Oncogene (2005) 24, 336–343
& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc
Lyn deficiency reduces GATA-1, EKLF and STAT5, and induces extramedullary stress erythropoiesis Evan Ingley1, David J McCarthy1, Jessica R Pore1, Mohinda K Sarna1, Aini S Adenan1, Michael J Wright1, Wendy Erber2, Peta A Tilbrook1 and S Peter Klinken*,1 1
Laboratory for Cancer Medicine, Western Australian Institute for Medical Research and Centre for Medical Research, The University of Western Australia, WA, Australia; 2PathCentre, The Queen Elizabeth II Medical Centre, Nedlands, WA, Australia
In vitro studies have implicated the Lyn tyrosine kinase in erythropoietin signaling. In this study, we show that J2E erythroid cells lacking Lyn have impaired signaling and reduced levels of transcription factors STAT5a, EKLF and GATA-1. Since mice lacking STAT5, EKLF or GATA-1 have red cell abnormalities, this study also examined the erythroid compartment of Lyn/ mice. Significantly, STAT5, EKLF and GATA-1 levels were appreciably lower in Lyn/ erythroblasts, and the phenotype of Lyn/ animals was remarkably similar to GATA-1low animals. Although young adult Lyn-deficient mice had normal hematocrits, older mice developed anemia. Grossly enlarged erythroblasts and florid erythrophagocytosis were detected in the bone marrow of mice lacking Lyn. Markedly elevated erythroid progenitors and precursor levels were observed in the spleens, but not bone marrow, of Lyn/ animals indicating that extramedullary erythropoiesis was occurring. These data indicate that Lyn/ mice display extramedullary stress erythropoiesis to compensate for intrinsic and extrinsic erythroid defects. Oncogene (2005) 24, 336–343. doi:10.1038/sj.onc.1208199 Published online 1 November 2004 Keywords: erythropoiesis; erythropoietin; Lyn; STAT5; GATA-1; EKLF
Introduction Binding of erythropoietin (Epo) to its cognate receptor on the surface of immature erythroid cells initiates a series of intracellular signaling events, including the JAK/STAT, the ras/raf/MAP kinase and PI-3 kinase/ Akt pathways (Carroll et al., 1991; Damen et al., 1993, 1995; Witthuhn et al., 1993; Gouilleux et al., 1995; Wakao et al., 1995; Bao et al., 1999; Lecoq-Lafon et al., 1999). While JAK2 is widely regarded as the primary *Correspondence: SP Klinken, Laboratory for Cancer Medicine, Western Australian Institute for Medical Research and Centre for Medical Research, The University of Western Australia, Level 6, MRF Building, Rear 50 Murray Street, Perth 6000, Western Australia; E-mail:
[email protected] Received 9 July 2004; revised 23 August 2004; accepted 20 September 2004; published online 1 November 2004
kinase involved in Epo signaling (Witthuhn et al., 1993), members of the Src family of tyrosine kinases have also been implicated in intracellular signaling cascades (Tilbrook et al., 1997, 2001; Chin et al., 1998; Kubota et al., 2001). Indeed, Src itself is activated by Epo (Kubota et al., 2001), and is capable of promoting STAT5 phosphorylation (Okutani et al., 2001). We have used the J2E cell line to examine erythroid differentiation in response to Epo in vitro (Klinken et al., 1988; Busfield and Klinken, 1992; Tilbrook et al., 1997, 2001). A mutant subclone (J2E-NR) failed to differentiate upon exposure to Epo (Klinken and Nicola, 1990; Tilbrook et al., 1996a) due to the absence of the Src family member Lyn (Tilbrook et al., 1997). Ectopic expression of Lyn in J2E-NR cells overcame the inability to mature with Epo stimulation, as the cells regained the capacity to synthesize hemoglobin and undergo morphological maturation (Tilbrook et al., 1997). Conversely, introduction of a dominant-negative Lyn into parental J2E cells suppressed Epo-induced differentiation, and reduced the levels of the key erythroid transcription factors GATA-1 and EKLF (Tilbrook et al., 2001). Other studies have confirmed the association of Lyn with the Epo receptor and its participation in downstream signaling, including phosphorylation of STAT5 (Chin et al., 1998; Arai et al., 2001). Lyn is also activated by Epo in CD34 þ cells maturing along the erythroid lineage (Harashima et al., 2002), and can phosphorylate protein Band 3 in erythrocytes (Brunati et al., 2000). The importance of several molecules to erythropoiesis has been confirmed in vivo using genetically modified mice, for example, elimination of the gene for Epo, or the Epo receptor, is lethal as embryos die due to a failure of definitive erythropoiesis (Wu et al., 1995). Similarly, JAK2-deficient mice die in utero with severe anemia (Neubauer et al., 1998; Parganas et al., 1998), while ablation of GATA-1 also results in prenatal death due to a lack of functional red cells (Pevny et al., 1991). Interestingly, mice engineered to express lower levels of GATA-1 (McDevitt et al., 1997) undergo compensatory extramedullary erythropoiesis in an attempt to maintain red cell levels (Vannucchi et al., 2001). EKLF/ mice, on the other hand, produce enucleated erythrocytes with insufficient hemoglobin and die of a lethal b-thalassemia (Nuez et al., 1995; Perkins et al., 1995). Although
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
337
STAT5a/b/ mice do not display gross defects in their erythroid compartment (Teglund et al., 1998), fetal anemia and an expansion of splenic erythropoiesis in adult mice due to decreased survival of early erythroblasts have been reported (Socolovsky et al., 1999; Socolovsky et al., 2001). Lyn-deficient mice have significant B lymphoid abnormalities, which culminate in autoimmune disease (Hibbs et al., 1995; Nishizumi et al., 1995; Chan et al., 1997), consistent with the involvement of Lyn in B-cell receptor signaling (Corey and Anderson, 1999). They also develop splenomegaly and extramedullary hemopoiesis (Harder et al., 2001). Although the number of red blood cells in Lyn/ mice appears normal (Nishizumi et al., 1995), it has been reported that old Lyn-deficient mice have elevated Ter119-positive erythrocytes in their spleens (Satterthwaite et al., 1998). In this study, we examined the expression profile of transcription factors in J2E-NR cells and Lyn/ erythroblasts. Unexpectedly, the levels of STAT5, EKLF and GATA-1 were reduced significantly in both J2E-NR cells and erythroid precursors from Lyndeficient mice. Considering the involvement of Lyn in Epo signaling in vitro (Tilbrook et al., 1997, 2001; Chin et al., 1998; Arai et al., 2001; Harashima et al., 2002), Lyn/ mice were then examined for other alterations to the erythroid compartment in vivo. Our data demonstrate that Lyn-deficient mice display extramedullary red cell production, characteristic of a stress erythropoietic response.
Results Transcription factors are reduced in J2E-NR cells Having observed that a dominant-negative Lyn construct reduced GATA-1 and EKLF levels in J2E cells (Tilbrook et al., 2001), expression of erythroid transcription factors was examined in J2E-NR cells, which fail to express Lyn (Busfield and Klinken, 1992; Tilbrook et al., 1996a). Immunoblotting experiments demonstrated that GATA-1 and EKLF protein content fell significantly in the mutant J2E clone lacking Lyn (Figure 1a). Interestingly, the levels of STAT5a were also lower in J2E-NR cells. The decrease in STAT5a expression in J2E-NR cells prompted a closer investigation of other STAT proteins in these cells. Figure 1b shows that the parental J2E cells, as well as the J2E-NR clone, contained equivalent amounts of STAT1a, STAT1b, STAT5b and STAT6. In contrast with the failure of J2E-NR cells to express STAT5a, the levels of STAT3 were raised 2–3-fold. These data indicate that the presence of Lyn can influence the amount of specific transcription factors within erythroid cells. Numerous studies have shown STAT5 activation upon Epo stimulation (Damen et al., 1995; Gouilleux et al., 1995; Wakao et al., 1995; Gobert et al., 1996; Klingmuller et al., 1996; Iwatsuki et al., 1997; Gregory et al., 1998), while other reports have indicated STAT1
Figure 1 J2E-NR cells have reduced erythroid transcription factor levels. Immunoblotting of J2E and J2E-NR cell lysates. (a) Blots were probed for expression of STAT5a, STAT5b, EKLF, GATA-1 and Lyn. Blots were reprobed for v-raf as a loading control. (b) Blots were also probed for the STAT family members STAT1a, STAT1b, STAT3, STAT5a, STAT5b and STAT6
and STAT3 are phosphorylated in response to Epo treatment (Ohashi et al., 1995; Penta and Sawyer, 1995; Kirito et al., 1998). It is worth noting that several STATs are present in J2E cells (Figure 1b), but only STAT5 isoforms were phosphorylated after Epo stimulation (Figure 2a); although STAT1a, STAT1b, STAT3 and STAT6 were clearly detected in J2E cells, they were not phosphorylated following exposure to Epo (data not shown). Significantly, STAT5 phosphorylation was almost eliminated in Epo-stimulated J2E-NR cells (Figure 2a). Despite abundant STAT5b present in these cells (Figure 1a and b), little phosphorylation was detected in the absence of STAT5a. Significantly, we have previously shown that J2E and J2E-NR cells have no significant difference in either total amount of JAK2 or level of Epo-induced phosporylation of JAK2 (Tilbrook et al., 1996a).
Figure 2 J2E-NR cells have reduced STAT5 phosphorylation and DNA-binding activity. (a) J2E and J2E-NR cells were stimulated with Epo for up to 60 min before cell lysates were prepared and STAT5 immunoprecipitated (IP). The level of tyrosine phosphorylation (p-Tyr) of STAT5 was determined by immunoblotting. Densitometric analysis of the immunoblots is included showing STAT5 tyrosine phosphorylation (arbitrary units) relative to the amount of STAT5 immunoprecipitated. (b) Nuclear extracts from unstimulated and Epo-stimulated J2E and J2E-NR cells were prepared for electrophoretic mobility shift assays. Anti-STAT5a and anti-STAT5b antibodies were used to super shift (ssSTAT5) the specific STAT5-DNA complex Oncogene
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
338
Electrophoretic mobility shift assays were then performed to examine the impact of STAT5 phosphorylation on DNA binding. Figure 2b shows that limited STAT5 association with DNA target elements was observed in J2E-NR cells compared with wild-type J2E cells. Collectively, these observations indicate that the absence of Lyn in J2E-NR cells had a marked effect on STAT5a content, as well as STAT5 phosphorylation and DNA-binding capacity. Reduced transcription factors in Lyn/ erythroid cells To explore whether the phenomenon of reduced transcription factor levels in J2E-NR cells was restricted to immortalized cells, splenic erythroblasts from Lyn/ mice were isolated and examined for STAT, EKLF and GATA-1 content. Phenylhydrazine was used to induce a surge of erythropoiesis within the spleen, resulting in more than 80% of the cells being Ter119-positive erythroblasts. Strikingly, decreased levels of GATA-1, EKLF, STAT5a and STAT5b were detected in these Lyn-deficient splenic erythroblasts (Figure 3a); the decrease in these transcription factors varied between 20 and 80% in different animals. These data supported the observations that STAT5a, EKLF and GATA-1 were noticeably lower in J2E-NR cells (Figure 1a). Moreover, these results showed that the absence of Lyn affects STAT5b content in primary erythroblasts. Like J2E-NR cells, Lyn/ erythroblasts retained STAT1, STAT3 and STAT6 (data not shown). The phosphorylation status of STAT5 was then examined in Lyn/ erythroblasts. The data presented in Figure 3b indicate that, in addition to lower STAT5 levels, phosphorylation of STAT5 after Epo stimulation was attenuated in Lyn-deficient cells. Thus, Lyn has two effects on STAT5 in erythroid cells viz. it affects the
Figure 3 Lyn-deficient splenic erythroblasts have reduced erythroid transcription factor levels. (a) Cell lysates were prepared from the spleens of Lyn þ / þ and Lyn/ mice 3 days after phenylhydrazine injection, and immunoblotted with antibodies against STAT5a, STAT5b, EKLF, GATA-1 and Lyn. Blots were reprobed for MAP kinase as a loading control. (b) Cell lysates were prepared as in (a) up to 120 min after Epo stimulation and immunoprecipitated (IP) with anti-STAT5 antibodies. Fivefold more lysate was used for the immunoprecipitation of STAT5 from Lyn/ cells to compensate for reduced STAT5 expression. The level of tyrosine phosphorylation (p-Tyr) of STAT5 was determined by immunoblotting. Densitometric analysis of the immunoblots is included showing STAT5 tyrosine phosphorylation (arbitrary units) relative to the amount of STAT5 immunoprecipitated Oncogene
amount of STAT5 protein and influences the degree of STAT5 phosphorylation. Interestingly, there is no significant difference in the total amount or level of Epo-induced tyrosine phosphorylation of JAK2 between Lyn/ and Lyn þ / þ splenic erythroblasts (data not shown). Lyn/ mice display erythroid abnormalities GATA-1/, GATA-1low, EKLF/ and STAT5a/b/ mice display various forms of deficiencies within the erythroid compartment (Pevny et al., 1991; Nuez et al., 1995; Perkins et al., 1995; Socolovsky et al., 1999, 2001; Vannucchi et al., 2001). Since Lyn/ mice had reduced levels of GATA-1, EKLF and STAT5 (Figure 3), Lyndeficient animals were then examined for erythroid defects. Initially, peripheral red blood cells from young adult mice were investigated, but no differences in hematocrits (Figure 4a), red cell numbers, hemoglobin content or red cell volume (data not shown) were observed between knockout and wild-type animals. However, there was a significant increase in the percentage of circulating reticulocytes detected in Lyn/ mice (Figure 4b). Interestingly, as the mice aged their hematocrits fell and the animals became anemic (Figure 4a). These observations suggested that a perturbation of the erythroid compartment was occurring in these animals. A close examination of bone marrow from young adult Lyn-deficient mice was undertaken to determine if this was the site of defective erythropoiesis. However, colony assays revealed that the erythroid burst-forming units (BFU-E) and colony forming units (CFU-E)were normal, and that the number of Ter119-positive cells was unaltered (data not shown). Significantly, dysmorphic erythroblasts were observed in the bone marrow of Lyn/ mice (Figure 5a). These cells were grossly enlarged compared with normal erythroid precursors and were reminiscent of megaloblasts. Surprisingly, florid erythrophagocytosis was detected in the Lyn-deficient bone marrow with numerous monocytes and macrophages engulfing enucleate erythroid
Figure 4 Lyn-deficient mice develop anemia and reticulocytosis. (a) Hematocrits (Hct) were determined from wild-type and knockout animals at 20 and 60 weeks of age (n ¼ 13). (b) Levels of circulating reticulocytes (Retic) were detected in Lyn þ / þ and Lyn/ mice (n ¼ 9). The mean7s.d. is presented. Statistically significant (two-way ANOVA) values are indicated by *Pp0.05 and **Pp0.01
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
339
Figure 5 Erythroid abnormalities in Lyn-deficient bone marrow. (a) Wright’s Giemsa-stained slides of cytocentrifuged bone marrow cells from Lyn þ / þ and Lyn/ mice. Open arrows show grossly enlarged erythroid cells, while closed arrows show normal differentiating erythroblasts. (b) Wright’s Giemsa-stained slides of cytocentrifuged bone marrow cells from Lyn/ mice. Arrows show erythrophagocytosis by macrophages and monocytes
Figure 7 Lyn-deficient spleens contain high levels of erythroid progenitors. (a) Flow cytometric analysis of spleen cells from Lyn þ / þ and Lyn/ mice. Cells were stained with anti-Ter119-phycoerythrin and anti-CD117-fluorescein isothiocyanate (c-kit) antibodies (n ¼ 9). (b) Methylcellulose colony assays for BFU-E and CFU-E from spleens of Lyn þ / þ and Lyn/ mice. Numbers represent CFUE per 2 104 spleen cells plated and BFU-E per 2 105 spleen cells. The mean7s.d. is presented (n ¼ 3). Statistically significant (twoway ANOVA) values are indicated by *Pp0.05 and **Pp0.01
Figure 6 Lyn-deficient spleens contain elevated erythroblasts. (a) Wright’s Giemsa-stained slides of cytocentrifuged spleen cells from Lyn þ / þ and Lyn/ mice. Open arrows show erythroblasts at different stages of differentiation. (b) Neutral benzidine/Wright’s Giemsa-stained slide of cytocentrifuged spleen cells from Lyn/ mice. Hemoglobin stains yellow/brown. (c) Proliferation of splenic erythroblasts from Lyn/ and Lyn þ / þ mice in response to different doses of Epo. The mean7s.d. is presented
cells (Figure 5b). Thus, morphological and physiological abnormalities involving the erythroid lineage were detected in Lyn/ bone marrow. Splenic erythropoiesis in Lyn/ mice As stress erythropoiesis occurs in the spleen (Bauer et al., 1999; Socolovsky et al., 2001), the spleens of Lyn/ mice
were examined next. Unlike the bone marrow, a marked increase in erythroblasts at various stages of maturation was seen in the spleens of Lyn-deficient animals (Figure 6a). It is worth noting that a proportion of abnormally large erythroblasts were also detected in the spleen. However, the immature splenic erythroid cells were able to synthesize hemoglobin, as they stained positive with neutral benzidine (Figure 6b). In addition, the erythroblasts displayed a normal dose–response curve in response to Epo stimulation (Figure 6c). The number of erythroid cells found in the spleens of young adult Lyn/ mice was then quantitated by flow cytometry. A significant 3–4-fold rise in total Ter119positive erythroid cells was observed in splenic preparations from Lyn-deficient mice (Figure 7a). Owing to the increased number of cells in Lyn/ spleens, this equates to a 4–5-fold increase in total Ter119-positive cells per Oncogene
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
340
mouse spleen. Similarly, a fourfold elevation in less mature c-kit/Ter119 double-positive erythroid cells per spleen was detected. Thus, an expansion of the erythroid compartment contributes to the splenomegaly that occurs in Lyn-deficient mice. Colony assays were also performed to determine whether the levels of erythroid progenitors within the spleen were altered in Lyn/ mice. Figure 7b shows that BFU-E and CFU-E levels were markedly elevated in the spleens of Lyn-deficient mice. In contrast with the few BFU-E or CFU-E detected in wild-type spleens, large numbers of progenitors were present in Lyn/ spleens. This extramedullary red cell production in the spleen indicates that the animals were undergoing persistent stress erythropoiesis, comparable with GATA-1low and STAT5a/b/ mice (Socolovsky et al., 2001; Vannucchi et al., 2001).
both wild-type and knockout mice (Figure 8a); notably, however, the hematocrits and circulating reticulocyte levels of Lyn-deficient mice recovered faster than their wild-type counterparts (Figure 8a,b). Moreover, splenic BFU-E and CFU-E numbers were raised significantly in knockout animals (Figure 8c). Like GATA-1low mice, the state of constant stress erythropoiesis experienced by Lyn/ mice probably explains the reduced reaction time to phenylhydrazine-induced anemia. Lyn-deficient mice were also exposed to repeated injections of exogenous Epo. Figure 8d shows the rapid reticulocyte response by Lyn/ mice, prior to the rise in hematocrits, which peaked at 67% on day 7. These responses to the physiological stresses imposed upon Lyn-deficient mice are remarkably similar to the data obtained with GATA-1low mice (Vannucchi et al., 2001).
Lyn/ mice respond faster to phenylhydrazine
Discussion
low
/
Both GATA-1 and STAT5a/b mice display altered responses to phenylhydrazine treatment and exogenous Epo challenge (Socolovsky et al., 1999; Vannucchi et al., 2001). Therefore, Lyn/ mice were injected with phenylhydrazine, and a rapid anemia was observed in
Figure 8 Lyn-deficient mice have an accelerated response to erythropoietic stress. (a) Mice were treated with phenylhydrazine (Phz) on consecutive days as indicated (arrows). Hematocrits (Hct) were determined over 8 days postphenylhydrazine (n ¼ 9). (b) Blood samples were also analysed for circulating reticulocytes (Retic). (c) Methylcellulose colony assays for BFU-E and CFU-E from spleens of Lyn þ / þ and Lyn/ mice injected with phenylhydrazine as in (a) and analysed on day 1 (post-treatment). Numbers represent CFU-E per 2 104 plated spleen cells and BFU-E per 2 105 spleen cells (n ¼ 3). (d) Mice were treated with Epo (10 U/ mouse/day, first day of injection is designated day 0) for 5 consecutive days as indicated (arrows), and reticulocytes (Retic) determined over 9 days. The mean7s.d. is presented. Statistically significant (two-way ANOVA) values are indicated by *Pp0.05 and **Pp0.01 Oncogene
Previous studies have shown that J2E-NR cells, which lack Lyn, have impaired intracellular signaling in response to Epo (Tilbrook et al., 1996a, 1997). In this study, we demonstrate that the absence of Lyn in this mutant J2E clone results in markedly reduced levels of transcription factors GATA-1, EKLF and STAT5a. Similarly, the concentration of these transcription factors was much lower in Lyn/ erythroblasts. These data suggest that in erythroid cells Lyn plays an important role in regulating transcription factor levels. This function appears to be supplementary to Lyn’s role of initiating intracellular signaling cascades in response to Epo (Tilbrook et al., 1997, 2001; Corey and Anderson, 1999). The observation that Lyn can affect transcription factor content is consistent with the observation that introduction of a dominant-negative Lyn into J2E cells suppressed GATA-1 and EKLF levels (Tilbrook et al., 2001). Conversely, overexpression of Lyn in the immature R11 erythroid cell line enhanced GATA-1 and EKLF levels appreciably, and promoted spontaneous differentiation along the erythroid pathway (Sarna et al., 2003). Thus, alteration of Lyn activity within erythroid precursors influences key erythroid transcription factor levels, independent of Epo stimulation. The concentration of GATA-1 within hemopoietic progenitor cells plays a significant role in directing lineage fate (Kulessa et al., 1995; Yamaguchi et al., 1998; Heyworth et al., 2002). This transcription factor plays a crucial role in the viability, proliferation and differentiation of red blood cells (Tsai et al., 1989; Pevny et al., 1991; Weiss et al., 1994; Rylski et al., 2003), and the levels of GATA-1 have a major effect on erythroid cells – reducing GATA-1 content in GATA-1low mice perturbs the erythroid compartment (McDevitt et al., 1997; Vannucchi et al., 2001), while overexpression of GATA-1 in transgenic mice impedes differentiation and results in a fatal anemia (Whyatt et al., 2000). STAT5 is also important for erythroid precursor cells as it
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
341
activates the antiapoptotic molecule Bcl-XL (Socolovsky et al., 1999), while EKLF is an erythroid-specific transcription factor that affects globin and heme enzyme genes for hemoglobin synthesis (Miller and Bieker, 1993; Nuez et al., 1995; Perkins et al., 1995; Spadaccini et al., 1998; Coghill et al., 2001). It is significant, therefore, that the GATA-1, EKLF and STAT5 content were reduced in Lyn/ erythroid cells, and that an impairment of erythropoiesis was observed in Lyn-deficient mice. A notable feature of this study was the parallel between the erythroid phenotypes displayed by Lyn/ and GATA-1low mice (Vannucchi et al., 2001). The decrease in GATA-1 protein in Lyn-deficient mice is comparable with the reduction in GATA-1low mice (McDevitt et al., 1997), which may explain the similar phenotypic alterations, that is, both mice strains display extramedullary stress erythropoiesis, indicating a compensatory mechanism to overcome erythropoietic defects. Like GATA-1low animals (Vannucchi et al., 2001), large numbers of erythroid progenitors and precursors were detected in the spleens of Lyn/ mice. The altered responses of Lyn/ mice to phenylhydrazine and exogenous Epo were also similar to GATA-1low animals. These data indicate that the reduction of GATA-1 levels in Lyn/ mice is likely to contribute to the red cell problems encountered by Lyn-deficient animals. The decrease in STAT5 protein in Lyn/ mice may also add to the difficulties in erythropoiesis, since STAT5a/b/ mice develop erythroid abnormalities (Socolovsky et al., 1999, 2001). Moreover, the reduction in Epo-stimulated phosphorylation of STAT5 observed in Lyn-deficient erythroblasts supports the proposition that Lyn plays a role in STAT5 phosphorylation (Chin et al., 1998). Taken together, the decrease in GATA-1 and STAT5 strongly indicate that an intrinsic defect exists within Lyn/ erythroid cells. The erythrophagocytosis observed in Lyn-deficient animals suggests an additional, extrinsic defect in erythropoiesis. As histiocytosis and histiocytomas are known to induce erythrophagocytosis, it is significant that Lyn/ mice accumulate atypical myelomonocytic cells and develop monocyte/macrophage malignancies (Harder et al., 2001), which could account for the abnormal engulfment of erythrocytes in these animals. It is also conceivable that as autoimmune disease emerges in Lyn/ mice, they may develop an autoimmune hemolytic anemia, which is typically characterized by reticulocytosis, erythrophagocytosis and extramedullary erythropoiesis. It is likely, therefore, that the combination of intrinsic and extrinsic erythroid defects generates compensatory extramedullary erythropoiesis, which ultimately leads to anemia as the animals age. An interesting observation to emerge from this study is that numerous STAT proteins were present in the J2E erythroid line; however, only STAT5 was phosphorylated and activated following Epo induction. This result is compatible with the recent observations that Epo only activates STAT5 in primary erythroblasts (Oda et al., 1998). As STAT3 has been implicated as a negative regulator of red cell maturation (Kirito et al., 1998), it is
tempting to speculate that different STATs are activated at discrete stages to modulate erythroid differentiation. In summary, the data presented in this manuscript reveal that Lyn affects erythroid transcription factor levels and that an underlying perturbation of red cell production occurs in Lyn/ mice. These observations show that the effects of Lyn are not restricted to in vitro models of erythroid maturation, and that this tyrosine kinase does indeed play a crucial role in erythropoiesis in vivo.
Materials and methods Mice Lyn/ and Lyn þ / þ mice (129/BL6 background) (Hibbs et al., 1995) were generally analysed between 10 and 12 weeks of age. Anemia and polycythemia were induced by treatment with either phenylhydrazine (60 mg/kg body weight) (Klinken et al., 1987) or rhEpo (10 U/mouse) (Vannucchi et al., 2001), respectively. Cell culture Hemopoietic parameters Heparinized microcapillary tubes were used for hematocrit determination, and reticulocyte content enumerated after methylene blue staining. Bone marrow and splenic cell morphology were examined microscopically, following cytocentrifugation and Wright’s Giemsa staining with, or without, neutral benzidine staining (McLeod et al., 1974). Flow cytometry was employed to assess cell surface expression of Ter119 (phycoerythrin-conjugated antimouse Ter119, BD Biosciences) and c-kit (fluorescein isothiocyanate-conjugated anti-mouse CD117, BD Biosciences), as detailed previously (Sarna et al., 2003) using an Epics XL/ MCL flow cytometer (Beckman-Coulter, Palo Alto, CA, USA). BFU-E and CFU-E were determined using methylcellulose cultures, as described elsewhere (Tilbrook et al., 2001). Cell line parameters J2E cells are derived from murine fetal liver erythroid cells immortalized with the J2 retrovirus expressing v-raf and v-myc (Klinken et al., 1988). J2E-NR cells are an Epo nonresponsive subclone of J2E cells generated spontaneously (Klinken and Nicola, 1990) and lack expression of Lyn (Tilbrook et al., 1997). J2E and J2E-NR cells were grown in Dulbecco’s modified Eagle’s medium/5% fetal bovine serum, and differentiation was initiated with Epo (5 U/ml) (Klinken et al., 1988; Busfield and Klinken, 1992). Viability was determined by eosin dye exclusion (Tilbrook et al., 1996b) and hemoglobin synthesis by benzidine staining (Klinken et al., 1988; Busfield and Klinken, 1992). Proliferation was assayed by [3H]thymidine incorporation as described previously (Busfield and Klinken, 1992) using a modification of the method established by Krystal (1983). Immunoprecipitation and immunoblotting Cells were analysed for immunoprecipitation and immunoblotting, essentially as reported previously (Ingley et al., 2000, 2001) by lysis in 20 mM Tris-HCl (pH 8.0), 120 mM NaCl, 1.0% Nonident P-40, 10 mM b-glycerophosphate, 10 mM NaF, 1 mM Na3VO4, 1 mM EDTA, 1 mM EGTA, 10 mM benzamidine, 1 mM PMSF, 10 mg/ml aprotinin. For co-immunoprecipitations, clarified cell lysates were incubated with antibodies to Lyn (SC-15G, SC-15, Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) or STAT5a/STAT5b (Santa Cruz BiotechnoOncogene
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
342 logy Inc., SC-1081, SC-835) for 2 h at 41C, then collected with protein G–Sepharose beads for 16 h before washing and analysis by immunoblotting. Additional antibodies used for immunoblotting were anti-MAP kinase, anti-v-raf, anti-phosphotyrosine and anti-GATA-1 (Santa Cruz Biotechnology Inc., SC-154, SC-133, SC-7020, SC-265), anti-STAT1, anti-STAT3 and anti-STAT6 (Transduction Laboratories, Lexington, KY, USA). Anti-EKLF antibodies were generously provided by Dr Jim Bieker (Mount Sinai School of Medicine, NY, USA). Secondary antibodies were coupled to horseradish peroxidase (Amersham Biosciences, UK) and detected by enhanced chemiluminescence (Amersham Biosciences). Electrophoretic mobility shift assays Nuclear extracts were prepared according to the method of Ramsay et al. (1992) from equal numbers of J2E and J2E-NR
cells, and DNA-binding reactions were conducted as we have described previously (Spadaccini et al., 1998). Briefly, 5 mg of nuclear extract was incubated with a 32P-end-labeled oligonucleotide (50 -AGA TTT CTA GGA ATT CAA ATC-30 ) corresponding to the STAT5 motif in the b-casein promoter (Oda et al., 1998). The STAT5 complex was identified using specific supershifting anti-STAT5 antibodies as described (Oda et al., 1998). Acknowledgements We thank Dr Margaret Hibbs for making the Lyn/ mice available and Shane Meakins for excellent technical assistance. This work was supported by grants from National Health and Medical Research Council (Grant 139008, 303101), the Cancer Foundation of Western Australia and the Medical Research Foundation of Royal Perth Hospital.
References Arai A, Kanda E, Nosaka Y, Miyasaka N and Miura O. (2001). J Biol Chem, 276, 33282–33290. Bao HF, Jacobs-Helber SM, Lawson AE, Penta K, Wickrema A and Sawyer ST. (1999). Blood, 93, 3757–3773. Bauer A, Tronche F, Wessely O, Kellendonk C, Reichardt HM, Steinlein P, Schutz G and Beug H. (1999). Genes Dev., 13, 2996–3002. Brunati AM, Bordin L, Clari G, James P, Quadroni M, Baritono E, Pinna LA and Donella-Deana A. (2000). Blood, 96, 1550–1557. Busfield SJ and Klinken SP. (1992). Blood, 80, 412–419. Carroll MP, Spivak JL, McMahon M, Weich N, Rapp UR and May WS. (1991). J. Biol. Chem., 266, 14964–14969. Chan VW, Meng F, Soriano P, DeFranco AL and Lowell CA. (1997). Immunity, 7, 69–81. Chin H, Arai A, Wakao H, Kamiyama R, Miyasaka N and Miura O. (1998). Blood, 91, 3734–3745. Coghill E, Eccleston S, Fox V, Cerruti L, Brown C, Cunningham J, Jane S and Perkins A. (2001). Blood, 97, 1861–1868. Corey SJ and Anderson SM. (1999). Blood, 93, 1–14. Damen JE, Mui AL-F, Puil L, Pawson T and Krystal G. (1993). Blood, 81, 3204–3210. Damen JE, Wakao H, Miyajima A, Krosl J, Humphries RK, Cutler RL and Krystal G. (1995). EMBO J., 14, 5557–5568. Gobert S, Chretien S, Gouileux F, Muller O, Pallard C, Dusanter-Fourt I, Groner B, Lacombe C, Gisselbrecht S and Mayeux P. (1996). EMBO J, 15, 2434–2441. Gouilleux F, Pollard C, Dusanter-Fourt I, Wakao H, Haldosen L-A, Norstedt G, Levy D and Groner B. (1995). EMBO J., 14, 2005–2013. Gregory RC, Jiang N, Todokoro K, Crouse J, Pacifici RE and Wojchowski DM. (1998). Blood, 92, 1104–1118. Harashima A, Suzuki M, Okochi A, Yamamoto M, Matsuo Y, Motoda R, Yoshioka T and Orita K. (2002). Blood, 100, 4440–4445. Harder KW, Parsons LM, Armes J, Evans N, Kountouri N, Clark R, Quilici C, Grail D, Hodgson GS, Dunn AR and Hibbs ML. (2001). Immunity, 15, 603–615. Heyworth C, Pearson S, May G and Enver T. (2002). EMBO J., 21, 3770–3781. Hibbs ML, Tarlinton DM, Armes J, Grail D, Hodgson G, Maglitto R, Stacker SA and Dunn AR. (1995). Cell, 83, 301–311. Ingley E, Chappell D, Poon SY, Sarna MK, Beaumont JG, Williams JH, Stillitano JP, Tsai S, Leedman PJ, Tilbrook PA and Klinken SP. (2001). J. Biol. Chem., 276, 43428–43434. Oncogene
Ingley E, Sarna MK, Beaumont JG, Tilbrook PA, Tsai S, Takemoto Y, Williams JH and Klinken SP. (2000). J. Biol. Chem., 275, 7887–7893. Iwatsuki K, Endo T, Misawa H, Yokouchi M, Matsumoto A, Ohtsubo M, Mori KJ and Yoshimura A. (1997). J. Biol. Chem., 272, 8149–8152. Kirito K, Uchida M, Takatoku M, Nakajima K, Hirano T, Miura Y and Komatsu N. (1998). Blood, 92, 462–471. Klingmuller U, Bergelson S, Hsiao JG and Lodish HF. (1996). Proc. Natl. Acad. Sci. USA, 93, 8324–8328. Klinken SP, Holmes K, Fredrickson TN, Erner SM and Morse HC. (1987). J. Immunol., 139, 3091–3098. Klinken SP and Nicola NA. (1990). Leukemia, 4, 24–28. Klinken SP, Nicola NA and Johnson GR. (1988). Proc. Natl. Acad. Sci. USA, 85, 8506–8510. Krystal G. (1983). Exp. Hematol., 11, 649–660. Kubota Y, Tanaka T, Kitanaka A, Ohnishi H, Okutani Y, Waki M, Ishida T and Kamano H. (2001). EMBO J., 20, 5666–5677. Kulessa H, Frampton J and Graf T. (1995). Gene Dev., 9, 1250–1262. Lecoq-Lafon C, Verdier F, Fichelson S, Chretien S, Gisselbrecht S, Lacombe C and Mayeux P. (1999). Blood, 93, 2578–2585. McDevitt MA, Shivdasani RA, Fujiwara Y, Yang H and Orkin SH. (1997). Proc. Natl. Acad. Sci. USA, 94, 6781–6785. McLeod DL, Shreeve MM and Axelrad AA. (1974). Blood, 44, 5176–5534. Miller IJ and Bieker JJ. (1993). Mol. Cell. Biol., 13, 2776–2786. Neubauer H, Cumano A, Muller M, Wu H, Huffstadt U and Pfeffer K. (1998). Cell, 93, 397–409. Nishizumi H, Taniuchi I, Yamanashi Y, Kitamura D, Ilic D, Mori S, Watanabe T and Yamamoto T. (1995). Immunity, 3, 549–560. Nuez B, Michalovich D, Bygrave A, Ploemacher R and Grosveld F. (1995). Nature, 375, 316–318. Oda A, Sawada K, Druker BJ, Ozaki K, Takano H, Koizumi K, Fukada Y, Handa M, Koike T and Ikeda Y. (1998). Blood, 92, 443–451. Ohashi T, Masuda M and Ruscetti SK. (1995). Blood, 85, 1454–1462. Okutani Y, Kitanaka A, Tanaka T, Kamano H, Ohnishi H, Kubota Y, Ishida T and Takahara J. (2001). Oncogene, 20, 6643–6650. Parganas E, Wang D, Stravopodis D, Topham DJ, Marine JC, Teglund S, Vanin EF, Bodner S, Colamonici OR,
Extramedullary erythropoiesis in Lyn/ mice E Ingley et al
343 van Deursen JM, Grosveld G and Ihle JN. (1998). Cell, 93, 385–395. Penta K and Sawyer ST. (1995). J. Biol. Chem., 270, 31282–31287. Perkins AC, Sharpe AH and Orkin SH. (1995). Nature, 375, 318–322. Pevny L, Simon MC, Robertson E, Klein WH, Tsai SF, D’Agati V, Orkin SH and Costantini F. (1991). Nature, 349, 257–260. Ramsay RG, Ishii S and Gonda TJ. (1992). J. Biol. Chem., 267, 5656–5662. Rylski M, Welch JJ, Chen YY, Letting DL, Diehl JA, Chodosh LA, Blobel GA and Weiss MJ. (2003). Mol. Cell. Biol., 23, 5031–5042. Sarna MK, Ingley E, Busfield SJ, Cull VS, Lepere W, McCarthy DJ, Wright MJ, Palmer GA, Chappell D, Sayer MS, Alexander WS, Hilton DJ, Starr R, Watowich SS, Bittorf T, Klinken SP and Tilbrook PA. (2003). Oncogene, 22, 3221–3230. Satterthwaite AB, Lowell CA, Khan WN, Sideras P, Alt FW and Witte ON. (1998). J. Exp. Med., 188, 833–844. Socolovsky M, Fallon AE, Wang S, Brugnara C and Lodish HF. (1999). Cell, 98, 181–191. Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C and Lodish HF. (2001). Blood, 98, 3261–3273. Spadaccini A, Tilbrook PA, Sarna MK, Crossley M, Bieker JJ and Klinken SP. (1998). J. Biol. Chem., 273, 23793–23798. Teglund S, McKay C, Schuetz E, van Deursen JM, Stravopodis D, Wang D, Brown M, Bodner S, Grosveld G and Ihle JN. (1998). Cell, 93, 841–850.
Tilbrook PA, Bittorf T, Busfield SJ, Chappell D and Klinken SP. (1996a). J. Biol. Chem., 271, 3453–3459. Tilbrook PA, Bittorf T, Callus B, Busfield SJ, Ingley E and Klinken SP. (1996b). Cell Growth Differ., 28, 175–181. Tilbrook PA, Ingley E, Williams JH, Hibbs ML and Klinken SP. (1997). EMBO J., 16, 1610–1619. Tilbrook PA, Palmer GA, Bittorf T, McCarthy DJ, Wright MJ, Sarna MK, Linnekin D, Cull VS, Williams JH, Ingley E, Schneider-Mergener J, Krystal G and Klinken SP. (2001). Cancer Res., 61, 2453–2458. Tsai S, Martin DIK, Zon LI, D’Andrea AD, Wong GG and Orkin SH. (1989). Nature, 339, 446–451. Vannucchi AM, Bianchi L, Cellai C, Paoletti F, Carrai V, Calzolari A, Centurione L, Lorenzini R, Carta C, Alfani E, Sanchez M, Migliaccio G and Migliaccio AR. (2001). Blood, 97, 3040–3050. Wakao H, Harada N, Kitamura T, Mui AL and Miyajima A. (1995). EMBO J., 14, 2527–2535. Weiss MJ, Keller G and Orkin SH. (1994). Genes Dev., 8, 1184–1197. Whyatt D, Lindeboom F, Karis A, Ferreira R, Milot E, Hendriks R, de Bruijn M, Langeveld A, Gribnau J, Grosveld F and Philipsen S. (2000). Nature, 406, 519–524. Witthuhn BA, Quelle FW, Silvennoinen O, Yi T, Tang B, Miura O and Ihle JN. (1993). Cell, 74, 227–236. Wu H, Lui X, Jaenisch R and Lodish HF. (1995). Cell, 83, 59–67. Yamaguchi Y, Zon LI, Ackerman SJ, Yamamoto M and Suda T. (1998). Blood, 91, 450–457.
Oncogene
