Background
MM is a B cell differentiated tumor characterized by clonal proliferation of tumor cells [
1‐
3]. MM is a heterogeneous disease with different clinical characteristics [
4]. By recognizing genetic mechanism and mutation, a normal plasma cell transited into the following disease stages: monoclonal gammopathy of undetermined significance, smouldering myeloma, myeloma and plasma cell leukaemia [
5]. The International Staging System (ISS) uses the staging criteria to divide MM into three phases, combining serum albumin levels with β2-microglobulin to determine the prognosis of MM patients [
6]. Revised International Staging System (R-ISS) is a simple and effective prognostic staging system that combines three prognostic tools (ISS: International Staging System, CA: chromosomal abnormalities, LDH: lactate dehydrogenase) to better assess patient prognosis [
7]. Gene expression profiling (GEP) is important for revealing MM molecular heterogeneity of different patients. The molecular classification established by UAMS (The University of Arkansas for Medical Sciences) based on GEP data [
8]. The molecular basis of MM is defined by unsupervised clustering analysis of mRNA expression profiles, which are divided into seven molecular subtypes [
9]. Recurrence is a major problem in most MM patients. Among all MM patients, early recurrence and poor prognosis are about 10% to 15% [
10].
1q21 amplification is the most common chromosomal aberration in MM and is considered a high-risk genetic feature [
11]. The function of the BCAR3 (Breast Cancer Antiestrogen Resistance 3) gene is varied. BCAR3 can promote cell migration, proliferation, and BCAR3 was identified as a molecular linkage between PTP and Cas [
12‐
14]. BCAR3 protein participates in the signaling pathway of EGF through its SH2 domain, leading to cell cycle progression, and BCAR3 itself belong to a mitogenic signaling pathway [
15,
16]. BCAR3 and p130Cas were associated with anti-estrogen in breast cancer, Rac activation. BCAR3 can regulate the Src signal transmission of BCAR3-p130 (cas) complex dependence mode [
17‐
19]. During the initial stage of gonadal development, BCAR3 in gonad development is very important and is expressed in XY gonads [
20]. Studies have shown that the BCAR3 gene is the first spontaneous mutation associated with cataracts caused by lens compression. This new cataract model could provide further knowledge about the function of BCAR3 protein [
21]. In the study of tamoxifen in the treatment of metastatic breast cancer, high expression of BCAR3 is related to good progression-free survival, and the expression level of BCAR3 in primary breast tumors is relatively low, which is related to the survival rate of distant metastasis [
14]. Ovarian cancer is a disease characterized by tumor heterogeneity, which is difficult to be diagnosed and treated. Studies have shown that inhibiting the expression of BCAR3 gene can inhibit the cell proliferation of ovarian cancer [
22]. However, there is no study reporting the relationship between BCAR3 and MM so far. By integrating data of 1878 MM patients, we found that BCAR3 gene is closely related to MM.
Methods
Data source
In our study, gene expression microarrays of 1878 MM patients (1930 samples) were derived from Gene Expression Omnibus database, including datasets GSE24080 (559 samples) [
23], GSE82307 (66 samples) [
24], GSE19784 (308 samples) [
25], GSE83503 (585 samples) [
26], GSE9782 (238 samples) [
27], GSE39754 (136 samples) and GSE19554 (38 samples) [
28,
29]. The criteria for patient selection in our research were stated. 1) All MM patients with the published high throughput gene expression data. 2) All the patients should have some information such as clinical features, biochemical examination, karyotype, therapy or therapy response. The study was approved by the Human Research Ethics Committee of Peking University third hospital. The research was conducted in accordance International Conference on and the Declaration of Helsinki.
Microarray analysis
All microarray data were analyzed, and the significantly abnormal expression genes were systematically screened as predictive biomarkers. The different expression of genes between BCAR3-low to BCAR3-high group were also analyzed and ranked by foldchange values (log2, FC > 0.8 or < − 0.8, P < 0.05).
We retrieved GSE24080 (559 samples) from the NCBI GEO database and analyzed the expression of BCAR3 among different ISS stages, different molecular subtypes, different 1q21 amplification levels, and different survival (EFS and OS). We retrieved GSE82307 (66 samples) from the NCBI GEO database. We analyzed the expression of BCAR3 before and after recurrence in the same patient. We retrieved GSE19554 (38 samples) in 19 MM patients from the NCBI GEO database, and we analyzed BCAR3 expression between the baseline (before chemotherapy) and pre-1st (after induction of chemotherapy, before bone marrow transplantation) in the same sample. We retrieved GSE19784 (308 samples) from the NCBI GEO database and analyzed the expression of BCAR3 in different molecular subtypes. We retrieved GSE83503 (585 samples) from the NCBI GEO database and analyzed the expression of BCAR3 in the recurrence and non-recurrence groups. We retrieved GSE9782 (238 samples) from the NCBI GEO database and analyzed the expression of BCAR3 between the different treatment response of bortezomib and dexamethasone. All patients from GSE39754 (136 samples) were treated with vincristine, adriamycin, and dexamethasone (VAD), followed by autologous stem cell transplantation (ASCT). Then measuring “therapeutic response” after ASCT. Analysis of BCAR3 expression in each treatment response compared to the average of all treatment responses. Therapeutic response: complete response (CR), very good partial response (VGPR), partial response (PR), no response, stable disease (NR), no response, progressive disease (Prog).
Gene ontology (GO) analysis
Use the DAVID to analyze the 559 samples (dataset GSE24080), and find out the enrichment pathways for different expressed genes between BCAR3-low and BCAR3-high group [
30]. The results were ranked by the P value (− log10, P < 0.05).
Statistical analyses
Statistical analyses were performed by R software v3.1.3 (ggplot2 and survminer package). The log-rank test and cox regression multivariate analysis were used for survival analysis. P value < 0.05 was considered statistically significant.
Discussion
For MM patients with age less than 65 years, high-dose treatment with autologous stem cells is a first-line effective treatment [
31]. But MM is a kind of incurable disease and a B-cell differentiated tumor, whose character is cloned plasma cell proliferations, kidney failure, anemia, dissolve the bony lesions, hypercalcemia and infection [
32‐
34]. Free monoclonal immunoglobulins in serum or urine were found to be related with malignant plasma cell marrow infiltration [
35,
36]. BCAR3 is a protein-coding gene, diseases associated with BCAR3 include estrogen resistance, breast cancer, and cataracts. However, the relationship between BCAR3 and MM has not been studied. Therefore, we analyzed the expression level of BCAR3 gene in patients with MM, and found that high BCAR3 expression has better prognosis.
Different mechanisms for BCAR3 in cancer were reported. (1) The interaction between BCAR3 and cas is blocked by the L744E/R748E mutation of BCAR3, the activity of Rac1 is decreased, and the tension of RhoA may be increased, which provides stable adhesion and slows disassembly, thus inhibiting tumor progression. In contrast, BCAR3 and cas interact to activate Rac1, rapidly breaking down adhesion and resulting tumor invasion [
16]. (2) High expression of AND-34/BCAR3 activates the activity of Rac and Pak1, thereby activating the CyclinD1 promoter, making anti-estrogen resistance and progression of breast cancer cells. In contrast, a dominant negative for Rac and Pak1 inhibits the activation of the CyclinD1 promoter [
37]. (3) It is believed that breast tumors initially depend on estrogen for development. Through experiments, BCAR3 is closely related to in vitro anti-estrogen resistance. There is an inverse relationship between BCAR3 expression and ER (estrogen receptor) protein expression in breast and ovarian cells. Tumor cell growth inhibition and hormonal signal blockade are due to the competition of ER by antiestrogens. When BCAR3 is low expressed, ER is highly expressed and inhibits tumor progression [
38]. (4) TGF-β/Smad is a signaling pathway of BCAR3 in breast invasive tumors, and BCAR3 inhibits the conduction of TGFβ/Smad signaling when highly expressed, thereby inhibiting tumor progression [
14]. So, overexpression of BCAR3 in breast cancer cells can promote cell migration and invasion in most researches. However, in some researches, overexpression of BCAR3 in breast cancer cells can inhibit cell migration and invasion. In our research, we found that BCAR3 is closely related to the immune response pathway (Additional file
1: Figure S3B). All the 11 different expressed genes such as CCL18, CD27, CD74, CTSW and CXCL12 were up-regulated in the BCAR3-high group compared with the BCAR3-low group, which suggest that in the immune response pathway was activated (Additional file
1: Figure S4). For example, CXCL12 is a gene in the immune response pathway. The expression of CXCL12 inhibits the metastasis and growth of primary breast cancer [
39]. So, high expression of BCAR3 may inhibit MM growth through up-regulated genes in the immune response pathway.
High dose autologous bone marrow transplantation has been reported and achieved good results [
40]. Pre-relapse mortality was significantly lower than post-relapse [
41]. The survival rate of patients after relapse is reduced. Our study shows that in patients with MM, low expression of BCAR3 at diagnosis can predict early relapse. In addition, MM is a clonal plasmacytoma that can be judged by monoclonal immunoglobulin (IgA, IgG) or increased light chain levels [
42]. The heavy/light chain assay can predict the prognosis and surveille disease in patients with myeloma after treatment [
43]. We also studied serum IgA and serum IgG as well as light chain and further analyzed the expression of BCAR3 at different stages of each serotype. In serum FLC, there was no significant regularity in the expression of BCAR3 between stages I, II and III. However, in the serum IgA group and the serum IgG group, the expression of BCAR3 of stage I MM was higher than that of stage II and stage III.
BCAR3 is an independent prognostic factor for MM, which can reflect the survival rate of patients with MM. We have some important evidence to support this view: First, patients with high BCAR3 expression had higher EFS and higher OS. Second, MM patients with BCAR3 had higher expression before relapse and lower expression after relapse. Last, B2M, LDH, HGB, ASPC, BMPC and MRI have significant differences in the BCAR3-low and BCAR3-high groups (Additional file
2: Table S2). It was found that high expression of BCAR3 gene predicted good prognosis.
The study has some shortcomings. For example, the molecular mechanism of BCAR3 gene in MM is not in-depth studied. We can also further study from this aspect. In addition, MM may be further divided into low risk, moderate risk and high risk combined the expression level of BCAR3 and other known biomarkers, which is helpful to evaluate the survival of patients. These shortcomings needs to be further studied.
Authors’ contributions
HMJ, XLY, WYL and XRZ conceived the project. WLZ, YSL, XNL, YZ and XH analyzed the data. WLZ, YSL, XNL, XH, YZ, WF, ZZY, PY, JW, KH, XRZ, WYL, XLY and HMJ contributed towards the interpretation of the data. All authors read and approved the final manuscript.