Immunoglobulin superfamily member 10 is a novel prognostic biomarker for breast cancer

Cell culture

The breast cancer cell lines: MDA-MB-231, MCF-7, BT-549, ZR-75-30, SKBR-3, and T47D (ATCC, Manassas, VA, USA) were maintained as previously described (Zhang et al., 2019). The normal mammary epithelial cell line MCF-10A was also maintained as previously described (Debnath, Muthuswamy & Brugge, 2003). All cell lines were cultured in a humidified incubator at 37 °C with an atmosphere containing 5% CO₂.

Patients with breast cancer and tissue samples

TCGA data were utilized as previously described (Qiu et al., 2018). In the present study, we analyzed IGSF10 expression in 1,095 patients with breast cancer in TCGA database. We included 700 patients with breast cancer who had complete RNA-seq data and complete clinical information to analyze the clinical correlation between IGSF10 expression and breast cancer. The following clinical information was collected: age, tumor size, lymph node status, tumor, node, metastasis (TNM) stage, estrogen receptor (ER) status, progesterone receptor (PR) status, human epidermal growth factor receptor 2 (HER2) status, and follow-up information.

Breast cancer tissue samples were collected as previously described (Li et al., 2018). Specifically, we collected 52 pairs of breast tumor and adjacent normal tissues from patients with breast cancer during surgery between 2014 and 2016 at The First Affiliated Hospital of Chongqing Medical University. The collected tissues were used for real-time quantitative polymerase chain reaction (RT-qPCR) and immunohistochemistry (IHC) analyses. All specimens were stored in liquid nitrogen. The collection and use of the tissues were approved by the Institutional Ethics Committees of the First Affiliated Hospital of Chongqing Medical University. The approval number allocated to this study by the Institutional Ethics Committees is 2017 Research Ethics (2017-012).

RNA isolation and RT-qPCR

As described in a previous study (Qiu et al., 2018), we followed the manufacturer’s instructions and extracted the total RNA using TRIzol reagent (Life Technologies Inc., Gaithersburg, MD, USA). RT-qPCR of 21 paired tissues was performed with an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) to examine IGSF10 expression. Relative quantification of the expression of the IGSF10 mRNA was standardized to the expression levels of GAPDH. The following primer pairs were used in the present study:

Forward primer (IGSF10): 5′-TTGGAGTTTGCCTGATGGAAC-3′;

Reverse primer (IGSF10): 5′-CGCTACCCCAACTTTGTTGAAG-3′;

Forward primer (GAPDH): 5′-GGAGCGAGATCCCTCCAAAAT-3′;

Reverse primer (GAPDH): 5′-GGCTGTTGTCATACTTCTCATGG-3′.

IHC

The procedure used for IHC was described in a previous study (Li et al., 2018). An anti-IGSF10 rabbit polyclonal antibody (ab197671, 1:100, Abcam), a secondary antibody (ZSGB 1:100 SPN9001) and HRP (ZSGB 1:100 SPN9001) were used. Thirty-one paired tissues were subjected to IHC. The IHC staining intensity scoring criteria were as follows: 0, none; 1, weak; 2, medium; and 3, strong. The scoring criteria for the proportion of positive tumor cells were as follows: 0, <5%; 1, 5%–25%; 2, 26%–50%; 3, 51%–75%; and 4, >75%. An overall score was derived by multiplying the intensity and percentage scores.

Bioinformatics analyses

The expression of IGSF10 in different subtypes of breast cancer was analyzed using UALCAN, a web portal for evaluating gene expression in different tumor subtypes stratified according to the various clinicopathological features of patients in TCGA database (Chandrashekar et al., 2017).

The expression of the IGSF10 mRNA in different breast cancer datasets was evaluated using Oncomine gene expression array datasets (Rhodes et al., 2004). The cutoff P-value and absolute fold change were defined as 0.01 and 2, respectively.

The relationship between IGSF10 expression and the prognosis of patients with breast cancer presenting different molecular subtypes was analyzed using a Kaplan–Meier plotter (http://kmplot.com/analysis/) (Lánczky et al., 2016). The Affymetrix probe set ID of IGSF10 is 230670_at. Patients were automatically stratified into IGSF10-high and IGSF10-low groups according to the mean expression of the IGSF10 mRNA.

GSEA

This method was described in previous study (Jiao et al., 2018). We performed a GSEA (http://software.broadinstitute.org/gsea) to explore the association between IGSF10 expression and biological processes/pathways according to the instructions of the user guide. We performed the GSEA using a microarray dataset (GSE1456) and TCGA microarray dataset.

Additional statistical analyses

All statistical analyses were performed using SPSS software (version 23.0). OS and RFS were calculated by constructing Kaplan–Meier curves. The differences between two groups were evaluated using Student’s t test. Significance was set to a P-value less than 0.05.

The expression of IGSF10 in breast cancer and its clinicopathological features

We examined the expression of the IGSF10 mRNA in 1095 patients with breast cancer in TCGA database. Based on our results, the IGSF10 mRNA was expressed at higher levels in adjacent normal tissues than in breast cancer tissues (Fig. 1A). We then detected the differences in IGSF10 expression in 21 paired tissue samples using RT-qPCR. Consistent with the results from TCGA database, IGSF10 expression was substantially downregulated in breast cancer tissues (Fig. 1B; Table S1). We collected 31 pairs of breast cancer and corresponding normal tissues and performed IHC. The staining scores of the breast cancer tissues were significantly lower than the adjacent normal tissues (Figs. 1C–1G). Finally, we examined the expression of the IGSF10 mRNA in breast cell lines. IGSF10 was expressed at higher levels in the normal breast epithelial cell line MCF10A than in the breast cancer cell lines (Fig. 1E; Table S2).

Seven hundred patients with breast cancer in the TCGA cohort were analyzed to further confirm the correlation between IGSF10 expression and breast cancer (Table S3). IGSF10 expression correlated with age (P < 0.001), tumor size (P = 0.003), and TNM stage (P = 0.03) (Table 1).

High IGSF10 expression correlated with a better prognosis for patients with breast cancer

The associations of IGSF10 expression with overall survival (OS) and relapse-free survival (RFS) were evaluated using Kaplan–Meier survival curves. Patients in TCGA dataset were stratified by the median IGSF10 mRNA expression level (Table S4). Patients with high IGSF10 expression were significantly more likely to experience prolonged OS (hazard ratio (HR) = 0.63, 95% confidence interval (CI) [0.41–0.97], P < 0.05) (Fig. 2A) and RFS (HR = 0.53, 95% CI [0.30–0.93], P < 0.05) (Fig. 2B) than patients with low IGSF10 expression. Subsequently, we used the UALCAN database to further evaluate the prognostic value of IGSF10 by stratifying patients into different molecular subtypes. Decreased levels of the IGSF10 mRNA were observed in luminal, HER2-positive, and triple-negative breast cancer samples compared with normal samples (Fig. 3A). Low IGSF10 expression was significantly correlated with a shorter OS of patients with basal (HR = 0.44, 95% CI [0.22–0.86], P = 0.013), luminal A (HR = 0.47, 95% CI [0.25–0.88], P = 0.017), and HER2+ (HR = 0.28, 95% CI [0.09–0.81], P = 0.012) breast cancer subtypes (Figs. 3B–3E). However, a significant relationship was not observed between the expression of IGSF10 and OS of patients with the luminal B subtype (HR = 0.61, 95% CI [0.3–1.23], P = 0.17) (Fig. 3D). The multivariate Cox regression analysis of TCGA patients with breast cancer showed that IGSF10 was an independent prognostic factor for OS (HR = 1.793, 95% CI [1.141–2.815], P = 0.011) and RFS (HR = 2.298, 95% CI [1.317–4.010], P = 0.003) (Table 2).

Potential biological roles and signaling pathways related to IGSF10

Potential mechanisms and signaling pathways that may be related to the ability of IGSF10 to regulate the development of breast cancer were explored by conducting a GSEA. According to the median value of IGSF10 expression in the microarray dataset (GSE1456) and TCGA dataset, we assigned patients to two groups. Nine gene sets were enriched in the GSE1456 dataset and 16 gene sets were enriched in TCGA dataset (P < 0.05; false discovery rate (FDR) < 0.25) (Figs. 4A and 4B; Table S5). Interestingly, IGSF10 expression was positively correlated with several cancer-related biological processes, including DNA repair (HALLMARK_DNA_REPAIR), cell cycle (HALLMARK_G2M_CKECKPOINT), and glycolysis (HALLMARK_GLYCOLYSIS) pathways in both datasets (Figs. 4C–4E). The PI3K/Akt/mTOR and mTORC1 signaling pathways were also associated with IGSF10 (Figs. 4F–4G). Moreover, in TCGA dataset, the transforming growth factor-β (TGF-β) signaling pathway (HALLMARK_TGF_BETA_SIGNALING), epithelial mesenchymal transition (EMT) (HALLMARK_EPITHELIAL_MESENCHYMAL_TRANSITION) and tumor necrosis factor (TNF) signaling pathway (HALLMARK_TNFA_SIGNALING_VIA_NFKB) were significantly enriched in the IGSF10-low group (Fig. S1A). These results indicated a possible mechanism underlying the role of IGSF10 in the tumorigenesis of breast cancer.