Downregulated expression of ARHGAP10 correlates with advanced stage and high Ki-67 index in breast cancer

RNA extraction and reverse-transcription quantitative polymerase chain reaction

Paired tumor tissues and adjacent normal tissues used in RT-qPCR were obtained from patients who were diagnosed by Department of Pathology and underwent breast cancer surgery at Department of Endocrine and Breast Surgery, the First Affiliated Hospital of Chongqing Medical University. This study was approved by the Institutional Ethics Committees of the First Affiliated Hospital of Chongqing Medical University (#2017-012). Specimens were collected and stored at −80 °C until utilized. Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The concentration of total RNA was measured with NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Reverse transcription was performed using Go-Taq polymerase (Promega, Madison, WI, USA). A SYBR Green PCR Master Mix kit (Invitrogen) for RT-qPCR was used with Applied Biosystems 7500 (Applied Biosystems, Foster City, CA, USA). GAPDH was used as internal control. Relative expression was determined by 2^(−Δt). Forward primer of ARHGAP10 was TGTGGAACCTATGCTGTCAT and reverse primer of ARHGAP10 was GACTTGCTCGTTTGTGGTC. Forward primer of GAPDH was GGAGTCAACGGATTTGGT and reverse primer of GAPDH was GTGATGGGATTTCCATTGAT.

Immunohistochemistry

Formalin-fixed and paraffin-embedded tissue samples were obtained from the Department of Endocrine and Breast Surgery and the Department of Pathology, First Affiliated Hospital of Chongqing Medical University from 2012 to 2017. Specimens were sliced into 4 μm thick section. None of the patients in this cohort received anti-tumor treatment before surgery. This study was approved by the Institutional Ethics Committees of the First Affiliated Hospital of Chongqing Medical University (#2017-012). The slides were deparaffinized with fresh xylene for 30 min and rehydrated with graded ethanol for 25 min, then washed with PBS. Antigens were retrieved in a microwave oven at 100 °C for 20 min. Slides were blocked with 3% hydrogen peroxide for 12 min after cooling to ambient temperature, followed by washing in PBS, blocking with normal goat serum for 15 min, incubation with primary antibody against ARHGAP10 (55139-1-AP; Proteintech Group, Wuhan, China) at dilution of 1:250 overnight at 4 °C, and rinsing with PBS. Slides were incubated with biotinylated goat anti-rabbit IgG at 37 °C for 15 min, rinsed with PBS, incubated with horseradish-peroxidase labled streptomycin, and then stained with DAB and hematoxylin respectively. Slides were finally immersed in differentiation liquid. All slides were dehydrated with graded ethanol and xylene series, then fixed with neutral resin and imaged under a microscope. The IHC results were scored according to the proportion of positively stained cells and staining intensity. The final score was determined by the product of staining intensity (0: negative; 1: weak; 2: moderate; 3: strong) and the percentage of positive cells (0: <5%; 1: 5%–25%; 2: 26%–50%; 3: 51%–75%; 4: >75%). A final score <8 was considered as low expression, and ≥8 was considered as high expression.

Cell lines

The human normal breast epithelial cell line MCF10A and the human breast cancer cell lines MCF7, BT-549, MDA-MB-468 and MDA-MB-231 were obtained from American Type Culture Collection. MCF10A were cultured as previously described (Debnath, Muthuswamy & Brugge, 2003), and other breast cancer cell lines were cultured in RPMI-1640 medium (Gibco-BRL, Karlsruhe, Germany), supplemented with 10% fetal bovine serum (Gibco-BRL) in a 5% CO₂ humidified atmosphere at 37 °C.

Antibodies and western blotting

Total cell lysates were lysed in lysis buffer (Beyotime) containing PMSF and an inhibitor cocktail on ice. Protein concentration was measured using the BCA protein assay kit (Pierce, Rockford, IL, USA). Aliquots containing 30 µg protein separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, then transferred to polyvinylidene difluoride membranes and blocked with 5% non-fat milk for 1 h at room temperature. Membranes were incubated with the corresponding primary antibody (ARHGAP10 at the dilution of 1:2000 and GAPDH at the dilution of 1:5000) at 4 °C overnight, washed with 0.1% Tween20 in TBS, and then incubated with secondary antibody. Primary rabbit anti-ARHGAP10 and anti-GAPDH were purchased from Proteintech Group (55139-1-AP; Wuhan, China) and BIOSS (bs-10900R, China), respectively. The secondary antibody HRP-conjugated goat anti-rabbit IgG was purchased from Abbkine (A21020, China).

Bioinformatics resource and data

The significance of ARHGAP10 for RFS in breast cancer was plotted via Kaplan–Meier Plotter (http://www.kmplot.com) (Gyorffy et al., 2010). A total of 3955 breast cancer patients were divided into high and low expression groups according to the median expression of ARHGAP10.

Oncomine (http://www.oncomine.org), an online database of the transcript level of certain genes (Rhodes et al., 2004), was used to examine the differences in gene expression between breast cancer tissues and normal mammary tissues. The following search parameters were used: ARHGAP10, Cancer vs. Normal Analysis, Breast Cancer, Clinical Specimen, ordered by under-expression, p-value 0.01 and fold change 2.

GEPIA (http://gepia.cancer-pku.cn/), an online website for RNA sequencing expression data of various tumor and normal samples from TCGA and the GTEx projects (Tang et al., 2017), was used for the analysis.

CpG islands were determined by online tool EMBOSS Cpgplot. CpG islands were defined as sequence ranges where the Obs/Exp value was greater than 0.6 and the GC content was greater than 50% within a length more than 200bp (Gardiner-Garden & Frommer, 1987).

Breast cancer data of 1,097 breast cancer cases and 113 normal cases from TCGA were downloaded from https://cancergenome.nih.gov/.

GSEA was performed using the c2.cp.kegg.v6.2symbols.gmt gene set with GSEA software. A total of 1,097 breast cancer samples were divided into two groups according to the median expression of ARHGAP10, namely the ARHGAP10-low group and ARHGAP10-high group. A nominal p-value <0.05 and false discovery rate (FDR) >0.25 were regarded as statistically enriched for a certain gene set.

Statistical analysis

Statistical analyses were performed with SPSS 23.0 (Chicago, IL, USA) and GraphPad Prism 7 software (San Diego, CA, USA). Two tailed Student’s t-test was used to examine the statistical relationship between two cohorts. The chi-square test, Bonferroni correction and Fisher’s exact test were used to examine correlations between clinicopathological parameters and the expression level of ARHGAP10. A p < 0.05 was considered statistically significant except in Bonferroni correction. The p value using Bonferroni correction was represented as p_b.

ARHGAP10 is downregulated at the transcriptional level in breast cancer tissues

The expression of ARHGAP10 was analyzed using data downloaded from TCGA and Curtis Breast in Oncomine (Curtis et al., 2012). The results showed that ARHGAP10 was downregulated according to TCGA (p < 0.001) (Fig. 1A) and Curtis Breast of Oncomine (p < 0.001) (Fig. 1B).

RT-qPCR analysis of 30 paired breast cancer and corresponding adjacent normal tissues was performed to confirm the expression pattern of ARHGAP10 in breast cancer. The result showed that ARHGAP10 was decreased significantly at transcriptional level when compared with paired normal adjacent tissues (p < 0.001) (Fig. 1C).

Intriguingly, ARHGAP10 was not only downregulated in breast cancer, but was also reduced in other cancers including bladder urothelial carcinoma, cervical squamous cell carcinoma and endocervical adenocarcinoma, colon adenocarcinoma, esophageal carcinoma, kidney chromophobe, liver hepatocellular carcinoma, lung adenocarcinoma, lung squamous cell carcinoma, prostate adenocarcinoma, rectum adenocarcinoma, testicular germ cell tumor, uterine corpus endometrial carcinoma and uterine carcinosarcoma according to GEPIA (Tang et al., 2017) (Fig. 1D).

ARHGAP10 protein expression is downregulated in breast cancer tissues

The protein expression of ARHGAP10 was analyzed in 30 paired tumor tissues and corresponding adjacent normal tissues by IHC. ARHGAP10 was located at both the cytoplasm and nucleus (Figs. 2A–2H). The IHC staining results, as determined by the product of the proportion of positive staining cells and staining intensity, showed that low expression of ARHGAP10 was more frequent in breast cancer tissues than in non-cancerous tissues (p < 0.001) (Fig. 2I).

The expression of ARHGAP10 in breast cell lines was evaluated by RT-qPCR and western blotting. The results demonstrated that ARHGAP10 was significantly lower in MDA-MB-231, MDA-MB-468 and BT-549 when compared with MCF10A, and ARHGAP10 showed relatively low-expression trend in MCF7, though it was of no statistical significance at mRNA level (Figs. 2J and 2K). The expression trend of ARHGAP10 in breast cancer cell lines was roughly consistent with that in the clinical tissue samples.

ARHGAP10 expression correlates with Ki-67 index and cTNM stage

To gain insight into the significance of ARHGAP10 in breast cancer, a total of 190 breast cancer samples were analyzed. In these samples, 71% (135/190) of tumor tissues showed low expression of ARHGAP10. Previous studies of ARHGAP10 in cancer did not examine the relationship between the expression of ARHGAP10 at protein level and clinicopathological parameters, which are important for patient prognosis and therapy options. We therefore divided patients into different groups by age, tumor size, ER status, PR status, HER-2 status, Ki-67 index (Ki-67 ≤20% was considered as low Ki-67 index, Ki-67 >20% was considered as high Ki-67 index), lymph node metastasis, pathological grade and cTNM stage according to the medical records with stratified expression level of ARHGAP10. ARHGAP10 expression was significantly correlated with Ki-67 index (p = 0.015) and clinical TNM stage (p = 0.005). Low expression of ARHGAP10 was significantly more frequent in cases with relatively high Ki67 index and advanced clinical stage (p_b = 0.001). However, ARHGAP10 expression was not significantly correlated with age (p = 0.287), tumor size (p = 0.217), ER status (p = 0.366), PR status (p = 0.245), HER2 (p = 0.167), pathological grade (p = 0.379) and lymph node metastasis (p = 0.209). Clinicopathological characteristics and expression level of ARHGAP10 of these breast cancer samples were summarized in Table 1.

Low expression of ARHGAP10 is associated with poor RFS and response to chemotherapy

The relationship between ARHGAP10 and RFS was analyzed using Kaplan–Meier Plotter in 3,955 breast cancer cases (Gyorffy et al., 2010). The results showed that low expression of ARHGAP10 was associated with a lower RFS rate after dividing patients according to the median ARHGAP10 expression (p = 0.0015) (Fig. 3A). This suggested that ARHGAP10 could be a predictor of breast cancer prognosis.

GSE45898 is a gene expression profile of breast cancer patients with contrary responses to epirubicin, cyclophosphamide and docetaxel chemotherapy (Gruosso et al., 2016). In this dataset, breast cancer patients’ biopsies were analyzed before therapy, and the responses to chemotherapy were classified as Good and Poor. Differentially expressed mRNAs were analyzed using GEO2R, and ARHGAP10 was identified as a differently expressed transcript with statistical significance. ARHGAP10 expression was significantly lower in the Poor group and higher in the Good group (p = 0.006) as shown in Fig. 3B. We therefore hypothesized that there was a relationship between the expression of ARHGAP10 and chemotherapy response, and the combination of epirubicin, cyclophosphamide and docetaxel may be less effective in breast cancer patients with low expression of ARHGAP10.

ARHGAP10 is involved in cancer-related signaling pathways in breast cancer.

To further examine the biological function and associated signal pathways that ARHGAP10 possibly involved in, GSEA was performed (Subramanian et al., 2005) using data downloaded from TCGA. A total of 1,097 breast cancer samples were divided into two groups by the median expression of ARHGAP10, namely ARHGAP10-low group and ARHGAP10-high group. The GSEA results indicated that protein export, nucleotide excision repair and base excision repair gene sets were markedly enriched in the ARHGAP10-low group (Figs. 4A–4C). Focal adhesion, JAK-STAT and actin cytoskeleton gene sets were evidently enriched in the ARHGAP10-high group (Figs. 4D–4F) with a nominal p-value <0.05 and FDR <0.25. The normalized enrichment scores of the signal sets were summarized in Fig. 4G.