Antitumor effect of a transducible fusogenic peptide releasing multiple proapoptotic peptides by caspase-3

Recently, several types of fusogenic peptides that are able to treat tumors efficiently have been generated. For example, a “homing” peptide was fused to a programmed cell death–inducing peptide to selectively target angiogenic endothelial cells (1). Additionally, an apoptosis-inducing peptide was fused to a protein transduction domain to ensure high transduction efficiency (2, 3), and a peptide targeting the specific protein kinase, casein kinase, which was identified by phage display screening, was fused to a cell penetrating peptide from the HIV-TAT protein to induce tumor cell apoptosis (4). Functional peptides used for the development of fusogenic antitumor peptides are generally classified into three functional groups: “targeting peptides” (5, 6), “transduction peptides” (7–11), and “killing peptides” (1, 12, 13). Any combination of these three groups can be designed, and it has been suggested that all of these peptides can be used to treat solid tumors or as an adjuvant therapy in conjunction with conventional therapies. To develop a more potent and efficient antitumor peptide, we designed TK3, which is a peptide composed of multiple proapoptotic peptides that are each separated by the caspase-3 substrate, DEVD, as well as a protein transduction domain to enable rapid and efficient transduction into cells. Therefore, when the peptide transduces into cells via the protein transduction domain, it activates caspase-3, which in turn cleaves the DEVD of the peptide, causing the release of more proapoptotic peptides. As expected, TK3 was more potent in terms of inhibition of tumor cell proliferation in vitro and tumor growth in vivo than TK1, which has only a single apoptosis-inducing peptide. Therefore, in this study, we described the development of a more potent antitumor peptide that functions based on a system that amplifies the caspase-3 activation signal.

The following oligonucleotides were used to prepare the peptides. 5′-CTAGCGGAGGTAAGCTAGCGAAGCTAGCGAAGAAGCTAGCGAAGCTAGCGAAGA-3′ and 5′-AGCTTCTTCGCTAGCTTCGCTAGCTTCTTCGCTAGCTTCGCTAGCTTACCTCCG-3′ were used to prepare (KLAKLAK)₂. Mutant (KLAKLAK)₂ DLSLARLATARLAI was prepared using 5′-GGCTAGCGGAGGTGATCTGTCACTGGCACGTCTGGCACTAGCACGTCTGGCAATCA-3′ and 5′-AGCTTGATTGCCAGACGTGCTAGTGCCAGACGTGCCAGTGACAGATCACCTCCGCTAGCCTGCA-3′, and tandem repeat (KLAKLAK)₂-DEVD-GG was prepared using 5′-GATCTAAGCTCGCGAAGCTCGCGAAGAAGCTCGCGAAGCTCGCGAAGGATGAAGTAGATGGAGGTGGATCCA-3′ and 5′-AGCTTGGATCCACCTCCATCTACTTCATCCTTCGCGAGCTTCGCGAGCTTCTTCGCGAGCTTCGCGAGCTTA-3′. Restriction enzyme sites that were incorporated to facilitate ligation are italicized.

pTK was constructed in the following manner to express the basic domain (amino acids 49-57) of HIV-1-TAT as a fusion protein with KLAKLAKKLAKLAK [(KLAKLAK)₂]. First, a double-stranded oligonucleotide encoding the nine amino acids of TAT, RKKRRQRRR, was cloned into pET30a(+) vector (Novagen) within the NdeI and EcoRV sites to generate pTAT. Next, pcDNA-(KLAKLAK)₂-Myc was generated by inserting a double-stranded oligonucleotide encoding GG-(KLAKLAK)₂ into pcDNA3.1/myc-His(-)A (Invitrogen) within the NheI and HindIII sites. The (KLAKLAK)₂-Myc sequence was then amplified by PCR (sense primer: 5′-AAAACTGCAGGCTAGCGGAGGTAAG-3′ and antisense primer: 5′-AAAACTCGAGATTCAGACCTCTTC-3′) from pcDNA-(KLAKLAK)₂-Myc and subcloned into pTAT within the PstI and XhoI sites to generate pTK. Next, pmTK was generated by replacing (KLAKLAK)₂ with DLSLARLATARLAI. The complete enhanced green fluorescent protein (GFP) gene sequence was then amplified from plasmid pEGFP-C1 (Clontech) by PCR (sense primer: 5′-AAAAGAGCTCCCGCTAGCGCTAACC-3′ and antisense primer: 5′-AAACTCGAGCGATCTGAGTCCGGACTT-3′) followed by subcloning into pTK, pET28a(+) (Novagen), and pmTK within the SacI and XhoI sites to generate pTK-GFP, pEGFP, and pmTK-GFP, respectively. A double-stranded oligonucleotide encoding (KLAKLAK)₂-DEVD-GG was then inserted into pTAT-GFP within the BamHI and SacI sites to construct pTK1-GFP followed by the tandem insertion of two more copies of (KLAKLAK)₂-DEVD-GG into pTK1-GFP within the BamHI and SacI sites one by one to construct pTK3-GFP.

The fusion proteins were expressed in the BL21 (DE3) pLysS (Novagen) bacterial strain and then induced with 1 mmol/L isopropyl-l-thio-β-d-galactopyranoside. After bacterial cells were harvested and sonicated with binding buffer [20 mmol/L Tris-HCl (pH 8.0), 500 mmol/L NaCl, 5 mmol/L imidazole (pH 7.9), 8 mol/L urea, 1 mmol/L phenylmethylsulfonyl fluoride, and 0.5 mmol/L DTT], the suspensions were clarified by centrifugation (12,000 rpm for 10 min at 4°C) and then subjected to purification on a Ni-NTA column (Qiagen) under denatured conditions. After washing by the stepwise addition of increasing imidazole concentrations to remove the high background of contaminating bacterial proteins, the target peptides were eluted using an elution buffer [20 mmol/L Tris-HCl, 500 mmol/L NaCl, 300 mmol/L imidazole (pH 7.9), and 8 mol/L urea] followed by desalting with a disposable PD-10 Sephadex G-25 column (Amersham Biosciences). The fusion proteins were used immediately after purification.

Chinese hamster ovary cells (CHO; KCLB 10061) were cultured at 37°C and 5% CO₂ in MEM-α supplemented with 10% fetal bovine serum (FBS; Life Technologies). B16F10 mouse melanoma cells and human breast cancer cells (MCF-7) were cultured in RPMI 1640 (Life Technologies) supplemented with 10% FBS. Human umbilical vein endothelial cells (HUVEC) were isolated from umbilical cords by collagenase treatment and cultured in M199 medium (Sigma-Aldrich) supplemented with 20% FBS.

Cells grown on coverslips in 24-well culture plates to 70% to 80% confluency were treated with the indicated concentrations of proteins. After 2 h of incubation, the cells were washed three times with PBS and then fixed in methanol/acetone (1:1) for 5 min at room temperature. The cells were then washed again with PBS and a solution of SlowFade solution (Molecular Probe) was added. A coverslip was overlaid and the fluorescence was analyzed using a fluorescence microscope (Zeiss Axioplan2; Carl Zeiss).

CHO cells (2 × 10⁵ per well) grown in six-well culture plates were treated with the indicated concentrations of proteins. After 2 h of incubation, the cells were harvested, washed three times with cold PBS (pH 7.4), and then resuspended in lysis buffer [20 mmol/L Tris-HCl (pH 8.0), 137 mmol/L NaCl, 2 mmol/L EDTA, 2 mmol/L Na₃VO₃,10% glycerol, 1% Triton X-100, and 1 mmol/L phenylmethylsulfonyl fluoride]. After centrifugation (12,000 rpm for 10 min at 4°C), the supernatant (30 μg protein/lane) was separated by electrophoresis on a 12% SDS-polyacrylamide gel and then transferred onto a nitrocellulose membrane. The membrane was then blocked with TBS-T (50 mmol/L Tris-HCl, 150 mmol/L NaCl, and 0.1% Tween 20) containing 3% skim milk for 1 h and incubated with horseradish peroxidase–linked anti-His antibody (Invitrogen) for 16 h at 4°C followed by detection with enhanced chemiluminescence reagent (Amersham Biosciences). Anti-β-actin (Sigma) and horseradish peroxidase–labeled anti-mouse immunoglobulin (Santa Cruz Biotechnology) were used. Protein concentrations were determined according to the manufacturer's instructions (Bio-Rad).

The measurement of viable cell numbers was carried out using a mitochondrial reduction activity assay (14). Briefly, cells (3 × 10³ per well) were grown in 96-well culture plates overnight, at which time the medium was replaced with fresh medium containing 2% FBS. The cells were then treated with TK or mTK at concentrations ranging from 0.5 to 15 μmol/L for 3 h. In the case of TK1 or TK3, cells were treated for 24 h with concentrations ranging from 0.2 to 2 μmol/L. After incubation, 0.5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; 50 μL) solution (Sigma) was added to each well. Cells were then lysed with DMSO and quantified by measuring the absorbance at 570 nm using a microplate reader 550 (Bio-Rad). A bromodeoxyuridine (BrdUrd) incorporation assay was then done as described previously (15). Briefly, cells were incubated in the presence or absence of TK or mTK in fresh medium containing 2% FBS for 3 h. After replacing the medium with fresh medium, BrdUrd was added and incubated for 20 h. The amount of incorporated BrdUrd was then assessed using a Cell Proliferation Assay Kit (Oncogene) following the manufacturer's instructions.

Cells (2 × 10⁵ per well) were grown in six-well culture plates overnight and then incubated on plates in the presence of 10 μmol/L TK or mTK in fresh medium containing 2% FBS for 6 h followed by labeling with FITC-conjugated Annexin V (Annexin V-FITC; Santa Cruz Biotechnology) for 20 min in the dark. Cells were then washed and analyzed by fluorescence-activated cell sorting at 488 nm in a FACSCalibur flow cytometry system (BD Biosciences) equipped with a 5 W argon laser.

To study the cleavage of TK3 by caspase-3, 1 μg TK3-GFP was incubated with various amounts of recombinant caspase-3 (Upstate) for 60 min or with 50 ng caspase-3 for various time periods in 6 mmol/L Tris-HCl (pH 7.5), 1.2 mmol/L CaCl₂, 5 mmol/L DTT, 1.5 mmol/L MgCl₂, and 1 mmol/L KCl. Reaction mixtures (17 μL/lane) were then separated by SDS-PAGE and transferred onto a nitrocellulose membrane followed by immunoblotting with horseradish peroxidase–linked anti-His antibody.

The caspase-3 assay of cell lysates was conducted as described previously. Briefly, B16F10 cells were grown to confluency and then treated with various concentrations of TK1 or TK3 for 3 h. Both the cells and the supernatants were centrifuged (1,500 rpm at 4°C), washed three times with cold PBS (pH 7.4), and then resuspended in lysis buffer. The lysates were then centrifuged at 12,000 rpm for 10 min at 4°C, and the supernatants were subjected to immunoblotting with an anti-caspase-3 antibody (Cell Signaling) and with a goat anti-rabbit horseradish peroxidase secondary antibody (Santa Cruz Biotechnology). The same membrane was then stripped and reprobed using a monoclonal anti-β-actin antibody (Sigma).

An in vitro endothelial tube formation was done as described previously (16). Briefly, 100 μL Matrigel (Chemicon) was added to each well in 96-well culture plates and allowed to polymerize. HUVEC suspension (4 × 10⁵/mL; 100 μL) was then added to each well coated with Matrigel. After incubation for 2 to 3 h at 37°C, the cells were treated with the indicated concentrations of proteins for 3 to 5 h at 37°C. The cells were then photographed, and the number of branch points in five fields was counted (×100) for each sample.

The intratumoral microvessel density was analyzed on sections of paraffin-embedded B16F10 tumor using a rat anti-mouse CD31 monoclonal antibody (BD Biosciences). Immunoperoxidase staining was conducted using the Vectastain avidin-biotin complex Elite reagent kit (Vector Laboratories). Sections were counterstained with hematoxylin. Microvessel density was assessed initially by scanning the tumor at low power followed by identification of eight areas containing the maximum number of discrete microvessels and by counting the individual microvessels under low magnification (×100).

B16F10 cells (1 × 10⁶) were injected s.c. into 6 week-old male C57BL/6 mice in each flank. When 100 to 200 mm³ tumors were observed, 50 μL of 120 μmol/L TK, mTK, or PBS was injected into the tumors once a day for 7 days. Each injection was directed at multiple sites within the tumors and this could make the final peptide concentration within tumor ∼30 μmol/L. TK1- or TK3-treated tumors were used to compare their activities. The tumors were then measured with a caliper and the respective volumes were estimated using the following formula: length × (width)² / 2. After 8 days, the mice were sacrificed and the tumors were excised and frozen in OCT embedding medium (Tissue-Tek, Sakura Finetechnical). Terminal deoxynucleotidyl transferase–mediated nick end labeling staining of the tumors was then done according to the manufacturer's instructions (Roche). The tumors were also counterstained with methyl green to reveal their histologic architecture. To explore whether TK-GFP could transduce into tumor cells in vivo, cryosectioned tumors (5 μm) that were counterstained with 4′,6-diamidino-2-phenylindole were confirmed using a fluorescence microscope and a confocal inverted microscope (Leica DM IRB; Leica Microsystems). All experiments were done in accordance with recommendations for the proper use and care of laboratory animals.

All values are expressed as mean ± SE. The statistical significance of differential findings between experimental and control groups was determined using a Student's t test. P < 0.05 was considered statistically significant.

We generated constructs that expressed TAT fusion proteins in bacterial cells. TK contains nine amino acid residues of HIV-1-TAT followed by (KLAKLAK)₂, a proapoptotic peptide. In mTK, a non-α-helix-forming peptide, DLSLARLATARLAI, was used instead of (KLAKLAK)₂. GFP was linked with TK (TK-GFP) or mTK (mTK-GFP) to monitor the transduction of TAT fusion peptides (Fig. 1A). CHO cells were then treated with TK-GFP, mTK-GFP, or GFP for 2 h, after which TK-GFP and mTK-GFP were readily detected in the cells by fluorescence microscopy (Fig. 1B, b and c, top). Although TK-GFP and mTK-GFP could transduce the cells, GFP that did not contain TAT could not (Fig. 1B, a). B16F10 cells were also susceptible to TK-GFP or mTK-GFP transduction, as expected (Fig. 1B, b and c, bottom). Western blot analysis showed that TK-GFP transduction occurred in a concentration-dependent fashion and that mTK-GFP was also transduced into cells (Fig. 1C).

The proapoptotic peptide, (KLAKLAK)₂, has been known to disrupt mitochondrial membranes (1). Therefore, to determine whether transduced TK retains its proapoptotic activity, we tested cell viability using a MTT assay. Considerable cell death was observed in CHO, B16F10, and MCF-7 cells (Fig. 2A) that were treated with TK for 3 h, which lead to a decrease in the percent viability when compared with cells treated with mTK as a negative control. We also measured cell proliferation using a BrdUrd incorporation assay and found that the proliferation of CHO and B16F10 cells (Fig. 2B) treated with TK was significantly inhibited, whereas the proliferation of cells treated with mTK was unchanged. To confirm that the apoptosis was induced by TK, we labeled apoptotic cells with Annexin V-FITC and analyzed them by fluorescence-activated cell sorting and found that the amount of Annexin V–positive cells increased in response to TK in a time-dependent manner (Fig. 2C).

To explore whether TK-GFP could transduce into tumor cells in vivo, TK-GFP or GFP was intratumorally administrated into B16F10 melanoma tumors implanted in syngeneic C57BL/6 mice. After 30 min, the tumors were excised, cryosectioned, and analyzed by fluorescence and confocal microscopy. TK-GFP was detected in the cytoplasm and nuclei of tumor cells (Fig. 3A,, e); however, GFP was not detected (Fig. 3A, b). 4′,6-Diamidino-2-phenylindole staining to identify cells was then conducted (Fig. 3A, a and d) and merged with TK-GFP and GFP (Fig. 3A, c and f). Serial z-sections of a single-cell image observed using confocal microscopy confirmed that TK-GFP transduced into tumor cells (Fig. 3B, 1-10).

To determine whether transduced TK could induce tumor cell apoptosis and inhibit tumor growth in vivo, we injected TK intratumorally daily for 8 days. As shown in Fig. 3C, administration of TK led to a striking reduction in tumor sizes compared with the size of tumors in mTK- or PBS-treated controls, and this difference was significant after the third day of treatment. As treatment continued, the differences between TK- and mTK- or PBS-treated groups became more pronounced until P < 0.001. To determine whether apoptosis was indeed the mechanism of tumor reduction in vivo, terminal deoxynucleotidyl transferase–mediated nick end labeling staining of the tumor sections was conducted. Terminal deoxynucleotidyl transferase–mediated nick end labeling staining revealed that treatment with TK induced apoptosis throughout ∼70% of the total tumor volume, whereas mTK and PBS did not (Fig. 3D). No significant differences in body weight were observed between the groups (data not shown).

To design a more powerful and efficient antitumor peptide, we introduced more (KLAKLAK)₂ peptides together with DEVD, a substrate peptide for caspase-3. TK1 and TK3 were named according to the numbers of (KLAKLAK)₂ peptides that they contained. In some experiments, GFP was fused to each peptide, which were named TK1-GFP and TK3-GFP (Fig. 4A). To confirm that TK3 was cleaved by caspase-3, we incubated TK3-GFP with purified caspase-3. TK3-GFP was cleaved by caspase-3 in both dose- and time-dependent manners as indicated by the production of bands with lower molecular masses (Fig. 4B and C). To confirm that the cleavage of TK3 occurs in the cells, we transduced TK3-GFP into B16F10 cells. As shown in Fig. 4D, TK3-GFP protein into B16F10 cells was cleaved to produce an expected band.

To compare the proapoptotic activities of TK1 and TK3, we measured caspase-3 activity in B16F10 cells that had been treated with each peptide. As shown in Fig. 5A, caspase-3 was more prominently activated by TK3 than TK1. The result of a MTT assay also showed that TK3 was more potent than TK1 in inhibiting cell growth. The effect of 0.2 μmol/L TK3 is slightly better than 2 μmol/L TK1 (Fig. 5B). To compare the activities of TK3 and TK1 in vivo, we injected TK1 and TK3 into tumors daily. As shown in Fig. 5C, daily administration of TK3 led to a dramatic reduction of tumor volume compared with TK1, and daily injection of TK3 induced nearly complete inhibition of tumor growth. Strikingly, a single injection of TK3 significantly inhibited tumor growth, and three injections of TK3 almost completely blocked tumor growth. No significant differences in body weight were observed between the groups (data not shown).

Angiogenesis is essential for tumor growth; therefore, endothelial cells are also target cells for cancer therapy. The level of the in vivo antitumor effect of TK3 suggested that it may also inhibit angiogenesis; therefore, we tested the effects of TK3 on endothelial cell growth and vessel-like tube formation. To determine if TAT fusion proteins could tranduce into endothelial cells (HUVEC) in vitro, HUVEC were treated with TK-GFP, mTK-GFP, or GFP for 2 h. As shown in Fig. 6A,, a, fluorescence was readily detectable in the cells treated with TK-GFP and mTK-GFP; however, it was not detected in cells treated with GFP. We then confirmed transduction of TK-GFP into endothelial cells in vivo by intratumorally administering TK-GFP into B16F10 melanoma tumors implanted in syngeneic C57BL/6 mice. After 30 min, the tumors were excised, cryosectioned, and analyzed by fluorescence microscopy. TK-GFP was detected in perimeter of blood vessels, penetrating into the tumor (Fig. 6A, b).To determine if transduced TK also retained its proapoptotic activity in endothelial cells, a MTT assay was done. As shown in Fig. 6B, HUVEC incubated for 9 h with TK showed considerable cell death when compared with cells treated with mTK. To determine whether TK proteins inhibited angiogenesis, we examined their ability to disrupt endothelial cell tube formation in Matrigel. TK1 and TK3 inhibited endothelial tube formation in a dose-dependent manner, and TK3 was more potent than TK1. Unlike TK1 or TK3, mTK did not inhibit tube formation (Fig. 6C). To determine whether the reduced size of TK3-treated tumors coincided with reduced neovascularization, we used representative PBS-, TK1-, or TK3-treated tumors to quantify the density of microvessels. Similar to the data obtained from the tumor growth inhibition assay, treatment with TK3 led to a dramatic reduction of CD31⁺ blood vessels when compared with treatment with TK1 (Fig. 6D).

A variety of approaches for the delivery of apoptosis-inducing peptides have been attempted (1–3); however, most of those approaches have focused on targeted delivery or efficient intracellular transduction. In this study, however, we designed a fusogenic peptide, TK3, using an approach that focused on transduction as well as potency, which was enhanced intracellularly by a caspase-3-mediated positive feedback loop. TK3 is composed of TAT followed by three repeats of (KLAKLAK)₂-DEVD. Once TK3 transduces into cells, it activates caspase-3, which in turn cleaves DEVD to release more (KLAKLAK)₂ peptides. Therefore, one TK3 molecule releases three apoptosis-inducing peptides intracellularly. We used TAT for the intracellular delivery of the peptide because it is able to facilitate concentration-dependent delivery of fusion proteins into nearly all cells being treated (10, 17). We found that TK peptides rapidly transduce ∼100% of cells in vitro and that TK uptake occurred in most of the cells in areas of the tumors that were injected with TK peptides, suggesting that TK peptides can transduce efficiently both in vitro and in vivo. TAT fusion proteins have an advantage at protein purification because it can be done regardless of denaturation. Misfolding of TAT fusion proteins could enhance their transduction into cells, where they are folded correctly and retain their bioactivities (17). All of our TK peptides that were purified under denaturation conditions appeared to be correctly folded in the inside cells because their apoptosis activities were not affected and their GFP forms were actively fluorescent in cells both in vitro and in vivo. The amphipathic peptide, (KLAKLAK)₂, does not disrupt the plasma membrane of eukaryotic cells. If internalized, however, it can disrupt the negatively charged mitochondrial membrane, resulting in apoptosis (18). Because deregulated cell proliferation of tumors coupled with suppressed apoptotic sensitivity after many treatments often renders tumors more progressive and resistant to any treatment, a strong and efficient initial treatment is required. Although (KLAKLAK)₂ has a powerful apoptosis-inducing activity, its antitumor effect is limited in vivo as shown in this article. Therefore, TK3, which was shown to have a stronger antitumor effect than TK1, was generated. TK3 induced cell growth inhibition at a 10-fold lower concentration than TK1 in vitro, and conversion of procaspase-3 to caspase-3 also occurred after administration of a significantly lower concentration of TK3 than TK1. This significant difference in potency does not seem to occur simply as a result of the difference in (KLAKLAK)₂ numbers at the same molar concentration, however, because the presence of three times the (KLAKLAK)₂ peptide numbers cannot account for >10-fold difference in activity. This difference does not appear to be due to a difference in their transduction efficiency either because both peptides use the same TAT domain, which transduces the protein in a dose-dependent manner, suggesting that a single transduced TK3 molecule more efficiently mediates apoptosis than three transduced TK1 molecules. However, the precise mechanism by which this result occurs remains unclear at present. Interestingly, the in vivo effect of TK3 is more dramatic. We found that TK3-injected tumors were almost completely inhibited after 4 days of TK3 injection, whereas TK1-injected tumors were still growing, although their growth rates were much slower than those of PBS-injected tumors. TK3 also induced apoptosis in HUVEC, which suggests that TK3 can induce apoptosis of both tumor cells and vascular endothelial cells. Destroying both tumor cells and tumor vascular endothelial cells may result in a synergistic effect in vivo. In addition, a previous report (19) suggested that apoptosis plays a role in priming the immune response; therefore, TK3 may enhance the cellular immune response against tumor antigens by triggering massive apoptosis in the tumor milieu.

Although intratumoral injection of TK3 shows an excellent antitumor activity in vivo, we need to develop a way to deliver TK3 systemically. TK3 can be linked to a tumor-specific target moiety such as RGD sequence that has been successfully used for the delivery of a short peptide (20, 21). It can be also delivered to the tumor site by “passive tumor targeting” with the help of particles that can carry TK3 by making physical complex or by chemical conjugation (22–25). Taken together, our data show that a peptide that amplifies the caspase-3 activation signal by transducing tandem repeats of (KLAKLAK)₂-DEVD sequences into cells exhibited potent antitumor activities both in vitro and in vivo. Therefore, this study serves as the prototype for a peptide-based drug with the potential to reduce tumor burden when used alone or in combination with conventional therapies.

No potential conflicts of interest were disclosed.