Phage display biopanning, an affinity selection technique that selects peptides binding to a given target, was conducted according to standard procedures (19). A peptide phage display library was commercially constructed (Ph.D.-12 Phage Display Peptide Library; New England Biolabs, Inc., Beverly, MA, USA) and used in subsequent experiments. The recombinant human CD105 protein (R&D Systems, Inc., Minneapolis, MN) at 100 µg/ml in sterile phosphate-buffered saline (PBS) was added to individual sterilised MaxiSorp plates (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and incubated overnight at 4°C. The plate was washed six times with Tris-buffered saline/0.1% Tween-20 (TBST) and then blocked for 1 h at 4°C using 1% bovine serum albumin (BSA) in PBS. M13 phage display libraries were allowed to bind for 1 h at room temperature, and the unbound phages were washed away 10 times with TBST. Subsequently, the binding phages were eluted by an elution buffer, and the eluted phages were then amplified for another round of biopanning. For each round of biopanning, the titering of M13 phage was calculated. The recovery rate was calculated as follows: Recovery rate = Titering output/titering input.

Subsequent to three rounds of biopanning, the target M13 phages were enriched. A total of 16 monoclonal phages were selected randomly for the ELISA binding essay. Recombinant human CD105 protein at 100 µg/ml in sterile PBS was added to 16 individual sterilised wells of MaxiSorp plates and incubated overnight at 4°C. Additionally, 16 individual sterilised wells were coated using 1% BSA as control. The plate was washed six times with TBST and blocked for 1 h at 4°C using 1% BSA in PBS. Next, the 16 monoclonal phages were amplified and added to the individual wells with CD105 protein and 1% BSA and incubated for 1 h at room temperature. Following repeated washing of the MaxiSorp ELISA plate with 0.1% TBST, bound phages were detected by incubation with mouse anti-M13 phage antibody (HRP conjugated; cat. no. ab6188; 1:20; Abcam, Cambridge, UK). The ABTS/H₂O₂ substrate (internally prepared) was used to measure the amount of bound HRP, and the absorbance was monitored at 405 nm. The inserted DNA sequences of 13 positive affinity monoclonal phages were determined using the primer 5′-CCCTCATAGTTAGCGTAACG-3′ (New England Biolabs, Inc.).

Based on aforementioned procedures, 13 peptides were synthesised through solid-phase peptide synthesis using Fmoc chemistry by Chinese Peptide Company (Jiangsu, China). An extra fluorescein-5-isothiocyanate (FITC) was linked at the amino (N) terminus of all peptides for labelling. Prepared peptides were identified by mass spectrometry (MS), and the purity (>95%) was assayed by high-performance liquid chromatography (HPLC; Fig. 1). The FITC-labelled peptides were stored at −20°C. For further usage, peptides were dissolved in sterilised water at 1 mg/kg and diluted according to the experimental requirements.

MNNG/HOS and Cal27 cell lines were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). MNNG/HOS cells were cultured in Dulbecco's modified Eagle's medium (DMEM)/F12 (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) supplemented with 5% foetal bovine serum (FBS), while Cal27 cells were cultured in DMEM (Sigma-Aldrich; Merck KGaA) supplemented with 10% FBS. The cells were plated onto 10 cm² flasks and maintained at 37°C in 5% CO₂ with 90% relative humidity. In order to investigate the CD105 expression in MNNG/HOS and Cal27 cells, the cell lines were then examined by reverse transcription-polymerase chain reaction (RT-PCR), immunofluorescence and flow cytometry.

A NanoDrop 1000 (Thermo Fisher Scientific, Inc., Wilmington, DE, USA) was used to measure the amount of RNA by spectrophotometry. A total of 300 ng total RNA was extracted from tumour cells, and converted to complementary DNA using the ReverTra Ace^® qPCR RT kit (Toyobo Life Science, Osaka, Japan) for PCR analysis. The semi-quantitative RT-PCR analysis was performed using the Takara Ex Taq^® kit (Takara Bio, Inc., Otsu, Japan). The PCR reaction conditions were performed as follows: 95°C for 5 min, and 30 cycles of 95°C for 30 sec, 60°C for 30 sec and 72°C for 1 min, following by a final elongation at 72°C for 10 min and held at 16°C for 10 min. The CD105 primers used were synthesised by Synbio Technologies (Suzhou, China) and were as follows: CD105 forward, 5′-CACCACAGCGGAAAAAGGTG-3′ and reverse, 5′-GCCGGTTTTGGGTATGGGTA-3′; and GAPDH forward, 5′-AATGGGCAGCCGTTAGGAAA-3′ and reverse, 5′-GCGCCCAATACGACCAAATC-3′). PCR products were then analysed by electrophoresis on a 1% agarose gel and visualised by ethidium bromide staining under UV illumination in Iquant Capture RT ECL (version 1.0.1; GE Healthcare, Chicago, IL, USA).

Following collection in a 0.25% (w/v) trypsin solution with 0.025% (w/v) ethylene diamine tetraacetic acid (EDTA), MNNG/HOS and Cal27 cells were seeded in a 6-well flask with slides at a density of 1×10⁵ cells/well and incubated at 37°C in 5% CO₂ with 90% relative humidity overnight. On the second day, the cells were fixed in 4% paraformaldehyde at room temperature for 15 min, and non-specific binding sites were blocked by 5% BSA (Guangzhou Xiang Bo Biological Technology Co., Ltd., Guangzhou, China) for 1 h at room temperature. The cells were then incubated with the CD105 antibody (cat. no. MA1-19408; 1:200; Invitrogen; Thermo Fisher Scientific, Inc.) overnight at 4°C. Unbound antibodies were eluted by washing three times with PBS. Subsequently, the cells were further treated with donkey anti-mouse IgG (H+L) highly cross-adsorbed Alexa Fluor 568 secondary antibody (cat. no. A10037; 1:1,000; Invitrogen; Thermo Fisher Scientific, Inc.) in blocking buffer for 1 h at room temperature. Following further washing with PBS, the nucleus was counterstained by DAPI, and the samples were visualised under confocal laser scanning microscopy (CLSM; Leica Microsystems GmbH, Wetzlar, Germany). For polypeptide labelling, the MNNG/HOS/HOS and Cal27 cells were incubated in polypeptide at a concentration of 100 µg/ml in 1 ml PBS for 30 min at 4°C prior to fixation.

Subsequent to collection in a 0.25% (w/v) trypsin solution with 0.025% (w/v) EDTA, MNNG/HOS and Cal27 cells were incubated with the CD105 antibody (cat. no. 130-102-819; 1:10; Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) at 4°C for 10 min. The cell pellets were washed three times in PBS, suspended in 500 µl PBS and then analysed using a FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA). For polypeptides labelling experiments, the MNNG/HOS/HOS or Cal27 cells were incubated in 100 µg/ml peptides in 100 µl PBS at 4°C for 30 min. Subsequent to further washing with PBS, cell pellets were suspended in 500 µl PBS and then analysed using a FACSCalibur flow cytometer.

ELISA binding assay of nABP296 to CD105 protein was performed, with CD106 protein (R&D Systems, Inc.) used as a control. Recombinant human CD105 protein and CD106 protein at a concentration of 100 µg/ml in sterile PBS were added to six individual sterilised MaxiSorp plates and incubated overnight at 4°C. The plates were washed six times with TBST and blocked for 1 h at 4°C using 1% BSA in PBS. Subsequently, nABP296 peptide was added at a concentration of 100 µg/ml and incubated for 1 h at room temperature. The unbound peptide was removed by PBS washing for 10 times. The absorbance was measured using a VICTOR X5 Multilabel plate reader (PerkinElmer, Inc., Singapore) at 490 nm, and the results are reported as the optical density.

For the cytotoxicity assay, 1×10⁵ cells/well were seeded in 96-well plates in 100 µl complete culture medium. Subsequent to culturing overnight at 37°C in a 5% CO₂ atmosphere, the cells were incubated in different concentrations of nABP296 solution ranging between 0.5 and 5 µMol. At 24 and 48 h, 10 µl Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan) solution was added to the cells and incubated for another 1 h. The absorbance was then determined at 450 nm using the VICTOR X5 Multilabel plate reader, and the survival rate was calculated.

Based on the aforementioned experiments, the nABP296 peptide was used in the in vivo experiment. The animal use protocol was reviewed and approved by the Animal Ethical and Welfare Committee of Sun Yat-sen University (Guangzhou, Guangdong, China). Briefly, MNNG/HOS cells were subcutaneously transplanted into 5-week-old BALB/c Nod mice at 1×10⁷ cells/ml in DF. When the volume of the xenograft tumours reached ~200 mm³, 1 mg/ml nABP296 in 100 µl sterile Milli-Q water was intravenously administered to the mice. The mice were observed using the IVIS Spectrum In Vivo Imaging system (PerkinElmer, Akron, OH, USA), and then images were captured after 1 h of administration. Next, the mice were euthanised, tumours were excised from the MNNG/HOS tumour-bearing mice and images were captured using the IVIS Spectrum In Vivo Imaging system. Tumours and hearts were freeze-sectioned into 6 µm and fixed in 100% acetone for 15 min. Following further washing with PBS, the nuclei of the sections were stained with DAPI. The results were analysed by CLSM (Leica Microsystems GmbH).

All patients participating in this experiment provided informed consent prior to tissue collection and agreed to the use of their samples in scientific research. The described experiments were approved by the Ethics Committee of The Hospital of Stomatology at Sun Yat-sen University (approval no. ERC-2017-11). A total of 2 patients (osteosarcoma patient: Male, 33 years old, hospitalised for 15 days in July 2017; tongue carcinoma patient: Male, 25 years old, hospitalised for 16 days in October 2017) who were treated in Guanghua School of Stomatology (Hospital of Stomatoloty, Sun Yat-sen University) and diagnosed with osteosarcoma and tongue carcinoma by the Department of Pathology were included in these experiments. Osteosarcoma and tongue carcinoma tissues were obtained from the patients during surgery.

Tissue slices derived from a xenograft tumour model, an osteosarcoma patient and a tongue carcinoma patient were used for histological staining. The tumour samples were sectioned (6-µm thick) using a cryostat microtome (Microm HM560; Thermo Fisher Scientific, Inc.). Non-specific binding sites were blocked by incubation with 5% BSA for 1 h at room temperature. Next, histological sections were incubated using a combination of 50 µMol nABP296 peptide and mouse monoclonal CD105 antibody (cat. no. MA1-19408; 1:200; Invitrogen; Thermo Fisher Scientific, Inc.) in blocking buffer at 4°C overnight. Unbound peptide and antibody were eluted by washing three times with PBS, following which cells were further treated with donkey anti-mouse IgG (H+L) highly cross-adsorbed Alexa Fluor 568 secondary antibody (cat. no. A10037; 1:1,000; Invitrogen; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. Following further washing with PBS, the nucleus was stained with DAPI, and the results were analysed under CLSM (Leica Microsystems GmbH).

Statistical analysis was performed using the GraphPad Prism software (version 5; GraphPad Software, Inc., La Jolla, CA, USA). One-way analysis of variance was used to evaluate the survival rate of MNNG/HOS cells at 24 and 48 h. P-values of <0.05 were considered to denote statistically significant differences.

The recovery rate of three rounds of biopanning is shown in Table I. Subsequent to three rounds of biopanning, the recovery rate of M13 phage was raised from 4.9×10⁻⁶ to 4.5×10⁻⁵. M13 phage binding to recombinant human CD105 protein was enriched. To identify a monoclone phage that can bind to CD105, ELISA was performed to M13 phage single clones (A-P). A total of 13 monoclonal phages were confirmed to have high affinity for CD105, as shown in Table II. The sequences and characteristics of the 13 positive monoclonal phages are listed in Table III in the range between nABP295 and nABP307.

Based on the results of RT-PCR, fluorescence and flow cytometry (Fig. 1), the MNNG/HOS cell line was found to be CD105-positive and the Cal27 cell line was observed to be CD105-negative. These two cell lines were then used in subsequent experiments to identify the peptide with high binding affinity for CD105 expressed in the cell membranes.

By flow cytometry, two peptides, namely nABP296 and nABP297, with high affinity for the CD105-positive MNNG/HOS cell line were identified (Fig. 2A and B). To determine whether these two peptides had a specific affinity for CD105, the binding of nABP296 and nABP297 to the CD105-positive MNNG/HOS and CD105-negative Cal27 cell lines was analysed by a flow cytometry assay. As shown in Fig. 2C, nABP296 exhibited higher binding efficiency to MNNG/HOS than Cal27 cells, while nABP297 did not exhibit different binding efficiencies between MNNG/HOS or Cal27 cells. According to the results of the immunofluorescence assay, nABP296 co-localised with the CD105 antibody in the MNNG/HOS cell line and had no binding affinity for Cal27 cells (Fig. 2D). Furthermore, the ELISA experiment of nABP296 binding to CD105 and control proteins revealed that nABP296 had a higher affinity for CD105 protein in comparison with that for the control protein (Fig. 2E). Finally, nABP296 exhibited no cytotoxicity on the MNNG/HOS cell line after 24 and 48 h (Fig. 2F). Taken together, nABP296 can selectively bind to the CD105-positive MNNG/HOS cell line and is biocompatible with this cell line, suggesting that nABP296 may be able to target human osteosarcoma in vivo. The chemical structures of nABP296 and the HPLC/mass spectrometry results are shown in Fig. 3A and B, respectively.

Following incubation of osteosarcoma histological sections derived from an MNNG/HOS xenograft tumour model and an osteosarcoma patient with nABP296, green fluorescence indicated nABP296 peptide and red flourescence indicated CD105 antibody. Results showed that the MNNG/HOS xenograft tumour model and osteosarcoma derived from the patient were successfully co-labelled with the nABP296 peptide and CD105 antibody (Fig. 4A and B). By contrast, absence of CD105 antibody and only weak nABP296 fluorescence were observed in the tongue carcinoma section (Fig. 4C). The results indicated that nABP296 was able to bind to osteosarcoma derived from a xenograft tumour model and a patient in vitro. Thus, nABP296 may be used in the diagnosis of osteosarcoma in place of CD105 in certain situations.

As shown in Fig. 5A, at 1 h after intravascular administration of nABP296 to MNNG/HOS tumour-bearing mice, fluorescence was detected in the MNNG/HOS xenograft tumour. Fluorescence detected in the tumour was higher compared with that in the surrounding tissues, whereas there was no fluorescence detected in the NC group. Furthermore, ex vivo imaging of tumours that were excised from the MNNG/HOS tumour-bearing mice was performed at 1.5 h. It was observed that FITC-labelled nABP296 was accumulated in the xenograft tumour (Fig. 5B). Fluorescence was detected in frozen tumour histological sections, while there was no FITC fluorescence in the heart sections (Fig. 5C). These results supported the high affinity and specificity for nABP296 peptide for the identified MNNG/HOS xenograft tumour in vivo.