Unique Roles for Grb2-Associated Binding (Gab) Adaptors in Hematopoietic Stem Cell Homeostasis and Repopulation

Abstract

Homeostasis of hematopoietic stem cells (HSCs) is essential for the continuous replenishment of the hematopoietic system throughout the entire lifespan of an organism and is a tightly regulated process, controlled by intrinsic and extrinsic signals. Although a variety of molecules controlling HSC homeostasis are known, it remains unclear how signaling is regulated. The growth factor receptor-bound protein 2 (Grb2)-associated binding (Gab) proteins are a family of Pleckstrin homology domain containing adaptor proteins that dock with receptor tyrosine kinases (RTKs) and non-RTKs to control cell activation of ERK and PI3K pathways. The murine Gab family includes Gab1, Gab2, and Gab3. Gab1 and Gab2 are expressed throughout hematopoiesis including HSCs but Gab3 is limited mostly to macrophages, dendritic cells, and T-cells. Previously we reported that Gab2 is intrinsically required for hematopoietic cell responses to early-acting cytokines, with defective hematopoiesis in Gab2-/- mice (obtained from Toshio Hirano, Osaka, Japan; Nishida K et al, Blood, 99, 2002). Little is known about redundancy among Gabs in hematopoiesis and it is possible that roles for Gab1 and Gab3 have been obscured by the prominence of Gab2. Using Gab1flox/flox (Bard-Chapeau EA et al, Nat. Med. 11, 2005) and Gab3-/- mice (obtained from Pam Hankey-Giblin, Penn. State; Seiffert M et al, Mol. Cell. Biol. 23, 2003) we set out to make compound mutant mice. Gab3 knockout mice do not have reported hematopoietic phenotypes. First, we generated Gab2/3 double knockout mice. In the peripheral blood, we observed greater B-lymphocyte and erythroid deficiency compared to wild-type or Gab2-/- mice. In non-competitive bone marrow transplantation, Gab2/3-/- bone marrow was sufficient to reconstitute relatively normal myeloid hematology in irradiated recipients but resulted in significantly reduced donor derived B- and T-lymphocytes relative to wild type or Gab2-/- donor bone marrow cells, which were the same. In competitive bone marrow transplantation against Boy J (CD45.1) we observed sharp reductions in multilineage engraftment of Gab2/3-/- (13-15 fold) compared to Gab2-/- (4-6 fold) mice while Gab3-/- mice were normal. These data combined with the limited expression pattern for Gab3 suggest a novel critical extrinsic role in hematopoiesis. Despite severe multilineage engraftment defects, we surprisingly observed that Gab2/3-/- mice had significantly increased numbers of KLS cells (~2-fold) as well as LT-HSCs (KLSCD150+CD48-, ~3-fold) compared to the wild-type controls. Associated with the increased HSC numbers were cell cycle changes observed by staining with Pyronin Y/Hoechst 33342. During steady-state, Gab2/3-/- KLS cells were less quiescent with a lower percentage of cells in G0 (Gab2/3-/- vs WT: 40.5 ± 6.1 vs 57.4 ± 11.4, N=4, p=0.04) and more cells in G1 (Gab2/3-/- vs WT: 19.5 ± 0.7 vs 13.5 ± 3.5, N=4, p=0.01) phase of the cell cycle. Interestingly, KLS cells from Gab2/3-/- mice were less apoptotic than wild-type control (1.05 ± 0.49% in Gab2/3-/- (N=7) vs 5.97 ± 1.40% in WT KLS (N=10), p<0.001). Together these indicate that the HSCs were expanded but functionally defective. To further explore the mechanisms for Gab2/3-/- defects, we found that Gab1 expression levels were increased 2-fold. Since germline Gab1 deficiency results in embryonic lethality, Gab1 floxed alleles were deleted conditionally using Vav1-Cre to generate viable adult Gab1/2/3-/- mice. Real-time PCR confirmed the average of 94% Gab1 deletion efficiency (N=10) and Western analysis showed no detectable level of Gab1 in splenocytes. Interestingly, Gab1 deletion reversed the KLS expansion observed in the Gab2/3-/- mice indicating that Gab1 has effects that oppose Gab2/3 at the KLS level. However, Gab1/2/3-/- mice had severe defects in multilineage engraftment in competitive repopulation (39-52 fold). Altogether, our data indicate that Gab2/3 play positive roles while Gab1 plays a negative role in HSC homeostasis but all three Gabs are essential for multilineage repopulation. Further studies of Gab1 and Gab3 function in intrinsic and extrinsic regulation of HSC biology and repopulation are ongoing. DisclosuresNo relevant conflicts of interest to declare.