Induction of Paraptotic Cell Death in Cancer Cells by Triptycene–Peptide Hybrids and the Revised Mechanism of Paraptosis II

In previous work, we reported on iridium(III) (Ir(III)) complex-peptide hybrids as amphiphilic conjugates (IPH-ACs) and triptycene-peptide hybrids as amphiphilic conjugates (TPH-ACs) and found that these hybrid compounds containing three cationic KK(K)GG peptide units through C6–C8 alkyl linkers induce paraptosis II, which is one of the nonapoptotic programmed cell death (PCD) types in Jurkat cells and different from previously reported paraptosis. The details of that study revealed that the paraptosis II induced by IPH-ACs (and TPH-ACs) proceeds via a membrane fusion or tethering of the endoplasmic reticulum (ER) and mitochondria, and Ca2+ transfer from the ER to mitochondria, which results in a loss of mitochondrial membrane potential (ΔΨm) in Jurkat cells. However, the detailed mechanistic studies of paraptosis II have been conducted only in Jurkat cells. In the present work, we decided to conduct mechanistic studies of paraptosis II in HeLa-S3 and A549 cells as well as in Jurkat cells to study the general mechanism of paraptosis II. Simultaneously, we designed and synthesized new TPH-ACs functionalized with peptides that contain cyclohexylalanine, which had been reported to enhance the localization of peptides to mitochondria. We found that TPH-ACs containing cyclohexylalanine promote paraptosis II processes in Jurkat, HeLa-S3 and A549 cells. The results of the experiments using fluorescence Ca2+ probes in mitochondria and cytosol, fluorescence staining agents of mitochondria and the ER, and inhibitors of paraptosis II suggest that TPH-ACs induce Ca2+ increase in mitochondria and the membrane fusion between the ER and mitochondria almost simultaneously, suggesting that our previous hypothesis on the mechanism of paraptosis II should be revised.


■ INTRODUCTION
−8 Although many types of anticancer drugs are used in hospitals, there is a need for the development of anticancer drugs with various mechanisms of action to avoid drug resistance against specific anticancer agents.−11 Later, it was reported that paraptosis is induced in breast cancer cell lines (MDA-MB-435S and MCF-7) and colorectal cancer cell lines (DLD-1 and RKO) by celastrol, 12,13 in MDA-MB-231 and in MDA-MB-435S by curcumin derivatives, 14−16 and in ovarian cancer cells (A2780, SK-OV-3 and HO-8910 cells) by morucin. 17owever, many aspects of paraptosis are poorly understood and a deep understanding of this process could lead to new strategies for the treatment of conditions such as cancers, autoimmune diseases, and viral infections.
The cytotoxicity of 1a was found to be more potent than that of either 1b or 1c, suggesting that the cytotoxicity of 1a− 1c is dependent on the number of cationic peptide units and their net cationic charge. 24These results suggest that IPH-ACs and TPH-AC may represent new types of anticancer drugs and that unraveling the induction mechanism of paraptosis would contribute to the development of novel anticancer agents.
A more detailed mechanistic study of paraptosis induced by 1−3 in Jurkat cells was carried out 22−26 and compared with that induced by celastrol (Scheme 2), which had been reported as a naturally occurring triterpenoid and a paraptosis inducer. 12,13Due to the observation of cell death induced by celastrol and IPH-ACs (TPH-ACs), cell death induced by celastrol was classified as paraptosis I, which negligibly involves membrane fusion between the ER and mitochondria, and cell death induced by IPH-ACs and TPH-ACs was referred to as paraptosis II, which is associated with membrane fusion between the ER and mitochondria. 25,26In addition, we found that TPH-ACs exhibit potent cytotoxicity against various cell lines like HeLa-S3 (human cervix carcinoma) and A549 (human Caucasian lung carcinoma) cells, but the cytotoxicity of IPH-ACs against HeLa-S3 and A549 cells is weak.
In this work, we decided to conduct mechanistic studies of the general mechanism of paraptosis II in HeLa-S3 and A549 cells as well as that in Jurkat cells.In addition, we designed and synthesized novel TPH-ACs 4−7 that possess linear or cyclic peptides containing cyclohexylalanine (Cha) in addition to K or arginine (R), because the localization of IPH-ACs in mitochondria had been observed in our previous works 25 and it had been reported that Cha-containing peptides exhibit high localization in mitochondria. 27−30 The results of the experiments described in this manuscript using fluorescence Ca 2+ probes in mitochondria and cytosol (Rhod-2/AM and Rhod-4/AM, respectively), fluorescence probes of mitochondria and the ER (MitoTracker Green and ERTracker Red, respectively) suggest that the paraptosis II induced by TPH-ACs in Jurkat, HeLa-S3, and A549 cells proceeds via the Ca 2+ transfer possibly from the ER to mitochondria and fusion (or tethering) between the ER and mitochondria almost simultaneously.In addition, the experiments using some inhibitors of Ca 2+ transport system from the ER and mitochondria such as carbonyl cyanide 3-chlorophenylhydrazone (CCCP), which is an uncoupling reagent and an inhibitor of mitochondrial Ca 2+ uptake and had been found to inhibit paraptosis II, 2-aminophenyl borate (2-APB), an antagonist of InsP 3 (inositol 1,4,5-trisphosphate) receptor, and 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid (DIDS), which is an inhibitor of the voltage-dependent anion channel (VDAC), were carried out.As a result, we would like to add some revisions to our previous hypothesis in this manuscript.
Moreover, the cytotoxicity of TPH-ACs against IMR90 cells, a model of normal cells, is reported to be weaker than that against the aforementioned three cancer cell lines.This result is attributed to a slower Ca 2+ increase in mitochondria and a weaker interaction between the ER and mitochondria than those in cancer cell lines even after treatment with 4, as proven via costaining experiments using ERTracker Red and MitoTracker Green.The results are reported herein.
MTT Assays.Jurkat, HeLa-S3, A549 cells (2.0 × 10 5 cells/ mL) and IMR90 cells (1.0 × 10 5 cells/mL) were incubated in the presence of 3−7 for 1 h, and etoposide, cisplatin and celastrol for 24 h (50 μL) in cell culture medium under CO 2 at 37 °C in 96-well plates (BD Falcon).MTT reagent (5 mg/ mL) in PBS (10 μL) was then added to the cells.After incubation at 37 °C for 4 h, a formazan lysis solution (10% sodium dodecyl sulfate (SDS) in 0.01 N HCl) (100 μL) was added and the resultant solution was incubated overnight under the same conditions, which was followed by measurement of the absorbance at 570 nm using a microplate reader (ARVO, PerkinElmer).
MTT Assays of Jurkat Cells with 4 in the Presence of Zn(NO 3 ) 2 .Jurkat cells (2.0 × 10 5 cells) were incubated in RPMI 1640 medium containing solution of 4 (final concentration = 0.16−10 μM) with Zn(NO 3 ) 2 (final concentration = 15 μM) for 1 h.MTT reagent (5 mg/mL) in PBS (10 μL) was then added to the cells.After incubation at 37 °C for 4 h, a formazan lysis solution (10% sodium dodecyl sulfate (SDS) in 0.01 N HCl) (100 μL) was added and the resultant solution was incubated overnight under the same conditions, which was followed by measurement of the absorbance at 570 nm using a microplate reader (ARVO, PerkinElmer).
MTT Assays of Jurkat Cells with 4 in the Presence of ATP and ADP.Jurkat cells (2.0 × 10 5 cells) were incubated in RPMI 1640 medium containing a solution of 4 (0.16−10 μM), etoposide (0.39−25 μM) and celastrol (0.16−10 μM) with ATP and ADP (100 μM) for 1 h.MTT reagent (5 mg/mL) in PBS (10 μL) was then added to the cells.After incubation at 37 °C for 4 h, a formazan lysis solution (10% sodium dodecyl sulfate (SDS) in 0.01 N HCl) (100 μL) was added and the resultant solution was incubated overnight under the same conditions, which was followed by measurement of the absorbance at 570 nm using a microplate reader (ARVO, PerkinElmer).
For the synthesis of 5, the cyclization reaction of 14 was carried out between Lys(Boc) at the N-terminus and Gly at the C-terminus by using HBTU, HOBt, and DIEA followed by deprotection of the benzyl group via catalytic hydrogenations using H 2 , Pd/C, and CH 3 COOH in MeOH and THF to give cyclic peptide 11 without racemization during the cyclization reaction.Condensation reactions of 11 with 9 and successive deprotection afforded 5.
We examined the stability of 4 containing linear peptides and 5 having the corresponding cyclic peptides against trypsin, which is a proteolytic enzyme that specifically hydrolyzes on the C-terminal side of R and K residues.We incubated 4 and 5 with trypsin (5 U/mL) for 1 h at 37 °C and analysis was conducted via RP-HPLC.As shown in Figure S1, considerable degradation of 4 was observed, while negligible decomposition of 5 was detected, which suggests a higher stability for 5 than that of 4 against trypsin digestion.
Evaluation of the Cytotoxicity of TPH-ACs Against Jurkat, HeLa-S3, A549 Cells and Normal (IMR90) Cell Lines.The cytotoxicities of 3, 4, 5, 6, and 7 against Jurkat cells, HeLa-S3 cells, A549 cells, and IMR90 cells (human Caucasian fetal lung fibroblasts) was examined by using propidium iodide (PI), a fluorescent DNA intercalator and a detector of dead cells. 31The cells (2.0 × 10 5 cells) were incubated with 3−7 for 1 or 2 h at 37 °C, which was followed by treatment with PI (1 μg/mL) for 15 min.As shown in Figure 1, cell death was observed in Jurkat, HeLa-S3, and A549 cells with similar morphological changes after incubation with 4 for 1 or 2 h at 37 °C under 5% CO 2 .Figures S2−S4  g) and (j) Bright-field images of HeLa-S3 cells, (h) and (k) emission images of PI, and (i) overlay images of (g) and (h), and (l) overlay images of (j) and (k).(m) and (p) Bright-field images of A549 cells, (n) and (q) emission images of PI, (o) overlay images of (m) and (n), and (r) overlay images of (p) and (q).Excitation was at 540 nm for propidium iodide.Scale bar (black) is 10 μm.Biochemistry μM, 24 h) and cisplatin (100 μM, 24 h) (the structures of these compounds are shown in Chart S1), were observed by fluorescence microscopy after staining with PI (Figure 2).The results indicated that the morphological change was different from that induced by the TPH-ACs presented in Figures 1 and S2−S4.
Our previous work strongly indicated that paraptosis induced by celastrol in Jurkat cells involves negligible membrane fusion between the ER and mitochondria.In contrast, the paraptosis induced by TPH-ACs (and IPH-ACs) is characterized by membrane fusion between the ER and mitochondria, as previously reported. 25,26These facts allowed us to classify the examples of paraptosis induced by celastrol and TPH-ACs (IPH-ACs) as paraptosis I and paraptosis II, respectively.Morphological change in HeLa-S3 and A549 cells induced by celastrol (100 μM, 24 h) was observed using fluorescence microscopy.It should be noted that HeLa-S3 and A549 cells were detached from the well after treatment with celastrol, possibly because celastrol inhibits cell adhesion through the inhibition of VEGFR2 (vascular endothelial growth factor receptor 2), as described by Huan and Simons. 32,33Therefore, the detached cells in culture medium were collected by centrifugation, treated with PI, and then observed via microscopy (Figure S5).
The cytotoxicity of TPH-ACs 3−7 against Jurkat, HeLa-S3, and A549 cells was evaluated via MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide) assay.The cells (2.0 × 10 5 cells/mL) were incubated with 3−7 in culture medium containing 10% fetal bovine serum (FBS) for 1 h at 37 °C under 5% CO 2 and then treated with the MTT reagent and the results are summarized in Figure 3 and Table 1.The EC 50 values of 3−7 were determined to be 0.7−2.2μM against Jurkat cells, 1.6−8.5 μM against HeLa-S3 cells, and 0.2−9.5 μM against A549 cells, respectively, and it was found that 4 exhibits a higher cytotoxicity than 3 against these cell lines.For comparison, the EC 50 values for 3−7 against IMR90 cells (human Caucasian fetal lung fibroblasts), a model of a normal  cell line, were determined to be >12.5 μM under the incubation conditions described above (Figure 3d).The EC 50 value of 4 against IMR90 cells was decreased to 4.4 μM after incubation for 5 h, suggesting slow induction of cell death in IMR90 cells by 4.
Effect of Inhibitors of Intracellular Events on Cell Death Induced by TPH-ACs.In our previous papers, we reported that the cell death induced by 1−3 is considerably inhibited by carbonyl cyanide 3-chlorophenylhydrazone (CCCP), which is an uncoupling reagent and an inhibitor of mitochondrial Ca 2+ uptake, 34−37 while cell death was negligibly inhibited by benzyloxycarbonyl-VAD(OMe)-fluoromethylketone (Z-VAD-fmk, a broad caspase inhibitor), 38−40 necrostatin-1 (a RIPK-1 inhibitor and necroptosis inhibitor), 41−43 and 3-methyladenine (3-MA), an inhibitor of autophagosome formation that functions by inhibiting the action of type III phosphatidylinositol 3-kinases (PI3K) 44−46 (the structures of these inhibitors are shown in Chart S2).This finding allowed us to conclude that the TPH-AC-induced cell death is paraptosis rather than apoptosis, necroptosis, or autophagy. 26n the present work, we conducted MTT assays of HeLa-S3 and A549 cells treated with 4 and celastrol in the presence of Z-VAD-fmk, necrostatin-1, 3-MA, CCCP and trifluoromethoxy carbonyl cyanide phenylhydrazone (FCCP), which is an uncoupler of oxidative phosphorylation in mitochondria. 47urkat, HeLa-S3, and A549 cells were incubated with these inhibitors and then treated with 4 for 1 h and celastrol for 24 h at 37 °C under 5% CO 2 .As shown in Figure 4, the 4-induced cell death in Jurkat, HeLa-S3 and A549 cells was considerably inhibited by CCCP, but negligibly inhibited by Z-VAD-fmk, necrostatin-1, and 3-MA, which strongly suggests that apoptosis, necroptosis, and autophagy are unlikely in these cancer cell lines.In addition, celastrol-induced cell death was negligibly inhibited by any inhibitors, as shown in Figure S6.
Changes in Calcium Concentrations in Mitochondria and Cytosol.Since it was reported that the transfer of Ca 2+ from the ER to mitochondria is crucial in paraptosis II in Jurkat cells induced by 3, 26 we conducted the measurement of Ca 2+ concentrations in mitochondria and cytosol in Jurkat, HeLa-S3, and A549 cells after treatment with 4 by flow cytometric analysis.Jurkat, HeLa-S3, and A549 cells were preincubated with Rhod-2/AM (5 μM), a probe of mitochondrial Ca 2+ , 48 or Rhod-4/AM (5 μM), an indicator of cytosolic Ca 2+ , 49 for 30 min and then treated with 4 (2.5, 5,  Figure 5 shows the emission intensity of Rhod-2/AM and Rhod-4/AM in Jurkat cells and that of Rhod-2/AM in HeLa-S3 and A549 cells after treatment with 4 for 0, 15, 30, 45, and 60 min, which suggests that 4 promotes an increase in mitochondrial Ca 2+ concentrations.
In addition, we examined the time-dependent change of fluorescence emission from Rhod-2/AM and Rhod-4/AM on fluorescence microscopy and microplate reader after treatment with 4. The considerable enhancement of red emissions from Rhod-2/AM was observed in Jurkat cells after incubation with 4 in about 30 min, as shown in Figures S7 and S8a, which is parallel to the observation shown in Figure 5a.In Figure 5b as well as Figure S8a, moderate∼weak enhancement was observed in the fluorescence intensity of Rhod-4/AM.
We also examined the effect of 4 on the emission of Rhod-2/ AM and Rhod-4/AM in HeLa-S3 and A549 cells, as shown in Figures S7, S8b and S8c.The enhancement of the red emission from Rhod-2/AM was observed in HeLa-S3 and A549 cells after incubation with 4 for 1 h, while there was a negligible enhancement of emission from Rhod-4/AM.These results are almost parallel to the observation in Figures 5c−f, confirming the generality of the Ca 2+ overload in mitochondria rather than in the cytosol in these cancer cells.Figure S8d shows a somewhat slower emission enhancement from Rhod-2/AM in IMR90 cells and this point is described below.
It has been reported that celastrol promotes an increase in Ca 2+ concentration in the cytosol. 12,13,22Therefore, we measured the emission changes of Rhod-2/AM and Rhod-4/ AM for about 4 h after addition of celastrol, and observed a moderate enhancement of Rhod-2/AM rather than Rhod-4/ AM (Figure S8e).
For comparison, we examined the time-dependent fluorescence changes of Rhod-2/AM and Rhod-4/AM in HeLa-S3 and A549 cells after addition of celastrol (a paraptosis I inducer) and cisplatin (an apoptosis inducer) on fluorescence microscopy, in which a mitochondrial and cytoplasmic Ca 2+ overload in HeLa-S3 cells and a mitochondrial Ca 2+ overload in A549 cells proceeded in 2−6 h, while the cytoplasmic Ca 2+ concentration was negligibly changed by celastrol (Figures S9  and S10).On the other hand, a weak enhancement of the red emissions from Rhod-2/AM was observed in HeLa-S3 cells after incubation with cisplatin, as shown in Figure S11, which indicated a small increase in mitochondrial Ca 2+ .

Induction of Membrane Fusion Between Mitochondria and the ER in Jurkat, HeLa-S3, and A549 Cells by 4.
Next, we conducted costaining experiments of the ER and mitochondria in Jurkat, HeLa-S3, and A549 cells with ERTracker Red, and MitoTracker Green, which are selective probes of the ER and mitochondria, respectively.Jurkat, HeLa-S3, and A549 cells were treated with MitoTracker Green and ERTracker Red for 1 h, respectively, and then incubated with 4 for 15 min (Jurkat and A549), or for 5 and 10 min (HeLa-S3).As shown in Figure 6a−e (Jurkat), 6k-o (HeLa-S3), and 6z-ad (A549), a weak overlap of the emission from MitoTracker Green and ERTracker Red was observed before the addition of 4, and a considerable overlap was observed after incubation with 4 as shown in Figure 6f−j (Jurkat), 6p-y (HeLa-S3), and 6ae-ai (A549).In these experiments, faster degradation of the ER was observed in HeLa-S3 than in Jurkat and A549 cells. 50

Biochemistry
The emission intensity profiles of MitoTracker Green and ERTracker Red are indicated by the green and red curves, respectively, in Figure 7 between points I and II in the corresponding images of Figure 6 (more intensity profiles at other points are presented in Figure S12).For comparison, we also examined the effect of celastrol on the interaction between the ER and mitochondria in HeLa-S3 and A549 cells.As shown in Figure S13a− z) and (ae) Bright-field images of A549 cells, (aa) and (af) emission images of MitoTracker Green, (ab) and (ag) emission images of ERTracker Red, (ac) overlay images (z−ab), (ad) overlay images (z−ac), (ah) overlay images (ae-ag), and (ai) overlay images (ae-ah).Excitation at 473 nm for (b), (g), (l), (q), (v), (aa) and (af) and at 559 nm for (c), (h), (m), (r), (w), (ab) and (ag).Exposure time was 20 μs/pixel.Scale bar (black) is 10 μm.trisphosphate) receptor (InsP 3 R), the voltage-dependent anion channel (VDAC), and mitochondrial Ca 2+ uniporter (MCU) involved in Ca 2+ transport from the ER and mitochondria at their interface. 51− 53 We previously reported that Ruthenium Red (RuRed), (an inhibitor of MCU), 54 and 2-aminophenyl borate (2-APB), an antagonist of InsP 3 R, (the structures of these inhibitors are shown in Chart S3), 55 inhibit paraptosis II induced by TPH-ACs (Scheme 4) to some extent. 26−64 For this purpose, the effect of an inhibitor of VDAC, 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid (DIDS) 65−69 (Chart S3) on 4-induced paraptosis II was checked.Jurkat, HeLa-S3, and A549 cells were preincubated with DIDS (75, 100, or 250 μM for these cancer cell lines, respectively) for 3 h and then treated with 4 for 1 h.The results of MTT assays indicate that DIDS considerably inhibits paraptosis II induced by 4 in Jurkat (Figure 9a), HeLa-S3 (Figure 9b), and A549 cells (Figure 9c).The inhibitory effect of 2-APB and RuRed on the 4-induced paraptosis II was found to be rather weak in this work.It should also be mentioned that it was reported that thapsigargin, which is an inhibitor of Ca 2+ transport between the sarco-endoplasmic reticulum (SR/ ER) negligibly inhibit paraptosis II induced by 3, one of our previous TPH-ACs. 26e compared the effects of 2-APB, RuRed, and DIDS on paraptosis I induced by celastrol with paraptosis II induced by 4. Jurkat cells were preincubated with 2-APB (50 μM), RuRed (75 μM), and DIDS (75 or 100 μM) and then treated with celastrol (5 μM) for 24 h.As shown in Figure 9d, these inhibitors did not significantly stop cell death induced by celastrol.
We checked the direct interaction of 4 and celastrol with DIDS and found the formation of weak precipitates in a mixture of 4 and DIDS in water (containing ca.1% DMSO) at neutral pH, possibly due to the electrostatic interaction between polycationic 4 (possibly 12+) and dianionic DIDS (DIDS 2− ). 70Therefore, we could not exclude the possibility of the suppression of paraptosis II by the complexation of TPH-ACs with DIDS inside cancer cells and hence DIDS was not used for further mechanistic studies described below.It is considered that interaction between anionic celastrol (due to the anionic form of carboxylate) with DIDS 2− is very weak.
Observation of Fe 2+ and Zn 2+ Ions in Paraptosis Induced by 4. Apoptosis inducers are known to cause a concentration change in Fe 2+ and Zn 2+ ions in Jurkat cells during apoptotic processes. 71Therefore, we examined the effect that Zn 2+ ions exert on the cytotoxicity of 4 against Jurkat cells.A mixture of 4 and Zn(NO 3 ) 2 was incubated for 1

Biochemistry
labile Fe 2+ in mitochondria) 72,73 and zinquin (a probe for intracellular Zn 2+ ) 74−79 (the structures of these probes are shown in Chart S4).The enhancement of the green fluorescence emission from Mito-FerroGreen shown in Figure S16 and blue emission from zinquin shown in Figure S17 suggests that increases in both free-Fe 2+ and free-Zn 2+ during apoptosis were induced by cisplatin and etoposide. 50,80On the other hand, the intensity of these emissions was negligibly enhanced in Jurkat cells treated with 4 and celastrol.These facts imply that the change in intracellular free-Zn 2+ and free-Fe 2+ was negligible during both paraptosis I (by celastrol) and paraptosis II (by TPH-ACs) and that the destruction of Ca 2+ homeostasis is important in these paraptosis processes.
Effect of ATP and ADP on Paraptosis Induced by 4.An increase in ATP concentrations in the mitochondria is known to enhance the transport of Ca 2+ from the extracellular space into mitochondria. 81−86 Therefore, we examined the effect of ATP uptake into cancer cells by agonists of P2X, which are ATP ligand-gated ion channel receptors.Because ADP and ATP had been established as a natural agonist and a competitive antagonist of P2Y1 which are G protein-coupled receptors, respectively, 81−86 we conducted MTT assays of Jurkat cells with 4 and celastrol, paraptosis inducers, and etoposide, an apoptosis inducer, in the presence of ATP and ADP.As shown in Figure S18, the EC 50 values of compound +ATP and compound+ADP approximated those of the compounds alone.The data suggests that ATP and ADP (extracellular) exert negligible effects on the cytotoxicity of these cell-death inducers.
Mechanistic Study Regarding Cancer Cell/Normal Cell Selectivity of TPH-ACs.TPH-ACs exhibit lower levels of toxicity in IMR90 cells than that in cancer cell lines, as described above (Table 1).Next, we then conducted costaining experiments of the ER and mitochondria in IMR90 cells with ERTracker Red and MitoTracker Green.IMR90 cells were treated with MitoTracker Green and ERTracker Red for 1 h, respectively, and were then incubated with 4 for either 15 or 60 min.As shown in Figure S19 and in Figure 10, a weak overlap of the emission from MitoTracker Green and ERTracker Red was observed both before and after the addition of 4 (more points for the intensity profile are presented in Figure S20).One more possibility suggests the occurrence of smaller numbers of VDAC and MCU in normal cells 87,88 or slower response of these Ca 2+ channels, as shown in Figure S8e, than that in cancer cells.
In addition, we examined the time-dependent fluorescence changes in either Rhod-2/AM or Rhod-4/AM in IMR90 cells using fluorescence microscopy and a microplate reader after treatment with 4 for 60 min.As shown in Figure S8d, enhancement of the red emissions from Rhod-2/AM in IMR90 cells is induced more slowly than in other cancer cell lines (see also Figure S21).We assume that the differences between cancer cells and normal cells account for the cancer cell selectivity of TPH-ACs.
Possible Mechanisms of Paraptosis II Induced by TPH-ACs in Jurkat, HeLa-S3, and A549 Cells.The aforementioned mechanistic studies are summarized as follows.
(1) TPH-ACs 4−7 (Scheme 1) were newly designed and synthesized in this work.The stability of 5 with cyclic peptide units is higher than that of 4 with corresponding linear peptide units (Figure S1).We found that 4 exhibits a higher level of cytotoxicity compared with that of 3 against these cancer cell lines, which was possibly due to the effect of a somewhat hydrophobic Cha residue in the peptide units of 3.
(2) TPH-ACs 4−7 have potent cytotoxicity against Jurkat, HeLa-S3, and A549 cells, as examined by PI-staining experiments and MTT assays.These TPH-ACs induce cell death after treatment for 1 h with morphological changes similar to that induced by our previous TPH-ACs such as 3 (Figures S2−4), albeit with EC 50 values that were smaller than those of 3 (Figure 3a−c and Table 1).Moreover, the cytotoxicity of TPH-ACs against IMR90 cells, a model of a normal cell line, is weaker than that against the cancer cells (Figure 3d and Table 1).
(3) TPH-ACs such as 4 induce membrane fusion (or tethering) between the ER and mitochondria, as indicated by costaining experiments of the ER and mitochondria in HeLa-S3 and A549 cells as well as in Jurkat cells, which is suggested to be common phenomena in paraptosis II processes (Figures 6 and 7).On the other hand, a weaker overlap was observed in cancer cell lines after incubation with celastrol than after incubation with 4 (Figures S13 and 8 in the text).It was reported by D'Eletto, M. et al. that the ER and mitochondria undergo membrane fusion by stimulating with transglutaminase 2 (TG2). 89However, the effect of a TG2 inhibitor, dansylcadaverine, on paraptosis II was negligible in our previous work. 264) The 4-induced paraptosis II in Jurkat, HeLa-S3 and A549 cells is considerably inhibited by CCCP (an inhibitor of mitochondrial Ca 2+ uptake) 36 and DIDS (an inhibitor of VDAC), but negligibly inhibited by inhibitors of apoptosis, necroptosis, and autophagy (Figures 4 and 9).On the other hand, the cell death induced by celastrol, which is known to induce a previously reported type of paraptosis (paraptosis I), was negligibly inhibited by these agents (Figure S6).
(5) TPH-AC 4 considerably promotes an increase in Ca 2+ concentrations in mitochondria as observed using flow cytometry, fluorescence microscopy, and a microplate reader (Figures 5, S7, and S8).Observation of Ca 2+ overload in mitochondria of Jurkat, HeLa-S3, and A549 cells by flow cytometry (Figure 5) and a microplate reader (Figure S8) strongly support the enhancement of Ca 2+ concentrations in mitochondrial rather than in cytoplasm.(6) To analyze the inhibitory effect of CCCP on paraptosis II induced by 4 in Jurkat, HeLa-S3, and A549 cells, the change of Ca 2+ concentrations in mitochondria (by Rhod-2/AM) and in cytosol (by Rhod-4/AM) of these cancer cells was measured on flow cytometry after the pretreatment with CCCP prior to the addition of 4 (Figure S22).When Jurkat cells were pretreated with CCCP (see Figures S22a, b, g, h), the emission from Rhod-2/AM was enhanced by CCCP in the initial 15 min and then suppressed later (please compare Figure 5a in the text with Figure S22b), although broad signals were observed with respect to the Ca 2+ concentrations in mitochondria and cytosol regardless of the presence or absence of 4 (see Figures S22a, b,  g, h).In HeLa-S3, and A549 cells, the emission from Rhod-2/ AM was suppressed by CCCP (Figure 5c in the text vs Figure S22d and Figures 5e vs S22f).The effect of DIDS on paraptosis II is not conclusive, because of the finding of direct salt formation of 4 with DIDS in aqueous solution at neutral pH, as described above.Therefore, DIDS was not used for further mechanistic studies.(7) It should be mentioned that the results of flow cytometry of Jurkat cells, which were treated with 4 in the presence of CCCP, shown in Figure S22 (especially, Figures S22g and S22h) show broad and complicated spectra, possibly due to the detection of different cells which include Ca 2+ at different concentrations.Then, we have decided to use a microplate reader, which was expected to detect the total emission from Rhod-2/AM (mitochondrial Ca 2+ ) and Rhod-4/ AM (cytosol Ca 2+ ) in Jutkat, HeLa-S3 and A549 cells after addition of 4 in the presence of CCCP and 2-APB.As summarized in Figure 11, emission enhancement of Rhod-2/ AM after the addition of 4 was moderately inhibited by CCCP (red dashed curves), while the effect of 2-APB (black dashed curves) was not so strong.On the other hand, the effect of CCCP and 2-APB on the emission of Rhod-4/AM was negligible, because the change of emission from Rhod-4/AM was negligible even after the addition of 4 (black curves with black filled squares).Although complete inhibition of Ca 2+ transfer from the ER to mitochondria by CCCP was not observed, these results may support the important roles of mitochondrial Ca 2+ uptake in paraptosis II induced by 4.
(8) The costaining experiments of the ER and mitochondria in Jurkat cells after treatment with 4 in the presence of CCCP were also conducted.As summarized in Figure S23, the overlap of the emission from MitoTracker Green and ERTracker Red by 4 (2.5 μM) was considerably inhibited by CCCP (40 μM). 90Consideration on aforementioned results together with Figure 11 in the text (weak inhibitory activity of CCCP against mitochondrial Ca 2+ overload) and Figure S23 strongly suggests that CCCP exhibits inhibitory effect on the-ER-mitochondria fusion (or tethering) and the Ca 2+ transfer from the ER to mitochondria.(9) We measured the mitochondrial membrane potential (MMP, ΔΨ m ) in Jurkat cells using 1,1′,3,3,3′,3′-hexamethy-lindodicarbocyanine iodide (DilC1(5)), 91 which is a probe that responds to ΔΨ m .Jurkat cells were stained with DilC1(5) (500 nM) for 30 min and then treated with 4 (2.5 μM) for 10−60 min for the observation on fluorescence microscopy.As shown in Figure S24, the emission of DilC1(5) starts to decrease in ca.20−40 min after addition of 4, while weakly CCCP inhibits the decrease in the emission of DilC1(5) by 4. 92 Although more detailed experiments will be required in the next work, these results suggest a possibility that the ΔΨ m values decrease in later steps of paraptosis II than Ca 2+ transfer to mitochondria and membrane fusion between the ER and mitochondria.
(10) Scheme 5 summarizes our revised proposal to explain the mechanism of paraptosis II induced by 4 in Jurkat, HeLa-S3, and A549 cells and the reaction points of two paraptosis II inhibitors, DIDS and CCCP.As described above, 4 induces Ca 2+ overload from the ER to mitochondria in these cells and induce membrane fusion between the ER and mitochondria almost simultaneously in 10−20 min after addition of 4. After that, the ΔΨ m values are decreased.It is suggested that CCCP exhibits inhibitory effect on the Ca 2+ transfer from the ER to mitochondria and the ER-mitochondria fusion (or tethering) and this point is yet to be studied.Inhibition of paraptosis II by DIDS could be explained by its complexation with 4 (and IPH-ACs such as 1a) due to the electrostatic interaction between them.Most importantly, it was concluded that TPH-ACs such as 4 induce paraptosis II via similar mechanism at least in these three cancer cell lines.(11) The cytotoxicity of TPH-ACs against IMR90 cells is weaker and slower (EC 50 value of 4 is lowered to 4.4 μM after incubation for 5 h) than that against the aforementioned three cancer cell lines (Figure 3d, Table 1, and Scheme 5).This result is attributed to slower mitochondrial Ca 2+ increase in IMR90 cells than that in cancer cells (Figure S8d) and weaker interaction between the ER and mitochondria than that in cancer cell lines even after treatment with 4, as proven by the costaining experiments using MitoTracker Green and ERTracker Red (Figures S19 and 10).
(12)Although intracellular free-Fe 2+ and free-Zn 2+ ions during apoptosis in Jurkat cells were increased, the effect of intracellular Fe 2+ and Zn 2+ ions during paraptosis induced by 4 and celastrol was negligible (Figures S16 and S17).The effects of ATP and ADP, which are an agonist and an antagonist of P2Y1 on paraptosis II, were also negligible (Figure S18).

■ CONCLUSIONS
In this paper, we report the results of detailed mechanistic studies of the paraptosis that is induced by triptycene-peptide hybrids as amphiphilic conjugates (TPH-ACs) in Jurkat, HeLa-S3, and A549 cells.We confirmed that the paraptosis II induced by TPH-ACs is different from the cell death induced by celastrol, which induces paraptosis I in our definition.TPH-AC-induced cell death is inhibited by CCCP, 2-APB, and DIDS in these cancer cell lines, while celastrolinduced paraptosis I is not affected by these inhibitors.The results of experiments strongly suggest that TPH-ACs induce a transfer of Ca 2+ into mitochondria possibly from the ER and membrane fusion of the ER and mitochondria in Jurkat, HeLa-S3, and A549 cells, resulting in the induction of paraptosis II in these three cancer cell lines, which is the main point in our hypothesis on paraptosis II.Although the details of mechanism of the inhibition of paraptosis II by CCCP is not fully understood, there is a high possibility that Ca 2+ from the ER to mitochondria and the membrane fusion between the ER and mitochondria are inhibited by CCCP.
To the best of our knowledge, there is no approved anticancer drug that induces paraptosis II in cancer cells.As described in the Introduction, paraptosis (paraptosis I and II) are not fully understood and its deep understanding could lead to new strategies for the treatment of not only cancers but also autoimmune diseases and other related diseases.At the same time, we assume that the finding of potent inhibitors of intracellular events is very important for their mechanistic study.For example, Z-VAD-fmk is used as one of useful caspase inhibitors and contributes to characterization and mechanistic study of apoptosis.In this work, CCCP was found as a potent inhibitor of paraptosis II and the finding of more potent and selective inhibitors would contribute to the mechanistic studies of paraptosis.
A detailed mechanistic study concerning selective toxicity against cancer cell lines over normal cells (IMR90 cells in this study) suggests the time-dependent toxicity of TPH-ACs against cancer cells over normal cells, which could be one of the future strategies in cancer chemotherapy and in controlling toxicity of anticancer drugs.
We conclude that the results reported in this study provide useful information for not only the mechanistic studies of paraptosis, but also the development of drugs with the potential to target cancer cells while minimizing the side effects on normal cells.
The stability of 4 and 5 after treatment with trypsin (Figure S1), fluorescence microscopic images of Jurkat (Figure S2), HeLa-S3 (Figure S3), and A549 cells (Figure S4) after treatment with 3, 4, 5, 6, and 7, fluorescence microscopic images of HeLa-S3, and A549 cells after treatment with celastrol (Figure S5), results of the MTT assays of HeLa-S3 and A549 cells after treatment with celasterol in the presence of Z-VAD-fmk, necrostatin-1, 3-MA, and CCCP (Figure S6), fluorescence microscopic images of Jurkat, HeLa-S3, and A549 cells treated with Rhod-2/AM or Rhod-4/AM in the presence of 4 (Figure S7), time-dependent change of fluorescence emission from Rhod-2/AM or Rhod-4/AM in Jurkat, HeLa-S3, A549, and IMR90 cells after addition of 4 and celastrol (Figure S8), fluorescence microscopic images of HeLa-S3 (Figure S9) and A549 cells (Figure S10) stained with Rhod-2/AM or Rhod-4/AM after the treatment with celastrol, fluorescence microscopic images of HeLa-S3 cells treated with Rhod-2/AM or Rhod-4/AM in the presence of cisplatin (Figure S11), and A549 cells treated with MitoTracker Green, and ERTracker Red in the presence of celastrol (Figure S13), emission intensity profiles of MitoTracker Green and ERTracker Red in Figure S13 (Figure S14), the results of the MTT assay of Jurkat cells treated with 4 in the presence of Zn(NO 3 ) 2 (Figure S15), fluorescence microscopic images of Jurkat cells treated with Mito-FerroGreen in the presence of cisplatin, etoposide, 4 and celastrol (Figure S16), fluorescence microscopic images of Jurkat cells treated with zinquin in the presence of cisplatin, etoposide, 4, and celastrol (Figure S17), results of the MTT assays of Jurkat cells after treatment with 4, celastrol, and etoposide in the presence of ATP and ADP (Figure S18), typical fluorescence confocal microscopy images of IMR90 cells treated with MitoTracker Green, and ERTracker Red in the presence of 4 (Figure S19), emission intensity profiles of MitoTracker Green and ERTracker Red obtained from Figure S19 (Figure S20), fluorescence microscopic images of IMR90 cells treated with Rhod-2/AM or Rhod-4/AM in the presence of 4 (Figure S21), the results of flow cytometry analysis of Jurkat, HeLa-S3, and A549 cells after treatment with Rhod-2/AM and Rhod-4/AM in the presence of 4 and CCCP (Figure S22), typical fluorescent confocal microscopy images of Jurkat cells stained with MitoTracker Green, and ERTracker Red after the pretreatment with CCCP and 4 and the emission intensity profiles of MitoTracker Green and ERTracker Red (Figure S23), fluorescence microscopic images of Jurkat cells treated with DilC1(5) and 4 in the presence of CCCP (Figure S24), and the chemical structures of compounds, inhibitors and probes (Charts S1−S4) (PDF) ■ AUTHOR INFORMATION

Figure 4 .
Figure 4. Results of an MTT assay of Jurkat (a), HeLa-S3 (b), and A549 cells (c) after the treatment with 4 in the presence of Z-VADfmk, necrostatin-1, 3-MA, CCCP, FCCP and then 4 at the indicated concentrations of these agents.
Figure 7a (Jurkat), Figure 7c (HeLa-S3), and Figure 7f (A549) show a partial overlap of the emission from MitoTracker Green and ERTracker Red before treatment with 4 and Figures 7b (Jurkat), Figures 7d−e (HeLa-S3), and Figure 7g (A549) show their considerable overlap, indicating that 4 induces tethering or membrane fusion between the ER and mitochondria in all of these cells.
e along with Figure 8a (HeLa-S3) (also in Figure S13k−o along with Figure 8c (A549)), a weak overlap of the emission from MitoTracker Green and ERTracker Red was observed before the addition of celastrol.A weaker overlap was observed after incubation with celastrol than after incubation with 4, as shown in Figure S13f−j along with Figure 8b (HeLa-S3) (also in Figure S13p−t along with Figure 8d (A549 cells).More intensity profiles at other points are shown in Figure S14.Effect of Inhibitors of Intracellular Ca 2+ -related Events Between the ER and Mitochondria.The aforementioned results suggest that mitochondrial Ca 2+ uptake is promoted by 4 in the paraptotic II processes.Scheme 4 shows the proposed relationship of InsP 3 (inositol 1,4,5-

Scheme 4 .
Scheme 4. Relationship Between InsP 3 Receptor, VDAC, MCU and Mitochondrial Ca 2+ uptake and Their Inhibitors in the Proposed Processes of Paraptosis II

Figure 9 .
Figure 9. (a−c) The results of MTT assays of Jurkat (a), HeLa-S3 (b), and A549 cells (c) pretreated with 2-APB and DIDS and then treated with 4 at the indicated concentrations.(d) The results of MTT assays of Jurkat cells pretreated with 2-APB, RuRed, and DIDS prior to addition of celastrol.

Figure 10 .
Figure 10.Emission intensity profiles of MitoTracker Green (green curves) and ERTracker Red (red curves) in IMR90 cells before and after treatment with 4 (5 or 50 μM) from point I to point II in Figure S19d (a), Figure S19i (b), and Figure S19n (c), respectively.

Figure 11 .
Figure 11.Time-dependent change of fluorescence emission from Rhod-2/AM (left side) and Rhod-4/AM (right side) (excitation at 540 nm and emission at 590 nm) in Jurkat (a), HeLa-S3 (b), and A549 cells (c) measured on microplate reader after the pretreatment with CCCP (40 μM for Jurkat, 160 μM for HeLa-S3 and A549 cells) and 2-APB (100 μM for Jurkat, HeLa-S3 and A549 cells) prior to the addition of 4 (2.5 μM for Jurkat, 5 μM for HeLa-S3, and 10 μM for A549 cells).Black plain curves with filled squares indicate emission from Rhod-2/AM and Rhod-4/AM in the absence of CCCP and 2-APB.Black dashed curves with open squares indicate emission from Rhod-2/AM and Rhod-4/AM in the presence of 2-APB.Red dashed curves with filled triangles indicate emission from Rhod-2/AM and Rhod-4/AM, respectively, in the presence of CCCP at the concentrations described above.A.u. is arbitrary unit.