The SET Oncogene, a Potent PP2A Inhibitor, Is Elevated in CLL and Antagonism of SET Induces Apoptosis.

Even though we have treatments for CLL, it remains an incurable leukemia. We need new and better treatments for this disease. The Akt kinase is usually constitutively activated (phosphorylated) in CLL, and it acts to maintain CLL cell viability. The tumor suppressor protein phosphatase 2A is important in deactivation of Akt, the mitogen activated protein kinases (MAPK) p38, JNK, ERK, and NFkB (through IkK). We developed apoE-mimetic peptides that potently decrease phosphorylation of Akt and MAPKs, decrease TNF and nitric oxide synthase expression, and display anti-inflammatory activity in vitro and in vivo by antagonism of SET, a potent physiological inhibitor of PP2A. Others have reported that PP2A activity is reduced and that SET is overexpressed in cells of chronic myelocytic leukemia patients. Increased SET expression with consequent decreased PP2A activity leads to dysregulated kinase signaling. We show here that SET is also overexpressed in CLL cells, and that SET antagonist apoE-mimetic peptides kill CLL cells.

Patients were from the Duke University and V.A. Medical Centers, and normal controls were from the community. Control normal PBMC were isolated by ficoll-Hypaque centrifugation, and CD19+ CLL or normal B cells were purified using negative selection with antibodies. We determined cytotoxicity using the MTS colorimetric assay. ApoE-mimetic peptides were prepared by chemical synthesis. Western blotting was used with anti-SET and anti-beta-actin antibodies. Apoptosis assays were performed with the annexin-V:propidium iodide staining method.

Samples from 17 CLL patients and 5 normal volunteers were examined by Western blotting. SET protein levels were 6.2-fold higher in CLL cells than in normal B cells. The apoE-mimetic COG compounds are peptides of 17 to 34 amino acids derived from the ligand-binding region of the apolipoprotein E holoprotein, and they act by binding to SET and preventing the inhibition of PP2A. This results in a net increase in PP2A activity in CLL cells. We have examined 11 COG peptides. Each of the 11 peptides displayed some cytotoxicity for CLL cells in vitro, irrespective of the patients' stages and other good or bad prognostic findings. 12 of 17 CLL patients were Rai stage 0 at presentation, and 5 were stage 1 or 2. They had been followed 4.2 yr (median; range 1.0 – 24.1 yr). 2 of 16 were CD38 positive, and 6 of 15 were Zap-70 positive. Of 15 analyzed, 6 had unmutated IgVH gene. 11 of 17 patients had not been treated. Peptide COG449 was the most potent, while a control peptide with an inverted apoE sequence had no activity. COG449 induced cell death in a dose-dependent fashion in all patients' samples, with a mean ED50 of 80 nM. The ED50 of COG449 for normal B cells was very high (> 10,000 nM). Annexin-V staining indicated that apoptosis was induced at concentrations in good agreement with the ED50 for cytotoxicity of the compounds tested. In vivo studies in normal mice using COG449 show no toxicity even at doses of 100 mg/kg when delivered by subcutaneous injection.

We demonstrated SET overexpression in CLL cells and that apoE-mimetic peptides bind SET to de-inhibit PP2A. This results in apoptosis and death of CLL cells in vitro with high efficacy and potency (low nanomolar ED50s). CLL cells are killed preferentially compared to normal PBMC and B cells. Preliminary studies show that the peptide is nontoxic in normal mice. Trials in CLL patients will help determine the efficacy in vivo.

Christensen:Cognosci Inc.: Employment. Oddo:Cognosci Inc.: Employment. Vitek:Cognosci Inc.: Employment, Equity Ownership.