Neonatal Megakaryocytes Do Not Mature Normally In the Absence of Thrombopoietin In Vivo: Potential Role of Developmental Differences In CXCR4 Expression

Abstract 88

Thrombopoietin (Tpo) and its receptor (c-mpl) constitute the main regulatory axis of megakaryocyte (MK) proliferation and maturation. Accordingly, adult Tpo and c-mpl knockout (KO) mice exhibit an 85% reduction in MK concentration, although the residual 10–15% MKs are ultrastructurally normal. The phenotype of newborn Tpo or c-mpl KO mice has not been well characterized, but we and others have described substantial molecular differences between neonatal and adult megakaryocytopoiesis, including several pathways that mediate the high proliferative rate of neonatal MK progenitors. Recently, two groups reported neonates with c-mpl mutations associated with congenital amegakaryocytic thrombocytopenia, who in the neonatal period exhibited normal numbers of immature appearing marrow MKs. The fact that these infants were severely thrombocytopenic at that time suggests that their MKs did not produce platelets normally. These reports, coupled with our recent observation that Tpo mediates the cytoplasmic maturation of human neonatal MKs, led us to hypothesize that, during fetal and neonatal life, Tpo-independent pathways predominantly stimulate MK proliferation, while MK maturation is Tpo-dependent. To test this hypothesis, we studied the characteristics of MKs generated in vivo in neonates in the absence of c-mpl. Since we have previously demonstrated that the liver is the main site of megakaryocytopoiesis in newborn mice, we evaluated MKs in the livers of c-mpl KO and WT mice (both C57BL/6) on day of life 1 and 3. As a first step, we quantified MKs immunohistochemically stained with an anti-vWF antibody, and found that MKs in the liver of newborn c-mpl KO mice were reduced by approx. 70%. Next, we examined the ultrastructure of these liver MKs by transmission electron microscopy, and categorized c-mpl KO MKs (n=28) and WT MKs (n=32) as stage I (immature), stage II (abundant alpha-granules and a developing demarcation membrane system, DMS), or stage III (platelet producing MKs, with an open DMS). According to these criteria, 50% of WT MKs were stage II, and 50% were stage III. In contrast, 22% of c-mpl KO MKs were stage I, 57% were stage II, and only 21% were stage III. Furthermore, significant ultrastructural abnormalities were found in 70% of c-mpl KO MKs, including decreased numbers of platelet granules, a very disorganized appearing closed demarcation membrane system, and/or an abnormally wide peripheral zone. Since MKs in adult mice mature normally in the absence of Tpo, we then hypothesized that our findings reflected a downregulation of Tpo-independent pathway(s) mediating MK maturation in neonates. In that regard, we recently found that the microRNA miR9 was expressed at 10- to 14-fold higher levels in murine fetal and neonatal compared to adult MKs. Since CXCR4 (the receptor for SDF-1) is a predicted target of miR9, and in view of recent studies characterizing the role of the SDF-1/CXCR4 axis as a Tpo-independent pathway that stimulates MK maturation, we evaluated CXCR4 protein expression in cultured MKs derived from murine fetal liver (E13.5), newborn liver, and adult bone marrow, by Western Blot. As predicted, CXCR4 protein levels were significantly lower in fetal and neonatal compared to adult MKs (p=0.003). To evaluate the significance of these findings in humans, we then quantified miR9 and CXCR4 protein levels in cord blood-derived and adult peripheral blood-derived human MKs (n=3 per group). Consistent with the murine findings, we found that miR9 levels were approximately 20-fold higher and CXCR4 protein levels were significantly lower in human neonatal compared to adult MKs (p<0.05 for both). Finally, to determine whether miR9 regulates CXCR4 protein expression, Meg-01 cells were nucleofected with miR9 or Cy3 (control). As hypothesized, up-regulation of miR9 resulted in a significant reduction in CXCR4 protein levels compared to control cells (p=0.02). In conclusion, our findings indicate that MKs in the neonatal period do not mature normally in the absence of Tpo, presumably due to a deficiency in Tpo-independent pathway(s) of MK maturation at this developmental stage. Our data also identified a developmental downregulation of CXCR4 protein expression by miR9 in fetal and neonatal MKs. Given the role of the SDF/CXCR4 axis mediating Tpo-independent MK maturation, this provides a potential mechanism to explain the c-mpl KO findings.