Tumor-specific CD8 T cell characterization in HR+ breast cancer reveals an impaired antitumoral response in patients with lymph node metastasis

Abstract

Most breast cancers express the estrogen receptor (ER), but the immune response of hormone receptor-positive (HR⁺) breast cancer remains poorly characterized. Here, dendritic cells loaded with tumor lysate are used to identify tumor-reactive CD8 T cells, which are detected in most HR⁺ breast cancer patients, especially those with early-stage tumors. When present, the circulating antitumor CD8 response contains cytotoxic T cells with diverse specificity and T cell receptor (TCR) repertoire. Additionally, patients with blood cancer-specific T cells have significantly more CD8 tumor-infiltrating lymphocytes (TILs). Moreover, tumor-reactive TCR sequences are detected in the tumor, but at a significantly lower proportion in patients with lymph node involvement. Our data suggest that HR⁺ breast cancer patients with lymph node metastasis lack tumor-specific CD8 T cells with capacity to infiltrate the tumor at significant levels. However, early-stage patients have a diverse antitumor CD8 response that could be harnessed to develop immunotherapeutic approaches for late-stage HR⁺ patients.

Keywords: CDR3; CTA; ER+ breast cancer; T cell receptor; T cell repertoire; antitumor CD8 T cells; cancer testis antigen; metastatic luminal breast cancer; moDC; tumor lysate; tumor-specific T cells.

Graphical abstract

Figure 1

A circulating tumor-reactive CD8 T cell response can be found in most early-stage HR+ breast cancer patients (A) Schematic of proliferation assay and representative flow cytometry dot plot. moDCs were generated from blood monocytes, loaded with tumor lysates and cocultured with blood T cells. (B) Graph showing relative T cell proliferation against the tumor lysate, in patients with different histological subtypes (ductal n = 14; lobular n = 7; mucinous n = 2). (C) Relative proliferation against the tumor lysate in patients divided by the presence of lymph node metastasis (yes n = 10; no n = 12; p = 0.033; Mann-Whitney test). (D) Graph of the tumor lysate-induced proliferation in patients with different histological subtypes and lymph node metastasis status (ductal no metastasis n = 5; ductal with lymph node metastasis n = 8; other subtype no metastasis n = 7; other subtype with lymph node metastasis n = 2). ∗p < 0.05.

Figure 2

Tumor-reactive CD8 T cells can kill tumor cells and contain CTA-specific T cells (A) CD107a and TNF-⍺ expression on Br1 tumor-reactive T cell clones cultured in the presence of different breast cancer cell lines in three independent experiments (n = 3). Data are represented as the mean ± SEM. Kruskal-Wallis test with Dunn’s multiple-comparison tests. Br1 clone 1: MDA-MB-231 vs. unstimulated p = 0.0277; Br1 clone 2: HCC1937 vs. unstimulated p = 0.0269. (B) CD107a and TNF-⍺ expression on Br1 tumor-reactive T cell clone 1 cocultured with MDA-MB-231 in the presence of antibodies to block MHC-I, HLA-A2, or an isotype control antibody (n = 1). (C) Killing ability of Br1 tumor-reactive T cell clones by Incucyte at different effector:target (E:T) ratios. One-way ANOVA with Tukey’s multiple-comparison tests on the values after 36 h coculture (n = 3). MDA-MB-231 + Br1 clone 1: 2:1 vs. 1:1 p = 0.0268; 1:1 vs. 1:2 p = 0.0163; 2:1 vs. 1:2 p = 0.0007. HCC1937 + Br1 clone 2: 2:1 vs. 1:1 p = 0.0032; 2:1 vs. 1:2 p = 0.0008. Data are represented as the mean ± SD. (D) Killing ability of Br1 tumor-reactive T cell clone 1 (green) against MDA-MB-231 tumor cells or clone 2 (red) against HCC1937 tumor cells by flow cytometry analysis (n = 1). (E) CD107a and TNF-⍺ expression on Br23 tumor-reactive T cell clone 3 in three independent experiments (n = 3) and clone 4 (n = 1) cultured in the presence of different breast cancer cell lines. Data are represented as the mean ± SEM. Kruskal-Wallis test with Dunn’s multiple-comparison tests. Br23 clone 3: MCF-7 vs. unstimulated p = 0.0074; MDA-MB-231 vs. unstimulated p = 0.0316. (F) Killing ability of Br23 tumor-reactive T cell clone 3 (blue) and 4 (gray) against cell line MCF-7 (top), MDA-MB-231 (middle), and SUM159PT (bottom). Br23 clone 3: n = 3; Br23 clone 4: n = 2. Data are represented as the mean ± SD. (G) CD107a and TNF-⍺ expression on Br1 tumor-reactive T cell lines cultured in the presence of overlapping peptide pools for 10 different tumor-associated antigens. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05.

Figure 3

The circulating tumor-reactive CD8 T cell population consists of a diverse, but mostly private, repertoire (A) Number of unique CDR3 beta sequences in the tumor-reactive blood CD8 T cell lines. Each dot represents a different patient. Patients were divided by histological subtypes (ductal n = 7; lobular n = 6; mucinous n = 2; p = 0.091; Kruskal-Wallis test) or lymph node metastasis (yes n = 9; no n = 5; p = 0.7972; Mann-Whitney test). (B) Diversity of the tumor-reactive TRBs calculated by the Shannon index and separated by patient’s histological subtypes (ductal n = 7; lobular n = 6; mucinous n = 2; p = 0.7062, Kruskal-Wallis test) or lymph node metastasis (yes n = 9; no n = 5; p = 0.0829; Mann-Whitney test). (C) V-J rearrangement Circos plots of the T cell receptor beta chain (TRB) in tumor-reactive T cell lines derived from each patient. Each plot represents the distribution of V (red outer arc) and J (blue outer arc) gene segment usage, with the connecting lines indicating specific V-J gene rearrangements for each TCR clonotype. The width of connecting lines reflects the relative abundance of that specific clonotype, with the color corresponding to the used TRBV segment. In parentheses next to each patient ID is the subtype of the tumor (L, lobular; D, ductal; M, mucinous). Samples are grouped based on the presence or absence of lymph node metastasis. (D) Heatmap showing the number of shared CDR3 alpha (left) and beta (right) chains among the tumor-reactive CD8 T cells from different patients. (E) TRB cluster analysis of tumor-reactive CD8 T cells using the GLIPH2 algorithm. Each color represents a different patient, and the size correlates with the abundance of the clone in the T cell line. The link highlights TCRs that are similar based on global alignment of the CDR3β.

Figure 4

Tumor-reactive T cells can infiltrate the tissue, but patients with lymph node metastasis have fewer tumor-reactive TILs (A) Representative CD8 immunohistochemistry (IHC) staining in a patient with (tumor-reactive-positive) and without (tumor-reactive-negative) a detected blood tumor-reactive CD8 T cell response. (B) Percentage of the total tumor area with CD8 staining in ductal breast cancer patients with (n = 6) and without (n = 6) a detected blood tumor-reactive T cell response (p = 0.0801; two-tailed unpaired t test). (C) Ratio of CD8 T cells present in the tumor compared to the paratumor in breast cancer patients with (n = 14) and without (n = 6) a detected blood tumor-reactive T cell response (p = 0.0200; Mann-Whitney test). (D) Schematic showing that tumor-reactive TCRs were identified in the tumor by comparing it to blood tumor-reactive TCRs and finding the CDR3 fully matched sequences. (E) Graph showing the percentage of TRA sequences that fully matched tumor-reactive TRAs, in patients with (n = 5) and without (n = 10) lymph node metastasis (p = 0.0416; Mann-Whitney test). (F) Graph showing the percentage of TRB sequences that fully matched tumor-reactive TRBs, in patients with (n = 5) and without (n = 10) lymph node metastasis (p = 0.0077; Mann-Whitney test). (G) Graph showing the percentage of TRB sequences that fully matched pathogen-specific TRBs in the VDJdb, in patients with (n = 5) and without (n = 10) lymph node metastasis (p = 0.5201; Mann-Whitney test). ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05.

Figure 5

CD8 TILs from lymph node-positive HR+ breast cancer patients show reduced neoantigen-specific transcriptional signature in two independent scRNA-seq datasets (A) Schematic of the analysis of the scRNA-seq datasets. The CD8 T cells from HR+ breast cancer patients were selected and submitted to a neoantigen-specific gene module analysis. (B) Neoantigen-specific score of each individual CD8 T cell from patients with and without lymph node analysis. Dataset 1: yes n = 1,995; no n = 2,001; p < 2.22e−16; dataset 2: yes n = 6,944; no n = 11,660; p = 6.1e−14. (C) Percentage of CD8 T cells that occupy the cluster with the highest neoantigen specificity score in patients with and without lymph node metastasis. Dataset 1: cluster LAG3; yes n = 7; no n = 4; p = 0.0421; dataset 2: cluster 3; yes n = 5; no n = 7; p = 0.2658. Two-tailed unpaired t test. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05.