Lyn kinase promotes erythroblast expansion and late-stage development

Abstract

Lyn kinase is known to modulate the formation and function of B cells, monocytes, and mast cells. However, Lyn-/- mice also develop erythrosplenomegaly, and cases for both negative and positive erythropoietic actions of Lyn recently have been outlined. In phenylhydrazine-treated Lyn-/- mice, extramedullary splenic erythropoiesis was hyperactivated, but this did not lead to accelerated recovery from anemia. Furthermore, ex vivo analyses of the development of bone marrow-derived Lyn-/- erythroblasts in unique primary culture systems indicated positive roles for Lyn at 2 stages. Late-stage Lyn-/- erythroblasts exhibited deficit Ter119(pos) cell formation, and this was paralleled by increased apoptosis (and decreased Bcl-xL expression). During early development, Lyn-/- erythroblasts accumulated at a Kit(pos)CD71(high) stage, possessed decreased proliferative capacity, and were attenuated in entering an apparent G1/S cell-cycle phase. In proposed compensatory responses, Lyn-/- erythroblasts expressed increased levels of activated Akt and p60-Src and decreased levels of death-associated protein kinase-2. Stat5 activation and Bcl-xL expression, in contrast, were significantly decreased in keeping with decreased survival and developmental potentials. Lyn, therefore, is proposed to function via erythroid cell-intrinsic mechanisms to promote progenitor cell expansion beyond a Kit(pos)CD71(high) stage and to support subsequent late-stage development.

Figure 1.

Phenylhydrazine-induced extramedullary erythropoiesis in Lyn^-/- mice is enhanced but does not assist recovery from anemia. (A) Lyn^+/+ and Lyn^-/- mice (8-12 weeks) were treated with phenylhydrazine (PHZ). At day 3.5, splenocyte preparations were assayed for frequencies of Ter119^pos erythroblasts. Data shown (including several-fold increases in CD71^pos and Ter119^pos Lyn^-/- erythroblasts) are representative of 3 independent experiments (Table 1). (B-C) Frequencies of splenic CFUe's and BFUe's also were determined for phenyhydrazine-treated Lyn^-/- and control mice (means ± SD, n = 3) together with increases in spleen weight (means ± SE, n = 3 per group). (D) Possible effects of enhanced splenic erythropoiesis in phenylhydrazine-treated Lyn^-/- mice on red cell mass were assessed based on mean hematocrits (± SE, n = 6). For comparison, mice expressing a knocked-in EpoR-H allele also were analyzed. For EpoR-H mice (but not Lyn^-/- mice), mean hematocrits at day 15 differed significantly from wild-type controls. ^*P < .01.

Figure 2.

Bone marrow-derived Lyn^-/- erythroblasts falter in their development to Ter119^pos erythroblasts. (A) Defects in Lyn^-/- bone marrow-derived erythroblast development at a Ter119^pos stage. Following 2.5 days of culture in SP34-EX medium, expanded Lyn^+/+ and Lyn^-/- erythroid progenitor cells were shifted to a transferrin, insulin, and Epo-containing differentiation medium. At day 1.5, frequencies of maturing erythroblasts were assayed based on Ter119 marker expression (i). Frequencies of CD71^highTer119^pos copositive cells also were assayed (bottom panels, circled populations). Overall frequencies of Ter119-positive and CD71⁺ erythroblasts from 3 independent experiments (n = 3 mice per experiment) are also graphed (i; mean ± SE). (B) CD71 and Ter119 marker expression among expanded erythroblasts also was assessed prior to differentiation. During this short-term expansion, marker profiles for Lyn^+/+ and Lyn^-/- populations were essentially equivalent. (C) Lyn^-/- erythroblasts exhibit decreased survival at a Ter119^pos stage. Bone marrow-derived erythroblasts from Lyn^-/- and Lyn^+/+ mice were expanded and shifted to differentiation conditions. At 24 hours of culture, frequencies of annexin V-positive cells among the maturing erythroblasts were assayed. Increased death among Lyn^-/- erythroblasts also was confirmed by propidium iodide staining (data not shown). (D) In these cell populations (and at 24 hours of differentiation), levels of Bcl-xL and of activated PY416-p60-Src (p-Src) expression also were analyzed (by Western blotting) together with Lyn and GAPDH.

Figure 3.

System for the ex vivo development of adult bone marrow-derived erythroblasts. (A) Outlined are steps used in the serum-free expansion and MACS-based isolation of marrow-derived erythroid progenitor cells. (B) Flow cytometric, cytospins, and colony-forming features of isolated erythroid progenitor cells. (i) Upon expansion for 4.5 days, cultures routinely were composed of approximately 50% to 60% Kit^posCD71^high and approximately 40% to 50% Kit^negCD71^high erythroblasts. (ii-iii) Wright-Giemsa staining of isolated Kit^negCD71^high and Kit^posCD71^high erythroblasts. (iv) Colony morphologies of Kit^posCD71^high cells cultured for 2.5 days in 0.9% methylcellulose with Epo at 5 U/mL. (C) Differentiation capacities of expanded erythroblasts. Kit^posCD71^high cells were isolated (by MACS) from expansion cultures and transferred to BSA, insulin, and transferrin (BIT)-containing medium. At 24 hours of culture, few cells continued to express Kit (CD117) while more than 50% routinely differentiated to Ter119^pos erythroblasts (lower right panel, cytospin preparation).

Figure 4.

Early-stage Lyn^-/- erythroid progenitor cells accumulate at a Kit^posCD71^high stage of development and are deficient in Epo and SCF proliferative responsiveness. (A) At the indicated time points, Kit^posCD71^low, Kit^posCD71^high, and Kit^negCD71^high erythroblast formation was assayed. At 48 hours, 2 subpopulations of Kit^pos cells reproducibly were detected (and are designated cohorts I and II). For Lyn^-/- erythroblasts, note accumulations at a Kit^posCD71^high stage (especially at 72 and 120 hours) and corresponding deficits in Kit^negCD71^high cell formation. Data illustrated are representative of 3 independent experiments (and additional representative experiments are illustrated in Figure S1, available at the Blood website; see the Supplemental Figures link at the top of the online article). (B) Also graphed are ratios of Kit^pos/Kit^neg erythroblast formation for wild-type versus Lyn^-/- progenitors at 48, 72, and 120 hours of culture. (C) Over an extended period of culture, expansion capacities of Lyn^-/- and control Lyn^+/+ cultures also were assayed by direct cell counts. Note the deficient expansion of Lyn^-/- erythroblasts (observed in 3 independent experiments). (D) Levels of Epo- and SCF-induced ³HdT incorporation in expanded Lyn^-/- and control erythroblasts. Expanded Kit^posCD71^high erythroblasts were isolated by lineage depletion and MACS, and cultured for 24 hours in SP34-EX medium in the presence of Epo and/or SCF as indicated. Cultures then were pulsed with ³HdT, and mean incorporation rates (± SE) were determined.

Figure 5.

At a Kit^negCD71^high stage of development, Lyn^-/- erythroid progenitor cells fail to efficiently enter an apparent G1/S cell cycle phase and undergo increased apoptosis. (A) Cultures of expanded Lyn^+/+ and Lyn^-/- erythroblasts were analyzed for Kit and CD71 marker expression, and were costained for DNA content with DRAQ5. DRAQ5 staining distributions are shown for total cell populations (i), for Kit^negCD71^high cells (ii), and for Kit^posCD71^high cells (iii). Note the limited frequencies of Kit^negCD71^high Lyn^-/- erythroblasts in the S-phase peak compared directly with Lyn^+/+ controls (arrows, ii). (B) Increased apoptosis of Lyn^-/- erythroblasts at a Kit^negCD71^high stage. Bone marrow-derived Lyn^-/- and wild-type control erythroblasts were cultured in SP34-EX (with SCF and Epo at nonlimiting concentrations). At 120 hours of culture, cells were coanalyzed for Kit (CD117) and CD71 expression, and for annexin V positivity. Selectively within a Kit^negCD71^high population, Lyn^-/- cells exhibited significantly increased levels of annexin V staining (33.4% vs 14.2%, or 220% over Lyn^+/+ controls). This is illustrated in a representative flow cytometric profile (i) and in histograms as mean values (± SE) for triplicate analyses (ii) (results shown are representative of 2 independent experiments). (C) Also analyzed (by Western blotting) were levels of GATA-1 expression (and GAPDH) in Kit^posCD71^high as well as Kit^negCD71^high Lyn^-/- and Lyn ^+/+ erythroblasts (as isolated by MACS).

Figure 6.

Altered DAPK2 expression and Akt activation in Lyn^-/- erythroblasts. (A) Lyn^-/- and wild-type erythroblasts were expanded from bone marrow preparations in SP34-EX media to yield 90% or more CD71^high populations. Kit^pos erythroblasts were isolated via MACS, washed, deprived of cytokines for 6 hours, and then exposed to SCF plus Epo (150 ng/mL, 5 U/mL) for the indicated intervals. Levels of DAPK2 in Kit^posCD71^high cells then were assayed by Western blotting and quantitatively by densitometry (bottom panels). (B) Lyn^-/- and wild-type erythroblasts were expanded from bone marrow preparations in SP34-EX media, and Kit^posCD71^high erythroblasts were isolated. Cells were then washed and cultured in the absence of SCF and Epo. At the indicated intervals of subsequent exposure to SCF plus Epo (150 ng/mL, 5 U/mL), levels of phospho-Akt were determined. In repeated independent experiments, Kit^negCD71^high erythroblasts also were analyzed (together with Kit^posCD71^high cells) (Figure S4).

Figure 7.

Attenuated Stat5 activation and late-stage-specific decreases in Bcl-xL expression in Lyn^-/- erythroblasts. (A) Stat5 activation in Lyn^-/- erythroblasts is attenuated. Bone marrow-derived progenitor cells from Lyn^-/- and Lyn^+/+ mice were expanded in SP34-EX medium. Kit^negCD71^high and Kit^posCD71^high subpopulations of erythroblasts were isolated (by MACS), and washed cells were incubated in the absence of cytokines. Cells then were exposed to SCF plus Epo (150 ng/mL, 5 U/mL) and lysed at the indicated intervals. Levels of phospho- and total Stat5 then were assayed by Western blotting. In Lyn^-/- erythroblasts, levels of phospho-Stat5 were significantly decreased. In addition, levels of an approximately Mr 130 000 factor (^*), which crossreacted with the phospho-Stat5-specific antibody used in the present studies, also were diminished. (B) Bcl-xL expression is up-modulated in Kit^posCD71^high cells but decreased in Kit^negCD71^high and later-stage erythroblasts. In these erythroblast populations (i-ii), and in expanded Lyn^-/- and Lyn^+/+ erythroblasts that were differentiated for 1.5 days (iii), Bcl-xL levels also were analyzed (together with GAPDH).