Cell and tissue chip

Breast cancer cell lines (MDA-MB-231, MCF-7, MDA-MB-453 and BT549) and human normal mammary epithelial cell line (MCF-10 A) were purchased from BeNa, which were cultured in the DMEM medium supplemented with 10% FBS. All of the cells were cultured in the incubator containing 5% at 37 °C.

The tissue chip was purchased from Shanghai Outdo Biotech Company (China), containing 143 breast cancer tissues and 32 para-carcinoma tissues. Besides, the pathological characteristics of patients with breast cancer, such as age, grade, and tumor infiltrate, were collected and subjected statistical analysis.

Bioinformatic analysis

GDC download tool was used to download the RNAseq counts document of breast cancer (BRCA) from TCGA database, containing 1097 breast cancer samples and 113 normal samples, and TCGAbiolinks package of R studio was used to download the clinical information of TCGA-BRCA patients. The estimate the dispersion in DEseq2 was used for data standardization, and the Benjamini-Hochberg (BH) was utilized to adjust P value. Differential analysis of gene expression was performed by DEseq2 and the screening criteria was as follows: |Fold Change| > 1.3 and P < 0.05.

The clinical information of breast cancer patients in TCGA database was from cbioportal (https://www.cbioportal.org/). The expression of TEX19 or CDK4 was divided into high expression group and low expression based on the optimal cut point, and then the difference of overall survival or progression-free survival between the two groups was analyzed by log rank test.

Immunohistochemical staining (IHC)

Tissue samples were prepared into 5 μm tissue sections, and underwent antigen repair through citrate buffer solution. Tissue sections were blocked by 3% H2O2 and then by 5% serum. Tissue sections were incubated in the primary antibody solution at 4 °C overnight, and washed 3 times by 1 × PBST for 5 min each time. Tissue sections were incubated in the secondary antibody solution at 37 °C for 1 h and washed 3 times by 1 × PBST. DAB solution and tissue sections were incubated in the dark for 5 min, and the restained with hematoxylin (Baso) for 10–15 s. Tissue sections were washed by water and then separated by alcohol for 1–2 s. Finally, neutral gum was used for closed tissue sections, and the staining was observed under a microscope (Olympus) and photographed. IHC results were scored in terms of positive cells and staining intensity. IHC score = positive cell score × staining color intensity score. The higher the score, the higher the expression of protein to be detected. The relevant information of antibodies used in this experiment was provided in Supplementary Table 1.

Lentivirus vector construction and cell infection

Construction of RNA interference lentivirus vector: Three RNAi target sequences were designed using TEX19 or CDK4 gene as templates, which were provided in Supplementary Table 2. RNAi sequences were synthesized single-stranded DNA oligo, and then annealed to form double-stranded DNA after being bathed at 90 °C for 15 min. Double-stranded DNA was attached to the linearized vector, and the ligated product was then transformed into the competent Escherichia coli cells. Subsequently, the plasmids were extracted according to the instructions for EndoFree Maxi Plasmid Kit (TIANGEN). The plasmids carrying the RNAi sequence were co-infected with the lentivirus packaging helper plasmids to 293T cells. ShTEX19 or shCDK4 lentivirus plasmids were extracted 72 h later.

Construction of overexpressed lentivirus vector: Using TEX19 gene as the template, the primer amplification sequence was designed. The synthesized primers were configured into PCR reaction system and PCR was carried out. The PCR products were connected to the linearized vector, and then co-transfected into the competent Escherichia coli cells. Subsequently, the plasmid extraction was performed using the EndoFree midi Plasmid Kit (TIANGEN). The plasmids carrying the TEX19 gene overexpression sequence were co-infected with the lentivirus packaging helper plasmids to 293T cells. TEX19 overexpression lentivirus plasmids were extracted 72 h later.

Cell infection by lentivirus: The infection solution was added to the culture dish of healthy growing breast cancer cells, along with the gene knockdown or overexpressed lentiviral plasmids (carrying green fluorescent protein), and then incubated in an incubator at 37 °C for 18 h. After the newly prepared medium was replaced, the culture was continued for 72 h, and the florescence was observed under a fluorescence microscope to evaluate the infection efficiency.

Real time qPCR (qPCR)

Total RNA in breast cancer cells was abstracted by Trizol (Sigma) according to the kit operation instruction. cDNA was attained by reverse transcription of RNA under the instruction of Hiscripr QRT supermix for qPCR (+ gDNA WIPER) (Vazyme). The reaction system was prepared with SYBR Green mastermics (AceQ qPCR STBR Green master mix, Vazyme), forward and reverse primes, cDNA and other reagents in a certain proportion, and then Real Time PCR was performed in a two-step method on the Real Time PCR instrument (ABI), followed by the preparation of the dissolution curve. Formula 2−△△Ct was used to calculate the mRNA levels of the gene. Primers used in qPCR assay were provided in Supplementary Table 3.

Western blotting (WB)

After breast cancer cells were lysed, proteins in cell lysates were extracted. The protein concentration was determined with BCA Protein Assay Kit (HyClone-Pierce). 20 µg protein was suffered from SDS-PAGE, and then transferred to the PVDF membrane. PVDF membranes were incubated with TBST solution containing 5% skim milk at room temperature for 1 h, and then incubated with primary antibody solution at 4 °C overnight. PVDF membranes were cleaned by TBST for 3 times with 5 min each time. PVDF membranes and secondary antibody solution were co-incubated at room temperature for 1 h. After being washed by TBST, immobilon Western chemiluminescent HRP Substrote (Millipore) was utilized for color development of PVDF membranes and chemiluminescence was carried out using a Chemiluminescence get imaging system (GE). Antibodies used in western blotting assay were provided in Supplementary Table 1.

Celigo cell counting assay

Breast cancer cells infected with lentivirus were digested with trypsin (Sangon Biotech (Shanghai) Co., Ltd), re-suspended into cell suspension, and inoculated into 96-well plate with 2000 cells per well. Starting the next day, Celigo (Nexcelom) scanned the target 96-well plate at the same point in time for 5 consecutive days to obtain the image. The scanned images were counted by Image J software and the cell proliferation curve was plotted.

Flow cytometry

Flow cytometry was performed to detect cell apoptosis. When cell coverage in 6-well plate reached about 85%, breast cancer cells were re-suspended to cell suspension. After cleaning the cells with D-Hanks (pH = 7.2 ∼ 7.4) precooling at 4 °C, the cells were washed again with 1 × binding buffer. Cells were suspended with 200 µL 1 × binding buffer and incubated with 10 µL Annexin V-APC (eBioscience) at room temperature in the dark for 10–15 min. Afterwards, apoptosis levels were detected by flow cytometry (Millipore).

Flow cytometry was conducted to determine cell cycle. When the cells in each group grew to about 80% coverage, the cells were digested with trypsinase and collected in a 5 mL centrifuge tube. The cells were washed with PBS (pH = 7.2 ∼ 7.4) precooling at 4 °C for one time, and fixed in 70% ethanol precooling at 4 °C for at least 1 h. After washing the cells with PBS, the cells were re-suspended with the prepared cell staining solution, and the cell cycle distribution was detected by flow cytometry (Millipore). The proportion of cell staining solution was as follows: 40 × PI solution (2 mg/mL) (Sigma):100 × RNase solution (10 mg/mL):1 × PBS = 25:10:1000.

Human apoptosis antibody array

Human Apoptosis Antibody Array (ab134001) was purchased from abcam, with a total of 43 apoptosis markers. According to the instruction, Human Apoptosis Antibody Array could simultaneously detect the concentration of 43 apoptosis markers in MDA-MB-231 cells of the TEX19 knockdown group and the shCtrl group. Total proteins were extracted from MDA-MB-231 cells infected with shTEX19 or shCtrl lentivirus after lysis. The protein concentration was determined with BCA Protein Assay Kit. The membrane was incubated with 2 mL 1 × Blocking buffer at room temperature for 30 min, and then incubated with 1.2 mL samples at 4 °C overnight. After that, the membrane was incubated in 1 mL of 1 × Biotin-conjugated Anti-Cytokines and 1.5 mL of 1 × Streptavidin-HRP successively. Finally, protein levels were detected under the instruction of Chemiluminescent Detection protocol.

Transwell assay

The Transwell upper chamber (3422, corning) was incubated with 100 µL serum-free medium for 2 h. The medium in upper chamber was replaced with 100 µL cell suspension (containing 50,000 breast cancer cells). Afterwards, the upper chamber was transferred to the lower chamber containing 600 µL medium with 30% FBS for incubation for 24 h. The medium was removed from the upper chamber and the non-metastatic cells were gently swabbed with a cotton swab. After that, the upper chamber was soaked in 400 µL Gimsa solution for 5 min. The film was rinsed with water and dried in air. Finally, the cell metastasis was observed under a microscope (Olympus) and photographed.

Wound healing

Lentivirus-infected breast cancer cell suspensions were added to 96-well plate, 40,000 cells per well. The next day, the 96-well plates were gently pushed upward from the lower central part with 96 Wounding Replicator (VP scientific), and rinsed 3 times with serum-free medium. Medium with 0.5% FBS was added and cells were cultured in an incubator containing 5% CO2 at 37 °C. The plates were scanned with Cellomics (Thermo) at the right time according to the degree of healing. The scanning time of MDA-MB-231 cells was 0 h, 8 h, and 24 h. The scanning time of BT549 cells was 0 h, 24 h, and 48 h. Cellomics was employed to evaluate cell migration by analyzing cell area from the same field of view at different time points.

Subcutaneous tumorigenesis model in nude mice

4-week-old BALB/c nude mice (female) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. Nude mice were randomly divided into control group (shRNA group) and experimental group (shTEX19 group) with 10 nude mice in each group. 200 µL cell suspension (4 × 106 MDA-MB-231 cells infected with shTEX19 or shRNA) was injected subcutaneously into nude mice. From day 25 onwards, the long and short diameter of the tumors were measured every 10 days with a Vernier caliper, and the tumor volume was calculated to plot the tumor growth curve. Tumor volume: V = π/6 × L × W × W. V represented tumor volume; L represented the long diameter for the tumor; W represented the short diameter of the tumor. On the last day of the experiment, 0.7% sodium pentobarbital (10 µL/g) (SIGMA) was injected intraperitoneally, and then anesthetized nude mice were imaged using the small animal living imaging system (BERTHOLD TECHOLOGIES). Fluorescence was observed and photographs were taken. The tumors were removed from the sacrificed nude mice, weighed and measured in volume, and then photographed with a digital camera (SONY). Immunohistochemical staining was used to detect the expression of Ki67 protein in tumor tissues, as described above. The antibodies used in this experiment were provided in Supplementary Table 1.

GeneChip and ingenuity pathway analysis (IPA)

Total RNA in breast cancer cells (MDA-MB-231) was abstracted by Trizol (Sigma) according to the kit operation instruction. According to the instruction manual of Affymetrix’s gene chip sequencing instrument, the whole gene expression profile chip experiment based on GeneChip was performed on the total RNA. The original chip data was preprocessed including KNN function missing value filling, data normalization, and data cleaning. In the process of hierarchical clustering significance difference analysis of preprocessed data using R studio, a linear model based on empirical Bayesian distribution was used to calculate the P-value of significant difference level, and Benjamini-Hochberg method was employed to correct the significant difference level (FDR). Significant difference of gene screening standard was: |Fold Change| ≥ 1.3 and P-value < 0.05. Based on the differentially expressed genes, the classical pathway enrichment analysis and molecular interaction network analysis were performed using IPA.

Co-immunoprecipitation (Co-IP)

After breast cancer cells (MDA-MB-231 and BT549) were lysed, proteins in cell lysates were extracted. The protein concentration was determined with BCA Protein Assay Kit. The lysate containing 1.2 mL total protein and the reference antibody were incubated at 4 °C overnight under rotating conditions. Then the incubation product and 200 µL beads were incubated at 4 °C for 2 h under rotating conditions, followed by centrifugation at 2000 × g centrifugal force for 1 min. The IP product was cleaned 2 times with IP lysate and subjected SDS-PAGE. The proteins were transferred to the PVDF membrane. PVDF membranes were incubated with TBST solution containing 5% skim milk at room temperature for 1 h, and then incubated with primary antibody solution at 4 °C overnight. PVDF membranes were cleaned by TBST for 3 times with 5 min each time. PVDF membranes and secondary antibody solution were co-incubated at room temperature for 2 h, and washed by TBST. Chemiluminescence was carried out using a Chemiluminescence get imaging system. Antibodies used in western blotting assay were provided in Supplementary Table 1.

Protein stability assay

Cycloheximide (CHX, 0.2 mg/mL, S7418, Selleck), an inhibitor of protein synthesis, was used to treat MDA-MB-231 and BT549 cells with TEX19 knockdown or SKP2 overexpression to evaluate protein stability. After 0, 3, 6, and 12 h of CHX treatment of cells, total proteins in MDA-MB-231 cells were extracted. Then 20 µg total protein was subjected to western blotting to detect CDK4 protein levels. Information about relevant primary and secondary antibodies was provided in Supplementary Table 1.

Ubiquitination assay

MDA-MB-231 and BT549 cells were infected with shTEX19 or SKP2 overexpression lentivirus for 24 h, and incubated with ubiquitin-proteasome pathway inhibitor MG-132 (20 µM, HY-13,259, MEC) for 6 h. Total proteins in MDA-MB-231 cells were extracted. 20 µg total protein was subjected to western blotting to detect CDK4 protein levels. Another1.0 mg total protein and antibody were incubated at 4 °C overnight, and then incubated with 20 µL beads at 4 °C for 2 h. The protein-antibody-beads complex was subjected to western blotting, and the levels of ubiquitin were determined by the ubiquitin antibody. Information about relevant primary and secondary antibodies was provided in Supplementary Table 1.

CCK8 assay

MDA-MB-231 and BT 549 cells in the logarithmic growth phase were suspended and added to the 96-well plate at 100 µL per well (2000 cells). At the 1st, 2nd, 3rd, 4th and 5th day, one 96-well plate was removed for CCK8 detection. Specifically, 10 µL CCK8 reagent (Sigma) was added to each well and incubated with cells for 4 h. After the 96-well plates were oscillated for 2 min, the OD values were detected at 450 nm wavelength by using a microplate reader (M2009PR, Tecan infinite), and the fold change of OD value was calculated.

Colony formation assay

MDA-MB-231 and BT549 cells were inoculated in 6-well plates with 2 mL cell suspension per well, containing 1000 cells. The cells were cultured for 8 days, during which the medium was changed every 3 days. 1 mL 4% paraformaldehyde was added per well to fix cells for 30–60 min. The cells were washed once with PBS and incubated with 500 µL GIEMSA solution (Shanghai Dingguo Biological Technology Co., Ltd.) for 10–20 min. After cleaned several times with ddH2O, a digital camera was used to take pictures and the clones was counted.

Statistical analysis

Sign test were employed to evaluate whether TEX19 gene expression was statistically different between breast cancer tissues and para-carcinoma tissues. Chi-square Test and Mann-Whitney U analysis were used to analyze the signification of TEX19 expression at different levels in different pathological characteristics, and Spearman correlation analysis was conducted to analyze the correlation between the expression level of TEX19 in cancer tissues and pathological characteristics. T-test was used for statistical analysis between the other two groups, and One-way ANOVA was utilized for statistical analysis among multiple groups. P < 0.05 represented significant difference.