The pancreatic tumor microenvironment of treatment-naïve patients causes a functional shift in γδ T cells, impairing their anti-tumoral defense

Abstract

Pancreatic ductal adenocarcinoma (PDAC) presents a unique challenge for researchers due to its late diagnosis caused by vague symptoms and lack of early detection markers. Additionally, PDAC is characterized by an immunosuppressive microenvironment (TME), making it a difficult tumor to treat. While γδ T cells have shown potential for anti-tumor activity, conflicting studies exist regarding their effectiveness in pancreatic cancer. This study aims to explore the hypothesis that the PDAC TME hinders the anti-tumor capabilities of γδ T cells through blockade of cytotoxic functions. For this reason, we chose to enroll PDAC treatment-naive patients to avoid the possibility of therapy modifying the TME. By flow cytometry, our research findings indicate that the presence of γδ T cells among CD45+ cells in tumor tissue is lower compared to CD66+ cells, but higher than in blood. Circulating Vδ1 T cells exhibit a terminal effector memory phenotype (TEMRA) more than Vδ2 T cells. Interestingly, Vδ1 and Vδ2 T cells appear to be more prevalent at different stages of tumor development. In our in vitro culture using conditioned medium derived from Patient-derived organoids ;(PDOs), we observed a shift in expression markers in γδ T cells of healthy individuals toward an activation and exhaustion phenotype, as confirmed by scRNA-seq analysis extracted from a public database. A deeper understanding of γδ T cells in PDAC could be valuable for developing novel therapies aimed at mitigating the impact of the pancreatic tumor microenvironment on this cell population.

Keywords: Fine-needle biopsies; PDAC; gamma delta T cells; immune checkpoints; immunotherapy; patient derived organoids.

Figure 1.

Infiltrating and circulating γδ T cells show heterogeneous frequencies in PDAC patients compared with non-PDAC patients. a) representative dot plots of PDAC-infiltrating γδ T cells and circulating Vδ1 and Vδ2 T cells from the same patients, and the relative gating strategies. b) Cumulative graph comparing individual values of infiltrating leukocytes (CD45+ cells) from PDAC patients and, in particular, CD66b+ cells and γδ TCR+ cells. The analysis of CD66b cells was obtained using the gating strategy reported as follows: SSC-A/FSC-A dot plot> single cells> live cells> CD45+ cells/SSC-A> CD66b+/FSC-A. c) representative violin plots comparing infiltrating (black contour line) and circulating (red contour line) γδ T cells and their subsets from PDAC patients and non-PDAC patients (gray violin). Black circles indicate γδ TCR+ T cells, black squares indicate Vδ1 T cells, black triangles indicate Vδ2 T cells. d) Representative paraffin section of human PDAC patients were stained using a mAb specific for TCRγ/δ. Pictures were acquired with 20X and 40X magnification, the latter was used to evaluate staining specificity. In 20X magnification scale bar represents 100 μm. 11 PDAC patients and 4 Non-PDAC patients analyzed. Statistical differences were assessed using unpaired student’s t-test with Mann-Whitney test, and data are expressed as median (continuous line) and quartiles (dashed lines).

Figure 2.

Immunocharacterization of γδ T cells from tissue and blood of PDAC patients compared to non-PDAC patients show distinctive profiles. a) representative violin plots of the phenotypes of circulating Vδ1 and Vδ2 T cells (red contour line) from PDAC patients (5 samples). Black squares indicate Vδ1 T cells, black triangles indicate Vδ2 T cells. Statistical differences were assessed using one-way ANOVA, Tukey’s multiple comparisons test and data are expressed as median (continuous line) and quartiles (dashed lines). b) representative pseudocolor-plots and counter plots of PDAC-circulating Vδ1 and Vδ2 T cells from the same patients and the relative gating strategies. c) representative violin plots of TIGIT+ or TIM3+ infiltrating γδ T cells (black contour line) and comparing PDAC patients and non-PDAC patients (gray violin) (8 samples PDAC patients and 4 Non-PDAC patients analyzed). d) representative violin plots of TIGIT+ or TIM3+ circulating Vδ1 and Vδ2 T cells (red contour line) and comparing PDAC patients and non-PDAC patients (gray violin) (8 samples PDAC patients and 4 Non-PDAC patients analyzed). e) violin plot comparing NKp46+ γδ T cells from tissue (black contour line) and blood (red contour line) from PDAC patients (6 samples PDAC patients and 4 Non-PDAC patients analyzed). Statistical differences were assessed using unpaired student’s t-test with Mann-Whitney test, and data are expressed as median (continuous line) and quartiles (dashed lines).

Figure 3.

Deep analysis of the expression markers in different clinical tumor stages reveals the early involvement of γδ T cells subsets. a) representative histogram showing frequency of infiltrating γδ T cells and b) the circulating subsets of Vδ1 and Vδ2 T cells at different tumor stages (from T2 to T4) and c) and d) of infiltrating lymph nodes (N) classified by TNM score, obtained by flow cytometry analysis. In b) and d) the red contour line indicates the analysis of circulating subsets of γδ T cells related to the clinical stages. Statistical differences were assessed using unpaired Kruskal-Wallis test, data shown are median and SEM. e) graphical representation of the analysis shown in F-H in which the immune characterization evaluating specific markers expression is correlated with different tumor stages detected by EUS-FNBs and confirmed by pathological anatomy report; f) cumulative box-plot floating bar with minimum and maximum and median as the middle line of TIGIT+ or TIM3+ γδ T cells from tumor tissue at different tumor stages (black contour line). g) and h) show cumulative data for circulating Vδ1 and Vδ2 (red contour line) with minimum and maximum and median as the middle line. The figures represent the cumulative data of 11 different samples distributed across the various tumor stages. Statistical differences were assessed by using unpaired student’s t-test with Mann-Whitney test.

Figure 4.

In vitro assay demonstrated that PDAC-TME could be selectively responsible for the activation of γδ T cell subsets and their exhaustion. a) graph illustrating the scheme of the experimental procedure of in vitro assays that describes the use of PDAC PDO (from 3 PDAC patients at late stages of tumor) from which we obtained CM after 4 days in culture, and blood from five healthy donors. The experiments were performed putting γδ T cells in culture 1 on at 37°C and 5% CO₂ with RPMI and in CM, as indicated in material and methods section. b) histogram plot shows median of percentage of TIM3, LAG3 and PD1 expressing γδ T cells in culture with RPMI (white histogram) and in CM (gray histogram). c) histogram plot shows median of percentage of CD45RA and CD27 expressing γδ T cells in culture with RPMI (white histogram) and in CM (gray histogram) identifying phenotype. d) histogram plot shows median of percentage of CD25 and CD69 expressing γδ T cells in culture with RPMI (white histogram) and in CM (gray histogram). Statistical differences were assessed using unpaired student’s t-test with Mann-Whitney test.

Figure 5.

Subpopulations identified in infiltrating γδ T cells by scRNA transcriptomics. a) UMAP clustering visualization of γδ T cells subpopulations in the analysis, b) in Normal and c) Tumor tissue. d) Stacked barplot showing the proportion of γδ T cells subclusters identified in the analysis, e) in Normal and f) in Tumor tissue.

Figure 6.

Functional analysis of γδ T cells subpopulations. GSEA using the Hallmark gene sets in GD2_b vs GD2_a cells (a), and in each cluster within tumor vs normal tissue (b). Tiles are colored according to the Normalized Enrichment Score (NES). c) Expression level of selected genes in each cluster vs all the others within the same condition. Tiles are colored according to the Average logFoldchange (Avg log2FC) and stars were used to indicate the statistical significance (p <0.001 (***), < 0.01 (**), < 0.05(*)).

Figure 7.

Visualization pseudotemporal trajectory analysis. Pseudotime trajectory analysis and sample density plot of γδ T cells calculated starting from GD1 (a-d) and GD2_a (e,f) clusters. A color gradient from black to yellow represents different pseudotime levels, with black indicating the earliest time point and yellow indicating the latest time point (a,c,e Density plots represent the distribution of cells in normal and tumor conditions along the pseudotime (b, d, f). In the transition from normal to tumor microenvironment, GD1 cells are distributed along the trajectory to GD2_b cells (a), as do GD2_a (c). Additionally, GD1 and GD2_a are distributed along the same trajectory in a bidirectional manner (b).