Thrombopoietin (Tpo) acts via its cell surface receptor (MPL/TpoR) to promote megakaryocyte progenitor survival, proliferation and differentiation via activation of JAK2, STAT5, STAT3, STAT1, ERK1,2-MAPK-kinase and PI-3'-kinase/Akt. In addition, Tpo signaling in the bone marrow niche is required for maintaining hematopoietic stem cells (HSCs). Recently we showed that in cells expressing high JAK2, such as megakaryocytes or engineered cell lines, Tpo induced an anti-proliferative and pro-senescence physiologic effect. Down-modulation of TpoR in myeloproliferative neoplasms (MPNs) progenitors might confer resistance to such anti-proliferative effects in megakaryocytes, thus contributing to their uncontrolled proliferation.

Here we report that Tpo activates tyrosine phosphorylation and transcriptional activation of STAT2, and of a complex containing STAT2, STAT1 and IRF-9. This is specific for type I interferon (IFN) and is associated with anti-proliferative and antiviral effects. We detected Tpo induced STAT2 activation in platelets of thrombocytopenia patients treated in vivo with Tpo or Tpo agonists as well as in human CD34+-derived megakaryocytes. In cell lines reconstituted with MPL/TpoR and JAK2, Tpo behaved like an IFN by protecting cells from subsequent viral infection and inducing IFN-stimulated gene expression. We asked whether Tpo-induced STAT2 activation depended on type I IFN receptors, on TYK2 or on STAT1. Using primary cells from mice where these genes have been knocked-out we show that Tpo indeed activates STAT2 via TpoR itself, and does neither require IFNAR1, IFNAR2 nor TYK2, demonstrating that this effect is direct.

By using cell lines expressing different levels of JAK2 we show that Tpo exerts a bimodal effect. At low JAK2 levels Tpo induces mainly STAT5, STAT3 and STAT1 activation, while at high JAK2 levels (comparable to those in UT7 cells or primary megakaryocytes) Tpo induces STAT2/STAT1/IRF9 activation, in addition to STAT5/STAT3/STAT1 activation. Interestingly, the STAT2 signaling is dominant and induces an anti-proliferative effect, which can be inhibited by shRNA targeting STAT2, or by mutation of distal receptor tyrosines. Moreover, using STAT2-/- mice we show that STAT2 activation by Tpo is relevant for the steady-state levels of platelets, with knock-out mice having significantly higher platelet numbers. Recovery studies after 5-FU injection and examination of HSCs and early progenitors showed that STAT2 acts as a negative regulator of the stimulatory effects of Tpo on myelopoiesis.

We recently identified inactive and active dimeric conformations of TpoR. Here we tested which dimeric receptor conformation activates STAT2 using dimeric fusion proteins where a dimeric coiled coil (Put3) was fused to the transmembrane segment. While cc-TpoR-I was found to recapitulate the classical stimulatory effect of Tpo on STAT5, STAT3 and STAT1 and platelet formation in vivo, here we show that STAT2 activation is induced by another dimeric conformation, represented by cc-TpoR-IV, which is the strongest JAK2 activator. cc-TpoR-IV also induces strong cell-to-cell adhesion possibly reminiscent of the intercellular HSC interactions to the bone marrow niche cells. Increasing JAK2 protein levels leads to stabilization of a Tpo-activated TpoR dimer that assumes a conformation that corresponds to cc-TpoR-IV. Thus, JAK2 levels act on fine-tuning the dimeric conformation of the TpoR dimer, switching it from a classical proliferative cytokine to an IFN-like cytokine, which is anti-proliferative.

The new Tpo-induced signaling pathway we describe is of very high relevance for MPN pathogenesis and treatment with IFN-I. Our data indicate that HSCs and megakaryocytes where TpoR-JAK2 V617F signal would be primed for IFN action, and might explain why IFN-I targets for inhibition the mutated HSCs, eventually decreasing the allele burden. Last but not least, autoimmune and viral-induced thrombocytopenia patients are treated with Tpo mimetics, and we could detect STAT2 activation in certain patients after in vivo Tpo treatment. Differences in the ratios of classical versus STAT2 activation might be at the basis of differences in response to Tpo of such patients.

Disclosures

Constantinescu:Shire: Consultancy, Speakers Bureau; Amgen: Consultancy, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Personal Genetics: Scientific Adv Board, Scientific Adv Board Other, Speakers Bureau; Dafra Pharma: Consultancy, Scientific Advisory Board Other.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution