Decoding the Immune Landscape of Pathogenic T Cells in Acute Graft-Versus-Host Disease By Single-Cell RNA-Seq

Introduction

Graft-versus-host disease (GvHD) is the major cause of non-relapse mortality in allogeneic hematopoietic cell transplantation (alloHCT), and about 40-60% cases would progress into steroid refractory (SR) aGvHD resulting in dismal outcomes. The pathophysiology of aGvHD is complex, and donor derived T cells are identified as the culprits of aGvHD. However, the relationships among distinct T cell subsets and molecular mechanisms underlying aGvHD remain incompletely understood.

Methods

We employed a single-cell RNA sequencing (scRNA-seq) and T cell receptor (TCR) sequencing (scTCR-seq) to profile 49288 T cells in peripheral blood isolated from alloHCT recipients developed into aGvHD (n=9) or not (as control, n=5). Then, samples from SR-aGvHD patients were further investigated to explore the molecular mechanism of pathogenic T cell subsets in SR-aGvHD.

Results

A total of thirteen unique T cell subsets with distinct molecular properties were identified, including four clusters of CD8⁺ T cells (CD8⁺ TN, CD8⁺ TCM, CD8⁺ TEM and CD8⁺ TEFF), five clusters of CD4⁺ T cells (CD4⁺ TN, CD4⁺ TCM, Th1, Th17 and Treg), NK cell cluster and three clusters of other T cells (IFN-T, γδT and MAIT). The composition shifted towards more-developed CD8⁺ TCM, CD8⁺ TEM, CD4⁺ TCM and Th1 cells in aGvHD group due to rapid cell cycling. The expansion of CD8⁺ TEM subpopulation was confirmed by multiparameter flow cytometry. The CD8⁺ TEFF and Th1 were considered to be pathogenic in aGvHD as they exhibited aGvHD-related cytokine/granzyme expression and signaling pathway activation. Upregulated genes in aGvHD-derived-Th1 cells were related to chemotaxis (CCL3, CXCR6), cytokine production (type I interferon, IL-23) and JAK-STAT signaling activation. CD8⁺ TEFF cells showed activity of chemotaxis (CCL3, CXCR6), granzyme and cytokine expression (GZMB, GZMA, TNF-α, IL-6, IL-10, IL-12) and MAPK/MTOR/NF-kappaB activation. The different composition of cytokine expression and activated signaling pathways suggested distinct functions of each subset. Furthermore, CD8⁺ TEM represented four traits related to a range of T-cell differentiation stages, spanning rapid cell cycling (transcription factor E2F family), chemotaxis (CCL3, CXCR6), cytotoxicity (GZMA, GZMB, and GZMK) and exhaustion (PDCD1, LAG3, HAVCR2 and TIMD4), suggesting CD8⁺ TEM as a transient stage which might further develop into CD8⁺ TEFF for aGvHD pathogenesis (Figure 1).

To elucidate the molecular mechanism of SR-aGvHD, we dissected the composition and functional heterogeneity of T-cell subsets between SR- and steroid sensitive (SS)-aGvHD. CD8⁺ TEM expanded to be the largest subset which resulted in the increase of CD8⁺ TEFF proportion in SR-aGvHD. CD8⁺ TEM and CD8⁺ TEFF showed enrichment of GvHD pathogenic scores and up-regulated expression of inflammation transcription factors STAT1 and STAT3 in SR-aGvHD. Of note, the expression of steroid receptor NR3C1 was downregulated in CD8⁺ TEM subset, as well as the steroid target gene ANXA1, compared with counterparts. Furthermore, both CD8⁺ TEM and CD8⁺ TEFF cells exhibited enrichment of JAK-STAT, MAPK, NF-kappaB, and PI3K-AKT signaling activation (Figure 2).

Conclusions

The effector T-cell subsets CD8⁺ TEFF and Th1 were co-activated and undertook distinct expression profiles and signaling pathways, which suggests their different roles in aGvHD pathogenesis. CD8⁺ TEM, characterized by proliferative, cytotoxicity, chemotaxis and exhaustion capacities, provided functional repertoire for CD8⁺ TEFF. Reduced NR3C1 expression and activation of by-pass survival pathways of CD8⁺ TEM might provide persistent differentiation origin of CD8⁺ TEFF to mediate resistance to steroid therapy. Collectively, our data described the immune landscape of aGvHD and provided novel therapeutic targets.

No relevant conflicts of interest to declare.