Anti-Cancer Peptide PNC-27 Kills Cancer Cells by Unique ...

• © 2024 by the Association of Clinical Scientists, Inc.

1. Victor Adler3,
2. Richard D. Feinman2,
3. Anna Miller2,
4. Miriam Silberstein3,
5. Ehsan Yazdi4⇑ and
6. Matthew R. Pincus1⇑

1. 1Department of Pathology, Brooklyn, NY, USA
2. 2Department of Cell Biology, SUNY Downstate Medical Center, Brooklyn, NY, USA
3. 3Touro College, Brooklyn, NY, USA
4. 4Nomocan Pharmaceuticals, New York, NY, USA

1. Address correspondence to Matthew R. Pincus; Department of Pathology, SUNY Downstate Medical Center, 450 Clarkson Avenue, Brooklyn, NY 11203, USA; e mail: matthew.pincus{at}downstate.edu
2. Address correspondence to Ehsan Yazdi: Nomocan Pharmaceuticals, 310 East 67th Street, New York, NY 10065, USA; e mail: ehsan.yazdi{at}nomocan.com

Abstract

Objective We have previously shown that the anti-cancer peptide PNC-27 kills cancer cells by co-localizing with membrane-expressed HDM-2, resulting in transmembrane pore formation causing extrusion of intracellular contents. We have also observed cancer cell mitochondrial disruption in PNC-27-treated cancer cells. Our objectives are to determine: 1. if PNC-27 binds to the p53 binding site of HDM-2 (residues 1-109) in the cancer cell membrane and 2. if this peptide causes selective disruption of cancer cell mitochondria.

Methods For aim 1, we incubated MIA-PaCa-2 human pancreatic carcinoma cells with PNC-27 in the presence of a monoclonal antibody against the amino terminal p53 binding site of HDM-2 to determine if it, but not negative control immune serum, blocks PNC-27-induced tumor cell necrosis. For the second aim, we incubated these cells with PNC-27 in the presence of two specific dyes that highlight normal organelle function: mitotracker for mitochondria and lysotracker for lysosomes. We also performed immuno-electron microscopy (IEM) with gold-labeled anti-PNC-27 antibody on the mitochondria of these cells treated with PNC-27.

Results Monoclonal antibody to the p53 binding site of HDM-2 blocks PNC-27-induced cancer cell necrosis, whereas negative control immune serum does not. The mitochondria of PNC-27-treated cancer cells fail to retain mitotracker dye while their lysosomes retain lysotracker dye. IEM of the mitochondria cancer cells reveals gold particles present on the mitochondrial membranes.

Conclusions PNC-27 binds to the p53 binding site of HDM-2 (residues 1-109) inducing transmembrane pore formation and cancer cell necrosis. Furthermore, this peptide enters cancer cells and binds to the membranes of mitochondria, resulting in their disruption.

Introduction

PNC-27 is the prototype of a new class of peptides that has been found to kill cancer cells while having no effect on untransformed cells. The sequence of this peptide was designed to contain residues 12-26 of the p53 tumor suppressor protein. This sequence contains the binding site for HDM-2, the target of p53 [1]. Binding of p53 to HDM-2 results in ubiquitination of p53 and its hydrolysis in the proteosome, thereby limiting its ability to induce apoptosis and block proliferation of cancer cells [2]. Although PNC-27 was designed by our research group to block the binding of p53 to HDM-2 in the nuclei of cancer cells as a decoy peptide, it was observed that it rapidly kills cancer cells without p53-induced apoptosis. Furthermore, surprisingly, PNC-27 killed cancer cells that were homozygously deleted for p53 [3,4]. High resolution transmission electron micrographs revealed that PNC-27 induced transmembrane pores in the cancer cells’ membranes [5]. Importantly, PNC-27 had no such effect on a wide variety of normal (untransformed) cells [6-8].

Investigation of the basis of this process revealed that a wide variety of cancer cells express HDM-2 in their cell membranes, whereas normal cells express either low levels of or no HDM-2 in their membranes. Both cancer cells and normal cells express high levels in their nuclei [9]. Fluorescent labeled antibodies to PNC-27 and to HDM-2 show colocalization of PNC-27 and HDM-2 in the membranes of cancer cells but not in normal cells [9]. The fact that HDM-2 is the target of PNC-27 was revealed in experiments in which untransformed human MCF-10-2A breast epithelial cells that were insensitive to PNC-27 were transfected with plasmids expressing different forms of HDM-2, including full-length HDM-2, with a membrane localization peptide on its carboxyl terminus, full length HDM-2, truncated HDM-2 that lacked the amino terminal 1-109 sequence which functions as the binding domain for p53, and empty vector [9]. Of the cells expressing these constructs, only the ones transfected with the full length HDM-2 containing the membrane localization peptide expressed HDM-2 in their membranes and became susceptible to PNC-27 [9].

We have performed high resolution immuno-scanning electron microscopic (SEM) studies of the pores induced by PNC-27 in cancer cell membranes [10]. We incubated two different cancer cell lines A2058 (human melanoma cells) and MIA-PaCa-2 (metastatic human pancreatic cancer cells) with anti-PNC-27 and anti-HDM-2 antibody systems labeled with 6 nm and 15 nm gold labels, respectively, and found that PNC-27-HDM-2 complexes form in multiple duplicate pairs surrounding well-defined transmembrane pores [10].

We have further tested PNC-27 and its shorter homologue peptide, PNC-28 (containing p53 residues 17-26 linked to the transmembrane penetrating peptide) in vivo. In one set of studies, we have found that PNC-28 administered continuously via implanted intraperitoneal pumps to nude mice resulted in eradication of a highly metastatic pancreatic cancer cell line (BMRPA1-TUC-3) transplanted subdermally. These mice survived with no exhibition of off-target effects and showed weight gain that was statistically identical to that of sham-treated control mice [11].

Recently, an independent investigation of the in vivo effects of PNC-27 on acute myelogenous leukemia (AML) was performed at the City of Hope Medical Center [12]. In these studies, tumor stem cell-enriched samples were obtained from nine different patients and placed into cell culture. PNC-27 was found to be cytotoxic to these cells but not to normal leukocytes [12]. All nine cell lines expressed HDM-2 in their membranes, and it was found that the IC50 of PNC-27 for cell killing decreased linearly with the level of membrane expression of HDM-2 [12].

Each of these cell lines was transplanted into the bone marrows of nude mice. These mice exhibited high white cell counts and were then treated via daily intra-peritoneal injection of PNC-27 for three weeks. The white cell counts of these mice normalized by the end of this period. Bone marrow cells from these treated nude mice were transplanted into the marrows of untreated nude mice. The white cell counts of these mice remained normal indicating eradication of the leukemia cells including the stem cells. Both cohorts of mice showed markedly prolonged Kaplan-Meier survival curves compared with those of sham-treated mice. The results of these in vivo studies suggest that PNC-27 has great potential in treating human cancers [12].

In these latter studies, it was found that the function of HDM-2 in the cell membranes was to bind to the cell adhesion molecule, cadherin 3, which is involved in cell-cell contacts that result in inhibition of cell division [12]. Binding of HDM-2 to cadherin 3 results in the ubiquitination of cadherin 3 resulting in its degradation in the proteosome, a finding that corroborates the results of prior studies on expression of HDM-2 in metastatic breast cancer cells [13].

In our electron microscopic studies of cancer cells treated with PNC-27 [10], we noticed that, in addition to the presence of transmembrane pores, mitochondria appeared to exhibit features of organelle damage [10]. Since these features were observed after short incubations of cancer cells with PNC-27, mitochondrial dysfunction seemed to be associated with the effects of PNC-27 molecules that entered the cells possibly through the transmembrane pores. In this present study, therefore, we investigated the effects of PNC-27 on cancer cell organelles and present evidence that this peptide induces selective mitochondrial damage while not affecting other organelles such as lysosomes. In addition, we present definitive evidence that PNC-27 interacts with HDM-2 in its amino terminal binding domain (residues 1-109).

Materials and Methods

Peptides. PNC-27, H-Pro-Pro-Leu-Ser-Gln-Glu-Thr-Phe-Ser-Asp-Leu-Trp-Lys-Leu-Leu-Lys-Lys-Trp-Lys-Met-Arg-Arg-Asn-Gln-Phe-Trp-Val-Lys-Val-Gln-Arg-Gly-OH, was synthesized using solid phase methodology (Biopeptides Corp., La Jolla, CA) and was >95% pure by HPLC and mass spectrographic analysis. The bold-type sequence corresponds to amino acid residues 12-26 of the HDM-2-binding domain of human p53 while the italicized sequence corresponds to the membrane residency peptide (MRP) segment. The negative control peptide, PNC-29, containing the X13 peptide from cytochrome P450 (bold-type) attached to the MRP (italics), H-Met-Pro-Phe-Ser-Thr-Gly-Lys-Arg-Ile-Met-Leu-Gly-Glu-Lys-Lys-Trp-Lys-Met-Arg-Arg-Asn-Gln-Phe-Trp-Val-Lys-Val-Gln-Arg-Gly-OH, was likewise synthesized by solid phase methodology, was >95% pure and was purchased from Biopeptide Corp as above. For the experiments stock solutions of 20mg peptide/ml Dulbecco’s phosphate buffered saline (PBS) were prepared of each peptide which was distributed in volumes of 0.05 to 0.1 ml into sterile protease-free vials and stored frozen at −80°C. The stock was used for up to three months without loss of activity whereby repeated freezing and thawing of individual vials was strictly avoided.

Antibodies. Anti-H(M)DM-2 rabbit polyclonal antibody, N-20; sc-813 with epitope mapping at the N-terminus (residues 1-102) of MDM2 and mouse monoclonal anti-H(M)DM2 OP145 (CalBiochem OP145) with epitope mapping to amino terminal amino acid residues 19-50 of HDM-2 were purchased from Calbiochem (San Diego, CA, USA). Anti-MRP rabbit polyclonal IgG, was raised against the MRP sequence of PNC-27 (Lys-Lys-Trp-Lys-Met-Arg-Arg-Asn-Gln-Phe-Trp-Val-Lys-Val-Gln-Arg-Gly-OH) by the Pocono Rabbit Farm (Canedensis, PA, USA). Its specificity for the MRP sequence was confirmed by demonstration of binding of this peptide to the antibody which was competitively reduced by the presence of PNC-27 that contains this sequence. The synthetic peptide corresponding to the p53 12-26 sequence of PNC-27 lacking the MRP sequence, however, failed to compete with the MRP peptide for binding to this antibody. Goat anti-rabbit (GαR)IgG was obtained from Zymed (South San Francisco, CA, USA). Non-specific control rabbit polyclonal - Isotype Control (ab37415) IgG (R Ig) was obtained from Abcam (Waltham, MA, USA).

Cells. A2058 (human melanoma), MiPaCa-2 (human pancreatic cancer) and AG13145 (normal fibroblast) cells were obtained from the American Type Culture Collection (ATCC; Manassas,VA) and cultured in cDMEM which is DMEM with 10%FBS and Penicillin-Streptomycin (100U/100 μg/mL). BMRPA1 (normal rat pancreatic acinar cells) were produced in our laboratory and cultured as described previously [3].

Assays. Assay for 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). MTT assay was performed using the guidelines of the manufacturer of the assay kit (Promega, Madison, WI). Briefly, 5x103 cells were seeded into the wells of a 96-well Tissue Culture Plates (TCP) and allowed to adhere over night in presence of 100 μl of complete medium (DMEM or RPMI containing 10-15% FBS) when they were treated with the concentrations of peptide (All peptides were sonicated twice for 30 sec.). After 24h incubation with the peptide, 10 μl of MTT substrate mix (Promega) was added to each well and incubated at 37°C. After 4h, 100 μl of MTT stop solution was added to each well and absorbance was read at Optical Density of 570 nm in SpectraMax 90. The values obtained for each set of triplicates were calculated for the mean and the standard error from the mean (SEM). Since the calculations showed that the SEM in virtually all experiments remained below 5%, their reproduction is not visible in the respective graphs.

Lactate Dehydrogenase (LDH) Assay. Approx. 7000 cells grown in 96 well tissue culture plates (TCP) over night were treated in triplicate wells with PNC-27 and PNC-29 peptides at various concentrations as described in Results. All peptides were sonicated twice for 30 sec. After 30min of incubation with the peptides titrated in Dulbecco’s Phosphate-Buffered Saline (PBS) at 37°C, a volume of 50 μl of the supernatant was collected from each well and distributed into a second set of 96-well TCP for the LDH Cytotoxicity Assay (Promega, WI) as the described previously [2,4]. Briefly, 50 μl of the assay buffer – substrate mix were added to each sample well of the second 96-well TCP. After another 30min reaction period, stop solution was added to each well and the samples were read at OD 490nm in an Enzyme-Linked ImmunoSorbent Assay (ELISA) reader. The values obtained for each set of triplicates were calculated for the mean and the standard error from the mean (SEM).

Cell Imaging

Real-Time Live Cell Imaging Using Spinning Disc Confocal Microscopy. Cells (5-10x104) were seeded and grown overnight in 35mm TC dishes containing glass coverslip (#1.5) bottoms (Mattek, MA, USA). The next day, the cells were prepared for three different conditions of real-time studies: Experiments using condition one (I) were performed without preloading the cells with a marker. Experiments using the second condition (II) required loading the cell’s mitochondria or lysosomes with a fluorescent dye prior to microscopy. In the experiments using the third condition (III), competition experiments, cells were pre-incubated with antibody to either HDM-2 or MRP of PNC-27 followed by the successive addition of PI and PNC-27.

Experimental condition I: Cells were washed with PBS and 1ml warm (37°C) PBS was added to the 35 mm TCD which was placed immediately into the specimen carrier of the microscope inside an environmental chamber equipped with an automated temperature setting at 37°C. After bringing the cells into optical focus, a Differential Interference Contrast (DIC) picture was taken to document the untreated cell’s morphology at T=0 min. The nuclear dye Propidium Iodide (PI, 491 nm, Molecular Probes, Invitrogen, CA) was added to a final concentration of 5 M in PBS and cellular staining was recorded at 491 nm in 10-30 sec intervals for up to 15 min when PNC-27 was added at a predetermined concentration. At this time another DIC picture was taken of the culture and the recording of the appearance of any fluorescence of the cells was continued for up to 15 min. Another group of TCDs containing the same cancer cells served as controls and were treated with PNC-29 or were left untreated. Normal primary human AG13145 fibroblasts (5x104 cells/35mm TCD) were used as additional controls and exposed to the same concentration of PNC-27 and their response recorded. For experimental condition I, the spent medium in the dishes was replaced with medium without FBS, and Mitotracker dye (516 nm, Green FM, M7514, Molecular Probes, Invitrogen) was added to a final concentration of 150 nM and incubation of the cells was continued in the 37°C incubator. After 30 min to 45 min, the preloaded cells in TCD were transferred to the microscope specimen carrier, a DIC picture and the staining of the cells mitochondria were recorded and PNC-27 or PNC-29 was added to the medium and recording of the cell’s reactions was begun within one minute and an image was taken every 10 seconds.

Experimental condition II: Different set of cells were also pre-stained with Lysosensor dye (505 nm, L7535 Green DND-189, Molecular Probes, Invitrogen, CA) at a final concentration of 2 M at 37°C. Control TCDs were incubated with an equal amount [21] of dimethyl sulfoxide (DMSO) the solubilizer of the lysosensor dye. After 45 min the cells were washed, transferred into PBS, dishes were placed onto the microscope at which time (T=0) DIC and fluorescent images were recorded prior PNC-27 addition to the cells. During incubation with PNC-27 images were taken every 10 sec for 20 min.

Experimental condition III. Cells were placed into PBS and preincubated with either anti-HDM-2 (RHDM-2) antibody (100 μg/ml) or anti-MRP of PNC-27 antibody (RaMRP IgG) (100 μg/ml). In control TCDs cells were left either untreated or incubated with rabbit non-specific polyclonal antibodies (RM IgG) (100 μg/ml) or with Human Serum Albumin (200 μg/ml) (Sigma, St. Louis, MO, USA). The specific Ab concentrations applied had been determined in earlier experiments using the LDH release assay. After incubation for 45 to 60 min at room temperature, the TCDs were transferred to the microscope, the cells were brought into focus in each of 5 different areas on the cover slip and were observed virtually simultaneously enabling the recording of no less than 100 cells for each time point. The cell’s morphology was recorded using a 40x dry Plan-Apo/0.11-0.23 FN objective (T=0 min). At this time PI was added to a final concentration of 5 mM. After 1 min, pictures of the cells were recorded for 2 min at 30 sec intervals at 491 nm to establish the number of cells in each of the 5 successive areas that would take up PI in absence of any peptide treatment. At this time PNC-27 was added and recording was continued at 561 nm at 10 sec intervals for up to 15 min or 20 min when the final DIC picture was taken before the incubation was stopped.

Immediately after the end of the recording period, the cells in each of the five separate areas that had been recorded for each time point were analyzed for PI uptake and the number of PI positive cells was noted. The number of PI positive cells found within the total number of cells present in the five areas per recorded time point is expressed as percent PI positive (dead) cells. Real-time live cell imaging of cells studied in experimental conditions I and II was performed in a spinning disk Confocal Laser Microscope (Zeiss/Perkin Elmer, CT, USA) using an Inverted Zeiss Axiovert 200 linked to a Hamatsu Orca ER cooled CCD Camera and 491 nm and 561nm lasers. Image acquisition and analysis were done using MetaMorph software (Molecular Devices, Downingtown, PA). Real-time live cell imaging of cells studied in experimental conditions III was performed in a spinning disk Confocal Mercury Lamp IX2UCB Microscope (Olympus IX81) linked to a Hamatsu Orca R2 cooled CCD Camera. Image acquisition and analysis were done with Slidebook 4.2 software.

Electron Microscopy

Immunogold TEM: This procedure was carried out as described previously [10]. Briefly, cells (1X 106) were grown in 6-well TCD overnight; when spent medium was removed, the cells were then washed with PBS and treated at 37°C with PNC-27 in PBS. After 10 min, the cells were washed and fixed in 3% buffered paraformaldehyde supplemented with 0.1% glutaraldehyde for 1.5 h. After extensive washing, quenching of free aldehyde groups with glycine (1%) (30 min) followed by NaBH4 (1%) (30 min) in PBS was performed, followed by washing in PBS. The cells were then incubated overnight at 4°C with anti-MRP antibody (5ug/mL) described above. After removal of unbound Ab and washing with PBS, the cells were incubated for 6 h with 6 nm gold-conjugated GαM F(ab’)2 or 6 nm gold-conjugated RaM F(ab’)2. After extensive washing, cells were post-fixed in 1% glutaraldehyde in cacodylate buffer (0.113M, pH7.2) overnight when the fixed cells were scraped into PBS, centrifuged into a pellet, washed in cacodylate buffer and post-fixed in 1% osmium tetraoxide. Following embedding and sectioning, the thin sections were stained with uranyl acetate and examined in a Zeiss EM10 TEM.

Competition Experiments. In order to further determine whether PNC-27 binds to the amino terminal domain of HDM-2 in the cancer cell membrane, we performed competition experiments in which PNC-27 was incubated with A2058 cells in the presence of different potential competing agents that included the anti-HDM-2 amino-terminal HDM-2 binding antibodies, sc-813, N20 and OP145, and anti-MRP antibody that binds to PNC-27. Cells (7000 cells/well) were seeded in 96-well dishes and were grown overnight. The next day, the spent medium was removed and triplicate sets of cells were incubated with serial dilutions of anti-H(M)DM2 N20 and OP145 Ab. The dilutions were prepared in PBS spanning concentrations from 0.1 mg/mL to 100 mg/mL. Another set of triplicate cells was incubated with serial dilutions of non-specific Rabbit IgG (R Ig) and human serum albumin (Sigma, St. Louis, MO, USA) in PBS, respectively, with concentrations from 0.1 mg/mL to 100 mg/mL. An additional triplicate set of cells was incubated in PBS only. Cells were incubated for 30min at 37°C when PNC-27 was added to a final concentration of 75 μg/ml. The cells in a total volume of 0.1ml of PBS-PNC-27-Competitor combination were incubated at 37°C water bath for another 30 min when the LDH assay on cell supernatants and lysates was performed as described above.

Results

Membrane-Expressed HDM-2 binds to PNC-27 at its N-terminal sequence containing the HDM-2 binding domain of p53. In previous studies [9], we found that MCF-10-2A untransformed human breast cells were not sensitive to PNC-27, but when transfected with a plasmid expressing HDM-2 with a membrane localization signal resulting in expression of HDM-2 in the cell membrane, they were killed by this peptide. In contrast, when these cells were transfected with a plasmid that expressed membrane-localized truncated HDM-2 lacking the p53 binding domain residues 1-109, they were not sensitive to PNC-27. These results suggest that PNC-27 binds to the p53 binding domain of HDM-2, although absence of this domain could inactivate another downstream binding site for the peptide.

Therefore, to further to determine if the 1-109 segment of HDM-2 is the site of binding of PNC-27, we employed an anti-HDM-2 antibody, whose specificity was for this domain of HDM-2, as a potential competitive inhibitor of PNC-27. For this purpose, we incubated A2058 cells with PNC-27, alone or in the presence of two different antibodies against the HDM-2 binding domain (residues 1-109): N-20 (residues 1-102) and OP145 (residues 19-50). We also tested a rabbit polyclonal antibody directed against the MRP sequence of PNC-27. In addition, as a control, we incubated A2058 cells with non-specific rabbit IgG followed by treatment with PNC-27. The expectation was that the specific antibodies would block the cytotoxic effects of PNC-27, while the non-specific IgG would not interfere with the binding of PNC-27 to membrane-expressed HDM-2.

We evaluated cytotoxicity in two ways: by incubation of treated cells with propidium iodide (PI) that stains only the nuclei of membrane-damaged cell and by measuring LDH release, indicative of membrane damage [3-9]. A2058 cells that were pre-incubated with N-20 anti-HDM-2 (Figure 1A) or with anti-MRP (Figure 1B) had dramatically reduced PI nuclear staining compared to cells incubated with PNC-27 alone (Figure 1C). Pre-incubation of cells with non-specific rabbit IgG (Figure 1D) did not block PI nuclear staining, supporting the conclusion that PNC-27 interacts with HDM-2 in its 1-109 amino terminal domain. In distinction, the other site-specific antibody, OP145, that binds to HDM-2 residues 19-50, completely blocked PI nuclear staining (Figure 1E).

Figure 1.

Antibodies against the p53 binding domain of HDM-2 specifically block the cytotoxic effects of PNC-27. (A) Human melanoma A2058 cells were pre-incubated with an HDM-2-specific polyclonal Ab (sc-813; N20) directed against the p53 binding site of HDM-2 (residues 1-102) followed by treatment with PNC-27 in the presence of PI. (B) PNC-27 pre-incubated with PNC-27-specific rabbit polyclonal Ab (anti-MRP) then incubated with A2058 cells in presence of PI. (C) PNC-27 was incubated with A2058 cells in presence of PI and in the absence of any blocking Ab. (D) A2058 cells pre-incubated with non-specific rabbit IgG as a control followed by PNC-27 treatment in presence of PI. (E) Human melanoma A2058 cells were pre-incubated for 20 min with another anti-H(M)DM-2 antibody, OP145, directed against the p53 binding site of HDM-2 (residues 19-50) followed by treatment with PNC-27 in presence of PI. Left panel shows the DIC image of the cells after each treatment with intact cells in A and B and destroyed cells in C and D. Right panel shows representative confocal images of dead cells stained with PI from eight sections totaling 100 cells.

Figure 2 shows the results of incubation of different agents on the release of LDH from A2058 cells incubated with PNC-27. Cells pre-incubated with OP145 at increasing titers and treated with PNC-27 (50 μg/mL) (red circles) show a steep decline at 0.5μg/100μl OP145 antibody titer. The efficiency of PNC-27 cytotoxicity reaches >90% at a concentration of OP145 at 1.0μg/0.1ml incubation medium (or 10 μg/ml). In contrast, even at the highest concentrations of 100 μg/ml, neither BSA (green circles) nor non-specific rabbit polyclonal antibodies (blue circles) affected PNC-27 mediated cytotoxicity.

Figure 2.

PNC-27-mediated cytotoxicity is dependent on the availability of plasma membrane HDM-2 sites. Human melanoma A2058 cells were pre-incubated with increasing concentrations of αHDM2 mAb OP145 (Blue triangles), non-specific rabbit polyclonal IgG (Green) or BSA (Red) as controls, followed 30 min later by treatment with PNC-27 (50 μg/ml) for 30 min. LDH assay was then performed to measure the release of the enzyme from leaky and dead cells into the spent cell supernatant to evaluate PNC-27 cytotoxicity. Each point represents average cytotoxicity ±SEM of 3 experiments. Maximum SEM was 5 percent.

PNC-27 Induces Mitochondrial Damage in Cancer Cells. TEM analysis of cells treated with PNC-27. MIAPaCa-2 cells were treated with PNC-27 (150 μg/ml) or negative control peptide, PNC-29 (150 μg/ml) for 5 min followed by extensive washing to remove unbound peptide. Cells were fixed for TEM as described in the Methods section. Untreated cells were processed in the same fashion as controls. Figure 3A shows a typical untreated MIA-PaCa-2 cell showing intact mitochondria seen as the intracellular ellipsoidal dark-staining bodies [18]. However, the same cells treated with PNC-27 showed large white gaps in their mitochondria as shown in Figure 3B (blue arrow in Figure 3B), representing degenerating mitochondrial bodies [18]. Higher magnification studies revealed the presence of multiple pores in the membranes of MIA-PaCa-2 cells treated with PNC-27 [18]. Figure 3C shows that the mitochondria of MIA-PaCa-2 cells incubated with the negative control peptide, PNC-29, appear normal as in the untreated cells shown in Figure 3A. Figure 3D shows a typical normal human fibroblast from AG13145 cell line treated with PNC-27. The mitochondria appear to be intact which is compatible with our previous finding that PNC-27 does not affect these cells [9].

Figure 3.

TEM of MIA-PaCa-2 cancer cells treated with PNC-27 shows disrupted plasma membrane and mitochondria. (A) shows an untreated MIA-PaCa-2 cancer cell in which mitochondria, that are the intracellular ellipsoidal bodies with dark staining, are all intact. (B) shows a MIAPaCa-2 cancer cell subjected to treatment with PNC-27. The mitochondria are seen to have large white areas such as labeled with the blue arrow representing degenerating mitochondria. Most of the mitochondria in this section show these vacated spaces. (C) shows a typical MIA-PaCa-2 cell treated with the negative control peptide, PNC-29. All the mitochondria appear to be intact with no vacated spaces. (D) shows a typical normal human fibroblast from the AG13145 cell line treated with PNC-27. The mitochondria of these cells were found to be intact. Frames A and B are reproduced from ref. 18 by permission.

Disruption of mitochondria of cancer cells treated with PNC-27. Mitotracker and lysosensor are two polyaromatic heterocyclic compound acidic dyes demonstrated to have a high specificity for binding to intact mitochondria and lysosomes, respectively [14]. We assessed the viability of mitochondria in cells with live cell imaging of pre-stained mitochondria using spinning disc confocal fluorescence microscopy to evaluate the time at which mitochondria are disrupted. We incubated MIA PaCa-2 cells in glass bottom TCD with MitoTracker (150 nM, 45 min, 37oC), which becomes detectable and defines the morphology of mitochondria once the dye accumulates in the organelles. The cell-permeant MitoTracker probes contain a mildly thiol-reactive chloromethyl moiety for labeling mitochondria. The dye passively diffuses across the plasma membrane and accumulates in active mitochondria. MitoTracker is readily sequestered by functioning mitochondria and is easily washed from cells. The dye is also rapidly lost from the mitochondria once the organelles experience a loss in membrane potential or are damaged otherwise [14].

Figure 4A shows untreated MIAPaCa-2 cells indicating healthy, active, and mitochondria stained with Mitotracker. These cells were then treated with control peptide PNC-29 (150 μg/ml) for 15min. Figure 4B shows the last frame of the 15 min PNC-29 treatment time course experiment where mitochondria remain stained and unchanged morphologically. In contrast, when these cells were treated with PNC-27 (150 μg/ml) for another 15 min, the mitochondria were observed to swell within 5 min, become round by 6 min, and then rupture within another 1-2 min (7-9 min after start of peptide treatment). Figure 4C shows the last frame of a 15 min live imaging time course during which the cells were recorded in 10 sec intervals.

Figure 4.

PNC-27 causes selective damage to mitochondria but not lysosomes. (A-C) Human pancreatic cancer MIAPaCa-2 cells were pre-stained with Mitotracker dye followed by either no treatment (A), PNC-29 (B) or PNC-27 (C) treatment. The disruption of mitochondria and leakage of the dye occurs only in PNC-27-treated cells (Figure 4C) where it is clear that the punctate green fluorescence intensity (yellow arrow) is significantly lower than in the corresponding cells (blue arrows) either untreated or treated with negative control PNC-29 peptide. The white arrow shown in Figure 4C shows low intensity, diffuse fluorescence due to diffusion of mitotracker dye into the cytoplasm from damaged mitochondria not seen in either Figures 4A or 4B. (D-G) Control normal human fibroblasts (AG13145 Cells) preincubated with Mitotracker and PI, followed by PNC-27 treatment. (D) Untreated AG13145 cells incubated with Mitotracker dye showing high levels of punctate green fluorescence suggesting mitochondrial integrity. (E) Same cells in 4D incubated with propidium iodide (PI) showing no entry of this dye into the cell suggesting retention of membrane integrity and no cell death. (F) AG13145 cells treated for 30 min with PNC-27 showing strong punctate green fluorescence suggesting mitochondrial integrity that appears almost identical to that found for the untreated cells in 4D. (G) AG13145 cells treated with PNC-27 for 30 min in the presence of PI showing no entry into cells suggesting retention of membrane integrity and no cell death. (H-K) MIAPaCa-2 cells pre-stained with Lysosensor dye followed by PNC-27 treatment. (H) Confocal image of untreated MIA-PaCa-2 cells incubated with lysosensor dye showing intense punctate red fluorescence indicating intact lysosomes. (I) Differential interference contrast (DIC) microscopic image of untreated MIA-PaCa-2 cells showing live, intact cells. (J) Confocal image of MIA-PaCa-2 cells treated with PNC-27 for 30min and incubated with lysosensor dye showing intense punctate red fluorescence indicating intact lysosomes even after PNC-27 treatment. (K) Differential interference contrast (DIC) microscopic image of MIA-PaCa-2 cells incubated with PNC-27 for 30 min showing dead and damaged cells.

The cells can be seen to leak the Mitotracker dye into the cytoplasm and the medium, as revealed by a loss of punctate intense green fluorescence and diffusion of the green fluorescence into the cytoplasm and even outside the cell membrane. Corresponding studies (not shown) of the cells in Figure 4A-C that were incubated with propidium iodide (PI) showed no entry into untreated cells (Figure 4A) or cells treated with negative control peptide PNC-29 (Figure 4B) but strong entry into cells treated with PNC-27 (Figure 4C).

The time course experiments allowed recording of the exact timing of the mitochondrial disruption at 5 min post PNC-27 treatment. In contrast, when Ag13145 primary human fibroblasts that had Mitotracker pre-stained mitochondria were treated with PNC-27 in the presence of PI, no disruption of mitochondria and no Mitotracker leakage was observed (compare Figure 4D&F). Furthermore, no PI staining of the fibroblast nuclei was observed, indicating that these cells were intact and healthy (Figure 4E&G).

PNC-27 disruption of cellular lipid bilayers is selective. Since PNC-27 induces pore formation in the plasma membrane and also disrupts mitochondrial integrity, we asked whether the effect of PNC-27 on mitochondria is specific or whether other intracellular organelles containing lipid bilayer membranes are affected. For this purpose, we chose lysosomes since the dye, Lysosensor (described in the Materials and Methods section), stains these organelles selectively.

Figure 4H shows untreated MIAPaCa-2 cells that had been stained with Lysosensor. Lysosomes appear as round and densely stained red dots throughout the cells. Moreover, the DIC image of the cells shows that they are morphologically intact (Figure 4I). These cells were then treated with PNC-27 (150 μg/ml) for up to 30min. As shown in Figure 4J, the lysosomes stay intact, the concentrated Lysosensor is engulfed in lysosomes, and no leakage of the dye was observed. However, as shown in Figure 4K, the cells are morphologically disrupted, which indicates their death. These observations show that PNC-27 is selective in disruption of lipid bi-layers as it causes plasma membrane and mitochondrial membrane disruption most likely due to its ability to interact with a receptor likely HDM-2. This suggests that the HDM-2 that we have found in the plasma membrane may also be expressed in the mitochondrial membrane. The absence of an effect of PNC-27 on lysosomal membranes suggests that HDM-2 is not expressed in the membranes of these organelles.

To evaluate the hypothesis that PNC-27 causes pore formation in the mitochondria, we performed immunogold TEM analysis of mitochondria of cancer cells that had been treated with PNC-27 and evaluated mitochondrial membranes for PNC-27 staining. MIA PaCa-2 cells treated with PNC-27 (150 μg/ml) for 30 min and stained with anti-PNC-27 Ab (RαMRP IgG) were subjected to immunogold TEM preparation as described in the Materials and Methods section. Multiple sections were analyzed. A typical section is shown in Figure 5, showing 6 nm gold particles representing PNC-27 on the mitochondrial but not lysosomal membranes which represents staining for PNC-27. This finding indicates the selective intracellular localization of PNC-27 on the mitochondrial membranes of treated cells. This finding may help explain the adverse effect of PNC-27 treatment on the integrity of mitochondrial membrane as described above in the leakage of Mitotracker dye from mitochondria and their destruction in cancer cells treated with PNC-27.

Figure 5.

PNC-27 localizes to the mitochondrial membrane of MIAPaCa-2 cells. Immunogold TEM of human pancreatic cancer MIAPaCa-2 cells treated with PNC-27 for 15-20 min and stained for with anti-MRP primary antibody and 6 nm gold-conjugated secondary antibody. Disrupted mitochondria observed by TEM are shown. Numerous small black dot-like gold particles identifying mitochondrial-bound PNC-27 are seen to occur in multiple mitochondria. The orange arrow in the highest magnification view shown in the right-most frame points to mitochondrial surface-bound PNC-27 which could be seen on multiple other high magnification views.

Discussion

In this report, we have provided evidence supporting previous results that PNC-27 co-localizes with HDM-2 in the plasma membrane of cancer cells and that PNC-27 binds to the amino terminal domain of HDM-2 [9-10]. In addition, we provide new evidence that PNC-27 causes disruption of the mitochondria of cancer cells by binding directly to the mitochondrial membrane. The latter result bears on recent interest in the metabolic aspects of cancer biology, notably in the role of the Warburg effect, that many cancer cells show increased ratio of glycolysis to respiration than normal cells even in the presence of oxygen [15]. A large literature exists supporting the role of metabolic control in cancer physiology and treatment (Reviews: [16,17] The data suggest the possibility that, in addition to, or as part of its effects on cell lysis, the PNC-27/HDM-2 system exerts effects on the control of energy metabolism. This idea is supported by our previous demonstration that ketone bodies, which inhibit glycolysis, enhance the effects of PNC-27 in cell culture [18].

Binding of PNC-27 to HDM-2. In the competition experiments reported here, we used an antibody, OP145, that binds to HDM-2 in its amino terminal domain and therefore the antibody would compete with PNC-27 for binding to this domain. Figure 1 showed that this antibody at a titer of 1μg/mL blocks PNC-27-induced tumor cell necrosis. Negative controls, rabbit and mouse immune sera, even at high titers, did not interfere with the action of this peptide. As a positive control, an anti-MRP antibody directed against the leader sequence strongly blocked PNC-27 from inducing tumor cell necrosis of the A2058 cells. The results confirm that PNC-27 binds to HDM-2 in its amino terminal domain.

PNC-27 causes disruption of the mitochondria of cancer cells and binds directly to mitochondrial membranes. Our previous TEM studies of cancer cells incubated with PNC-27 revealed not only pores in the cell membranes but also significant disruption of mitochondria which appeared in deformed shapes within 5 min of PNC-27 treatment [3,5]. Other cell organelles such as lysosomes and Golgi bodies did not appear damaged. The current results support this conclusion. It is instructive to compare the intracellular effects of PNC-27 with those of the lytic protein, streptolysin O (SLO), which also induces transmembrane pore formation but is much less selective than PNC-27 since it causes lysis of normal cells. Like PNC-27, SLO damages mitochondria [19]. However, in contrast to PNC-27, SLO also causes disruption of lysosomes and Golgi bodies [20,21]. Thus PNC-27 is significantly more selective in its intracellular effects.

Acknowledgements

This work was supported in part by the Nutrition and Metabolism Society.

References

1. 1.↵
2. 2.↵
3. 3.↵
4. 4.↵
5. 5.↵
6. 6.↵
7. 7.
8. 8.↵
9. 9.↵
10. 10.↵
11. 11.↵
12. 12.↵
13. 13.↵
14. 14.↵
15. 15.↵
16. 16.↵
17. 17.↵
18. 18.↵
19. 19.↵
20. 20.↵
21. 21.↵