HEB Restrains Effector Gene Expression during Early CD8+ Memory Precursor T Cell Differentiation

Abstract

Memory T cells are essential for maintaining long-term adaptive immunity. Memory cell precursors and short-lived effector cells emerge from undifferentiated naïve T cells directly downstream of TCR signaling but little is known about how this lineage choice is regulated at the molecular level. The transcription factor HEB is known to be an important regulator of thymic T cell development, but how it functions in peripheral T cell differentiation is poorly understood. We assessed the role of HEB in the differentiation of memory-like T cell precursors by inducing TCR signaling in CD8 T cells in the context of memory-polarizing cytokines or inflammatory conditions and found that CD8 T cells from HEB-deficient mice underwent accelerated differentiation as compared to WT cells. Transcriptomic analysis revealed aberrant upregulation of immune response genes and decreased expression of genes promoting stemness from the earliest stages of post-TCR signal activation and persisting throughout the course of differentiation. In addition, acute viral infection of HEB cKO mice resulted in enhanced memory precursor cell formation and increased effector functionality. Therefore, we have identified HEB as a central participant in the gene regulatory networks that regulate early CD8 memory T cell differentiation and effector gene expression. This study showed that naïve CD8 T cells lacking HEB exhibit increased TCR signal strength and loss of signatures of stem-ness, revealing a role for HEB in promoting immune memory.

Keywords: CD8 T cells; E proteins; T cell differentiation; gene regulation; memory precursor T cell.

Figure 1. HEB deficiency augments TCR-dependent CD8⁺ T cell differentiation in vitro

(A) Schematic overview of in vitro experimental setup. Splenocytes from WT and HEB cKO mice were enriched for naïve CD8⁺ T cells using magnetic-activated cell sorting (MACS) prior to stimulation in culture with CD3/CD28 microbeads and cytokines (IL-7, IL-15). Microbeads were added for 48 hours and then removed, whereas IL-7 and IL-15 were present throughout the culture period. (B) Representative flow cytometry plots of CD8 T cells from anti-CD3/CD28 stimulated cultures depicting the proportions of naïve/T_SCM-like cells (CD62L⁺CD44⁻), T_CM-like (CD62L⁺CD44⁺), and T_EM-like (CD62L⁻CD44⁺) cells out of all CD8⁺ T cells at days 0, 2, and 3 of culture. (C) Representative flow cytometry plots of depicting the proportions of T_SCM-like cells out of all CD62L⁺CD44⁻ cells (CD122⁺CXCR3⁺) at days 0, 2, and 3 of culture. (D) Quantification of the percentages of naïve/T_SCM-like, CM, EM, and SCM cells within the indicated populations. (E) Absolute numbers of naïve/T_SCM-like, T_CM-like, T_EM-like, and T_SCM-like cells at days 0, 2, and 3 of culture. Data are representative of two independent experiments with three or more mice per group. All data are depicted as mean ± SEM. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05 as determined by unpaired Student’s t test. Salmon bars = WT, aqua bars = HEB cKO.

Figure 2. HEB cKO cells exhibit rapid differentiation prior to proliferation

Splenocytes from WT and HEB cKO mice were enriched for naïve CD8⁺ T cells using magnetic-activated cell sorting (MACS) prior to stimulation in culture with CD3/CD28 microbeads and cytokines (IL-7, IL-15). Microbeads were added for 48 hours and then removed, whereas IL-7 and IL-15 were present throughout the culture period. At days 2 and 3 of culture, cells were analyzed by flow cytometry. (A, B) Percentages of all CD8+ T cells that have divided as assessed by dilution of CFDA in WT and HEB cKO cells in culture over time, with representative histograms (A) and quantification (B). (C) Proliferation (CFDA) and differentiation of T_SCM-like cells (CXCR3) at specified days of culture. (B) Quantification of the percentages of cells T_SCM-like cells that have differentiated (CXCR3⁺) but not yet divided (CDFA⁻). Data are representative of two independent experiments with three mice per group. All data are depicted as mean ± SEM. **** p < 0.0001, *** p < 0.001, ** p < 0.01, as determined by unpaired Student’s t test. Salmon bars = WT, aqua bars = HEB cKO.

Figure 3. RNA sequencing confirms the in vitro derived CD8 T cell subset identities and reveals the impact of HEB deficiency on the transcriptional landscape

In vitro-derived CD8 T cell subsets of naïve, T_SCM-like (SCM), T_CM-like (CM), and T_EM-like (EM) cells were sorted from day 3 cultures and subjected to bulk RNA sequencing in duplicate. (A) Plot displaying principal component analysis (PCA) comparing CD8 T cell subsets from WT and HEB cKO cultures along the PC1 and PC4 axes. Cell type and genotype (status) are depicted on the plot and in the legend (right). (B) Log-normalized counts of naïve- (top) and effector-associated (bottom) T cell genes in WT CD8 T cell subsets, as indicated in the legend (right). (C) Heatmap depicting relative expression of subset-associated T cell genes in WT and HEB cKO CD8 T cell subsets as indicated below plot and in legend (right). Data are representative of four independent experiments with three or more mice per group, pooled into two replicates per cell type. KO = HEBcKO.

Figure 4. Lymphocyte activation genes are elevated in culture-derived T_SCM from HEB cKO mice relative to WT

In vitro-derived CD8 T cell subsets of naïve, T_SCM-like (SCM), T_CM-like (CM), and T_EM-like (EM) cells sorted from day 3 cultures and subjected to bulk RNA sequencing were analyzed for differential gene expression. (A) Volcano plots showing differentially expressed genes (DEGs) in WT versus HEB cKO samples for each CD8 T cell subset (naïve, SCM, CM, EM) as indicated on the plot. Red dots represent significant genes with FDR and adjusted p-value <0.05, log2foldchange <−0.5 or >0.5. (B) Heatmap illustrating expression of DEGs in WT and HEB cKO samples by cell type and genotype, as denoted by legend (left). Genes were filtered on adjusted p-value <0.01, log2foldchange>0.58 (1.5x) or <−0.58 (−1.5x). Rows were scaled and clustered prior to gene ontology analysis. (C) Heatmap showing CD8 T cell subset expression of specific genes involved in Type I IFN signaling (GO:0035456; GO:0060337) in WT and HEB cKO mice, shown in duplicate. (D) Heatmap showing expression of specific genes involved in IFNγ production and lymphocyte activation (GO:0032609; GO:0046649) in WT and HEB cKO CD8 T cell subsets. Data are representative of four independent experiments with three or more mice per group, pooled into two replicates per cell type. Significance was determined by the Wald test in the DESeq2 package. KO = HEBcKO.

Figure 5. HEB deficiency results in aberrant expression of immune-associated transcripts

In vitro-derived CD8 T cell subsets of naïve, T_SCM-like (SCM), T_CM-like (CM), and T_EM-like (EM) cells sorted from day 3 cultures and subjected to bulk RNA sequencing. (A) Log-normalized counts of select effector- and naïve-associated transcripts in culture-derived WT and HEB cKO CD8 T cell subsets according to cell type. For each gene, naïve T cell values are depicted in upper left corner, T_SCM in upper right, T_CM in lower left, and T_EM in lower right. WT values are in salmon and HEB cKO values (KO) are in aqua, as indicated in legend (right). (B) Hallmark (left) and Biocarta (right) pathway analysis using fgsea illustrating up- and downregulated pathways in HEB cKO T_SCM cells relative to WT T_SCM cells. Red bars indicate significantly enriched pathways (adjusted p-value <0.05). Blue dots next to the pathway names associated with T cell receptor signalling and activation indicate significantly higher representation in HEB cKO cells. (C) Heatmaps showing expression of genes involved in NFAT (left) and PI3K/MTOR/AKT (right) signalling that are significantly enriched in HEB cKO (KO) T_SCM cells relative to WT T_SCM cells. Data are representative of four independent experiments with three or more mice per group, pooled into two replicates per cell type. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05 as determined by the Wald test in the DESeq2 package to determine differential expression. KO = HEB cKO.

Figure 6. HEB cKO mice show enhanced MPEC formation and enhanced cytokine responses during acute LCMV infection

(A) Schematic overview of in vivo experimental setup for LCMV infection and analysis. WT and HEB cKO mice were injected intraperitoneally with LCMV Armstrong and analyzed by flow cytometry eight days post-infection. (B) Representative flow cytometry plots (left) and quantification (right) of the percentages of LCMV gp33-tetramer (Tet)⁺ CD8 cells in the spleens of WT and HEB cKO mice 8 days post-infection. C) Representative flow cytometry plots of MPEC (IL7R⁺KLRG⁻) and SLEC (IL7R⁻KLRG⁺) cells (left) and quantification (right) of MPEC cells within the splenic CD8⁺Tet⁺ population. (D) Representative flow cytometry plots (left) and quantification (right) of T_SCM (CD62L⁺Sca1⁺) cells within the splenic CD8⁺Tet⁺CD44⁻ population. (E, F) Cells were harvested 8 days post-infection, re-stimulated with PMA/ionomycin, and analyzed for co-expression of IFNγ and CD107a (E), and co-expression of IFNγ and TNF (F), in the CD8⁺Thy1⁺ populations of WT and HEB cKO mice. (G) Mice were injected with BrdU 7 hours prior to anti-BrdU staining to analyze proliferation. Representative histogram (left) and quantification (right) are shown for BrdU incorporation into CD8⁺Tet⁺ cells. H. Representative histogram (left) and quantification (right) of Annexin V staining to measure apoptosis in CD8⁺Tet⁺ cells. Each dot represents an individual mouse. All data in graphs are depicted as mean ± SEM. ** p < 0.01, * p < 0.05 as determined by unpaired Student’s t test. Salmon bars = WT, aqua bars = HEB cKO.

Figure 7. High dimensionality reduction analysis reveals clusters of effector and memory CD8 T cell subsets revealing an increase in memory formation in LCMV-infected HEB cKO mice.

WT and HEB cKO mice were injected intraperitoneally with LCMV and analyzed by flow cytometry eight days post-infection with a 15-color panel of antibodies (CD8, gp33-tetramer, CD27, CD44, CD62L, CD127, CXCR3, CXCR6, CX3CR1, Gzma, Gzmb, KLRG1, Ly6C, Sca-1, TCF-1) plus Viability dye (16 colors total). Flow cytometry files were concatenated by genotype, and the CD8⁺Tet⁺ populations from WT and HEB cKO cells were merged and subjected to high dimensionality reduction using FlowSOM. (A, B) tSNE plots showing clustering profiles of merged data for total (WT + HEBcKO) cells (A), and for cells from WT (left) or HEB cKO (right) mice (B). (C, D). Cluster identities were assigned using a heatmap (C) and a set of feature plots (D) showing the distribution and intensity of each marker in clusters containing all (WT + HEB cKO) cells. Marker combinations used to assign identities and supporting references are shown in Table 1. (E) Frequencies of cluster-assigned cells out of all cells within each genotype (WT vs HEB cKO). TDEC = terminally differentiated effector cells, EECs = early effector cells, LLEC = long lived effector cells, Pre-T_RM = pre-resident memory T cells, Pre-T_RM/T_EM = precursors with both resident memory T cell and effector memory T cell potential, Pre-T_CM = pre central memory T cells, Pre-mem = pre-memory cells. Each dot represents an individual mouse. All data in graphs are depicted as mean ± SEM. ** p < 0.01, * p < 0.05 as determined by unpaired Student’s t test, ns = non-significant. Salmon bars = WT, aqua bars = HEB cKO.

Figure 8. HEB cKO CD8 T cells exhibit enhanced response to TCR signaling.

The impact of HEB absence on events downstream of TCR signaling was assessed by measuring the upregulation of CD69, CD44, PD-1, and CD25 in an in vitro TCR stimulation assay. A) Naive (CD62⁺CD44⁻) CD8 T cells were sorted from WT or HEB cKO splenocytes and cultured in plates coated with anti-CD3 and anti-CD28 and supplemented with IL-2. Cells were stained with CSFE and harvested for analysis after 3 days. B) Cells undergoing the earliest stages of TCR-mediated activation were examined by gating on CSFE⁻CD44⁻ (Gen 0) and CFSE⁻CD44⁺ cells (Gen 0.5), and additional gates were applied to obtain cells after 1, 2, 3, and 4 cell divisions (Generations 1–4). C) CD69 geometric means in each CSFE generation for WT (blue) and HEB cKO (red) cells. D) Quantification and significance of CD69 geometric mean differences between WT and HEB cKO cells in Gen 0, 0.5, and 1. Data shown in (D) shows replicates and significance for data shown as a line graph in (C). E, F. Representative flow plots (E) and quantification (F) of cells co-expressing CD69 and CD25 in Gen 0 (left) and Gen 0.5 (right) cells from WT (top) or HEB cKO (bottom) mice. G, H. Representative flow plots (G) and quantification (H) of PD-1 expression in Gen 0 (left) and Gen 0.5 (right) cells from WT (top) or HEB cKO (bottom) mice. Note that cells increased in size, as indicated by an increase in the forward scatter profile, as they gained expression of PD1 in Gen 0 of HEB cKO but not WT cells. Gen = generation. All data in bar graphs are depicted as mean ± SEM. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05 as determined by unpaired Student’s t test. Salmon bars = WT, aqua bars = HEB cKO.