merely br Results Our previous work

Results Our previous work showed that severe bacterial sepsis induced by P. aeruginosa or by its LPS causes a TLR4-dependent dysfunctional expansion of HSC and hematopoietic progenitor merely (HSPC) (Rodriguez et al., 2009). To determine whether these changes are MYD88 or TRIF dependent, we performed LPS challenge in mice lacking MYD88 (MYD88−/−) or TRIF (TRIF−/−). Similar to our previous findings, wild-type (WT) mice responded to LPS with a significant increase in the frequency and absolute number of BM Lin−SCA-1+c-KIT+ (LSK) cells, a subset enriched in HSPC (∼4-fold; Figures 1A and 1C). Surprisingly, TRIF−/− mice exhibited no such expansion of LSK cells, whereas MYD88−/− mice behaved in a manner similar to WT mice (Figures 1A and 1C). The LSK pool can be further dissected by immunophenotypic analysis into distinct subsets (Figure 1B) (Kiel et al., 2005; Wilson et al., 2008), consisting of: long-term HSCs (LT-HSC; Lin−SCA-1+c-KIT+CD150+CD48−), which have the ability to self-renew; multipotential progenitor 1 cells (MPP1; Lin−SCA-1+c-KIT+CD150+CD48+), which include short-term HSCs (ST-HSC); and MPP2 cells (Lin−SCA-1+c-KIT+CD150−CD48+). WT and MYD88−/− mice responded to LPS with a noticeable increase in LT-HSC and a robust expansion (5-fold) of MPP1 and MPP2, in both frequency and absolute number (Figures 1D–1F). In contrast, TRIF−/− mice did not show any noticeable increase in LT-HSC, MPP1, or MPP2 subsets. Of note, following LPS challenge the representation of LT-HSC and MPP within the LSK population shifted in similar manner in all genotypes, with increased representation of MPP1 cells at the expense of LT-HSC (Figure 1B); however, as the total LSK population dramatically increased in WT and MYD88−/− mice, the LT-HSC frequency in the Lin− cell pool and their total absolute number were increased in these genotypes. Overall, there was a trend for increased absolute numbers of LT-HSC (LSK CD150+CD48−) in WT mice following LPS, though not statistically significant. Increases in LT-HSC cell numbers were also observed when CD34 and FLK2 were used to define LT-HSC and ST-HSC (Figure S1A), but not when a more stringent definition of the LT-HSC pool was adopted by using SLAM markers in combination with CD244 and CD229 (Oguro et al., 2013) (Figure S1B). To determine whether these changes were also seen in a bacterial sepsis model, we utilized a mouse model of cecal ligation and puncture (CLP)-induced polymicrobial peritonitis (Ferreira et al., 2014). Responses noted in the LPS model were to a large extent recapitulated in the CLP model (Figure 1G). Modest differences between the models are likely due to the higher complexity of bacterial sepsis compared with LPS challenge. Collectively, these data show that both LPS and bacterial sepsis result in a dramatic increase in MPP1 and MPP2 subsets, associated with no or modest expansion of LT-HSC, and that TRIF, but not MYD88, is required for this process.
Discussion TLR4 activation via MYD88 or TRIF pathways plays a central role in regulating host defense against bacterial infection (Poltorak et al., 1998; Beutler, 2000; Weighardt and Holzmann, 2007). However, the specific contribution of MYD88 and TRIF in regulating the BM response to sepsis is poorly understood. In this study, we dissected these two pathways using knockout mice lacking MYD88 or TRIF function in combination with P. aeruginosa LPS-induced endotoxemia (Rodriguez et al., 2009) or CLP-induced peritoneal polymicrobial sepsis (Ferreira et al., 2014). Our observations unveil two distinct mechanisms mediating the effects of severe sepsis on BM homeostasis. MYD88 activation is critical for myelosuppression occurring in BM during sepsis, whereas TRIF activation mediates permanent injury of HSC (Figure 7E). Based on our previous findings (Rodriguez et al., 2009), we initially hypothesized that the myelosuppression occurring during sepsis was caused by dysfunctional expansion of HSPC (LSK) associated with a block of differentiation and inability to produce adequate myeloid precursors. However, our studies using MYD88−/− and TRIF−/− mice suggest that myelosuppression and HSPC dysfunction can occur as two independent processes. Analysis of BM response to LPS challenge or severe sepsis showed that TRIF deficiency prevented LSK expansion, but not the loss of myeloid progenitors. In contrast, MYD88 deficiency preserved myeloid cells but did not abolish the LSK expansion. This suggests that TRIF and MYD88 have cell-specific effects.