TLR1/2-stimulated DCs “prime” HSCs via IL-1β

Bone marrow dendritic cells (BMDCs) are increasingly recognized as an important cellular component of the hematopoietic stem cell (HSC) perivascular niche. In this issue of Blood, Li et al¹ report that BMDCs respond to Toll-like receptor 1/2 (TLR1/2) agonist stimulation by producing interleukin 1β (IL-1β), which is sensed by HSCs, leading to their expansion.

Mature hematopoietic cell production during both homeostasis and stress relies on a fine-tuned process of environmental sensing and intracellular signaling by HSCs within their microenvironment, also referred to as the “HSC niche.” Integration of cellular extrinsic signals such as inflammatory or pathogen-associated molecular pattern molecules (PAMPs) into cellular intrinsic molecular circuits determines whether HSCs remain quiescent or engage in multilineage differentiation and/or self-renewal divisions to generate the blood cells in demand, while at the same time maintaining a sufficient HSC pool size.² HSCs can be activated via direct PAMP sensing by TLRs.³ In addition, PAMPs are detected by multiple BM cell types, leading to inflammatory cytokine production that ultimately determines HSC function.⁴ However, which cells, and possibly which specific perivascular niche cells are responsible for PAMP sensing and which inflammatory signals function as messengers to fine-tune HSC and hematopoietic stem and progenitor cell responses, is still not clear. In this issue of Blood, Li et al identify BMDCs as major cellular sensors of TLR1/2 agonists and show that this sensing results in increased BMDC IL-1β production, signaling HSCs to transiently expand their pool (see figure).

To identify the mouse BM myeloid phagocyte populations with the highest TLR expression levels, the authors preformed RNA-sequencing on DCs (CX3CR1^high MHCII^high CD11c^high Gr-1^– B220^–), macrophages (CX3CR1^– MHCII^high CD169⁺ Gr-1^– B220^–), and monocytes (CX3CR1^high CD115⁺ Gr-1⁺ B220^–) and observed that DCs exhibit the highest levels of Tlr1/2 and inflammatory Il1b expression. In vivo exposure to a TLR1/2 agonist (PAM3CSK4) resulted in an overall decrease of BM myeloid phagocytic populations, decreased osteoblast/mesenchymal stromal cell/sinusoidal endothelial cell numbers, and increased mobilization and expansion of phenotypically defined HSCs due to higher HSC cycling activity. Importantly, using a “conventional” DC-specific deletion of the TLR pathway member Myd88 (Zbtb46-Cre, Myd88f/f mice), Li et al show that only HSC expansion is directly dependent on DC-specific TLR sensing. This is a major finding, with 2 key implications: (1) it excludes major cell-autonomous effects of the TLR1/2 agonist in BM HSC proliferation, challenging the notion that HSCs are, at least partially, direct sensors of this specific type of PAMP⁵; and (2) it links HSC expansion downstream of TLR1/2 to a perivascular niche cell type, which had previously been shown for the sensing of microbiota-derived circulating bacterial DNA by BM CX3CR1⁺MHCII^hi mononuclear cells during steady-state hematopoiesis.⁴ Importantly, whether DCs can also sense steady-state levels of microbiota-derived TLR1/2 agonists and not just hyper-physiological doses of PAM3CSK4 remains to be determined.

To address the mechanism by which DCs contribute to PAM3CSK4-induced HSC expansion, Li et al performed RNA-sequencing on sorted BMDCs 24 hours after treatment with PAM3CSK4. The authors identified Il1b upregulation as a potential candidate, thereby linking DCs to HSC expansion. Indeed, IL-1β chronic exposure was shown to cause HSCs and multipotent progenitor cell expansion and enhanced HSC myeloid differentiation, while impairing HSC reconstitution capacity.⁶ To test if IL-1β was the molecular driver of HSC proliferation, the authors exposed IL-1R1^−/− mice to PAM3CSK4 and observed reduced HSC expansion compared with wild-type (WT) animals, strongly suggesting that TLR1/2-induced HSC and progenitor cell expansion is dependent on IL-1 signaling. Of note, there was still a minor but significant expansion of LSK (Lin^– Sca-1⁺ C-Kit⁺) and multipotent progenitor 3 (MPP3) in IL-1R1^−/− mice after PAM3CSK4 exposure, suggesting that additional inflammatory mediators may be involved. Moreover, as expected from chronic IL-1β exposure studies, while PAM3CSK4-treated WT HSCs lost repopulation capacity, their IL-1R1^−/− counterparts were able to maintain repopulation capacity upon secondary transplantation.

Finally, the authors cultured sorted murine WT HSCs with conditioned media from PAM3CSK4-stimulated BMDCs and observed that IL-1β, present in the conditioned media, was indeed able to increase the Lin^– C-Kit⁺ population cell number. As further confirmation of this finding, exposure of human DCs, but not monocytes, to PAM3CSK4 resulted in increased IL-1β levels. This finding suggests that PAM3CSK4 treatment induces an expansion of hematopoietic progenitors, at least in part, by increasing IL-1β expression in BMDCs. Overall, this study establishes BMDC-mediated IL-1β production downstream of TLR1/2 activation as a key molecular bridge between PAMP-sensing and HSC responses. These findings align with and extend other recent data, in which we observed an age-dependent increase in circulating microbiota-derived TLR4/8 agonists. These agonists are sensed by BM inflammatory monocytes and granulocytes, causing increased IL-1 production that ultimately leads to HSC pool expansion and an increased myeloid differentiation bias.⁷ Integration of these findings highlights IL-1 as a major inflammatory cytokine induced downstream of TLR signaling within different BM myeloid cell types influencing HSC proliferation and myelopoiesis.

Moreover, the findings by Li et al implicate TLR1/2 and IL-1 signaling in the development of hematologic dysfunction. Indeed, the authors show increased TLR1 and IL-1β expression in BMDCs, CD14⁺ monocytes, and CD16⁺ monocytes in a small number of human myelodysplastic syndrome (MDS) samples. Although this finding needs validation in larger cohorts, it aligns with previous reports on increased TLR1/2 signaling in MDS hematopoietic progenitors.⁸ Moreover, altered IL-1 signaling has also been implicated in MDS pathology and development of acute myeloid leukemia (AML). Expression of IL-1 accessory protein (eg, IL1RAP, a IL-1R1 coreceptor) is increased in high-risk MDS and AML blasts and associates with poor overall survival in patients with AML.⁹ It is therefore tempting to speculate that in pre-leukemic states such as MDS (and potentially also in clonal hematopoiesis), enhanced TLR1/2 sensing can lead to higher IL-1 production; this might favor the selection of mutant HSCs over WT counterpart cells, leading to expansion of inflammation-adapted mutant clones and further accelerating progression to full leukemia. The findings by Li et al further highlight the need to understand whether leukemia-founding mutations might not only be present in BMDCs but also might alter and possibly enhance their inflammatory responses, as observed for DNMT3a and TET2 mutant myeloid cells.¹⁰ 

Overall, the findings by Li et al identify BMDCs as active players of the HSC niche with a critical role in PAMP sensing and IL-1–dependent regulation of HSC expansion. Future studies will need to unravel the precise contribution of multiple PAMP-sensing cell types as well as the molecular pathways operating in various infectious/inflammatory contexts and their overall implications for host defense responses, hematologic aging, and disease development.

Conflict-of-interest disclosure: The authors declare no competing financial interests.