Disruption of SND1–MTDH Interaction by a High Affinity Peptide Results in SND1 Degradation and Cytotoxicity to Breast Cancer Cells In Vitro and In Vivo

Breast cancer is the most common cancer in women, causing 626,679 deaths worldwide in 2018 (1). Staphylococcal nuclease domain-containing protein 1 (SND1) is an oncoprotein found to be overexpressed in breast, prostate, lung, colorectal carcinomas, hepatocellular carcinomas, and malignant glioma, especially in advanced and metastatic cases (2–4). SND1 downregulation was found in renal cell carcinoma (5).

SND1 is a multifunctional protein with multiple interacting partners (2). SND1 is an endonuclease and a transcriptional cofactor that binds to metadherin (MTDH), a proposed oncogene regulated by c-myc (6). Overexpression of MTDH has been found in many cancers, including breast, prostate, liver, and esophageal cancers. Functional studies suggest that MTDH could promote the tumorigenic and metastatic potential of tumor cells through regulating multiple signaling pathways, including PI3K/Akt, NF-κB, Wnt/β-catenin, and MAPK pathways (7). Interaction of SND1 with MTDH has been demonstrated to contribute to the initiation and progression of breast cancer, making this interaction a promising target for breast cancer treatment (8). Cocrystal structure of SND1-MTDH complex (PDB: 4QMG) has been solved, revealing the important residues of the binding interface. A 22-mer MTDH peptide (amino acid 386–407, SSADPNSDWNAPAEEWGNWVDE) was determined to be the minimum binding motif to interact with SN1/2 domain of SND1. W394 and W401 within this 22-mer MTDH were found to be the essential amino acids in the interaction with SND1 (9). In breast cancer, knockout of MTDH, knockdown of SND1, or disrupting their interaction by mutation of the two important tryptophan compromised tumorigenic potential of breast cancer cells in vivo (8).

The effect of SND1 inhibitors in liver cancer has also been investigated. Deoxythymidine 3′,5′-bisphosphate (pdTp) was a strong competitive inhibitor of staphylococcal nuclease. pdTp showed a specific inhibition of the nuclease activity of SND1 (10) and it could inhibit the proliferation of liver cancer cell in vitro (11) and tumor growth in vivo (12).

The goal of this study was to discover novel peptides that can disrupt SND1–MTDH interaction using phage display technology and to test whether these peptides can inhibit breast cancer growth. SN1/2 domain (amino acid 16–339) of SND1 is reported to be responsible for interacting with MTDH (9). Here, SN1/2 domain of SND1 was used as bait to discover SND1-interacting peptides. Cytotoxicity of SND1-interacting peptide was tested after fusing it to a cell-penetrating peptide. The possible mechanisms of the cytotoxicity were also investigated. The effect of SND1-interacting peptides was also tested on breast cancer xenograft mouse model.

The cDNA encoding full-length SND1 was purchased from DNASU (https://dnasu.org/DNASU/). The cDNA encoding SN1/2 fragment was obtained by PCR with 5′ primer: CGGGATCCGTCCCCACCGTGCAGCGGGGCA and 3′ primer: CCGGAATTCTTACTTTTGGTCCAAATTAGCTGTG. SN1/2 cDNA fragment was cloned into a modified version of pET24b, which was cloned with GST-6xHis-HRV-3C protease cleavage site (a kind gift of Prof. Yanxiang Zhao, Department of Applied Biology and Chemical Technology, the Hong Kong Polytechnic University, Hong Kong). Escherichia coli (E. coli) BL21-competent cells were transformed with the recombinant plasmid, cultured in LB medium supplemented with 50 μg/mL kanamycin, and induced with 0.2 mmol/L IPTG at 26°C to express GST-6xHis-SN1/2 recombinant protein.

GST-6xHis-SN1/2 protein was purified by HisTrap Column (GE Healthcare Life Sciences) followed by digestion with human rhinovirus 3C protease (self-prepared) at 4°C overnight. The digested protein mixtures (SN1/2 and GST-His tag) were loaded onto HisTrap column and SN1/2 protein was eluted by Tris buffer (50 mmol/L Tris, 150 mmol/L NaCl, 10 mmol/L KCl, and 10 mmol/L MgCl₂, pH 8.0).

The procedure of phage display screening was modified on the basis of the manufacturer's instructions (#E8110S, NEB). A petri dish (90 × 15 mm, Biologix) was coated with SN1/2 before screening. The dish was incubated with 100 μg/mL SN1/2 protein at 4°C overnight prior to blocking with 0.1 mol/L NaHCO₃ (pH 8.6) containing 5 mg/mL BSA at 4°C for 1 hour. Phage display library was then added onto the SN1/2-coated petri dish and incubated at room temperature for 1 hour. Unbound phages were removed by washing with TBST [50 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, and 0.1% (v/v) Tween-20]. Bound phages were eluted with glycine-HCl (pH 2.2) and amplified. Phages with high binding affinity to SN1/2 were harvested after four rounds of biopanning. Tween-20 was increased to 0.5% (v/v) in TBST washing buffer in rounds 2–4.

For phage ELISA, 96-well plate (Nunc 96-well optical-bottom plate) was coated with 100 μg/mL SN1/2 protein at 4°C overnight prior to blocking with 0.1 mol/L NaHCO₃ (pH 8.6) containing 5 mg/mL BSA at 4°C for 1 hour. Serial dilutions of single phages were added onto the plate for interaction. After incubation at room temperature for 2 hours with shaking, the plate was washed, followed by incubation with horseradish peroxidase (HRP)-conjugated anti-M13 mAb at 0.2 μg/mL (Sino Biological, #11973-MM05T, RRID:AB_285792) in blocking buffer at room temperature for 1 hour with agitation. The plate was washed before the addition of HRP substrate, ABTS (Thermo Fisher Scientific, #34026) in 50 mmol/L sodium citrate, pH 4.0. Thirty percent of H₂O₂ was used as stop solution. Color product was measured at OD₄₁₀.

For peptide ELISA, 96-well plate (PerkinElmer, Optiplate-96F) was coated with 10 μg/mL SN1/2 protein at 4°C overnight prior to blocking. A serial dilution of peptides in binding solution (0.02% TBST) was added onto the plate. After incubation for 1 hour at 37°C, the plate was washed with washing buffer (0.02% TBST) before fluorescence detection using CLARIOstar microplate reader (λ_excitation = 485 nm and λ_emission = 535 nm). For competitive peptide ELISA, SN1/2-coated plate was preincubated with serial dilutions of competitive peptides for 1 hour at 30°C, followed by washing. 22-mer 5-FAM-MTDH peptide (3.5 μmol/L) together with serial dilutions of competitive peptides was then added onto the plate for competition for 1 hour. Plate was washed before fluorescence detection.

Peptides were synthesized by Pepmic using standard solid-phase peptide synthesis chemistry and purified typically to 95% purity as the acetate form.

5-FAM-MTDH peptide (5-FAM-SSADPNSDWNAPAEEWGNWVDE) and 5-FAM-CPP-4-2 peptide (5-FAM-RRRKKRRQRRRQFDYDHFLMWYS) were synthesized with 5-FAM fluorophore labeled on the N-terminus of the peptides. Biotinylated 4-2 peptide was synthesized by biotinylation on the N-terminus of peptide 4-2. Other peptides were synthesized with amidation on the C-terminus and acetylation on the N-terminus.

Stable transfection of MDA-MB-231-GFP-Red-FLuc cell line with pCMV6-Entry-SND1-cMyc-DDK plasmid (#RC200059, OriGene) was conducted using Lipofectamine 3000 Reagent (Thermo Fisher Scientific) according to the instructions of the manufacturer. After transfection, cells were selected with G418 at a concentration of 2 mg/mL for 2 weeks.

MDA-MB-231-GFP-Red-FLuc cell lysate was prepared with modified RIPA lysis buffer (10 mmol/L Tris, 75 mmol/L NaCl, and 0.5% NP40, pH 7.4) supplemented with 1 mmol/L PMSF and 1 × Protease Inhibitor (#11836153001, Roche). For coimmunoprecipitation (co-IP) assay, anti-MTDH antibody (Thermo Fisher Scientific, #40-6500, RRID:AB_2533475) and rabbit IgG isotype control antibody (Abcam, catalog no., ab171870, RRID:AB_2687657) were preincubated with protein G agarose (sc-2002) for 3 hours before co-IP. Peptide 4-2 was added to the cell lysate for preincubation for at least 2 hours before immunoprecipitation. For peptide pull-down assay, biotinylated 4-2 peptide was preincubated with Streptavidin Agarose Breads (Thermo Fisher Scientific) for 3 hours before pull-down. Competitive peptides were added to the cell lysate for preincubation at 4°C for at least 2 hours before pull-down. Co-IP and pull-down assay were conducted by mixing the treated beads and cell lysates overnight at 4°C with rotation.

Isothermal titration calorimetry (ITC) was performed using a microCal PEAG-ITC (MicroCal Inc.). To detect the interaction between protein SDN1 and peptide 4-2, 400 μmol/L of SND1(16-339) in the syringe was titrated into the sample cell containing 40 μmol/L 4-2 peptide. All samples were prepared in a buffer containing 50 mmol/L Tris, pH 8.0, and 150 mmol/L NaCl. The titration consisted of 20 injections of 2 μL, with 200 seconds equilibration between injections. The data were analyzed using Origin 7.0 (Origin, RRID:SCR_014212).

Western blot analysis was conducted according to the common protocol. The primary antibodies (1:1,000–3,000) used were anti-MTDH antibody (Thermo Fisher Scientific, #40-6500, RRID:AB_2533475), anti-SND1 antibody (sc-271590), anti-β-actin antibody (Santa Cruz Biotechnology, sc-47778, RRID:AB_626632).

All animal studies were approved by Animal Subjects Ethics Subcommittee (no. 13/32) and conducted at the Central Animal Facilities of The Hong Kong Polytechnic University (Hung Hom, Hong Kong).

A total of 18 BALB/c nude mice (4–6 weeks old female, with body weight of 16–20 g, Charles River Laboratories) were subcutaneously inoculated with 5 × 10⁶ of MDA-MB-231-GFP-Red-FLuc cells in 1:1 PBS/Matrigel (Corning). Tumors were allowed to develop to the size of around 100 mm³ before mice were randomly grouped into three groups (n = 6). Different groups of mice were intraperitoneally injected with 50 mg/kg of CPP-4-2 peptide [molecular weight (MW) 3,326 g/mol], 45.5 mg/kg of CPP-4-2AAA peptide (negative control peptide, MW 3,027 g/mol), or PBS, respectively, once daily for 30 days. Tumor volumes were measured every 2 days using vernier calipers. Tumor volume was calculated according to the formula: V = 0.5 × length × width × width. Tumors were excised and tumor weight was measured at the end of the experiment. Statistical analysis was done using two-way ANOVA or Student t test as indicated in the Results.

During in vivo efficacy study, toxicity symptoms, like loss of appetite, slowness in activity, or treatment-related mortality, were recorded. A body weight loss of more than 15% for 3 consecutive days was considered as treatment-related toxicity and the mice would be euthanized.

SN1/2 domain (amino acid 16–339) of SND1 was used as bait for phage display screening. SN1/2 domain of SND1 was expressed in E. coli and purified as described in Materials and Methods. A single band of approximately 40 kDa was found after purification, indicating the purity of SN1/2 (Fig. 1A, left). The monomeric status of SN1/2 domain was confirmed by size exclusion chromatography-multi-angle light scattering (SEC-MALS; Fig. 1A, right). A major peak with average MW ranging from 20 to 50 kDa was found, suggesting that there was no aggregation of SN1/2 (40 kDa).

After four rounds of biopanning using SN1/2 as bait in phage display screening, phages were isolated and peptide sequences of the phages were determined. A total of 20 phages were picked and sequenced (Fig. 1B). Sequence of peptide 4-1 was repeated in 4-18. Sequence of peptide 4-2 was repeated in 4-3, 4-4, 4-7, 4-9, and 4-10. Interestingly, an unusually high percentage of tryptophan (W, 14.2%) and tyrosine (Y, 11.3%) was observed in the peptide sequences. Tryptophan and tyrosine are relatively rare amino acid in vertebrates with observed frequencies of 1.3% and 3.3%, respectively (13, 14). Phage ELISA was performed to confirm the binding of phages toward SN1/2. Phage 4-2, 4-8, 4-13, 4-16, and 4-19 showed SN1/2 binding in phage ELISA study (Fig. 1C).

SND1 as a multifunctional protein has multiple interacting partners (2). The interaction between SND1 and one of its partner proteins, MTDH, was shown to be important in breast cancer initiation and progression (8). Here, SND1–MTDH interaction was studied using a 22-mer fluorescently labeled 5-FAM-MTDH peptide (5-FAM-SSADPNSDWNAPAEEWGNWVDE; ref. 8). 5-FAM-MTDH peptide showed dose-dependent binding to SN1/2 (Fig. 1D). Peptide 4-2 and 4-8 could compete with 5-FAM-MTDH peptide binding to SN1/2, with EC₅₀s of 52.1 ± 8.4 and 12.3 ± 2.8 μmol/L, respectively (Fig. 1E). An MTDH 11-mer peptide composed of the middle part of MTDH 22-mer peptide (9) was also included in this competition. MTDH 11-mer peptide was less effective in competing with 5-FAM-MTDH 22-mer binding to SN1/2 compared with peptide 4-2 and 4-8 (Fig. 1E). Other peptides could not compete with 5-FAM-MTDH 22-mer binding to SN1/2 (Supplementary Fig. S1). This result suggested that peptide 4-2 and 4-8 could disrupt MTDH 22-mer peptide from binding to SN1/2, possibly by binding to the SN1/2-MTDH 22-mer–interacting region. Peptide 4-2 and 4-8 were better competitors of SN1/2–MTDH 22-mer interaction compared with MTDH 11-mer peptide. Peptide 4-8 precipitated heavily in tissue culture medium after fusion with cell penetrating peptide (CPP), so it was suspended from this stage.

We investigated the interaction of peptide 4-2 and SN1/2 by using ITC. The binding affinity between peptide 4-2 and SN1/2 was determined to be 3.1 μmol/L by ITC (Fig. 2A).

To investigate the binding of peptide 4-2 to SND1 and the essential amino acids of peptide 4-2 in SND1 interaction, biotinylated peptide 4-2 and three mutants of peptide 4-2 with potential critical amino acids (tryptophan and tyrosine) changing to alanine were synthesized (Fig. 2B, left). Tryptophan and tyrosine were selected for mutagenesis because they were unusually abundant in SN1/2-interacting peptides (Fig. 1B). Pull-down assays of SND1 by biotinylated 4-2 peptide was conducted with or without the competing peptides. Figure 2B, right, shows that biotinylated peptide 4-2 could pull-down SND1. Peptide 4-2, 4-2Y4A, and 4-2Y11A could compete for the pull-down of SND1 by biotinylated 4-2, whereas 4-2W10A could not. This result suggested that peptide 4-2 could interact with SND1 and W10 is an essential amino acid of peptide 4-2 in interacting with SND1, whereas Y4 and Y11 are not (Fig. 2B, left).

Next, we investigated whether peptide 4-2 could also disrupt SND1-MTDH complex. co-IP experiment showed that MTDH could pull-down SND1 (Fig. 2C, left). Treatment with peptide 4-2 could disrupt SND1–MTDH interaction, whereas treatment with peptide 4-2W10A could not (Fig. 2C, left). The quantitation of the co-IP result is shown in Fig. 2C, right. This result suggested that SND1 could interact with MTDH and peptide 4-2 could disrupt SND1–MTDH interaction with W10 as an essential amino acid.

To investigate the effect of peptide 4-2 in cytotoxicity, a cell permeable 4-2 peptide (CPP-4-2, RRRKKRRQRRRQFDYDHFLMWYS) was constructed by fusing RR-TAT (15), a hybrid CPP, at the N-terminus of peptide 4-2 to facilitate its penetration into mammalian cells.

We then investigated the expression of SND1 in different breast cancer cell lines (Fig. 3A). All the three breast cancer cell lines tested showed expression of SND1 and MTDH.

CPP-4-2 peptide exhibited cytotoxicity to breast cancer cell lines MDA-MB-231-GFP-Red-FLuc, MCF7, and MDA-MB-468, with an IC₅₀ value of 22.4 ± 1.0, 18.7 ± 0.2, and 15.9 ± 6.2 μmol/L, respectively (Fig. 3B and C). Comparable amount of 5-FAM-CPP-4-2 was detected in MDA-MB-231-GFP-Red-FLuc and MCF7 cells, whereas MDA-MB-468 accumulated around 2-fold more 5-FAM-CPP-4-2 (Supplementary Fig. S2). To summarize, CPP-4-2 could kill breast cancer cells in vitro.

We then investigated whether tryptophan and tyrosine of CPP-4-2 were also important in the cytotoxicity to breast cancer cells. We also included a triple alanine–mutant CPP-4-2AAA, where Y4, W10, and Y11 were all mutated to A as a negative control. CPP-4-2Y4A (IC₅₀ = 23.9 ± 2.5 μmol/L) and CPP-4-2Y11A (IC₅₀ = 22.2 ± 1.6 μmol/L) were equally cytotoxic as CPP-4-2 (IC₅₀ = 23.6 ± 2.5 μmol/L) toward MDA-MB-231-GFP-Red-FLuc cells (Fig. 3D and E). However, CPP, CPP-4-2W10A, and CPP-4-2AAA were not cytotoxic, with IC₅₀s > 40 μmol/L (Fig. 3D and E). This result suggested that W10 was an essential amino acid of CPP-4-2 in its cytotoxicity to MDA-MB-231-GFP-Red-FLuc cells, whereas Y4 and Y11 were not.

We investigated the mechanism by which CPP-4-2 exerted its cytotoxicity to breast cancer cells. As shown in Fig. 4A, CPP-4-2 downregulated SND1 of three breast cancer cell lines, MDA-MB-231-GFP-Red-FLuc, MCF7, and MDA-MB-468. It is reported that in breast cancer cells, MTDH could stabilize SND1 through their interaction. MTDH could no longer protect SND1 from degradation when MTDH–SND1 interaction was disrupted (8). The CPP-4-2–mediated degradation of SND1 in breast cancer cells (Fig. 4A) may be due to the disruption of SND1–MTDH interaction by peptide 4-2 (Fig. 2C). The effect on SND1 by the mutants of CPP-4-2 was also investigated. CPP-4-2, CPP-4-2Y4A, and CPP-4-2Y11A could downregulate SND1, whereas CPP-4-2W10A and CPP-4-2AAA showed no effect on SND1 downregulation (Fig. 4B, left and right). CPP-4-2 peptide and its mutants showed no effect on MTDH expression (Supplementary Fig. S3).

SND1 was overexpressed in MDA-MB-231-GFP-Red-FLuc cells (Fig. 4C). SND1 overexpression was found to be able to partially reverse CPP-4-2–induced cytotoxicity to MDA-MB-231-GFP-Red-FLuc cells (Fig. 4D), suggesting that the CPP-4-2–induced cytotoxicity to breast cancer cells was mediated by the degradation of SND1. We tried to overexpress MTDH in MDA-MB-231-GFP-Red-FLuc cells. The overexpression of MTDH in this cell line was not significant. We could not investigate whether MTDH overexpression could protect the cell from the cytotoxicity induced by CPP-4-2 peptide.

The activities of peptide 4-2 and its mutants are summarized in Fig. 4E. Peptide 4-2 could kill breast cancer cells, interact with SND1, disrupt SND1–MTDH interaction, and induce SND1 degradation. W10 was an essential amino acid in all these activities.

Because CPP-4-2 could kill breast cancer cell in vitro, we investigated the efficacy of CPP-4-2 peptide in breast cancer xenograft mouse model. MDA-MB-231-GFP-Red-FLuc cells were inoculated subcutaneously into BALB/c nude mice. CPP-4-2 peptide could inhibit the growth of breast cancer cells in xenograft mouse model when compared with PBS solvent control and CPP-4-2AAA negative control (Fig. 5A). No body weight loss was observed during the treatment period of 30 days (Fig. 5B). No toxic symptoms, like loss of appetite, slowness in activity, or treatment-related mortality, were observed during treatment period. Tumors were excised at the end of the experiment. Tumor volume in CPP-4-2 group was significantly smaller than that in PBS or CPP-4-2AAA group (Fig. 5C). Tumor weight was also found to be significantly lower in CPP-4-2 group compared with CPP-4-2AAA and PBS group (Fig. 5D). The expression of SND1 in the tumors derived from MDA-MB-231-GFP-Red-FLuc cells was investigated. As shown in Fig. 5E, left and right, CPP-4-2 could degrade SND1 in vivo, whereas CPP-4-2AAA and PBS could not.

Importantly, there was a clear association between SND1 degradation and tumor volume in CPP-4-2 group, strongly suggesting that the tumor inhibition by CPP-4-2 peptide was due to SND1 degradation (Fig. 5C and E). Tumor number 5 in CPP-4-2 group was the smallest in volume and with the highest amount of SND1 degradation. To summarize, CPP-4-2 peptide could inhibit breast cancer growth in vivo possibly by degrading SND1, whereas CPP-4-2AAA could not.

After four rounds of screening, an unusually high percentage of tryptophan and tyrosine (Fig. 1B) was displayed in the randomly picked single-phage clones. Tryptophan and tyrosine are rare amino acids in nature (13, 14, 16). Two important tryptophans in MTDH (W394 and W401) were found to occupy the binding pockets of SND1-MTDH interface (9). These two tryptophans in MTDH were critical in SND1–MTDH interaction and also in promoting breast cancer initiation and progression (9). One could speculate that the tryptophans found in SN1/2-interacting peptides may play a similar role as the tryptophans in SND1–MTDH interaction. Peptide 4-2 may be an MTDH-like peptide that could interact with SND1 and disrupt SND1–MTDH interaction.

Peptide 4-2 was demonstrated to disrupt SND1–MTDH interaction in both ELISA and co-IP assays. Mutagenesis study suggested that W10 of peptide 4-2 was essential in mediating the cytotoxicity to breast cancer cells, SND1 interaction, SND1-MTDH disruption, and SND1 degradation. These results suggested that peptide 4-2 may be an MTDH-like peptide that could interact with SND1 and disrupt SND1–MTDH interaction, resulting in SND1 degradation and breast cancer cell death. W10 was essential in all the above activities of peptide 4-2.

MTDH is reported to support the survival of breast cancer cells by interacting with and stabilizing SND1. When SND1–MTDH interaction was disrupted, MTDH could no longer protect SND1, which resulted in the degradation of SND1 and reduced cell survival (8).

Here, we have used phage display to identify an SN1/2-interacting peptide 4-2, and this peptide could disrupt SND1–MTDH interaction in ELISA and co-IP assays. CPP-4-2 peptide could degrade SND1 in breast cancer cells both in vitro and in vivo. Overexpression of SND1 could reduce CPP-4-2–mediated cytotoxicity in breast cancer cells. These results suggested that the degradation of SND1 induced by CPP-4-2 peptide was the reason for breast cancer cell death. CPP-4-2 peptide mediated cytotoxicity to breast cancer cells possibly by inducing SND1-MTDH disruption and subsequent degradation of SND1. W10 was an essential amino acid in the activities of CPP-4-2 peptide (Fig. 6).

In summary, we have discovered a high affinity peptide 4-2, by phage display screening, that can interact with SND1 and this peptide 4-2 induced cytotoxicity to breast cancer cells both in vitro and in vivo possibly through interacting with SND1, disrupting SND1–MTDH interaction, and resulting in SND1 degradation. W10 was an essential amino acid in the activities of peptide 4-2.

No disclosures were reported.

P. Li: Conceptualization, formal analysis, methodology, writing-original draft, writing-review and editing. Y. He: Investigation, methodology. T. Chen: Investigation, methodology. K.-Y. Choy: Investigation, methodology. T.S. Chow: Investigation, methodology. I.L.K. Wong: Investigation, methodology. X. Yang: Methodology. W. Sun: Investigation, methodology. X. Su: Investigation, methodology. T.-H. Chan: Conceptualization, formal analysis, project administration, writing-review and editing. L.M.C. Chow: Conceptualization, formal analysis, supervision, funding acquisition, investigation, project administration, writing-review and editing.

Sincere thanks are given to the University Research Facility in Life Sciences of PolyU for their kind support in experimental techniques and equipment. Sincere thanks are also given to PolyU studentship for their nice and generous support for this project. This project was partly supported by the Area of Strategic Importance grant from the Hong Kong Polytechnic University (1-ZE22) and Collaborative Research Fund by the University Grants Committee of Hong Kong (C5012-15E).

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.