Abstract
Previous experiments in vitro have emphasized the important role of a5b1 integrin/fibronectin interactions in terminal stages of erythroid differentiation (JCB1987, 105:3105), whereas in vivo experiments with genetically deficient mice (JI2000, 165:4667) and recent in vitro ones emphasized the important contribution of a4b1 integrin in the expansion of fetal erythroid progenitors (JCB2007, 177:871) or for optimal responses post stress in adult animals (MCB2003, 23:9349). However, no abnormalities in erythropoiesis were reported in a model of conditional ablation of b1 integrins post-transplantation (Blood2006, 108:1857). Therefore, it has not been clear to what extent each of the two major b1 integrins (a4b1 and a5b1) alone or in combination is critical for expansion and/or terminal erythroid differentiation of adult cells at homeostasis and/or after stress.
We have made detailed and parallel observations comparing erythropoiesis in two genetic models with conditional ablation of b1 or a4 integrins at homeostasis and after phenylhydrazine (PHZ)-mediated stress. Basal erythropoiesis in b1-, a4-deficient and control mice as assessed by hematocrit levels and total nucleated erythroid cells (Ter119+) in BM and spleen was similar. Furthermore, both b1 and a4-deficient mice showed an increase in circulating progenitors (1275±230 CFC/ml PB, 2446±256 CFC/ml PB, respectively) over controls (338±113 CFC/ml PB). However, post PHZ-induced hemolytic stress there was a dramatic difference in outcomes of b1-deficient, but modest differences in a4-deficient mice compared to controls. Survival of b1-deficient mice by day 6 post PHZ was 33% compared to 100% in a4-deficient and control groups. In b1-deficient animals, no significant increase in spleen cellularity (153±26×106 and194±64×106 cells/spleen at day 0 and 6 post PHZ, respectively) was detected and the expansion of total erythroid precursors (CD71hi,Ter119+) in the spleen was minimal (from 2.08×106 to 10.8×106 cells/spleen at day 6). In contrast, in a4-deficient and control mice by the same time spleen cellularity increased respectively by 3 and 8 fold, and erythroid precursors expanded by 400 and 2,500 fold. Of interest, BM response to PHZ was not significantly different among all groups.
To test whether the splenic response was cell-autonomous or environmentally controlled we compared PHZ response in wild type recipients reconstituted with b1-ablated (Cre+b1D/D) or with control (Cre-b1f/f) BM cells. Recipients of b1-ablated cells had an impaired response compared to recipients of control cells, which was somewhat intermediate to that seen in non-transplanted b1-deficient animals; by day 6 post PHZ, spleen cellularity was 300±24×106 cells/spleen and erythroid precursors expanded by 130 fold in recipients of b1-ablated BM cells compared to 859±159×106 cells/spleen and 900 fold precursor increase in control recipients. These data suggest that both erythroid and their environmental cells were responsible for the reduced survival and poor spleen response in b1-deficient mice. The target environmental cells (fibroblasts, endothelial cells, macrophages) and/or matrix involved will be the focus of future studies. It is of interest that in contrast to splenic response, the increased release of progenitors from BM seen in animals reconstituted with b1D/D cells was as high as that seen in non-transplanted b1- deficient animals and with the same qualitative characteristics, suggesting this alteration in biodistribution of progenitors is cell autonomous.
Taken together, our data suggest that
a combined expression of b1 integrins in erythroid and cells in their microenvironment is critical for survival and optimal splenic response to a PHZ-induced stress in adult mice; release of progenitors seen at homeostasis in both b1 and a4 models is cell autonomous with a preferential erythroid progenitor release from BM seen only in b1-deficient but not in a4-deficient mice; in contrast to results with fetal liver cells showing a critical role of a4b1 but not a5b1 integrin for proliferative expansion of erythroid cells, in adults a5b1 expression in erythroid and environmental cells in the spleen assumes a more critical role.
Our data expand the current knowledge on the distinct dependency of a4b1 vs a5b1 integrins in basal vs stress erythropoiesis and bridge previously divergent information from in vitro and in vivo experiments.
In vivo experiments do not support the charge zipper model for Tat translocase assembly
