Data Source and Online Analysis Tool
Gene expression data with clinical information (Workflow Type: HTSeq-TPM and HTSeq-FPKM) was acquired from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/)[12]. Data of samples without sufficient clinical information were excluded. 8626 pan-cancer tissues and 713 corresponding paracancerous normal tissues from the TCGA-ALL dataset, 1083 tumor tissues and 111 paracancerous normal tissues from TCGA-BRCA dataset, including 110 paired samples were enrolled in this study. R (3.6.3 version) with ggplot2 package (3.3.3 version), survminer package (0.4.9 version), survival package (3.2–10 version), pROC package (1.17.0.1 version), RMS package (6.2–10 version) and GSVA package (1.34.0version) were used to process the original data and generate some figures and tables. Three online databases, including the LinkedOmics (http://www.linkedomics.org/login.php), the TISIDB (http://cis.hku.hk/TISIDB/index.php), and the Kaplan Meier Plotter (http://kmplot.com/analysis/) were applied in this study.
Differentially Expression
We proceeded with Logistics regression analysis of the correlation between GREB1L mRNA expression and clinical characteristics in BRCA. We compared the differential GREB1L expression levels between tumors and corresponding paracancerous normal tissues in 23 human tumor types by Wilcoxon rank sum test. In BRCA cohort, we analyzed differential GREB1L expression levels between tumor tissues and paracancerous normal tissues. The expression levels in paired samples were explored additionally. Results were showed in scatter plots respectively.
Prognostic Value
According to the median value of GREB1L expression, patients were classified as low- and high-expression groups. TISIDB website was used to create a bar plot to compare the association between GREB1L expression and overall survival (OS) across 30 human malignancies. Data were processed by Log Rank test. Kaplan-Meier curves of OS and relapse-free survival (RFS) between GREB1L low- and high-expression groups were carried out based on gene chip data from the Kaplan-Meier plotter database. Adopted statistical approach was also the Log Rank test. Forest plot of the multivariate Cox regression analysis was processed to show the prognostic factors in BRCA. We calculated the p-value, hazard ratio (HR) and 95% confidence interval (CI) of every potential predictor. Prognostic factors with HR > 1 and p-value < 0.05 were risk factors of BRCA prognosis. Meanwhile, those with HR < 1 and p-value < 0.05 were regarded as protective factors.
Co-expression Networks And Gene Set Enrichment Analysis
To predict the potential biological mechanism of GREB1L in BRCA, we used the LinkFinder module in the LinkedOmics website (http://www.linkedomics.org/login.php) to research the co-expression network of GREB1L in TCGA- BRCA database. Then, in the LinkInterpreter module of the same portal, GSEA was utilized to identify terms significantly related (FDR < 0.05) to GREB1L co-expression genes. The analysis contains four aspects, including Gene Ontology biological process (GO-BP), Gene Ontology cellular component (GO-CC), Gene Ontology molecular function (GO-MF), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. We utilized Pearson correlation as a statistical approach, and false discovery rate (FDR) < 0.05 was defined as significantly related or enriched.
Specimens And Cells
Breast tissue specimens were obtained from Qilu Hospital of Shandong University (Qingdao). The specimens were frozen in liquid nitrogen immediately after surgical resection. All breast tissues were collected according to the protocol approved by the Ethics Committee of Shandong University Qilu Hospital (Qingdao). MCF10A, MCF7, Hs578T, ZR-75-1, MDA-MB-231, MDA-MB-453 and SK-BR-3 cells were preserved by our laboratory. Hs578T was cultured by DuIbecco’s Modified EagIe’s Medium (DMEM) (BI, C3113) containing 10µg g/mL insulin. MCF10A was cultured by DMEM/F12 containing 5% horse serum, 20ng/ml epidermal growth factor, 0.5µg/ml hydrocortisone, 10µg/ml insulin and 1% non-essential amino acids (Procell, CM-0525). ZR751 was cultured by Roswell Park Memorial Institute (RPMI)-1640 (Procell, PM150110) medium. MDA-MB-453 was cultured in Leibovitz's L-15 (Procell, PM151010) medium. MDA-MB-231, MCF7, and SK-BR-3 were cultured in DMEM medium. All cell cultures required 10% fetal bovine serum (FBS) (BI, C04001) and 1% penicillin-streptomycin solution (Procell, PB180120).
Lentivirus Construction And Transfection
Lentiviruses carrying short hairpin RNA (shRNA) against GREB1L were constructed by Genechem Company (Shanghai, China). The RNAi sequence targeting human GREB1L was 5′- GCGTTTGGTATCACTGTGTAT-3′. The negative control sequence was 5′-TTCTCCGAACGTGTCACGT-3′. Viruses were transfected by HitransG P according to manufacturer’s instructions.
Transwell Migration And Invasion Assays
Transwell migration and invasion assays were performed using 24-well insert transwell chambers (Corning, #3422). About 5╳104 MCF7 cells and 3.5╳104 MDA-MB-231 cells were resuspended in 200ul DMEM without FBS and seeded in the upper chamber. DMEM containing 20% FBS was added to the bottom wells to stimulate migration or invasion. For migration assay, the seeding MCF7 cells were incubated for 24 hours (231 cells for 12 hours). For the invasion assay, the MCF7 cells were incubated for 24 hours (231 cells for 18 hours). After incubation, the cells on the upper surface of the chamber were wiped off with a cotton swab, then rinsed with PBS, and the cells on the lower surface of the chamber were fixed with methanol, stained with 0.1% crystal violet. Then counting at 100╳ with microscope. For the invasion assay, the top chamber was coated with matrigel (ABW, 082704).
Cell Counting Kit-8 (Cck8) Cell Proliferation Assays
About 3╳103/well MDA-MB-231 or 4╳103/well MCF7 cells were seeded in 96-well plates, 10ul CCK8 (Dojindo, CK04) reagent was added to each well before measurement, and after incubating for 2 hours in the 37℃ incubator, the absorbance was measured at OD450nm. Measure once every 24 hours.
Clone Formation Assays
Approximately 1╳103 cells were seeded in six-well plates. After 2 ~ 3 weeks of treatment with purinomycin, cells were washed three times with PBS, then fixed with methanol and stained with 0.1% crystal violet.
Western Blot
RIPA lysis buffer (Solarbio, #R0010) was used to extract total cell protein. PMSF (Solarbio, P0100) was added to lysis buffer. Lysates were separated on 6% and 10% acrylamide gel. Then the proteins were transferred from gel to PVDF membrane (Immobilon-P, IPVH00010). Blocking buffer ( Boster, AR0041) were used to block blots. The anti-GREB1L (1:300, ATLAS ANTIBODIES, HPA041647) and the anti-β-Tubulin (1:1000, absin, abs830032) were used as the primary antibodies. Anti-β-Tubulin were used as an internal control. HRP-goat anti-mouse IgG (1:5000, Earthox, E030110) and HRP-goat anti-rabbit IgG (1:5000, Earthox, E030120) were used as the secondary antibody.
Immunohistochemistry
Immunohistochemistry (IHC) was carried out as described previously [13] with anti-GREB1L (1:300, ATLAS ANTIBODIES, HPA041647) in the paraffin-embedded breast cancer tissue sections.
Statistical analysis
All experiments were repeated three times more. All quantitative data were presented as the means ± SD. We used a standard two-tailed unpaired t-test for the statistical analysis of the two groups. When p < 0.05 was considered to be statistically significant. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns representd no significance.