Increased PD-1+Foxp3+ γδ T cells associate with poor overall survival for patients with acute myeloid leukemia

Abstract

Problems: γδ T cells are essential for anti-leukemia function in immunotherapy, however, γδ T cells have different functional subsets, including regulatory cell subsets expressing the Foxp3. Whether they are correlated with immune-checkpoint mediated T cell immune dysfunction remains unknown in patients with acute myeloid leukemia (AML).

Methods: In this study, we used RNA-seq data from 167 patients in TCGA dataset to analyze the correlation between PD-1 and FOXP3 genes and these two genes' association with the prognosis of AML patients. The expression proportion of Foxp3⁺/PD-1⁺ cells in γδ T cells and two subgroups Vδ1 and Vδ2 T cells were performed by flow cytometry. The expression level of FOXP3 and PD-1 genes in γδ T cells were sorted from peripheral blood by MACS magnetic cell sorting technique were analyzed by quantitative real-time PCR.

Results: We found that PD-1 gene was positively correlated with FOXP3 gene and highly co-expressed PD-1 and FOXP3 genes were associated with poor overall survival (OS) from TCGA database. Then, we detected a skewed distribution of γδ T cells with increased Vδ1 and decreased Vδ2 T cell subsets in AML. Moreover, significantly higher percentages of PD-1⁺ γδ, Foxp3⁺ γδ, and PD-1⁺Foxp3⁺ γδ T cells were detected in de novo AML patients compared with healthy individuals. More importantly, AML patients containing higher PD-1⁺Foxp3⁺ γδ T cells had lower OS, which might be a potential therapeutic target for leukemia immunotherapy.

Conclusion: A significant increase in the PD-1⁺Foxp3⁺ γδ T cell subset in AML was associated with poor clinical outcome, which provides predictive value for the study of AML patients.

Keywords: Foxp3; PD-1; acute myeloid leukemia; outcome; overall survival; γδ T cells.

Figure 1

The correlation and the prognostic between the expression of PD-1 and FOXP3 genes in AML from the TCGA (The Cancer Genome Atlas). (A) Correlation of the expression levels of PD-1 and FOXP3. (B–D) The optimal cutoff values were based on the median gene expression levels, the FOXP3 and PD-1 genes were divided into high expression (red line) and low expression (blue line) groups, which were plotted in Kaplan-Meier curves (top) with the number at risk AML patients (bottom). Kaplan-Meier curves are shown for single FOXP3 high or low expression, single PD-1 high or low expression, and co-high or co-low expression of PD-1/FOXP3. (B) FOXP3 ^high vs. FOXP3 ^low, 24-month OS: 36% vs. 53% P = 0.368. (C) PD-1 ^high vs. PD-1 ^low, 24-month OS: 35% vs. 54% P = 0.009. (D) PD-1 ^high FOXP3 ^high vs. PD-1 ^low FOXP3 ^low, 24-month OS: 35% vs. 65% P = 0.018.

Figure 2

Distribution and frequency of Foxp3 and PD-1 expression in γδ, Vδ1, and Vδ2 cell subsets in PB from patients with de novo AML. (A, B) γδ T cells were gated by CD3⁺ T cells, further identifying Vδ1 and Vδ2 cell subsets from a healthy donor and a patient with de novo AML by flow cytometry analysis. (C) Comparison of the percentage of γδ, Vδ1, and Vδ2 cell subsets in AML patients compared with HIs. (D) The pie chart representing the distribution of Vδ1 and Vδ2 cells in AML patients and healthy individuals (HIs). (E) Detection of Foxp3 expression in γδ, Vδ1, and Vδ2 cell subsets from a healthy donor and an AML patient. (F) Comparison of the percentage of Foxp3⁺ γδ, Vδ1, and Vδ2 cell subsets in AML patients compared with HIs. (G) Detection of PD-1 expression in γδ, Vδ1, and Vδ2 cell subsets from a healthy donor and an AML patient. (H) Comparison of the percentage of PD-1⁺ γδ, Vδ1, and Vδ2 cell subsets in AML patients compared with HIs. (I) Heatmap representing the frequency of the Foxp3⁺ and PD-1⁺ γδ T subsets in patients with AML (A1-A21) compared with HIs (H1-H15).

Figure 3

Distribution and frequency of co-expression with PD-1 and Foxp3 in γδ, Vδ1, and Vδ2 cell subsets. (A) Detection of co-expression with PD-1 and Foxp3 in γδ, Vδ1, and Vδ2 cell subsets from a patient with de novo AML by flow cytometry analysis. (B) Comparison of the percentage of PD-1⁺Foxp3⁺ in γδ, Vδ1, and Vδ2 cell subsets from de novo AML patients compared with HIs. (C) Heatmap representing the frequency of the PD-1⁺Foxp3⁺ γδ T subset in HIs (H1-H15) and patients with de novo AML (A1-A21).

Figure 4

Correlation and OS analysis of PD-1 and Foxp3 in γδ T cells from de novo AML patients. (A) Correlation of the frequency of the Foxp3⁺, PD-1⁺, and PD-1⁺ Foxp3⁺ γδ T subsets between de novo AML patients and HIs. (B) The expression levels of the FOXP3 and PD-1 genes in γδ T cells from de novo AML patients compared to HIs. (C) Correlation analysis of the FOXP3 and PD-1 genes. (D–F) Correlation of the OS of de novo AML patients stratified by high and low Foxp3⁺ γδ T subsets, high and low PD-1⁺ γδ T subsets and high and low PD-1⁺ Foxp3⁺ γδ T subsets. (D) Foxp3^{+ high} vs. Foxp3^{+ low}, 24-month OS: 36% vs. 50% P = 0.883. (E) PD-1^+high vs. PD-1^+low, 24-month OS: 30% vs. 53% P = 0.179. (F) PD-1⁺ Foxp3^+high vs. PD-1⁺ Foxp3^+low, 24-month OS: 20% vs. 64% P = 0.034.

Figure 5

Overview of alterations in γδ T cell subsets in patients with AML.