Breast cancer (BC) is the leading cause of cancer and the most common cancer in women worldwide, affecting approximately 12% of females during their lifetimes (1). BC is very heterogenic and classified into distinct molecular subtypes based on hormone receptors, namely, estrogen receptor (ER)/progesterone receptor (PR), and growth factors, including human epidermal growth factor receptor 2 (HER2) and Ki-67 as a proliferation marker (2). Triple-negative BC (TNBC) is a subtype of BC that lacks the expression of ER, PR, and HER-2, is generally aggressive, has high rates of relapse, and results in a decreased overall survival (3, 4). TNBC accounts for 10%–20% of all BCs (5). Given that TNBC lacks the expression of ER/PR/HER2, the application of endocrine therapy or targeted therapy against HER2 is difficult. Chemotherapy has become the main treatment mode, but it generally presents a poor efficacy. Standardized TNBC treatment regimens are still lacking. Therefore, the development of new TNBC treatment strategies has become an urgent clinical need (4). Understanding the molecular profiles of TNBC is critical for the development of new therapeutic options to prevent the progression of metastatic illness and eventually improve the survival of this patient population (6).

TNBC can be categorized into various molecular subtypes, including by the six subtypes by Lehmann (7) or the four subtypes by Burstein (8), based on gene expression profiling of tumor samples (4). Burstein’s four subtypes include luminal androgen receptor (LAR), mesenchymal (MES), basal-like immunosuppressed (BLIS), and basal-like immune-activated (BLIA) subtypes, with BLIA displaying the upregulation of genes controlling B-cell, T-cell, and natural killer cell functions with distinct prognoses (8). The efficacy of existing treatment regimens, therapeutic drugs, and targeted treatment regimens on TNBC subtypes varies (4). Immunotherapies [e.g., programmed cell death protein 1 (PD-1) or Programmed death ligand 1 (PD-L1) inhibitors] may be useful treatments for BLIA tumors, whereas VTCN1-immuno-regulator may be effective treatments for BLIS tumors (9).

Novel biomarkers have been successfully identified by genetic co-expression network (GCN) analysis in BCs (10–24). GCN is an undirected graph, where each node corresponds to a gene, and a pair of nodes is connected with an edge if there is a significant co-expression relationship between them (25). In our previous study (26), we combined the methods of GCN and gene expression profiling to identify six novel immunoglobulin-related gene modules (IGHA1, IGHD, IGHG1, IGHG3, immunoglobulin lambda constant 2 (IGLC2), and IGLJ3) associated with favorable prognosis for TNBC patients. The mRNA expression data sets of 920 BC tumor tissue samples were from the National Center for Biotechnology Information GEO data sets (recurrence n = 354, no recurrence n = 566, and average follow-up time of four data sets = 6–9 years).

IGLC2 had the most significant effects out of six genes; its hazard ratios (HRs) in relapse-free survival (RFS) and distant metastasis-free survival (DMFS) were 0.64 (p = 0.038) and 0.13 (p = 0.025), respectively. These six immunoglobulin genes were involved in the tumor microenvironment of B lymphocytes, which play important roles in BC prognosis, especially in TNBC (27). Growing evidence indicates that variants of immunoglobulin segments are associated with the prognosis of BC (including TNBC) (28, 29). The stromal immunoglobulin kappa chain (IGKC) serves as an immunologic biomarker of initiation, prognosis (30, 31), and treatment response of BCs and other cancers (32). Among the six discovered immunoglobulin genes, IGLC2 has the most significant effect on prognosis, but only its association with lymphoid neoplasia (33) and amyloidosis has been discussed (34). To the best of our knowledge, no study has evaluated the prognostic effect and function of IGLC2 in TNBC. Therefore, we aimed to analyze the function and prognostic value of IGLC2 in TNBC patients.

To our knowledge, this research is the first study to confirm the function and beneficial prognostic effect of IGLC2 on the progression of TNBC. The knockdown of IGLC2 expression increased the proliferation, migration, and invasion of MDA-MB-231 cells. A high IGLC2 gene expression increased the RFS and DMFS in TNBC patients. The prognostic value of IGLC2 was stronger in predicting DMFS than RFS. IGLC2 had a strong prognostic value for lymph node-negative TNBC patients but had no significant effect on lymph node-positive and Grade 3 TNBC ones. IGLC2 was expressed highly in BLIA molecular subtype which was significant with unadjusted p value but became insignificant after correction for multiple comparisons. The results of pathway enrichment analysis indicated that IGLC2 may influence the proliferation and metastasis of TNBC through the PI3K-Akt, MAPK, and ECM–receptor interaction pathways.

The mRNA expressions of IGLC2 and PD-L1 were positively correlated. IGLC2 was more specific to TNBC prognosis, molecular subtypes, and clinical phenotypes than PD-1/PD-L1. Thus, IGLC2 has potential as a prognostic biomarker for the identification of TNBC patients who can benefit the most from immune therapy of PD-1/PD-L1 inhibitor agents. In addition, we confirmed that IGLC2 was expressed in MDA-MB-231 cells rather than simply generated by immune cells, as mentioned in previous studies (26).

IGLC2 is a protein-coding gene. Its protein constitutes the constant region of the immunoglobulin heavy chain. IGLC2 shows a potentially high mutation burden in a pan-cancer context (24, 47). Evidence indicates that the overexpression of immunoglobulin gene signatures (48) leads to a better prognosis of overall survival and disease-free survival in the ER−/HER2− subgroup and TNBCs (49). Other studies showed that the overexpression of a seven-gene module (C1QA, XCL2, SPP1, TNFRSF17, LY9, IGLC2, and HLA-F) has a better prognosis in ER-negative BCs. The downregulation of this module confers a high risk of distant metastasis (HR: 2.02, p = 0.009) that is independent of lymph node status and lymphocytic infiltration (50). Bianchini’s team (51) discovered that B-cell/plasma metagene dominated by immunoglobulin (IGL, IGKC, IGHC3 IGHA1, and IGHG3) has an independently prognostic value in ER-negative BCs, indicating that a high B-cell/plasma metagene score is correlated to favorable DMFS. We observed that IGLC2 and IGKC were highly correlated with each other with a Pearson correlation r close to 1 ( Figure S9 ). IGKC is an independent prognostic biomarker for TNBC patients and a marker of the humoral immune system (31). Tumor-infiltrating plasmablasts and plasma cells were identified as sources of IGKC expression (52). Therefore, we speculated that IGLC2 is a biomarker of the humoral immune system and associated with tumor-infiltrating lymphocytes (TILs). High expressions of IGLC2 reflect the activated humoral immune system.

Immunoglobulin genes are generally regarded to be produced by plasma or B cells rather than tumor cells (53, 54). In this study, IGLC2 was also expressed in tumor cells and played important roles in the prognosis of TNBC. This finding may explain the controversial effects of immunoglobulins on TNBC; Yang (55) stated that high immunoglobulin expression in BCs is correlated with the malignancy and American Joint Committee on Cancer stages of cancers (55), which are contradictory to our findings. This situation implies that various types and origins of immunoglobulins may exert different roles in cancers and thus warrants further study to understand the mechanisms of tumor cells. Growing evidence has indicated that immunoglobulins are not only produced by mature B lymphocytes or plasma cells but also by various normal cell types at immune privileged sites and neoplasms, including non-hematopoietic human cancer cells (55–57) and BC (55). Babbage’s team (53) found rearranged V_H transcripts in the most commonly used BC cell lines. They guessed that BC tumor cells in vivo acquired extraneous genes from neighboring cells and kept them in their genome, or malignant epithelial cells may have initiated the required cascade of complex molecular events to rearrange the VH genes (53).

Rakha’s team observed that tumor size, lymph node stage, and androgen receptor are clinical prognostic markers in TNBC (58). In the lymph node-positive subgroup, the size and androgen receptor retained their prognostic significance. However, in the lymph node-negative tumor subgroup, basal phenotype is the sole prognostic marker identified (58). Thus, the prognostic factors of TNBC varied with lymph node status. Molecular subtypes played major roles in the prognosis of lymph node-negative TNBC patients. We observed that IGLC2 had a strong beneficial prognostic effect on lymph node-negative TNBCs but had no significant influence on lymph node-positive ones. In addition, the IGLC2 expression was associated with TNBC molecular subtypes. Our finding echoed Rakha’s results.

We also observed the high expression levels of PD-1 in lymph node-positive TNBC ( Figure S10 ). PD-1 plays a vital role in inhibiting immune responses. The PD-1/PD-L1 axis inhibits T-cell activation, proliferation, and survival and cytotoxic secretion within cancer cells (59). We speculated that the immune response was suppressed in lymph node-positive TNCB. Thus, IGLC2 no longer reflected the effective prognostic effect of immune response against tumors. Therefore, no significant prognostic value was observed with IGLC2 in lymph node-positive TNBC.

The tumor microenvironment is a complex system formed by distinct and interacting cell populations, and its composition is related to cancer prognosis and response to clinical treatment (60). Immune modules were predictive of the pathological complete response to neoadjuvant chemotherapy in ER-/HER2-BCs (48, 61–63). B cells are present and activated in approximately one quarter of BCs and represent up to 40% of the TIL population in several BCs (49). The success of monoclonal antibody-based immunotherapy indicates the potential for harnessing the humoral immune response in BC treatment (64–67). Immunotherapy has become a promising treatment for TNBC. However, the immune checkpoint inhibitor monotherapy targeting PD-1/PD-L1 shows a mild response for TNBC patients (68). Nevertheless, promising results were found in the combination treatment of immune checkpoint inhibitors and chemotherapy in clinical trials (69). PD-L1 is the only biomarker applied in clinical practice for the selection of patients who are likely to respond to PD-1/PD-L1 immune checkpoint inhibitors. The elevated PD-L1 expression of TNBC tumors predicted an improved response to PD-1/PD-L1 immune checkpoint inhibitor treatment (45). However, the definition of “PD-L1-positive” population in clinical practice remains challenging (69). In addition, assessing whether a cancer is “immune activated” or “immune inactivated” is difficult (70). Therefore, finding new molecular biomarkers for the prediction immunotherapy response for TNBC is an urgent issue.

TNBC is the subtype of BCs most related to TIL infiltration and PD-1/PD-L1 expression (70). Among the TNBC subtypes, BLIA is enriched in immune-response genes, immune activated, and can benefit the most from immune checkpoint inhibitor treatment (69). We observed that the IGLC2 expression was significantly associated with PD-L1 expression and expressed in higher levels in BLIA subtypes.

He’s team identified three TNBC subtypes, namely, Immunity High (Immunity_H), Immunity Medium (Immunity_M), and Immunity Low (Immunity_L) subtypes. Immunity_H is characterized by greater immune cell infiltration and anti-tumor immune activities and better survival prognosis compared with the other subtypes. A high immunity is positivity associated with PD-L1 levels (71). We speculated that IGLC2 is a biomarker of immune cell infiltration and anti-tumor immune activities. Therefore, we observed the positive correlation between IGLC2 and PD-L1. In addition, IGLC2 and PD-1/PD-L1 belong to the immunoglobulin superfamily, and IGLC2 is more specific to TNBC. Compared with PD-1/PD-L1, IGLC2 was more correlated with the prognosis of TNBC patients and more variable with TNCB clinical phenotypes. Thus, IGLC2 may be a potential biomarker for the identification of TNBC patients who are immune activated and will benefit from the current immune therapy.

The results of GO pathway enrichment analysis of IGLC2 indicated that IGLC2 may influence the migration of TNBC via the pathways of receptor binding and ECM organization. Insulin-like growth factor binding was one of the notable pathways identified. Insulin-like growth factor binding protein-3 (IGFBP-3) drives an oncogenic pathway in human TNBC cell lines (7) involving the activation of tyrosine kinase receptor epidermal growth factor receptor (EGFR) and lipid kinase sphingosine kinase (SphK) (72) and is associated with poor prognosis (73). IGFBP-3 promotes the growth of TNBC cells by increasing the EGFR signaling, which is mediated by SphK1, and the combined inhibition of EGFR and SphK1 has potential as an anticancer therapy in TNBC in which EGFR and IGFBP-3 expression is high (74).

In the results of KEGG pathway enrichment analysis, IGLC2 may influence the progression of TNBC via the PI3K-Akt signaling pathway, MAPK signaling pathway, ECM–receptor interaction, HCM, DCM, and ARVC. The findings of MAPK signaling and PI3K-Akt signaling were the same with He’s study, that is, these pathways are hyperactivated in TNBC subtype with high immunity (71). The PI3K/AKT/mammalian target of rapamycin pathway is the most frequently altered pathway in BC (75) and TNBC (76, 77). This pathway has been studied to identify promising new targets for the treatment of TNBC (76, 78–80). Ras-MAPK pathway activation promotes immune evasion and is related to the resistance to conventional chemotherapy in TNBC (81). Several studies have analyzed the targeted inhibitors of the Ras/MAPK pathway to identify potential treatment targets in TNBC (81–85). Increasing evidence emphasizes the crucial role of the ECM in BC progression, invasion, and metastasis (86). The ECM–receptor interaction pathway is associated with a poor prognosis, high metastatic risk (87, 88), and high incidence of chemotherapy resistance of BCs (89). Cardiomyopathy is a common adverse effect of chemotherapeutic agents (i.e., trastuzumab and doxorubicin) for BC (90–92), which may partially explain why IGLC2 is related to HCM, DCM, and ARVC. IGLC2 may serve as a potential biomarker to monitor or reduce cardiomyopathy in BC chemotherapy.

There were some limitations in the study. We only used one TNBC cell line MDA-MB-231 to validate IGLC2. The biological processes related with cell death involved by IGLC2 should be analyzed. Thus, we used multiple independent mRNA data sets from TNBC tissues to support our findings. Our team is still working on the functional analysis of IGLC2. The abovementioned limitations will be included in our future work.

The suppression of IGLC2 gene expression increases the proliferation and migration of MDA-MB-231 cell lines. The function pathways may be involved in PI3K-Akt, MAPK, and ECM–receptor interaction. The high expression of IGLC2 mRNA was related to favorable RFS and DMFS of TNBC patients. IGLC2 exhibited a strong beneficial prognostic effect on lymph node-negative TNBC patients but had no prognostic value for lymph node-positive TNBC patients. The combination evaluation of IGLC2 and clinical lymph node status can provide a precise prognosis prediction of TNBC patients. IGLC2 is positively correlated with PD-L1 and specific to TNBC. IGLC2 may be a potential biomarker for identifying TNBC patients who can benefit the most from immune checkpoint blockade treatment.

Publicly available datasets were analyzed in this study. This data can be found here: The validation data from the mRNA microarray and survival information are available in the online tool KM plotter (http://kmplot.com/analysis/). GSE76275 was from NCBI GEO database available at https://www.ncbi.nlm.nih.gov/geo.

Data curation, Y-TC; Formal analysis, Y-TC, Y-JC and W-ZL; Funding acquisition, C-YY, M-CL and J-MH; Investigation, Y-TC and Y-JC; Methodology, Y-TC and W-ZL; Resources, G-SL, C-YY, J-MH, C-MC, M-CL, H-MH and V-ST; Supervision, C-MC; Validation, Y-TC; Writing – original draft, Y-TC; Writing – review & editing, Y-TC, W-CT, G-SL, Y-JC, M-CL, J-MH, C-CW, G-SL, J-CY, C-MC and H-MH.

This work was supported by the Songshan Branch of Tri-Service General Hospital, National Defense Medical Center [grant numbers 807SB109602] and Taichung Armed Forces General Hospital [grant numbers TCAFGH-E-110040].

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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.