A mitochondrial targeted fusion peptide exhibits remarkable cytotoxicity Free

Molecularly targeted anticancer agents are rapidly emerging; however, many cancers ultimately show drug resistance. Consequently, combination therapies are common places to improve treatment outcomes. The essential role of mitochondria to mediate programmed cell death has led to the recent attention towards antitumor agents that target mitochondrial functions (1, 2).

Certain cationic amphipathic peptide sequences are able to interact with the negatively charged prokaryotic plasma membrane and cause lipid matrix distortion (3, 4). These peptides are normally nontoxic to eukaryotic cells, as the mammalian plasma membrane is mainly composed of zwitterionic phospholipids. Recently, an L-configuration amphipathic antimicrobial peptide sequence, KLAKLAKKLAKLAK (KLA), was reported as an antibacterial agent (5). This nontoxic peptide can be further modified by fusion of a transduction domain to enhance its cellular uptake (6). Once inside the cell, KLA is cytotoxic and was shown to disrupt the negatively charged mitochondrial membrane. Subsequent apoptosis was triggered via the release of cytochrome c (7, 8). Other studies have also shown that the attachment of a tumor homing motif (RGD-4C) to KLA reduced the growth of breast carcinoma in nude mice (9). Recently, a bifunctional peptide has been designed, combining an anti-HER-2 peptide and KLA, in order to simultaneously perturb the growth factor receptor signaling and the mitochondrial activity (10). This peptide showed selective internalization into HER-2-overexpressing human breast cancer cells both in vitro and in vivo.

A major obstacle for therapeutic peptides is that amide bonds are subject to protease digestions. The metabolic stability of different peptides have been studied in detail (11–13) and their half-lives in vitro vary from minutes to days, depending on the peptide sequences and the cells studied. In the present study, we have created a fusion of the D-forms of KLA (kla) and an arginine-rich cell-penetrating peptide (D-hepta-arginine) to increase cellular uptake. Two unstructured glycine residues were inserted as a spacer. The advantage of using hepta-arginine (r7) as a delivery vector is that the D-configuration has better membrane-crossing properties (14), and at the same time, is more likely to resist protease digestion. The resulting conjugate, r7-kla (Fig. 1A; Table 1), shows enhanced kinetic of cellular uptake, and efficient apoptosis in cells. In vivo evaluation also suggests that r7-kla could lead to a new approach for anticancer drug design.

All solvents were purchased from Fisher Scientific (Fair Lawn, NJ). 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate, N-hydroxybenzotriazole, N,N-diisopropylethylethylamine, and N-methylpyrrolidone were the products of Applied Biosystems (Foster City, CA). AnaSpec (San Jose, CA) supplied all the amino acids for peptide synthesis. Rink amide resin was from Novabiochem (San Diego, CA). Annexin V FITC conjugate (AnxV-FITC) solution, anisole, cyclophosphamide monohydrate, cis-diammineplatinum (II) dichloride (cisplatin), doxorubicin, ethanedithiol, methotrexate, piperidine, propidium iodide solution (PI), thioanisole, and trifluoroacetic acid were purchased from Sigma-Aldrich (St. Louis, MO). Hoechst 33342 and paclitaxel were purchased from Molecular Probes (Eugene, OR). DMEM and RPMI 1640 were purchased from Mediatech (Herndon, VA).

Peptide synthesis was done on an automated peptide synthesizer (FastMoc, 0.1 mmol, ABI 433A; Applied Biosystems) employing the traditional N_α-Fmoc strategy. A mixture of 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (10 equivalents), N-hydroxybenzotriazole (10 equivalents), and N,N-diisopropylethylethylamine (20 equivalents) in N-methylpyrrolidone (15 mL) was used as the coupling cocktail. All amino acid building blocks (10 equivalents) were attached to the Rink amide resin (162 mg, 0.1 mmol) by stepwise elongation. The N_α-Fmoc protecting groups were removed by 20% (v/v) of piperidine in N-methylpyrrolidone (15 mL). After synthesis, the peptides were cleaved in a mixture of trifluoroacetic acid (1 mL), thioanisole (250 μL), ethanedithiol (150 μL), and anisole (100 μL) for 3 hours at room temperature. The resins were filtered out and the filtrates were collected. The peptides were then precipitated by methyl-tert-butyl ether (15 mL) at 4°C, dried, and further purified with reversed phase high-pressure liquid chromatography (Ranin, Worburn, MA). The purified fractions were lyophilized, yielding white foams with >95% homogeneity. All final products were characterized by analytic high-pressure liquid chromatography and MALDI-TOF mass spectrometry (Tufts Protein Chemistry Facility, Boston, MA; Table 1).

All tumor cells lines [HT-1080, HCT-116, HT-29, MDA-MB-468, HeLa, PC-3, MIA PaCa-2, and LL/2 (LLC1)] were obtained from American Type Culture Collection (Rockville, MD) and were cultured in 5% CO₂ at 37°C. HT-1080 fibrosarcoma cells were incubated in DMEM supplemented with 10% (v/v) fetal bovine serum, glucose (4.5 g/L), l-glutamine (4 mmol/L), nonessential amino acids (0.1 mmol/L), sodium bicarbonate (1.5 g/L), penicillin (50,000 units/L), and streptomycin (0.05 g/L). Jurkat T lymphocytes (clone E6-1) were cultured in RPMI 1640 supplemented with 10% (v/v) fetal bovine serum and l-glutamine (4 mmol/L) in the absence of antibiotics.

The cytotoxicity of peptides and drugs was monitored by MTS assay (CellTiter 96 AQ_ueous; Promega, Madison, WI). HT-1080 cells (5,000/well) were seeded on a 96-well cell culture cluster plate (Costar, Corning, NY) for 12 hours. Cells were incubated with different concentrations (ranging from 0 to 500 mmol/L) of peptides, cyclophosphamide, cisplatin, doxorubicin, methotrexate, and paclitaxel in culture medium (100 μL) for different time points (1, 4, 8, 16, and 24 hours) at 37°C in 5% CO₂. MTS solution (20 μL) was then added to individual wells and allowed to further incubate for 1 hour at 37°C. Samples were prepared in triplicate and the absorbance was measured at 490 nm. IC₅₀ values were calculated using Prism (version 4.0a, GraphPad Software, Inc., San Diego, CA). Paired t test analysis (one-tailed) was done by comparing the data (at specific concentrations) with the cells treated with kla.

HT-1080 cells (10,000/well) were seeded on microscope cover glasses (Fisher Scientific) in 24-well cell culture cluster plates (Costar) for 24 hours at 37°C in 5% CO₂. Cells were then treated with peptide r7-kla (10 μmol/L) in culture medium (1 mL) for 2 minutes at room temperature and washed once with PBS (1 mL). For staining, a solution of Hoechst 33342 (1 μmol/L), AnxV-FITC (125 ng), and PI (500 ng) in HEPES/NaOH binding buffer (10 mmol/L, pH 7.5; 1 mL) containing NaCl (140 mmol/L) and CaCl₂ (2.5 mmol/L) was added to the samples. The cells were incubated for 10 minutes at 37°C and washed once with culture medium (1 mL). For mitochondrial staining, DePsipher kit (Trevigen, Gaithersburg, MD) was used according to the manufacturer's instruction. Cover glasses were then transferred to microscope slides (Superfrost, Pittsburgh, PA) and observed under the epifluorescence microscope (Nikon Eclipse 80 I, Melville, NY). Images were recorded on a CCD camera (Photometrics Cascade-512B, Tucson, AZ) interfaced with a computer. Data was analyzed by IPLab software (Scanalytics, Inc., Fairfax, VA). Images from the same excitation and emission channels were adjusted to the same window settings.

Jurkat cells were concentrated by centrifugation for 5 minutes at 1,000 rpm to 2 × 10⁶ cells/mL. Cells (200 μL) were then incubated in culture medium containing r7-kla (6 μmol/L) for different time points (in minutes), washed with RPMI 1640 (10 mL) and further centrifuged for 5 minutes. Cell pellets were resuspended in 400 μL of HEPES/NaOH buffer (10 mmol/L, pH 7.5) containing NaCl (140 mmol/L) and CaCl₂ (2.5 mmol/L). AnxV-FITC (2.5 μL, 50 μg/mL) and PI (4 μL, 100 μg/mL) were added to the samples and allowed to incubate for 10 minutes at room temperature. Control experiments were done without r7-kla incubation. Samples were analyzed by flow cytometry using a Becton Dickinson FACSCalibur (San Jose, CA), using the FL-1 channel to detect the AnxV-FITC and FL-2 channel for PI.

All animal studies were approved by the institutional animal care committee. Female athymic nude mice (6–7 weeks old, 21–23 g) were purchased from Charles River (Wilmington, MA). All animals were quarantined for 1 week. Tumors were implanted s.c. under i.p. anesthesia [ketamine, 90 mg/kg (BenVenue, Bedford, OH) and xylazine, 10 mg/kg (Phoenix, St. Joseph, MO)]. Cells (3 × 10⁷ cells/mL) in DMEM (50 μL) were injected into both hind legs of each mouse. When the tumor sizes were ∼5 mm in diameter, peptide r7-kla (100 μg) in sterilized normal saline solution (50 μL) was injected into the tumor sites. For the control tumors, only saline was administered. Sixteen hours postinjection, the animals were sacrificed and the tumor tissues were excised, paraffin-embedded, and sectioned into 7 μm and stained with H&E (Sigma-Aldrich) to reveal the histologic architecture. Terminal nucleotidyl transferase–mediated nick end labeling staining (Chemicon, Temecula, CA) was done according to the manufacturer's instructions. Primary rabbit polyclonal antibody against human cleaved caspase-3 (Asp175; Cell Signaling, Beverly, MA) was used for the detection of activated caspase-3. Staining was visualized by using 3,3′-diaminobenzidine chromogenic substrate system from the EmVision kits (DAKO, Carpinteria, CA).

Cell-penetrating peptides have been widely used for the delivery of biologically active compounds in in vitro and in vivo disease models (15). Among different cell-penetrating peptides, arginine-rich sequences are known to have high cellular uptake (16, 17), and their internalization mechanism was recently proposed to involve binding to the cell surface heparan sulfate, and subsequently by endocytosis (18). To determine the cytotoxic effects of r7-kla, we initially screened eight different tumor cell lines for proliferation studies by MTS assay. Cell survival rates were determined 24 hours following peptide treatments. We observed that the killing effects of r7-kla to different tumors were nonselective (Table 2). The IC₅₀ values are within the range of 3 to 25 μmol/L.

HT1080 fibrosarcoma was then chosen to confirm the toxicity of r7-kla. This cell line could develop multiple drug resistance to doxorubin and paclitaxel (19). The determined IC₅₀ value of r7-kla was 3.54 μmol/L (Fig. 1B), whereas individual kla and r7 sequences both had IC₅₀ values of >100 μmol/L. Equal amounts of kla and r7 were separately added to the cells, and the obtained IC₅₀ value remained at >100 μmol/L, indicating that the cytotoxiciy of r7-kla was not caused by the synergic effect of mitochondrial-disrupting and cell-penetrating domains.

The D-form r7-kla has a slightly lower IC₅₀ value than its L-counterpart, R7-KLA (IC₅₀ = 7.21 μmol/L). The differences in the IC₅₀ values obtained for r7-kla and R7-KLA could be attributed to (a) the higher cellular uptake of r7 when compared with the R7 (14), (b) the degradation of R7-KLA by protease digestions, and (c) the difference of mitochondrial membrane disruption properties. Further detailed experiments are required to confirm these hypotheses.

We next studied the kinetics of cell deaths over 24 hours. As expected, r7-kla showed a very fast kinetic of cell killing, especially in the first 8 hours of drug treatment (Table 3). In contrast, other clinically used anticancer agents exhibited cytotoxic effects after 8 hours of incubation and at higher doses. The IC₅₀ values of all screened compounds at 24 hours were: r7-kla < doxorubicin < paclitaxel < methotrexate. Other anticancer drugs such as cisplatin and cyclophosphamide were unable to kill HT1080 cells at the tested conditions.

Next, we did a series of experiments to elucidate the mechanism of cell death. After incubation with r7-kla (10 μmol/L), as shown in Fig. 2A, the treated cells detached from the slide within 5 minutes and showed typical signs of apoptosis. These cells shrunk in size and had altered plasma membranes including blobbing. Their DNA showed “comet” patterns with Hoechst staining. To distinguish early apoptosis from late apoptosis/necrosis, r7-kla-treated and nontreated cells were costained with AnxV-FITC and PI. Without r7-kla treatment, cells were predominantly AnxV negative and PI negative (AnxV⁻/PI⁻). Both early and late apoptosis was induced by incubation with r7-kla (Fig. 2A). The single positive (AnxV⁺/PI⁻) cells were identified as early apoptotic, because they exposed phosphatidylserine on the outer leaflet of the plasma membrane for AnxV binding. AnxV⁺/PI⁺ cells were considered late apoptotic/necrotic because their membrane was permeable to PI.

The mitochondrial-disruptive ability of r7-kla was further evaluated using a commercially available cationic dye, DePsipher, which is commonly used to monitor mitochondrial membrane potential losses during apoptosis (20). This dye enters cells readily and gives a bright red fluorescence within healthy mitochondria, whereas its fluorescence turns green when the mitochondrial membrane is damaged. After 5 minutes of r7-kla incubation, unhealthy HT-1080 cells showed primarily green fluorescence, and the disappearance of red aggregates suggested that DePsipher no longer accumulated onto the mitochondrial membrane in which the electrochemical gradient had collapsed (Fig. 2B).

As mentioned above, r7-kla exhibited its full cytotoxicity in cell culture in <1 hour and few differences were seen at later time points. To further study the kinetics of cell death, flow cytometry (fluorescence-activated cell sorting) was used to collect real-time information in Jurkat T lymphocyte. After staining with AnxV-FITC and PI, three distinct populations (Fig. 3A) were identified: double negative cells (AnxV⁻/PI⁻), single positive cells (AnxV⁺/PI⁻), and double positive cells (AnxV⁺/PI⁺), corresponding to viable, early apoptotic, and late apoptotic/necrotic cells, respectively. Prior to r7-kla addition, ∼93% of cells were healthy. Sixteen minutes after incubation with r7-kla (8 μmol/L), the number of viable cells had decreased to 6%. Conversely, the apoptotic and necrotic populations had increased to 48% and 42%, respectively. After 30 minutes, the majority of cells had become necrotic. Dose-dependent cell death was studied by using the same method (Fig. 3B). After 20 minutes of incubation with 2 μmol/L r7-kla, >70% of cells were viable. On the other hand, >70% of cells had turned necrotic with 12 μmol/L r7-kla. Therefore, 8 μmol/L r7-kla was chosen to follow the cell kinetics during 45 minutes. After 2 minutes of r7-kla treatment, most cells were nonviable and the apoptotic distribution had increased to 63% (Fig. 3C). At longer times, the apoptotic population started declining and reached a minimum of 5% at 45 minutes. During this experimental time frame, the necrotic population increased inversely with the kinetic profile of apoptotic cells. At a r7-kla concentration of 8 μmol/L, the transition from the early apoptotic to the necrotic state was completed within 30 minutes of peptide incubation.

We next tested whether the new fusion peptide could be used in vivo to treat human HT1080 fibrosarcoma xenografts. Tumors were implanted into both hind legs of each mouse (n = 4). When tumors reached ∼5 mm in size, a single dose of r7-kla (100 μg) or saline was injected into each tumor site. The mice were then sacrificed and tumors were removed for immunohistologic studies (Fig. 4). The treated tumor showed significant tissue loss and extensive necrotic areas. Moreover, ∼80% of the cell residuals at the injection sites were stained positive to terminal nucleotidyl transferase–mediated nick end labeling and caspase-3.

Our data confirm that the cytotoxicity of KLA can be significantly enhanced through cellular internalization. Unlike the traditional lock-and-key approach, the antitumoral effect of KLA is based on physical disruption of the membrane bilayer and is therefore expected to be less susceptible to multiple drug resistance. With the D-configuration of KLA (kla) attached to a hepta-arginine cell-penetrating domain (r7), the synthetic peptide, r7-kla, showed better cytotoxicity against the HT1080 human fibrosarcoma cell line compared with the clinically used antitumor agents evaluated in the current study. The peptide induces apoptosis both in vitro and in vivo. In addition, we showed the distinct cytotoxic property of r7-kla. The instant toxicity of r7-kla could be an advantage, especially by early inhibition of cell proliferation. However, the fusion of cell-penetrating domain r7 to kla had compromised the killing kinetic with cellular selectivity. Recently, we have encapsulated this sequence into a peptide-based biomaterial system for functional drug delivery (21). Overall, we believe that the encapsulation of this fusion peptide into nanoparticles containing specific targeting moieties may be useful for the systemic treatment of tumors.

The authors thank Dr. Mikael Pittet for FACS analysis.