Effects of Interferon alpha on mouse and human hematopoietic stem cells

Abstract

Maintenance of the hematopoietic system is dependent on hematopoietic stem cells (HSCs). During homeoastasis HSCs are quiescent. However upon injury, HSCs can efficiently be activated, leading to repair of the system. Signals leading to the activation of quiescent HSCs are still largely unknown. Recently our group has shown that administration of IFNα leads to activation of mouse HSCs in vivo. This is mediated by activation of IFNAR/STAT1 signaling followed by the up-regulation of Sca-1, however the exact mechanism of cell cycle activation remains unclear. To get further insight into this process we performed microarray analysis on HSCs after treatment with IFNα. This screen identified several candidate genes, which are potentially involved in HSC activation, including cell cycle regulators like p57KIP2, Maged1 and Reprimo as well as cytokines like Ccl5 and Cxcl10, which are key regulators of inflammatory responses. Furthermore we identified interferon response genes like Ifitm1, Ifitm3, Iigp1, Iigp3 or Ddx58, which were previously linked to regulation of proliferation in different contexts. Along these studies we uncovered that Ifitm1 and Ifitm3 expression is highly enriched within hematopoietic stem and progenitor cells both on the RNA as well as on the protein level. Moreover expression is further induced by IFNα. However mice lacking the Ifitm family show normal hematopoiesis and normal HSC numbers and cycling behavior of HSCs in homeostatic conditions. Ifitm-deficient HSCs are capable to self-renew and differentiate similar to wild-type HSCs. This suggests that the Ifitm protein family is dispensable for HSCs during homeostasis. Notably Ifitm-deficient HSCs are also efficiently activated by IFNα, similar to wild type HSCs. Microarray analysis of HSCs from Ifitm deficient mice, both during homeostasis and after administration of IFNα, showed no differences in the expression profiles, indicating a role for the Ifitm family as terminal effectors rather than regulatory proteins within HSCs. During our study it was shown by others that the Ifitm family is a potent viral restriction factor in endothelial cells. We are currently investigating whether Ifitm proteins have a similar role in the immune defense of HSCs. Thus far it is still unclear whether human HSCs are similarly activated by IFNα as mouse HSCs. To elucidate this we established a xenotransplantation model with human cord blood cells, which allows testing of the effects of IFNα on human HSCs in vivo. Surprisingly, unlike mouse HSCs, human HSCs are not activated by IFNα in this model. Notably human HSCs in this model are already less quiescent during homeostasis compared to their mouse counterparts. In the mouse also the bacterial endotoxin LPS can induce cell cycle activation in HSCs. Surprisingly LPS similarly activates human HSCs in our model. Gene expression analysis showed a high overlap between the genes induced in mouse and human HSCs after LPS treatment, while IFNα only affected cell cycle regulatory genes in murine HSCs. One explanation for this phenotype could be an impaired interaction of murine stromal niche cells with human HSCs and we are currently investigating this in more detail. Finally we examined the effect of IFNα on quiescent leukemic stem cells (LSCs). While IFNα is known to activate normal mouse HSCs, it is unclear whether also LSCs are similarly affected. To address this we investigated the effects of IFNα on LSCs in a mouse model for chronic myeloid leukemia. Surprisingly LSC were less efficiently activated compared to normal HSCs. This can be explained by down-regulation of the IFNAR by BCR-ABL kinase activity, which was previously described in vitro. This also highlights the importance of exact timing of LSC activation and treatment in combination therapy approaches. Our group is currently using this model to further identify and optimize new possible combination therapies.