Functional memory T cells are derived from exhausted clones and expanded by checkpoint blockade [TCR-seq]

Bulk TCR-seq Data Associated with the manuscript: Functional memory T cells are derived from exhausted clones and expanded by checkpoint blockade

Manuscript Abstract: 

Immune checkpoint blockade can facilitate tumor clearance by T cells, resulting in long term patient survival. However, the capacity of exhausted CD8+ T cells (Tex), present during chronic antigen exposure, to form memory after antigen clearance remains unclear. Here, we performed longitudinal single cell RNA/T cell receptor sequencing and ATAC-sequencing on antigen-specific T cells after the peripheral clearance of chronic lymphocytic choriomeningitis virus infection. These data revealed the formation of a robust population of memory CD8+ T cells that transcriptionally, epigenetically, and functionally resemble central memory T cells (Tcm) that form after clearance of acute infection. To lineage trace the origin and memory recall response of Tex-derived memory clones, we utilized T cell receptor sequencing over the course of primary infection and rechallenge. We show that chronic Tcm are a clonally distinct lineage of Tex derived from progenitor exhausted cells, persist long-term in the absence of antigen, and undergo rapid clonal expansion during rechallenge. Finally, we demonstrate that αPD-L1 immune checkpoint blockade selectively expands clones which form Tcm after clearance. Together, these data support the concept that chronically stimulated T cells form bona fide functional memory T cells through an analogous differentiation pathway to acutely stimulated T cells, which may have significant implications for enhancing immune memory to cancer through checkpoint blockade and vaccination.

Bulk TCR-seq data are from four (4) datasets associated with the manuscript:

1. TCR-seq of GP33+ CD8+ T cell subsets from LCMV-Cl13 infection 100+ dpi
2. TCR-seq of PD1+ Tex subsets from the blood and spleen 28 dpi with LCMV-Cl13
3. TCR-seq of CD8+ T cell subsets from LCMV-Cl13 or LCMV-Arm infection 100+ dpi prior to transfer of cells and LCMV-Arm rechallenge 
4. Longitudinal TCR-seq of CD8+ T cell subsets from LCMV-Cl13 infection 21, 35, and 100+ dpi +/- αPD-L1 treatment

Methods Details

(1) TCR-seq of GP33+ CD8+ T cell subsets from LCMV-Cl13 infection 100+ dpi

OVERVIEW: TCR sequencing was performed on sorted populations of GP33+ T cell subsets at 126 days post-infection. GP33+ cells were also sorted from blood of the same at 28 days post-infection.

DETAILS: Peripheral blood was collected at 28 days post-infection by retro-orbital bleed, and spleen  was collected at 126 days post-infection.  CD8+ T cells were enriched by magnetic separation, and peripheral blood was treated with chloride-potassium bicarbonate for red blood cell lysis. Cells were stained with viability dye, GP33 tetramer, and surface antibodies. Live CD8+ GP33+ T cells were sorted based on surface protein expression using a BD FACSAria into RPMI supplemented with 10% FBS. The cells were centrifuged, the supernatant was removed, and dry pellets were snap-frozen. Genomic DNA extraction was performed by Adaptive Biotech (Seattle, WA).  TCR-seq was performed by Adaptive via their immunoseq platform at the survey depth. Adaptive performed intial data alignment and processing. 

(2) TCR-seq of PD1+ Tex subsets from the blood and spleen 28 dpi with LCMV-Cl13

OVERVIEW: CD8+ T cells populations from the spleen and blood of the same mouse were sorted 28 days post-infection with LCMV-Cl13.

DETAILS: Mice were infected with LMCV-Arm or LCMV-Cl13 and sacraficed at 28 dpi. Spleen and peripheral blood (collected via cardiac puncture) were collected and enriched for CD8+ T cells using magnetic separation. Cells were stained with viability dye, GP33 tetramer, and surface antibodies. Live CD8+ GP33+ T cells were sorted based on surface marker expression using a BD FACSAria into RPMI supplemented with 10% FBS. The cells were centrifuged, the supernatant was removed, and dry pellets were snap-frozen. Genomic DNA extraction was performed by Adaptive Biotech (Seattle, WA).  TCR-seq was performed by Adaptive via their immunoseq platform at the survey depth. Adaptive performed intial data alignment and processing. 

(3) TCR-seq of CD8+ T cell subsets from LCMV-Cl13 or LCMV-Arm infection 100+ dpi prior to transfer of cells and LCMV-Arm rechallenge 

OVERVIEW: To track T cell clones during rechallenge, CD8+ T cells were collected from the spleens of LCMV-Cl13 immune or LCMV-Arm immune mice. Half of the cells were transferred into new host mice and challenged with LCMV-Arm. The remaining pre-rechallenge cells were subset-sorted, and bulk TCR sequencing was performed. These data are associated with scRNA/TCR-seq of the same T cell clones after rechallenge (GEO: GSE285414) 

DETAILS: Splenocytes were harvested at 151 days post-infection from mice infected with LCMV-Cl13 or LCMV-Arm, and CD8+ T cells were enriched using magnetic separation. Approximately half of the cells were transferred for rechallenge, while the remaining cells underwent TCR sequencing. CD8+ T cells that were not transferred were stained with viability dye, GP33 tetramer, and surface antibodies. Live CD8+ T cell subsets were sorted based on a combination of surface antigens using a BD FACSAria into DNA/RNA Shield (Zymo) and snap-frozen. RNA extraction was performed by iRepertoire (Huntsville, AL). iRepertoire performed TCR beta chain sequencing via Amp-RepSeq+1 Platform (Mouse TCR beta, long read). iRepertoire performed intial data alignment and processing. Reads were UMI normalized.

(4) Longitudinal TCR-seq of CD8+ T cell subsets from LCMV-Cl13 infection 21, 35, and 100+ dpi +/- αPD-L1 treatment

OVERVIEW: T cells were isolated from blood at 21 and 35 days post-infection and from the spleen of the same mice 100 days post-infection. T cell subsets were sorted at each timepoint, and bulk TCR sequencing  was performed on the sorted cells. Mice were treated six times with either aPDL1 or PBS from days 22 to 34.

DETAILS: Mice were infected with LCMV-Cl13 and dosed with 200ug aPDL1 (B7-H1) or PBS by IP injected every 2-3 days from 22-34 dpi. Peripheral blood was collected by retro-orbital bleed at 21 and 35 days post-infection, and spleen  was collected at 100 days post-infection. Splenocytes were enriched for CD8+ T cells using magnetic separation, and peripheral blood was treated with chloride-potassium bicarbonate for red blood cell lysis. Cells were then stained with viability dye, GP33 tetramer, and surface antibodies. Live CD8+ T cells were sorted based on surface protein expression using a BD FACSAria and directly into Zymo RNA lysis buffer. The cells were snap-frozen and stored at -80°C. RNA extraction was performed by iRepertoire (Huntsville, AL). iRepertoire performed TCR beta chain sequencing via Amp-RepSeq+1 Platform (Mouse TCR beta, long read). iRepertoire performed intial data alignment and processing. Reads were UMI normalized.

Samples: 

Note: We often sorted >4 cell populations at once, so we could not sort all cell subsets at all times, due to this we report the total number of cells in a subset (calculated based on absolute count beads) to normalize TCR-seq counts back to absolute frequency of a given clone.

Also availble for download in file: `TCR-seq Samples.xlsx`