Inhibition of Mitogen-Elicited Signal Transduction and Growth in Prostate Cancer with a Small Peptide Derived from the Functional Domain of DOC-2/DAB2 Delivered by a Unique Vehicle

Differentially expressed in ovarian cancer-2/disabled 2 (DOC-2/DAB2) is a potent growth inhibitor for prostate cancer cells by suppressing several mitogen-elicited mitogen-activated protein kinase (MAPK) pathway (1–3). Our data indicate that the proline-rich domain in the COOH terminus of DOC-2/DAB2 is the key functional domain responsible for this activity. Thus, DOC-2/DAB2 could be a potential therapeutic agent for prostate cancer. Recent development of cell-permeable peptide (CPP) as a delivery vehicle in combination with functional peptide from tumor suppressor offers a new avenue of cancer therapy (4) because CPP is able to carry different cargos, such as proteins/peptide, DNA/RNA, liposome, and nanoparticles. In addition, recent studies (5, 6) indicated that CPP-conjugated small peptide did not cause many undesirable side effects in the preclinical animal model. Because the potential application of CPP on prostate cancer has not yet explored, we decided to identify the most efficient CPP delivery system using several prostate cancer cell lines. Based on our knowledge from studying DOC-2/DAB2 protein, we synthesized a fusion peptide by conjugating CPP with the functional proline-rich domain derived from DOC-2/DAB2 sequence and showed that it exhibited the same biochemical and biological activities as the native protein. Data from this study have provided a good rationale to use small peptide as a tool for studying its biological function and to develop into a potential therapeutic agent in prostate cancer cells.

Peptide synthesis. TAT (^FITC-G-RKKRRQRRR; ref. 7), penetratin (PENE; ^FITC-G-RQIKIWFQNRRMKWKK; ref. 7), KALA (^FITC-G-KLALKLALKALKAALKLA; ref. 7), homopolyers of l-arginine R11 (^FITC-G-R₁₁) or l-lysine K11 (^FITC-G-K₁₁; ref. 8), R11AAL (^FITC-G-R₁₁-GGG-FQLRQAALVASRKGE), and R11PPL (^FITC-G-R₁₁-GGG-FQLRQPPLVPSRKGE) were synthesized by automated peptide synthesizer using the standard solid-phase chemistry, purified by reverse-phase high-performance liquid chromatography, and analyzed by mass spectrometry. The amount of peptide was determined by mass spectrometry and normalized by fluorescence intensity.

Determination of peptide uptake and its subcellular localization. All prostate cancer cell lines were maintained as described previously (3). To determine the uptake efficiency of CPP, 1 × 10⁵ cells per well were plated in a 12-well plate. Next day, different concentrations of FITC-tagged peptides were incubated with cells for 30 minutes or indicated time. The total cell number was determined and cells were lysed in Tris [50 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, 5 mmol/L EDTA, 1% Triton X-100]. The fluorescence intensity was examined by fluorometer (excitation, 490-500 nm; emission, 515-525 nm). To determine the subcellular localization of CPP, after 30-minute incubation, cells were fixed with 4% paraformaldehyde in PBS plus 4′,6-diamidino-2-phenylindole (DAPI; 1 μg/mL) counterstaining (Sigma, St. Louis, MO) and examined under fluorescence microscope.

Cell growth and apoptosis assay. Cells were plated in 96-well plate at a density of 10³ per well in T-medium supplemented with 5% charcoal-stripped fetal bovine serum (FBS) for overnight. Cells were incubated with or without indicated peptides (5 μmol/L) for 3 hours before adding 5% FBS or 10 nmol/L dihydrotestosterone. The peptides were changed every 2 days. The cell proliferation was determined 4 days after the initial peptide treatment by using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Roche, Indianapolis, IN). The cell apoptosis was done using Cell Death Detection ELISA^PLUS (Roche) according to the manufacturer's manual and the positive control from the kit was include to ensure the quality of this assay.

Western blot analysis. Cells were plated in six-well plates (2 × 10⁵ per well) overnight. Next day, cells were incubated with peptides (5 μmol/L) for 3 hours, 10 ng/mL epidermal growth factor (EGF; Upstate, Charlottesville, VA), or 5% FBS was added to the culture and cells were harvested at the indicated time. The cell lysate was subjected to 10% SDS-PAGE and then probed with phosphorylated extracellular signal-regulated kinase 1/2 (ERK1/2; αpERK; Cell Signaling, Beverly, MA) or ERK2 (αERK2; Santa Cruz, Santa Cruz, CA) antibody. The protein intensity was detected using an enhanced chemiluminescence detection kit (Amersham, Piscataway, NJ).

In vitro pull-down assay. As described previously (2), cell lysate was prepared in 0.5 mL lysis buffer; after a low-speed spin, 0.4 mL supernatant was collected and incubated with 30 μL glutathione S-transferase (GST)-Src homologue 3 (SH3) fusion protein with glutathione-Sepharose overnight at 4°C. The increasing concentrations of peptide (10, 30, and 100 μmol/L) for R11PPL or R11AAL were also added during the incubation. Next day, pellet was washed twice with lysis buffer and dissolved in the sample buffer and then subjected to Western blot analysis probed with antibodies against son of Sevenless (SOS; αSOS1; Upstate).

Screening of the efficiency of CPP uptake by prostate cancer cell lines. To determine the uptake of CPP, fluorometer was used to measure the intensity of each CPP per cell basis. Among five commonly used CPPs, R11 had exhibited the highest uptake (at least 6-fold higher than other CPPs) among four prostate cancer cell lines tested (Fig. 1A); in contrast, K11 had the lowest uptake efficiency by prostate cancer cell lines. In summary, the uptake efficiency of five different CPPs tested is R11 ≫ KALA ≥ TAT ≥ PENE ≥ K11 among four prostate cancer cells. Using fluorescence-activated cell sorting technique, we were able to confirm that R11 could give rise to 100% of positive staining cells at the lowest concentration compared with the other four CPPs (data not shown).

Moreover, we determined the uptake rate of R11 (5 μmol/L) uptake by cells. The fluorescence intensity reached plateau at 10-minute incubation for four prostate cancer cell lines (Fig. 1B). The half-maximal intensity can be reached at 5-minute incubation for LNCaP, C4-2, and LAPC4 cells. It seemed that PC3 cell had much rapid uptake rate (i.e., 5-minute) than the other prostate cancer cells (Fig. 1B). By pulsing cells with R11 (5 μmol/L) for 30 minutes, the in vitro half-life of R11 in LNCaP, C4-2, LAPC4, and PC3 was determined as 23.4, 24.1, 22.8, and 24.0 hours, respectively (Fig. 1C). These data conclude that R11 is a rapid delivery vehicle for a variety of prostate cancer cells with its in vitro half-life of ∼23 to 24 hours.

Design of DOC-2/DAB2-derived peptide conjugated with R11. Our previous data indicated that three proline-rich domains in DOC-2/DAB2 have their specific affinity to various SH3-containing proteins (1, 2). Nevertheless, the second proline-rich domain (i.e., PPL) can interact with more SH3-containing proteins. We synthesized R11PPL by conjugating FITC-R11 with PPL as the COOH terminus separated by three glycines. For control peptide R11AAL, all the proline was replaced with alanine.

The uptake of R11PPL and R11AAL in four different prostate cancer cells was examined and the data (Fig. 2A) indicated that the uptake of R11AAL and R11PPL by these cells was very similar. However, R11 seemed to have higher uptake than R11AAL or R11PPL; it may be due to the length of peptide and/or charge/mass ratio. In terms of subcellular localization of R11AAL and R11PPL in prostate cancer cells, both peptides exhibited a similar pattern compared with R11 in these four prostate cancer cells (Fig. 2B); the majority of peptides were localized in the cytosol, with few nuclear staining seen in PC3 cells. Noticeably, in LAPC4 cells, an intense staining of peptides was associated with plasma membrane.

Mechanism of action of R11PPL. To examine whether R11PPL has a similar growth-inhibitory effect as the native protein, both LNCaP and C4-2, DOC-2/DAB2-negative cell lines, were used. As shown in Fig. 3A and B, both serum and dihydrotestosterone were potent mitogens for both cells. In the presence of R11PPL, either serum- or dihydrotestosterone-induced cell proliferation of LNCaP and C4-2 cells was attenuated; in contrast, the control peptide R11AAL did not have any effect. Thus, effect may be due to growth arrest because no apoptosis was detected in R11PPL treated C4-2 cells (Fig. 3C).

It is known that DOC-2/BAD2 can interrupt the interaction of adaptor molecules, such as growth factor receptor bound protein 2 (Grb2) with SOS, which results in inhibiting downstream MAPK activation (1). In addition, the interaction between Grb2 and SOS was inhibited by the presence of native DOC-2/DAB2 protein (Fig. 4A). Consistently, a dose-dependent inhibitory effect was observed in R11PPL but not in R11AAL treatment. These data indicate that R11PPL can behave like its native protein to sequester SH3-containing protein during signal transduction.

To study the effect of R11PPL peptide on MAPK activation, a time course study was carried out using C4-2 cells treated with EGF or serum (Fig. 4B). R11PPL but not R11 or R11AAL treatment could suppress the EGF- or serum-induced ERK2 phosphorylation. In addition, serum-induced ERK2 phosphorylation was also studied in C4-2 cells. The data indicate that the treatment of R11PPL peptide can suppress EGF- or serum-elicited MAPK phosphorylation.

Using small peptide derived from tumor suppressor now becomes an emerging technology for cancer therapy (4). Small peptide can be engineered to mimic a selected function of tumor suppressors, such as p53 (9, 10) and VHL (11). Small peptide can also potentially evade immune system because it is less immunogenic. Moreover, in addition to many practical obstacles of delivering the whole protein into target cells, using peptide derived from a well-characterized functional domain of tumor suppressor gene may also have benefit for not eliciting any other unexpected effects (5). This could be important because DOC-2/DAB2 is a large protein containing many other functional domains that are not fully characterized.

From the structural/functional analysis of DOC-2/DAB2, we have shown that DOC-2/DAB2 can selectively bind to the SH3 domain-containing proteins, such as Grb2, c-Src, Fgr, and Nck but not to Crk or Spectrin (1, 2). PPL seems to be more potent by binding more SH3-containing proteins than the other two proline-rich domains. We also show that the proline-rich domain in DOC-2/DAB2 is involved in suppressing mitogen-elicited MAPK activation (1, 2) often associated with high-grade prostate cancer (12, 13). Thus, we decided to explore the biological activity of PPL in prostate cancer cells first from this study. Our data clearly indicate that the R11PPL but not R11AAL peptide exhibits a similar biological and biochemical activities of its native DOC-2/DAB2 protein in blocking growth factor–elicited signal transduction, such as Grb2 binding to SOS and then downstream MAPK activation (Fig. 4), which resulted in growth arrest but not cell apoptosis in DOC-2/DAB2-negative prostate cancer cells (Fig. 3). Interestingly, R11PPL exhibited similar efficiency of uptake in immortalized normal prostate epithelium or prostate cancer expressing endogenous DOC-2/DAB2 but did not cause any growth-inhibitory effect (Supplementary Data S1). Thus, these data indicate that R11PPL can restore the function of DOC-2/DAB2 in modulating Grb2-SOS-MAPK signal axis in prostate cancer cells with the lost DOC-2/DAB2 expression. Further studying the effect of other two proline-rich domains will provide better understanding of the role of these domains in DOC-2/DAB2 protein function.

Proline-rich domains have been used as a tool to analyze the signal transduction (14). As a ligand with respect to SH3 domain, two classes of proline-rich sequence have been proposed based on alanine-scanning mutagenesis, phage display, combination of chemistry, and high-resolution structure determination. The consensus sequence for class I or II ligand is defined as RxxPxΦP or PxΦPxR (where x is any amino acid and Φ is hydrophobic amino acid), respectively (15). Based on the location of arginine in each class and its binding to the acidic cluster of SH3 domain, it is believed that class I ligand binds SH3 domain in the opposite orientation as class II ligand does. It seems that the PPL sequence (RQPPLVPSR) derived from DOC-2/DAB2 has a unique composition because it contains an overlapping consensus sequence from both classes. Although the functional significance of this unique structure is not completely understood, further study of this structure of PPL peptide should provide better understanding the interaction mode between proline-rich and SH3 domains. For example, it is still unclear that how specific pairwise interaction between proline-rich and SH3 domain can be achieved during each specific signaling transduction, because the binding affinity between proline-rich and SH3 domain is generally low (K_i, 10⁻⁶−10⁻⁴ mol/L). It is possible that PPL can engage simultaneously in multiple interactions with several SH3 domain-containing proteins and such dynamic equilibrium is necessary to elicit a meaningful biological output (15). Nevertheless, PPL should be a good model to test this hypothesis because PPL can bind to more SH3 domains from different proteins than two other proline-rich domains from DOC-2/DAB2 (1, 2).

Polyarginine has been shown as a CPP based on the sequence similarity with TAT and other CPPs (16) with more efficient cell uptake (17). However, the uptake efficiency of polyarginine is highly cell type dependent (8). For example, R11-PPL failed to inhibit the growth of bladder cancer cells without endogenous DOC-2/DAB2 expression, which may be due to the significantly lower uptake of R11 by bladder cancer cells than prostate cancer cells (Supplementary Data S2). These data also suggest that peptide targeting can be further manipulated by certain dose range among different cell type. Nevertheless, R11 seems to be the best efficient delivery system among the peptides tested in several prostate cancer cell lines. The uptake can be detected within 5 minutes in prostate cancer cells shortly exposed to CPP (Fig. 1B), which is consistent with previous finding (18). Although the key delivery mechanism of polyarginine is not well understood, macropinocytosis as well as oligosaccharide composition of cell membrane proteins has been proposed (8). It has been reported that the majority of polyarginine peptides remain inside in cells with a little leak out and they seem to be intact (16). From the in vitro half-life study (Fig. 1C), we have observed a longer half-life in prostate cancer cells compared with other reports (16, 19), indicating that R11 may be more stable in prostate cancer cells than other cell types.

Taken together, this polyarginine delivery system conjugated with a functional motif derived from the native DOC-2/DAB2 protein is a potent growth inhibitor in prostate cancer cells. The outcome of this study provides a unique tool for analyzing signal transduction pathway in prostate cancer and it can be further developed into a therapeutic agent for DOC-2/DAB2-negative prostate cancer.

Grant support: U.S. Army grant DAMD17-03-2-0033.

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 Dr. Hongtao Yu for helpful discussion in the article.