Our study demonstrates that HSCs can undergo a transition from a dormant state to an active state during an intracellular bacterial infection. This conclusion is based on our observations that LSK cells isolated from the bone marrow during infection exhibited increased proliferation, reduced engraftment, and a loss of long-term repopulating potential, all characteristics of activated HSCs. Although it has been demonstrated that HSCs can be activated by IFNγ during chronic mycobacterial infection 
, our study demonstrates that HSC activation can also occur during acute bacterial infection, and, can be reversed quite rapidly. We demonstrate that this transition from dormancy to activation is IFNγ-dependent, supporting the notion that IFNγ acts on HSCs during inflammation. Similar changes in HSCs have been shown to occur following IFNα expression, induced by poly I:C treatment in vivo 
. Thus, an emerging paradigm is that IFNs control the basal state of the dormant HSC population. As most infections trigger the production of IFNs, we postulate that the transient activation of dormant HSCs is a common occurrence.
Among hematopoietic cells, expression of the α chain of the IFNγ receptor is highest in dormant LT-HSCs 
, consistent with the observation that HSC activation by IFNγ occurs via direct signaling on this cell population. Although we have not shown in this study that IFNγ acts directly on HSCs during ehrlichiosis, such a conclusion is strongly supported by our previous data, and other published studies that demonstrated that IFNγ-responsive genes, such as Adar
, are activated in and required by HSCs in response to inflammatory-stress 
. IFNγ signaling under homeostatic conditions does not appear to be essential, as IFNγ and IFNγR-deficient mice exhibit normal hematopoiesis and bone marrow cellularity.
We hypothesize that, during infection, HSCs transition from dormancy to activity in order to promote the expansion of more differentiated progenitor cells, such as myeloid-biased MPPs. This interpretation is consistent with our previous findings that intracellular ehrlichial infection induced myelopoiesis 
. In that study we demonstrated that direct IFNγ signaling in promyelocytes resulted in the expression of transcription factors essential for the proper differentiation of granulocytes and monocytes. The changes we have observed are not limited to the ehrlichia. We have observed similar changes to the hematopoietic compartment during influenza and Mycobacterium tuberculosis
infections (KM and GW, unpublished data), and similar IFNγ–directed alterations in hematopoiesis have been reported during malarial infection 
. IFNγ signaling has been considered to be detrimental during chronic infection 
, but we propose that IFNγ signaling is essential for activating the hematopoietic response to infection, and for the production of innate immune cells required for combating infection.
Hematopoietic stem and progenitor cells express Toll like receptors (TLRs), and it has been shown that these cells can be directed to differentiate upon direct interaction with TLR ligands 
. In addition, chronic exposure to LPS, which can stimulate cells via TLR4, results in phenotypic and functional changes within the HSC population 
. It has been proposed that circulating HSCs and progenitor cells express TLRs so that they can respond directly to pathogens, and differentiate in situ
, in non-lymphoid tissues 
. E. muris
is not known to encode canonical TLR-ligands, so it is likely that IFNγ provides an alternative mechanism to activate HSCs.
Prospective isolation of HSCs based on cell surface marker expression has been an important focus of hematopoiesis research, and has relied on the use of cell surface markers to facilitate the enrichment of cells with the most potent self-renewal and differentiation potential. During infection, expression of inflammatory cytokines, such as IFNγ, can modulate the expression of cell surface proteins used to identify stem and progenitor cells 
. We have shown that the LSK population, which contains HSCs, undergoes phenotypic and functional changes during ehrlichial infection, similar to findings obtained in studies of viral, parasitic, and other bacterial infections 
. Initially, we observed that the number of LSK CD34− CD135− cells increased during infection, which suggested that the LSK population contained more HSCs. However, our studies demonstrated that the infection-induced LSK cells exhibited poor engraftment, indicating that the LSK CD34− CD135− cell population contained few LT-HSCs. We showed that the decreased engraftment potential of the LSK population on day 8 post-infection correlated with a decrease in the number of dormant HSCs, where dormant HSCs were characterized as LSK CD150+ CD48− CD135− CD34− cells. By day 16 post-infection, when the dormant HSC population was found to recover, we observed a similar increase in LSK engraftment potential. In a model of cecal ligation-induced bacterial sepsis, LT-HSC cells were shown to expand significantly within the bone marrow 
, findings which are in apparent contradiction to our own. However, in the study by Scumpia et al.
, the authors defined LT-HSCs as LSK CD150+ CD135− cells. Whereas we, too, found that this population expanded during ehrlichiosis, we observed high expression of CD48 on these cells. Thus, we conclude that the LSK CD150+ CD135− CD48+ cells represent more differentiated MPPs, not LT-HSCs. Thus, additional phenotypic markers may be required to distinguish LT-HSCs from more differentiated progenitor cells under inflammatory conditions.
Our findings, together with the studies from other laboratories, demonstrate that infections can mediate profound changes to HSCs by driving, via interferons, their transient conversion from a state of dormancy to activity. It will be important to gain a better understanding of how IFNγ mediates the activation of dormant HSCs, how IFNγ signaling is regulated in these cells, and how this process impacts host defense during infectious diseases of public health significance.