Introduction
Autologous stem cell transplant (auto-SCT) is unparalleled in its ability to induce clinical and endoscopic remission for the treatment of refractory Crohn’s disease (CD). Auto-SCT is unique as a cellular therapy aimed to reset immune pathophysiology to a pre-disease state using hematopoietic stem cells (HSC). Despite the efficacy of this therapy, the mechanism of auto-SCT and characterization of the HSCs and the reconstitution of the peripheral and mucosal immune system in refractory CD remains unknown. We used single cell RNA sequencing (scRNAseq) for deep immunophenotyping of the stem cell graft and the peripheral and mucosal immune response to auto-SCT.
Methods
Patients with CD were enrolled in a Phase IIa study of auto-SCT (NCT03219359). 16 patients were transplanted (2018-2023). Paired blood and intestinal samples were taken prior to and 6 months post-SCT and additional blood samples were taken after mobilization and engraftment. Fresh isolated single cells from each were processed for scRNAseq using 10x Genomics 3’ Gene Expression.
Results
At 6 months from auto-SCT, 15/16 patients had an endoscopic response (SES-CD decrease ≥ 50%) and 11/16 were in endoscopic remission (total SES-CD ≤ 4, no sub-score >1). The scRNAseq analysis of the blood after mobilization and engraftment demonstrates the predominant emergence of myeloid progenitors and HSC populations which through RNA velocity analysis with scVelo demonstrates prominent ontologic relationship between HSCs and myeloid progenitors suggestive of a myeloid bias within the stem cell graft. Further analysis of gene enrichment in the HSCs demonstrate substantial changes in growth factor signaling pathways and epigenetic regulators after transplant. In the intestine, CellphoneDB analysis of ligand and receptor interactions demonstrate a strong myeloid-stromal-glial network with increased interactions between epithelial and other myeloid clusters post-SCT.
Conclusion
For the first time, using scRNAseq, we characterized the peripheral blood, intestinal mucosa and the stem cell graft in auto-SCT. We show that auto-SCT may be effective secondary to the HSCs in the graft rather than the mobilization and conditioning regimen alone. There is a predominance of myeloid progenitors and HSCs with myeloid bias in the graft that mirrors the prominent shift of the myeloid compartment in the mucosa, as well as a change in growth factor signaling and epigenetic regulation in the HSCs that may be critical to the efficacy of auto-SCT in restoring mucosal signaling pathways and mucosal healing. These studies reveal potential mechanisms through which SCT heals the intestinal mucosa in a manner unique to this therapeutic modality that is unachieved through targeting inflammatory pathways alone, and there may be epigenetic changes in HSCs that contribute to the manifestation of medically refractory disease.

Figure 1. a. scRNAseq UMAP of blood samples at baseline, stem cell collection and engraftment. b. RNA velocity curves using scVelo of blood at stem cell collection with myeloid progenitor clusters denoted by dashed box and HSC clusters denoted by solid box outline. c. Differentially expressed genes in HSC cluster 24 from baseline to 6 months post-SCT were identified and significantly upregulated genes were analyzed by Enrichr to identify significantly enriched pathways (wikipathways). Volcano plot highlights a significant increase in the hematopoietic gene regulation pathway (genes in pathway listed). d. Analysis of gene regulatory network that underlies significantly increased gene expression for HSC cluster 24 shows EP300 is a key gene regulator.

Figure 2 a. scRNAseq UMAP of all intestinal tissue samples with major call populations annotated. b. CellphoneDB heatmap of scRNAseq intestinal clusters at baseline and 6 months post-SCT with highlighted clusters with notable change after SCT in red denoting epithelial clusters (3, 4, 14) and in black denoting clusters from myeloid (17), stromal (19) and glial (21) clusters.

Sayali Talware

Appearances