Synuclein-γ Targeting Peptide Inhibitor that Enhances Sensitivity of Breast Cancer Cells to Antimicrotubule Drugs Free

Breast cancer development and progression involve abnormality of multiple genes through genetic and epigenetic alterations. Differential DNA sequencing and in situ hybridization linked the aberrant expression of synuclein-γ (SNCG; ref. 1), also referred breast cancer specific gene 1 (2), to the disease progression of breast cancer. SNCG, along with the synuclein-α (SNCA) and synuclein-β, belongs to a family of intrinsically disordered proteins (3). Recently, it has been recognized that intrinsically disordered proteins are more common than previously thought and can play important biological roles. SNCG mRNA and protein are not expressed in normal breast tissue or tissues with benign breast diseases but are abundantly expressed in a high percentage of invasive and metastatic breast carcinomas (2, 4, 5). A series of functional studies have shown that SNCG expression significantly stimulates proliferation (6–9), invasion, and metastasis of breast cancer cells (10).

SNCG has been shown to interact with BubR1, a mitotic checkpoint kinase required for the prevention of cell mitotic divisions following severe cell damage or mutation (11). In MCF7 breast cancer cells with SNCG overexpression, nocodazole and taxol, the common microtubule inhibitors, have much reduced efficacy. This has been attributed to the SNCG-BubR1 interaction (11, 12). The inhibitory effects of SNCG on checkpoint function are also reduced when the cellular expression level of BubR1 is increased by ectopic overexpression (13). Additional evidence shows that the interaction of BubR1 with CENP-E, which is critically required for the mitotic checkpoint signaling, is impaired in the presence of SNCG. This suggests that SNCG may inhibit the BubR1 function through the interference of BubR1-CENP-E interaction (13). Taken together, it may be hypothesized that BubR1 is a cellular target of SNCG, and the interaction of SNCG-BubR1 may represent a novel mechanism for inactivation of the mitotic checkpoint (11–13).

Currently, microtubule inhibitors are used as the first-line chemotherapeutic agents to treat patients with advanced or metastatic breast cancer (14, 15). However, response rates to this class of drugs vary significantly. These microtubule-disrupting agents are thought to arrest cells in mitosis by triggering the mitotic checkpoint activation, resulting in cells arrested in the mitotic phase without entering anaphase (16). Prolonged treatments with these agents lead to cell death by undergoing apoptosis (17). Because the working mechanism of antimicrotubule drugs relies heavily on the normal function of the mitotic checkpoint machinery in which BubR1 is a critical component (17), the inhibitory effect of SNCG on BubR1 function may explain the induced resistance of breast cancer cells to microtubule inhibitors after exogenous expression of SNCG (18).

In this study, we have designed a novel peptide based on the conserved residues of a single ankyrin-repeat motif and investigated its interaction with SNCG both in vitro and in breast cancer cell lines. We have further examined the molecular mechanism by which the ANK peptide overrides SNCG-mediated drug resistance. Therapeutic application of this peptide may represent a possible future strategy to intervene against overexpression of SNCG in breast cancer and to enhance the effectiveness of anticancer drugs.

Cloning, expression, and purification of SNCG. The SNCG coding region was amplified from the human cDNA clone (Genbank accession no. AF017256) using primers 5′-GCGGATCCATGGATGTCTTCAAGAAGGGC-3′ (sense) and 5′-GAGCGGCCGCAGTCTCCCCCACTCTGGGCCTC-3′ (antisense) and was subsequently subcloned as a BamHI-NotI fragment into the pGEX-4T-3 vector (Amersham Biosciences, Arlington Heights, IL) in the correct reading frame to express the glutathione S-transferase (GST)-SNCG fusion protein. Recombinant GST-SNCG was first purified using GSTrap column as per manufacturer's instructions (Amersham Biosciences) followed by further purification using G-200 gel filtration column on AKTA fast protein liquid chromatography (Amersham Biosciences). Purity of GST-SNCG and SNCG was analyzed using 12% SDS-PAGE (Supplementary Fig. S1). The average yield was ∼22 mg of GST-SNCG and 10 mg of GST cleaved SNCG per liter of Escherichia coli culture.

Nuclear magnetic resonance. Isotopically labeled SNCG was expressed in M9 minimal media containing 1 g/L ¹⁵N-ammonium chloride and purified as described above. Final nuclear magnetic resonance (NMR) samples were in 10 mmol/L sodium phosphate (pH 7), 7.5% D₂O. ¹H-¹⁵N heteronuclear single quantum coherence (HSQC) experiments were done on a Varian Inova 800 MHz spectrometer and processed using NMRPipe (19).

Circular dichroism and fluorescence spectroscopy. ANK peptide was prepared by solid-phase synthesis (Supplementary Methods). SNCG-ANK interaction was monitored using far-UV circular dichroism (CD) recorded on a rapid-scanning monochromator fitted with a CD module (RSM 1000, Olis, Inc., Bogart, GA). Molar ellipticity [𝛉] was calculated according to the formula [𝛉] = [𝛉] × 100 / (nlc), where n represents the number of amino acids in the protein, l represents the path length of the cuvette in centimeters, and c represents the concentration in millimolar. ANS binding experiments were carried out using an LS50B fluorescence spectrometer (Perkin-Elmer, Norwalk, CT). The “phase diagram” method analysis of spectroscopic data, which is extremely sensitive for the detection of intermediate states (20) of protein during binding of ligand, was used to observe association of the ANK peptide with SNCG (Supplementary Methods).

Surface plasmon resonance and isothermal titration calorimetry. Surface plasmon resonance (SPR) experiments were done as described previously (21) using Biacore3000 (Biacore, Inc., Piscataway, NJ), detailed in the Supplementary Methods. Titration calorimetry measurements were done with a VP-isothermal titration calorimetry (ITC) calorimeter (MicroCal, Inc., Northampton, MA). Binding variables, such as the number of binding sites, the binding constant, and the binding enthalpy of bound ligand, were determined by fitting the experimental binding isotherms.

Cloning and expression of enhanced green fluorescent protein–fused ANK. The sense sequence of the 105-bp single-stranded oligonucleotide corresponding to the AA sequence of the ANK peptide with a stop codon was as follows: AAGGGCAACAGTGCCCTTCACGTAGCCTCACAGCATGGCCACCTTGGATGCATACAGACCTTGGTTAGATATGGAGCAAATGTCACCATGCAGAACCACGGGTGA. This single-stranded oligonucleotide was used as the template in a PCR reaction to produce a double stranded DNA fragment with the 5′ primer (AAGGGCAACAGTGCCCTTC) and the 3′ primer (TCACCCGTGGTTCTGCATG). The DNA fragment was cloned into an expression vector pEGFP-C2 to produce an enhanced green fluorescent protein (EGFP)-ANK fusion protein. Cytoplasmic expression of EGFP-ANK fusion protein was confirmed by green fluorescent signals in cells transfected with pEGFP-ANK. The correct molecular mass of EGFP-ANK was verified by Western blotting with anti-GFP antibody.

Detection of ANK-SNCG interaction and mitotic index in breast cancer cell lines. T47D, MCF7-neo, and MCF7-SNCG clones were transfected with pEGFP-ANK plasmid or the control plasmid pEGFP by FuGENE 6 transfection reagent (Roche, Indianapolis, IN). Cell lysate was prepared as described previously (11, 12) followed by coimmunoprecipitation using 10 μg of goat anti-SNCG antibody. Western blotting was done with a monoclonal anti-GFP antibody (Santa Cruz Biotechnology, Santa Cruz, CA; 1:500 dilution) and then subsequently probed with 1:100 dilution of a goat anti-SNCG polyclonal antibody (Santa Cruz Biotechnology). The mitotic index was determined followed by the treatment with 0.5 μmol/L taxol as per standard protocol (11).

GST pull-down assay. GST-SNCG (20 μg) protein or GST alone was added to pre-equilibrated slurry of glutathione-Sepharose 4B and incubated for 60 min. The unbound protein was then washed thrice with PBS containing 1% Triton X-100. MCF-7 cell lysate (1 mg) was mixed with the beads and incubated for 1 h at 4°C. After extensive washes, the beads were collected, boiled in sample buffer, separated on 12% SDS gel, and analyzed by Western blotting as described above.

BubR1-SNCG association in the absence and presence of the ANK peptide. COS7 cells were cotransfected with pCS2-BubR1 (11, 13), pCI-SNCG, and pEGFP-ANK plasmid or with the control plasmid pEGFP by FuGENE 6 transfection reagent (Roche). Two days after transfection, cells were lysed, and immunoprecipitation with anti-SNCG antibody was done as described above. The proteins in immunoprecipitate complexes and total cell lysates were analyzed by Western blotting using a monoclonal anti-Myc antibody (1:500 dilution; BD Biosciences, San Jose, CA), anti-GFP antibody, and anti-SNCG antibody.

Microinjection of the ANK peptide and immunofluorescence microscopy. MCF7-SNCG and MCF-Neo cells as describe previously (11) were grown to 80% confluence, split with trypsin, and seeded on a 22-mm glass coverslip treated with human fibronectin (10 mg/mL) into 60-mm diameter dishes containing cell maintenance medium for 24 h and used for the experiments. To exclude the possibility that microinjection might affect cell cycle and proliferation, MCF7-Neo and MCF7-SNCG cells were microinjected with goat IgG as a control. Injections were done at a constant flow by varying the injection pressure P_c between 40 and 70 units. The microscope stage was kept at 37°C. Approximately 25 to 30 cells were microinjected in each experiment (n = 3). Thirty minutes after microinjection, the coverslips were transferred to medium containing 0.5 μmol/L nocodazole and incubated for additional 24 h. Medium without nocodazole was used for control. In the immunofluorescence experiments, cells were fixed as described (12), and mitotic index determined following standard protocol (11).

Immunostaining of phosphorylated histone (H3). The MCF7-SNCG and MCF7-neo cells cultured on coverslips were transfected by pEGFP or pEGFP-ANK respectively for >24 h and treated with 0.5 μmol/L taxol (Sigma, St. Louis, MO) for 18 h. The mitotic index was calculated according to the proportion of the number of H3 staining GFP-positive cells to total GFP-positive cells.

Flow cytometry. MCF7-SNCG cells were cultured, transfected, and fixed as described previously followed by probing with 2 μg/mL rabbit polyclonal anti-phosphorylated H3 (Ser¹⁰; Upstate, Charlottesville, VA) at room temperature. After two washes with PBS, cells were treated with allophycocyanin-labeled goat anti-rabbit secondary antibody (Molecular Probes, Carlsbad, CA), 50 μg/mL propidium iodide, and 100 μg/mL RNase. Cells were analyzed for cell cycle and anti–phosphorylated H3 population on a FACSCalibur flow cytometer using CellQuest software (Becton Dickinson, San Jose, CA). The mitotic index was calculated according to the proportion of the number of double H3-GFP–positive cells to the total GFP-positive cells.

Design of the single ankyrin peptide. The in vivo interaction of SNCA with multiple ankyrin repeats containing synphilin-1 was shown previously (22). In light of the high sequence homology between SNCA and SNCG (55.9%), it is conceivable that a peptide similar to the ankyrin repeats could interact with SNCG. To investigate this possibility, we designed a 34-residue peptide (KGNSALHVASQHGHLGCIQTLVRYGANVTMQNHG), which we shall call the ANK peptide. The underlined residues in the ANK peptide represent the nine conserved ankyrin residues (Supplementary Fig. S2).

SNCG functions as a natively unfolded protein and interacts with the ANK peptide in vitro. Experiments were first done to confirm the previous observation (3) that SNCG is intrinsically disordered. The ¹H-¹⁵N HSQC spectrum of native SNCG was characterized by narrow line widths and limited ¹H chemical shift dispersion (Fig. 1A). This is indicative of conformational averaging within a rapidly inter-converting ensemble, which shows that SNCG is intrinsically disordered. The far-UV CD spectrum of SNCG was characteristic of an intrinsically disordered polypeptide chain as well, as evidenced by the absence of bands in the 210- to 230-nm region and an intensive minima at ∼196 nm (Fig. 1B). Interestingly, no significant conformational change could be detected in SNCG upon titration of the ANK peptide into SNCG solution (Fig. 1B). Demonstration of SNCG association with the ANK peptide using NMR titrations was inconclusive due to the high concentrations of protein and peptide required and the precipitation of the ANK peptide, which is not highly soluble, under these conditions.

To further probe the association of SNCG with the ANK peptide, we have used several biophysical approaches. We observed the maximal intrinsic tyrosine fluorescence of SNCG at 345 nm, whereas addition of the ANK peptide resulted in a pronounced blue shift in λ_max, analyzed using the method of parametric dependencies (Supplementary Fig. S3). It has been shown that the dependence I(λ₁) = f[I(λ₂)] is linear for the all-or-none transition between two different conformations. Whereas the nonlinearity of this function reflects the sequential character of structural transformations, each linear portion describes the individual all-or-none transition. Supplementary Figure S3 shows that the I₃₁₀ versus I₃₇₀ plot has at least three linear regions, suggesting a complex mechanism of the ANK peptide binding to SNCG involving at least three distinct stages, possibly including de-aggregation of peptide.

ANS fluorescence was used to study the conformational properties of SNCG and the SNCG-ANK interaction (Fig. 2A). Changes in ANS fluorescence intensity are frequently used to monitor formation of partially folded intermediates during protein unfolding and refolding (23) and to analyze the hydrophobic surfaces of proteins (24). Figure 2A shows that SNCG did not interact with ANS, confirming that this protein is intrinsically disordered and does not possess large solvent-exposed hydrophobic patches that are typical of folded conformations. Addition of the ANK peptide induced significant changes in ANS fluorescence properties, including a dramatic increase in the fluorescence intensity and a blue shift in λ_max. This suggests that the ANK peptide can interact with ANS, potentially as an oligomer. Importantly, formation of the SNCG-ANK complex is accompanied by an ∼20% decrease in ANS fluorescence intensity of the only peptide that is affected by the presence of SNCG, suggesting the association between SNCG and the ANK peptide.

Kinetics of the SNCG-ANK interaction. Using immobilized ANK peptide and a GST-SNCG analyte, SPR showed effective binding of the SNCG to the ANK peptide. Figure 2B shows sensograms measured at several GST-SNCG concentrations. Sensograms were deconvoluted globally, using a simultaneous fit for both association (k_a, M⁻¹ s⁻¹) and dissociation (k_d, s⁻¹). Global fits were obtained for the two-phase reaction (conformation change) model with rate constants of k_a1 (2.59 × 10⁻³ M⁻¹ s⁻¹), k_d1 (0.014 s⁻¹), K_a2 (3.32 × 10⁻³ M⁻¹ s⁻¹), k_d2 (8.02 × 10⁻⁴ s⁻¹). Overall, the apparent K_d was found to be 73.4 μmol/L for the SNCG-ANK peptide interaction. There was no significant decrease in analyte binding capacity to the ANK peptide for the duration of the experiments, and the active ANK peptide surfaces were very stable. No binding was observed when GST was used as the analyte (data not shown), which confirmed that SNCG was responsible for the observed interaction between GST-SNCG and the ANK peptide.

ITC was also used to thermodynamically characterize the SNCG-ANK interaction (Fig. 2C). Because the thermodynamic properties of an interaction often reflect its structural characteristics, it was interesting to note an endothermic phase followed by an exothermic phase. No visible precipitation was observed in the titration mixture. The thermodynamics of the peptide titration in GST-SNCG solution exhibited excellent agreement with ideal binding (Fig. 2C), indicating the presence of one high-affinity and multiple low-affinity sequential binding sites. The heat of dilution was measured by titrating the ANK peptide in buffer alone. This was generally small and subtracted from each binding titration curve. A measured K_d of 70 ± 5 μmol/L and ΔH of 2.174 ± 0.2 kcal/mol for the first binding site and in the severalfold lower affinity range for other sequential binding sites were observed. These results were in very good agreement with SPR and fluorescence analysis, validating the observed interaction in solution and in the immobilized form.

ANK peptide binds to SNCG in breast cancer cell lines. To observe the association of the ANK peptide with SNCG in breast cancer cell lines, we cloned nucleotide sequence corresponding to the ANK peptide into pEGFP-C2 vector (pEGFP-ANK) and expressed the ANK peptide as a fusion protein with its NH₂-terminal linked to EGFP. Plasmids pEGFP-ANK and pEGFP were separately transfected into cells. Western blotting using anti-GFP antibody showed specific signals of EGFP as a 33-kDa protein and EGFP-ANK as an ∼37-kDa fusion protein (Fig. 3A). The intracellular localization of EGFP-ANK was shown by strong green fluorescent signals in cytoplasm, which is colocalized with red signals of SNCG (Fig. 3B). To examine the in vivo binding of ANK to SNCG, we did coimmunoprecipitation experiments (Fig. 3C). MCF7-SNCG (lanes 3 and 4) cells were transfected with pEGFP-ANK (lane 3) or the control vector pEGFP (lane 4). In this experiment, MCF7-neo cells were included as negative controls (lanes 1 and 2). Two days after transfection, cell lysates were prepared, and immunoprecipitation with anti-SNCG antibody was conducted. The presence of EGFP-ANK in SNCG immunoprecipitate complexes obtained from different transfected cells was detected by Western blot using anti-GFP antibody. The membrane was subsequently reprobed with anti-SNCG antibody to show equal amounts of SNCG in immunoprecipitate complexes of MCF7-SNCG cell lysates. Figure 3C, (top) shows that when anti-SNCG IP complexes were probed with GFP antibody, only EGFP-ANK but not EGFP alone is found to coprecipitate with SNCG (lane 3) in MCF7-SNCG cells. EGFP-ANK was also not in the immunoprecipitate complexes of neo cells. Figure 3C, (bottom) shows the immunoblotting of anti-SNCG and anti-GFP in total cell lysates. The results of coimmunoprecipitation clearly show the direct intracellular binding of the ANK peptide to SNCG. These results were further validated using GST pull-down assay. We generated and purified GST fusion proteins that contain the full-length (amino acid 127) SNCG (Supplementary Fig. S1). GST pull-down assay was done with purified GST fusion proteins immobilized on glutathione-Sepharose beads and MCF7-neo cell lysate transfected with pEGFP-ANK or pEGFP alone. After intensive washings, proteins bound to the beads were eluted and analyzed by Western blotting using anti-GFP antibody. Figure 3D shows that EGFP-ANK did not bind to GST (lane 3). However, a strong band of EGFP-ANK was detected with GST-SNCG (lane 2). The pull-down assay also showed no interaction of EGFP alone with GST-SNCG (lane 1). Taken together, these results confirmed the specific association of SNCG with the ANK peptide in cell lines.

ANK peptide disrupts SNCG association with BubR1 and releases SNCG-mediated drug resistance. Our previous studies have identified BubR1 as a cellular target of SNCG through yeast two-hybrid and coimmunoprecipitation (11). The binding of SNCG to BubR1 is believed to result in the compromised mitotic checkpoint function as shown by increased resistance of cancer cells to antimicrotubule drugs (11). We were interested to determine whether the binding of ANK to SNCG would directly affect SNCG interaction with BubR1. In further experiments, we transfected COS7 cells with pCS2-BubR1, pCI-SNCG, pEGFP-ANK, or the control vector pEGFP at equal molar ratios. Two days after transfection, cells were harvested for immunoprecipitation with anti-SNCG antibody. The presence of the myc-tagged BubR1 in immunoprecipitate complexes was detected by Western blot using anti-Myc antibody. The membrane was then sequentially probed with anti-GFP antibody and followed with anti-SNCG antibody. Figure 4 (left) shows that when anti-SNCG immunoprecipitate complexes were probed with myc antibody, the myc-BubR1 was found to coprecipitate with SNCG in the absence of EGFP-ANK, and myc-BubR1 signal was not detected in the immunoprecipitate complex of cells expressing the ANK peptide. In contrast, anti-GFP antibody only detected the GFP signal in the immunoprecipitate complex obtained from cells expressing EGFP-ANK and not from cells expressing GFP only. Figure 4 (right) illustrates the results of Western blots using total cell lysates, showing similar expression levels of these proteins. These results clearly indicate that the binding of the ANK peptide to SNCG interrupted the interaction of SNCG with BubR1.

To determine whether the ANK peptide mediated disruption of SNCG-BubR1 association in vivo released the inhibition of mitotic arrest, we used microinjection of the ANK peptide into the stable breast cancer cell line MCF7 expressing SNCG (MCF7-SNCG) and the control SNCG-negative MCF7-Neo cell line. Microinjection efficiency and viability were excellent when cells were injected with goat IgG as a control, showing that background effects of microinjection and antibody were minimal in both cell lines. Significant numbers of cells survived after treatment with the ANK peptide by microinjection in both cell lines (Supplementary Fig. S4). After microinjection, cells were treated with 0.5 μmol/L nocodazole for 24 h, and the fraction of mitotically arrested cells was counted after 4′,6-diamidino-2-phenylindole (DAPI) staining. As shown in Fig. 5, when the cells without ANK injection were treated with nocodazole, MCF7-Neo cells exhibited a significantly higher degree of mitotic arrest compared with MCF7-SNCG cells. In contrast, injection of the ANK peptide into MCF7-SNCG cells significantly (P < 0.05) increased the level of mitotically arrested cells to a level comparable with that in MCF7-Neo cells (Fig. 5B , gray columns). These results are indicative of the ANK peptide functioning as a SNCG inhibitor.

To further show the function of the ANK peptide in restoring the impaired mitotic checkpoint control in SNCG-expressing cells, MCF7-neo, MCF7-SNCG, and T47D cells were transfected with pEGFP-ANK or the control vector pEGFP. Two days after transfection, cells were exposed to taxol for 18 h, and numbers of mitotic arrested and apoptotic cells were counted after DAPI staining. Expression of EGFP-ANK in neo cells did not significantly alter the mitotic index or the percentage of apoptotic cells. However, expression of this peptide in SNCG cells consistently caused an ∼2-fold increase in the number of mitotic arrested cells and a 5-fold increase in apoptotic cell populations (Fig. 6A). These cells were also probed with rabbit polyclonal anti–phosphorylated H3, and the mitotic index was calculated according to the proportion of the number of H3-stained EGFP-positive cells to total EGFP-positive cells (Fig. 6B). This result corroborated with increases in sensitivity to taxol by overexpressing EGFP-ANK in MCF7-SNCG and T47D cells shown by DAPI-stained cells. Similar results were achieved using flow cytometric analysis of EGFP-ANK containing MCF7-SNCG cells treated with 0.5 μmol/L taxol. Cell cycle analysis of MCF7-SNCG cells transfected with EGFP-ANK revealed almost a 4-fold increase in apoptotic/dead cell population in comparison with cells transfected with the control vector pEGFP (Fig. 6C and D). Furthermore, anti–phosphorylated H3 staining showed a 3.5-fold increase in positive phosphorylated histones in comparison with MCF7-SNCG cells not containing EGFP-ANK (Fig. 6C and D). Taken together, these results confirmed our finding that overexpression of EGFP-ANK improves the sensitivity of MCF7 breast cancer cells to antimicrotubule drug treatment.

The exogenous expression of oncogenic protein SNCG in breast cancer cells has been linked to stimulated proliferation, increased cell invasion and metastasis, and resistance to nocodazole-elicited apoptosis (6–8, 11, 13). One of the molecular mechanisms underlying the oncogenic functions of SNCG is its interaction with BubR1 and the consequential inhibition of BubR1-mediated mitotic checkpoint functions (11, 13). SNCG belongs to a family of intrinsically disordered proteins, which have a wide range of significance in living systems, but their structural and functional characteristics are still poorly understood (3, 25–29). Our CD and NMR results confirm the previous observations that SNCG is intrinsically disordered (3). In the present study, we have designed and characterized a 34-amino-acid ANK peptide based on the conserved residues of single ankyrin motif, which is the repeating unit of multiple ankyrin domain known as the protein-protein interaction module (30).

Indeed, we observed the in vitro SNCG-ANK interaction using a number of biophysical approaches. Interestingly, upon interaction, far UV CD showed that SNCG undergoes no significant conformational changes and remains disordered. This contrasts to the coupled binding and folding that has often been observed for other intrinsically disordered proteins (31). Similar to the SNCG-ANK interaction is the case of the calmodulin-binding fragment of caldesmon (CaD136), which binds a single calmodulin molecule with relatively high affinity while undergoing only very moderate folding (32). These SNCG and calmodulin studies represent a newly recognized class of intrinsically disordered proteins that are able to bind their ligands while remaining disordered and affirm the biological relevance of intrinsic disorder. Intrinsic tyrosine fluorescence provides a convenient means of monitoring the binding between the peptide and SNCG. The fact that the addition of the ANK peptide was accompanied by a pronounced blue shift in the fluorescence emission spectra, which reflects the normalization of tyrosine fluorescence, can be attributed to the distortion in the residual SNCG structure induced by binding with the ANK peptide. In SNCA, paramagnetic relaxation enhancement NMR experiments have shown that there are similar long-range residual tertiary contacts involving the charged COOH-terminal region and the middle section of the protein (33, 34). These observations lead to conclusion that binding of the ANK peptide did not induce folding of SNCG, suggesting that SNCG can function biologically in its intrinsically disordered state. Furthermore, SPR results provide the direct evidence of the SNCG-ANK interaction (Fig. 2A), which suggest a good fit to bivalent analyte model. For ITC, deconvolution of isotherm data indicates that the observed heat peaks are the results of cumulative contributions from bindings to at least two thermodynamically distinct binding sites (Fig. 2B). In fact, it is not unusual to observe simultaneous endothermic and exothermic reaction while protein has binding site in two different states (35, 36). Importantly, the analysis of binding-induced changes in tyrosine fluorescence using the phase diagram method suggests a mechanism of the ANK peptide binding to SNCG in which the complex formation involves at least three distinct stages (Supplementary Fig. S3).

We investigated the cellular effects of the SNCG-ANK interaction using the endogenously SNCG-overexpressing cells T47D, MCF7-SNCG stable expressing cell line, and its control cell line MCF7-Neo. When cell lines were treated with nocodazole or taxol, MCF7-SNCG and T47D displayed a much lower mitotic index than MCF7-Neo, indicating an impaired mitotic checkpoint function. However, in the presence of synthetic ANK peptide microinjected into the cells, MCF7-SNCG cells exhibited similar response to nocodazole as MCF7-Neo cells (Fig. 5), indicating that the inhibitory effects of SNCG on mitotic checkpoint function was released by the ANK peptide. We have recently shown similar findings when SNCG expression was knocked down using antisense RNA (12), providing independent support for the notion of ANK-mediated disruption of SNCG complex leading to breast cancer cell sensitivity to taxol-induced mitotic arrest. To understand the molecular mechanism of ANK function, we have further shown that the binding of the ANK peptide to SNCG prohibited the interaction of SNCG with BubR1, thereby releasing the inhibitory effect of SNCG on BubR1-mediated mitotic checkpoint function. This is a very important observation as various types of cancers have mitotic checkpoint defects that in some cases could be caused by silencing of BubR1 through mutations (37). BubR1 can prevent the uncontrolled cell division observed in highly infiltrating breast cancers and therefore is a desired target for controlling these cancers, whereas cells not expressing BubR1 are able to override mitotic checkpoint controls and continue to progress through cell cycle (38). Because BubR1 is a critical component of the mitotic checkpoint control, it could be a potential cellular target of oncogenic proteins, such as SNCG, that induce tumor progression. Given that the SNCG-BubR1 interaction is the shown leading factor that overrides the effect of nocodazole and taxol (11, 12), minimization of this interaction would be useful. Using colocalization, coimmunoprecipitation, and GST pull-down experiments, we concluded that EGFP-ANK binds to the SNCG intracellularly (Fig. 3). Furthermore, we show that when expressed in MCF7-SNCG cells, EGFP-ANK, but not the EGFP alone, restored the mitotic checkpoint function, whereas expressing EGFP-ANK in MCF7-neo cells had little effect on mitotic index or the percentage of apoptotic cells (Fig. 6). These findings suggest that the ANK peptide functions as a SNCG inhibitor.

Although we cannot exclude the possibility that the ANK peptide may also interact with SNCA, this scenario is either unlikely or has minimal effect in the breast cancer. In both MCF7-SNCG and MCF7-Neo cells, alteration of background level of SNCA expression has not been reported. The fact that MCF7-Neo cells are not affected by ANK peptide suggests the lack of any detectable involvement of SNCA. Taken together, we believe that SNCG is the molecular target for the ANK peptide, which can be directly related to the mitotic arrest mediated by nocodazole and taxol. The observed ANK peptide binding with SNCG and the resultant competition with the SNCG-BubR1 interaction in cell lines represent a novel mechanism by which cancer cells increases sensitivity to antimicrotubule drugs.

In conclusion, we have designed and characterized a novel peptide based on a single ankyrin motif. The peptide associates with SNCG, which remains disordered. To our best knowledge, this is the first report that a single ankyrin motif has been shown to interact with any protein, in contrast to commonly observed interactions involving multiple ankyrin repeats. The interaction overrides the effects of elevated SNCG levels, by competing with the SNCG-BubR1 interaction, and results in the nocodazole- and taxol-mediated mitotic arrest. This finding may be useful in devising future strategies to intervene against overexpressed SNCG in breast cancer and enhance the effectiveness of anticancer drugs. In light of our recent new findings showing prominent expressions of SNCG in eight different types of human cancers (39), SNCG expression status may have a broad effect in patient responses to antimicrotubule chemotherapy, and the ANK peptide will be an important molecular tool to study the tumorigenic functions of SNCG in a wide range of human cancers in addition to breast cancer.

An NMR study of SNCG has been recently published (40).

Grant support: Canadian Institutes of Health Research (Z. Jia), Department of Veterans Affairs, Office of Research and Development, Medical Research Service, U.S. Army Medical Research and Material Command grants breast cancer 010046 and breast cancer 033154 (J. Liu), and Natural Sciences and Engineering Research Council of Canada Postgraduate Scholarship (J.A. Marsh).

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. Steve Smith (Queen's University) for critical reading of the article and Kim Munro (Protein Function Discovery Facility, Queen's University) for his technical support.