Targeting Glycolysis and MAPK Pathway: A Combined Pre-Clinical Approach on Acute Myeloid Leukemia

We have demonstrated that selective MEK inhibitors (MEK-I) significantly inhibit in vitro the growth of AML cell lines and primary AML blasts (Ricciardi et al., JMM 2012). However, these effects were mostly related to the inhibition of cell cycle progression, while apoptosis induction requires higher concentrations of the inhibitor and longer times of exposure. Among the many downstream effectors of MAPK pathway, metabolism is one of the key aspects that can be targeted at several levels for therapeutic perspectives (Ricciardi et al., Blood 2015). In this study, we have explored on AML cells the functional effects of combining the MEK1/2 inhibitor, PD0325901 (PD), with a metabolic modulator, dichloroacetate (DCA), which acts as pyruvate dehydrogenase kinase (PDHK) inhibitor. AML cell lines (OCI-AML3, U937, MOLM13, HL60) and AML primary samples were exposed to PD (2.5-1000nM) and DCA (0.5-2mM), alone or in combination.

PD dose-dependently inhibited cell growth of OCI-AML3, HL60 and MOLM13, showing a constitutive activation of RAS/RAF/MEK/ERK pathway. This activation was not seen in U937, thus exhibiting resistance to PD. DCA shows a pro-apoptotic activity on AML cells, only at the highest concentration. The combination between PD and DCA synergistically enhanced the antiproliferative effect, with a combination index (CI) ranging between 0.37 and 0.41 in OCI-AML3, as measured by isobologram analysis from MTT data. Conversely, in U937 the combination resulted antagonistic. Co-administration of non- or sub-toxic concentrations of DCA (2mM) and of PD (10nM) resulted in a strong increase of cell death in OCI-AML3 (Annexin V). The PD/DCA combination significantly increased Annexin V+ cells, at 72 hours, (65.7±25.0%), compared to single compounds: 16.4±8.7% with 10nM PD (p=0.03) and 28.0±12.4% with 2mM DCA (p=0.04). Similar results were obtained in HL60 and MOLM13 (data not shown). In the U937 cell line, resistant to PD, the levels of apoptosis observed with DCA alone were not further enhanced by the combination: 6.19±1.7% (1000nM PD), 23.63±6.7% (2mM DCA), 31.28±9.9% (PD/DCA). At a protein level, densitometric analysis of Western blot carried out on OCI-AML3 demonstrated that PD/DCA was able to further increase PD-induced ERK inhibition (from 40% to 54%), while DCA alone, as expected, had no effect. Real time metabolic data, obtained by a Seahorse XF24 on OCI-AML3, showed that the combination of PD and DCA affected mitochondrial metabolism: basal respiration and ATP production were impaired by 37.34±10.4% (p=0.012) and by 56.31±10.0% (p=0.003), respectively, compared to control. On the other hand, both maximal respiration and spare respiratory capacity were increased by 28.80±9.5% (p=0.025) and 68.78±14.2% (p<0.001), respectively. The effects of the PD/DCA combination were then examined on AML primary samples (n=3). Preliminary results indicate that PD/DCA combination effectively killed AML cells with a net apoptosis induction at 72 hours (Annexin V+ cells: 31.79±5.9% with PD/DCA, 9.95±5.6% with 10nM PD and 8.48±7.3% with 2mM DCA) in 2 out of 3 patients. In the third sample, which proved hypersensitive to PD (58.73% with 10nM PD at 72h), the combination was ineffective (52.48%).

These preliminary results suggest that the combination of a signal transduction and a metabolic inhibitor spares mitochondrial machinery while impairing at the same time its activity, thus resulting in an enhanced antiproliferative and pro-apoptotic effect on AML cells.