A Synthetic Cell-Penetrating Dominant-Negative ATF5 Peptide Exerts Anticancer Activity Against a Broad Spectrum of Treatment-Res

Published OnlineFirst April 28, 2016; DOI: 10.1158/1078-0432.CCR-15-2827

Cancer Therapy: Preclinical Clinical Cancer Research A Synthetic Cell-Penetrating Dominant-Negative ATF5 Peptide Exerts Anticancer Activity against a Broad Spectrum of Treatment-Resistant Cancers Georg Karpel-Massler1, Basil A. Horst2, Chang Shu1, Lily Chau1, Takashi Tsujiuchi3, Jeffrey N. Bruce3, Peter Canoll1, Lloyd A. Greene1, James M. Angelastro4, and Markus D. Siegelin1

Abstract

Purpose: Despite significant progress in cancer research, many led to diminished levels of antiapoptotic Bcl-2 family members tumor entities still have an unfavorable prognosis. Activating Mcl-1 and Bcl-2. In line with this, CP-d/n-ATF5-S1 synergisti- transcription factor 5 (ATF5) is upregulated in various malignan- cally enhanced tumor cell apoptosis induced by the BH3- cies and promotes apoptotic resistance. We evaluated the efficacy mimetic ABT263 and the death ligand TRAIL. In vivo, CP-d/ and mechanisms of the first described synthetic cell-penetrating n-ATF5-S1 attenuated tumor growth as a single compound in inhibitor of ATF5 function, CP-d/n-ATF5-S1. glioblastoma, melanoma, prostate cancer, and triple receptor– Experimental Design: Preclinical drug testing was performed negative breast cancer xenograft models. Finally, the combina- in various treatment-resistant cancer cells and in vivo xenograft tion treatment of CP-d/n-ATF5-S1 and ABT263 significantly models. reduced tumor growth in vivo more efficiently than each reagent Results: CP-d/n-ATF5-S1 reduced the transcript levels of on its own. several known direct ATF5 targets. It depleted endogenous Conclusions: Our data support the idea that CP-d/n-ATF5-S1, ATF5 and induced apoptosis across a broad panel of treat- administered as a single reagent or in combination with other ment-refractory cancer cell lines, sparing non-neoplastic cells. drugs, holds promise as an innovative, safe, and efficient anti- CP-d/n-ATF5-S1 promoted tumor cell apoptotic susceptibility neoplastic agent against treatment-resistant cancers. Clin Cancer Res; in part by reducing expression of the deubiquitinase Usp9X and 1–14. 2016 AACR.

Introduction ATF5 is an example that has been identified as a potential target for cancer treatment but for which no specific therapy has Although recent significant responses for cancer therapy have been developed (4, 5). Activating transcription factor 5 (ATF5; been achieved in some malignancies, these are often initially also termed ATFx) is a member of the activating transcription impressive, but unfortunately not durable (1, 2). For other malig- factor/cyclic AMP–responsive element-binding (ATF/CREB) nancies, such as high-grade primary brain cancers, the prognosis is family. A common feature of this family is the presence of a unfavorable even with treatment (3). Therefore, efforts continue basic leucine zipper (bZIP) domain that promotes DNA bind- to identify new potential targets for tumor-specific treatments as ing via the basic region and interactions with other proteins well as novel therapeutic strategies to exploit these targets. through the leucine zipper (5–7). ATF5 binds to several different promoter elements including the nutrient-sensing report 1Department of Pathology & Cell Biology, Columbia University Medical element (NRSE) and a novel motif to regulate gene transcrip- 2 Center, New York, New York. Department of Dermatology, Columbia tion (6, 8). Full-length ATF5 appears to be rapidly degraded via University Medical Center, New York, New York. 3Department of Neu- rosurgery, Columbia University Medical Center, New York, New York. the proteasome and it has the unusual property that it is among 4Department of Molecular Biosciences, University of California, Davis a small group of proteins that are selectively translated when School of Veterinary Medicine, Davis, California. eiF2a is phosphorylated (9, 10). Note: Supplementary data for this article are available at Clinical Cancer ATF5 protein levels are increased in a variety of human Research Online (http://clincancerres.aacrjournals.org/). malignancies, including glioblastoma, breast, pancreatic, lung, Current address for L. Chau: Department of Neurology, SUNY Downstate and colon cancers (11). In contrast, with few exceptions (liver, Medical Center, Brooklyn, New York, New York. prostate, and testis), ATF5 expression is low in normal tissue of Corresponding Authors: Markus D. Siegelin, Department of Pathology & Cell the respective organs. In several tumor types, including glio- Biology, Columbia University Medical Center, 630 West 168th Street, P&S 15-401, blastoma and non–small cell lung cancer, ATF5 expression New York, NY 10032. Phone: 212-305-1993; E-mail: [email protected] negatively correlates with survival (12, 13). In cell culture ; and [email protected] and James M. Angelastro, Department of Molecular studies, ATF5 promotes survival by counteracting apoptosis in Biosciences, University of California One Shields Avenue, Davis, CA 95616. pro-B lymphocytes deprived of IL3 or in HeLa cells after growth Phone: 530-752-1591; E-mail: [email protected] factor withdrawal (14). In addition, ATF5 regulates transcrip- doi: 10.1158/1078-0432.CCR-15-2827 tion of antiapoptotic B-cell leukemia 2 (Bcl-2) and of Bcl-2 2016 American Association for Cancer Research. family member, myeloid cell leukemia-1 (Mcl-1), presumably

www.aacrjournals.org OF1

Karpel-Massler et al.

Reagents Translational Relevance CP-d/n-ATF5-S1, mutated CP-d/n-ATF5-S1, and Penetratin Targeting treatment-resistant malignancies remains a major were purchased from CS Bio. Recombinant TRAIL was from challenge in oncology. In this study, we introduce a novel Peprotech. ABT263 was from Selleckchem. therapeutic compound targeting activating transcription factor 5 (ATF5) through utilization of a novel cell-penetrating pep- Cell culture tide, termed CP-d/n-ATF5-S1. CP-d/n-ATF5-S1 displays broad Cells were grown as described previously (20, 21). Cells were anticancer activity against glioblastoma, triple receptor–neg- obtained from the ATCC or Cell Line Services and authenticated ative breast cancer, prostatic carcinoma, pancreatic cancer, by the manufacturer. No cell line authentication was performed melanoma, non–small cell lung carcinoma, and hematologic by the authors and details are found in the Supplementary malignancies. Notably, we provide evidence that these anti- Section. neoplastic effects are not only observed in vitro, but are also seen in six different animal models of glioblastoma, melano- Cell viability assays ma, prostatic adenocarcinoma, triple receptor–negative breast To examine cellular proliferation, MTT assays were performed cancer, and pancreatic carcinoma without any detectable as described previously (21). toxicity. Finally, CP-d/n-ATF5-S1 sensitizes tumor cells for BH3-mimetics and extrinsic apoptotic stimuli in vitro and in Measurement of apoptosis and mitochondrial membrane ﬁ vivo. Taken together, CP-d/n-ATF5-S1 is a novel highly ef ca- potential cious anticancer compound with minimal toxicity and poten- Annexin V/propidium iodide (PI), PI, and JC-1 stainings were tially warrants clinical testing in patients. performed as described previously (20, 22).

Western blot analysis Protein expression was determined by Western blot analysis as thereby promoting tumor cell survival (13, 15). Conversely, described previously (23). interference with ATF5 expression or activity yields a marked induction of apoptosis in glioblastoma cells in vitro and in vivo Transfections of siRNAs without affecting astrocytes (16, 17). Moreover, in a transgenic siRNAs were transfected as described previously (22, 24). murine model in which endogenous glioblastomas were inducedbyaPDGF/sh-p53–expressing virus, activation of a cDNA synthesis and RT-PCR dominant/negative (d/n)-ATF5 blocked tumor formation and cDNA synthesis and RT-PCR were performed as described resulted in regression of formed tumors (17). Antineoplastic previously (23). activity of d/n-ATF5 was also reported for breast cancer cells and pancreatic cancer cells in vitro (11, 15, 18). These findings Subcutaneous xenograft models thus suggest ATF5 as a promising target for a tailored anticancer Subcutaneous xenografts were implanted as described previ- therapy. ously (20). To provide a potential means to target ATF5 in vivo, we designed a d/n-ATF5 linked to a cell-penetrating domain (Penetratin; Statistical analysis ref. 19). This recombinant peptide passes the blood–brain barrier, Statistical significance was assessed by Student t test using Prism enters tumor cells, and exerts antineoplastic activity in a rodent version 5.04 (GraphPad). P 0.05 was considered statistically transgenic glioma model. In this study, we assessed the activity significant. and mechanism of action of a similar peptide (CP-d/n-ATF5-S1) that was further modified to reduce its size and that was generated Results synthetically. In in vitro studies and in in vivo murine xenograft models, CP-d/n-ATF5-S1 shows apoptosis induction over a broad CP-d/n-ATF5-S1 range of recalcitrant human malignancies without apparent CP-d/n-ATF5-S1 is a synthetic 67-amino acid peptide that was fi effects on nontransformed cells. A novel mechanism of action engineered to cross cellular membranes and to speci cally inter- was found in which the peptide reduces expression of the deu- fere with the survival-promoting actions of ATF5 (Fig. 1A). The biquitinating enzyme Usp9X, which in turn leads to depletion of N-terminal end has a 16-amino acid Penetratin domain that Mcl-1 and Bcl-2 and to consequent apoptotic death. The latter facilitates cellular penetration (25, 26). A dominant/negative findings led us to rationally design and carry out in vitro and in vivo sequence follows in which the DNA-binding domain of ATF5 is tests of several potential combination therapies with CP-d/n- substituted by an amphipathic sequence with a leucine repeat at ATF5-S1 that had enhanced efficacy compared with either agent every seventh residue and then by the human ATF5 bZIP domain fi – alone. truncated after the rst valine (26 29). Parallel work has dem- onstrated that a similar recombinant tagged peptide passes the – Materials and Methods blood brain barrier, enters intact cells both in vivo and in vitro, and promotes selective death of glioma cells (19). Four independent Ethics statement batches of the peptide (including one under GMP conditions) All procedures were in accordance with Animal Welfare Reg- have had comparable activity. For control purposes, peptides were ulations and approved by the Institutional Animal Care and Use also synthesized with a penetratin domain alone and in which key Committee at Columbia University Medical Center (New York, leucine residues were mutated to glycine in the d/n portion to NY). reduce binding to potential partners (Fig. 1A).

OF2 Clin Cancer Res; 2016 Clinical Cancer Research

Peptide-Based Inhibition of ATF5 for Cancer Therapy

A CP-d/n-ATF5-S1

H2N-RQIKIWFQNRRMKWKK LEQRAEELARENEELLEKEAEELEQENAE LEGECQGLEARNRELKERAESV-COOH ATF5 Leucine zipper Penetratin Dominant/negative-sequence truncated after first valine Mutated CP-d/n-ATF5-S1

L G L G LL GG L G L G L G L G

H2N-RQIKIWFQNRRMKWKK GEQRAEEGARENEEGGEKEAEEGEQENAE GEGECQGGEARNREGKERAESV-COOH ATF5 Leucine zipper Penetratin Dominant/negative-sequence truncated after first valine

Penetratin

H2N-RQIKIWFQNRRMKWKK-COOH

B C D T98G T98G T98G MDA-MB-436 0 h 45 min 72 h 72 h 100 Cycloheximide + + + + CP-d/n-ATF5-S1 – – CP-d/n-ATF5-S1 – + – + ATF5 ATF5 50

Actin Actin Relative protein expression (%) 0 0 h 45’15’ h 3 h1.5 Cycloheximide + +++ + +++ ++ CP-d/n-ATF5-S1 –+++–+–+–– E

T98G T98G T98G 200 # 200 250 24 h 24 h 24 h # 150 72 h 150 72 h 200 72 h # 150 100 100 * 100 * * 50 50 50 (% of control) (% of Mcl-1 mRNA Bcl-2 mRNA (% of control) (% of 0 0 mRNA Usp9X control) (% of 0 pen 50 200100 pen 50 200100 pen 50 200100 CP-d/n-ATF5-S1 (μmol/L) CP-d/n-ATF5-S1 (μmol/L) CP-d/n-ATF5-S1 (μmol/L)

F G H Penetratin 200 μmol/L CP-d/n/ATF5-S1 50 μmol/L All glioblastoma samples n = 99 T98G n U87MG ATF5-amplified (≥ 2.2 copies) = 20 P < 0.001 ATF5-deleted (≤ 1.8 copies) n = 19 150 Log-rank P values: P < 0.001 Amplified vs. deleted: 0.043 Amplified vs. all: 0.266 100 100 μm Deleted vs. all: 0.012 50

Cellular viability (%) 0 – 20010050 CP-d/n-ATF5-S1 (μmol/L) CP-d/n/ATF5-S1 100 μmol/L CP-d/n/ATF5-S1 200 μmol/L

Figure 1. A, graphical representation showing the sequences of CP-d/n-ATF5-S1, mutated CP-d/n-ATF5-S1, and penetratin. B, T98G glioblastoma and MDA-MB-436 breast cancer cells were treated for 72 hours with increasing concentrations of CP-d/n-ATF5-S1 under reduced serum conditions to mimic the nutrient-deprived state of tumor cells in the tumor tissue (1.5% FBS) followed by Western blot analysis for ATF5. Actin Western blot analysis was performed to conﬁrm equal protein loading. Arrow indicates a speciﬁc band of ATF5. C, T98G glioblastoma cells were treated with CP-d/n-ATF5-S1 or solvent for 48 hours before adding 10 mg/mL cycloheximide and Western blot analysis for ATF5 and actin. D, graphical representation following densitometric analysis of the experiment described under C using ImageJ (NIH, Bethesda, MD; http://imagej.nih.gov/ij). (Continued on the following page.)

www.aacrjournals.org Clin Cancer Res; 2016 OF3

Karpel-Massler et al.

CP-d/n-ATF5-S1 depletes endogenous ATF5 Annexin V–positive cells, thus indicating an apoptotic response Western blot analyses revealed that treatment of cultured across a wide and diverse panel of solid and nonsolid cancer cells. tumor cell lines (T98G, MDA-MB-436, and GBM12) with CP- Moreover, this effect was attenuated by treatment with the pan- d/n-ATF5-S1 leads to a dose-dependent reduction of endoge- caspase inhibitor z-VAD-Fmk in T98G cells (Supplementary nous ATF5 protein levels by 3 days (Fig. 1B and Supplementary Fig. S6), which is consistent with studies indicating that interfer- Fig. S1B). In T98G cells, this effect was present after 48 hours, but ence with ATF5 function or expression in tumor cells promotes not 24 hours (Supplementary Fig. S1A). RNAseq analysis of apoptosis (9, 12). T98G glioblastoma cells treated with CP-d/n-ATF5-S1 for 1 to To verify whether the effect of the peptide on survival is 3 days showed no significant alteration of ATF5 transcript levels specifically related to the dominant-negative domain and not to (data not shown). Comparison of endogenous ATF5 levels in the cell-penetrating domain, we treated T98G cells either with T98G cells treated with or without CP-d/n-ATF5-S1 in presence Penetratin alone or CP-d/n-ATF5-S1. In contrast to CP-d/n-ATF5- of cycloheximide indicates that the peptide significantly S1, Penetratin did not markedly increase the fraction of Annexin decreases ATF5 protein stability (Fig. 1C and D and Supplemen- V–positive cells (Fig. 2D and E). Moreover, neither Penetratin nor tary Fig. S1C). In contrast, treatment with Penetratin peptide did CP-d/n-ATF5-S1 resulted in a significant induction of apoptosis in not affect ATF5 stability (Supplementary Fig. S1D and S1E). human fetal astrocyte cultures (Fig. 2F and G), suggesting that CP- Thus, one action of CP-d/n-ATF5-S1 is loss of endogenous ATF5 d/n-ATF5-S1 possesses specificity toward cancer cells. caused at least in part by enhanced turnover. CP-d/n-ATF5-S1 leads to dissipation of mitochondrial CP-d/n-ATF5-S1 interferes with transcription of known ATF5 membrane potential and activates caspase-9 downstream targets Next, we addressed whether activation of apoptosis by CP-d/n- CP-d/n-ATF5-S1 treatment for 24 hours resulted in downregu- ATF5-S1 is mediated at least in part through a mitochondrial lation of three known ATF5 target genes [Mcl-1 (13), Bcl-2 (15, pathway. JC1 staining revealed that peptide treatment leads to 30), and asparagine synthetase (8)] at the mRNA level (Fig. 1E and a marked reduction of mitochondrial membrane potential Supplementary Fig. S2). However, for Mcl-1 and Bcl-2, this (Fig. 2H) and cleavage (activation) of caspase-9 (Fig. 2I) suggest- decrease was transitory with mRNA levels returning to baseline ing that CP-d/n-ATF5-S1 induces apoptosis in part through the by 72 hours, presumably by compensatory mechanisms. In con- mitochondrially driven apoptotic pathway. trast, CP-d/n-ATF5-S1 did not decrease mRNA levels of Usp9X, a gene not described as transcriptionally regulated by ATF5 CP-d/n-ATF5-S1 downregulates antiapoptotic Bcl-2 and Mcl-1 (Fig. 1E). proteins Because our findings pointed toward involvement of mito- CP-d/n-ATF5-S1 promotes apoptotic cell death across a wide chondria in apoptosis driven by CP-d/n-ATF5-S1, we next focused panel of treatment-resistant human cancer cell lines on expression of Bcl-2 family proteins. ATF5 is reported to ATF5 is expressed in a variety of human cancers including regulate transcription of antiapoptotic Bcl-2 and Mcl-1 (13, glioblastoma (11, 15, 18). In silico analysis of the Rembrandt 30). However, at least in T98G cells, there was a rebound of dataset for glioblastoma shows a significantly worse overall Mcl-1 and Bcl-2 mRNA expression by 3 days of peptide treatment survival in patients harboring an amplification of the ATF5 gene (Fig. 1E), so it was important to assess protein levels at this time as compared with those with 1.8 copies (Fig. 1F). Several studies well. As shown in Fig. 3A and Supplementary Fig. S7A, Mcl-1 was have also reported an inverse relationship between ATF5 protein consistently downregulated in all lines tested (U87MG, T98G expression and glioblastoma patient survival (5). glioblastoma, and H1975 non–small cell lung cancer, PANC-1 Inhibition of ATF5 function or expression has marked anti- pancreatic carcinoma, A375 melanoma, and PC3 prostate cancer) neoplastic effects in vitro and in vivo (11, 15, 16, 18). Initially, to at 72 hours of peptide treatment and in some cases, by 48 hours. assess the activity of CP-d/n-ATF5-S1, T98G, U87MG glioblasto- Bcl-2 protein was similarly downregulated in all but the PC3 cell ma, and HL-60 myeloid leukemia cells were treated for 72 hours line. Expression of Bcl-xL, a third member of the antiapoptotic Bcl- with increasing concentrations of the peptide. CP-d/n-ATF5-S1 2 family, was altered in some lines (U87MG and T98G), but not in yielded a dose-dependent antiproliferative effect as determined others. Despite a rebound of Mcl-1 and Bcl-2 mRNA levels in by MTT assay (Fig. 1G and Supplementary Fig. S3) as well as T98G cells by 3 days of CP-d/n-ATF5-S1 treatment, expression of marked changes in cellular morphology observed by light micros- the corresponding proteins was still decreased in these cells after copy (Fig. 1H). To further assess the mechanism of such effects, we 6 days of peptide treatment (Supplementary Fig. S1F). performed Annexin V/PI staining across a variety of therapy- refractory human cancer cell lines after treatment with increasing CP-d/n-ATF5-S1 downregulates Bag3 and Usp9X proteins concentrations of CP-d/n-ATF5-S1 for 48 hours. As shown in Fig. The observed decreases in Mcl-1 and Bcl-2 proteins at 72 hours 2A–C, Supplementary Figs. S4A, S4C, and S5A, the peptide promoted by CP-d/n-ATF5-S1 under conditions in which mRNA yielded a strong and dose-dependent increase in the fraction of levels appear to be unaffected led us to next assess whether the

(Continued.) E, T98G glioblastoma cells were treated for the indicated durations with 100 mmol/L Penetratin (pen) or increasing concentrations of CP-d/n-ATF5-S1 prior to performing qRT-PCR for Mcl-1, Bcl-2, and Usp9X. Columns, mean; error bars, SD. , P < 0.05 versus treatment with Penetratin for 24 hours; #, P < 0.05 vs treatment with pen for 72 hours. F, in silico analysis on the survival of glioblastoma patients based on the ampliﬁcation status of the ATF5 gene (National Cancer Institute, 2005; REMBRANDT home page, http://rembrandt.nci.nih.gov). G, T98G glioblastoma cells were treated with increasing concentrations of CP-d/n-ATF5-S1 under low serum conditions (1.5% FBS). After 72 hours, a MTT assay was performed. Data presented are representative for at least two independent experiments. Columns, means; error bars, SEM. H, representative microphotographs of U87MG cells at 40 magniﬁcation after 72 hours of treatment with CP-d/n-ATF5-S1. Morphologic changes such as decreased lengths and numbers of cell processes are especially seen in those cells treated with 200 mmol/L CP-d/n-ATF5-S1. Black arrow tip points at blebs.

OF4 Clin Cancer Res; 2016 Clinical Cancer Research

Peptide-Based Inhibition of ATF5 for Cancer Therapy

A 150 MDA- U87MGT98G U251 LN229 MGPP-3 SF188 PC3 DU145 PANC-1 H1975 A375 HL-60 SU-DHL4 K562 Raji MB-436 100

Cellular viability (%) Cellular viability 0 CP-d/n-ATF5-S1 50 μmol/L - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - CP-d/n-ATF5-S1 100 μmol/L - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - + - - - +- - - + - - - + - - - + - - - + - - - + - CP-d/n-ATF5-S1 200 μmol/L ------+ - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + B C

CP-d/n-ATF5-S1 (μmol/L) - 50 100 200 CP-d/n-ATF5-S1 (μmol/L) - 50 100 200

6.36 0.7521.62.573.322.737.58 21.9 13.41.72 12.91.08 45.70 82.00.075 U87MG HL-60

81.6 42.228.647.32.7391.24.42 35.1 3.5581.3 2.1283.9 6.6347.6 6.2911.6 8.12 0.2123.05.337.1512.210.0 4.73 11.36.52 6.78 9.4937.5 20.9 61.110.4 T98G SU-DHL4 78.8 3.0041.829.94.0376.63.06 92.1 1.0481.1 0.6483.1 0.3041.3 1.6526.9 6.36 4.7616.010.39.815.713.05 18.5 0.61 18.951.515.924.02.687.96 68.9 U251 K562 86.7 51.417.955.820.164.43.92 25.3 88.4 10.66.4826.28.1665.13.05 1.63 2.45 4.2611.510.44.112.944.54 22.5 6.52 66.011.145.110.619.613.2 18.6 LN229 Raji

90.2 59.28.4669.62.0390.92.82 14.0 78.2 15.40.1143.70.5769.22.10 0.039 4.53 1.462.133.780.608.190.57 10.4 MGPP-3 D E 94.1 54.810.184.02.2788.90.84 33.3 T98G T98G 1.80 0.6351.82.3314.61.5311.6 17.0 Penetratin CP-d/n-ATF5-S1 100 100 μmol/L PC3 100 μmol/L 3.59 7.07 5.01 33.6 50 79.3 14.223.122.829.054.87.37 68.1 Cellular viability (%) viability 2.05 1.1267.81.1161.51.396.52 45.8 Penetratin +– 84.5 4.81 24.2 37.2 CP-d/n-ATF5-S1 –+ DU145 F hu astrocytes G hu astrocytes 87.4 25.517.114.017.519.64.06 27.7 Penetratin CP-d/n-ATF5-S1 4.06 0.3950.31.7627.36.859.41 21.1 200 μmol/L 100 μmol/L 100 5.81 9.46 3.24 11.5 PANC-1 50 Cellular 84.0 15.627.021.016.649.22.56 62.9 (%) viability 10.3 Penetratin +– Propidium iodide 79.5 5.23 75.0 6.32 9.0935.116.214.010.39.61 51.3 CP-d/n-ATF5-S1 –+ MDA- Annexin V MB-436 82.9 34.42.6146.13.3872.31.20 5.19 H I 2.10 3.9621.61.7811.22.497.41 44.4 T98G T98G H1975 CP-d/n- +– CP-d/n-ATF5-S1 CP-d/n-ATF5-S1 ATF5-S1 85.3 28.125.351.36.1180.25.15 23.5 Control 50 μmol/L 100 μmol/L 0.65 84.0 1.11 46.2 0.19 40.8 CP9 1.06 6.6221.51.969.680.643.56 37.9 CF A375 fluorescence JC-1 red 0.22 15.1 0.48 52.2 0.16 58.9 Actin 90.8 39.312.763.817.172.64.56 16.1

Propidium iodide JC-1 green VAnnexin fluorescence

Figure 2. A, U87MG, T98G, U251, LN229, MGPP-3 (transge