Bone marrow microenvironment in autoimmune hemolytic anemia: from trephine biopsy to single cell RNA sequencing

Abstract

The role of bone marrow (BM) compensatory response in autoimmune hemolytic anemias (AIHAs) is emerging and inadequate reticulocytosis has been associated with more severe disease and adverse outcomes. However, few is known about the BM immunologic microenvironment composition in these diseases. Here we investigated BM features in a large cohort of 97 patients with autoimmune hemolytic anemia (AIHA) and observed a high prevalence of hypercellularity, dyserythropoiesis, reticulin fibrosis, and T-cell infiltration (65%, 29%, 76%, and 69% of patients, respectively). These findings were associated with inadequate bone marrow compensation, more severe anemia at onset, and need of multiple treatments. In a subset of warm type AIHA patients we investigated BM microenvironment by single-cell RNA sequencing. We found distinct immune cell profiles across disease stages (diagnosis, remission, relapse). In particular, upregulation of inflammatory response pathways was noted in CD8 + , CD4 + , and monocyte subsets during relapse compared to diagnosis and remission. Moreover, by single-cell TCR sequencing, we found small T cell clones at diagnosis that may either disappeared or expanded at remission. Disappearing clones exhibited a naive CD8+ phenotype and were more likely to respond to glucocorticoid treatment. Expanding clones showed upregulation of cytotoxic T cell markers and may play a role in the transition to a chronic/relapsing phase. Finally, cytokine gene expression differed across disease phases. At relapse, pro-inflammatory cytokines such as TNF-alpha, IL-1, and IL-6 were upregulated in CD4+ and CD8 + T cells, while TGF-beta was downregulated, potentially in an attempt to counteract the transition to chronic phase. This is the largest study evaluating BM histology and clinical characteristics, and the first evaluation of BM microenvironment by single-cell RNA sequencing in AIHA. We showed a complex scenario encompassing T-cell infiltration, clonality, and up/down-regulation of cytokine genes, associated with a more severe and relapsing disease.

Fig. 1

Single-cell RNA sequencing landscape of AIHA patients. a Schematic workflow for scRNA-seq of AIHA and healthy donors samples. The panel was created by Biorender.com. b Donut charts representing the distribution of cells in AIHA patients according to clinical stratification (Red: diagnosis; Orange: relapses/remission; blue: remission). Numerical values are reported in the legend for category. c Uniform Manifold Approximation and Projection (UMAP) representation of cells (N = 54,830) from the 9 AIHA single-cell RNA-seq data. Clusters of cells are colored by clinical stratification. Color legend as in (b). d Dot plot displaying the top ten marker genes that distinguish the clinical stratification. The X-axis lists the clinical category, while the Y lists gene names. Circle size corresponds to the number of cells in the category expressing the gene of interest, while shade correlates with the level of expression. e Cartoon depicting the deregulated expression profiles of AIHA patients, obtained from 1:1 DE analyses by Wilcoxon test. Color code as in Fig. 1b. Original plot results represented in Figure Supplementary Fig. 3

Fig. 2

The bone marrow-associated cell repertoire of AIHA patients. a Single-cell uniform manifold approximation and projection (UMAP) plot of the AIHA patients. Cells are color-coded by computationally determined cell clusters. b Donut charts representing the distribution of immune cells (T CD8 + , T CD4+ and monocytes) in AIHA patients according to clinical stratification (Red: diagnosis; Orange: relapses/remission; blue: remission). Numerical values are reported in the legend for category. c Dot plot displaying the top ten marker genes that distinguish the clinical stratification, across immune cell repertoire (T CD8 + , T CD 4 + and monocytes). The X-axis lists the clinical category, while the Y lists gene names. Circle size corresponds to the number of cells in the category expressing the gene of interest, while shade correlates with the level of expression. d Cartoon depicting the deregulated expression profiles of AIHA patients, obtained from 1:1 DE analyses by Wilcoxon test, across immune cell repertoire (T CD8 + , T CD 4 + and monocytes). Color code as in Fig. 1b. Original plot results represented in Figure Supplementary Fig. 5

Fig. 3

Transcriptional phenotype of CD8+ cell repertoire in bone marrow aspirates of AIHA patients. a Distribution of the CD8 T clone size among small, medium and large cell clones in bone marrow aspirates from three matched patients at diagnosis and remission. Colors represent group size from small to large size. b Alluvial plots showing the distribution of T cell clones dynamics from diagnosis to remission samples, according to the T cell clones. c Dot plot displaying the top ten marker genes of CD8 + T cells clones that distinguish the diagnosis and remission. The X-axis lists the clinical category, while the Y lists gene names. Circle size corresponds to the number of cells in the category expressing the gene of interest, while shade correlates with the level of expression. d Dot plots showing the Wilcoxon enriched hallmarks of CD8 + T cells clones of AIHA patients at diagnosis and remission. Gene-ratio on the y-axis, counts as dot size and color represent padj values in z-score