Overexpression of Stat3α, but Not Statβ, in Myeloid Cells Results in Neutrophil Expansion through Reduced Apoptosis and Modulation of a Unique Set of Apoptosis Pathway Genes.

Normal neutrophil development requires G-CSF signaling, which includes activation of signal transducer and activator of transcription (STAT) 3. The role of Stat3 in myelopoiesis is controversial--studies in murine myeloid cell lines indicate that Stat3 inhibits apoptosis and promotes myeloid differentiation, while results from transgenic mouse models are conflicting. One possible explanation for this confusion is that the two Stat3 isoforms co-expressed in myeloid cells, α (p92) and β (p83), which are derived from a single gene by alternative mRNA splicing, have distinct biological functions. Stat3α contains a 55 residue C-terminal transactivation domain (TAD) and is transcriptionally active while Stat3β lacks the TAD, is transcriptionally inactive and contains 7 unique C-terminal residues that encode prolonged (>3 hr) nuclear retention. The effects of altered expression of only one isoform in myeloid cells lines or within the myeloid lineage of mice have not been determined. We conditionally overexpressed either Stat3α or Stat3β in the murine myeloid cell line 32Dcl3, using a two-vector system--one encoding a transactivator that requires glucocorticoid and a tetracycline derivatives for nuclear translocation and DNA binding, respectively, and a second vector encoding Stat3α or Stat3β downstream of a tetracycline response element. Two clones showing overexpression of each isoform following exposure to dexamethasone (100 nM) and doxycycline (1 μg/ml) (dex/dox) were expanded for further studies: α5, α6, β1, and β15. Cells were induced with dex/dox (I), or left uninduced (U), for 48 hr, followed by G-CSF simulation (GS; 100 ng/ml). At 7 days, I/GS α5 cells generated 15.5-fold more neutrophils than U/GS α5 cells (p<0.05). Neutrophil expansion of similar magnitude was also seen for the I/GS α6 cells, but not for either β overexpressing clone, nor for the single-vector control clone 3A4, and was due to an increase in both the percentage of neutrophils and total cell number. The percentage of annexin V positive I/GS α5 and α6 cells were reduced 50–80% compared to U/GS cells (p<0.05), while flow cytometry analysis using the Cell Census System (Sigma) revealed no difference in proliferation rates. Affymetrix gene expression analysis (GeneChip Mouse Expression Set 430 2.0 arrays using ANOVA with a threshold False Discovery Rate<0.05) of 4 sets each of I/GS α5 cells, U/GS α5 cells, I/GS 3A4 cells and U/GS 3A4 cells identified several apoptosis pathway genes that were upregulated in I/GS α5 cells compared to the other 3 groups including nucleoside transporter Slc28a2 (CNT2; 13.5-fold; cytoplasmic accumulation of nucleotides recently has been shown to inhibit the intrinsic apoptosis pathway) and Evi1 (4-fold). In addition, several pro-apoptotic genes were downregulated including caspase 6 (2.6 fold) and caspase 8 (1.8 fold). Quantitative RT-PCR confirmed the microarray differences between I/GS α5 and U/GS α5 cells, including a 600-fold increase in Slc28a2 mRNA, and revealed a much smaller change, or no change, in expression of these genes in I/GS β15 vs. U/GS β15 cells. Thus, G-CSF activation of Stat3α, but not Stat3β, in 32D cells inhibits apoptosis and promotes neutrophil expansion perhaps by modulating a unique set of apoptosis pathway genes; these results support the hypothesis that each Stat3 isoform may contribute uniquely to myelopoiesis.