High Expression of the Ca²⁺-Binding Proteins S100A8 and S100A9 Cause Glucocorticoid Resistance in MLL-Rearranged Infant Acute Lymphoblastic Leukemia.

Abstract 729

MLL-rearranged Acute Lymphoblastic Leukemia (ALL) in infants (i.e. children <1 year of age) represents an aggressive and difficult to treat type of leukemia, displaying cellular resistance to several chemotherapeutics, especially to glucocorticoids like prednisolone. As prednisolone response is highly predictive for clinical outcome, it is of utmost importance to unravel the mechanism underlying resistance to this drug.

To gain insights in the prednisolone resistance mechanism, we compared gene expression profiles (Affymetrix HU133plus2) from prednisolone-resistant and prednisolone-sensitive MLL-rearranged infant ALL patients. This gene signature revealed that multiple genes involved in calcium signaling were up-regulated in prednisolone-resistant samples. Most pronounced up-regulation was observed for the S100 protein family members S100A8 and S100A9, which are calcium-binding proteins, that function in a complex. Quantitative RT-PCR (TaqMan) analyses confirmed that S100A8 and S100A9 mRNA expression was ∼100-fold higher in prednisolone-resistant cells compared to patients sensitive to this drug (p=0.008).

Glucocorticoids are known to induce apoptosis by releasing Ca²⁺ from the endoplasmic reticulum (ER) into the cytosol and towards the mitochondria. Within mitochondria, elevated Ca²⁺ levels induce cytochrome c release, triggering apoptosis. Since S100A8/A9 are capable of binding free cytosolic Ca²⁺, we hypothesized that over-representation of these cytosolic proteins may prevent Ca²⁺ to reach the mitochondria and forestall apoptosis.

To test our hypothesis, we first co-incubated prednisolone-sensitive MLL-rearranged ALL cells with prednisolone and EGTA or BAPTA-AM. Both agents represent Ca²⁺chelators and scavenge free cytosolic Ca²⁺, thereby mimicking the Ca²⁺binding by S100A8/A9. Flow cytometry analyses showed that both EGTA and BAPTA-AM inhibited the free cytosolic calcium released by prednisolone. Cytotoxicity tests demonstrated that these prednisolone-sensitive cells became more resistant to prednisolone after co-incubation with either EGTA or BAPTA-AM. Next we asked whether enforced over-expression (using retroviral expression vectors) of S100A8 and/or S100A9 in prednisolone-sensitive MLL-rearranged ALL cells could also inhibit free cytosolic Ca²⁺and induce prednisolone resistance. Indeed, both S100A8 and S100A9 were capable of binding free cytosolic Ca²⁺ released by prednisolone, accompanied by a 30-40% increase in cell survival when either S100A8 or S100A9 were over-expressed alone. Simultaneous over-expression of both S100A8 and S100A9 almost completely reversed the prednisolone-sensitive into a prednisolone-resistance phenotype.

In conclusion, these findings implicate that high expression of S100A8 and S100A9 contributes to prednisolone resistance in MLL-rearranged infant ALL.