Background
BC is one of the most prevalent malignancies in women [
1]. The latest cancer statistics showed that BC topped the list of cancer morbidity in Chinese women, with a mortality rate ranking fourth [
2]. Most studies agree that intricate multi-gene networks are involved in the occurrence, development, and metastasis of BC in synergy, accompanied by mutations and/or abnormal genetic activities, such as proto-oncogene activation, apoptosis inhibition of tumor suppressor genes, abnormalities in gene expressions [
3‐
5]. For personalized diagnosis, treatment, and prognosis evaluation of BC, arduous tasks are ahead of us to unveil the underlying genetic mechanisms behind the onset and evolution of BC.
BATF-2, also known as suppressor of activator protein-1 regulated by interferon (
SARI), is a recently discovered tumor suppressor gene using subtractive hybridization in 2008 [
6]. It is a type I interferon (IFN) inducible protein with a leucine zipper and an activator protein (AP)-1 transcription factor family member and has specific structural characteristics to activate transcription factors [
6]. Pieces of evidence show that the BATF-2 gene, though expressed in a variety of normal tissue cells (e.g., melanocytes, astrocytes, pancreatic mesothelial cells, and prostate epithelial cells), can selectively inhibit the growth of tumor cells [
6,
7]. Strikingly, recent studies found its roles in the occurrence and development of various malignancies. For example,
BATF-2 mRNA expression is down-regulated in chronic myeloid leukemia (CML) patients compared to healthy individuals, and BCR-ABL chimeric protein participates in the inhibition of
BATF-2 gene expression [
8]. Abnormalities in BATF-2 and cellular communication network factor 1 (CCN1) expressions and their correlation are closely associated with malignant behaviors of colorectal cancer cells, affecting the prognosis of patients [
9,
10]. Moreover, in other tumors such as hepatocellular carcinoma (HCC) [
11], non-small-cell lung cancer (NSCLC) [
12], esophageal squamous cell carcinoma (ESCC) [
13] and gastric cancer (GC) [
14], the down-regulation of BATF-2 expression is associated with a poor prognosis. Wang et al. found that BATF-2 regulated the epithelial-mesenchymal transition (EMT) and lung adenocarcinoma (LUAD) metastasis [
15]. However, little is known about the roles of BATF-2 in BC. Studies on the expression patterns and clinical implications of BATF-2 in BC are needed.
In contrast to preclinical and clinical studies that are time-consuming, bioinformatics has provided convenience or high efficiency for studies of genetic activity in cancers since the Human Genome Project paved the way [
16]. As gene and protein reaction networks consist of voluminous interactions, bioinformatics is a useful tool for the studies of genomics, proteomics, and other fields [
17]. In this work, we sought to explore the expression patterns and diagnostic and prognostic implications of
BATF2 mRNA and protein expressions in BC using bioinformatics, which were subsequently verified in the serum, serum-derived exosomes, and cancer tissues of BC patients using qRT-PCR and IHC.
Discussion
The occurrence of breast cancer (BC) is a female-prevalent malignancy with massive involvement of intricate proto-oncogene networks and tumor suppressor gene inactivation [
3‐
5]. Current evidence supports
BATF2 as a tumor suppressor gene in various malignancies. Su ZZ et al. first reported that BATF2 overexpression in malignant glioma, melanoma, and prostatic cancer cell lines strongly inhibited the growth and apoptosis of cancer cells, without harm to the survival of noncancerous cells [
6]. Ma H et al. found that the low BATF2 expression was positively correlated with the occurrence and development of liver cancer, and strikingly, all patients with down-regulated BATF2 expression had a poor prognosis [
11]. Li et al. reported that BATF2 expression in prostate cancer was significantly associated with clinicopathological features such as serum PSA levels, clinical stage, and distant metastasis, which could be a critical player in the recurrence and progression of prostate cancer [
26]. Consistently, other studies also suggest that a low BATF2 level is a risk factor for the poor prognosis in non-small cell lung cancer; BATF2 deletion promotes the EMT process, leading to LUAD cell invasion and metastasis [
15].
BATF2 mRNA expression was also significantly down-regulated in cancerous tissues of colorectal cancer: patients negative for BATF2 protein expression often exhibit a poor grade of tumor differentiation, deep invasion, a higher TNM-stage, and a short period of postoperative survival, with significant correlations [
10]. The study of CML showed that CML patients often had lower serum
BATF2 mRNA expression levels than healthy individuals; the down-regulation of
BATF2 gene expression is related to
BCR-ABL inhibition and participates in the occurrence and development of CML [
8]. These mentioned studies imply that BATF2 can be used as a prognostic indicator of patients, a monitoring sensor for tumor therapy, and a potential target in gene therapy.
This work initially assessed BATF2 mRNA and protein expressions as diagnostic and prognostic biomarkers in BC using bioinformatics. Subsequently, these expressions and their clinical implications were fully confirmed in the serum, exosome, and cancer tissue samples of BC patients using qRT-PCR and IHC analyses. In the first step, we comprehensively analyzed
BATF2 mRNA and protein expressions using the expression data from GEPIA and UALCAN and HPA data from TCGA and GTEX. We found the expressions of
BATF2 mRNA and protein in BC tissues were at low-to-moderate levels. The average expression level of
BATF2 mRNA in healthy controls was slightly higher than that in BC cancer tissues, but there was no statistical difference. By contrast, BATF2 protein was mainly located in the nucleus of BC cancer cells based on HPA analysis, with a low-to-moderate level in protein expressions in 7 cases out of 12 cases. The correlation analysis revealed that
BATF2 mRNA was co-expressed with
TAP1,
STAT1, and
PSMB9 in BC. Based on the GEPIA database,
BATF2 mRNA expression was negatively associated with the AR expression and positively correlated with
BRCA2,
Mki67, and
TP53 expressions, with non-significant relationships with
HER2,
EGFR,
AFP and
TP73 expressions. Some studies have confirmed that BATF2 expression is negatively correlated with CCN1 expression and regulates the biological behaviors of cancer cells via regulating CCN1 expression in vivo [
9].
We further evaluated the relationship between the differentially expressed
BATF2 mRNA levels and BC prognosis in UALCAN [
19], OSbrca [
21], Kaplan-Meier Plotter [
22], GEPIA [
18], and other databases. Most databases yielded a non-significant correlation between
BATF2 mRNA expression and the prognosis, but the subgroup analyses uncovered the significantly prolonged OS of TNBC patients with high
BATF2 mRNA expressions versus the low expression group.
BATF2 mRNA expression levels and menopause status were also associated with the survival of BRCA patients. However, information on BATF2 expression and other biomarkers for BC prognosis is currently needed. In external validations, we determined the expression and prognostic value of BATF2 in BC patients using tissue microarray and IHC analysis. The results showed that BATF2 was mainly located in the nucleus of cancer cells of BC, which was consistent with the results of HPA analysis, yet with a lower positive rate of BATF2 protein expression. And all validation samples positive for BATF2 protein showed its expressions at low-to-moderate levels (according to the IHC staining score). Correlations analysis showed that BATF2 protein expression was positively correlated with AR expression, which is in line with the correlation analysis results based on GEPIA. The survival analysis based on tissue microarray data showed that patients with high BATF2 expressions had a longer OS. Therefore, a high BATF2 expression in BC can be a protective factor for the prognosis of patients. However, the survival analysis of BATF2 expressions in HPA database showed that BATF2 expression (high =254 versus low = 821) yielded a
P value of 0.053 in predicting the OS of the breast invasive carcinoma patients. Further investigations are still needed to verify the prognostic significance of BATF2 in BC.
Exosomes are extracellular nanovesicles (30-150 nm) fabricated via a series of regulatory processes, as simplified by “endocytosis - fusion - exocytosis” [
26]. Recent studies have shown that exosomes act as carriers containing miRNA, mRNA, DNA fragments, proteins, and other bioactive substances, involving in various physiological and pathological processes [
27,
28]. It has been proven that exosomes are enriched in the peripheral blood, urine, saliva, ascites, amniotic fluid, and other body fluids; and that tumor-derived or tumor-related exosomes even participate in the regulation of tumor occurrence and development [
29]. The quantification of tumor exosomes can assist in early diagnosis, curative effect evaluation, and the prognosis of tumor patients [
30]. Our previous studies reported that the clinical implications of serum and exosomal
LDHC gene (a CTA molecule) expressions in BC and HCC, serving as an assistant for diagnosis, efficacy evaluation, and recurrence monitoring [
23,
24]. In BC patients, the positive rates of
BATF2 mRNA expressions in the serum and exosomes were 45.00 and 41.67%, respectively, versus 57.14 and 51.80% in healthy controls. Both serum and exosomal
BATF2 showed the AUCs for BC diagnosis of higher than 0.85, which exhibited promising diagnostic values. Similarly, Roe JK et al. reported the
BATF2 transcript level as a single sensitive biomarker in differentiating active pulmonary and extracellular TB from healthy individuals [
30]. Our study, for the first time, confirmed the expression and diagnostic value of serum and exosomal
BATF2 in BC, providing preliminary evidence for further research on the clinical application of BATF2 in BC patients.
Besides, we discovered that the prognosis assessment based on BATF2 tissue microarray did not yield the results of
BATF2 mRNA as a molecular index for BC prognosis prediction. Several explanations can be considered. Firstly, the survival analysis was conducted based on BATF2 protein expressions using IHC scores. Secondly,
BATF2 mRNA expression levels can be inconsistent with its protein levels due to protein posttranslation modifications, and different analyses for protein and mRNA expressions may also contribute to inconsistent results. It is reported that the linear relationship between mRNA and protein expression levels is only about 0.4 to 0.5. For instance, a study by Antonis Koussounadis et al. reported that there was merely a weak correlation between all their measured mRNA and protein expression levels (r = 0.08,
n = 579,
P = 0.07) [
31]. Thus, different measurement analyses, statistical methods, and expression levels cohesively result in inconsistent results.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.