Transduction of TAT-NF-YA Fusion Protein into K562 Human Chronic Myelogenous Leukemia Cells Upregulates HOXB4 Activity.

Protein transduction to redirect hematopoietic stem cell (HSC) fate for clinical therapy is an attractive alternative to genetic modification, in that the stem cell genome is not permanently modified. Indeed, published studies suggest that protein transduction with HOXB4, similar to retroviral overexpression of HoxB4, can increase HSC competitive repopulating ability. We have recently found that the trimeric transcription factor NF-Y, by modulation of its regulated subunit NF-YA, controls the expression of several HSC activating genes, including the Hox4 paralogs HoxB4, HoxC4, HoxD4 as well as p27, telomerase, Notch1, and LEF1. We have therefore now asked whether we can deliver NF-YA to hematopoietic cells by fusion with the protein transduction domain of the HIV-1 TAT protein, and modeled the molecular and biological activity of this protein by its effects on the HoxB4 promoter in vitro and in vivo. First, we designed and tested three fusion constructs expressing human NF-YA: with an N-terminal 6-histidine leader, a C-terminal 6xHis tag, and a GST tag at the N-terminus of NF-YA. Recombinant proteins bearing either the N- or C-terminal His-tag were poorly expressed under diverse growing and induction conditions, and were highly contaminated with bacterial host proteins. GST-fusion protein showed a much higher expression level and no host protein contamination, so all experiments were performed using GST-constructs. Next, electrophoretic mobility shift assays (EMSA) demonstrated that purified GST-TAT-NF-YA binds to HoxB4 promoter probes identically to endogenously translated NF-YA. NIH 3T3 fibroblasts, or K562 cells cultured in 0.1–2mM GST-TAT-NF-YA for 1.5h, but not without recombinant protein, showed bright cytoplasmic and nuclear fluorescence upon staining with anti-GST. To test the ability of GST-TAT-NF-YA to activate the HoxB4 promoter in live cells, GST-TAT-NF-YA’s effect was measured both on transfected HOXB4 promoter reporter constructs and on endogenous HoxB4 mRNA. First, when GST-TAT-NF-YA was incubated with K562 cells electroporated with HoxB4 promoter/luciferase, along with plasmids encoding the NF-Ya partners NF-YB, NF-YC, USF1, USF2 (and Renilla luciferase for transfection control), normalized luciferase activity calculated as a % relative to control HoxB4-luc activity increased in an NF-YA concentration-dependent fashion. More pronounced effects were seen when GST-TAT-NF-YA protein was added to stably transfected HOXB4-luc K562 cell lines, with up to 2.7 fold increased HoxB4-luc activity versus control cells. Importantly, addition of GST-TAT-NF-YA protein to unmanipulated K562 cells increased expression of endogenous HoxB4 mRNA 2-2.5X, as measured by real-time PCR. Taken together, these data demonstrate that NF-YA protein can be biochemically, efficiently and functionally delivered to hematopoietic cells via TAT-mediated technology, providing a promising alternative to gene transfer to stimulate multiple HSC-activating genes.