In our quest to reveal fundamental functions of hematopoietic stem cells (HSCs) in vivo, we first devised an HSC-specific genetic switch which could turn on a fluorescent marker in HSCs without the need for isolation or in vitro manipulation of HSCs. While transplant-based experiments led to the long-standing view that few HSCs generate many blood and immune cells frequently, our experiments revealed a very different situation: We found, that many HSCs contribute to hematopoiesis, but that individual HSCs are rarely active and produce only a small output under normal steady state conditions. We could also derive differentiation rates emerging from HSCs to progenitors of the lymphoid and myeloid lineages. These experiments were all based on the propagation of a single-color label, a fluorescent protein. While this method reveals important information about differentiation rates of HSCs as outlined above, it cannot uncover the relationships of HSC products, or show whether individual HSCs give rise to different cell types (HSC fate types).
To overcome this principal limitation of conventional fate mapping, we developed a new, highly versatile cellular barcoding system, termed Polylox. In combination with specific or ubiquitous genetic switches (Cre recombinase mouse lines), Polylox can be used to introduce several hundred thousand different inheritable barcodes into the cells' genomic DNA, again without cell isolation or in vitro manipulation. Importantly, the use of Cre recombinase renders this recombination system inducible, which is key the for the time- and tissue-specific generation of the barcodes.
Using Polylox barcoding, we obtained the following key results:
1. The hematopoietic system shows a basic structure with separate lymphoid and myeloid-erythroid branches. This applies to both embryonic and adult hematopoiesis.
2. We identified three major fate types of HSCs: Multilineage HSCs, myeloid-erythroid-restricted HSCs and differentiation-inactive HSCs. There is also a rare subset of HSCs that are restricted to megakaryocytes (blood platelets).
3. These HSC fate patterns are stable from embryonic development throughout adult life.
4. To directly link fates (barcoding) with transcriptome (RNA sequencing), we have developed a new version of Polylox, termed PolyloxExpress. Here, barcodes are expressed as RNA molecules and can therefore be read together with the transcriptome. Experiments using PolyloxExpress revealed the 'position' of single HSC clones with defined fates in the transcriptional landscape of stem and progenitor cells. In this manner, we could derive HSC fate-associated transcriptional signatures.
5. In the course of an infection, there is an increased demand for differentiated effector cells, to which hematopoiesis must respond and adapt. It was commonly assumed that the source of this regeneration is HSCs, as they show elevated proliferation under stress conditions. There was, however, no direct evidence for enhanced HSC output in response to infection. Using fate mapping and cellular barcoding, we have now shown in a clinically relevant polymicrobial sepsis model, that not HSCs but the progenitors downstream of HSCs are rapidly increasing their output to replenish the needed myeloid cells (granulocytes and monocytes).
Collectively, using novel fate mapping and in vivo barcoding technologies, we deconvoluted HSC fates, and uncovered physiological properties of hematopoiesis under normal and under stress conditions.
