Electrostatic complementarity of B-cell receptor CDR3s and TP53-mutant amino acids in breast cancer is associated with increased disease-free survival rates

Breast cancer has long been characterized as a B-cell-related disease with regard to the immune response. For example, data have indicated that B-cell receptor (BCR) recombination read recoveries from breast cancer exome (WXS) files revealed an increased disease-free survival (DFS) rate.¹ However, the antigens stimulating the B-cell response have not been well characterized, particularly via immunogenomics analyses. Thus, we evaluated the net charge per residue (NCPR) for CDR3s, for BCRs expressed on tumor-infiltrating lymphocytes (TILs), to assess its electrostatic complementarity with TP53-mutant amino acids (AAs), because the CDR3 loop domain is the most important part of the BCR polypeptide for antigen-binding specificity.²

Two algorithms, termed the extended and abbreviated approaches (Supporting Online Material (SOM), Methods), were used to determine complementarity scores (CSs) based on the CDR3-mutant TP53 AA electrostatic complementarities. Both algorithms assessed complementarity using the NCPR for the BCR CDR3s and the values for the TP53 mutants based on the difference between the amino acid charges of mutant and wild-type TP53 (SOM, Methods; Tables S1-S5).

Kaplan-Meier (KM) analysis of the DFS for complementary and noncomplementary case IDs revealed statistically significant DFS distinctions according to the case IDs from the CSs determined by both the extended and abbreviated approaches (Fig. 1a, b, respectively), The case IDs representing BCR CDR3-mutant TP53 complementarity correlated with an increased DFS rate. Case IDs showing complementarity based on the abbreviated approach exhibited a statistically significant increase in the DFS rate when compared to all remaining BRCA case IDs (Fig. 1c); moreover, there was no significant distinction between the noncomplementary case IDs versus all remaining case IDs (Fig. S1) using the abbreviated approach. The complementary case IDs representing the extended approach showed a trend towards increased DFS rates in comparison with all remaining IDs (p = 0.078, Fig. S2). Additionally, the noncomplementary case IDs examined with the extended approach had DFS rates similar to those of all remaining case IDs (Fig. S3).

Fig. 1.

Kaplan-Meier (KM) DFS analyses for TCGA-BRCA case IDs representing BCR CDR3-mutant TP53 combinations. a Comparison of the DFS rates for case IDs representing complementary (black) versus noncomplementary (gray) BCR CDR3-TP53-mutant combinations using the extended approach (SOM, Methods; p-value = 0.0292). b Comparison of the DFS rates for case IDs representing complementary (black) versus noncomplementary (gray) BCR CDR3-TP53-mutant combinations using the abbreviated approach (SOM, Methods; p-value = 0.0126). c Comparison of the DFS rates for case IDs representing complementary (black) BCR CDR3-TP53-mutant combinations using the abbreviated approach versus those of all remaining (gray) case IDs (p-value = 0.0330). d Box and whisker plots of RNASeq values used to compare gene expression in BRCA tumor samples representing complementary and noncomplementary CDR3-mutant TP53 combinations. See SOM for the reference set of the following B-cell genes. The means of the RNASeq values for the genes being compared for the complementary (black) and noncomplementary (gray) tumor samples using the extended approach are as follows. CD19 (i): complementary (mean, 145.95); noncomplementary (mean, 64.77). CD22 (ii): complementary (369.96); noncomplementary (194.42). CD72 (iii): complementary (182.02); noncomplementary (132.78). CD79A (iv): complementary (1004. 24); noncomplementary (540.59). CD79B (v): complementary (336.96); noncomplementary (206.87). MS4A1 (vi): complementary (672.25); noncomplementary (325.24). TNFRSF13B (vii): complementary (38.50); noncomplementary (19.01). TNFRSF13C (viii): complementary (21.18); noncomplementary (12.29). TNFRSF17 (ix): complementary (137.95); noncomplementary (80.64). e KM DFS analysis of ovarian cancer (OV) case IDs based on the recovery of BCR CDR3s from OV WXS files. Comparison of the DFS for case IDs representing the recovery of IGH, IGK or IGL BCR CDR3s (n = 29) versus that of all remaining case IDs (n = 470; p-value = 0.022). f KM DFS analysis based on the physico-chemical properties of OV for WXS-based BCR CDR3s. Comparison of the DFS for case IDs from IGH and IGL for the top half of tumor samples with CDR3 AA sequences representative of higher aromaticity (n = 10) versus that for the bottom half (n = 9) of tumor samples with CDR3 AA sequences representative of lower aromaticity (p-value = 0.0176)

To determine whether there was a difference in gene expression between the complementary and noncomplementary case ID tumors, we obtained the RNASeq values for three categories of genes according to the extended approach groupings: (i) immune function genes, (ii) proliferation effector genes,³ and (iii) apoptosis effector genes³ (Table S6). The analyses revealed a statistically significant difference in the RNASeq values for nine genes (Fig. 1d; Table S7), all of which were related to or highly specific for B-cells. To determine whether the differences in the gene expression levels could be due to increased B-cell activation rather than the presence of an increased number of cells, the RNASeq values were normalized to the BCR recombination read recoveries (Table S8), which have been shown to be proportional to lymphocyte infiltration.⁴ The results indicated no significant difference between the complementary and noncomplementary case IDs with regard to the RNASeq values that were normalized to the BCR recombination read recoveries (Table S8), indicating that the increase in the B-cell specific RNASeq values was likely due to an increase in the number of B-cells present in the CDR3-mutant TP53 complementary cases.

The CSs for the CDR3-mutant TP53 combinations were recalculated using the extended and abbreviated approaches with the CDR3 AA sequences represented by IGH recombination reads obtained from TCGA-BRCA RNASeq files generated previously.⁵ RNASeq data for IGH CDR3s that were identical to the recovered IGH CDR3s in the WXS file were eliminated from further consideration to maintain the highest standard for establishing a replicative value when assessing the RNASeq-based IGH CDR3s. Next, the proportion of case IDs representing complementary RNASeq-based, CDR3-mutant TP53 CSs was determined for the set of case IDs representing complementary CSs when the scores were obtained for WXS-based CDR3s. For example, for the extended CS calculation approach, out of 42 RNASeq-based CSs that were complementary, 38 (90%) represented case IDs that also had a complementary CS, as determined according to the WXS-based CDR3s (Tables S9, S10, S11). Overall, this approach indicated the similarity of the RNASeq-based CDR3s and the WXS-based CDR3s obtained from the complementarity scoring process.

The assessment of a variety of clinical parameters potentially overlapping CDR3-mutant TP53 complementarity revealed only an inverse correlation with the fraction of genome altered (Table S12, p = 0.01), suggesting that the complementary CSs were not a surrogate for increased mutational burden.

To determine whether BRCA BCR CDR3s from complementary BCR CDR3-mutant TP53 combinations could potentially be relevant for ovarian cancer, we first determined whether the recovery of BCR recombination reads reflected increased survival rates within the TCGA-OV dataset; this was revealed to be the case (Fig. 1e). Additionally, KM curves based on the aromaticity of OV cancer BCR CDR3 AA sequences, which was assessed as described in reference,⁶ showed that case IDs with IGH and IGL CDR3s with higher aromaticity had improved DFS (Fig. 1f), indicating the relevance of chemical and antigen interactions to the BCRs involved in ovarian cancer. We next determined whether the OV and BRCA datasets contained common TP53 mutants. The comparison of TP53 mutants in the OV and BRCA datasets revealed that approximately 20% of the OV TP53 mutants were the same as the TP53 mutants in the BRCA BCR CDR3-mutant TP53 complementary set using the extended approach (Table S13), suggesting that the more abundant and readily detectable BRCA-related BCR CDR3s could be of relevance to the study and eventual treatment of ovarian cancer.

Overall, the hypothesis that tumor-resident BCR CDR3-mutant TP53 complementarity would be associated with improved survival was consistent with the results obtained above using multiple approaches. Thus, so-called electrostatic complementarity may result in an increase in the immune response that is mediated by B-cells, resulting in an increase in the DFS rate. Such an immune response is most likely associated with an increase in the number of infiltrating B-cells, which in turn would suggest the hypothesis that the increased number of B-cells is due to the stimulation of cell division mediated by BCR activation by TP53 mutants. However, it must be emphasized that the data presented above represented a correlative study and cannot lead to conclusions regarding cause and effect. Additionally, the study does not provide an indication of the complete connection between BCR CDR3-mutant TP53 complementarity and tumor cell death, although such B-cell-related mechanisms of tumor cell death, such as antibody-dependent cell cytotoxicity, have been well studied and documented.⁷

Competing interests

The authors declare no competing interests.

Supplementary information

The online version of this article (10.1038/s41423-019-0328-8) contains supplementary material.