Fli-1 overexpression in erythroleukemic cells promotes erythroid de-differentiation while Spi-1/PU.1 exerts the opposite effect

Abstract

The ETS transcription factors play a critical role during hematopoiesis. In F-MuLV-induced erythroleukemia, Fli‑1 insertional activation producing high expression of this transcription factor required to promote proliferation. How deregulated Fli‑1 expression alters the balance between erythroid differentiation and proliferation is unknown. To address this issue, we exogenously overexpressed Fli‑1 in an erythroleukemic cell harboring activation of spi‑1/PU.1, another ETS gene involved in erythroleukemogenesis. While the proliferation in culture remains unaffected, Fli‑1 overexpression imparts morphological and immunohistochemical characteristics of immature erythroid progenitors. Fli‑1 overexpression in erythroleukemic cells increased the numbers of erythroid colonies on methylcellulose and reduced tumorigenicity as evidenced by increase latency of erythroleukemogenesis in mice inoculated with these cells. Although all transplanted mice developed enlargement of the spleen and liver due to leukemic infiltration, Fli‑1 overexpression altered the hematopoietic phenotype, significantly increasing the expression of regulatory hematopoietic genes cKIT, SCA-1, CD41 and CD71. In contrast, expression of Spi‑1/PU.1 in a Fli‑1 producing erythroleukemia cell line in which fli‑1 is activated, resulted in increased proliferation through activation of growth promoting proteins MAPK, AKT, cMYC and JAK2. Importantly, these progenitors express high levels of markers such as CD71 and TER119 associated with more mature erythroid cells. Thus, Fli‑1 overexpression induces a de-differentiation program by reverting CFU-E to BFU-E erythroid progenitor activity, while Spi‑1/PU.1 promoting maturation from BFU-E to CFU-E.

Figure 1

Exogenous expression of Fli-1 in SFFV-induced erythroleukemic cell line DP17-17. (A) Fli-1 and Spi-1/PU.1 expression levels in HB60-5, DP16-1, DP17-17, CB7, CB3 and HB22.2 cell lines. (B) Exogenous expression of Fli-1 in DP17-17 cells as detected by western blot analysis. β-actin was used as a loading control. (C) Unchanged proliferation rate of transduced DP17-17 cells as determined by trypan-blue dye exclusion. (D) Gross morphological examination revealed that Fli-1 overexpression increases the proportion of the non-adherent cell population under normal culture conditions, compared to the MigR1 empty vector transduced cells.

Figure 2

Exogenous expression of Fli-1 alters morphology of transduced DP17-17 cells. (A) May-Grunwald Giemsa stained cytospin preparations of double-sorted DP17-17 cells transduced with the MigR1 and MigR1-Fli-1 vectors. (B) Erythroid differential counts from stained cytospin preparations of transduced DP17-17 cells were performed in a blinded manner by a clinical hematopathologist. The bar graph represents percentages of total cells analyzed. R1 represents the proerythroblast; R2 represents the early basophilic erythroblast, and R3 the late basophilic erythroblast stage (42). (C and D) Stained cytospin preparations (C) and flow cytometric analysis (D) of untreated and DMSO-treated DP17-17, DP17-17 MigR1 and DP17-17 Fli-1.

Figure 3

DP17-17 cells overexpressing Fli-1 display increased numbers of erythroid colonies. (A) Western blot analysis of Fli-1 expression after treatment of the indicated cells for 3 day with 2% DMSO. Densitometry was used to determine the level of Fli-1 downregulation by DMSO. (B) Transduced DP17-17 cells were plated in triplicate on methylcellulose media in the presence of cytokines. Erythroid colonies were quantified by staining with benzidine following 12 days of culture. Representative data indicating benzidine positive staining of methylcellulose cultures on day 12. (C) Analysis of individual methylcellulose colonies. (D) Cellular cytospins and May-Grunwald-Giemsa stains of sampled colonies from transduced DP17-17 cells display characteristic morphologies of mid to late stage erythroblasts.

Figure 4

DP17-17 cells overexpressing Fli-1 express markers of megakaryocytic cells. (A) Flow cytometry analysis demonstrated a higher expression of CD41 and cKIT expression in DP17-17 Fli-1 than the control DP17-17 vector cells. (B) qRT-PCR analysis for expression of GATA1 (B), SCL/TAL1 (C) and RUNX1 (D) in DP17-17 Fli-1 and DP17-17-vector cells. β-actin waas used as loading control. ^*P<0.05; ^**P<0.005.

Figure 5

Increased disease latency in mice inoculated with Fli-1 overexpressing DP17-17 erythroleukemic cells. (A) The mean survival rate of eight-week-old DBA/2J mice i.v injected with 1×10⁶ (n=6), 1×10⁵ (n=6), or 1×10⁴ (n=6) DP17-17 cells transduced with MigR1 or MigR1-Fli-1 retroviruses. ^*P<0.05. (B) Frequency of hematopoietic cell surface marker expression, as a percentage (% ± standard deviation) of total GFP-positive cells isolated from the spleens and livers of CB3 (10⁶) recipient DBA/2J mice, where Fli-1; n=7 and MigR1 control; n=9.

Figure 6

CB3 cells transduced with exogenous Spi1/PU.1 express markers of more mature erythroid progenitors. (A) Expression of Spi-1/PU.1 in CB3 cells accelerates the growth of these cells in culture when compared to CB3-vector cells. (B) Expression of the indicated protein in untreated CB3 (N/T), CB3-vector, CB3-Spi-1/PU.1 and DP17-17 cells. β-actin is used as loading control. (C) May-Grunwald Giemsa stained cytospin preparations of CB3-Spi-1 and CB3-vector cells transduced with the MSCV-Spi-1 and empty vector plasmids. (D) Flow cytometric analysis of CB3-Spi-1 and CB3-vector cells using the indicated antibodies. (E) A proposed model of erythroid de-differentiation and differentiation by Fli-1 and Spi-1/PU.1, respectively. In this model, expression of Fli-1 in DP17-17 cells (CFU-E like progenitors) induces a de-differentiation program resulting in generation of cells resembling BFU-E progenitors. In contrast, expression of Spi-1/PU.1 in CB3 cells (BFU-E like progenitors) promotes differentiation to cells resembling CFU-E like progenitors.