This article was submitted to Molecular Diagnostics and Therapeutics, a section of the journal Frontiers in Molecular Biosciences
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Breast cancer is most commonly observed in female patients with cancer, which is a main cause of death induced by cancer. The proportion of BRC in all newly diagnosed cancers (
As a multiple class of transcripts, lncRNA has more than 200 bases and no protein-coding potential (
Here, we presented lncRNA ST7-AS1 as a positive prognostic biomarker for BRC patients in public databases such as TCGA. We also investigated the distinctive genomic alterations and functional networks associated with lncRNA ST7-AS1 expression and evaluated its role in tumor immunity. Our work could potentially reveal more direct evidence for hypothesizing lncRNA ST7-AS1 expression as a potential treatment target or a prognostic biomarker in risk stratification and provide insights into the molecular mechanisms for BRC patients.
We obtained the lncRNA ST7-AS1 expression (1,065 cases for the workflow type: HTSeq-FPKM and HTSeq-Counts) and clinical data of BRC projects from TCGA. Those without RNA sequencing data and an overall survival of at least 30 days were excluded. We then converted level 3 HTSeq-FPKM data into TPM (Transcripts Per Million) reads for subsequent analyses. As shown in
Association between lncRNA ST7-AS1 expression and clinicopathologic features in the validation cohort.
Characters | Level | Low expression of ST7-AS1 | High expression of ST7-AS1 |
|
Test |
---|---|---|---|---|---|
N | 533 | 532 | |||
T stage (%) | T1 | 118 (22.2%) | 157 (29.6%) | 0.002 | |
T2 | 325 (61.1%) | 290 (54.7%) | |||
T3 | 64 (12.0%) | 73 (13.8%) | |||
T4 | 25 (4.7%) | 10 (1.9%) | |||
N stage (%) | N0 | 245 (46.9%) | 262 (50.0%) | 0.419 | |
N1 | 175 (33.5%) | 174 (33.2%) | |||
N2 | 66 (12.6%) | 50 (9.5%) | |||
N3 | 36 (6.9%) | 38 (7.3%) | |||
M stage (%) | M0 | 470 (97.3%) | 419 (98.4%) | 0.396 | |
M1 | 13 (2.7%) | 7 (1.6%) | |||
Pathologic stage (%) | Stage I | 82 (15.6%) | 98 (18.9%) | 0.116 | |
Stage II | 299 (57.1%) | 307 (59.3%) | |||
Stage III | 131 (25.0%) | 107 (20.7%) | |||
Stage IV | 12 (2.3%) | 6 (1.2%) | |||
PR status (%) | Negative | 169 (34.0%) | 169 (32.8%) | 0.738 | |
Positive | 328 (66.0%) | 346 (67.2%) | |||
ER status (%) | Negative | 108 (21.7%) | 129 (25.0%) | 0.248 | |
Positive | 390 (78.3%) | 388 (75.0%) | |||
HER2 status (%) | Negative | 269 (73.3%) | 279 (82.5%) | 0.004 | |
Positive | 98 (26.7%) | 59 (17.5%) | |||
PAM50 (%) | Basal | 82 (15.4%) | 108 (20.3%) | <0.001 | |
Her2 | 52 (9.8%) | 30 (5.6%) | |||
LumA | 249 (46.7%) | 302 (56.8%) | |||
LumB | 141 (26.5%) | 61 (11.5%) | |||
Normal | 9 (1.7%) | 31 (5.8%) | |||
Histological type (%) | Infiltrating ductal carcinoma | 418 (85.8%) | 339 (71.8%) | <0.001 | |
Infiltrating lobular carcinoma | 69 (14.2%) | 133 (28.2%) | |||
Race (%) | Asian | 33 (7.0%) | 27 (5.4%) | 0.059 | |
Black or African American | 73 (15.5%) | 106 (21.0%) | |||
White | 366 (77.5%) | 371 (73.6%) | |||
Menopause status (%) | Peri | 18 (3.8%) | 21 (4.4%) | 0.734 | |
Post | 353 (73.5%) | 340 (71.4%) | |||
Pre | 109 (22.7%) | 115 (24.2%) | |||
Anatomic neoplasm subdivisions (%) | Left | 287 (53.8%) | 266 (50.0%) | 0.232 | |
Right | 246 (46.2%) | 266 (50.0%) | |||
TP53 status (%) | Mut | 179 (36.4%) | 156 (33.6%) | 0.408 | |
WT | 313 (63.6%) | 308 (66.4%) | |||
PIK3CA status (%) | Mut | 172 (35.0%) | 142 (30.6%) | 0.172 | |
WT | 320 (65.0%) | 322 (69.4%) | |||
Age (median [IQR]) | 59.00 [49.00,68.00] | 58.00 [48.00,66.00] | 0.179 | Nonnorm |
Next, we used a total of 15 PCR samples of BRC patients admitted to the Harbin Medical University Cancer Hospital from February 2020 to June 2020. The study protocol has obtained the approval of the Ethics Committee of Harbin Medical University Cancer Hospital and is in conformity with the Declaration of Helsinki.
Based on the manufacturer’s protocol, we used the Trizol reagent (Invitrogen) to isolate total RNA from clinical samples. Total RNA was reversely transcribed into cDNA for PCR amplification. By using the SYBR Green I real-time detection kit (Cwbio, Beijing, China), we performed real-time quantitative PCR (RT-PCR) on a CFX96 Detection System (Bio-Rad, California, United States). Specific primers for lncRNA ST7-AS1 and GAPDH were as follows: LncRNA ST7-AS1 forward primers: 5′-ACC CTA CTC TGC CTC CCT TAT C-3′, reverse primers: 5′-TAG CAT CTG CCA CCC AAA TC-3′; GAPDH forward primers: GAA GGT GAA GGT CGG AGT CA, reverse primers: TTG AGG TCA ATG AAGG GGTC.
Based on the multivariate Cox analysis results, we established a nomogram to predict the prognosis of BRC patients. According to the prognosis model, we calculated each patient’s risk score as the total score of each parameter, which could predict the prognosis of BRC patients. The accuracy estimation of nomogram prediction was obtained from a calibration plot. It was found that the bias-corrected line in the calibration plot was close to the ideal curve (Keynesian cross), indicating a strong consistency between predicted values and observed values. The nomogram discrimination was determined using a concordance index (C-index), and 1,000 resamples were used in calculation by bootstrap approach. In this study, all statistical tests were two-tailed, with a statistical significance level of 0.05.
The GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed for high and low lncRNA ST7-AS1 expression groups on the DEGS using the clusterProfiler package (
Using the clusterProfiler package (
We used a GSVA R package to quantify the BRC immune infiltration of 24 tumor-infiltrating immune cells in tumor samples through ssGSEA. According to the 509 gene signatures of 24 tumor-infiltrating lymphocytes (TILs) (
Search Tool for the Retrieval of Interacting Genes (STRING;
R (Version 3.5.1) was used for all statistical analyses and plots. The correlations between clinicopathological characteristics and lncRNA ST7-AS1 expression were evaluated using the Chi-squared test, Fisher exact test, Kruskal–Wallis (KW) test, Wilcoxon signed-rank test, Wilcoxon rank sum test, and logistic regression. Kaplan-Meier method was adopted to draw survival curves [hazard ratio (HR), 95% CI]. Through univariate and multivariate analysis combined with Cox logistic regression models, other clinical factors impacting the survival and the lncRNA ST7-AS1 expression level were found. All reported
To have a better understanding of the correlation of lncRNA ST7-AS1 expression in cancer, we showed the lncRNA ST7-AS1 across all TCGA differential expressions between tumor and adjacent normal tissues (
LncRNA ST7-AS1 expression in patients with BRC. Expression of lncRNA ST7-AS1 in different types of tumors in the TCGA database
In the cohort, 532 cases showed high lncRNA ST7-AS1 expression and 533 cases showed low ST7-AS1 expression. Correlation analysis was carried out to identify the clinicopathologic features and ST7-AS1 expression level. As shown in
Correlation with lncRNA ST7-AS1 expression and clinicopathologic characteristics. Correlation with lncRNA ST7-AS1 expression and clinicopathologic characteristics, including
According to univariate analysis using logistic regression, as a categorical dependent variable, the expression of lncRNA ST7-AS1 (based on the median value) was correlated with poor clinicopathological prognosis. Decreased lncRNA ST7-AS1 expression in BRC was significantly correlated with pathologic stage (OR = 0.74 for Stage III/IV vs. Stage I/II,
LncRNA ST7-AS1 expression associated with clinical pathological characteristics (logistic regression).
Characteristics | Total (N) | Odds Ratio (OR) |
|
---|---|---|---|
T stage (T3 & T4 vs. T1 & T2) | 1,062 | 0.92 (0.67–1.28) | 0.636 |
N stage (N1 & N2 & N3 vs. N0) | 1,046 | 0.88 (0.69–1.13) | 0.321 |
M stage (M1 vs. M0) | 909 | 0.60 (0.23–1.49) | 0.287 |
Pathologic stage (Stage III & Stage IV vs. Stage I & Stage II) | 1,042 | 0.74 (0.56–0.99) | 0.040 |
PR status (positive vs. negative) | 1,012 | 1.05 (0.81–1.37) | 0.689 |
ER status (positive vs. negative) | 1,015 | 0.83 (0.62–1.11) | 0.219 |
HER2 status (positive vs. negative) | 705 | 0.58 (0.40–0.83) | 0.003 |
Histological type (infiltrating lobular carcinoma vs. infiltrating ductal carcinoma) | 959 | 2.38 (1.72–3.30) | <0.001 |
TP53 status (Mut vs. WT) | 956 | 0.89 (0.68–1.16) | 0.371 |
PIK3CA status (Mut vs. WT) | 956 | 0.82 (0.63–1.08) | 0.152 |
We used Cox model to conduct the univariate analysis of prognostic factors for OS. There was a correlation between low expression of lncRNA ST7-AS1 and poor OS (
Univariate regression and multivariate survival model of prognostic covariates in patients with BRC.
Total(N) | HR (95% CI) univariate analysis |
|
HR (95% CI) multivariate analysis |
|
|
---|---|---|---|---|---|
T stage (T3 & T4 vs. T1 & T2) | 1,061 | 1.673 (1.152–2.429) | 0.007 | 2.102 (0.976–4.528) | 0.058 |
N stage (N1 & N2 & N3 vs. N0) | 1,045 | 2.145 (1.497–3.073) | <0.001 | 1.315 (0.653–2.648) | 0.443 |
M stage (M1 vs. M0) | 909 | 4.327 (2.508–7.465) | <0.001 | 2.121 (0.636–7.075) | 0.221 |
Pathologic stage (Stage III & Stage IV vs. Stage I & Stage II) | 1,041 | 2.519 (1.787–3.549) | <0.001 | 2.411 (1.058–5.493) | 0.036 |
PR status (negative vs. positive) | 1,011 | 1.312 (0.931–1.849) | 0.120 | ||
ER status (negative vs. positive) | 1,014 | 1.420 (0.983–2.052) | 0.062 | 2.931 (1.632–5.262) | <0.001 |
HER2 status (negative vs. positive) | 705 | 0.621 (0.378–1.019) | 0.059 | 1.303 (0.661–2.566) | 0.445 |
Age (>60 vs. <=60) | 1,064 | 2.036 (1.468–2.822) | <0.001 | 3.388 (1.743–6.584) | <0.001 |
Race (White vs. Asian & Black or African American) | 975 | 0.880 (0.593–1.306) | 0.526 | ||
Histological type (infiltrating lobular carcinoma vs. infiltrating ductal carcinoma) | 959 | 0.860 (0.546–1.355) | 0.516 | ||
Anatomic neoplasm subdivisions (right vs. left) | 1,064 | 0.776 (0.559–1.077) | 0.130 | ||
Menopause status (post vs. pre & peri) | 955 | 2.405 (1.445–4.002) | <0.001 | 2.374 (0.986–5.714) | 0.054 |
TP53 status (Mut vs. WT) | 955 | 1.218 (0.858–1.730) | 0.269 | ||
PIK3CA status (Mut vs. WT) | 955 | 1.015 (0.696–1.479) | 0.938 | ||
ST7-AS1 (high vs. low) | 1,064 | 0.576 (0.414–0.801) | 0.001 | 0.541 (0.304–0.962) | 0.037 |
Based on the median expression of lncRNA ST7-AS1, the cohorts included the low expression subgroup and the high expression subgroup. Based on the KM Plotter in TCGA, the OS of patients with high expression level of ST7-AS1 was significantly longer than that of patients with low expression level of ST7-AS1 [HR = 0.58 (0.41–0.80),
Association between lncRNA ST7-AS1 expression and BRC patients’ outcomes.
Prognostic values of differential expression of lncRNA ST7-AS1 in different subgroups, including anatomic neoplasm subdivision-right, menopause status-post, ER-positive, PR-positive, HER2-positive, infiltrating ductal carcinoma, infiltrating lobular carcinoma, age >60°years, race-white, M stage-M0, N stage N2 and N3, T stage-T2, T stage-T3 (all
Prognostic value of differential expression of lncRNA ST7-AS1 in different subgroups. Prognostic value of differential expression of lncRNA ST7-AS1 in different subgroups, including
In order to provide clinicians with predicted prognosis of BRC patients quantitatively, we established the nomogram combining lncRNA ST7-AS1 and independent clinical risk factors (pathologic stage, ER status, and age). Higher total points on the nomogram for OS, progression-free interval (PFI), and DSS, respectively, indicated a worse prognosis (
Construction and performance validation of the lncRNA ST7-AS1-based nomogram for BRC patients. Nomogram to predict
Based on the calibration curve of nomograms for OS, PFI, and DSS, the predictions conformed well to observations in all patients, and the test showed no deviation from the perfect fit. The nomogram had a C-index of 0.750 and contained 1,000 bootstrap replicates (95% CI: 0.727–0.773) for OS. We also found PFS (C-index: 0.703, CI: 0.676–0.730) and DSS (C-index: 0.778, CI: 0.747-0.809). It was found that the bias-corrected line in the calibration plot was close to the ideal curve (Keynesian cross), indicating a strong correlation between predicted values and observed values (
The DESeq2 package was used to analyze the data from TCGA in R (adjusted
Identifying differentially expressed genes between high and low expression of lncRNA ST7-AS1 groups.
ClusterProfiler was used for GO and KEGG enrichment analyses of the functions of lncRNA ST7-AS1 associated DEGs in BRC. The top GO enrichment items were classified into three functional groups. Molecular functions (MF) mainly involve receptor ligand activity, peptidase regulator activity, endopeptidase regulator activity, peptidase inhibitor activity, and endopeptidase inhibitor activity. The cellular components (CC) were mainly transcription regulation by intermediate filament cytoskeleton, intermediate filament, keratin filament, and cornified envelope. Epidermal development, skin development, epidermal cell differentiation, keratinocyte differentiation, keratinization, and cornification were genes related to biological processes (BP). Based on KEGG enrichment analysis, we found the correlation between retinol metabolism, metabolism of xenobiotics by cytochrome P450, drug metabolism–cytochrome P450,
Functional annotation and prediction of signaling pathways.
We conducted GSEA between high lncRNA ST7-AS1 expression set and low lncRNA ST7-AS1 expression set to identify different signaling pathways in BRC, which revealed marked differences (FDR<0.25, adjusted
Enrichment plots from the GSEA. GSEA results showing
The Spearman correlation was used to show the correlation of ST7-AS1 expression level (TPM) with immune cell infiltration level quantified by ssGSEA in the BRC tumor microenvironment. We found that ST7-AS1 expression was negatively correlated with the abundances of immunocytes (Th2 cells, Tgd, macrophages, Tcm, etc.) and positively correlated with abundances of innate immunocytes (pDCs, NK cells, B cells, CD8 T cells, etc.) (
The lncRNA ST7-AS1 expression was correlated with immune infiltration in the tumor microenvironment.
LncRNA is emerging as a tumor suppressor or oncogene in a number of tumors, as well as a potential therapeutic molecular target or biomarker with prognostic value (
Numerous studies have shown that lncRNA expression level plays an important part in the occurrence and progression of cancer (
In order to make a more accurate prognostic prediction, nomograms have been established, which had a better performance than conventional staging systems in some cancers. In our study, a nomogram was established using the results of multivariate Cox analysis which combined lncRNA ST7-AS1 with independent clinical risk factors (pathologic stage, ER status, and age). The calibration plot showed a good consistency between actual values and predicted values for 1-, 2-, and 3-year OS, DFI, and DSS. We constructed the model based on a complementary perspective of each tumor to provide personalized scores for individual patients.
There is another new finding from this study; that is, lncRNA ST7-AS1 participates in the cell cycle, apoptosis, and DNA repair. Genomic instability and mutagenesis belong to basic characteristics of cancer cells. Kinases and their associated signaling pathways conduce to the stabilization and repair of genomic DNA (
Equally important, our study used ssGSEA and Spearman correlation to reveal the association between lncRNA ST7-AS1 expression and immune infiltration levels in BRC. We found the association between lncRNA ST7-AS1 and T helper cells and DC cells. Many tumors, including lung cancer, glioma, cervical cancer, BRC, gastric cancer, and colorectal cancer, all have Th1/Th2 balanced drift in the body, and Th2 cells are often dominant, which may be related to the immune escape of tumors (
Although the approach used in this study helps to understand the relationship between lncRNA ST7-AS1 and BRC, some limitations still exist. Firstly, to fully and comprehensively elucidate the specific role of lncRNA ST7-AS1 in BRC development, various clinical factors, such as details of patients’ treatments, should be considered. However, because the experiments are carried out in different laboratories, the public database is lack of such information or the treatments are inconsistent. Secondly, the number of healthy subjects used as controls differs significantly from that of cancer patients in this study, so further studies should be carried out to maintain a balanced sample size. In conclusion, limitations still exist in retrospective studies, especially the inconsistency of interventions and the lack of certain information. Therefore, future prospective studies are needed to eliminate analysis bias of this study which is retrospective in nature. Because this study was based on RNA sequencing data from the TCGA database, we could not elucidate the lncRNA ST7-AS1 expression at the protein level, nor could we clearly evaluate the direct mechanism by which lncRNA ST7-AS1 participates in the development of BRC. Therefore, the direct mechanism in BRC should be further studied.
To conclude, our study showed decreased expression of lncRNA ST7-AS1 in BRC samples. Furthermore, low expression of lncRNA ST7-AS1 was associated with adverse outcomes. We also established a nomogram using lncRNA ST7-AS1 to predict 1-, 3-, or 5-year survival in patients with BRC. In terms of biological functions, we showed the involvement of lncRNA ST7-AS1 in cell cycle, DNA repair, and immune cell infiltration in the BRC immune microenvironment. We found the correlation of lncRNA ST7-AS1 with T helper cells and DC cells. In conclusion, these results indicate that lncRNA ST7-AS1 has a potential role in BRC as a prognostic marker and therapeutic target.
Publicly available datasets were analyzed in this study. This data can be found here: The Cancer Genome Atlas (
The studies involving human participants were reviewed and approved by the Ethics Committee of Harbin Medical University Cancer Hospital and are in conformity with the Declaration of Helsinki. The patients/participants provided their written informed consent to participate in this study.
QZ conceived and designed the experiments. ZZ, HZ, DL, XZ, and JW analyzed the data. ZZ wrote this manuscript. All authors read and approved the final manuscript.
This study was financially supported by Grants 81730074 and 81672599 from National Natural Science Foundation of China and the National Science Foundation of Heilongjiang Province of China for Returnees (Grant No. LC2017037).
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.