Fig 2: Single-cell RNA sequencing (scRNA-seq) delineates transcriptionally distinct states within CD4+ and CD8+ TRM cell populations(A) Schematic of two scRNA-seq experiments. Biopsies of small intestinal transplant tissue were collected from subjects at endoscopy, then dissociated to isolate intestinal lymphocytes. These were sorted by fluorescence-activated cell sorting (FACS), first with bulk sorting of donor-derived T cells in experiment 1, then with index sorting of donor-derived CD103− and CD103+ T cells in experiment 2, before scRNA-seq library preparation and sequencing using the 10× Genomics platform (experiment 1) or the Smart-Seq2 protocol (experiment 2).(B) t-Distributed Stochastic Neighbor Embedding (tSNE) plot of 974 donor-derived T cells from a single subject showing four transcriptionally distinct populations of conventional CD4+ and CD8+ T cells.(C) Dot plot of key gene identifiers for the four clusters showing two clusters of CD8+ T cells and two clusters of CD4+ T cells, including one containing cells expressing IL2RA and FOXP3, consistent with a regulatory T cell phenotype. Dot size indicates the proportion of cells in which the gene is expressed.(D) tSNE plots showing the expression of key genes upregulated in CD8+ population 2 (top row) or upregulated in CD8+ population 1 (bottom row).(E) Heatmap indicating hierarchical clustering of gene expression of CD8+ donor-derived T cells from population 1 and population 2. Cell labels above indicate cluster: red, population 1; green, population 2. Genes of interest are highlighted.(F) Dot plot showing the expression of 13 genes previously associated with tissue residency in human CD69+ T cells (Kumar et al., 2017). Dot size indicates the proportion of cells in which the gene is expressed.

Fig 3: Intrahepatic lymphocytes Are Characterized by High Ex Vivo Autophagy Levels(A) The dimensionality reduction algorithm tSNE was applied to flow cytometry data (single cell expression values from total live CD45+ singlet lymphocytes for: CD3, CD4, CD8a, CD19, CD103, CD69, pan-?d T cell receptor (TCR), pan-aß TCR, CD161, CD56, and LC3) to generate a two-dimensional map of lymphocytes from paired PBMC (left) and IHL (middle) samples from two individuals colored by intensity of LC3 or by lymphocyte subset (right; example gating Figure S1A).(B) Histograms (gated on CD8+; ± bafilomycin A1 [bafA1] treatment, 0.1 µM; FMO for LC3) and summary data for LC3 staining of paired peripheral (PBMCs; black) and intrahepatic (IHLs; red) T cells (23 biological replicates).(C and D) Example of gating, histograms, and summary data for LC3 staining of CD161-, CD161mid, and mucosal-associated invariant T cells (MAITs; CD161hi Va7.2+; 11–14 biological replicates) (C) and CD19+ (B cells) and CD56+ (NK cells) lymphocytes (10 biological replicates) (D).Cells were treated with bafA1 unless otherwise stated (unblocked data in Figure S1) (A, C, and D). Wilcoxon paired t test (B and D). For pairwise multiple comparisons (within PBMC/IHL comparisons) Friedman test (ANOVA) with Dunn’s post hoc test (C). For multiple unpaired comparisons (between PBMC and IHL for a given subset) Kruskal-Wallis (ANOVA) with Dunn’s post hoc test. Bars at mean (B, C, and D). *p < 0.05, **p < 0.005, ****p < 0.0001.

Fig 4: CD103+ and CD103- CD8+ T cells display distinct phenotypes in healthy and transplanted intestine(A) Representative expression of CD161, CD7, CD127 (IL7R), ß2-integrin (ITGB2), and granzyme K on CD103- (red) and CD103+ (black) donor-derived T cells from the intestinal transplant graft.(B and C) Phenotypic analysis of CD8+ T cell populations in healthy small intestine.(B) Representative flow cytometry plots and histograms of CD8+ T cell phenotype from spectral flow cytometry, with expression of the following markers plotted against CD103: CD69, CD161, CD7, CD127, ß2-integrin, granzyme K, KLRG1, and Ki-67. Blue, CD69-CD103- cells; red, CD69+CD103- cells; black, CD69+CD103+ cells.(C) Proportion of positive cells (CD161, KLRG1, and Ki-67) or MFI (CD7, CD127, ß2-integrin, and granzyme K) of CD8+ T cells, categorized by CD69 and CD103 expression, in small intestinal biopsies from healthy control subjects (n = 10). Mean percentage or MFI represented by bars. Connecting lines represent populations from the same subject.(D–I) Phenotypic analysis of recipient-derived CD8+ T cell populations infiltrating the intestinal transplant graft.(D) The proportion of recipient-derived CD8+ T cells co-expressing CD69 and CD103 in intestinal transplant grafts, categorized by time after transplant (n = 35; 16 subjects).(E–I) Proportion of CD161+ cells (E), MFI of CD127 (F), ß2-integrin (G), and granzyme K (H), or proportion of Ki-67+ cells (I) of recipient-derived CD8+ T cells, categorized by CD69 and CD103 expression and time after transplant in intestinal transplant grafts (n = 23; 12 subjects). Mean percentage or MFI represented by bars. Black lines connect populations from the same subject.Statistical analysis performed with one-way ANOVA with Tukey’s multiple-comparison test. *p = 0.05, **p = 0.01, ***p = 0.001, ****p = 0.0001.

Fig 5: Long-lived, conventional CD4+ and CD8+ T cell, but not unconventional T cell, populations can persist for at least 5 years in the human intestine(A) Representative flow cytometry plot of HLA-A2 expression on T cells from the blood, the recipient native intestinal mucosa, and the intestinal transplant graft demonstrating identification of donor- and recipient-derived populations by HLA mismatch.(B) Representative chip cytometry image of intestinal graft mucosa demonstrating the presence of donor-derived (HLA-A3+, yellow arrows) and recipient-derived (HLA-A3-, white arrows) CD3+ T cells in the lamina propria. Cytokeratin (gray); CD3 (purple); HLA-A3 (green).(C) Percentage of recipient-origin CD3+ T cells in intestinal grafts, categorized by time after transplant (n = 37; 16 subjects; means ± SEM).(D) Percentage of recipient-origin CD3+ T cells in intestinal grafts, categorized by history of graft rejection (n = 37; 16 subjects; means ± SEM).(E) Flow cytometry plot of HLA-A3 expression on graft-derived T cells in one subject who demonstrated persistent donor chimerism in the intestinal graft 1,865 days (5 years and 1 month) after transplant.(F) Percentage of donor-origin CD3+ T cells in the blood of intestinal transplant recipients, categorized by time after transplant (n = 13; means ± SEM). Dashed line at 4% represents the cutoff for significant macrochimerism from prior studies (Fu et al., 2019; Zuber et al., 2015).(G) Conventional CD8+ and CD4+ T cell subsets in the small intestinal graft as a proportion of total T cells, categorised by time post-transplant (n = 39; 18 subjects; means ± SEM).(H) The percentage of recipient-derived T cells within conventional CD8+ and CD4+ T cell subsets in the intestinal graft, categorized by time after transplant (n = 37; 16 subjects; means ± SEM).(I) Unconventional non-Vd2+ ?d T cell, Vd2+ ?d T cell, and Va7.2+CD161+ CD8+ T cell (mucosal-associated invariant T cell) subsets in the small intestinal graft as a proportion of total T cells, categorized by time after transplant (n = 39; 18 subjects; means ± SEM).(J) The percentage of recipient-derived T cells infiltrating the intestinal graft within unconventional non-Vd2+ ?d T cell (n = 27; 12 subjects), Vd2+ ?d T cell (n = 25; 12 subjects), and Va7.2+CD161+ CD8+ T cell (n = 20; 12 subjects) subsets, categorized by time after transplant (means ± SEM).(K) Representative flow cytometry plot of CD69 and CD103 expression on donor- and recipient-derived CD8+ and CD4+ T cells in the intestinal graft.(L) Percentage of donor- and recipient-derived CD8+ and CD4+ T cells in the intestinal graft co-expressing CD69 and CD103 categorized by time after transplant (n = 35; 16 subjects; median marked with black line).For further analysis of surface marker expression, rare populations with fewer than 10 cells were excluded from the analysis (J and L). Statistical analysis performed with one-way ANOVA with Dunnett’s multiple-comparison test. *p = 0.05, **p = 0.01.