Alice A. Gibson

Supplementary Table S1. Percentage of OCT2 inhibition by compounds reported to reduce platinum-induced toxicities. Supplementary Table S2. Validation of human, rat, and murine overexpressed cells by evaluating their ability to accumulate known prototypical transport substrates. Supplementary Figure S1. Chemical structure of duloxetine. Supplementary Figure S2. IVIS imaging of tumor bearing mice. Supplementary Figure S3. Metabolites of duloxetine do not inhibit OCT2 function. Supplementary Figure S4. Duloxetine is an OCT2 inhibitor that extensively binds to extracellular membrane. Supplementary Figure S5. Duloxetine does not prevent peripheral neurotoxicity associated with vincristine and paclitaxel. Supplementary Figure S6. Paw withdrawal force measured by VFH instrument before the start of the treatment in wild-type (WT) and OCT1/2(-/-) mice (n=5-10 per group). Supplementary Figure S7. Activity of oxaliplatin in various colorectal cancer cell lines. Supplementary Figure S8. Effect of duloxetine on sciatic and caudal nerve velocity and amplitude.

Oxaliplatin-induced peripheral neurotoxicity (OIPN) is a debilitating side effect that afflicts approximately 90% of patients that is initiated by OCT2-dependent uptake of oxaliplatin in dorsal root ganglion (DRG) neurons. The antidepressant drug duloxetine has been used to treat OIPN, although its usefulness in preventing this side effect remains unclear. We hypothesized that duloxetine has OCT2-inhibitory properties and can be used as an adjunct to oxaliplatin-based regimens to prevent OIPN. Transport studies were performed in cells stably transfected with mouse or human OCT2 and in isolated mouse DRG neurons ex vivo. Wild-type and OCT2-deficient mice were used to assess effects of duloxetine on hallmarks of OIPN, endogenous OCT2 biomarkers, and the pharmacokinetics of oxaliplatin, and the translational feasibility of a duloxetine-oxaliplatin combination was evaluated in various models of colorectal cancer. We found that duloxetine potently inhibited the OCT2-mediated transport of several xenobiotic substrates, including oxaliplatin, in a reversible, concentration-dependent manner, and independent of species and cell context. Furthermore, duloxetine restricted access of these substrates to DRG neurons ex vivo and prevented OIPN in wild-type mice to a degree similar to the complete protection observed in OCT2-deficient mice, without affecting the plasma levels of oxaliplatin. Importantly, the uptake and cytotoxicity of oxaliplatin in tumor cell lines in vitro and in vivo were not negatively influenced by duloxetine. The observed OCT2-targeting properties of duloxetine, combined with the potential for clinical translation, provide support for its further exploration as a therapeutic candidate for studies aimed at preventing OIPN in patients with cancer requiring treatment with oxaliplatin.Significance:We found that duloxetine has potent OCT2-inhibitory properties and can diminish excessive accumulation of oxaliplatin into DRG neurons. In addition, pretreatment of mice with duloxetine prevented OIPN without significantly altering the plasma pharmacokinetics and antitumor properties of oxaliplatin. These results suggest that intentional inhibition of OCT2-mediated transport by duloxetine can be employed as a prevention strategy to ameliorate OIPN without compromising the effectiveness of oxaliplatin-based treatment.

Oxaliplatin-induced peripheral neurotoxicity (OIPN) is a debilitating side effect that afflicts approximately 90% of patients that is initiated by OCT2-dependent uptake of oxaliplatin in dorsal root ganglion (DRG) neurons. The antidepressant drug duloxetine has been used to treat OIPN, although its usefulness in preventing this side effect remains unclear. We hypothesized that duloxetine has OCT2-inhibitory properties and can be used as an adjunct to oxaliplatin-based regimens to prevent OIPN. Transport studies were performed in cells stably transfected with mouse or human OCT2 and in isolated mouse DRG neurons ex vivo. Wild-type and OCT2-deficient mice were used to assess effects of duloxetine on hallmarks of OIPN, endogenous OCT2 biomarkers, and the pharmacokinetics of oxaliplatin, and the translational feasibility of a duloxetine-oxaliplatin combination was evaluated in various models of colorectal cancer. We found that duloxetine potently inhibited the OCT2-mediated transport of several xenobiotic substrates, including oxaliplatin, in a reversible, concentration-dependent manner, and independent of species and cell context. Furthermore, duloxetine restricted access of these substrates to DRG neurons ex vivo and prevented OIPN in wild-type mice to a degree similar to the complete protection observed in OCT2-deficient mice, without affecting the plasma levels of oxaliplatin. Importantly, the uptake and cytotoxicity of oxaliplatin in tumor cell lines in vitro and in vivo were not negatively influenced by duloxetine. The observed OCT2-targeting properties of duloxetine, combined with the potential for clinical translation, provide support for its further exploration as a therapeutic candidate for studies aimed at preventing OIPN in patients with cancer requiring treatment with oxaliplatin.Significance:We found that duloxetine has potent OCT2-inhibitory properties and can diminish excessive accumulation of oxaliplatin into DRG neurons. In addition, pretreatment of mice with duloxetine prevented OIPN without significantly altering the plasma pharmacokinetics and antitumor properties of oxaliplatin. These results suggest that intentional inhibition of OCT2-mediated transport by duloxetine can be employed as a prevention strategy to ameliorate OIPN without compromising the effectiveness of oxaliplatin-based treatment.

Cellular viability and uptake of 8FcA by OATP1B1 following reverse transfection with 25 nM siRNA for various protein kinases. Percent uptake was determined relative to OATP1B1 uptake with control siRNA.

Assessment of LYN kinase expression, and function in the presence of various TKIs. (A) Detection of LYN protein expression in human hepatocyte and spleen lysates as determined by Western blot analysis. Prior to analysis, spleen (1:10) and liver (1:1) lysates were diluted. Protein expression of GAPDH was measured as a loading control. The image is a representative of 3 independent experiments. (B) LYN activity in the presence of various TKI concentrations (10-doses) was determined using a kinase assay. (C) Regression of LYN IC50 values of several TKIs and percent of LYN inhibition.

Purpose:OATP1B1 (SLCO1B1) is the most abundant and pharmacologically relevant uptake transporter in the liver and a key mediator of xenobiotic clearance. However, the regulatory mechanisms that determine OATP1B1 activity remain uncertain, and as a result, unexpected drug–drug interactions involving OATP1B1 substrates continue to be reported, including several involving tyrosine kinase inhibitors (TKI).Experimental Design:OATP1B1-mediated activity in overexpressing HEK293 cells and hepatocytes was assessed in the presence of FDA-approved TKIs, while rosuvastatin pharmacokinetics in the presence of an OATP1B1 inhibiting TKI were measured in vivo. Tyrosine phosphorylation of OATP1B1 was determined by LC/MS-MS–based proteomics and transport function was measured following exposure to siRNAs targeting 779 different kinases.Results:Twenty-nine of 46 FDA-approved TKIs studied significantly inhibit OATP1B1 function. Inhibition of OATP1B1 by TKIs, such as nilotinib, is predominantly noncompetitive, can increase systemic concentrations of rosuvastatin in vivo, and is associated with reduced phosphorylation of OATP1B1 at tyrosine residue 645. Using genetic screens and functional validation studies, the Src kinase LYN was identified as a potential regulator of OATP1B1 activity that is highly sensitive to inhibition by various TKIs at clinically relevant concentrations.Conclusions:A novel kinase-dependent posttranslational mechanism of OATP1B1 activation was identified and interference with this process by TKIs can influence the elimination of a broad range of xenobiotic substrates.

Influence of TKIs on OATP1B1-dependent uptake or substrate disposition. (A) Uptake of 3H-EβG (2 µM, 2 min) in OATP1B1 overexpressing HEK293 cells in the presence or absence of select individual FDA approved TKIs (10 μM) identified as positive or negative OATP1B1 inhibitors, following pre-treatment with TKIs for 15 minutes (n=6) (A). Data are presented as a percent of OATP1B1 activity compared to treatment with DMSO (Vehicle). *denotes P < 0.05 vs vehicle and all values represent mean {plus minus} SEM and the experiment was repeated on two separate occasions. (B) Time dependent plasma concentration of methotrexate following injection (10 mg/kg, i.v.) in the presence or absence of a 30-minute pre-treatment with nilotinib (37.5 mg/kg, p.o.). *denotes P < 0.05 vs vehicle and all values represent mean {plus minus} SEM of three male mice.

Mean {plus minus} SEM change of 8FcA uptake (Mean difference) in OATP1B1 overexpressing HEK293 cells between vehicle (DMSO) and exposure to FDA approved TKIs. Statistical significance was measured by ANOVA and Fisher's LSD test.

A structural model of OATP1B1. The model was generated by Phyre2 and presented as a cartoon. Y640 and Y645 are highlighted in blue and the image was produced with PyMOL.