The AKT1 isoform plays a dominant role in the survival and chemoresistance of chronic lymphocytic leukaemia cells

The pathophysiology of chronic lymphocytic leukaemia (CLL) is characterised by a dynamic equilibrium of resting and proliferative tumour cells. While CLL cells in the peripheral blood are mostly G0‐arrested, those residing in lymphoid organs have an activated signature due to supportive signals from diverse immune and stromal cell types (Herishanu et al, 2011). The clinical success of novel small molecule inhibitors targeting Bruton tyrosine kinase, such as Ibrutinib, and phosphatidylinositol‐3 kinase (PI3K), such as Idelalisib, strengthens the idea that signals downstream of the B cell receptor are critical for CLL development and progression. In this context, the protein kinase C (PKC) and PI3K pathways are indisputable chief players (for review see Woyach et al, 2012). We and others have shown that downstream of PI3K and PKC‐beta, the serine/threonine kinase AKT, also known as protein kinase B (PKB), regulates various signalling cascades involved in survival (Hofbauer et al, 2010; Zhuang et al, 2010). AKT is encoded by three distinct genes, namely AKT1, AKT2 and AKT3 (also termed PKB‐alpha, PKB‐beta, PKB‐gamma, respectively). AKT1 and AKT2 are ubiquitously expressed whereas AKT3 is mainly expressed in testes and brain (Yang et al, 2003). In this study, we aimed to gain insight into isoform‐specific expression of AKT, and the contribution of AKT1 and AKT2 to stromal and activated T cell‐mediated survival and chemoresistance in CLL. Peripheral blood samples from CLL patients were collected after informed consent was obtained in accordance with the Declaration of Helsinki and under the ethical approval of the Ethics Commission of the Province of Salzburg (415‐E/1287/4–2011, 415‐E/1287/8–2011). First, analysing basal AKT isoform expression in unstimulated purified CLL cells, we observed increased AKT2 mRNA and AKT2 protein expression as compared to AKT1 (Fig 1A and B). As we had previously noted increased AKT phosphorylation in CLL stromal cell co‐cultures (Hofbauer et al, 2010), we evaluated AKT phosphorylation kinetics and isoform contribution to this phenomenon. AKT was rapidly phosphorylated at serine 473 (pS473), a phosphorylation site crucial for full AKT activation (Sarbassov et al, 2005), and remained in the activated form for at least 24 h (Fig 1Ci). To assess the relative activation of AKT1 versus AKT2 in CLL cells co‐cultured with primary stromal cells, we analysed the phosphorylation at Ser473 (AKT1) and Ser474 (AKT2). We observed robust stromal cell‐induced AKT activation of both isoforms in CLL cells cultured in direct cell‐cell contact with stromal cells (Fig 1Cii), but not in CLL cells separated from the stromal layer by a transwell insert (Fig 1Di). AKT activation was associated with increased cell viability (Fig 1Dii). Figure 1 AKT1 is the dominant AKT isoform in stromal cell mediated chronic lymphocytic leukaemia (CLL) cell survival. (A) AKT1 and AKT2 mRNA expression in CD19‐positive selected CLL cells (n = 10) were measured by quantitative real‐time ... Next, we treated co‐cultured CLL cells with several AKT inhibitors. As there is no AKT1‐specific inhibitor available, we used the pan‐AKT inhibitors MK2206 and AiX, and the AKT2‐selective inhibitor, Akti‐2. The specificities of these inhibitors were confirmed by an AKT isoform‐specific pull‐down and subsequent in vitro kinase assay to detect phosphorylation of the AKT substrate GSK3A/B after MK2206 or Akti‐2 treatment of the Epstein‐Barr virus‐positive CLL patient‐derived MEC1 cells (Fig 1E). Applying the inhibitors to primary CLL cells co‐cultured with stromal cells indicated that the selective inhibition of AKT2 did not decrease cell viability, whereas pan‐AKT inhibition resulted in significantly decreased survival (Fig 1F), suggesting a dominance of AKT1 or a cooperation of both isoforms in maintaining cell vitality. Genetic manipulations in primary CLL cells are hard to achieve, therefore, to address these alternatives, we used a siRNA approach to target AKT1 or AKT2 in MEC1 cells. Successful and similar knockdown efficiencies were achieved in both experimental settings (Fig 1G). However, the transient knockdown of AKT1, but not AKT2, resulted in loss of cell viability, establishing AKT1 as the dominant AKT isoform (Fig 1H). Consistently, simultaneous knockdown of both isoforms did not further reduce viability compared to the single AKT1 knockdown (Fig 1H). Chronic lymphocytic leukaemia cells co‐cultured with activated T cells are rapidly activated, allowing us to mimic in vitro at least part of the proliferative processes that take place in lymph nodes (Asslaber et al, 2013). Under these co‐culture conditions, we observed significant transcriptional upregulation of both AKT1 and AKT2 within 24 h (Fig 2A), which was accompanied by phosphorylation of both isoforms on the protein level (Fig 2B). Figure 2 AKT1 phosphorylation and regulation by STAT3 upon T cell‐mediated activation results in fludarabine resistance. Peripheral blood mononuclear cells from chronic lymphocytic leukaemia (CLL) patients containing >5% T cells were activated ... We next studied AKT isoform activation in the context of the oncogene STAT3, a factor of clinical significance to CLL, which has been described to directly interact with the AKT1 promoter (Park et al, 2005; Hazan‐Halevy et al, 2010). Activation of CLL cells by T cells resulted in pronounced STAT3 signalling (pY(705)‐STAT3), which could be antagonised by treatment with the STAT3 inhibitor S3I‐201 (Fig 2C). STAT3 inhibition significantly decreased AKT1 mRNA expression (Fig 2D), indicating an interaction of these pathways upon CLL cell activation. Notably, relative AKT isoform transcription and protein expression in activated CLL cells was not altered upon treatment with the novel small molecule inhibitors Ibrutinib or Idelalisib (data not shown). These observations prompted us to investigate potential synergistic effects of AKT‐ or STAT3‐inhibition with conventional drugs used in the treatment of CLL. We recently observed that T cell activated‐CLL cells gain resistance towards fludarabine (Hofbauer et al, 2014). Consistently, CLL cells co‐cultured with activated T cells remained viable for up to 48 h even in the presence of fludarabine. However, exposure to the pan‐AKT or STAT3 inhibitor induced CLL cell apoptosis within 24 h (data not shown), indicating that inhibition of AKT or STAT3 is able to overcome CLL activation‐induced protection. Notably, after 48 h of culture, CLL cells exposed to the STAT3 inhibitor underwent strong apoptosis irrespective of the presence of fludarabine. Pan‐AKT inhibition, but importantly not AKT2 inhibition, resulted in decreased cell viability levels, which were significantly pronounced when combined with fludarabine (Fig 2E). Taken together, our results indicate a dominant role of AKT1 in microenvironment‐mediated CLL survival and chemoresistance. CLL patients could particularly benefit from targeting the predominant AKT1 isoform, which may also increase the response rate towards classical chemotherapeutics.