The immunotherapy candidate TNFSF4 may help the induction of a promising immunological response in breast carcinomas

Abstract

Immune checkpoint blockade, an immunotherapy, has been applied in multiple systemic malignancies and has improved overall survival to a relatively great extent; whether it can be applied in breast cancer remains unknown. We endeavored to explore possible factors that may influence immunotherapy outcomes in breast cancer using several public databases. The possible treatment target TNF superfamily member 4 (TNFSF4) was selected from many candidates based on its abnormal expression profile, survival-associated status, and ability to predict immune system reactions. For the first time, we identified the oncogenic features of TNFSF4 in breast carcinoma. TNFSF4 was revealed to be closely related to treatment that induced antitumor immunity and to interact with multiple immune effector molecules and T cell signatures, which was independent of endocrine status and has not been reported previously. Moreover, the potential immunotherapeutic approach of TNFSF4 blockade showed underlying effects on stem cell expansion, which more strongly and specifically demonstrated the potential effects of applying TNFSF4 blockade-based immunotherapies in breast carcinomas. We identified potential targets that may contribute to breast cancer therapies through clinical analysis and real-world review and provided one potential but crucial tool for treating breast carcinoma that showed effects across subtypes and long-term effectiveness.

Figure 1

Analyzing candidate immune checkpoint molecules targeted by blockade in carcinomas. Carcinomas throughout the body were enrolled for analysis, and multiple systems showed various immune checkpoint molecule patterns. PD-L1 (A), CD112 (B), TNFRSF14 (C), TNFSF4 (D), TNFSF18 (E), CD48 (F), and LGALS9 (G) were analyzed for abnormal expression in breast carcinomas. In each image, the specific quantitative value is shown at the top, and each dot represents the expression in samples. The bar graphs below were used for comparison, and bar height bar represents the median expression of the specific tumor type or normal tissue. Specifically, CD112, TNFSF4, TNFSF18, and LGALS9 were relatively overexpressed in breast carcinomas.

Figure 2

Selection of the most representative immune checkpoint molecules targeted by blockade. Expression signatures and correlations were studied using cBioPortal and GEPIA2. (A) Expression heat-map showing the expression of each candidate in a red bar, and the most representative genes are labeled with a red star. The red bars indicate actual data points, and the gray bars indicate that no specific data were available. (B) The analysis flowchart is shown. (C) Nearly all the immunotherapeutic targets showed expanded expression intervals referring to ESR1 and PGR, and TNFSF4, TNFSF18, and CD48 showed increased expression in breast carcinomas without ERBB2, ESR1, and PGR expression. Amplification or mutation of KRAS (D), TP53 (E), or ERBB2 (F) indicated worse predicted survival.

Figure 3

Clinical evaluations of potential therapeutic candidates. Data were acquired from the combined study cohorts of BREAST (METABRIC 2016), BREAST CANCER (MSK 2018), and BREAST INVASIVE CARCINOMA BREAST (TCGA PANCAN 2018). The clinical significance of the cluster including TNFSF18 (A), LGALS9 (B), TNFSF4 (C) and CD48 (D) was analyzed for the value of these molecules as immunotherapeutic targets. The correlations between KRAS and TNFSF18 (E), LGALS9 (F), TNFSF4 (G) or CD48 (H) were analyzed, and the positive correlation between TNFSF4 and KRAS strongly suggested the prospective therapeutic effects of using TNFSF4 as target for manipulation or blockade. (I) A close correlation between TNFSF4 and ALDH1A1 expression was identified. Stem cells with a positive ALDH1A1 phenotype had a shorter survival time (J) and shorter progression-free time (K).

Figure 4

TNFSF4-based immunotherapy may intersect with cancer stem cell signature repression. (A) Schematic figure illustrating the immune response. the immune system becomes suppressed when a carcinoma becomes aggressive. Later, when immune function inhibitors are blocked, active immune cells begin to infiltrate and exert cytotoxic activities against all tumor cells. Spearman correlations between TNFSF4 and immunoinhibitory factors (Y axis) across human cancers (X axis). The items in the column are listed in sequence as follows: ADORA2A, BTLA, CD160, CD244, CD274, CD96, CSF1R, CTLA4, HAVCR2, IDO1, IL10, IL10RB, KDR, KIR2DL1, KIR2DL3, LAG3, LGALS9, PDCD1, PDCD1LG2, PVRL2, TGFB1, TGFBR1, TIGIT, and VTCN1. The infiltrating lymphocyte functions and connective functional factors were analyzed, and both ALDH1A1 overexpression (B) and TNFSF4 overexpression (C) were correlated with more lymphocyte infiltration. Spearman correlations between TNFSF4 and kinds of lymphocytes (Y axis) across human cancers (X axis). The items in the column are listed in sequence as follows: ADORA2A, BTLA, CD160, CD244, CD274, CD96, CSF1R, CTLA4, HAVCR2, IDO1, IL10, IL10RB, KDR, KIR2DL1, KIR2DL3, LAG3, LGALS9, PDCD1, PDCD1LG2, PVRL2, TGFB1, TGFBR1, TIGIT, and VTCN1. However, infiltrating immune cells were suppressed by highly expressed immune inhibitors in cells with increased ALDH1A1 (D) or TNFSF4 expression (E). These results indicated that TNFSF4 blockade treatment could potentially reactivate the immune response and partially function through precisely inhibiting stem cells.

Figure 5

TNFSF4 blockade therapies could be assessed in the stem cell signature-associated mode. The transcriptomic and genomic profiles of pretreated tumor biopsies from responders and non-responders treated with anti-PDL1 and anti-PD1 antibodies were enrolled for analysis. In total, the responders tended to exhibit higher TNFSF4 expression levels, and each study could be reviewed by searching for a certain PMID number (red labeling). Assessment of real-world immunotherapeutic effects indicated that effects related to TNFSF4 (A) tended to imply better outcomes and that TNFSF4-associated TP53 (B), KRAS (C), and ERBB2 (D) all indicated better immunotherapy response. Increased ALDH1A1 expression indicated shorter disease-specific survival (E) and overall survival (F), and ALDH1A1 surprisingly correlated with higher therapeutic response ratios (G) through clinical data assessment. (H,I) ALDH1A1 was analyzed for its roles in predicting immunotherapy response, and the 5 subgroups C1 (N = 369), C2 (N = 390), C3 (N = 191), C4 (N = 92), and C6 (N = 40) were involved in assessing functional aspects. ALDH1A1 expression dominated in all the subtypes, participating in multiple immune reaction processes. (J,K) Associations between ALDH1A1 expression and molecular subtypes across human cancers were also identified, and the signature of increasing ALDH1A1 expression tended to be found in all kinds of breast carcinomas. Specifically, C1 represents wound healing, C2 represents IFN-gamma dominant, C3 represents inflammatory, C4 represents lymphocyte depleted, C5 represents immunologically quiescent and is not shown, and C6 represents TGF-b dominant.

Figure 6

Exploration of the putative clinical roles of TNFSF4. (A) IHC staining images are shown to clarify different expression patterns (left to right, in sequence, < 25%, 25–75%, and > 75%). A lymph node slide was set as a positive control, and an unstained slide was set as a negative control. (B) High RNA expression of TNFSF4 was universally identified in breast carcinoma, with testing and calculation based on FPKM, and the cutoff line is labeled, which was used for clinical predictions. (C,D) Higher TNFSF4 expression pointed to poorer survival outcomes. (E,F) Flow cytometry with FACSAria sorting was applied to isolate stem cells from ZR75-1, MCF-7, and MM-231 cells. (G,H) Stem cells with a CD44⁺/24⁻ or ALDH1A1⁺ phenotype were identified and isolated, and the TNFSF4 expression patterns in different cell lines were checked to illustrate the increased expression.