ISSCR 2022

Quality assurance of hematopoietic stem cells by macrophages determines adult stem cell clonality

Sam Wattrus

Tissue stem cells persist for a lifetime and differentiate to maintain homeostasis. Despite their importance, there are no described quality assurance mechanisms for newly formed stem cells. We followed hematopoietic stem/progenitor cell (HSPC) development in zebrafish embryos and observed intimate and specific interactions between macrophages and HSPCs. We found that 70% of nascent HSPCs were contacted for 5 - 45 minutes by macrophages in the niche. Macrophage-HSPC interactions either lead to removal of cytoplasmic material and stem cell division or complete HSPC engulfment. 81% of all HSPC divisions occurred within 30 minutes of macrophage interaction. To assess if interactions affected hematopoietic stem cell (HSC) clone number, we depleted embryonic macrophages in a brainbow color barcoding system. Macrophage depletion with either the irf8 morpholino or clodronate liposomes significantly reduced the number of HSC clones (14 vs 24.6 clones (p = 0.0002)). To identify signals involved in interactions we pursued few-cell proteomics of macrophages marked by uptake of fluorescent HSPC material. This revealed enrichment of three calreticulin paralogs (calr, calr3a, and calr3b). Calreticulin is typically an ER-resident chaperone protein but can also be displayed on the cell surface as an “eat me” signal. Antibody staining confirmed surface Calreticulin on HSPCs, and single cell RNA-seq identified macrophage expression of lrp1ab and c1qa, the canonical surface Calreticulin binding partners. Morpholino knock down of calr3a or calr3b reduced macrophage interactions as much as 2-fold (p = 0.0008) and overexpression of non-ER bound calr3a or calr3b increased interactions up to 6-fold (p = 0.0001). Knock down of calr3a or calr3b in brainbow embryos also reduced HSC clone number (15.6 vs 19.6 clones (p < 0.0001)). Single cell RNA-seq of irf8 morphants identified a population of HSPCs marked by FOXO activity and a loss of ERK/MAPK target genes, suggesting that reactive oxygen species (ROS) high cells accumulated in the absence of macrophages. ROS levels correlated with Calreticulin staining by flow cytometry (p < 0.0001) and treatment with an ROS inhibitor reduced macrophage-HSPC interactions (p < 0.05). Our work supports a model in which embryonic macrophages determine HSC clonality by monitoring stem cell quality.