BAG1 Overexpression Restrains the Anti-Apoptotic BCL2, MCL1 and HSP70 Proteins in Acute Myeloid Leukemia.

Abstract

AbstractAbstract 2492 BACKGROUND:Childhood Acute Myeloid Leukemia (AML) is a heterogeneous disease characterized by recurrent genetic aberrations; impaired apoptosis of leukemia cells can also contribute to disease development. BCL2-associated AthanoGene1 (BAG1) is a multifunctional protein preferentially supporting cell survival. Its aberrant expression was demonstrated in diverse cancer types. We previously reported high expression of BAG1 protein in leukemia cell lines. We showed that leukemic cell survival was impaired after silencing BAG1. Additionally, we confirmed that, in vitro, BAG1 acted in synergy with BCL2 and influenced MCL1 expression, both being important anti-apoptotic players in AML (Aveic et al. PlosOne 2011). AIM:We investigated BAG1 expression and role in pediatric AML samples collected at diagnosis, as well as its partner molecules BCL2, MCL1 and HSP70. We then evaluated whether BAG1 may affect patient survival. METHODS:Expression of BAG1 was studied at mRNA and protein levels, using RQ-PCR and Western Blot (WB). Initially, 10 patients with AML whose samples were collected at diagnosis and during therapy while they were in disease remission, were included. Subsequently, wide protein expression was performed by Reverse-phase protein array (RPPA) on a further cohort of 66 newly diagnosed AML patients enrolled into the AIEOP-LAM 2002 protocol. RESULTS:BAG1 mRNA resulted to be heterogeneously expressed in AML cases. Furthermore, BAG1 protein was highly expressed in leukemia samples when compared with healthy bone marrow (hbm), and with patient bm in remission, where BAG1 was undetectable by WB analysis. By RPPA, we confirmed the overexpression of BAG1 protein in 56/66 AML patients (84.8 %) with respect to hbm (113616 ± 66721 Arbitrary Unit (AU); mean fold increase: 2.36). Then, we studied its anti-apoptotic co-partner BCL2, whose expression was found to correlate with BAG1 over-expression (110033 ± 89117 AU, p<0.001). We hypothesized that high BAG1 and BCL2 expression may sustain anti-apoptotic signal in AML. We demonstrated that the active phosphorylated form of BCL2, as well as MCL1, resulted to be overexpressed in AML pediatric patients (BCL2S70 = 70387 ± 40864 AU; p<0.001; MCL1 = 112309 ± 49254 AU; p<0.001). BAG1 has a more specialized role in regulating protein degradation; such a specialized function is accomplished together with HSP70 that co-localized with BAG1 in complexes. HSP70 accumulates inside the cells after exposure to stress, promoting their survival, and, in contrast to normal cells, it is expressed abundantly in most cancers. We showed a positive correlation between HSP70 protein expression and BAG1, BCL2 and MCL1 overexpression (112420 ± 4289 AU; p<0.001). We subdivided AML patients according to BAG1 protein expression into quartiles and studied its relation with biological and clinical features. No significant correlation was found with karyotype, molecular genetics, FAB-morphology, age and white blood count. We highlighted a higher frequency of t(8;21) AML1-ETO translocated patients (who are stratified in standard risk group with a better prognosis in the AIEOP AML 2002 protocol) with low BAG1 expression [9/14 patients (64.3 %) of first quartile; p<0.001]. Studying the event-free survival through the Kaplan Meier method of patients with low (quartile 1+2) and high BAG1 expression (quartiles 3+4), we found a better outcome for patients belonging to the former group, as compared to the latter one (72 % EFS vs. 54% at 3y, p= 0.08, SR= 0.1). CONCLUSIONS:We demonstrate that BAG1 is overexpressed in the majority of childhood AML, and that this correlated with high levels of anti-apoptotic proteins such as BCL2, MCL1 and HSP70. We propose that an aberrant activation of the anti-apoptotic pathway in AML may play a crucial role in leukemia development, and it deserves further investigations. Disclosures:No relevant conflicts of interest to declare.