Martin Solders

Maintenance of tissue-specific immunity is important for immunological fitness, but its establishment has been difficult to assess in humans. Here, we investigated reconstitution of the human uterine immune system by studying women who underwent uterus solid organ transplantation or hematopoietic stem cell transplantation (HSCT). Through single-cell identification based on single-nucleotide polymorphisms and disparate human leukocyte antigen expression using single-cell RNA sequencing or high-parameter flow cytometry, donor versus recipient cell origin was determined, and we examined the gene expression states, surface marker profiles, and spatial organization of these cells in the endometrium. Endometrial immune cell reconstitution occurred after both uterus transplant and HSCT at the transcriptomic, phenotypic, and spatial levels. Reconstitution involved restoration of all major immune lineages with frequencies comparable to those of healthy controls, with preservation of canonical endometrial immune architecture. Recipient-derived immune cells replaced donor immune cells after uterus transplant, whereas HSCT resulted in near-complete donor-derived immune reconstitution, including formation of tissue-resident lymphocytes. This occurred despite tacrolimus-induced, calcineurin-mediated inhibition of the nuclear factor of activated T cells (NFAT) pathway, which affected de novo induction of tissue-residency features in vitro. In one patient, immune cells of male origin reconstituted the endometrium after HSCT. Collectively, these data provide insights into tissue immune system persistence and reconstitution capabilities in the uterus, an organ that undergoes continuous regeneration.

Allogeneic haematopoietic stem cell transplantation (HSCT) is the only curative therapy for patients diagnosed with acute myeloid leukaemia (AML) and myelodysplastic syndrome (MDS). γδ+ T‐cells, a rare subpopulation of T‐cells with both innate and adaptive anticancer functions, have been understudied and the role of different γδ+ T‐cell subsets after HSCT is unclear. Objectives: We aimed to characterise the γδ+ T‐cells reconstitution post‐HSCT, investigating their association with graft‐versus‐host disease (GvHD) treatment and occurrence, and cytomegalovirus (CMV) reactivation. Methods: Peripheral blood samples from AML and MDS patients and their respective donors were immunophenotyped by flow cytometry before and at several time points after HSCT. Additionally, next‐generation sequencing of the TRG locus was performed to assess clonal dynamics and repertoire diversity. Results: Early after HSCT, γδ+ T‐cells showed a similar phenotype compared to pre‐HSCT, and stable repertoires were observed up to 6 months post‐HSCT. Patients with acute GVHD presented a higher frequency of CD8 in γδ+ T‐cells, and γδ+ T‐cell and subsets frequencies were not affected by anti‐thymocyte globulin prophylaxis. Donor‐derived Vδ2+ T‐cell central memory phenotype was associated with a reduced risk of CMV reactivation in the recipient and with repertoire disturbance. Effector memory cells displaying a specific HLA‐DR+ CD86+ phenotype were associated with CMV reactivation and a higher proportion of hyperexpanded clonotypes prior to transplantation. Conclusion: Overall, Vδ1+ and Vδ2+ T‐cell phenotypic profiles and γδ TCR repertoire remain stable post‐HSCT. This stability is disrupted by CMV reactivation, which is potentially associated with γδ+ T‐cells of donor origin and γδ TCR repertoire.

Objectives: The clinical outcome after allogeneic haematopoietic stem cell transplantation (aHCT) relies greatly on the efficient recovery of T cells. Several studies have investigated the short‐term γδ T cell reconstitution and their role in clinical outcomes following haematopoietic stem cell transplantation. Nevertheless, their long‐term characteristics and impact have remained largely unknown.Methods: We analysed γδ T cells from 20 recipient/donor pairs at phenotypic, clonotypic and functional levels to assess their reconstitution ≥ 8 years (median 18 years) post‐transplantation using high‐parameter flow cytometry and next‐generation sequencing of the TCR γ‐chain.Results: γδ T cells displayed comparable phenotypic characteristics between recipients and matching donors. The Vδ2⁺ subset showed a more activated phenotype and cytokine production, while the Vδ1⁺ and non‐Vδ2 T cells maintained long‐term CMV control. TCR γ‐chain composition in long‐term survivors was largely restored, with no significant differences in gene segment usage or diversity. A small cohort of recipients with severe chronic graft‐versus‐host disease (GVHD) showed overrepresented donor‐derived private clonotypes. Furthermore, we also found elevated HLA‐DR⁺Vδ1⁺ T cells in recipients with severe chronic GVHD.Conclusion: Overall, γδ T cells reconstitute with a normalised repertoire, high functional capacity and sustained CMV control ability. An increased proportion of activated Vδ1⁺ T cells correlates with chronic GVHD severity, indicating a potential therapeutic target.

Proportion of immune cells in pancreatic tissues. The proportion of different immune subsets were studied by flow cytometry in central tissues (CT; blue), peripheral tissues (PT; purple), non-tumor tissues (NTT; brown), and patients’ PBMCs (black). A, Tumor and reactive stroma area calculated by QuPath in the three tissue compartments (n = 9). B, Total numbers of CD4+ and CD8+ T cells per mg of tissue (n = 40). C, Frequency of CD4+ and CD8+ T cells from CD45+ cells (CT, PT n = 35, NTT n = 25, blood n = 35). D, Memory phenotype of CD4+ and CD8+ T cells (CT n = 15, PT n = 14, NTT n = 6, blood n = 24). E, Proportion of CD69 on CD4+ and CD8+ T cells (CT n = 12, PT n = 11, NTT n = 5, blood n = 12). F, Proportion of ICOS on CD4+ and CD8+ T cells (CT n = 15, PT n = 14, NTT n = 5, blood n = 15). G, Proportion of granzyme B, perforin, and granzyme B-perforin coexpression on CD4+ and CD8+ T cells from central and peripheral tumor tissues, non-tumor tissues and patients’ PBMCs (n = 8). H, Proportion of MAIT cells from CD3+ T cells (CT n = 21, PT n = 20, NTT n = 13, blood n = 21). I, Proportion of CD19+ B cells from CD45+ cells (CT n = 32, PT n = 32, NTT n = 26, blood n = 32). A–C, F–I, Friedman test followed by Dunn test was used to evaluate significant difference between groups (CT, PT, NTT; F–I) and patients’ PBMCs if n > 5. *, P <0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

Effects of presurgical chemotherapy on the immune landscape. The immune landscape for patients that underwent upfront surgery (gray squares) was compared with the patients that underwent neoadjuvant chemotherapy followed by surgery (white squares). A, Kaplan–Meier survival curves comparing the two groups. B, Proportion of CD39+ CD103+ CD8+ T cells. C, Left, Proportion of CD4+ and CD8+ T cells out of CD45+ cells. Right, Ratio of cells in central tissues relative to peripheral and non-tumor tissues for CD4+ and CD8+ T cells out of CD45+ cells. Proportion of CD4+ T cells (D) and CD8+ T cells (E) expressing PD-1, TIM-3, CCR5, CXCR4, CXCR5, and CXCR3. F, Left, Proportion of CD19+ cells out of CD45+ cells. Right, Proportion of CD19 cells in central tissues relative to peripheral and non-tumor tissues. Mann–Whitney U test was used to detect statistically significant differences. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

Coinhibitory receptor expression on CD4+ and CD8+ tumor-infiltrating T cells. Proportion of CD4+ (A) and CD8+ (B) T cells expressing PD-1 (CT n = 54, PT n = 52, NTT n = 42), TIM-3 (CT n = 53, PT n = 51, NTT n = 41), LAG-3 (NTT n = 6, CT n = 16, PT n = 16), CTLA-4 (CT n = 9, PT n = 8,NTT n = 2), and coexpressing PD-1 and TIM-3 coinhibitory markers (CT n = 53, PT n = 51, NTT n = 41; C, left), flow cytometry representative dot plots on CD8+ T cells (right) to identify coexpression of PD-1 and TIM-3. D, Proportion of PD-1, TIM-3, PD-1TIM-3, LAG-3, and CTLA-4 in CD4+ and CD8+ T cells from patients’ PBMCs. E, Boolean gating analysis was performed on TIM-3, PD-1, and LAG-3 (n = 16). Pies (left) show median values on central tumor tissues (CT), peripheral tumor tissues (PT), and patients’ PBMCs (blood). F, Proportion of CD4+ and CD8+ T cells expressing 0, 1, 2, or 3 coinhibitory markers in central and peripheral tumor tissues. A–C, Friedman test followed by Dunn test was used to evaluate significant difference between groups (CT, PT, NTT) if n > 5. E, Wilcoxon matched pairs signed-rank test was used to detect statistically significant differences between CT and PT tissues. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

Overall survival curves for DP CD8 T cells and SPCD39, DP CD4 T cells.

Tumor-infiltrating CD39+CD103+ T cells. A, A t-SNE analysis showing cluster point plots (1–8) and sample distribution in central (n = 5) and peripheral (n = 5) tumor tissues on 10-parameters flow cytometry analyses. B, Distribution of each parameter as a single cell level. The color shows the median fluorescence expression (MFI) of each marker. Green circles highlight cluster 3 representing CD4+ CD39+ T cells. Pink circles highlight cluster 8 representing CD8+ CD39+ CD103+ T cells. C, Heat map showing the distribution of each parameter for every cluster (1–8). The color shows the MFI of each marker (red: high MFI; blue: low MFI). D, Left, Pie charts showing the median values of CD39 SP, DP, CD103 SP, and DN CD4+ and CD8+ T cells in central tumor tissues (CT) n = 39, peripheral tumor tissues (PT) n = 38, non-tumor tissues (NTT) n = 34 and patient PBMCs (blood; n = 38; right). Flow cytometry dot plots showing the gating strategy to identify CD39+ CD103− (CD39 SP), CD39+ CD103+ (DP), CD39− CD103+ (CD103 SP), and CD39− CD103− (DN) on CD8+ T cells. Proportion of CD4+ T cells SP CD39 (CT n = 39, PF n = 38, NTT n = 34; E) and DP CD8+ T cells (F) in CT n = 39 and PF n = 38, NTT n = 34. G, Kaplan–Meier survival curves were performed on patients with high and low frequency of DP CD8+ T cells out of CD8+ T cells in central, peripheral, and non-tumor tissues. log-rank test was performed to detect statistical significance. H, Forest plots representing survival HRs of the clinicopathologic parameters and DP CD8+ T cells in central pancreatic tumor tissues defined in multivariate Cox proportional hazards regression analysis. E and F, Friedman test followed by Dunn test was used to evaluate significant difference between groups. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant. T2, Tumor >2 cm; T3, Tumor > 4 cm; T4, Tumor involves coeliac axis, superior mesenteric artery and/or common hepatic artery; N0, No regional lymph node metastasis; N1, Metastases in one to three regional lymph nodes; N2, Metastases in four or more regional lymph nodes; M0, No distant metastasis; M1, Distant metastasis.

In pancreatic ductal adenocarcinoma, the infiltration of CD8+ T cells within the tumor microenvironment correlates with a favorable prognosis. However, a significant proportion of tumor-infiltrating T cells become trapped within the desmoplastic stroma and lack tumor reactivity. Here, we explored different T-cell subsets in pancreatic tumors and adjacent tissues. We identified a subset of CD8+ T cells, double positive (DP) for CD39 and CD103 in pancreatic tumors, which has recently been described to display tumor reactivity in other types of solid tumors. Interestingly, DP CD8+ T cells preferentially accumulated in central tumor tissues compared with paired peripheral tumor and adjacent non-tumor tissues. Consistent with an antigen encounter, DP CD8+ T cells demonstrated higher proliferative rates and displayed an exhausted phenotype, characterized by elevated expression of PD-1 and TIM-3, compared with CD39−CD103− CD8+ T cells. In addition, DP CD8+ T cells exhibited higher expression levels of the tissue trafficking receptors CCR5 and CXCR6, while displaying lower levels of CXCR3 and CXCR4. Importantly, a high proportion of DP CD8+ T cells is associated with increased patient survival. These findings suggest that DP CD8+ T cells with a phenotype reminiscent of that of tumor-reactive T cells are present in pancreatic tumors. The abundance of DP CD8+ T cells could potentially aid in selecting patients for pancreatic cancer immunotherapy trials.Significance:Patients with pancreatic cancer with a high proportion of CD39+CD103+ CD8+ T cells exhibiting a tumor-reactive phenotype have improved survival rates, suggesting their potential utility in selecting candidates for immunotherapy trials.

Effects of presurgical chemotherapy on the immune landscape. The immune landscape for patients that underwent upfront surgery (gray squares) was compared with the patients that underwent neoadjuvant chemotherapy followed by surgery (white squares). A, Kaplan–Meier survival curves comparing the two groups. B, Proportion of CD39+ CD103+ CD8+ T cells. C, Left, Proportion of CD4+ and CD8+ T cells out of CD45+ cells. Right, Ratio of cells in central tissues relative to peripheral and non-tumor tissues for CD4+ and CD8+ T cells out of CD45+ cells. Proportion of CD4+ T cells (D) and CD8+ T cells (E) expressing PD-1, TIM-3, CCR5, CXCR4, CXCR5, and CXCR3. F, Left, Proportion of CD19+ cells out of CD45+ cells. Right, Proportion of CD19 cells in central tissues relative to peripheral and non-tumor tissues. Mann–Whitney U test was used to detect statistically significant differences. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.