Journal: Cell reports. Medicine

Article Title: Excessive concentrations of kinase inhibitors in translational studies impede effective drug repurposing.

doi: 10.1016/j.xcrm.2023.101227

Figure Lengend Snippet: Figure 2. Successful drug repurposing is limited by poorly predictive in vitro metrics (A and B) Schematics for in vitro metrics used to predict clinical efficacy. (A) Fold change in IC50 is measured from dose-response data. The IC50 is the concentration of drug at which half the maximum cytotoxic effect is observed. The IC50 fold change is the ratio of mutant IC50 to wild-type IC50. (B) The free drug kill rate is measured from the cell response at a specific concentration: the drug’s known free Cave. The fraction of viable cells at the free Cave is transformed using a model of exponential growth to determine the free drug kill rate (a). For a derivation of the a term, see STAR Methods. (C) Schematic of the kinase domain of the BCR-ABL oncogene. All mutations used in our dose-response drug screen are annotated. (D) A heatmap indicating the IC50 fold change of 23 BCR-ABL variants and five ABL inhibitors. BCR-ABL variants were transduced in BaF3 cells, and IC50 values were determined by measuring the dose-response of each drug-variant pair in triplicate. Red indicates more resistant; blue indicates more sensitive. (E) Free drug kill rate of each drug-variant pair as determined using the drug’s in vivo unbound Cave. Dose response for each drug-variant pair was measured as in (D). The free drug kill rate was calculated using the transformation in (B). Red indicates more resistant; blue indicates more sensitive. (F) ROC curves for classifying clinical resistance in each drug-variant pair using either fold change IC50 (orange) or free drug kill rate (red). A binary classifier was constructed for either metric using data in (D) and (E). A drug-variant pair’s true class was determined from the clinical data. For a complete description of clinical classification, see Table S4.

Article Snippet: REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, peptides, and recombinant proteins Murine IL-3 PeproTech Cat# 213-13 Human serum albumin Sigma Cat# A9511 Alpha-1-acid glycoprotein Sigma Cat# G9885 Critical commercial assays CellTiter-Glo Promega Cat# G7570 p-ABL ELISA Cell Signaling Cat# 7903C BCA Assay Thermo Scientific Cat# 23225 Deposited data Raw data for BCR-ABL dose response curves This study https://github.com/pritchardlabatpsu/DrugRepurposing/ tree/main/ReferenceFiles/DoseResponseCurves Raw data for EGFR/KIT dose response curves This study https://github.com/pritchardlabatpsu/DrugRepurposing/ tree/main/ReferenceFiles/EGFR_KIT Experimental models: Cell lines HEK293T ATCC Cat# CRL-3216 BaF3 DSMZ Cat# ACC 300 Recombinant DNA pLVX-IRES-Puro Clontech Cat# 632183 pLVX-BCRABL-IRES-Puro Genscript Acc: EU216071.1 pLVX-EGFR-IRES-Puro Genscript Acc: NM_005228.4 pLVX-KIT-IRES-Puro Genscript Acc: NM_000222.3 Software and algorithms Code for data analysis and model construction This study Zenodo https://doi.org/10.5281/zenodo.8326122

Techniques: In Vitro, Concentration Assay, Mutagenesis, Transformation Assay, Variant Assay, In Vivo, Construct

Journal: Cell reports. Medicine

Article Title: Excessive concentrations of kinase inhibitors in translational studies impede effective drug repurposing.

doi: 10.1016/j.xcrm.2023.101227

Figure Lengend Snippet: Figure 3. A competitive binding model enables determination of the effective drug kill rate (A) Schematic of in vivo drug binding kinetics. The drug may bind to its intended target protein or to off-target serum proteins. (B) Biochemical expression and differential equations for alternative models of drug binding kinetics. In the noncompetitve binding reaction (top), drug can bind only to serum proteins, as in an equilibrium dialysis assay. In the competitive binding reaction, drug may bind to target or serum proteins, as in the in vivo setting. (C) Heatmap indicating the ratio (r) of drug not bound to serum (4) in the competitive vs. noncompetitive cases: r = 4competitive/4noncompetitive. Here, drug not bound to serum (4) refers to unbound drug in the noncompetitive model and unbound and target-bound drug in the competitive model. A high ratio (yellow region) indicates parameter spaces where equilibrium dialysis (noncompetitive setting) strongly underpredicts the available drug reaching its target in vivo (competitive setting). Species concentrations were determined from a closed-form steady-state solution of ordinary differential equations (ODEs) in (B); see STAR Methods for more details. (D and E) Cell viability of BCR-ABL-transformed BaF3 cells treated with 700 nM imatinib (D) and 50 nM nilotinib (E) in the presence of varying concentrations of HSA and AAG (purple), HSA alone (red), AAG alone (blue), or no serum proteins (gray horizontal line). Error bars are standard error of the mean for technical triplicates. (F) ELISA results for phospho-ABL signal in BCR-ABL-transduced HEK293T cells. Cells were dosed with imatinib in the presence (purple) or absence (gray) of HSA and AAG. The difference in response indicates that the addition of serum proteins directly interferes with imatinib’s inhibition of ABL activity. Error bars are standard error of the mean for technical triplicates. (G) The effective drug kill rate is the kill rate measured at the effective Cave. The effective Cave is determined using the drug’s serum shift factor (top), defined as the ratio of IC50 observed in the presence (purple) and absence (gray) of serum protein. The Cave measured in vivo is corrected by dividing by the serum shift factor to determine the effective Cave (bottom), and the effective drug kill rate is measured from the serum-free dose-response curve at this concentration. (H) A heatmap indicating the effective drug kill rate of each drug-variant pair. Red indicates more resistant; blue indicates more sensitive. (I) ROC curves for classifying clinical resistance in each drug-variant pair, using effective drug kill rate (green), fold change IC50 (orange), or free drug kill rate (red).

Article Snippet: REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, peptides, and recombinant proteins Murine IL-3 PeproTech Cat# 213-13 Human serum albumin Sigma Cat# A9511 Alpha-1-acid glycoprotein Sigma Cat# G9885 Critical commercial assays CellTiter-Glo Promega Cat# G7570 p-ABL ELISA Cell Signaling Cat# 7903C BCA Assay Thermo Scientific Cat# 23225 Deposited data Raw data for BCR-ABL dose response curves This study https://github.com/pritchardlabatpsu/DrugRepurposing/ tree/main/ReferenceFiles/DoseResponseCurves Raw data for EGFR/KIT dose response curves This study https://github.com/pritchardlabatpsu/DrugRepurposing/ tree/main/ReferenceFiles/EGFR_KIT Experimental models: Cell lines HEK293T ATCC Cat# CRL-3216 BaF3 DSMZ Cat# ACC 300 Recombinant DNA pLVX-IRES-Puro Clontech Cat# 632183 pLVX-BCRABL-IRES-Puro Genscript Acc: EU216071.1 pLVX-EGFR-IRES-Puro Genscript Acc: NM_005228.4 pLVX-KIT-IRES-Puro Genscript Acc: NM_000222.3 Software and algorithms Code for data analysis and model construction This study Zenodo https://doi.org/10.5281/zenodo.8326122

Techniques: Binding Assay, In Vivo, Expressing, Transformation Assay, Enzyme-linked Immunosorbent Assay, Inhibition, Activity Assay, Concentration Assay, Variant Assay