The Anti-Proliferative Effect of Immunotoxin Gemtuzimab Ozogamycin (GO) (Mylotarg) to Acute Myeloid Leukemia (AML) Cells Is Associated with the Level of Protein Kinase Syk Exspression.

AML cells express the cell surface antigen CD33 that serves as a down-regulator of cell growth. Anti-CD33 monoclonal antibody (mAb) coupled to a toxin is a licensed drug for the treatment of relapsed AML (Mylotarg). Syk is not only an essential element in several cascades coupling antigen receptors to cell responses, but also SYK is a tumor suppressor gene. Silencing of SYK by hypermethylation results in absence of Syk expression and unresponsiveness of tumor cells to various treatments.. Earlier we demonstrated that about 30% of the AML samples were Syk-negative and the response of leukemia cells to CD33 ligation correlated with Syk expression. Therefore, here we investigated whether or not the response of the AML cells to GO treatment also depends on the level of Syk expression. It is even more intriguing since 50% of Mylotarg is not conjugated to calicheamicin, and hence some of its activity is in fact due to anti-CD33 mAb signaling. Primary leukemia cells were analyzed for their level of the Syk expression (11 Syk positive and 6 Syk-negative) and then treated with GO. In both groups GO mediated dose-dependent inhibition of proliferation, however the level of inhibition in Syk-positive group was significantly higher (p<0.003). The largest differences were notices at low doses (1–50 ng/ml) of GO, while at higher doses (>100 ng/ml) this effect was diminished. Moreover, 8 of 11 (72.7%) samples in Syk-positive group were “good” responders to GO (inhibition>50%), while in Syk-negative group only 1 of 6 (16.6%) samples demonstrated similar response. This difference was statistically significant and suggesting a correlation between the level of Syk expression in AML cells and their response to Mylotarg. After treatment of Syk-negative AML samples with methylation inhibitor 5-aza-cytidine in 2 of 6 samples Syk protein expression was restored. However, in all 6 samples after 5-aza-CdR treatment GO inhibitory activity was significantly higher (p<0.005). Moreover, after 5-aza-cytidine treatment, these initially CD33 ligation non-responsive cells demonstrated high level (up to 70%) inhibition of proliferation upon anti-CD33 mAb treatment. Analysis of CD33 ligation mediated signaling in 3 Syk-positive samples (1 responsive and 2 unresponsive to GO treatment) revealed significant differences: in responsive AML sample Syk became phosphorylated and created complexes with CD33 and protein phosphatase SHP-1, while in non-responsive samples were no Syk/CD33 or Syk/SHP-1 complexes. This does imply that not only the physical presence of Syk, but also its functional activity is important for the adequate response of AML cells to Mylotarg. To create a model with down-regulated Syk we treated Syk-positive human AML cell lines with various doses of the Syk inhibitor Piceatannol and analyzed their proliferative response to Mylotarg. In AML cell lines treated with 10–100 μM of Piceatannol, GO inhibitory activity was significantly lower compared to untreated cells or even those treated by low doses of Piceatannol. This confirmed our observation that expression and functional activity of Syk is associated with the ability of AML cells to respond to GO treatment. We hypothesize that the pre-treatment level of Syk expression may be used as a prognostic factor for response to Mylotarg treatment. Screening for Syk expression in leukemia cell samples from patients before treatment might allow rational selection of patients for this expensive and toxic therapy and thus improve its success rate in the clinic.