Fig 1.
Ba/F3 EML4-ALKmutants exhibit variable dose-response to ABT-199 treatment.
Cell survival dose-response curves of Ba/F3 EML4-ALKmutants (WT, C1156Y, L1196M and G1202R) treated with increasing concentrations of ALK inhibitors (crizotinib, ceritinib and alectinib) and BCL2 inhibitor ABT-199. B) Table shows IC50 values (Mean ± SD) of crizotinib, ceritinib, alectinib and ABT199 treatment in four Ba/F3 EML4-ALKmutants cell models, calculated by parametric nonlinear regression in GraphPad Prism. C) The heat map shows the relative resistance to treatments taking crizotinib in WT cells as a reference; these values were calculated as the average IC50 ratio between all treatments in Ba/F3 EML4-ALKmutants and crizotinib treatment in Ba/F3 EML4-ALKWT. The data show that EML4-ALKWT and EML4-ALKG1202R are resistant to ABT-199 treatment, while EML4-ALKC1156Y and EML4-ALKL1196M are more sensitive. D) Bars representation of IC50 values in the different cell lines treated with the various inhibitors. Statistical comparisons were done by Anova and Tukey test; asterisks indicate significance: p< 0.05 (*p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).
Fig 2.
BCL2 antiapoptotic protein exhibits heterogeneous expression in Ba/F3 EML4-ALKmutants.
A) Heatmap show the expression of BCL2 in different treatment conditions: control, 100nM crizotinib, 50nM ceritinib, and 50nM alectinib, and at two different time points (24 and 48 hours). No differences in BCL2 expression are evident between the two time points. B) Graph bars show BCL2 expression normalized using β-actin as loading control at 48 hours. ANOVA with Tukey test is used for statistical comparisons C) Box plots show in summary the average of BCL2 expression in ALK mutants across all treatment conditions; WT cell model overexpressed BCL2 followed by G12022R, while C1156Y and L1196M exhibited BCL2 expression closely to baseline. D) Pearson correlation analysis between IC50 values and BCL2 expression levels (r: 0.962 and p: 0.0332). Overexpression of BCL2 exhibits resistance to ABT-199, while baseline expression shows sensitivity to ABT-199. Statistical values: significant p< 0.05 (*p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).
Fig 3.
Molecular docking showed that ABT-199 can fit into active site of ALKMutant.
A) At the center of figure, the tridimensional structure of molecular docking of ALKMutant receptor interacting with ABT-199 ligand (orange) simulated in Yasara software is shown. Blue square zoomed ALKwt binding to ABT-199: in yellow sticks we can see Cys1156, in green Leu1196 and in red Gly1202 amino acids (please see S1 for further details); Yellow square zoomed ALKC1156Y binding to ABT-199: in yellow sticks we can see Tyr1156 amino acid point mutation (please see S2 for further details); Green square zoomed ALKL1196M binding to ABT-199: in green sticks we can see Met1196 amino acid point mutation (please see S3 for further details); Red square zoomed ALKG1202R binding to ABT-199: in red sticks we can see Arg1196 amino acid point mutation (please see S4 for further details); ABT-199 can fit into active site of ALKMutant by hydrophobic and hydrogen bond interactions. B) Heat map shows the number of interactions between cdALK amino acid positions and ABT-199 in all ALKmutant, classified into hydrophobic, hydrogen bonds, and π cation interactions. These data were obtained from PLIP website by prior submission “pdb.” of the molecular docking structure.
Fig 4.
Amino acids interactions and binding energy describe that ABT-199 binds to ALKMutant.
A) Heat map shows the binding energy measured in Kcal/mol and calculated using YASARA™ software of ALKmutant proteins (cdALK and AF-ALK) docked with ATP, crizotinib, ceritinib, alectinib, or ABT-199. Values represent the average of independent calculations using two different ALK models obtained from Yasara. B) Anova graph show binding energies mean values normalized with ATP, (grouped by mutant) here we can see to ABT199 competition with ALKi in each cell mutant model. C) Anova graph show binding normalized with ATP (grouped by ligand), here we can see that mutants exerts resistance to each ALKi, same case we can see that ABT199 do not exhibit significant differences in each ALK mutant. Statistical values: significant p< 0.05 (*p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).
Fig 5.
ABT199 inhibits ALK mutant kinase activity and decreases it phosphorylation.
A) ALK kinase activity curve shows that ABT199 is able to inhibit ALK kinase activity (IC50: 5.5 μM). B) Graph bars show Anova and tukey test of pALK expression normalized to ALK. C) Heatmap show pALK expressions normalized after 48 hours treatment by ALKi or ABT199, and their combination (Combo 1: 300 nM ABT + 300 nM Crizotinib, Combo 2: 300 nM ABT + 50 nM Ceritinib, and combo 3: 300 nM ABT + 30 nM Alectinib). Statistical values: significant p< 0.05 (*p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).
Fig 6.
Loewe synergy-antagonism score analysis shows additive and synergistic effects after ALKi and ABT-199 combination treatment.
A) Heat map shows the Loewe synergy-antagonism scores (synergy >10 “blue”, antagonism < -10 “red”, additive from -10 up to 10 “white”) in four EML4-ALK mutant cell models treated with ABT-199 in combination with ALKi (crizotinib, ceritinib and alectinib). B) Dot plot showed that WT and G1202R exhibit wide range of additivity and synergism, respectively, while C1156Y and L1196M showed limited range of synergism and wide range of additivity, respectively.
Fig 7.
Combination therapy exhibited apoptosis in Ba/F3 EML4-ALKmutant cell models.
A) Plots show apoptosis assay by flow cytometry, and B) graph bars show Anova and tukey test. Regarding WT, C1156Y, and L1196M mutants, these results show that combination treatments lead to apoptosis by synergistic and additive effects, as proved by Loewe analysis. In G1202R cells, there are no significant differences between ABT-199 and combination treatments. Statistical values: significant p< 0.05 (*p<0.05; **p<0.01; ***p<0.001; ****p<0.0001).