ErbB2 is an excellent target for cancer therapies. Unfortunately, the outcome of current therapies for ErbB2-positive breast cancers remains unsatisfying due to resistance and side effects. New therapies for ErbB2-overexpressing breast cancers continue to be in great need. Peptide therapy using cell-penetrating peptides (CPP) as peptide carriers is promising because the internalization is highly efficient, and the cargoes delivered can be bioactive. However, the major obstacle in using these powerful CPPs for therapy is their lack of specificity. Here, we sought to develop a peptide carrier that could introduce therapeutics specifically to ErbB2-overexpressing breast cancer cells. By modifying the HIV TAT-derived CPP and conjugating anti-HER-2/neu peptide mimetic (AHNP), we developed the peptide carrier (P3-AHNP) that specifically targeted ErbB2-overexpressing breast cancer cells in vitro and in vivo. A signal transducers and activators of transcription 3 (STAT3)–inhibiting peptide conjugated to this peptide carrier (P3-AHNP-STAT3BP) was delivered more efficiently into ErbB2-overexpressing than ErbB2 low-expressing cancer cells in vitro and successfully decreased STAT3 binding to STAT3-interacting DNA sequence. P3-AHNP-STAT3BP inhibited cell growth in vitro, with ErbB2-overexpressing 435.eB breast cancer cells being more sensitive to the treatment than the ErbB2 low-expressing MDA-MB-435 cells. Compared with ErbB2 low-expressing MDA-MB-435 xenografts, i.p. injected P3-AHNP-STAT3BP preferentially accumulated in 435.eB xenografts, which led to more reduction of proliferation and increased apoptosis and targeted inhibition of tumor growth. This novel peptide delivery system provided a sound basis for the future development of safe and effective new-generation therapeutics to cancer-specific molecular targets. (Cancer Res 2006; 66(7): 3764-72)
- targeted therapy
Although exciting progress has been made toward understanding the molecular mechanisms underlying breast cancer development, many challenges remain. One is the development of safe and effective new-generation therapeutic agents targeting cancer-specific molecular alterations. The ErbB2 (or HER-2/neu) gene is a member of the epidermal growth factor receptor family and is overexpressed in about 30% of breast cancers; this was associated with the number of lymph node metastases and with a poor prognosis ( 1, 2). ErbB2 overexpression has been shown to increase the metastatic potential of human breast cancer cells ( 2– 4) and confer breast cancer cells an increased resistance to some chemotherapeutic agents ( 5– 7). Therefore, ErbB2 can serve as an excellent target for the development of novel cancer treatments. However, the outcome of current therapies against ErbB2 in breast cancers remains unsatisfying. For example, resistance to trastuzumab is quite significant in cancer patients despite the stringent criteria for selection of patients whose tumors express high levels of ErbB2 ( 8, 9). Therefore, development of new therapies for ErbB2-overexpressing breast cancers is urgent.
The signal transducers and activators of transcription (STAT) proteins are a family of transcription factors. In response to the binding of growth factors or cytokines to their receptors, STATs become phosphorylated on a tyrosine residue located near the COOH terminus (Tyr705 in STAT3), and then two phosphorylated STAT proteins form dimers through reciprocal phosphotyrosine-SH2 interactions, translocate into the nucleus, and bind to STAT-specific DNA response elements, thereby regulating gene expression ( 10, 11). STAT3 has been shown to be constitutively activated (++ to +++ nuclear staining) in 50% of breast carcinomas as determined by immunohistochemistry staining analysis ( 12). Results of studies using dominant-negative and activated mutant STAT3 and antisense oligonucleotides in relevant cell cultures, animal models, and patient samples have validated STAT3 as a cancer therapeutic target ( 13– 18). Along with others, we have found that ErbB2 can activate STAT3, and that this activation may contribute to ErbB2-induced transformation and tumor progression ( 19, 20). Hence, specifically targeting STAT3 in ErbB2-overexpressing breast cancer cells that contain activated STAT3 may effectively inhibit ErbB2-mediated malignant phenotypes.
The goal of anticancer therapy is to kill malignant cells while keeping the side effects to normal cells minimal. Traditional therapies like chemotherapy are often poorly specific to cancer cells, and the need to develop targeted therapies is pressing. However, the use of polypeptides or oligonucleotides as targeted therapeutic agents has always been a challenge because of their low biomembrane permeability or their relatively rapid degradation. The cell-penetrating peptides (CPP) hold great promise as the therapeutic agent carriers because of their ability to effectively cross cell membranes and to deliver bioactive cargoes into the cells ( 21). Recent studies have validated the potential of CPPs in cellular translocation of biologically active molecules, such as peptides, proteins, and oligonucleotides ( 21, 22). About 20 different peptide delivery carriers have been reported thus far, the majority being the cellular internalization domains of the respective proteins ( 22, 23). Some examples include the third α-helix of the homeodomain of Antennapedia, a Drosophila transcription factor ( 24, 25), the VP22 DNA-binding protein of herpes simplex virus-1 ( 22, 26), and the small protein transduction domains from the transduction and transactivation protein (TAT) of HIV ( 27, 28). Another study has elegantly showed the potential of TAT protein in the clinical application by showing that i.p. injection of the 120-kDa β-galactosidase protein, fused to TAT protein, results in delivery of the biologically active fusion protein to all tissues in mice ( 29). This result suggests that TAT-based CPPs may be developed into an effective delivery system for therapeutic peptides or proteins and may have great potential in therapeutic applications. However, the major impediment to use CPPs for therapies is their lack of specificity; for example, they deliver to both cancer and normal cells at similar level.
The purpose of this study is to develop a novel TAT-based, target-specific delivery system to introduce therapeutic peptides selectively into ErbB2-overexpressing breast cancer cells. We conjugated an ErbB2 extracellular domain-binding peptide with a TAT-derived CPP to achieve target specificity of delivery. This novel TAT-based peptide delivery system delivered a therapeutic peptide, STAT3BP, which is known to inhibit STAT3 signaling, into ErbB2-overexpressing breast cancer cells in vitro and in vivo, and specifically inhibited the growth of ErbB2-overexpressing breast tumor xenografts by blocking STAT3 activation. The results of this work provide new insight for designing new-generation targeted therapies for breast and other types of cancers.
Materials and Methods
Cell lines and cell cultures. Human breast cancer cell lines MDA-MB-435, SKBr3, BT474, and MCF7 were obtained from the American Type Culture Collection (Rockville, MD). Wild-type erbB2 transfectants of two of these cell lines, 435.eB and MCF7/HER-2, have been described previously ( 30, 31). Constitutively active erbB2 transfectant of MDA-MB-435, 435.VE, was also described previously ( 7). The cells were grown in DMEM containing a high level of glucose (DMEM/F-12; Life Technologies, Grand Island, NY) supplemented with 10% fetal bovine serum. The antibiotic G418 (Calbiochem, La Jolla, CA) was added in the medium (800 μg/mL) for the selection of ErbB2 transfectants.
Peptide synthesis. The peptides used included fluorescein-conjugated, 11-amino-acid TAT-derived peptide P1 (FITC-β-Ala-YGRKKRRQRRR) and its derivatives P2 (FITC-β-Ala-YGRKKRRQRR) and P3 (FITC-β-Ala-YGRKKRRQR). The control peptide P0 (FITC-β-Ala-YGKKKKKKKKK) was also used. All four peptides were synthesized by Ansynth Service B.V. (Berkel en Rodenrijs, the Netherlands). The biotin-conjugated TAT-derived peptides [biotin-YGRKKRRQR-G-FCDGFYACYKDV (with an intra-molecular disulfide bond) and biotin-YGRKKRRQR] were synthesized by the Peptide Core Facility at M.D. Anderson Cancer Center. The PY(PO3)L sequence-containing peptide with an intramolecular disulfide bond [5′-Fam-β-Ala-YGRKKRRQR-G-FCDGFYACYKDV-PY(PO3)L] was synthesized by SynPep Corp. (Dublin, CA) and New England Peptide, Inc. (Gardner, MA).
Antibodies and reagents. Anti-STAT3 antibodies were purchased from Santa Cruz Biotechnology (clone H-190, Santa Cruz, CA) and Upstate Biotechnology (Charlottesville, VA). An antibody to phosphorylated STAT3-Tyr705 was obtained from Cell Signaling Technology (Beverly, MA), and an antibody to ErbB2 was obtained from Calbiochem. Trastuzumab (Herceptin) was provided by Genentech, Inc. (San Francisco, CA). The antibodies used for immunohistochemical staining included rabbit anti-human c-erbB2-oncoprotein antibody and peroxidase-conjugated rabbit anti-FITC antibody (DAKO, Carpinteria, CA). Ki67 antibody clone SP6 was purchased from Lab Vision (Fremont, CA).
Western blot analysis. Total cell lysates were collected, and the proteins were subjected to electrophoresis and transferred to nitrocellulose membranes, which were then probed with primary antibodies. Horseradish peroxidase–conjugated secondary antibodies were detected using chemiluminescence, as described previously ( 32).
Immunohistochemistry. Immunohistochemical staining was done as described previously ( 9). Cell proliferation was assessed using an antibody against Ki67.
Immunofluorescence staining and fluorescence microscopy. Immunofluorescence staining and fluorescence microscopy were done as described previously ( 7). Briefly, the cells were cultured on chamber slides (Falcon, Lincoln Park, NJ), treated with the indicated custom peptides, and then fixed in 2% paraformaldehyde/PBS for 20 minutes at room temperature. For cells treated with fluorescein-conjugated peptides, slides were subjected to fluorescence microscopy after 4′,6-diamidino-2-phenylindole staining. For visualization of the peptides linked with biotin, cells were permeabilized with 0.3% Triton X-100/PBS for 10 minutes at room temperature and incubated with fluorochrome-labeled streptavidin overnight at 4°C. Slides were examined with a Nikon ECLIPSE E400 microscope (Nikon, Japan) and photographed with a Sensys digital camera (Photometrics Ltd., Tucson, AZ). The fluorescent haze was removed with a two-dimentional deconvolution algorithm in the MetaMorph Imaging System (Universal Imaging, Downingtown, PA).
Cell viability assay. Cell viability assay was done as described previously ( 9).
Electrophoretic mobility shift assays. The assays were carried out as described previously with minor modifications ( 33, 34). Briefly, nuclear extracts were prepared from 435.eB cells treated with the indicated peptides and incubated with radiolabeled high-affinity, sis-inducible elements (hSIE; 5′-GTGCATTTCCCGTAAATCTTGTCTACA-3′; Santa Cruz Biotechnology, Santa Cruz, CA). The hSIE probe was radiolabeled with T4 polynucleotide kinase (Roche Applied Science, Indianapolis, IN).
For the competition assays, unlabeled probes were used in 50-fold molar excess to the radiolabeled probe. Radiolabeled probes were added to the nuclear extracts, and the mixtures were incubated for 20 minutes at room temperature. Supershift assays to detect STAT3 binding were done by adding the respective antibodies (from Santa Cruz Biotechnology or Upstate Biotechnology) for 10 minutes at room temperature before the gel-shift procedure. The reaction complexes were resolved on a 4% nondenaturing polyacrylamide gel.
Animal experiments. Six-week-old female severe combined immunodeficient (ICR-SCID) mice were purchased from Taconic (Germantown, NY). MDA-MB-435 or 435.eB cells (2 × 106) in a mixture of 200 μL of PBS and Matrigel (BD Biosciences, Bedford, MA; 2:1 ratio) were injected s.c. into the mammary fat pad on either side of each mouse, respectively. When the established xenografts reached a size of at least 150 mm3, the mice were randomly divided into two groups of 10 to 12 mice each. Each group received i.p. injections of either P3-AHNP-STAT3BP or control P3-AHNP peptides (150 nmol/mouse; thrice per week) in PBS. The tumor diameters were serially measured with calipers thrice per week, and the tumor volumes were calculated with the following formula: volume = width2 × length × 0.5. After 1 month of treatment, the mice were euthanized. An unpaired Student's t test was used to assess statistical significance.
In situ terminal deoxynucleotidyl transferase–mediated nick end labeling assay. The breast cancer xenografts harvested from the female ICR-SCID mice treated with P3-AHNP-STAT3BP were embedded in paraffin blocks, and the sections were mounted on slides. Apoptosis was determined in situ by internucleosomal DNA fragmentation [terminal deoxynucleotidyl transferase–mediated nick end labeling (TUNEL)] as described previously ( 35).
Statistics. All in vitro experiments were repeated at least tree times. The results are described as mean ± SE. Statistical analysis was done using Prism software (GraphPad Software, San Diego, CA) and two-tailed Student's t test was used for comparisons between groups. P < 0.05 was designated as statistically significant.
Optimization and characterization of the cell-penetrating properties of TAT-derived peptides. An 11-amino-acid peptide derived from TAT sequence could efficiently deliver bioactive protein into all tissues in the mice; however, the delivery is not specific ( 29). To optimize the peptide delivery system for specific targeting, we tested whether the 11-amino-acid peptide could be shortened and still be effective in deliver therapeutic agents into cells. We reasoned that downgrading the nonspecific internalization ability of TAT sequence by decreasing its arginine (R) residues ( 36) may allow our latter modification of TAT-derived peptide for specific targeting into ErbB2-overexpressing cancer cells. In addition, shortening the TAT sequence may prevent or reduce the immunogenicity induced by the foreign peptides and the cytotoxicity due to possible nonspecific binding of peptides to cellular DNA. To this end, a panel of peptides derived from the HIV TAT protein was synthesized with deletions of different numbers of COOH-terminal arginine residues ( Fig. 1A ). P1, the 11-amino-acid peptide previously shown to internalize efficiently through the plasma membrane ( 29), was used as a positive control. A negative control peptide P0 was constructed with multiple lysine (K) residues in place of arginines. The fluorophore FITC was tagged at the NH2 terminus of the peptides to allow detection of the subcellular localization of the peptides.
We first examined the ability of these TAT-derived peptides to cross cell membranes in various breast cancer cells. ErbB2-overexpressing SKBr3 cells were incubated with medium containing individual P1, P2, P3, or P0 peptides, and the internalization of the peptides were detected by immunofluorescence staining ( Fig. 1B). As indicated by its strong fluorescence signals in both the cytosolic and nuclear compartments, P1 was highly efficient in cell penetration. Whereas P2 and P3 had a gradually decreased extent of penetration ( Fig. 1B), a finding that is consistent with the notion that arginine residues are crucial for cell penetration ( 36). When the cells were treated with same concentration of the peptides, each of P1, P2 (data not shown), and P3 peptide internalized to a similar degree into the MDA-MB-435, BT-474 ( Fig. 1C), and SKBr3 ( Fig. 1D, 20 nmol/L) cell lines, indicating the internalization is in a cell type–independent manner. When the SKBr3 cells were treated with different concentrations of P3 for 15 minutes, cellular internalization of the peptide showed to be concentration dependent and majority of the peptides were in the cytoplasmic compartment ( Fig. 1D). Taken together, although it is shortened, P3 peptide still internalized into multiple breast cancer cell lines, and its reduced nonspecific penetration may allow us to achieve target specificity when conjugated with a targeting moiety. Therefore, we primarily used this peptide in our latter studies.
Anti-HER-2/neu peptide-conjugated TAT-derived peptide achieves ErbB2-targeted delivery in vitro and in vivo. A 12-amino-acid anti-HER-2/neu peptide mimetic (AHNP) has been shown to specifically bind to ErbB2 with high affinity ( 37). To achieve ErbB2-targeting capacity, we covalently linked AHNP peptide to the COOH terminus of the P3 peptide (P3-AHNP, Fig. 2A ). A biotin was linked to the NH2 terminus of the peptide, which can be visualized with a streptavidin-fluorophore. Biotin-P3 was synthesized to serve as a control peptide. We then added 20 nmol/L P3-AHNP or P3 to the ErbB2 low-expressing MDA-MB-435 human breast cancer cells or their ErbB2 transfectant 435.eB cells and compared peptide penetration efficiency. After 15 minutes of incubation, P3-AHNP more efficiently internalized into ErbB2-overexpressing 435.eB cells, as indicated by higher fluorescence intensities, than it did into MDA-MB-435 cells ( Fig. 2B). P3-AHNP also internalized better than did P3 specifically into ErbB2-overexpressing 435.eB cells, whereas in MDA-MB-435 cells, the addition of the AHNP moiety to P3 did not increase the extent of peptide internlization (compare Fig. 2B and C). The quantification of the fluorescence intensities per unit area in the cytoplasmic compartments confirms that more P3-AHNP molecules penetrated into the 435.eB cells than in the MDA-MB-435 cells ( Fig. 2D).
AHNP was developed from the ErbB2-binding CDR3 loops of the recombinant humanized anti-HER-2 monoclonal antibody rhuMAb4D5 (trastuzumab), which binds to the extracellular domain of ErbB2 receptor ( 37). If the enhanced internalization into ErbB2-overexpressing cells resulted from ErbB2 recognition by AHNP, pretreating ErbB2-overexpressing breast cancer cells with trastuzumab should preoccupy the ErbB2 receptors and reduce the internalization of P3-AHNP to these cells. Indeed, when we pretreated ErbB2-overexpressing 435.eB and SKBr3 cells with trastuzumab, P3-AHNP translocation was reduced in both cell lines ( Fig. 2E), indicating that the ErbB2-targeting capacity of P3-AHNP was mediated at least in part by the AHNP moiety. Taken together, these data showed that P3-AHNP is capable of targeted delivery to ErbB2-overexpressing breast cancer cells.
Next, we investigated whether P3-AHNP also selectively targets ErbB2-overexpressing human breast tumors in an animal model. We established tumor xenografts of MDA-MB-435 and 435.eB breast cancer cells in the mammary fat pads on either side of 6-week-old female SCID mice. The mice were then treated with biotin-P3 or biotin-P3-AHNP through i.p. injections, and xenografts were harvested after 2 hours. Immunohistochemical staining using anti-biotin antibody showed that the signals of P3-AHNP were much stronger in the ErbB2-overexpressing 435.eB xenografts than ErbB2 low-expressing MDA-MB-435 xenografts ( Fig. 2F, left). In contrast, without the AHNP moiety, P3 internalized into both cell types at similarly lower levels ( Fig. 2F, right). These results indicate that the AHNP-linked P3 possesses ErbB2-targeting specificity both in vitro and in vivo. Thus, we have successfully developed a TAT-based peptide that specifically targets ErbB2-overexpressing breast cancer cells.
ErbB2-selective delivery of STAT3-blocking peptide by P3-AHNP into breast cancer cells. ErbB2 activates multiple downstream signaling pathways, including STAT3, which may contribute to ErbB2-induced cell transformation and tumor progression ( 19, 20, 38). Indeed, in 435.eB, 435.VE, and MCF/HER-2 breast cancer cell lines that express high levels of ErbB2, phosphorylation on the Y705 residue of STAT3 is much higher than that in the parental cell lines, indicating higher activation of STAT3 signaling pathway in these ErbB2-overexpressing cells ( Fig. 3A ). We reasoned that by blocking the activation of STAT3 in ErbB2-overexpressing cells using an P3-AHNP delivery vehicle, we may inhibit the malignant behavior of ErbB2-overexpressing, STAT3-activated breast cancer cells and thus could eventually provide therapeutic benefits to breast cancer patients with ErbB2-overexpressing tumors.
A STAT3 SH2 domain-binding phosphotyrosyl peptide, X(pY)L, has been shown to efficiently inhibit STAT3 activation in vitro ( 33). Although it could not internalize into cells by itself, when linked to a membrane-translocating sequence, it inhibited downstream gene regulation and the transformation of v-Src–transformed NIH3T3 cells ( 33). We conjugated this STAT3 binding peptide (STAT3BP) to the P3-AHNP vehicle and investigated whether the resulting peptide (P3-AHNP-STAT3BP) could achieve targeted delivery to ErbB2-overexpressing breast cancer cells, inhibit their STAT3 signaling, and exert a growth inhibitory effect. When applied to 435.eB cells in vitro, fluorophore-conjugated P3-AHNP-STAT3BP efficiently internalized into these cells in a concentration-dependent manner ( Fig. 3B). As we expected, the signals were stronger in the 435.eB cells than in the MDA-MB-435 cells, indicating the ErbB2-targeting specificity of the peptide.
To confirm that the targeting specificity of P3-AHNP-STAT3BP was closely associated with ErbB2 expression levels, we incubated the peptide with MDA-MB-435 and MCF7 breast cancer cells and their ErbB2 transfectants 435.eB and MCF7.HER-2, respectively. ErbB2 was counterstained with Texas Red. A higher degree of P3-AHNP-STAT3BP translocation was indeed associated with higher ErbB2 staining ( Fig. 3C). A similar translocation pattern was observed in two other ErbB2-overexpressing breast cancer cell lines, BT474 and SKBr3 ( Fig. 3D). These results indicate that P3-AHNP can selectively deliver STAT3-blocking peptide STAT3BP into multiple ErbB2-overexpressing breast cancer cell lines.
P3-AHNP-STAT3BP disrupts STAT3 activation in vitro. We next examined whether P3-AHNP-STAT3BP could disrupt STAT3 activation in living cells. ErbB2-overexpressing 435.eB cells were treated with P3-AHNP-STAT3BP or control P3-AHNP peptides for 6 hours, after which nuclear extracts were prepared and electrophoretic mobility shift assays were conducted to examine the STAT3 DNA-binding activity ( Fig. 3E). Unlike the control DMSO and P3-AHNP treatments, P3-AHNP-STAT3BP treatment reduced STAT3 DNA binding activities in a dose-dependent manner. Two antibodies that against two different epitopes of STAT3 both blocked and supershifted the peptide-DNA complex ( Fig. 3E, lanes 7 and 8), indicating that the peptide-DNA complex contained STAT3. These results showed that P3-AHNP-STAT3BP can efficiently block STAT3 activation in living cells.
To determine whether the P3-AHNP-STAT3BP peptide has tumor cell suppressing functions, we treated MDA-MB-435 and 435.eB cells with increasing concentrations (5, 50, or 100 μmol/L) of P3-AHNP-STAT3BP or control peptide P3-AHNP for 102 hours, and we measured the cell proliferation rates using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay ( Fig. 3F). With 50 or 100 μmol/L P3-AHNP-STAT3BP treatments, the survival rate of MDA-MB-435 cells decreased 50% or 54%, respectively. Notably, the 435.eB cells were more sensitive to the P3-AHNP-STAT3BP treatment, with inhibition rates of 65% (at 50 μmol/L) or 70% (at 100 μmol/L), respectively. In contrast, the suppressing effect of P3-AHNP was much less potent than that of P3-AHNP-STAT3BP as it did not significantly decrease survival of the cells at 50 μmol/L concentration but only did at 100 μmol/L concentration. To further confirm these findings, we did MTS assay on breast cancer cell line MCF7 and its ErbB2 transfectants MCF7.eB cells after P3-AHNP-STAT3BP treatment (Supplementary Fig. S1). P3-AHNP-STAT3BP inhibited the proliferation of both MCF7 and MCF7.eB cells and the inhibitory effect is more potent in MCF7.eB cells than in MCF7 cells. Therefore, P3-AHNP-STAT3BP selectively inhibits ErbB2-overexpressing tumor cell growth in vitro.
P3-AHNP-STAT3BP inhibits ErbB2-overexpressing tumor growth in vivo. Next, we investigated whether P3-AHNP-STAT3BP retains its ErbB2-targeting feature in vivo. The SCID mice bearing MDA-MB-435 and 435.eB tumor xenografts on either side of the mammary fat pads were i.p. injected with 150 nmol P3-AHNP-STAT3BP peptide. The mice were euthanized 6, 24, or 48 hours after the injection, and their tumor xenografts were harvested. Immunohistochemical staining for ErbB2 and P3-AHNP-STAT3BP peptide was done on these tumor samples and on untreated xenograft tumor samples. The P3-AHNP-STAT3BP peptide located predominantly in the ErbB2-overexpressing 435.eB tumors and less in the ErbB2 low-expressing MDA-MB-435 tumors or normal liver tissues on the same mouse ( Fig. 4A ). Notably, compared with the 6-hour time point, the intensities of P3-AHNP-STAT3BP peptide staining was much higher in 435.eB tumors at 48 hours after the injection, indicating a tendency to accumulate in the ErbB2-overexpressing 435.eB tumors, whereas in MDA-MB-435 tumors or liver tissues the intensities remained very low ( Fig. 4A). This indicates that the ErbB2-targeting power of the peptide carrier did not diminish but rather increased over the course of treatment.
STAT3 contributes to tumor cell malignancy at least in part by its antiapoptotic function ( 17). To examine whether P3-AHNP-STAT3BP selectively induce apoptosis in ErbB2-overexpressing cancer cells in vivo, we did in situ TUNEL staining on the tissue samples. Without peptide treatment, the TUNEL-positive staining cells were minimal in both MDA-MB-435 and 435.eB tumors (Supplementary Fig. S2A). However, P3-AHNP-STAT3BP treatment induced more apoptotic cells in ErbB2-overexpressing 435.eB tumors compared with ErbB2 low-expressing MDA-MB-435 tumors and the normal liver tissues, which showed only a small number of apoptotic cells after treatment ( Fig. 4B), indicating that P3-AHNP-STAT3BP preferentially induce apoptosis in ErbB2-overexpressing tumor tissues. P3-AHNP also induced apoptosis in ErbB2-overexpressing tumors, but the proapoptotic effect of P3-AHNP was much less potent than that of the P3-AHNP-STAT3BP ( Fig. 4B; Supplementary Fig. S2A).
We also used Ki67 labeling to measure the proliferation rate of the cells in the tumor xenografts. Similar to the results of TUNEL staining, Ki-67 staining showed that after P3-AHNP-STAT3BP or P3-AHNP treatment, the proliferation rate for the 435.eB tumors was lower than that of the MDA-MB-435 tumors, but both tumor types had similar high proliferation rates when left untreated ( Fig. 4C; Supplementary Fig. S2B), indicating that P3-AHNP and P3-AHNP-STAT3BP can selectively inhibit the proliferation of ErbB2-overexpressing breast cancer cells in vivo. Again, the antiproliferative effect of P3-AHNP was much less potent than that of the P3-AHNP-STAT3BP ( Fig. 4C; Supplementary Fig. S2B).
To examine whether P3-AHNP-STAT3BP can selectively inhibit ErbB2 overexpressing tumors in vivo, we i.p. injected 150 nmol of P3-AHNP-STAT3BP or control P3-AHNP peptides, respectively, into two groups of 12 female SCID mice bearing MDA-MB-435 and 435.eB tumor xenografts on either side of the mammary fat pads. Tumor diameters were serially measured as described in Materials and Methods. After 4 weeks of peptide treatment, the growth curves of the xenografts were plotted. In the MDA-MB-435 xenografts ( Fig. 4D, right), P3-AHNP-STAT3BP did not show further growth inhibition than did P3-AHNP. However, growth of the 435.eB xenografts ( Fig. 4D, left) was significantly inhibited by P3-AHNP-STAT3BP, indicating that P3-AHNP-STAT3BP can selectively inhibit ErbB2-overexpressing tumors in vivo. When comparing the xenografts treated with P3-AHNP, 435.eB showed a slower growth rate than that of MDA-MB-435, indicating that P3-AHNP also inhibited 435.eB tumor growth, which is consistent with the previous report ( 37), but to a significantly lesser extent than did P3-AHNP-STAT3BP.
In conclusion, P3-AHNP achieved targeted delivery of STAT3BP peptide to ErbB2-overexpressing 435.eB xenografts, resulting in increased apoptosis, decreased proliferation and, ultimately, growth suppression of 435.eB tumors.
TAT-derived peptides have been known for their highly efficient but nonspecific translocation into many cell types tested ( 28, 29, 39). Although these features enable TAT-derived peptides to deliver therapeutic agents into the cells efficiently, they also make it very difficult to achieve target-specific delivery. In this study, we convert the nonspecific TAT-derived, cell-penetrating peptide into an ErbB2-targeting delivery carrier by downgrading the nonspecific internalization efficiency of TAT sequence and conjugating an ErbB2-binding moiety to the TAT-derived sequence.
We showed that the shortened TAT-derived peptide P3 internalized to a similar extent into multiple cell lines tested ( Fig. 1C and D), a finding that is consistent with a previous report that TAT peptides might translocate through plasma membrane by a mechanism involving features common to all cell types ( 40). Comparing with the original TAT-derived sequence, the shortened peptide P3 has a reduced internalization capacity. With its attenuated internalization ability, P3 is a good starting sequence for building a target-specific peptide delivery vehicle.
The TAT-derived peptides we used in this study contain a potential nuclear localization sequence (NLS) Gly-Arg-Lys-Lys-Arg-Arg ( 40). Although the P1 and P2 peptides had a moderate to high extent of nuclear translocation, as expected, the P3 peptide accumulated predominantly in the cytosolic compartment despite the unaltered NLS in this peptide. This difference might allow further manipulation of the targeted subcellular delivery of the conjugated cargo by using different TAT-derived peptides (P1, P2, or P3) based on different delivering needs. Moreover, the addition of AHNP moiety rendered the translocation profile of P3 peptide ErbB2 selective both in vitro and in vivo. This observation has important implications for the clinical applications of cell-penetrating peptides, which to date have been limited by their inability to deliver bioactive molecules to specific targets.
We also showed that P3-AHNP carrier successfully delivered a STAT3-inhibiting peptide, STAT3BP, selectively to ErbB2-overexpressing cancer cells ( Fig. 3B-D). This differential delivery may explain why there was a great extent of growth inhibition in 435.eB xenografts as opposed to MDA-MB-435 xenografts ( Fig. 4D). Because the activation level of STAT3 in the 435.eB cells was much higher than that in the MDA-MB-435 cells ( Fig. 3A), the 435.eB cells might have become more dependent on the STAT3 pathway for survival or progression to malignancy or both. This greater activation level of STAT3 might contribute to the susceptibility of 435.eB cells to a disruption of STAT3 signaling by P3-AHNP-STAT3BP in vivo.
P3-AHNP-STAT3BP showed much stronger tumor suppressing activity than that of P3-AHNP ( Fig. 4D) in 435.eB tumors, showing the STAT3-targeting effect of STAT3BP is important to achieve the tumor-inhibiting capacity of P3-AHNP-STAT3BP. It also shows that the P3-AHNP is an excellent targeting vehicle for specific delivery of different therapeutics blocking ErbB2 oncogenic signaling pathways.
Our study proved that the “modular approach” in designing target-specific peptide therapy is feasible. Hypothetically, by replacing the AHNP moiety with a different molecule targeting another tumor surface marker, along with a relevant therapeutic agent conjugated to the resulting vehicle, we might be able to attack numerous pivotal cancer pathways. It might also be possible to design “tailored” anticancer drugs according to the specific tumor molecular marker profiles of individual patients, thus developing drugs that have higher efficacy, fewer side effects, and less resistance than currently available therapies have. Such modular design of these types of delivery system holds great promise for the future development of novel target-specific anticancer therapies.
In summary, in this proof of concept study, we have successfully developed a novel TAT-based peptide delivery system that can deliver in a target-specific manner a therapeutic peptide, STAT3BP, which inhibits STAT3 activation, and selectively inhibit the growth of ErbB2-overexpressing breast cancer cells in vitro and in vivo. This system might also be used to deliver various biochemically amendable therapeutics and might have great applications in development of new-generation targeted cancer therapy.
Grant support: U.S. Army Medical Research and Material Command grant DAMD-17-99-1-9271 (D. Yu); NIH grants P30-CA 16672 (M.D. Anderson Cancer Center), 1RO1-CA109570, 1RO1-CA119127-01, and PO1-CA099031 project 4 (D. Yu); and M.D. Anderson Cancer Center institutional research grant (M. Tan). K-H. Lan is a recipient of a predoctoral fellowship from the U.S. Army Breast Cancer Research Program Training Grant to M.D. Anderson Cancer Center (DAMD17-99-1-9264).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Drs. Yoichi Nagata and Zhiyong Ren for valuable discussions and technical assistance, Dr. Martin Campbell and the M.D. Anderson Peptide Core Facility for peptide synthesis, Priscilla Chauvin for assistance with equipments, and Carol Johnston for preparing the histology slides.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
M. Tan and K-H. Lan contributed equally to this study.
- Received August 3, 2005.
- Revision received December 19, 2005.
- Accepted January 25, 2006.
- ©2006 American Association for Cancer Research.