glucose assay HSC functions are severely compromised followi

HSC functions are severely compromised following acute and chronic TLR4 activation (Esplin et al., 2011; Rodriguez et al., 2009). We demonstrated that TRIF signaling is a critical mediator of HSC injury, since its deletion fully restores HSC functions following LPS exposure. Transplantation of LPS LSK glucose assay containing equal amounts of LT-HSC in WT, MYD88−/−, and TRIF−/− mice revealed significant differences in long-term engraftment of these genotypes. WT LT-HSC from mice challenged with LPS were significantly exhausted at 24 weeks after transplant and showed low ability to self-renew and contribute to all lineages in competitive repopulation assays. In contrast, LPS TRIF−/− LT-HSC maintained their self-renewal and multilineage contribution, similarly to PBS controls. Loss of MYD88 provided some protection to HSC and progenitors, as LPS MYD88−/− LSK cells exhibited better engraftment than the LPS WT LSK cells, but did not fully rescue their functions. Interestingly, enrichment in MPP and contaminating progenitors in the LSK pool did not confer any advantage in short-term engraftment of WT and MYD88−/− LSK donor cells. Collectively, our results show that HSC are qualitatively damaged by TRIF activation during LPS exposure and that this damage persists over the long term, and also in the presence of a healthy microenvironment. Strikingly, TRIF activation, unlike MYD88, did not directly cause myelosuppression. This cell-context-dependent effect does not seem to be mediated by differential expression of MYD88 or TRIF in the different subsets. Similar to the cell-context-dependent impact of other molecules (i.e., Notch), the distinct effects of TRIF and MYD88 are likely due to different requirements and molecular interactions of these two pathways in HSC and in myeloid progenitors. Potential mechanisms downstream of TRIF in HSC may involve STAT1, AKT, p38/ROS (Katsoulidis et al., 2005), or pathways activated by type I IFNs, which are known to damage HSC by induction of cell-cycle entry, apoptosis, and DNA damage (Baldridge et al., 2010; Pietras et al., 2014). However, mild induction of IFNs by LPS and lack of a protective effect in Ifnra1 mice following LPS suggest that this mechanism is not dominant in an LPS response. Analysis of expression of Spi1 and CebpA, two key transcription factors involved in the regulation of HSC and myeloid transcriptional programs (Hasemann et al., 2014; Staber et al., 2013), revealed that they were significantly downregulated in WT HSPC following LPS treatment, while their levels were preserved in the absence of MYD88 (Spi1) or TRIF (CebpA). Spi1 and CebpA expression levels continued to remain low in HSCs at 24 weeks following transplant, whereas they were restored in MYD88−/− and TRIF−/− HSPC. Thus, our study shows that acute exposure of HSPC to LPS endotoxemia is sufficient to induce permanent transcriptional changes even when the endotoxic environment is removed and cells are transplanted in healthy donors. Changes in histone methylation patterns in mature cells have been reported following sepsis (Carson et al., 2011), and we found changes in epigenetic modifications in HSPC after in vivo exposure to LPS (H.Z. and N.C., unpublished observations). Further studies are necessary to address the impact of epigenetic regulation of HSC on their function and differentiation during sepsis. In conclusion, the present study shows that TRIF and MYD88 uncouple effects of TLR4 signaling on HSC and myeloid progenitors during severe bacterial infection. This observation has potential clinical relevance. Despite the central role for TLR4 in sepsis, a recent randomized clinical trial failed to show survival improvement in patients with sepsis treated with a TLR4 antagonist (Opal et al., 2013). It is likely that complete inhibition of TLR4 signaling, with simultaneous inhibition of the MYD88 and TRIF pathway, also abrogates its protective effects, thus compromising the effectiveness of this therapeutic approach. The results presented here provide a guide to further dissect the independent effects of MYD88 and TRIF during response to severe bacterial infection and support the rationale for pursuing time-tuned selective silencing of one pathway (MYD88 or TRIF) to mitigate myelosuppression or stem cell injury. Finally, these observations may also provide insight into the impact of MYD88 and TRIF in HSC and myeloid progenitors during chronic inflammation related to aging and hematopoietic cell malignant transformation.