTX3-induced apoptosis. CTX3 treatment elicited Ca2+-dependent degradation of the protein phosphatase 2A (PP2A) catalytic subunit (PP2Acα) and phosphorylation of AMPKα. Overexpression of PP2Acα mitigated the CTX3-induced AMPKα phosphorylation. CTX3-induced autophagy was via AMPK-mediated suppression of the Akt/mTOR pathway. Removal of Ca2+ or suppression of AMPKα phosphorylation inhibited the CTX3-induced cell death. CTX3 was unable to induce autophagy and apoptosis in U937 cells expressing constitutively active Akt. Met-modiﬁed CTX3 retained its membrane-perturbing activity, however, it did not induce AMPK activation and death of U937 cells. These results conclusively indicate that CTX3 induces autophagy and apoptosis in U937 cells via the Ca2+/PP2A/AMPK axis, and suggest that the membrane-perturbing activity of CTX3 is not crucial for the cell death signaling pathway induction.

Keywords: cardiotoxin; Ca2+; PP2A; AMPK; autophagy; apoptosis

Key Contribution: The cytotoxicity of N. atra CTX3 on U937 cells is related to the Ca2+/PP2A/AMPK axis-elicited autophagy and apoptosis; Our data suggest a dissociation of CTX3-induced death signaling from its eﬀect on cell membrane damage.

1. Introduction Cobra cardiotoxins (CTXs) show a wide range of biological activities including hemolysis, cardiotoxicity, and disruption of muscular integrity [1–3]. Moreover, accumulated evidence has shown that CTXs induce apoptosis in cancer cells through the mitochondria-mediated apoptotic pathway, lysosome-mediated death pathway, or membrane-damaging eﬀects [4–8]. It is widely believed that the plasma membrane is a target of CTXs, but the membrane pore formed by CTXs may have a limited lifetime on the cell surface due to membrane reorganization [9]. Therefore, some studies indicate that the cytotoxicity of CTXs is related to their capability of cell penetration [4,7,10,11]. The internalized CTXs are suggested to elicit mitochondrial fragmentation [10,11] or lysosome disruption [4,7], and thereby induce the death of the myoblast cells, primary cortical neurons, or cancer cells. Notably, Langone

Toxins 2019, 11, 527; doi:10.3390/toxins11090527 www.mdpi.com/journal/toxins Toxins 2019, 11, x FOR PEER REVIEW 2 of 12

Toxinslysosome2019, 11 disruption, 527 [4,7], and thereby induce the death of the myoblast cells, primary cortical2 of 12 neurons, or cancer cells. Notably, Langone et al. [12] found that activated AMPK can attenuate the etcytotoxicity al. [12] found of Naja that activatedpallida CTX AMPK on myoblast can attenuate cells. AMPK the cytotoxicity has been ofdemonstratedNaja pallida CTX to be on involved myoblast in cells.mitochondrial AMPK has biogenesis been demonstrated [13,14] and to belysosomal involved biogenesis in mitochondrial [15]. Some biogenesis studies [13 have,14] and revealed lysosomal that biogenesisAMPK acts [15 as]. a Sometumor studies suppressor have or revealed oncogene that in AMPKdifferent acts cancer as a tumorcells [16]. suppressor Other studies or oncogene indicate inthat di ﬀAMPK-mediatederent cancer cells signaling [16]. Other elicits studies death indicate in cancer that AMPK-mediatedcells via autophagy signaling and/or elicitsapoptosis death [17]. in cancerTherefore, cells it via is autophagyintriguing to and explore/or apoptosis the role [17 of]. the Therefore, AMPK-mediated it is intriguing pathway to explore in the thecytotoxicity role of the of AMPK-mediatedCTXs. Prior studies pathway have inshown the cytotoxicity that human of myeloid CTXs. Prior leukemia studies cells have are shown highly that susceptible human myeloid to the leukemiacytotoxicity cells of are CTXs highly [4,7,18]. susceptible Therefore, to the we cytotoxicity investigated of whether CTXs [4 ,7the,18 AMPK-associated]. Therefore, we investigated pathway is whetheran important the AMPK-associated mediator in N. atra pathway CTX3-induced is an important death ofmediator human leukemia in N. atra U937CTX3-induced cells. death of human leukemia U937 cells. 2. Results 2. ResultsTreatment with CTX3 reduced the U937 survival of cells in a concentration- and time-dependent mannerTreatment with an with IC50 CTX3 value reduced of approximately the U937survival 150 nM for of cells a 4-h in treatment a concentration- (Figure and1A). time-dependent Hence, we used mannerthis dose with of CTX3 an IC 50tovalue study of the approximately mechanism of 150 itsnM cytotoxicity. for a 4-htreatment CTX3 treatment (Figure 1increasedA). Hence, the we number used thisof annexin dose of CTX3V-FITC to staining study the U937 mechanism cells (Figure of its 1B). cytotoxicity. In line with CTX3 this, treatment the CTX-treated increased cells the numbershowed ofdegradation annexin V-FITC of procaspase-3 staining U937 and PARP cells (Figure (Figure1 B).1C), In whereas line with the this, caspase the CTX-treated inhibitors inhibited cells showed the cell degradationdeath induced of procaspase-3by CTX3 (Figure and PARP1D). These (Figure results1C), whereas indicated the that caspase CTX3 inhibitors induces apoptosis inhibited thein U937 cell deathcells. induced by CTX3 (Figure1D). These results indicated that CTX3 induces apoptosis in U937 cells.

Figure 1. Cobra cardiotoxin (CTX)3 induced apoptotic death of U937 cells. (A)Eﬀect of CTX3 on the viabilityFigure 1. of Cobra U937 cells.cardiotoxin Cells were (CTX)3 incubated induced with apoptotic indicated death CTX3 of U937 concentrations cells. (A) Effect for 4 h.of (Inset)CTX3 on U937 the cellsviability were of treated U937 withcells. 150Cells nM were CTX3 incubated for indicated with indicated time periods. CTX3 Cell concentrations viability was for determined 4 h. (Inset) using U937 methlythiazolyldiphenyl-tetrazoliumcells were treated with 150 nM CTX3 bromide for indicated (MTT) time assay. periods. Results Cell are viability expressed was as determined the percentage using of cellmethlythiazolyldiphenyl-tetrazolium survival relative to the control. Each bromide value is the(MTT) mean assay.SD Results of three are independent expressed experimentsas the percentage with ± triplicateof cell survival measurements; relative to (B the) Flow control. cytometry Each va analyseslue is the of mean±SD annexin V-PI of three double independent staining CTX3-treated experiments cells.with U937triplicate cells measurements; were incubated (B with) Flow 150 cytometry nM CTX3 analyses for 4 h. Onof annexin the ﬂow V-PI cytometric double scatterstaining graphs, CTX3- thetreated left lowercells. U937 quadrant cells representswere incubated remaining with 150 live nM cells. CTX3 The for right 4 h. lower On the quadrant flow cytometric represents scatter the populationgraphs, the ofleft early lower apoptotic quadrant cells. represents The right remaining upper quadrantlive cells. representsThe right lower the accumulation quadrant represents of late apoptoticthe population cells; (ofC) early Western apoptotic blot analyses cells. The of right procaspase-3 upper quadrant and poly(ADP-ribose) represents the accumulation polymerase (PARP) of late degradationapoptotic cells; in CTX3-treated (C) Western blot cells. analyses U937 cells of procaspase-3 were incubated and with poly(ADP-ribose) 150 nM CTX3 for polymerase 4 h; (D) Viability (PARP) ofdegradation CTX3-treated in CTX3-treated cells was restored cells. by U937 pretreatment cells were with incubated caspase with inhibitors. 150 nM U937CTX3 cells for 4 were h; (D pretreated) Viability withof CTX3-treated 10 µM Z-VAD-FMK cells was (pan-caspase restored by inhibitor)pretreatment or Z-DEVD-FMK with caspase inhibitors. (caspase-3 inhibitor)U937 cells for were 1 h, pretreated and then incubatedwith 10 μM with Z-VAD-FMK 150 nM CTX3 (pan-caspase for 4 h. Each inhibitor) value is or the Z-DE meanVD-FMKSD of (caspase-3 three independent inhibitor) experiments for 1 h, and ± with triplicate measurements (* p < 0.05).

Toxins 2019, 11, x FOR PEER REVIEW 3 of 12

then incubated with 150 nM CTX3 for 4 h. Each value is the mean ± SD of three independent Toxins 2019experiments, 11, 527 with triplicate measurements (* p < 0.05). 3 of 12

To examine whether CTX3-induced apoptosis is related to mitochondrial dysfunction, the To examine whether CTX3-induced apoptosis is related to mitochondrial dysfunction, mitochondrial membrane potential (ΔΨm) of CTX3-treated cells was thus measured using TMRM the mitochondrial membrane potential (∆Ψm) of CTX3-treated cells was thus measured using TMRM fluorescence. CTX3 caused a marked loss of ΔΨm in U937 cells, as demonstrated by flow cytometry ﬂuorescence. CTX3 caused a marked loss of ∆Ψm in U937 cells, as demonstrated by ﬂow cytometry analysis (Figure 2A). Mcl-1, Bcl-2, or Bcl-xL suppression has been shown to induce mitochondrial analysis (Figure2A). Mcl-1, Bcl-2, or Bcl-xL suppression has been shown to induce mitochondrial permeability and ΔΨm loss [19]. Immunoblotting analyses showed that CTX3 induced permeability and ∆Ψm loss [19]. Immunoblotting analyses showed that CTX3 induced downregultion downregultion of Mcl-1, Bcl-2, and Bcl-xL expression in U937 cells (Figure 2B). Moreover, we used of Mcl-1, Bcl-2, and Bcl-xL expression in U937 cells (Figure2B). Moreover, we used 10-N-nonyl 10-N-nonyl acridine orange (NAO), that binds to cardiolipin on the mitochondrial membrane, to acridine orange (NAO), that binds to cardiolipin on the mitochondrial membrane, to measure the measure the mitochondrial mass. Compared to the untreated control cells, the CTX3-treated cells mitochondrial mass. Compared to the untreated control cells, the CTX3-treated cells showed a showed a reduction in mitochondrial mass (Figure 2C). These results suggested that mitochondrial reduction in mitochondrial mass (Figure2C). These results suggested that mitochondrial function is function is dysregulated in CTX3-treated U937 cells. As the cytotoxicity of CTXs is reported to be dysregulated in CTX3-treated U937 cells. As the cytotoxicity of CTXs is reported to be related to an related to an increase in [Ca2+]i [12], we analyzed [Ca2+]i in U937 cells after CTX3 treatment. Compared increase in [Ca2+]i [12], we analyzed [Ca2+]i in U937 cells after CTX3 treatment. Compared to the to the untreated cells, the CTX3-treated cells showed a notable increase in [Ca2+]i (Figure 2D). untreated cells, the CTX3-treated cells showed a notable increase in [Ca2+]i (Figure2D). Pretreatment Pretreatment with the intracellular calcium chelator, BAPTA-AM, inhibited the increases of [Ca2+]i with the intracellular calcium chelator, BAPTA-AM, inhibited the increases of [Ca2+]i and the death and the death of U937 cells induced by CTX3 (Figure 2E,F). Moreover, BAPTA-AM alleviated the of U937 cells induced by CTX3 (Figure2E,F). Moreover, BAPTA-AM alleviated the CTX3-induced CTX3-induced dissipation of ΔΨm (Figure 2G). These results suggested a causal role of Ca2+ in CTX3- dissipation of ∆Ψm (Figure2G). These results suggested a causal role of Ca 2+ in CTX3-induced death induced death of U937 cells. of U937 cells.

2+ FigureFigure 2.2. CTX3-inducedCTX3-induced increaseincrease inin [Ca[Ca2+]i]i elicitedelicited mitochondrialmitochondrial depolarization.depolarization. WithoutWithout specificspecific indication,indication, U937U937 cellscells werewere treatedtreated withwith 150150 nMnM CTX3CTX3 forfor 44 h.h. ((AA)) CTX3CTX3 inducedinduced dissipationdissipation ofof ∆ΨΔΨm. TheThe lossloss of of∆Ψ ΔΨmm was was analyzed analyzed by by flow flow cytometry.; cytometry.; (B) Western(B) Western blot analysesblot analyses of Bcl-2, of Bcl-2, Bcl-xL, Bcl-xL, and Mcl-1 and inMcl-1 CTX3-treated in CTX3-treated U937 cells.; U937 (C cells.;) CTX3 (C caused) CTX3 a reductioncaused a reduction in mitochondrial in mitochondrial mass.; (D) CTX3mass.; induced (D) CTX3 an 2+ increaseinduced inan [Ca increase]i. U937 in [Ca cells2+]i. were U937 incubated cells were with incubated 150 nM CTX3 with for150 indicated nM CTX3 time for periods. indicated Results time wereperiods. shown Results as fold-increase were shown in fluorescenceas fold-increase intensity in fluorescence compared with intensity the control compared group. with Each valuethe control is the mean SD of three independent experiments with triplicate measurements. (E)Effect of BAPTA-AM on group.± Each value is the mean ± SD of three independent experiments with triplicate measurements. CTX3-induced increase in [Ca2+]i in U937 cells. U937 cells were pretreated with 10 µM BAPTA-AM for (E) Effect of BAPTA-AM on CTX3-induced increase in [Ca2+]i in U937 cells. U937 cells were pretreated 1 h, and then incubated with 150 nM CTX3 for 4 h (mean SD, * p < 0.05).; (F)Effect of BAPTA-AM on with 10 μM BAPTA-AM for 1 h, and then incubated with± 150 nM CTX3 for 4 h (mean ± SD, * p < 0.05).; the viability of CTX3-treated cells (mean SD, * p < 0.05). (G)Effect of BAPTA-AM on CTX3-induced (F) Effect of BAPTA-AM on the viability± of CTX3-treated cells (mean ± SD, * p < 0.05). (G) Effect of the loss of ∆Ψm (mean SD, * p < 0.05). BAPTA-AM on CTX3-induced± the loss of ΔΨm (mean ± SD, * p < 0.05). Previous studies have revealed that AMPK has a role in suppressing the cytotoxicity of Previous studies have revealed that AMPK has a role in suppressing the cytotoxicity of N. pallida N. pallida CTX on C2C12 myoblast cells [12]. Moreover, AMPK can be activated by calcium CTX on C2C12 myoblast cells [12]. Moreover, AMPK can be activated by calcium signaling [17]. signaling [17]. AMPKα phosphorylation at Thr172 represents AMPK activation. Thus, the AMPKα AMPKα phosphorylation at Thr172 represents AMPK activation. Thus, the AMPKα phosphorylation phosphorylation was examined. As shown in Figure3A, CTX3 treatment increased the AMPK α was examined. As shown in Figure 3A, CTX3 treatment increased the AMPKα phosphorylation. phosphorylation. Pretreatment with BAPTA-AM inhibited the CTX3-induced AMPKα phosphorylation, Pretreatment with BAPTA-AM inhibited the CTX3-induced AMPKα phosphorylation, suggesting suggesting that Ca2+ stimulated AMPK activation in the CTX3-treated cells. In addition to apoptosis (Figure 3B), AMPK-mediated signaling can elicit autophagy in cancer cells [ 17]. Thus, we analyzed the Toxins 2019, 11, x FOR PEER REVIEW 4 of 12

Toxinsthat Ca20192+, 11stimulated, 527 AMPK activation in the CTX3-treated cells. In addition to apoptosis (Figure4 of3B), 12 AMPK-mediated signaling can elicit autophagy in cancer cells [17]. Thus, we analyzed the autophagy-relatedautophagy-related proteins, proteins, LC3 LC3 and and p62. p62. Figure Figure3C 3C shows shows that that the the amount amount of LC3IIof LC3II was was increased increased in thein the CTX3-treated CTX3-treated U937 U937 cells. cells. CTX3 CTX3 treatment treatment reduced reduced the p62 the protein p62 level.protein Flow level. cytometric Flow cytometric analyses, usinganalyses, the Cyto-IDusing theTM Cyto-IDautophagyTM autophagy detection kit,detection revealed kit, an revealed increase an in increase acidic vesicular in acidic organelles vesicular inorganelles the CTX3-treated in the CTX3-treated cells, compared cells, tocompared that in the to untreatedthat in the cells untreated (Figure cells3D). Further(Figure 3D). increase Further in theincrease population in the ofpopulation autophagic of cellsautophagic was observed cells wa whens observed U937 when cells wereU937 co-treated cells were with co-treated CTX3 with and chloroquineCTX3 and chloroquine (CQ) (Figure (CQ)3D). (Figure The inhibition 3D). The of autophagyinhibition of by autophagy 3-methyladenine by 3-methyladenine (3-MA) reduced (3-MA) the productionreduced the of production LC3II and restored of LC3II p62 and expression restored inp62 the expression CTX-treated in the U937 CTX-treated cells (Figure U9373E). CQ-induced cells (Figure inhibition3E). CQ-induced of lysosomal inhibition degradation of lysosomal increased degradation the expression increased of LC3IIthe expression and p62 inof theLC3II CTX3-treated and p62 in U937the CTX3-treated cells (Figure 3U937E). These cells results(Figure corroborated 3E). These re thesults CTX3-induced corroborated the autophagy CTX3-induced in U937 autophagy cells. Thus, in 3-MAU937 andcells. CQ Thus, protected 3-MA U937and cellsCQ protected from CTX3-induced U937 cells cellfrom death CTX3-induced (Figure3F). Pretreatmentcell death (Figure with CQ3F). alleviatedPretreatment CTX3-induced with CQ alleviated apoptotic CTX3-induced death (Figure apoptotic3G). These death results (Figure indicated 3G). These that results CTX3-induced indicated autophagythat CTX3-induced promoted autophagy the apoptotic promoted death ofthe U937 apoptotic cells. death of U937 cells.

Figure 3. CTX3 induced autophagy of U937 cells. Without specific indication, U937 cells were treated Figure 3. CTX3 induced autophagy of U937 cells. Without specific indication, U937 cells were treated with 150 nM CTX3 for 4 h. On the other hand, U937 cells were pre-treated with 10 µM BAPTA-AM, 10 µM with 150 nM CTX3 for 4 h. On the other hand, U937 cells were pre-treated with 10 μM BAPTA-AM, 3-methyladenine (3-MA) or 10 µM chloroquine (CQ) for 1 h, and then incubated with 150 nM CTX3 for 10 μM 3-methyladenine (3-MA) or 10 μM chloroquine (CQ) for 1 h, and then incubated with 150 nM 4 h. (A) Western blot analyses of AMPKα and phospho-AMPKα in U937 cells; (B)Effect of BAPTA-AM CTX3 for 4 h. (A) Western blot analyses of AMPKα and phospho-AMPKα in U937 cells; (B) Effect of on AMPKα phosphorylation; (C)Effect of CTX3 on LC3II and p62 expression; (D) Flow cytometry BAPTA-AM on AMPKα phosphorylation; (C) Effect of CTX3 on LC3II and p62 expression; (D) Flow analysis of acidic vesicular organelles in cells treated with CTX3 or CQ plus CTX3. CTX3-treated cells cytometry analysis of acidic vesicular organelles in cells treated with CTX3 or CQ plus CTX3. CTX3- were then stained with a Cyto-IDTM autophagy detection kit according to manufacturer’s protocol; (E) treated cells were then stained with a Cyto-IDTM autophagy detection kit according to manufacturer’s Effect of 3-MA and CQ on CTX3-induced the formation of LC3II and p62 downregulation; (F)Effect of protocol; (E) Effect of 3-MA and CQ on CTX3-induced the formation of LC3II and p62 3-MA and CQ on the viability of CTX3-treated cells. Cell viability was determined using MTT assay downregulation; (F) Effect of 3-MA and CQ on the viability of CTX3-treated cells. Cell viability was (mean SD, * p < 0.05). (G)Effect of CQ on CTX3-induced apoptosis of U937 cells. Apoptosis was determined± using MTT assay (mean ± SD, * p < 0.05). (G) Effect of CQ on CTX3-induced apoptosis of assessed in triplicate by annexin V-PI double staining followed by flow cytometry, and percentage U937 cells. Apoptosis was assessed in triplicate by annexin V-PI double staining followed by flow apoptosis is shown as percentage of annexin V-positive cells. Data represent mean SD (* p < 0.05). cytometry, and percentage apoptosis is shown as percentage of annexin V-positive± cells. Data Priorrepresent studies mean have ± SD shown(* p < 0.05). that activated AMPK-induced autophagy is related to inhibition of the Akt/mTOR pathway [17]. Thus, we next analyzed the phosphorylation of Akt, mTOR, and S6R (mTORC1Prior substrate)studies have in CTX3-treatedshown that activated U937 cells. AMPK-induc Treatmented with autophagy CTX3 reduced is related the to phosphorylation inhibition of the levelAkt/mTOR of Akt, mTOR,pathway and [17]. S6R Thus, in U937 we cells next (Figure analyzed4A). the To verifyphosphorylation that suppression of Akt, of themTOR, Akt pathway and S6R is(mTORC1 crucial for substrate) the cytotoxicity in CTX3-treated of CTX3, U937 the e ffcells.ect of Treatment CTX3 on with CA-Akt-overexpressing CTX3 reduced the phosphorylation U937 cells was examined.level of Akt, As mTOR, shown and in Figure S6R in4B, U937 transfection cells (Figure of CA-Akt 4A). To suppressed verify that the suppression cytotoxicity of ofthe CTX3 Akt pathway in U937 cells.is crucial Consistent for the withcytotoxicity this, the of overexpression CTX3, the effect of of CA-Akt CTX3 on inhibited CA-Akt-overexpressing the CTX3-induced U937 apoptosis cells was of U937examined. cells (Figure As shown4C). Figurein Figure4D shows4B, transfection that CTX3 of was CA-Akt unable suppressed to inhibit mTORthe cytotoxicity phosphorylation of CTX3 in in U937U937 cellscells. expressingConsistent CA-Akt.with this, Moreover,the overexpressi CA-Akton overexpressionof CA-Akt inhibited mitigated the CTX3-induced the effect of CTX3apoptosis on

Toxins 2019, 11, x FOR PEER REVIEW 5 of 12

Toxins 2019, 11, 527 5 of 12 of U937 cells (Figure 4C). Figure 4D shows that CTX3 was unable to inhibit mTOR phosphorylation in U937 cells expressing CA-Akt. Moreover, CA-Akt overexpression mitigated the effect of CTX3 on LC3IILC3II productionproduction and and p62 p62 downregulation downregulation in U937 in U937 cells. cells. These These results results highlighted highlighted that CTX3-induced that CTX3- autophagyinduced autophagy and apoptosis and apoptosis were mediated were mediated throughsuppression through suppression of the Akt of/mTOR the Akt/mTOR pathway. pathway.

Figure 4. CTX3 induced phosphorylation of Akt, mTOR and S6R in U937 cells. Without specific Figure 4. CTX3 induced phosphorylation of Akt, mTOR and S6R in U937 cells. Without specific indication, U937 cells were treated with 150 nM CTX3 for 4 h. (A)Effect of CTX3 on phosphorylation of indication, U937 cells were treated with 150 nM CTX3 for 4 h. (A) Effect of CTX3 on phosphorylation Akt, mTOR, and S6R; (B)Effect of constitutively active Akt (CA-Akt) on the viability of CTX3-treated of Akt, mTOR, and S6R; (B) Effect of constitutively active Akt (CA-Akt) on the viability of CTX3- cells. U937 cells were transfected with empty expression vector or CA-Akt, respectively. After 24 h treated cells. U937 cells were transfected with empty expression vector or CA-Akt, respectively. After post-transfection, the transfected cells were treated with 150 nM CTX3 for 4 h. (Inset) Western blot 24 h post-transfection, the transfected cells were treated with 150 nM CTX3 for 4 h. (Inset) Western analyses of Akt and phospho-Akt in CA-Akt-transfected cells; (C)Effect of CA-Akt on CTX3-induced blot analyses of Akt and phospho-Akt in CA-Akt-transfected cells; (C) Effect of CA-Akt on CTX3- apoptosis of U937 cells. Data represent mean SD (* p < 0.05). (D)Effect of CA-Akt on CTX3-induced induced apoptosis of U937 cells. Data represent± mean ± SD (* p < 0.05). (D) Effect of CA-Akt on CTX3- the formation of LC3II and p62 downregulation. induced the formation of LC3II and p62 downregulation. Previous studies have shown that protein phosphatase 2A (PP2A) negatively regulate AMPK activationPrevious [20,21 studies]. Moreover, have shown downregulation that protein of the phos PP2Aphatase catalytic 2A (PP2A) subunit negatively (PP2Ac) by regulate microRNA-429 AMPK activatesactivation AMPK [20,21]. in osteoblasticMoreover, do cellswnregulation [22]. We thus of analyzed the PP2A the cataly expressiontic subunit of the (PP2Ac) PP2A catalytic by microRNA- subunit, PP2Ac429 activatesα, in the AMPK CTX3-treated in osteoblastic cells. Ascells shown [22]. We in Figure thus analyzed5A, CTX3 the induced expression PP2Ac of αthedownregulation PP2A catalytic insubunit, U937 cells.PP2Ac However,α, in the quantitativeCTX3-treated RT-PCR cells. As analyses shown showed in Figure that 5A, the CTX3 PP2Ac inducedα mRNA PP2Ac levelsα remaineddownregulation unchanged in U937 after cells. CTX3 However, treatment quantitative (Figure5B). InhibitionRT-PCR analyses of proteasome showed activity that the by MG132PP2Acα mitigatedmRNA levels the CTX3-inducedremained unchanged PP2Acα afterdownregulation CTX3 treatment (Figure (Figure5C), suggesting 5B). Inhibition that CTX3of proteasome induced PP2Acactivityα bydegradation. MG132 mitigated Co-treatment the CTX3-induced with okadaic PP2Ac acid andα downregulation CTX3 showed a (Figure further 5C), increase suggesting in AMPK thatα phosphorylationCTX3 induced PP2Ac in U937α degradation. cells compared Co-treatment to that with with CTX3 okadaic treatment acid alone and (FigureCTX3 showed5D). Conversely, a further overexpressionincrease in AMPK of PP2Acα phosphorylationα abolished CTX3-inducedin U937 cells compared AMPKα phosphorylation to that with CTX3 (Figure treatment5E). These alone results(Figure indicate 5D). thatConversely, the suppression overexpression of PP2A elicited of AMPKPP2Acα activation abolished in the CTX3-induced CTX3-treated U937AMPK cells.α Moreover,phosphorylation overexpression (Figure of5E). PP2Ac Theseα inhibitedresults indicate the cell that death the induced suppression by CTX3 of (FigurePP2A 5elicitedF), indicating AMPK theactivation link between in the theCTX3-treated PP2A-AMPK U937 axis cells. and Moreover, CTX3 cytotoxicity. overexpression Figure of5G PP2Ac showsα thatinhibited BAPTA-AM the cell pretreatmentdeath induced restored by CTX3 PP2Ac (Figureα expression 5F), indicating in the the CTX3-treated link between cells. the PP2A-AMPK These results axis indicated and CTX3 the positioningcytotoxicity. of Figure Ca2+ as 5G a majorshows regulator that BA ofPTA-AM CTX3-induced pretreatment PP2Ac restoredα degradation. PP2Acα expression in the CTX3-treated cells. These results indicated the positioning of Ca2+ as a major regulator of CTX3- induced PP2Acα degradation.

ToxinsToxins2019 2019, 11, 11, 527, x FOR PEER REVIEW 6 ofof 1212

Figure 5. CTX3-induced PP2Acα degradation promoted AMPKα phosphorylation. Without specific Figure 5. CTX3-induced PP2Acα degradation promoted AMPKα phosphorylation. Without specific indication, U937 cells were treated with 150 nM CTX3 for 4 h. On the other hand, U937 cells were indication, U937 cells were treated with 150 nM CTX3 for 4 h. On the other hand, U937 cells were pre- pre-treated with 1 µM MG132, 10 nM okadaic acid (OA) or 10 µM BAPTA-AM for 1 h, and then treated with 1 μM MG132, 10 nM okadaic acid (OA) or 10 μM BAPTA-AM for 1 h, and then incubated incubated with 150 nM CTX3 for 4 h. (A) Western blot analyses of PP2Acα expression in CTX3-treated with 150 nM CTX3 for 4 h. (A) Western blot analyses of PP2Acα expression in CTX3-treated cells; (B) cells; (B) Quantitative analyses of PP2Acα mRNA level in CTX3-treated cells; (C)Effect of MG132 on Quantitative analyses of PP2Acα mRNA level in CTX3-treated cells; (C) Effect of MG132 on PP2Acα PP2Acα expression in CTX3-treated cells; (D)Effect of OA on CTX3-induced AMPKα phosphorylation. expression in CTX3-treated cells; (D) Effect of OA on CTX3-induced AMPKα phosphorylation. (E) (E)Effect of PP2Acα overexpression on AMPKα phosphorylation in CTX3-treated cells. U937 cells were Effect of PP2Acα overexpression on AMPKα phosphorylation in CTX3-treated cells. U937 cells were transfected with empty expression vector or pcDNA-PP2Acα, respectively. After 24 h post-transfection, transfected with empty expression vector or pcDNA-PP2Acα, respectively. After 24 h post- the transfected cells were treated with 150 nM CTX3 for 4 h. (F)Effect of PP2Acα overexpression on the transfection, the transfected cells were treated with 150 nM CTX3 for 4 h. (F) Effect of PP2Acα viability of CTX3-treated cells (mean SD, * p < 0.05). (G)Effect of BAPTA-AM on PP2Acα expression overexpression on the viability of CTX3-treated± cells (mean ± SD, * p < 0.05). (G) Effect of BAPTA-AM in CTX3-treated cells. on PP2Acα expression in CTX3-treated cells. Prior studies suggest that the plasma membrane is a target of CTXs [9]. Compared to CTX3, Met-modifiedPrior studies CTX3 suggest showed that a lower the plasma activity membrane on the release is a target of calcein of CTXs from [9]. liposomes Compared [23 ].to To CTX3, evaluate Met- whethermodified CTX3 CTX3 exerted showed its a lower cytotoxicity activity through on the rele membranease of calcein damage, from the liposomes activities [23]. of nativeTo evaluate and Met-modifiedwhether CTX3 CTX3 exerted were its analyzed cytotoxicity comparatively. through memb Met-modifiedrane damage, CTX3 the retained activities an of approximate native and 50%Met- activitymodified of CTX3 onwere inducing analyzed the comparatively. leakage of EYPC Me/EYSMt-modified/cholesterol CTX3 vesiclesretained (Figurean approximate6A), whereas 50% Met-modifiedactivity of CTX3 CTX3 on did inducing not reduce the leakage the survival of EY ofPC/EYSM/cholesterol U937 cells appreciably vesicles (Figure (Figure6B). Treatment 6A), whereas with Met-modifiedMet-modified CTX3CTX3 diddid not reduce induce th AMPKe survivalα phosphorylation of U937 cells apprecia in U937bly cells (Figure (Figure 6B).6C). Treatment We further with analyzedMet-modified CTX3- CTX3 and Met-modifieddid not induce CTX3-induced AMPKα phosphorylation membrane permeabilityin U937 cells of(Figure calcein-loaded 6C). We further U937 cells.analyzed Treatment CTX3- with and CTX3Met-modified for 10 min CTX3-induced notably increased membrane the cell permeability population withof calcein-loaded a reduced calcein U937 fluorescentcells. Treatment signal with (Figure CTX36D). for Exposure 10 min notably of calcein-loaded increased U937the cell cells population to CTX3 for with 20 a min reduced or up tocalcein 4 h resultedfluorescent in calcein signal release(Figure to 6D). a similar Exposure extent of calcein-lo (Figure6D),aded showing U937 cells that to CTX3 CTX3 did for not 20 further min or increaseup to 4 h theresulted membrane in calcein permeability release to of a U937 similar cells extent after (Figure treatment 6D), for showing 20 min. Comparedthat CTX3 did to CTX3, not further Met-modified increase CTX3the membrane showed a lowerpermeability activity onof inducingU937 cells the after membrane treatment permeability for 20 min. of the Compared calcein-loaded to CTX3, U937 cells.Met- However,modified aCTX3 marked showed release a oflower intracellular activity calceinon inducing was still the noted membrane after treatment permeability of the calcein-loadedof the calcein- U937loaded cells U937 with cells. Met-modified However, a CTX3 marked for release 20 min. of Considering intracellular that calcein Met-modified was still noted CTX3 after retained treatment the capabilityof the calcein-loaded to damage the U937 membrane, cells with these Met-modified results suggested CTX3 for that 20 themin. membrane-perturbing Considering that Met-modified activity of CTX3CTX3 could retained not the exclusively capability elucidate to damage its cytotoxicity the membrane, on U937 these cells. results suggested that the membrane- perturbingAs our activity data showed of CTX3 that co AMPKuld notα exclusivelyphosphorylation elucidate is associated its cytotoxicity with theon U937 cytotoxicity cells. of CTX3, we next analyzed the effect of compound C (AMPK inhibitor) on CTX3 cytotoxicity. Treatment with 0.5, 1, or 2 µM compound C reduced AMPKα phosphorylation and AMPKα expression in U937 cells (Figure7A). U937 cells exposed to 0.5 or 1 µM compound C showed a reduction in viability by approximately 20% and 30%, respectively (Figure7B). Thus, we only studied the e ffect of 0.5 and 1 µM compound C on CTX3 cytotoxicity. Combined treatment with 0.5 µM or 1 µM compound C and CTX3 protected the U937 cells from CTX3-induced cell death (Figure7B). Co-treatment with 0.5 µM compound C inhibited CTX3-induced AMPKα phosphorylation (Figure7C). Moreover, combination with 0.5 µM compound C mitigated the CTX3-induced the reduction in the mitochondrial mass, ∆Ψm

Toxins 2019, 11, x FOR PEER REVIEW 7 of 12

Toxins 2019, 11, 527 7 of 12

loss, and apoptosis (Figure7D–F). These results highlighted that CTX3-induced AMPK activation is Toxinscrucially 2019, 11,involved x FOR PEER in itsREVIEW cytotoxicity. 7 of 12

Figure 6. Effect of Met-modified CTX3 on AMPKα phosphorylation and membrane permeability of U937 cells. Without specific indication, U937 cells were treated with 150 nM CTX3 or Met-modified CTX3 for 4 h. (A) CTX3 and Met-modified CTX3 induced membrane permeability of EYPC/EYSM/cholesterol vesicles. Each value is the mean±SD of three independent experiments with triplicate measurements; (B) Effect of Met-modified CTX3 on the viability of U937 cells. (C) Effect of Met-modified CTX3 on AMPKα phosphorylation in U937 cells; (D) CTX3 and Met-modified CTX3 induced the release of calcein from calcein-loaded U937 cells.

As our data showed that AMPKα phosphorylation is associated with the cytotoxicity of CTX3, we next analyzed the effect of compound C (AMPK inhibitor) on CTX3 cytotoxicity. Treatment with 0.5, 1, or 2 μM compound C reduced AMPKα phosphorylation and AMPKα expression in U937 cells (Figure 7A). U937 cells exposed to 0.5 or 1 μM compound C showed a reduction in viability by approximately 20% and 30%, respectively (Figure 7B). Thus, we only studied the effect of 0.5 and 1 μFigureM compoundFigure 6. Effect 6. Eff Cect of on ofMet-modified CTX3 Met-modified cytotoxicity. CTX3 CTX3 Combined on AMPKAMPKαα phosphorylationtreatment phosphorylation with 0.5 and and μ membraneM membraneor 1 μM permeability compound permeability of C and of CTX3 protected the U937 cells from CTX3-induced cell death (Figure 7B). Co-treatment with 0.5 μM U937U937 cells. cells. Without Without specific specific indication,indication, U937 U937 cells cells were were treated treated with 150 with nM 150 CTX3 nM or Met-modifiedCTX3 or Met-modified CTX3 compoundfor 4 h. C (A inhibite) CTX3 andd CTX3-induced Met-modified CTX3 AMPK inducedα phosphorylation membrane permeability (Figure of7C). EYPC Moreover,/EYSM/cholesterol combination CTX3 for 4 h. (A) CTX3 and Met-modified CTX3 induced membrane permeability of withvesicles. 0.5 μM compound Each value is C the mitigated mean SD the of CTX3-induced three independent theexperiments reduction in with the triplicate mitochondrial measurements; mass, ΔΨm EYPC/EYSM/cholesterol vesicles.± Each value is the mean±SD of three independent experiments with loss, (andB)E ffapoptosisect of Met-modified (Figure 7D–F). CTX3 These on the viabilityresults highlighted of U937 cells. that (C)E CTX3-inducedffect of Met-modified AMPK CTX3 activation on is triplicate measurements; (B) Effect of Met-modified CTX3 on the viability of U937 cells. (C) Effect of cruciallyAMPK involvedα phosphorylation in its cytotoxicity. in U937 cells; (D) CTX3 and Met-modified CTX3 induced the release of Met-modifiedcalcein from CTX3 calcein-loaded on AMPK U937α phosphorylation cells. in U937 cells; (D) CTX3 and Met-modified CTX3 induced the release of calcein from calcein-loaded U937 cells.

Figure 7. Effect of compound C on the cytotoxic effects of CTX3 on U937 cells. Without specific indication, U937 cells were pretreated with 0.5 µM compound C concentrations for 1 h, and then

incubated with 150 nM CTX3 for 4 h. (A)Effect of compound C on AMPKα phosphorylation in U937 cells. U937 cells were pretreated with indicated compound C concentrations for 4 h. (B)Effect of compound C on the viability of CTX3-treated cells (mean SD, * p < 0.05). (C)Effect of compound C on AMPKα ± phosphorylation in CTX3-treated U937 cells.; (D)Effect of compound C on CTX3-induced reduction in mitochondrial mass. (E)Effect of compound C on CTX3-induced the loss of ∆Ψm (mean SD, ± * p < 0.05). (F)Effect of compound C on CTX3-induced apoptosis of U937 cells (mean SD, * p < 0.05). ±

Toxins 2019, 11, 527 8 of 12

3. Discussion Our data reveal that CTX3 exerts its cytotoxicity on U937 cells via the Ca2+/PP2A/AMPK signaling pathway, and that the membrane-perturbing activity of CTX3 is not crucial for its capability to induce death in U937 cells. In agreement, previous studies have shown the distinct mechanism responsible for the cytotoxicity and membrane-damaging activity of CTX3 [24]. Although the internalization of CTXs into cells is suggested to be involved in their cytotoxicity [4,7,9–11], other studies suggest that N. atra CTX3 induces apoptosis of human CD8+ T lymphocytes via a membrane receptor-mediated signaling pathway [25]. Considering that Met-modified CTX3 retains its membrane-perturbing activity but does not elicit the AMPK-mediated cell death pathway, we deduced that CTX3-induced activation of death signaling in U937 cells is not exclusively dependent on the process of toxin internalization. Some studies have proposed that internalization of N. atra CTX3 and N. mossambica mossambica CTXs elicits mitochondrial fragmentation in H9C2 myoblast cells, primary cortical neurons, or SH-SY5Y neuroblastoma cells [10,11]. Other studies have reported that an increase in lysosomal membrane permeability and the release of lysosomal cathepsin B elucidates the cytotoxic effect of N. atra CTX1 on breast cancer MCF-7 and leukemia K562 cells [7]. Feofanov et al. [4] suggested that the capability of internalized N. oxiana, N, kaouthia, and N. haje CTXs to disrupt lysosome determines their cytotoxicity in human lung cancer A549 cells and leukemia HL60 cells. In contrast, Wang et al. [26] found that the internalized N. atra CTX3 induces apoptosis in fetal rat primary neurons, whereas it induces necrosis in fetal rat primary cardiomyocytes and astrocytes. Other studies showed that N. nigricollis CTX, toxin γ, notably induces mitochondria-mediated death pathway in U937 cells, whereas it shows lower cytotoxicity against human peripheral blood mononuclear cells [27]. These results suggested that (1) CTX isotoxins do not share the same intracellular targets, and/or (2) the cytotoxic targets of CTXs are cellular context- and cell-type dependent. Thus, we deduce that the Ca2+/PP2A/AMPK axis identified in the present study cannot fully elucidate the mechanisms responsible for the cytotoxicity of CTX isotoxins in different types of cancer cells or normal cells. Nevertheless, previous studies have demonstrated that AMPK activators induce mitochondrial fragmentation [14,28]. Sekar et al. [29] found that the AMPK activator, P2X7, exerts its cytotoxicity in microglia cells via induction of lysosomal instability and cathepsin B release. Altogether, the involvement of AMPK in CTX-induced mitochondrial fragmentation and lysosomal disruption is worth investigating further. Previous studies have shown that metformin (AMPK activator) repressed the cytotoxicity of N. pallida CTX on C2C12 myoblast cells [12]. On the contrary, our data reveal that CTX3-induced AMPK activation elicits autophagy-dependent apoptosis in U937 cells. Supplementary Figure S1A shows that treatment with 1, 5, or 10 mM metformin all notably increase AMPKα phosphorylation. Co-treatment with 1 mM metformin further increased AMPKα phosphorylation in CTX3-treated cells, whereas combined treatment with metformin enhanced the CTX3-induced death of U937 cells (Supplementary Figure S1B,C). Similarly, treatment with 1, 5, or 10 µM A-769662 (AMPK activator) increased AMPKα phosphorylation ( Supplementary Figure S1D), and co-treatment with 5 µM A-769662 further increased the AMPKα activation in CTX3-treated cells (Supplementary Figure S1E). Consistently, co-treatment with A-769662 enhanced the cytotoxicity of CTX3 in U937 cells (Supplementary Figure S1F). Evidently, the AMPK activator did not inhibit the CTX3-induced death of U937 cells. Thus, the hypothesis that CTX isotoxins induce cellular context- or cell-type dependent death signaling remains a subject of future investigation. Prior studies have shown that α4 plays a crucial role in stabilizing PP2Ac and protects PP2Ac from poly-ubiquitination and degradation by MID1 [30–32]. We analyzed α4 and MID1 expression in CTX3-treated U937 cells, and found that CTX3 reduced α4 expression but not MID1 expression (Supplementary Figure S2A). Pretreatment with BAPTA-AM restored PP2Acα (Figure5G) and α4 expression in CTX3-treated cells (Supplementary Figure S2B). These results suggest that the CTX3-induced elevation in [Ca2+]i distorts α4-mediated PP2Acα stabilization. Toxins 2019, 11, 527 9 of 12

4. Conclusions In conclusion, we identiﬁed that the cytotoxicity of N. atra CTX3 on U937 cells is related to the Ca2+/PP2A/AMPK axis-elicited autophagy and apoptosis. Additionally, our data suggest a dissociation of CTX3-induced death signaling from its eﬀect on cell membrane damage.

5. Materials and Methods

5.1. Chemicals and Reagents Purification of N. atra CTX3 and modification of Met residues in CTX3 were performed according to the procedure described previously [33]. Calcein, egg yolk phosphatidylcholine (EYPC), egg yolk sphingomyelin (EYSM), cholesterol, MTT, MG132, chloroquine, 3-methyladenine, metformin, A-769662, and anti-β-actin antibody were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA); and calcein-AM, Fluo-4 AM, BAPTA-AM, and tetramethylrhodamine methylester (TMRM) were from Molecular Probes (Eugene, OR, USA). Cyto-IDTM autophagy detection kit was the product of Enzo Life Sciences Inc. (Farmingdale, NY, USA). Compound C was obtained from Cayman chemical Co (Ann Arbor, MI, USA). Annexin V-FITC/propidium iodide (PI) flow cytometry assay kit and 10-n-nonyl-acridinium-orange-chloride (NAO) were purchased from Invitrogen (Carlsbad, CA, USA); and okadaic acid, Z-DEVD-FMK and Z-VAD-FMK were from Calbiochem (San Diego, CA, USA). Antibodies separately against BCL2, MCL1, and PP2Ac were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA); antibodies separately against caspase-3, α4, MID1, PARP, BCL2L1, AMPKα, p-AMPKα, LC3B, p62, mTOR, p-mTOR, S6R, p-S6R, Akt, and p-Akt were from Cell Signaling Technology (Beverly, MA, USA). HRP-conjugated rabbit anti-mouse IgG and goat anti-rabbit IgG were the products of Pierce (Rockford, IL, USA). Cell culture supplements were from GIBCO/Life Technologies Inc. (Carlsbad, CA, USA).

5.2. Cell Culture Human acute myeloid leukemia U937 cells were cultured in RPMI-1640 medium containing 10% FCS (fetal calf serum), 1% sodium pyruvate, 2 mM L-glutamine, penicillin (100 units/mL), and streptomycin (100 µg/mL) incubating at 37 ◦C in an humidiﬁed incubator with 5% CO2 atmosphere.

5.3. Measurement of Mitochondrial Depolarization CTX3-treated cells were incubated with 2 nM TMRM for 20 min following by washing with PBS. Fluorescence intensity of TMRM was analyzed using ﬂow cytometry. The cell population with the loss of ∆Ψm showed a reduction in TMRM ﬂuorescence.

5.4. Measurement of Intracellular Ca2+ Concentration ([Ca2+]i) After incubation of CTX3-treated cells with 5 µL of 2 µM Fluo-4 AM for 1 h, the ﬂuorescence of [Ca2+]i was detected using Beckman Coulter ParadigmTM Detection Platform.

5.5. Measurement of Mitochondrial Mass

CTX3-treated cells were incubated with 10 nM NAO for 20 min at 37 ◦C. The NAO fluorescence was measured using a flow cytometer, and mitochondrial mass was evaluated using NAO fluorescent intensity.

5.6. Quantitative RT-PCR Quantitative RT-PCR was performed to detect the levels of PP2Acα mRNA using GoTag qPCR Master mix (Promega, Madison, WI). Primer sequences are listed in following: PP2Acα, 5’-TCGTTGTGGTAACCAAGCTG-3’ (F), 50-AACATGTGGCTCGCCTCTAC-3’ (R); GAPDH, 5’-GAAATCCCATCACCATCTTCCAGG-3’ (F), 5’-GAGCCCCAGCCTTCTCCATG-3’ (R). Toxins 2019, 11, 527 10 of 12

5.7. Transfection of DNA The pcDNA-PP2Acα and constitutively activated, myristoylated AKT (CA-Akt) plasmids were described in our previous studies [34,35]. Transfection of U937 cells with the plasmids was conducted using 4D-Nucleofector (Lonza, Cologne AG, Germany).

5.8. Membrane Permeability Assay EYPC/EYSM/cholesterol (mol/mol/mol, 49:21:30) vesicles encapsulating calcein at the self-quenching concentration (50 mM) was prepared according to previous report [22]. The lipid vesicles dispersed in 10 mM Tris-HCl-100 mM NaCl (pH 7.5) were used for analyzing membrane-damaging activity of CTX3. CTX3-induced the release of calcein was expressed as percentage of totally dequenched calcein ﬂuorescence after the addition of 0.2% Triton X-100. The calcein ﬂuorescence was measured using excitation and emission wavelength at 490 nm and 520 nm, respectively.

5.9. Leakage of Calcein-Loaded Cells

U937 cells were incubated with 5 µM calcein-AM for 1 h at 37 ◦C following by washing with 10 mM Tris-HCl-100 mM NaCl (pH 7.5). The cells were further incubated at 37 ◦C for 20 min for the conversion of non-fluorescent calcein-AM to fluorescent calcein inside the cells. After treatment with CTX3 or Met-modified CTX3, the fluorescent signal of calcein-loaded cells was analyzed using flow cytometer.

5.10. Other Tests Cell viability analysis using MTT assay, ﬂow cytometric analysis of apoptosis using annexin V-FITC/PI kit, qualiﬁcation of acidic vesicular organelles using Cyto-IDTM autophagy detection kit, and western blot analyses were performed according to the procedure previously described [36].

5.11. Statistical Analysis Data were analyzed using GraphPad Prism (GraphPad software, version 5.01, San Diego, CA, USA). All data are expressed as mean SD. Differences among the means were assessed using one-way ± analysis of variance (ANOVA) followed by Tukey’s multiple comparison test. Difference considered significant when p < 0.05.

Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6651/11/9/527/s1, Figure S1: Effect of metformin and A-769662 on CTX3-induced cell death and AMPKα phosphorylation. Without specific indication, U937 cells were treated with 150 nM CTX3 for 4 h. (A) Effect of metformin on AMPKα phosphorylation in U937 cells. U937 cells were incubated with indicated metformin concentrations for 4 h; (B) Metformin enhanced CTX3-induced AMPKα phosphorylation. U937 cells were co-treated with 1 mM metformin and 150 nM CTX3 for 4 h; (C) Metformin enhanced CTX3-induced cell death. U937 cells were co-treated with indicated metformin and CTX3 concentrations for 4 h (mean SD, * p < 0.05); (D) Effect of A-769662 on AMPKα phosphorylation in U937 cells. U937 cells were incubated with± indicated A-769662 concentrations for 4 h; (E) A-769662 enhanced CTX3-induced AMPKα phosphorylation. U937 cells were co-incubated with 5 µM A-769662 and 150 nM CTX3 for 4 h; (F) A-769662 enhanced CTX3-induced cell death. U937 cells were incubated with indicated A-769662 and CTX3 concentrations for 4 h; Figure S2: Effect of CTX3 on α4 and MID1 expression in CTX3-treated cells. Without specific indication, U937 cells were treated with 150 nM CTX3 for 4 h. On the other hand, U937 cells were pre-treated with 10 µM BAPTA-AM for 1 h, and then incubated with 150 nM CTX3 for 4 h. (A) Western blot analyses of α4 and MID1 expression in CTX3-treated cells; (B) Effect of BAPTA-AM on α4 expression in CTX3-treated cells. Author Contributions: J.-T.C., Y.-J.S., L.-J.W., C.-H.H. and Y.-C.L. performed the experiments; J.-T.C., Y.-J.S. and L.-S.C. analyzed the data; J.-T.C. and L.-S.C. designed the experiments and wrote the paper. Funding: This research was funded by the Ministry of Science and Technology, Taiwan, grant MOST108-2320-B-110-001-MY2 (to Long-Sen Chang). Conflicts of Interest: The authors declare no conflict of interest. Toxins 2019, 11, 527 11 of 12

References

1. Dufton, M.J.; Hider, R.C. The structure and pharmacology of Elapid cytotoxins. In Snake Toxins; Harvey, A.L., Ed.; Pergamon Press: New York, NY, USA, 1999; pp. 259–302. 2. Konshina, A.G.; Dubovskii, P.V.; Efremov, R.G. Structure and dynamics of cardiotoxins. Curr. Protein Pept. Sci. 2012, 13, 570–584. [CrossRef][PubMed] 3. Ogasawara, S.; Cheng, X.W.; Inoue, A.; Hu, L.; Piao, L.; Yu, C.; Goto, H.; Xu, W.; Zhao, G.; Lei, Y.; et al. Cathepsin K activity controls cardiotoxin-induced skeletal muscle repair in mice. J. Cachexia Sarcopenia Muscle 2018, 9, 160–175. [CrossRef][PubMed] 4. Feofanov, A.V.; Sharonov, G.V.; Astapova, M.V.; Rodionov, D.I.; Utkin, Y.N.; Arseniev, A.S. Cancer cell injury by cytotoxins from cobra venom is mediated through lysosomal damage. Biochem. J. 2005, 390, 11–18. [CrossRef][PubMed] 5. Chien, C.M.; Yang, S.H.; Yang, C.C.; Chang, L.S.; Lin, S.R. Cardiotoxin III induces c-jun N-terminal kinase-dependent apoptosis in HL-60 human leukaemia cells. Cell Biochem. Funct. 2008, 26, 111–118. [CrossRef][PubMed] 6. Vyas, V.K.; Brahmbhatt, K.; Bhatt, H.; Parmar, U. Therapeutic potential of snake venom in cancer therapy: Current perspectives. Asian Pac. J. Trop. Biomed. 2013, 3, 156–162. [CrossRef] 7. Wu, M.; Ming, W.; Tang, Y.; Zhou, S.; Kong, T.; Dong, W. The anticancer effect of cytotoxin 1 from Naja atra Cantor venom is mediated by a lysosomal cell death pathway involving lysosomal membrane permeabilization and cathepsin B release. Am. J. Chin. Med. 2013, 41, 643–663. [CrossRef] 8. Chaisakul, J.; Hodgson, W.C.; Kuruppu, S.; Prasongsook, N. Effects of animal venoms and toxins on hallmarks of cancer. J. Cancer 2016, 7, 1571–1578. [CrossRef][PubMed] 9. Wu, W.G.; Tjong, S.C.; Wu, P.L.; Kuo, J.H.; Wu, K. Role of heparan sulfates and glycosphingolipids in the pore formation of basic polypeptides of cobra cardiotoxin. Adv. Exp. Med. Biol. 2010, 677, 143–149. [PubMed] 10. Wang, C.H.; Wu, W.G. Amphiphilic β-sheet cobra cardiotoxin targets mitochondria and disrupts its network. FEBS Lett. 2005, 579, 3169–3174. [CrossRef] 11. Zhang, B.; Li, F.; Chen, Z.; Shrivastava, I.H.; Gasanoff, E.S.; Dagda, R.K. Naja mossambica mossambica cobra cardiotoxin targets mitochondria to disrupt mitochondrial membrane structure and function. Toxins 2019, 11, 152. [CrossRef] 12. Langone, F.; Cannata, S.; Fuoco, C.; Lettieri Barbato, D.; Testa, S.; Nardozza, A.P.; Ciriolo, M.R.; Castagnoli, L.; Gargioli, C.; Cesareni, G. Metformin protects skeletal muscle from cardiotoxin induced degeneration. PLoS ONE 2014, 9, e114018. [CrossRef][PubMed] 13. Marin, T.L.; Gongol, B.; Zhang, F.; Martin, M.; Johnson, D.A.; Xiao, H.; Wang, Y.; Subramaniam, S.; Chien, S.; Shyy, J.Y. AMPK promotes mitochondrial biogenesis and function by phosphorylating the epigenetic factors DNMT1, RBBP7, and HAT1. Sci. Signal 2017, 10, eaaf7478. [CrossRef][PubMed] 14. Herzig, S.; Shaw, R.J. AMPK: Guardian of metabolism and mitochondrial homeostasis. Nat. Rev. Mol. Cell. Biol. 2018, 19, 121–135. [CrossRef] 15. Eichner, L.J.; Brun, S.N.; Herzig, S.; Young, N.P.; Curtis, S.D.; Shackelford, D.B.; Shokhirev, M.N.; Leblanc, M.; Vera, L.I.; Hutchins, A.; et al. Genetic analysis reveals AMPK is required to support tumor growth in murine Kras-dependent lung cancer models. Cell Metab. 2019, 29, 285–302. [CrossRef][PubMed] 16. Liang, J.; Mills, G.B. AMPK: A contextual oncogene or tumor suppressor? Cancer Res. 2013, 73, 2929–2935. [CrossRef][PubMed] 17. Mihaylova, M.M.; Shaw, R.J. The AMPK signalling pathway coordinates cell growth, autophagy and metabolism. Nat. Cell Biol. 2011, 13, 1016–1023. [CrossRef] 18. Chiou, S.H.; Raynor, R.L.; Zheng, B.; Chambars, T.C.; Kuo, J.F. Cobra venom cardiotoxin (cytotoxin) isoforms and neurotoxin: Comparative potency of protein kinase C inhibition and cancer cell cytotoxicity and model of enzyme inhibition. Biochemistry 1993, 32, 2062–2067. [CrossRef] 19. Morciano, G.; Giorgi, C.; Balestra, D.; Marchi, S.; Perrone, D.; Pinotti, M.; Pinton, P. Mcl-1 involvement in mitochondrial dynamics is associated with apoptotic cell death. Mol. Biol. Cell 2016, 27, 20–34. [CrossRef] 20. Park, S.; Scheffler, T.L.; Rossie, S.S.; Gerrard, D.E. AMPK activity is regulated by calcium-mediated protein phosphatase 2A activity. Cell Calcium 2013, 53, 217–223. [CrossRef] Toxins 2019, 11, 527 12 of 12

21. Joseph, B.K.; Liu, H.Y.; Francisco, J.; Pandya, D.; Donigan, M.; Gallo-Ebert, C.; Giordano, C.; Bata, A.; Nickels, J.T., Jr. Inhibition of AMP kinase by the protein phosphatase 2A heterotrimer, PP2APpp2r2d. J. Biol. Chem. 2015, 290, 10588–10598. [CrossRef] 22. Guo, S.; Chen, C.; Ji, F.; Mao, L.; Xie, Y. PP2A catalytic subunit silence by microRNA-429 activates AMPK and protects osteoblastic cells from dexamethasone. Biochem. Biophys. Res. Commun. 2017, 487, 660–665. [CrossRef][PubMed] 23. Kao, P.H.; Lin, S.R.; Chang, L.S. Differential binding to phospholipid bilayers modulates membrane-damaging activity of Naja naja atra cardiotoxins. Toxicon 2009, 54, 321–328. [CrossRef] 24. Chen, K.C.; Kao, P.H.; Lin, S.R.; Chang, L.S. The mechanism of cytotoxicity by Naja naja atra cardiotoxin 3 is physically distant from its membrane-damaging effect. Toxicon 2007, 50, 816–824. [CrossRef][PubMed] 25. Su, S.H.; Su, S.J.; Lin, S.R.; Chang, K.L. Cardiotoxin-III selectively enhances activation-induced apoptosis of human CD8+ T lymphocytes. Toxicol. Appl. Pharmacol. 2003, 193, 97–105. [CrossRef] 26. Wang, C.H.; Monette, R.; Lee, S.C.; Morley, P.; Wu, W.G. Cobra cardiotoxin-induced cell death in fetal rat cardiomyocytes and cortical neurons: Different pathway but similar cell surface target. Toxicon 2005, 46, 430–440. [CrossRef][PubMed] 27. Liu, W.H.; Chang, L.S. Reactive oxygen species and p38 mitogen-activated protein kinase induce apoptotic death of U937 cells in response to Naja nigricollis toxin-γ. J. Cell. Mol. Med. 2009, 13, 1695–1705. [CrossRef] 28. Toyama, E.Q.; Herzig, S.; Courchet, J.; Lewis, T.L., Jr.; Losón, O.C.; Hellberg, K.; Young, N.P.; Chen, H.; Polleux, F.; Chan, D.C.; et al. AMP-activated protein kinase mediates mitochondrial fission in response to energy stress. Science 2016, 351, 275–281. [CrossRef] 29. Sekar, P.; Huang, D.Y.; Hsieh, S.L.; Chang, S.F.; Lin, W.W. AMPK-dependent and independent actions of P2X7 in regulation of mitochondrial and lysosomal functions in microglia. Cell Commun. Signal. 2018, 16, 83. [CrossRef] 30. Trockenbacher, A.; Suckow, V.; Foerster, J.; Winter, J.; Krauss, S.; Ropers, H.H.; Schneider, R.; Schweiger, S. MID1, mutated in Opitz syndrome, encodes an ubiquitin ligase that targets phosphatase 2A for degradation. Nat. Genet. 2001, 29, 287–294. [CrossRef] 31. Kong, M.; Ditsworth, D.; Lindsten, T.; Thompson, C.B. α4 is an essential regulator of PP2A phosphatase activity. Mol. Cell 2009, 36, 51–60. [CrossRef] 32. Jiang, L.; Stanevich, V.; Satyshur, K.A.; Kong, M.; Watkins, G.R.; Wadzinski, B.E.; Sengupta, R.; Xing, Y. Structural basis of protein phosphatase 2A stable latency. Nat. Commun. 2013, 4, 1699. [CrossRef] 33. Shi, Y.J.; Huang, C.H.; Lee, Y.C.; Wang, L.J.; Chiou, J.T.; Chang, L.S. Naja atra cardiotoxins enhance the protease activity of chymotrypsin. Int. J. Biol. Macromol. 2019, 136, 512–520. [CrossRef] 34. Liu, W.H.; Chou, W.M.; Chang, L.S. p38 MAPK/PP2Acα/TTP pathway on the connection of TNF-α and caspases activation on hydroquinone-induced apoptosis. Carcinogenesis 2013, 34, 818–827. [CrossRef] 35. Lee, Y.C.; Chen, Y.J.; Huang, C.H.; Chang, L.S. Amsacrine-induced apoptosis of human leukemia U937 cells is mediated by the inhibition of AKT- and ERK-induced stabilization of MCL1. Apoptosis 2017, 22, 406–420. [CrossRef] 36. Lee, Y.C.; Wang, L.J.; Huang, C.H.; Shi, Y.J.; Chang, L.S. ABT-263-induced MCL1 upregulation depends on autophagy-mediated 4EBP1 downregulation in human leukemia cells. Cancer Lett. 2018, 432, 191–204. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).