Abstract

Abstract 3835

Background:

Activation of the RAS pathway plays an important role in the pathogenesis of myeloid malignancies. RAS mutations occur in some 10% of MDS and acute myelogenous leukemia (AML) cases. We previously showed in mouse models of NRASD12/BCL-2 MDS/AML that RAS and BCL-2 co-operate in vivo. These genes co-localize at the plasma membrane in the MDS model (MRP8NRASD12/MMTVLTRTetoBCL-2) with increased apoptosis and in mitochondria in the AML post-MDS model (MRP8[NRASD12/BCL-2] with reduced apoptosis (Omidvar Cancer Res 2007). Here, we screened MDS patients for the cellular co-localizations of RAS:BCL-2 and the consequent apoptosis status.

Methods:

39 MDS and AML post-MDS patients were studied, including (WHO classification): 3 RCMD-RS, 3 RCMD, 1 RARS, 8 RAEB I, 9 RAEB II, 4 CMML I, 1 CMML II, 10 AML post-MDS and 2 normal individuals. 17 patients had cytogenetic abnormalities. Bone marrow (BM) mononuclear cells (MNC) were assayed for RAS:BCL-2 localization by confocal and immunofluorescence microscopy. IL-3 dependent mouse hematopoietic FDCP-1 cells were infected with NRASD12 or BCL-2 alone or in combination. Co-localization was measured using anti-NRAS, anti-BCL-2, anti-mitochondrial (Tom 20) and anti-ezrin or anti-wheat germ agglutinin (WGA) plasma membrane antibodies. Apoptosis in patient samples was detected using the Mebstain apoptosis kit and visualized using FITC-dUTP on cytospun cells; apoptosis in the mice was detected by single photon emission computed tomography (SPECT) using technicium labeled annexin-V (99mTc-AnnexinV). NRAS mutations were analyzed by exon direct sequencing or by WAVE, a denaturing high-performance liquid chromatography (DHPLC) based assay. Lineage negative Lin-/Sca-1+/c-Kit+ (LSK) spleen cells were purified by automacs followed by flow sorting. Western blots were probed with anti-caspase 3 and 9 specific antibodies.

Results:

RAS and BCL-2 were co-expressed in the BM (MNC) of all 39 cases. The intensity (low or high) of RAS:BCL-2 co-localization measured by confocal microscopy significantly correlated with % BM blasts (p=0.0283) and WHO classification (p<0.0001); and was particularly high in 11 of the 12 cases (2 RAEBII, 9 AML post-MDS) with >15% marrow blasts. The intensity of the complex was independent of karyotype and NRAS mutation (found in 3 of 30 (10%) patient samples). The 2 normal samples had low intensities of RAS:BCL-2 co-localizations and wild type NRAS. Further analysis of RAS and BCL-2 co-localizations was carried out in 13 of these patients. In 6 of the 7 patients with BM blasts <15%, more than 10% of cells with apoptotic features (APO+) were seen and the RAS:BCL-2 complex was found at the plasma membrane. On the other hand, 4 of the 6 patients with BM blasts ≥15% showed no apoptotic features (APO-) and a strong co-localization signal of the RAS:BCL-2 complex at the mitochondria. In the remaining 2 patients with BM blasts ≥15% both mitochondria and plasma membrane localization patterns were observed, suggesting a mixed cell population. LSK enriched spleen cells from the AML post-MDS mouse model, and the FDCP-1 cell lines expressing mutant NRAS or BCL-2 alone or in combination were subjected to ligand activated caspase-3 and 9 assays revealed that when BCL-2 is co-localized with NRASD12 at the plasma membrane the anti-apoptotic protective effect of BCL-2 appears to be lost (p<0.0001). This is consistent with the pro-caspase-9-mediated apoptotic features observed in the MDS model determined by Western blotting and further supported by in vivo SPECT apoptosis imaging observed in this model and in low risk MDS patients by FITC-dUTP assays.

Conclusion:

These findings illustrate the prognostic impact of the intensity of the RAS:BCL-2 complex in MDS/AML post-MDS patients, and further suggest that localization of RAS:BCL-2 at the plasma membrane correlates with less advanced disease while its presence at mitochondria correlates with more advanced MDS. Therapeutic targeting of the RAS:BCL-2 complex could potentially prevent transformation in MDS patients.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.