Using Genetic Cartography to Map Cell Lineage

Hematopoietic stem cells (HSC) are cellular factories that churn out blood and immune cells, but their production output varies significantly. This variation influences the success of bone marrow transplant therapies, which are heavily dependent on the quantity and type of blood cells in donor HSC transplantation.¹ Although gene expression influences cell lineage, the regulatory mechanism remains unclear.

“This is an example where therapeutic usage precedes our understanding of the fundamental mechanism. [HSC] are clinically proved to be useful, so it really drives further research to better understand their regulation,” remarked Rong Lu, a stem cell biologist at the University of Southern California. 

When investigating HSC regulation to improve immune regeneration, Lu and her team identified associations between gene activity that modulate the variety and amount of immune cell production.  Their findings, published in Science Advances, provide insights into how distinct genetic features can improve optimization of donor cell pools for therapies.²

Lu wondered whether differences across individual HSC were intrinsic or influenced by their environment. To test this, the researchers infected HSC with lentivirus carrying unique genetic barcodes to track individual cells. This method enabled Lu to measure the quantity and type of cellular production.³ Lu focused on granulocytes and B cells, which are the most abundant myeloid and lymphoid immune cells, respectively.

In a one-to-multiple transplantation experiment, they injected barcoded HSC clones into a primary mouse recipient. After recovering the same stem cells, the researchers transplanted them into several other mice. Even with radiation or chemo treatment before transplantation, the blood production levels in the secondary mice remained consistent; these results implied that an inherent factor dictates blood producing capacity.

“The key technical challenge is being able to connect the cell and cell production tracking with the gene expression of the corresponding cells,” said Lu. 

To tackle this question, the researchers measured the cellular outputs of the genetically tracked barcoded HSC. After using single-cell RNA sequencing (scRNA-seq), they labeled individual cells with another tag: a cDNA cellular barcode. To link the data from genetic tracking and scRNA-seq, Lu and her team generated “molecular bridges” through polymerase chain reaction to amplify the molecules containing both barcodes. Then they mapped single-cell lineage output to gene expression against a cDNA library to generate a molecular bridge database. 

Lu and her team used the database to quantify HSC lineage outputs across mice, measuring transcription levels of genes associated with blood cell production. They identified nearly 40 genes related to blood cell production with some linked to diseases that influence immune cell production. They found certain genes associated with the production of lymphoid cells and others that boasted a balance of various immune cells. 

To dig further into the associations between gene activity and the quantity and variety of immune cells produced, Lu examined genes across the whole range of a lineage output. Instead of simply categorizing these outputs as low or high, she conducted a more comprehensive analysis to catch genes that might have a sweet spot. 

“The results tell us that there’s some sort of mechanism or quantitative association patterns that we haven’t appreciated in the field, which is how the genes quantitatively control cell behavior,” said Lu.

The researchers found the genes correlated with the amount and type of blood cell production via four models of association: constant, unimodal, discrete, and multimodal. They identified some genes with constant association, where gene expression always associated with lymphocyte production. In discretely associated genes, lymphocyte production fell within a specific range of output levels. Most commonly, genes fell into the unimodal or multimodal category. In unimodal patterns, cell output is dependent on the gene expression state and resembles an on/off switch. However, in multimodal associations, gene activity occurs at multiple levels or thresholds for several distinct quantities of cell production.

“This study uses a clever technology that involves both sophisticated molecular biology and computational approaches. It’s a very novel approach to how we track and assess blood stem cells and their outputs,” remarked David Kent, a stem cell biologist at the University of York who was not involved in the study. “[This method] sets the stage to answer new questions, so I see this as a beneficial tool development paper.”

Through this work, Lu hopes to achieve the next level of precision medicine. “By understanding the difference between individual cells, we can make all the cells do the things that we want. This is really going to take advantage of natural variation across individual cells to manage whole cell populations,” said Lu. 

References

1. Lu R, et al. Clonal-level lineage commitment pathways of hematopoietic stem cells in vivo. Proc Natl Acad Sci U S A. 2019;116(4):1447-1456.
2. Jiang D, et al. Quantitative association between gene expression and blood cell production of individual hematopoietic stem cells in mice. Sci Adv. 2024 Jan;10(4):eadk2132.
3. Lu R, et al. Tracking single hematopoietic stem cells in vivo using high-throughput sequencing in conjunction with viral genetic barcoding. Nat Biotechnol. 2011;29(10):928-933.