Comprehensive characterization of B7 family members in breast cancer: B7-H5 switch reverses breast cancer from “immuno-cold” into “immuno-hot” status

Expression and scRNA-seq distribution of B7 family members in BrCa

The members of the B7 family and their aliases are shown in Table 1. We initially used the TCGA database to compare the expression of B7 members in normal and BrCa samples. Compared with normal tissues, eight B7 family members demonstrated differential expression in BrCa, including B7-1, B7-2, B7-DC, B7-H1, B7-H3, B7-H4, B7-H5, and B7-H6 (Fig. 1A). In addition, we enrolled 10 BrCa patients from the Wuxi Maternal and Child Health Hospital for scRNA-seq analysis to investigate the distribution of different cell types in BrCa TIME, including B cells, basal cells, endothelial cells, fibroblasts, luminal cells, mast cells, myeloid cells, pericytes, plasma cells, and T/natural killer (NK) cells (Fig. 1B). Next, we explored the association between B7 family member expression and scRNA-seq distribution in BrCa TIME. The cell distribution of B7-H7 was almost invisible. B7-1 and B7-2 were mainly concentrated in myeloid and B cells. The distribution of B7-H1, B7-H2, and B7-H6 was similar, and they were distributed in cell subsets depicted in the diagram. B7-DC was mainly present in myeloid cells and fibroblasts. B7-H4 was mainly found in luminal cells, whereas B7-H3 was significantly enriched in several subgroups except for T/NK and B cells. B7-H5 was largely distributed in all involved cell subsets, except in luminal cells (Fig. 1C).

Fig. 1figure 1

Expression and scRNA-seq distribution of B7 family members in BrCa (A) The expression of the B7 family in normal and BrCa tissues using the TCGA database. (B) The scRNA-seq analysis of 10 BrCa patients shows the distribution of different cell types in BrCa TIME. (C) The relationship between B7 family expression and scRNA-seq distribution in BrCa TIME.

Association of B7 family with clinicopathological parameters and survival in BrCa

We next analyzed the correlation between B7 family members and clinicopathological parameters in BrCa, including TNM stage, T stage, N stage, and M stage (Fig. 2). According to the TNM stage, B7-H3 demonstrated significant differential expression among different TNM stages (p < 0.05). In addition, based on the T stage, B7-DC (p<0.05), B7-H4 (p<0.0001), and B7-H7 (p<0.01) demonstrated significant differential expression among different T stages. In addition, the expression of B7-H3 (p < 0.05), B7-H5 (p < 0.001), and B7-H6 (p < 0.01) was related to the N stage, whereas the expression of B7-H3 (p < 0.05) correlated with the M stage. Further, we investigated the relationship between the B7 family and survival in BrCa, including overall survival (OS) and disease-free survival (DFS). Thus, no significant association was present between B7 family members and patients with OS and DFS (See Fig. 3).

Fig. 2figure 2

Association between B7 family and clinicopathological parameters and overall survival in BrCa. (A) Expression of 10 B7 family members among different TNM stages in BrCa. (B) Expression of 10 B7 family members among different T stages in BrCa. (C) Expression of 10 B7 family members among different N stages in BrCa. (D) Expression of 10 B7 family members among different M stages in BrCa

Fig. 3figure 3

Association between B7 family expression and survival in BrCa. (A) Association between B7 family expression and overall survival (OS) in BrCa. (B) Association between B7 family expression and disease-free survival (DFS) in BrCa

Functional enrichment and signaling pathways of B7 family members in BrCa

The median of B7 family members was used to perform differential analysis between low-expressed and high-expressed groups, whereas differentially expressed genes of B7 family members were presented using the volcano plot (Fig. 4A). Next, we conducted the gene ontology (GO) enrichment analysis of B7 family members, including biological process (BP), cellular component (CC), and molecular function (MF) (Fig. 4B). The gene set enrichment analysis (GSEA) enrichment analysis of B7 family members was performed for potential signaling pathways (Fig. 4C).

B7-1 was enriched in immune response-activating cell surface receptors, external side of the plasma membrane, and protein serine or threonine kinase activity, whereas B7-1 was enriched in ribosomes, oxidative phosphorylation, JAK–STAT signaling pathway, T cell receptor signaling pathway, NK cell-mediated cytotoxicity, and Parkinson’s disease. Similarly, B7-2 was enriched in leukocyte migration, the external side of the plasma membrane, and nucleoside-triphosphatase regulator activity, whereas B7-2 was enriched in ribosomes, cytokine, asthma, viral myocarditis, glycosylphosphatidylinositol (GPI) anchor biosynthesis, and oxidative phosphorylation. B7-H1 was enriched in the regulation of lymphocyte activation, the external side of the plasma membrane, and antigen binding, whereas B7-H1 was enriched in the ribosomes, oxidative spliceosomes, Leishmania infection, viral myocarditis, and cytokine–cytokine receptor interaction. B7-DC was enriched in leukocyte migration, endosome membrane, and cell adhesion molecule (CAM) binding, whereas B7-DC was enriched in oxidative phosphorylation, ribosome, CAMs, viral myocarditis, Leishmania infection, and Parkinson’s disease. B7-H2 was enriched in phagocytosis, immunoglobulin complex, and antigen binding, whereas B7-H2 was enriched in intestinal immune network for IGA production, melanogenesis, cytokine, cytokine receptor interaction, endocytosis, neurotrophin signaling pathway, and cell cycle. B7-H3 was enriched in extracellular structure organization, collagen-containing extracellular matrix, and CAM binding, whereas B7-H3 was enriched in basal cell carcinoma, leukocyte, neuroactive ligand–receptor interaction, taste transduction, pyrimidine metabolism, and amyotrophic lateral sclerosis. B7-H4 was enriched in extracellular structure organization, cell–cell junction, and CAM binding, whereas B7-H4 was enriched in the proteasome, base excision repair, DNA replication, melanogenesis, hedgehog signaling pathway, and leukocyte transendothelial migration. B7-H5 was enriched in the regulation of lymphocyte activation, cell–substrate junction, and CAM binding, whereas B7-H5 was enriched in systemic lupus erythematosus, CAMs, cytokine–cytokine receptor interaction, GPI anchor biosynthesis, nucleotide excision repair, and mismatch repair. B7-H6 was enriched in proteasomal protein catabolic process, mitochondrial inner membrane, and small GTPase binding, whereas B7-H6 was enriched in ribosomes, Parkinson’s disease, oxidative phosphorylation, taste transduction, axon guidance, and oocyte meiosis. B7-H7 was enriched in the regulation of lymphocyte activation, the external side of the plasma membrane, and antigen binding, whereas B7-H7 was enriched in spliceosomes, RNA degradation, primary immunodeficiency, GPI anchor biosynthesis, intestinal immune network for IGA production, and autoimmune thyroid disease.

Fig. 4figure 4

Functional enrichment and signaling pathways of B7 family members in BrCa. (A) The volcano plot depicts differentially expressed genes of B7 family members in BrCa. (B) The GO functional enrichment analysis of B7 family members in BrCa. (C) The GSEA signaling pathway analysis of B7 family members in BrCa

Genomic alterations of B7 family members in BrCa

We used the cBioPortal database to investigate the relationship between B7 family members and genomic alterations in BrCa. With respect to the genomic alteration frequency, the mutation ratios of B7-H5 and B7-H6 were relatively higher (up to 3%), whereas the mutation rates of B7-1 and B7-2 were lower, less than 1% (Fig. 5A). Genomic alterations of B7 family members in different BrCa-related researches were presented, including structural variant data, mutation data, and CNA data (Fig. 5B). In addition, the lollipop plot of B7 family members in BrCa is depicted (Fig. 5C).

Fig. 5figure 5

Genomic alterations of B7 family members in BrCa. (A) The mutation rate of B7 family members in BrCa. (B) The genomic alterations of B7 family members in different BrCa-related researches. (C) The lollipop plot of B7 family members in BrCa

Correlation between B7 family members and TIME in BrCa

We next analyzed the relationship between the B7 family and TIME in BrCa, including TIME scores, immune cell infiltration, and immune functions. TIME score included estimate score, immune score, and stromal score, whereas the immune cell infiltration included 23 kinds of immune cells. The immune functions consisted of antigen-presenting cell (APC) co-inhibition, APC co-stimulation, CCR, checkpoint, cytolytic activity, HLA, promotion of inflammation, MHC class I, parainflammation, T cell co-stimulation, T cell co-inhibition, type I interferon (IFN) response, and type II IFN response. According to the TIME scores, 10 B7 family members exhibited significant differences in estimate scores and immune scores, whereas 8 B7 family members displayed significant differences in stromal scores, except for B7-H2 and B7-H6 (Fig. 6A-C). In addition, the relationship between B7 family members and immune cell infiltration, and immune functions were presented (Fig. 7A-B).

Fig. 6figure 6

Correlation between B7 family members and TIME in BrCa. Relationship between B7 family members and TIME estimate score. (B) Relationship between B7 family members and TIME immune score. (C) Relationship between B7 family members and TIME stromal score

Fig. 7figure 7

Correlation between B7 family members and immune cell infiltration and immune functions. (A) Relationship between B7 family members and immune cell infiltration. (B) Relationship between B7 family members and immune function

Expression of B7-2, B7-H3, and B7-H5 in BrCa

We recruited a clinical cohort from the Wuxi Maternal and Child Health Hospital. The preliminary analysis revealed the correlation between the B7 family and the clinical stage. B7-H3 and B7-H5, two molecules with clinical research potential, as well as B7-2, a B7 family member with little research in BrCa, were selected for immunohistochemical and mIHC analyses. Figure 8A depicts representative images of B7-2, B7-H3, and B7-H5 expression in mIHC-stained tumor tissue. Representative images stained with B7-2, B7-H3, and B7-H5 were divided into low expression group and high expression group for display. DAPI staining results were used as a reference. Statistical analysis of H-score staining results in BrCa microarray tumor (n = 63) and adjacent tumor (n = 20) tissues demonstrated that the expression of B7-2 and B7-H3 in BrCa tissues was significantly higher than in adjacent tumor tissues, whereas no statistical difference was found in B7-H5 between adjacent tumor tissues and BrCa tissues (Fig. 8B). In addition, we used multiple subtypes of BrCa cells to further verify, whereas we used normal breast cells MCF-10 A as controls. The results of western blotting and qRT-PCR were consistent, indicating that B7-2 and B7-H3 were highly expressed in BrCa, and the expression of B7-H5 was low in BrCa (Fig. 8C-D).

According to the clinical cohort in Wuxi Maternal and Child Health Hospital, the relationship between clinical features and expression of B7-2, B7-H3, and B7-H5 was analyzed using the chi-square test or Fisher’s test. The statistical results demonstrated that B7-2, B7-H3, and B7-H5 in BrCa tissues were not statistically correlated with clinical features (Table 2).

Table 2 Association between B7-2, B7-H3, B7-H5, and clinical features, infiltration of CD8+immune cellsFig. 8figure 8

B7-2, B7-H3, and B7-H5 expression in BrCa. (A) Representative images of B7-2, B7-H3, and B7-H5 expression in BrCa tissues stained by mIHC. All representative images were divided into low expression group and high expression group. The DAPI staining results are listed as a reference. (B) The expression of B7-2, B7-H3, and B7-H5 in several BrCa cells was assessed by qRT-PCR. (C) The expression of B7-2, B7-H3, and B7-H5 in several BrCa cells was assessed by western blotting. Normal breast cell line MCF-10 A was used as a reference. Significance was calculated using Student’s t-test

B7-H5 is significantly correlated with CD8+cell infiltration and has the potential to predict immunothermal tumors

CD8+ immune cells were divided into three layers with high, medium and low expression. We used the clinical cohort in Wuxi Maternal and Child Health Hospital to study the relationship between infiltration of CD8+ immune cells and expression of B7-2, B7-H3, and B7-H5 using chi-square test or Fisher’s test. We found that B7-H5 significantly correlated with CD8+ cells (p < 0.001), as shown in Table 2. We demonstrated the representative images of B7-H5 and CD8 high, middle, and low layers. As depicted in Fig. 9A, the chip diagram (40 ×) uses a white box to select typical mIHC images, and four single-channel display images of DAPI\B7-H5\CD8+ (200 ×) are conducted. Results demonstrated that CD8+ infiltration increased significantly with an increase in B7-H5 expression. We believe that B7-H5 is a potential molecule in predicting thermal tumors, that is, immune invasive tumors.

We performed a cluster analysis to divide the expression of B7H5 and CD8 into cluster 1 and cluster 2, in which cluster 2 referred to high expression of B7H5 and high CD8+ (Fig. 10A-C). The risk classification of patients with BrCa was performed based on the expression of B7-H5 and CD8+, which was further assessed in terms of TIME, degree of immune cell infiltration, and immune functions (Fig. 10D-F).

Fig. 9figure 9

B7-H5 and CD8 + expression in BrCa. (A) Representative images uncovering B7-H5 and CD8+ expression in tumor tissues using mIHC staining. The images are divided into three levels: B7-H5 low and CD8+ low, B7-H5 mid and CD8+ mid, and B7-H5 high and CD8+ high. The DAPI staining results are used as a reference

Fig. 10figure 10

Association between B7-H5 expression and CD8+ immune cell infiltration in BrCa patients. (A-C) Risk classification of patients with BrCa according to B7-H5 and CD8+ expression. (D) TIME estimate score of risk classification in BrCa. (E) TIME immune score of risk classification in BrCa. (F) TIME stromal score of risk classification in BrCa. (G) Immune cell infiltration of risk classification in BrCa. (H) Immune functions of risk classification in BrCa

留言 (0)

沒有登入
gif