Autophagy describes a process of membrane trafficking where specialized compartments (autophagosomes) engulf damaged or dispensable organelles and target them to lysosomic degradation. Thus, autophagy has cytoprotective functions e.g. during starvation, but autophagy can also lead to caspase-independent cell death (PCD type II). ATG5 is a central player in autophagy and inactivating ATG5 severely impairs normal development. In hematopoiesis, ATG5 is essential for maturation of B lymphocytes as well as for T cell survival and proliferation. In recent years, several differentiation and death-inducing agents were shown to activate autophagy in acute myeloid leukemia (AML) cell line models, but the mechanisms involved are still poorly defined. We therefore decided to investigate the role of ATG5 and autophagy in all-trans retinoic acid (ATRA)-induced neutrophil differentiation of AML cells. We found that ATG5 mRNA was upregulated 5.9-, 3.8- and 3.4-fold after 6 days of ATRA-treatment in HL60, NB4 and HT93 AML cells, respectively. In contrast, PMA-induced macrophage differentiation of HL60 and U937 cells only slightly induced ATG5 mRNA by 1.3- and 1.5-fold, respectively, indicating a specific role for ATG5 in granulocyte differentiation. In line with the above observation in leukemic cell lines, we found that ATG5 mRNA levels were increased in 5/5 APL patients upon ATRA therapy. In addition, ATG5-ATG12 conjugates, hallmarks of the autophagic process, were markedly activated in ATRA-treated NB4 and HL60 cells compared to control cells as measured by Western blotting. In agreement, we found increased autophagic activity during myeloid differentiation as evidenced by two additional autophagy markers, i.e. conversion of light chain 3 B (LC3B)-I into LC3B-II and degradation of sequestosome 1 (SQSTM1, p62) protein by Western blotting. Inhibition of autophagy during ATRA treatment of NB4 and HL60 cells with 3-Methyladenine or Bafilomycin A significantly impaired neutrophil differentiation by 80% as measured by CD11b surface expression. Similarly, lentivirus-driven short hairpin (sh)RNA-mediated silencing of ATG5 in NB4 cells resulted in an 85% reduction of ATRA-induced neutrophil differentiation as measured by CD11b surface expression and by quantitative RT-PCR of the myeloid differentiation markers GCSFR, C/EBPε and lactotransferrin. Interestingly, inhibition of autophagy increased overall cell death during the myeloid differentiation process rather than reducing it as measured by reduction of tetrazolium salt (XTT assay). Enhanced cell death and reduced myeloid differentiation upon blocking ATG5 would suggest that autophagy is needed for maintaining myeloid differentiation. Further support for our hypothesis that ATG5 is needed for neutrophil development stems from our survey of ATG5 mRNA expression in primary hematopoietic cells. We found significantly higher ATG5 mRNA levels in granulocytes and macrophages (n=7) as compared to CD34+ hematopoietic progenitor cells from healthy donors (n=4; p=0.0424) as well as compared to primary AML patient samples at diagnosis (n=76; p=0.0003). ATG5 mRNA expression in CD34+ and AML patients was not significantly different (p=0.6669). In summary, we show a correlation of high ATG5 expression with terminal myeloid differentiation and of low ATG5 expression with a myeloid leukemic phenotype. Using chemical inhibitors of autophagy as well as RNAi technology to knockdown ATG5, we further provide evidence that ATG5 and consequently autophagy are essential for neutrophil development.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author