Bioinformatic analysis

Gene expression data and corresponding clinical information of glioma patients were downloaded from The Cancer Genome Atlas (TCGA, http://cancergenome.nih.gov), and Chinese Glioma Genome Atlas (CGGA, http://www.cgga.org.cn). The CGGA encompasses two mRNA sequence databases named CGGA325 and CGGA693, containing 325 and 693 samples, respectively. Bioinformatics analyses were realized with R Language. Survival and survminer R packages were used to conduct Kaplan–Meier (K-M) survival analyses. The survival ROC package was used to generate ROC curves at 1, 3, and 5 years using the K-M method. The correlation between CD13 expression and various clinical characteristics was plotted using beeswarm package. Univariate and multivariate Cox analysis were performed to evaluate independent prognostic factors. Gene Ontology (GO) analysis was conducted using clusterProfiler package to identify biological processes and pathways associated with the CD13 expression. Genes related to CD13 expression were identified using Pearson’s correlation analysis with a threshold of|r| ≥ 0.3.

Human specimens

Brain tissue samples of 16 patients were collected during therapeutic surgical treatment from 2013 to 2014 (Department of Neurosurgery, Charité-Universitätsmedizin Berlin, Germany). Specimens of four epilepsy patients, four astrocytoma WHO °III and eight GBM patients were investigated. Neuropathologists assessed brain tissues by standard histologic markers. Characteristics of patients are shown in Supplementary Table 1, Additional File 1. Approval for the study was given by the Ethics Committee of Charité-Universitätsmedizin Berlin (application number: EA4/065/13). Investigations were performed in accordance with the obligations of scientific working with patient material. Declaration of consent was obtained from all patients.

Immunofluorescence staining of human tissue

Following surgery, tissues were preserved in 4% paraformaldehyde (PFA) for 24 h and dehydrated in a serial dilution with rising concentrations of sucrose, subsequently. Samples were gently frozen with liquid nitrogen. Sections of 10 μm thickness were prepared. Sections were treated with Autofluorescence Eliminator Reagent (Millipore, Burlington, MA, USA) following the instructions of the manufacturer. Additionally, antigen retrieval was performed using Antigen Retrieval Reagent Universal 10x (R&D Systems) according to manufacturer's instructions. Blocking was carried out using 1% Casein/PBS + 0.1% Triton X-100 for 30 min at room temperature directly after antigen retrieval protocol. Slices were incubated with primary antibody for CD13 (rabbit, Cat #ab108310, 1/400; Abcam, Cambridge, UK) overnight at 4 °C. After several wash steps, secondary antibody was added: AlexaFluor488 anti-rabbit IgG (1/200; Dianova, Hamburg, Germany). After 1.5 h incubation at room temperature, slices were washed with PBS and water, and treated with Immunoselect Antifading Mounting Medium DAPI (Dianova).

Images were acquired by Zeiss Axio Observer Z1 fluorescence microscope (Zeiss Micro-Imaging GmbH) at room temperature. ImageJ 1.49v (NIH) was used to analyze images. Approximately 18–26 images for each patient at three different brain tissue areas were analyzed.

Cultivation of human glioblastoma cell lines

Human glioblastoma cell lines T98G, SF188, SF767, U373, U118, SF126, and U1242 were cultured in DMEM (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% FBS (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) and penicillin/streptomycin (100 U/mL). U87MG cells were cultured in DMEM containing 20% FBS and penicillin/streptomycin (100 U/mL). These cell lines were generously provided by Prof. Dr. Axel Ullrich from the Max Planck Institute of Biochemistry, Martinsried, Germany. The cells were maintained at 37 °C in a humidified atmosphere containing 5% CO2. The characteristics of these cell lines are presented in Supplementary Table 2, Additional File 2.

Quantitative real-time PCR

To extract RNA from human glioma cells, the PureLink® RNA Mini Kit (Invitrogen, Carlsbad, CA, USA) was used. RNA concentration was measured using a plate photometer (Infinite M200, Tecan, Männedorf, Switzerland). Reverse transcription was performed using PrimeScript™ RT reagent Kit (Takara Bio, Kusatsu, Japan). Real-time PCR amplifications were carried out in triplicate in a 20 µL reaction volume, using the TB Green® Premix Ex Taq Kit (Takara Bio, Kusatsu, Japan) with the human CD13 primer pairs (Fw: CATCCATCAGAGATGGCAGAC, Rev: TGCTGAAGAGATCGTTCTGG). Quantitative real-time PCR was performed using a 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) to quantify the target mRNA levels, which were then normalized to  18S (Fw: GGCCCTGTAATTGGAATGAGTC, Rev: CCAAGATCCAACTACGAGCTT) levels. The ΔΔCt method was used to analyze the data.

Protein extraction and western blot

Total proteins were extracted from human glioma cells using RIPA buffer (Thermo Fisher Scientific, Waltham, MA, USA) and heated for 5 min. 4–15% Mini-PROTEAN® TGX™ Precast Gels (BIO-RAD, California, USA) were used to separate the proteins which were then transferred onto PVDF membranes. The membranes were blocked for 20 min at room temperature using StartingBlock™ (PBS) Blocking buffer (Thermo Fisher Scientific, Waltham, MA, USA) and incubated at 4 °C overnight with primary antibodies including CD13 antibody (rabbit, Cat #ab108310, 1/500; Abcam, Cambridge, UK), BAX (D2E11) (rabbit, Cat#5023, 1/1000; Cell Signaling Technology, Danvers, MA, USA ), BCL-2 (mouse, Cat#15071, 1/1000; Cell Signaling Technology), NOXA (mouse, Cat#ab13654, 1/1000; Abcam), Caspase-3 (D3R6Y) (rabbit, Cat#14220, 1/500; Cell Signaling Technology) and GAPDH antibody (mouse, Cat #ab9484, 1/1500; Abcam). After washing several times with TBS-T (0.05% Tween20), the membranes were incubated with HRP-conjugated secondary antibodies (1/200; Dianova, Hamburg, Germany) for 2 h at room temperature. The blots were washed several times before the enhanced chemo-luminescence detection kit (ECL Advance) was applied and luminescence was measured.

Immunocytochemistry

Human glioma cells were seeded into 8-well chamber slides (ibidi, Gräfelfing, Germany) at different densities as indicated in Supplementary Table 3, Additional File 3. After 3 d of growth, cells were fixed with 4% paraformaldehyde in PBS for 20 min, followed by washing with PBS. Then, cells were blocked in 1% casein (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 1 h. Cells were incubated with CD13 antibody (rabbit, Cat #ab108310, 1/400; Abcam) and phalloidin (Cat #ab176753, 1/200; Abcam) for 2 h at room temperature. After washing twice with 0.5% casein/PBS, cells were incubated with Cy3 anti-rabbit IgG (1/200; Dianova) for 2 h, protected from light. The slides were then washed with PBS and water, and the chambers were removed. Finally, the slides were mounted with DAPI-containing media (Dianova). Images were acquired at 20-fold magnification using an inverse fluorescence microscope Zeiss Axio Observer Z1 (Carl Zeiss MicroImaging GmbH, Germany).

Flow Cytometry

All samples were measured on the BD FACSCanto II (BD Biosciences, Heidelberg, Germany) and evaluated with FlowJo software (Ashland, OR, USA).

Staining of CD13 (surface staining and intracellular staining)

Human glioma cells were trypsinized. For intracellular staining, the eBioscience Intracellular Fixation & Permeabilization Buffer Set (eBioscience, San Diego, CA) was used according to the manufacturer's instructions. Briefly, cells were fixed and permeabilized before being incubated with PE anti-human CD13 Antibody (mouse, Cat #310704, 1/20; BioLegend, London, UK) in Perm Buffer for 30 min at room temperature. For surface staining, cells were directly incubated with the PE anti-human CD13 Antibody in 0.5% BSA/PBS for 15 min on ice. After staining, cells were washed and resuspended in 0.5% BSA/PBS. DAPI was added before measurement to exclude dead cells.

Annexin-V staining

Human glioma cells were treated with 500 µg/mL bestatin (Cayman, Neratovice, Czech) or without drug for 48 h. After trypsinization and collection, 1 × 105 cells/sample were used for Annexin-V staining. Following two washes with PBS, cells were incubated with FITC Annexin V (mouse, Cat #640906, 1/20; BioLegend, London, UK) and DAPI in Annexin V Binding Buffer (BioLegend) for 15 min at RT protected from light.

Detection of reactive oxygen species (ROS)

Human glioma cells were treated with 500 µg/mL bestatin or without drug for 48 h. After trypsinization and collection, 1 × 105 cells/sample were used for ROS detection. CellROX™ Deep Red Flow Cytometry Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) was utilized according to the manufacturer's protocol. Briefly, CellROX® Deep Red Reagent (50mM) was incubated with cells at 37 °C for 1 h to detect ROS production.

Proliferation assay

Human glioma cells were seeded into a 96-well plate in a no-phenol red medium at various densities as indicated in Supplementary Table 3, Additional File 3. After 24 and 48 h of growth, the cells were treated with different concentrations of bestatin (0, 62.5, 125, 250, 500 µg/mL) for 24 and 48 h. The MTT Cell Proliferation Assay Kit (Cayman, Neratovice, Czech) was utilized according to the manufacturer's instructions. The absorbance was measured at 570 nm using a microplate reader (Infinite M200, Tecan, Männedorf, Switzerland). The experiment was replicated twice, with triplicate wells for each replication.

Scratch assay

Human glioma cells were collected and adjusted to different cell numbers as indicated in Supplementary Table 3, Additional File 3. A total of 70 µl of cells was added into the insert well of a Culture-Insert 2 Well 24 plate (ibidi, Gräfelfing, Germany), while 340 µl of cells was added around the insert. After 24 h of growth, cells were starved for up to 12 h in DMEM without supplements. The inserts were removed, and residual cells were washed away with PBS. Next, cells were treated with 250 µg/mL bestatin or without drug, and pictures were taken at 0 h, 12 h, and 24 h with 10x magnification. The scratch area was calculated using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Colony formation assay

Human glioma cells were seeded into 6-well plates (250 cells/well). After 24 h of growth, cells were treated with 250 µg/mL bestatin or without drug and cultured for approximately two weeks. Following washing with PBS, cells were fixed with methanol for two minutes and then stained with 2.3% crystal violet solution (Sigma-Aldrich, St. Louis, MO, USA) for 15 min. After several washes with distilled water, pictures were taken, and the number of colonies was counted using a microscope. A cell population with more than 50 cells was defined as a colony.

CD13 activity assay

Human glioma cells were treated with 500 µg/mL bestatin or without drug for 48 h. Cells were trypsinized and collected, and 1 × 106 cells/sample were used for CD13 activity assay. The Aminopeptidase N (APN/CD13) Activity Assay Kit (Fluorometric) (Abcam, Cambridge, UK) was utilized according to the manufacturer's protocol. A standard curve was constructed using AFC Standard. Samples were incubated with the substrate, and the fluorescence (Ex/Em = 384/502 nm) was measured in kinetic mode for 60 min at 37 °C using a microplate reader (Infinite M200, Tecan). Results were calculated as µU/mL of substrate cleaved.

Statistics

Statistical analyses were performed using GraphPad Prism 8 (San Diego, CA, USA), and data are presented as mean ± SD. Comparisons between groups were analyzed using two-tailed unpaired Student’s t-tests or one-way ANOVA with Bonferroni correction as indicated. A P-value < 0.05 was considered statistically significant.