Permissive role of thrombopoietin and granulocyte colony-stimulating factor receptors in hematopoietic cell fate decisions in vivo

Abstract

The question of whether extracellular signals influence hematopoiesis by instructing stem cells to commit to a specific hematopoietic lineage (instructive model) or solely by permitting the survival and proliferation of predetermined progenitors (permissive model) has been controversial since the discovery of lineage-dominant hematopoietic cytokines. To study the potential role of cytokines and their receptors in hematopoietic cell fate decisions, we used homologous recombination to replace the thrombopoietin receptor gene (mpl) with a chimeric construct encoding the extracellular domain of mpl and the cytoplasmic domain of the granulocyte colony-stimulating factor receptor (G-CSFR). This chimeric receptor binds thrombopoietin but signals through the G-CSFR intracellular domain. We found that, despite the absence of a functional mpl signaling domain, homozygous knock-in mice had a normal platelet count, indicating that in vivo the cytoplasmic domain of G-CSFR can functionally replace mpl signaling to support normal megakaryopoiesis and platelet formation. This finding is compatible with the permissive model, according to which cytokine receptors provide a nonspecific survival or proliferation signal, and argues against an instructive role of mpl or G-CSFR in hematopoietic cell fate decisions.

Figure 1

The mpl/G-CSFR chimera. (A) Domain swap and the resulting chimeric protein. Hatched box, transmembrane domain of the G-CSFR; solid box, mpl transmembrane domain. (B) Proliferation of BaF3 cells transfected with expression vectors carrying the indicated cDNAs in response to increasing TPO concentrations. (C) Strategy used for gene targeting of the mpl locus. Solid boxes represent exons. Exons 11 and 12 were replaced with a cDNA for the G-CSFR-signaling domain (hatched box), followed by a simian virus 40 polyadenylation signal (shaded box). Excision of the neo cassette by cre recombinase generates the floxed ki (fki) allele. A, ApaI recognition site. (D) Southern analysis of the mpl locus after ApaI digestion of DNA from the targeted ES clone (ES), compared with genomic DNA from mice before (+ neo), and after germ-line excision of the plox-neo gene (Δ neo). Numbers indicate sizes of restriction fragments in kilobases. The same membrane was successively probed with the three hybridization probes that are depicted in C.

Figure 2

Expression of chimeric mpl/G-CSFR. (A) Ribonuclease protection assay. Position of riboprobe used to distinguish wild-type from chimeric mpl/G-CSFR mRNA. Solid box, transmembrane domain. Arrows, length of protected fragments for the wild-type mpl and chimeric transcripts. (B) Expression of wild-type and chimeric mRNA in tissues of heterozygous +/ki mice: bm, bone marrow; sp, spleen; th, thymus; ln, mesenteric lymph node; br, brain; ki, kidney; lu, lung; ht, heart; li, liver; pa, pancreas. A riboprobe for mouse hypoxanthine phosphoribosyltransferase (hprt) was used as an internal control for RNA loading. (C) Expression of chimeric mRNA in spleens of ki and wild-type mice. On the left is the original mouse strain containing the neo cassette (+ neo); on the right is the mouse strain after germ-line excision of the plox-neo gene (Δ neo); fki, the floxed ki allele. (D) Flow cytometric analysis of Percoll-fractionated bone marrow cells. In the top row is shown the biotinylated anti-mpl with streptavidin-PE (thick line) versus isotype control (thin line). In the bottom row is two-color staining with anti-mpl/streptavidin-PE and fluorescein isothiocyanate-labeled anti-CD41. Numbers indicate percentages of cells in each quadrant.

Figure 3

Blood counts and bone marrow progenitor numbers. (A) Analysis of platelets and neutrophil granulocytes. Results represent the means ± SEM of nine mice for each genotype except for ki/ko mice, for which the means ± SEM of four mice are given. (B) Analysis of megakaryocyte progenitors (CFU-Meg), granulocytic progenitors (CFU-G), and granulocyte-macrophage progenitors (CFU-GM). Results represent the means ± SEM of three mice. ND, not determined.

Figure 4

Effects of the chimeric mpl/G-CSFR on megakaryopoiesis and granulopoiesis in ki/ki mice. (A) Predictions of the instructive model. Signals generated by the G-CSFR part of the chimeric receptor (open box) instruct the stem cell or early progenitor to commit to the granulocytic lineage (thick arrow). This results in increased numbers of granulocytic progenitors (CFU-G). Because the mpl signaling domain is absent in ki/ki mice, commitment to the megakaryocytic lineage is reduced (dashed lines). This results in a reduction of megakaryocyte progenitors (CFU-Meg) and platelets similar to the mpl ko/ko mice. N, normal. (B) Predictions of the permissive model. Commitment of stem cells and early progenitors is independent of cytokine signaling (dashed arrows). G-CSFR signals generated by the chimera can substitute for the absence of mpl signaling resulting in a normal megakaryopoiesis.