Hematopoietic defects in agrin-deficient (Musk-L;Agrn^−/−) mice. (A) Representative images of hematoxylin and eosin–stained sections of bone marrow from P5 control (ctrl) and Musk-L;Agrn^−/− mice. Scale bar, 200 μm. The histogram on the right shows the average absolute numbers of total CD45⁺ cells (ctrl, n = 20; Musk-L;Agrn^−/−, n = 10; ***P < .0001) (B-C) Absolute numbers of CD45⁺ cells in spleen (B) or blood (C) of control and Musk-L;Agrn^−/− mice (spleen: ctrl, n = 12; Musk-L;Agrn^−/−, n = 7; **P < .01; blood: ctrl, n = 7; Musk-L;Agrn^−/agn^−/−, n = 4; ***P < .0001). (D) Absolute numbers of CD45⁺ hematopoietic cells in control and Musk-L;Agrn^−/− bone marrows (ctrl, n = 12; Musk-L;Agrn^−/agn^−/−, n = 8; *P < .05, **P < .01). (E) MCM2 immunoreactivity of bone marrow sections of P5 control (ctrl; n = 10) and agrin-deficient (Musk-L;Agrn^−/−; n = 7) mice. Scale bar, 100 μm. The graph shows the analysis of the proliferation index (***P < .0001). (F) In vivo BrdU incorporation experiments. Bone marrow cells from control or Musk-L;Agrn^−/− mice were analyzed 2 hours after intraperitoneal injection. Left, representative flow cytometry profiles; right, statistical analysis. (ctrl, n = 10; Musk-L;Agrn^−/−, n = 4; ***P < .0001). In all panels, error bars represent SEM.