Fig 1.
A simplified view of the intrinsic apoptosis pathway and Bcl-2 family PPI network.
(a) The anti-apoptosis and pro-apoptosis proteins mutually inhibit each other as indicated. The BH3-only proteins activate the effector proteins, which go on to cause MOMP and subsequently cell death [80, 85–87, 90–99]. (b) Range of dissociation constant (KD) values reported in the literature for some of the BH3 peptides, truncated, and full length Bcl-2 family proteins [86]. White dashed cells indicate unreported KD values for the corresponding interaction.
Fig 2.
Protein network corresponding to the key interactions between the intrinsic apoptosis pathway and the integrated stress response (ISR).
Direct and indirect regulatory effects of venetoclax (ABT-199) and tedizolid on the coupled pathway are shown in the left panel. ABT-199 binds to Bcl-2 to form a protein-drug complex [5, 162]. In addition, both ABT-199 and tedizolid perturb ISR-related and non-ISR pathways, ultimately exerting both positive and negative regulatory effects on c-Myc and Chop (ISR effector protein) expression levels [26, 116]. The coupled apoptosis-ISR pathway is shown in the right panel and depicts the regulatory effects of the transcription factors c-Myc and Chop on the Bcl-2 family proteins, including both the pro-survival and pro-apoptosis effects of c-Myc [128, 130–132, 136–141], as well as the pro-apoptosis effects of Chop [143, 145, 146, 149, 152–157]. In addition, the interactions between the Bcl-2 family proteins [87–89] are shown, including the binding of anti-apoptosis proteins (Bcl-2, Mcl-1) with pro-apoptosis activator (Bim) and effector (Bax, Bak) proteins, as well as the binding of activator and effector proteins and the subsequent activation of the latter. The activated effector proteins cause Caspase-3 activation, which causes cleavage of the anti-apoptosis proteins and promotes apoptosis [80, 85–87, 90–99]. The pro-apoptosis species are coloured green, anti-apoptosis species are red, neutral species are white, c-Myc is orange, and Chop is blue.
Fig 3.
Simulated protein expression levels at t = 72 hour in comparison to experimentally-measured protein expression in MOLM-13 R2 cells.
Blue circles correspond to the average of experimental measurements and error bars correspond to one standard deviation. Plots illustrate total protein expression levels for Mcl-1, Bcl-2, Bim, Bax, Bak, c-Myc, and Chop, and active Caspase-3 levels only. All protein levels are normalized to the expression in the untreated system. (a) 400 nM venetoclax monotherapy, (b) 5,000 nM tedizolid monotherapy, (c) 400 nM venetoclax and 5,000 nM tedizolid combination therapy administered simultaneously. (d) Shows the predicted c-Myc-dependent regulatory functions for each Bcl-2 family protein in the model, normalized by the transcription factor-dependent maximal production rate for each protein. The boundaries separating distinct zones are illustrative and were chosen as approximately the concentration of c-Myc that separates states where pro-apoptotic c-Myc-dependent regulatory functions are predominantly saturated from states where the anti-apoptotic c-Myc-dependent regulatory functions are predominantly saturated (see Eqs. (A.18)-(A.22) in S1 Text).
Fig 4.
Cellular viability (a) and live cell population (b) in untreated and treated MOLM-13 R2 cells.
Solid lines show the results of the mathematical model, while data points correspond to the average of experimental measurements over the five-day window. Error bars correspond to one standard deviation obtained from experimental measurements.
Fig 5.
Results of local sensitivity analysis around the nominal parameter set for (a) the untreated system, (b) 400 nM venetoclax monotherapy, (c) 5,000 nM tedizolid monotherapy, and (d) 400 nM venetoclax and 5,000 nM tedizolid combination therapy. Plots show the relative sensitivity calculated using Eq (2) resulting from perturbations of ±1% to individual kinetic parameters or initial conditions in each treatment case. Top 20 kinetic parameters/initial conditions are shown in each plot. Asterisks denote kinetic parameters that affect the expression of the transcription factors c-Myc and Chop (see Ref. [116]).
Fig 6.
Cellular viability for 400 nM venetoclax treatment in combination with either 5,000 nM tedizolid treatment or an Mcl-1 inhibitor.
Solid lines indicate the results of the mathematical model. Experimental data points for venetoclax/tedizolid combination therapy are shown for reference.
Fig 7.
Results of pre-treatment before the administration of 400 nM venetoclax and 5,000 nM tedizolid combination therapy.
(a) Simulated response values, calculated using Eq (3), for 400 nM venetoclax administered prior to combination therapy and for 5,000 nM tedizolid administered prior to combination therapy, with the indicated pre-treatment windows. Combination therapy was simulated for 5 days following pre-treatment. (b) Simulated peak and end-point live cell counts during the combination therapy with different pre-treatment windows. In (a) and (b), a pre-treatment window of “0” corresponds to combination therapy without pre-treatment. (c) Simulated expression of Chop, free (inactive) Bak, and free Mcl-1 protein levels after the administration of 5,000 nM tedizolid monotherapy. Each protein was normalized to its total expression in the untreated system. (d) Active Caspase-3 levels, normalized to the level in the untreated system, observed in simulations of combination therapy without pre-treatment and with tedizolid administered 36 hours prior to combination therapy. Time is measured relative to the administration of venetoclax. (e) Experimentally-measured cellular viability for MOLM-13 R2 cells treated with venetoclax monotherapy, combination venetoclax/tedizolid therapy without pre-treatment, or combination therapy with tedizolid pre-treatment administered 48 hours prior, at the indicated venetoclax concentrations. Cellular viability was measured 72 hours after the administration of venetoclax and has been normalized to the cellular viability in the untreated cells. Data points show average and standard deviation of experimental measurements, and solid lines correspond to sigmoidal fits to the data (see text for further details).