Abstract
Cancer-type organic anion transporting polypeptide 1B3 (Ct-OATP1B3), a splice variant of the hepatic uptake transporter OATP1B3 (liver-type), is expressed in several tumor entities, including colorectal carcinoma (CRC) and breast cancer. In CRC, high OATP1B3 expression has been associated with reduced progression-free and overall survival. Several kinase inhibitors used for antitumor treatment are substrates and/or inhibitors of OATP1B3 (e.g., encorafenib, vemurafenib). The functional importance of Ct-OATP1B3 has not been elucidated so far. Human embryonic kidney 293 cells stably overexpressing Ct-OATP1B3 protein were established and compared with control cells. Confocal laser scanning microscopy, immunoblot, and proteomics-based expression analysis demonstrated that Ct-OATP1B3 protein is intracellularly localized in lysosomes of stably transfected cells. Cytotoxicity experiments showed that cells recombinantly expressing the Ct-OATP1B3 protein were more resistant against the kinase inhibitor encorafenib compared with control cells [e.g., encorafenib (100 µM) survival rates: 89.5% versus 52.8%]. In line with these findings, colorectal cancer DLD1 cells endogenously expressing Ct-OATP1B3 protein had poorer survival rates when the OATP1B3 substrate bromosulfophthalein (BSP) was coincubated with encorafenib or vemurafenib compared with the incubation with the kinase inhibitor alone. This indicates a competitive inhibition of Ct-OATP1B3–mediated uptake into lysosomes by BSP. Accordingly, mass spectrometry-based drug analysis of lysosomes showed a reduced lysosomal accumulation of encorafenib in DLD1 cells additionally exposed to BSP. These results demonstrate that Ct-OATP1B3 protein is localized in the lysosomal membrane and can mediate transport of certain kinase inhibitors into lysosomes, revealing a new mechanism of resistance.
SIGNIFICANCE STATEMENT This study describes the characterization of a tumor-associated splice variant (cancer-type organic anion transporting polypeptide; Ct-OATP1B3) of the liver-type OATP1B3 protein (Lt-OATP1B3). This variant is localized in lysosomes mediating resistance against kinase inhibitors that are substrates of this transport protein by transporting them into lysosomes and thereby reducing the cytoplasmic concentrations of these drugs. Therefore, the expression of the cancer-type OATP1B3 protein is associated with a better survival of cells, revealing a new mechanism of drug resistance.
Introduction
Transport proteins are important for the uptake, distribution, and excretion of xenobiotics, drugs, and endogenous substances in normal and cancerous tissues (Robey et al., 2018; Pizzagalli et al., 2021). Export proteins usually belong to the superfamily of ABC (ATP-binding cassette) transporters (Moitra and Dean, 2011; Locher, 2016), whereas uptake transporters are members of the SLC (solute carrier) transporter superfamily (Pizzagalli et al., 2021). Additional to their expression in all healthy tissues, including intestine, kidney, and liver, ABC and SLC transporters are also expressed in several cancerous tissues mediating the uptake or export of drugs or endogenous substrates into or out of cancer cells (Robey et al., 2018; Al-Abdulla et al., 2019; Zhang and Wang, 2020). Remarkably, during antitumor therapy, ABC transporters in particular have been characterized as important mediators of drug resistance when they are overexpressed in cancerous tissues (Robey et al., 2018), whereas in this context, the role of SLC transporters is currently under intensive investigation (Neul et al., 2016; Huang et al., 2020; Krchniakova et al., 2020; Sun et al., 2020).
The focus of this study is on a variant of the SLCO/SLC21 family member organic anion transporting polypeptide 1B3 (OATP1B3, gene symbol SLCO1B3). OATP1B3 has been characterized as an uptake transporter predominantly localized in the basolateral membrane of human hepatocytes (König et al., 2000). Several endogenous substances as well as widely prescribed drugs have been identified as substrates of this transporter (Seithel et al., 2008; Fahrmayr et al., 2010; Roth et al., 2012). Transported drugs include 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors (statins), antibiotics, and several antitumor therapeutic agents such as methotrexate, paclitaxel, and kinase inhibitors (e.g., vemurafenib) (Zimmerman et al., 2013; Kayesh et al., 2021).
In 2001, Abe et al. (2001) reported that OATP1B3 is also expressed in several cancerous tissues and cancer-derived cells. Nagai et al. (2012) finally demonstrated that not the liver-type OATP1B3 protein (Lt-OATP1B3) but an isoform of the Lt-OATP1B3 protein termed cancer-type OATP1B3 protein (Ct-OATP1B3) is expressed in human cancerous tissues. This isoform results from alternative splicing of the SLCO1B3 gene with the Ct-SLCO1B3 mRNA possessing a unique first exon (called exon 1*), which originates from an alternative transcriptional start site located in intron 3 (Fig. 1a) of the SLCO1B3 gene (Nagai et al., 2012). Ct-OATP1B3 protein encoded by the alternatively spliced Ct-SLCO1B3 mRNA lacks the first 28 amino acids of Lt-OATP1B3 [Fig. 1b (Sun et al., 2014)]. Sun et al. (2014) could discriminate between Lt- and Ct-SLCO1B3 mRNA and demonstrated that Ct-SLCO1B3 mRNA is expressed in 87.2% of investigated colorectal cancer tissues but only in 2.6% of the adjacent healthy tissues. So far, in vitro data regarding Ct-OATP1B3 protein transport function and subcellular localization are controversial (Imai et al., 2013; Thakkar et al., 2013; Sun et al., 2014). Interestingly, it has been demonstrated that the aminoterminal region of the OATP1B3 protein, which is present in Lt-OATP1B3 but absent in Ct-OATP1B3 (Fig. 1C), seems to be important for the correct insertion of Lt-OATP1B3 in the plasma membrane (Chun et al., 2017).
Expression of Ct-SLCO1B3 mRNA or Ct-OATP1B3 protein has been confirmed in different cancerous tissues, including colorectal (Abe et al., 2001; Lee et al., 2008), gastric (Abe et al., 2001), breast (Muto et al., 2007), prostate (Hamada et al., 2008; Pressler et al., 2011), and pancreatic cancer (Kounnis et al., 2011; Hays et al., 2013). Interestingly, a study with 278 colorectal tumor samples demonstrated a higher Ct-SLCO1B3 mRNA expression in lower stage tumors (Lockhart et al., 2008), whereas Lee et al. (2008) found no correlation between tumor stage and Ct-SLCO1B3 mRNA expression. Furthermore, some studies found that a high expression of Ct-OATP1B3 protein is associated with a better overall survival of colorectal cancer or endometrial cancer patients (Lockhart et al., 2008; Ogane et al., 2013; Tang et al., 2021), whereas other studies found that a high expression of Ct-SLCO1B3 mRNA is associated with a reduced progression-free survival in patients with advanced and metastatic colorectal cancer (Teft et al., 2015). Recently, a high SLCO1B3 gene expression was associated with lower overall survival in patients with colorectal cancer (Zhi et al., 2021). Taken together, the cancer-type variant of OATP1B3 protein is expressed in several cancerous tissues, but the localization, transport function, and possible role of expression needs to be elucidated. Therefore, the aim of this exploratory study is to gain insights into the subcellular localization, transport function, and possible functional role of Ct-OATP1B3.
Materials and Methods
Materials
[3H]-bromosulfophthalein (BSP, 14 Ci/mmol) was obtained from HARTMANN ANALYTIK GmbH (Braunschweig, Germany). [3H]-estradiol-17β-glucuronide (E217βG, 50 Ci/mmol) was purchased from American Radiolabeled Chemicals, Inc. (St. Louis, MO). Stably labeled [2H3]-clopidogrel was purchased from @rtMolecule (Poitiers, France). Unlabeled BSP was from AppliChem GmbH (Darmstadt, Germany); unlabeled E217βG and the ProteoExtract Native Membrane Protein Extraction Kit were from Sigma Aldrich (St. Louis, MO). Regorafenib was purchased from LC Laboratories (Woburn, MA). Encorafenib and vemurafenib were from MedChemtronica AB (Sollentuna, Sweden). Minimum essential medium, Dulbecco’s modified Eagle’s medium/F12(1:1), RPMI 1640, Dulbecco’s phosphate buffered saline, fetal bovine serum, hygromycin B (50 mg/mL), penicillin streptomycin solution, 0.05%-trypsin-EDTA solution, Pierce BCA Protein Assay Kit, CellLight Lysosomes-RFP, BacMam 2.0, and the SYTOX Green nucleic acid stain were obtained from Thermo Fisher Scientific (Dreieich, Germany). The NucleoSpin RNA Plus Kit was obtained from MACHEREY-NAGEL GmbH & Co. KG (Düren, Germany). The iScript cDNA Synthesis Kit was from Bio-Rad Laboratories GmbH (München, Germany). The Minute Plasma Membrane Protein Isolation and Cell Fractionation Kit was purchased from Invent Biotechnologies, Inc. (Plymouth, MN). Six-well, 12-well, and 96-well cell culture plates were from Greiner Bio-One (Frickenhausen, Germany). Oregon Green 488 carboxylic acid, succinimidyl ester (OG) was obtained from Invitrogen (Rockford, IL). The LightCycler FastStart DNA MasterPLUS SYBR Green I Kit was obtained from Roche Diagnostics GmbH (Mannheim, Germany). SPY555-DNA (SPY dyes series) was from Spirochrome AG (Stein am Rhein, Switzerland), and the pmScarlet-1_peroxisome_C1 plasmid (RRID:Addgene_85065; Addgene, Watertown, MA) was kindly provided by the Optical Imaging Center Erlangen (OICE). The µ-Slide 8-well glass bottom arrays were obtained by ibidi GmbH (Gräfeling, Germany). The CCK-8 Assay was purchased from GERBU Biotechnik GmbH (Heidelberg, Germany).
Antibodies
Polyclonal rabbit anti–human OATP1B3 antiserum (SKT, indicating the first three amino acids of the epitope: serine, lysine and threonine) directed against the carboxyterminal end of the human OATP1B3 protein (König et al., 2000) was obtained from the Division of Tumor Biochemistry of the German Cancer Research Center (Heidelberg, Germany). The goat anti–rabbit IgG horseradish peroxidase–labeled antibody (RRID:AB_2650489) was purchased from GE Healthcare Life Sciences (Buckinghamshire, UK). The goat anti–mouse IgG Alexa Fluor Plus 647 antibody (RRID:AB_2633277), the Alexa Fluor 568 goat anti–rabbit IgG antibody (RRID:AB_143157), the goat anti–mouse IgG (H+L) secondary antibody, and the HRP antibody (RRID:AB_2533947) were from Thermo Fisher Scientific (Dreieich, Germany). The Anti–alpha 1 Sodium Potassium ATPase antibody (RRID:AB_306023) was from Abcam plc. (Cambridge, UK). The monoclonal mouse anti–pan cadherin antibody (RRID:AB_476826), the mouse anti–β-actin monoclonal antibody (RRID:AB_476743), and the anti–lysosomal-associated membrane protein 1 (LAMP1) antibody produced in rabbit (RRID:AB_477157) were purchased from Sigma-Aldrich (St. Louis, MO).
Cell Culture
Human embryonic kidney 293 (HEK293; RRID:CVCL_0045) cells were obtained from the American Type Culture Collection and incubated at 37 °C and 5% CO2. The cell culture medium was minimum essential medium, supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. For antibiotic selection, hygromycin B (250 µg/mL) was added. The stable-transfected cell lines were routinely checked for expression of the respective transport protein. DLD1 (RRID:CVCL_0248) and T84 cells (RRID:CVCL_0555) were kindly provided by Dr. Britzen-Laurent and Dr. Naschberger (Department of Surgery, Erlangen, Germany). All human cell lines have been authenticated using Short Tandem Repeat (or single-nucleotide polymorphism) profiling within the last 3 years, and the experiments were performed with mycoplasma-free cells. For DLD1 cells, RPMI 1640 medium was used, and for T84 cells, Dulbecco’s modified Eagle’s medium/F12 (1:1) was used, both supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. Subcultivation was done twice a week using trypsin 0.05%–EDTA 0.02% solution.
Generation of Stably Transfected HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3 Cells
For the amplification of the Ct-SLCO1B3 cDNA, synthesized single-stranded cDNA of human colon tumor total RNA served as template, using oCt-OATP1B3-5′-For (5′-aactagcagatgttcttggcag-3′) and oOATP1B3-RT-Rev2 (5′-gcatagacttatccattggtcc-3′) as primers. Subsequent to TOPO-TA cloning, the Ct-SLCO1B3 cDNA was cloned into the expression vector pcDNA3.1 Hygro(-), resulting in the plasmid pCt-SLCO1B3.31_Hygro. To introduce the Kozak sequence around the start ATG, the pCt-SLCO1B3.31_Hygro plasmid and the primer pair oCt-OATP1B3.Kozak.For (5′-gccgccaccatgttcttggcagc-3′) and oOATP1B3-RT-Rev2 were used to amplify a modified cDNA. This cDNA was cloned into the vector pcR2.1.Topo and subcloned into the expression vector pcDNA3.1 Hygro(-), resulting in the plasmid pKozak-Ct-SLCO1B3.31_Hygro. Before transfection, all cDNAs were sequenced, and both plasmids contain a cDNA encoding the same Ct-OATP1B3 protein. Stable transfectants expressing the Ct-OATP1B3 protein were established in the same way as described before by Taghikhani et al. (2019), resulting in the HEK293 cell lines HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3.
Quantitative Polymerase Chain Reaction
Total RNA was isolated using the NucleoSpin RNA Plus Kit according to the manufacturer’s protocol. The concentration of the isolated RNA was photometricly measured. The iScript cDNA Synthesis Kit was used according to the manufacturer’s protocol to generate the first-strand cDNA using 1 µg total RNA per sample. For the quantitative reverse transcriptase PCR reaction, the LightCycler FastStart DNA MasterPLUS SYBR Green I Kit was used. Each sample consisted of 5 µL purified water, 2 µL solution I (Kit included), 1 µL forward and reverse primer, respectively, and 1 µL template single-stranded cDNA. Subsequent to the initial 10-minute denaturation step at 95 °C, the DNA was amplified during 45 cycles of 10-second denaturation at 95 °C, 10-second annealing at 64 °C, and 30-second elongation at 72 °C. The expression of each sample was calculated via linear regression and normalized to the respective expression of the housekeeping gene β-actin.
Immunoblot Analysis
The immunoblot analysis was performed as described earlier (Taghikhani et al., 2020). A protein amount of 30 µg was used for the whole cell homogenate. The plasma membrane fractions and cytosolic fractions were adjusted to an amount of 10 µg protein. For the detection of the OATP1B3 protein, the membrane was incubated with the polyclonal antiserum SKT (1:500) at 4 °C overnight. To detect the binding of the primary antibody, the goat anti–rabbit IgG horseradish peroxidase–labeled antibody (1:10 000) served as secondary antibody. To control sample loading, either the monoclonal mouse anti–human β-actin primary antibody (1:10,000) or the monoclonal mouse anti–pan cadherin antibody (1:2,000) was used. The goat anti–mouse IgG horseradish peroxidase–labeled antibody (1:2,000) was used as secondary antibody. For the analysis of the enriched lysosomal fraction, 10 µg protein was used and staining with the rabbit anti–LAMP1 antibody (1:1,000) served as loading control.
Immunofluorescence Microscopy
Cells were seeded at an initial cell number of 3 × 105 cells per well on poly-D-lysine–coated object slides. After 24 hours, cells were fixed using 70% ice-cold methanol solution and permeabilized with 0.4% Triton X-100 PBS solution. To block the fixed cells, 2% bovine serum albumin solution was added, and the cells were incubated with the polyclonal antiserum SKT (1:500) and the Anti–alpha 1 Sodium Potassium ATPase antibody (1:500) overnight at 4 °C. Alexa Fluor 568 antibody and Alexa Fluor Plus 647 antibody were used as secondary antibodies (1:2000). The nuclei were counterstained with SYTOX Green nucleic acid stain. Microscopy was performed on the Zeiss Spinning Disc Axio Observer Z1 at the OICE (Erlangen, Germany).
Subcellular Fractionation
To separate the plasma membrane fraction from the cytosolic fraction of the stable transfectants, 1 × 107 cells were seeded on 10 cm plates. Twenty-four hours later, protein expression was induced with the addition of sodium butyrate (final concentration 10 mM). Forty-eight hours after seeding, cells were washed with 2 mL PBS, and after the addition of 1.5 mL PBS, cells were transferred into a 2 mL microcentrifuge tube and centrifuged for 5 minutes at 600 g at 4 °C. For isolation, the Minute Plasma Membrane Protein Isolation and Cell Fractionation kit was used. The plasma membrane fraction was finally dissolved in 50 µL of the extraction buffer II of the ProteoExtract Native Membrane Protein Extraction kit. For immunoblot analysis, the plasma membrane fraction was lysed in 0.2% SDS solution containing the complete Tablets (Roche Diagnostics GmbH, Mannheim, Germany) as proteinase inhibitor cocktail.
Lysosome Enrichment
The lysosome fraction was isolated according to Mazzulli et al. (2011). For each sample, cells from three 10 cm plates (1 × 107 cells per plate) were pooled and resuspended in 600 µL sucrose-HEPES buffer (0.25 M sucrose, 10 mM HEPES pH 7.4, 0.1 M EDTA). The cell suspension was homogenized using the B.Braun Potter S homogenizer and subsequently centrifuged for 5 minutes at 6,800 g and 4 °C. The supernatant was saved, and the pellet was again resuspended in 600 µL sucrose-HEPES buffer, and homogenization and centrifugation were repeated. Both supernatants were combined and centrifuged for 10 minutes at 17,000 g and 4 °C. The supernatant was removed, and the pellet, containing the enriched lysosomal fraction, was saved until further use. For immunoblot analysis, the final pellet was lysed in 0.2% SDS solution containing proteinase inhibitor cocktail.
Protein Quantification by liquid chromatography tandem mass spectrometry
Protein abundance of OATP1B3 and Na/K-ATPase as reference protein was determined using a validated liquid chromatography tandem mass spectrometry-based targeted proteomics assay as described elsewhere (Drozdzik et al., 2019). To isolate the crude membrane protein fraction, the cell pellets from five 75 cm2 cell culture flasks per sample were pooled and isolated with the ProteoExtract Native Membrane Protein Extraction kit (Merck KGaA, Darmstadt, Germany) according to the manufacturer’s protocol. The protein concentration of the obtained crude membrane fractions was determined with the BCA assay. To analyze the amount of OATP1B3 protein in the plasma membrane fraction, the Minute Plasma Membrane Protein Isolation and Cell Fractionation kit was used as mentioned above. If necessary, membrane fractions were adjusted to a maximum protein amount of 2 mg/mL.
Subsequently, 100 µL of each membrane fraction was mixed with 10 µL dithiothreitol (200 mM; Sigma-Aldrich, Taufkirchen, Germany), 40 µL ammonium bicarbonate buffer (50 mM, pH 7.8; Sigma-Aldrich, Taufkirchen, Germany), and 10 µL ProteaseMAX (1% m/v; Promega, Mannheim, Germany) and incubated for 20 minutes at 60 °C (denaturation).After cooling down, 10 µL iodoacetamide (400 mM; Sigma-Aldrich, Taufkirchen, Germany) was added, and the samples were incubated in a darkened water quench for 15 minutes at 37 °C (alkylation). For protein digestion, 10 µL trypsin (trypsin:protein ratio: 1:40; Promega, Mannheim, Germany) was added, and samples were incubated in a water quench for 16 hours at 37 °C. Digestion was stopped by the addition of 20 µL formic acid (10% v/v; Sigma-Aldrich, Taufkirchen, Germany). All samples were stored at −80 °C until further processing. The samples were centrifuged one more time for 15 minutes at 16,000g and 4 °C. Finally, 50 µL of the supernatant was mixed with 25 µL isotope-labeled internal standard peptide mix (10 nM of each internal standard; Thermo Fisher Scientific, Dreieich, Germany). All sample preparations and digestion steps were performed using Protein LoBind tubes (Eppendorf, Hamburg, Germany). Protein quantification was conducted on a 5500 QTRAP triple quadrupole mass spectrometer (AB Sciex, Darmstadt, Germany) coupled to an Agilent Technologies 1260 Infinity system (Agilent Technologies, Waldbronn, Germany) using validated liquid chromatography tandem mass spectrometry methods as recently described (Gröer et al., 2013). The monitored peptides for OATP1B3 were ISITQIER, IYNSVFFGR, and NVTGFFQSLK, whereas LSLDELHR was used for Na/K-ATPase. For each peptide, 3 to 4 mass transitions have been monitored, and the absolute protein abundance was assessed by using the stable isotope method and considering the protein content of the sample (data given as pmol transporter protein per mg membrane protein).
Cellular Uptake Assays
Uptake experiments were performed as described earlier (König et al., 2012). In brief, cells were seeded at an initial cell number of 7 × 105 cells per well on poly-D-lysine–coated 12-well plates. After 24 hours, cells were induced with 10 mM sodium butyrate to obtain higher protein levels. Before uptake experiments, cells were washed with prewarmed uptake buffer (142 mM NaCl, 5 mM KCl, 1 mM K2HPO4, 1.2 mM MgSO4, 1.5 mM CaCl2, 5 mM glucose, and 12.5 mM HEPES, pH 7.3). Radiolabeled substrates were dissolved in uptake buffer, and unlabeled substances were added in the respective concentrations for the uptake experiments [BSP (1 µM) and E217βG (5 µM)]. The cells were incubated with the uptake solution for 10 minute and subsequently washed three times with ice-cold uptake buffer. After the cells were lysed with 0.2% SDS, the intracellular accumulation of radioactivity was determined by liquid scintillation counting, and protein concentrations were determined by a bicinchoninic acid assay (BCA Protein Assay Kit, Thermo Fisher Scientific, Bonn Germany). The uptake of 2 µM OG was measured accordingly (without radiolabeled OG), and intracellular OG was measured using the CLARIOstar (BMG LABTECH GmbH, Ortenberg, Germany) after 10 minutes incubation by pipetting 100 µL of each cell lysate and an Oregon Green dilution series in 0.2% SDS solution into a 96-well plate. The emission was measured using the preinstalled fluorescein settings.
Colocalization Analysis
For colocalization analysis, µ-Slide 8-well glass bottom arrays (ibidi GmbH, Gräfeling, Germany) were used and coated with 200 µL poly-D-lysine. Next, 9 × 104 HEK-Kz-Ct-OATP1B3 cells per well were seeded into the µ-Slide and incubated for 24 hours at 37 °C and 5% CO2. Afterward, either 20 µL of the CellLight lysosomes RFP, BacMam 2.0 solution per well was added or the cells were transfected with the pmScarlet-1_peroxisome_C1 plasmid by lipofection and incubated overnight. Finally, the cells were incubated with 1 µM OG solution for 1 hour, washed with PBS, and subsequently, 200 µL new culture medium was applied. OG fluorescence was measured using the preinstalled settings of the GFP channel; SPY555 was measured using the Cy3 channel of the Zen software (RRID:Addgene_85065; Carl Zeiss Microscopy GmbH, Jena, Germany). Microscopy was performed on the Zeiss Spinning Disc Axio Observer Z1 (Carl Zeiss Microscopy GmbH, Jena, Germany) at the OICE (Erlangen, Germany).
Cytotoxicity Analysis
HEK vector control (HEK-VC) and HEK-Kz-Ct-OATP1B3 cells were seeded at an initial cell number of 7 × 103 cells per well on poly-D-lysine–coated 96-well plates at a volume of 200 µL (for DLD1 cells 4 × 103 cells at 100 µL per well). After 24 hours, the kinase inhibitors (stock solutions dissolved in DMSO) were added to the respective cell culture medium. Cells were incubated for 72 hours after compound addition. For vemurafenib and regorafenib, the medium was removed, and the wells were carefully washed twice with 200 µL PBS solution. To lyse the cells, 100 µL of 0.2% SDS solution were added to each well, and the plate was shaken for 30 minutes on a 96-well plate shaker. After the lysis, 150 µL BCA solution was added, and the plate was incubated for 30 minutes at 37 °C. The absorption was measured at 560 nm using a Multiskan FC multiplate reader (Thermo Fisher Scientific Inc., Waltham, MA). The blank value was subtracted, and the absorption values were normalized with the average value of the control. Because encorafenib was interacting with the BCA assay, the CCK-8 assay was used to determine cytotoxicity. For this purpose, 10 µL CCK-8 substrate per well was added to the cells after 72 hour of compound treatment and incubated for 2 hours at 37 °C. The absorption was measured at 450 nm and normalized to the average absorption of the untreated cells. BSP was used as Ct-OATP1B3 transport inhibitor at 200 µM and was added additionally with the cytotoxic compounds, and the respective absorption values were normalized to the average absorption of cells treated with BSP only.
Quantification of Encorafenib in the Lysosomal Fraction by Mass Spectrometry
DLD1 cells were seeded at 1 × 107 cells on 10 cm plates. The cells were incubated either with 100 µM encorafenib alone or in combination with 200 µM BSP for 4 hours. Subsequently, the lysosomal fraction was enriched as described above with minor modifications. The lysosomal pellet was first resuspended in 50 µL PBS, and 2.5 µL of this suspension was used to determine the protein amount via BCA assay. The samples were again centrifuged for 10 minutes at 17,000 g and 4 °C, and the PBS was removed. Each pellet was lysed in 80% MeOH/water (including the internal standard [2H3]-clopidogrel) and afterward centrifuged for 10 minutes at 25,000 g at 4 °C, and 200 µL of the supernatant was transferred into autosampler glass vials (Agilent, Santa Clara, CA). The solvent was evaporated using a MULTIVAP nitrogen evaporator (Organomation, Berlin, MA), and the residue was reconstituted in 100 µL solution with 90% water (containing 0.5% formic acid) and 10% MeOH (with 0.1% formic acid). The samples were measured in triplicates via an in-house method with a Q-Exactive Focus mass spectrometer coupled to a Dionex UltiMate 3000 ultrahigh-performance liquid chromatography (both Thermo Fisher Scientific, Dreieich, Germany) (Kehl et al., 2021). The area of the encorafenib peak was normalized to the peak area of the internal standard and to the protein amount of the lysosomal fraction aliquot. The data are given as percent of the average amount of encorafenib in the DLD1 cells without added BSP.
Statistical Analysis
For all transport and cytotoxicity studies, at least two independent experiments were performed on different days, each measured in triplicates or quadruplets. All data are expressed as means ± S.D. The graphs and statistical analysis were done by using GraphPad Prism (RRID:SCR_002798; Version 5.01, 2007, GraphPad Software, San Diego, CA). The statistical significance was analyzed using either one-way ANOVA with Bonferroni adjusted post hoc tests (comparison between all pairs of columns; group size = 4 if more than two cell lines were used) or a two-tailed unpaired Student′s t test (comparison between two cell lines). For cytotoxicity experiments, P values were calculated for each tested kinase inhibitor concentration to demonstrate statistically different effects between HEK-VC cells and HEK-Kz-Ct-OATP1B3 cells. Due to the exploratory design of the experiment, the calculated P values are descriptive.
Results
Establishment of Cell Lines Recombinantly Overexpressing OATP1B3 Variants
Ct-SLCO1B3 cDNA was cloned by an reverse transcriptase PCR–based approach using human colon cancer total RNA as template. The cDNA was inserted into the expression vector pcDNA3.1 Hygro(-), resulting in the plasmid pCt-SLCO1B3.31_Hygro. Because it has been demonstrated that a Kozak consensus sequence around the start codon of the Ct-SLCO1B3 cDNA enhances protein expression (Imai et al., 2013), the original cDNA was mutated to contain a Kozak consensus sequence, resulting in the plasmid pKz-Ct-SLCO1B3.31_Hygro. Both plasmids were used to establish HEK293 cells recombinantly overexpressing the identical Ct-OATP1B3 protein (HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3). SLCO1B3 mRNA expression analysis was performed by quantitative reverse transcriptase PCR (Fig. 2A), showing an expression of 21.9 ± 3.04% and 56.8 ± 0.7% in HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3 cells, respectively (normalized to the expression of the housekeeping gene β-actin). Already established HEK293 cells recombinantly overexpressing the liver-type variant of the OATP1B3 protein (HEK-Lt-OATP1B3) had a SLCO1B3 mRNA expression of 119.7 ± 11.7% relative to the expression of β-actin (Seithel et al., 2007).
Immunoblot analysis using whole cell homogenates (Fig. 2B) demonstrated a fully glycosylated OATP1B3 protein in HEK-Lt-OATP1B3 cells, whereas in both cell lines expressing the Ct-OATP1B3 protein, a weak band was detectable around 75 kDa, representing the core-glycosylated form of the OATP1B3 protein (König et al., 2000). In the manuscript by König et al. (2000), the same antiserum was used. By using confocal laser scanning microscopy (Fig. 2C), we could confirm the known localization of the Lt-OATP1B3 (red staining) in the plasma membrane (sodium-potassium ATPase, green staining) of HEK-Lt-OATP1B3 cells, whereas in both cell lines expressing the Ct-OATP1B3 protein (red staining), a more intense intracellular staining could be detected in comparison with HEK-VC.
Cellular Uptake Assays
To investigate the function of the OATP1B3 proteins as uptake transporters (putatively) localized in the plasma membrane, cellular uptake assays (Taghikhani et al., 2017) were performed using the prototypic OATP1B3 substrates BSP (Fig. 3A), E217βG (Fig. 3B), and OG (Fig. 3C). All three substrates had a markedly higher uptake into HEK-Lt-OATP1B3 cells (P < 0.001 versus uptake into HEK-VC cells), but no differences in the uptake of HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3 cells could be detected compared with the uptake into HEK-VC control cells.
Proteomics Analysis
Using a proteomics analysis, the OATP1B3 protein amount in the isolated plasma membrane fraction (Fig. 4A) and in the crude membrane fraction (Fig. 4B) of the HEK293 cells was quantified. In the isolated plasma membrane fractions, OATP1B3 protein could be detected in HEK-Lt-OATP1B3 cells but not in HEK-VC, HEK-Ct-OATP1B3, and HEK-Kz-Ct-OATP1B3 (Fig. 4A). In the crude membrane fractions of HEK-Ct-OATP1B3 and HEK-Kz-Ct-OATP1B3 cells, OATP1B3 protein could be detected but not in HEK-VC cells (Fig. 4B). Due to the high protein amount, the quantification of Lt-OATP1B3 protein was omitted in the crude membrane fraction.
Subcellular Localization of Ct-OATP1B3
For subsequent experiments, only HEK-Kz-Ct-OATP1B3 cells were used as Ct-OATP1B3 overexpressing cells because of the higher SLCO1B3 mRNA expression. Next, the subcellular localization of the Ct-OATP1B3 protein was analyzed. In immunoblot analysis and in line with the proteomics analysis, there was no detectable protein band at 75 kDa in the isolated plasma membrane fraction of the HEK-Kz-Ct-OATP1B3 cells. However, a band could be detected in the cytosolic fraction containing intracellularly localized organelles (Fig. 4C). In contrast, a strong band at 130 kDa could be detected in HEK-Lt-OATP1B3 cells, whereas no protein was detectable in the cytosolic fraction of these cells. Next, we isolated the lysosomal fraction of HEK-Kz-Ct-OATP1B3 and HEK-VC cells and analyzed the protein abundance. Figure 4D shows an OATP1B3 protein band at 75 kDa in the enriched lysosomal fraction. Because the OATP1B3 staining using the polyclonal antiserum SKT also leads to unspecific protein binding (e.g., the protein bands in the cytosolic and lysosomal fractions at approximately 65 kDa) and the fact that no monoclonal anti–OATP1B3 antibody, which is able to detect both variants, is commercially available, a direct colocalization with lysosomal markers was not possible (data not shown). Therefore, we tried to confirm the lysosomal localization by transfecting the HEK-Kz-Ct-OATP1B3 cells either with the baculovirus-based CellLight lysosomes-RFP BacMam 2.0 system to stain the lysosomes or with the pmScarlet-1_peroxisome_C1 plasmid to stain the peroxisomes. Twenty-four hours after transfection, we incubated the cells for 1 hour with OG, a known substrate of Lt-OATP1B3 (Fig. 3C) (Izumi et al., 2016). Colocalization of the lysosomal marker with OG, resulting in a yellow color of the lysosomes, could be detected (Fig. 4E), whereas no colocalization could be detected between OG-stained vesicles and the stained peroxisomes (Fig. 4F; Supplemental Fig. 1), indicating lysosomal localization of the Ct-OATP1B3 protein.
Cytotoxicity of Kinase Inhibitors
To investigate the function of Ct-OATP1B3 localized in lysosomes, we used the BRAF inhibitor vemurafenib, substrate and inhibitor of Lt-OATP1B3 (Zimmerman et al., 2013; Kayesh et al., 2021) and the BRAF inhibitor encorafenib, approved for the treatment of metastatic colorectal cancer in combination with cetuximab. In addition, we used the vascular endothelial growth factor receptor inhibitor regorafenib (also approved for the treatment of metastatic colorectal cancer), which is not a substrate or inhibitor of Lt-OATP1B3 (https://www.accessdata.fda.gov/drugsatfda_docs/nda/2013/204369Orig1s000ClinPharmR.pdf; Ohya et al., 2015). HEK-Kz-Ct-OATP1B3 and HEK-VC cells were incubated with the respective kinase inhibitor at different concentrations, and the number of surviving cells was determined (Fig. 5).
Expression of the Ct-OATP1B3 protein led to better survival when cells were incubated with encorafenib (Fig. 5A). A slight effect of Ct-OATP1B3 expression was detected for vemurafenib (Fig. 5B; P = 0.051 at 100 µM), whereas the survival curves of both cell lines were nearly identical when treated with regorafenib (Fig. 5C). This indicates that Ct-OATP1B3 expression may confer resistance against kinase inhibitors and that this resistance is substrate dependent.
Ct-OATP1B3 in the Colorectal Cancer Cell Lines DLD1 and T84
Additionally, the Ct-SLCO1B3 mRNA and Ct-OATP1B3 protein analysis were performed in the two colorectal carcinoma cell lines T84 and DLD1. Both cell lines endogenously express Ct-SLCO1B3 mRNA (Fig. 6A). Furthermore, Ct-OATP1B3 protein expression could be verified in immunoblot (Fig. 6B) and proteomics analysis (Fig. 6C). The protein band of Ct-OATP1B3 could also be detected in the enriched lysosomal fractions of both cell lines (Fig. 6D).
In DLD1 cells, we inhibited the potential Ct-OATP1B3–mediated substrate transport into lysosomes in the cytotoxicity assays by adding the OATP1B3 substrate BSP (200 µM). When added in combination with the same kinase inhibitors used in HEK293 cell experiments, BSP led to an increase in the cytotoxic effects of encorafenib (Fig. 7A) and vemurafenib (Fig. 7B). Interestingly, for regorafenib, the addition of BSP led to a markedly higher survival in the DLD1 cells (P < 0.001 at 100 µM; Fig. 7C). To verify whether the increase in the cytotoxicity of encorafenib is due to an inhibition of the lysosomal sequestration, we incubated DLD1 cells with encorafenib (100 µM) alone or in combination with BSP for 4 hours, isolated the enriched lysosomal fraction, and quantified the amount of this kinase inhibitor. Due to transport inhibition, the addition of BSP led to a reduction of encorafenib in the enriched lysosomal fraction (P < 0.001; Fig. 7D).
Discussion
The aim of this study was to gain insights into the expression, localization, and function of the Ct-OATP1B3 protein and to clarify previously published inconsistent results regarding these issues. Therefore, we established stably transfected HEK293 cells recombinantly overexpressing this transport protein and investigated protein amount and localization, Ct-OATP1B3–mediated transport, and a possible function of Ct-OATP1B3 expression in transfected HEK293 cells and colorectal cancer cells. Our results demonstrate that Ct-OATP1B3 protein is localized in lysosomes mediating the transport of substances into these intracellular vesicles. Furthermore, expression of Ct-OATP1B3 protein led to a better survival of cells when treated with the kinase inhibitor encorafenib, used for the treatment of metastatic colorectal cancer.
Before Nagai et al. (2012) characterized the Ct-SLCO1B3-mRNA as a splice variant of the SLCO1B3 gene expressed in colorectal cancer tissue, several authors described OATP1B3 overexpression in different cancerous tissues without discriminating between the Lt- and Ct-OATP1B3 variant (Abe et al., 2001; Lee et al., 2008; Lockhart et al., 2008). Hence, these older studies may need re-evaluation on this background, and further studies should discriminate between both variants (Alam et al., 2018; Sun et al., 2020). For example, Alam et al. (2018) identified mRNA expression of both variants in breast cancer samples, whereas Thakkar et al. (2013) detected only Ct-SLCO1B3-mRNA in CRC samples. This discrimination of both variants could also help to explain the described varying SLCO1B3-correlated outcomes in breast cancer (Muto et al., 2007; Tang et al., 2021) and CRC (Teft et al., 2015; Zhi et al., 2021). Furthermore, the published results obtained with CRC cells transfected with the cDNA encoding the Lt-OATP1B3 protein (Lee et al., 2008; Niedermeyer et al., 2014) are difficult to interpret because the localization of Ct-OATP1B3 is altered due to the truncated aminoterminal end (Chun et al., 2017). This altered localization of the Ct-OATP1B3 protein could be verified with the present work since no Ct-OATP1B3 protein could be detected in the isolated plasma membrane fraction (Fig. 4A), but low amounts were detectable in the crude membrane fraction (Fig. 4B). In addition, this aminoterminal truncation seems to affect the glycosylation of the Ct-OATP1B3 protein (Fig. 2B), resulting in a band with a molecular mass of approximately 75 kDa (König et al., 2000; Ho et al., 2006). Similar results were obtained by analyzing Ct-OATP1B3 expression in the colorectal carcinoma cell line HCT-116 and the pancreatic cancer cell line Panc-1 (Thakkar et al., 2013) and by analyzing the N-linked glycosylation of Lt-OATP1B3 in nonalcoholic fatty liver disease (Clarke et al., 2017). Additionally, the results of the cellular uptake experiments (Fig. 3), which showed no difference between the Ct-OATP1B3 transfectants and HEK-VC, are in line with the intracellular localization of Ct-OATP1B3 and with the previously described lack of transport function (Thakkar et al., 2013).
Because unspecific antiserum binding of the polyclonal antiserum SKT and the low protein translation efficiency, especially in stably transfected HEK293 cells, limit the methods of protein detection, colocalization studies were used to identify the vesicles containing the Ct-OATP1B3 protein. Using the fluorescent dye and known OATP1B3 substrate OG, we analyzed the potential transport function of Ct-OATP1B3 protein located in intracellular vesicles (Fig. 4, E and F) and identified these vesicles as lysosomes by cotransfection with the lysosomal marker BacMam Lysosomes-RFP (Fig. 4E). Interestingly, this lysosomal localization may be of special interest for the possible function of Ct-OATP1B3 protein expression in cancer cells. Lysosomal sequestration has been described for various drugs used in antitumor therapy, especially some kinase inhibitors (e.g., nintendanib, sunitinib) (Krchniakova et al., 2020). This sequestration leads to a reduced cytoplasmic concentration of these drugs and consequently results in a higher cytotoxic resistance (Englinger et al., 2017). So far, this kind of transport has been attributed to the function of ABC transporters [e.g., ABCA3, P-glycoprotein] localized in lysosomal membranes (Chapuy et al., 2009; Yamagishi et al., 2013). Our cytotoxicity experiments (Fig. 5) demonstrated that overexpression of the SLC transporter Ct-OATP1B3 led to a better survival of cells when incubated with selected kinase inhibitors, suggesting that also SLC transporters located in intracellular vesicles may be important for this resistance mechanism.
Zimmermann et al. (2013) could demonstrate a considerably higher uptake of vemurafenib into stable-transfected HEK293 cells overexpressing Lt-OATP1B3 compared with the HEK293 control cell line. The other BRAF inhibitor encorafenib is not described as substrate for the Lt-OATP1B3 protein so far and did not show an enhanced uptake into HEK-Lt-OATP1B3 cells compared with the control cell line in our experiments (Supplemental Fig. 2), but both BRAF inhibitors are mentioned as inhibitors of Lt-OATP1B3–mediated transport by the Food and Drug Administration (https://www.accessdata.fda.gov/drugsatfda_docs/nda/2011/202429Orig1s000ClinPharmR.pdf; https://www.accessdata.fda.gov/drugsatfda_docs/nda/2018/210496Orig1s000MultidisciplineR.pdf). Notably, the tested concentrations of vemurafenib in this study were in the concentration range of its reported plasma trough concentrations in humans (Verheijen et al., 2017; Mueller-Schoell et al., 2021), underlining a potential in vivo relevance of Ct-OATP1B3. This was recently reported for OATP1B3 protein without discriminating between the Lt- and Ct-OATP1B3 protein (Kayesh et al., 2021). It should be noted that the cytoprotective effect of Ct-OATP1B3 expression is not a general effect for all kinase inhibitors as we could show with regorafenib, which is neither a substrate nor inhibitor of Lt-OATP1B3 (Fig. 5C) (https://www.accessdata.fda.gov/drugsatfda_docs/nda/2013/204369Orig1s000ClinPharmR.pdf; Ohya et al., 2015). Interestingly, the protective effect of Ct-OATP1B3 expression was diminished by the coapplication of the Lt-OATP1B3 substrate BSP together with encorafenib (Fig. 7A) or vemurafenib (Fig. 7B) in DLD1 cells endogenously expressing Ct-OATP1B3. In this experimental setup, the Ct-OATP1B3–mediated transport of both kinase inhibitors into lysosomes seems to be inhibited, resulting in an increased cytotoxicity of treated DLD1 cells. In line with these results, the amount of encorafenib in the enriched lysosomal fraction was reduced (Fig. 7D). It can be speculated that Ct-OATP1B3–mediated resistance could contribute to insufficient clinical effects of monotherapy of BRAF inhibitors in colorectal cancer (Prahallad et al., 2012; Mao et al., 2013; Yaeger et al., 2017).
Taken together, we could demonstrate that Ct-OATP1B3 protein, a splice variant of the liver-type uptake transporter Lt-OATP1B3, is localized in lysosomes and is capable of transporting Ct-OATP1B3 substrates into these vesicles. Furthermore, as shown for encorafenib, the substrate spectrum between the Ct-OATP1B3 and the Lt-OATP1B3 protein seems to be different, and further studies are necessary to gain insights into the substrate spectrum of the Ct-OATP1B3 protein. When treated with kinase inhibitors, the expression of Ct-OATP1B3 protein led to a better survival of cells by transporting these cytotoxic compounds into lysosomes, thereby reducing their cytoplasmic concentration. This is a new kind of a Ct-OATP1B3–mediated resistance mechanism against antitumor drugs and in line with the association of a higher Ct-SLCO1B3 expression and a reduced progression-free survival or overall survival of colorectal cancer patients (Teft et al., 2015; Zhi et al., 2021). Its importance for the therapy of other tumor entities expressing Ct-OATP1B3 or for other antitumor drugs that are substrates of this transport protein needs to be elucidated in the future. Furthermore, these results suggest that the expression of Ct-OATP1B3 protein in tumor tissues may serve as an important biomarker for patients treated with antitumor drugs.
Acknowledgments
We thank Olga Stelmakh and Claudia Hoffmann for their technical assistance. Microscopy analysis was performed with the support of the OICE and Dr. Philipp Tripal, Dr. Benjamin Schmidt, and Dr. Zoltán Winter. DLD1 and T84 cells were kindly provided by Dr. Britzen-Laurent and Dr. Naschberger (Erlangen, Germany).
Authorship Contributions
Participated in research design: Haberkorn, Oswald, Gessner, Taudte, Zunke, Fromm, König.
Conducted experiments: Haberkorn, Oswald, Kehl, Dobert.
Contributed new reagents or analytic tools: Oswald, Gessner, Taudte, Dobert, Zunke.
Performed data analysis: Haberkorn, Oswald, König.
Wrote or contributed to the writing of the manuscript: Haberkorn, Oswald, Kehl, Gessner, Taudte, Dobert, Zunke, Fromm, König.
Footnotes
- Received April 8, 2022.
- Accepted September 5, 2022.
This work was supported by the Wilhelm Sander-Stiftung (Grants 2019.050.1 and 2019.097.1) and by the DFG (Grant INST 90/1048-1 FUGG).
The authors declare no conflict of interest.
↵This article has supplemental material available at molpharm.aspetjournals.org.
Abbreviations
- ABC
- ATP-binding cassette
- BSP
- bromosulfophthalein
- CRC
- colorectal carcinoma
- Ct-OATP1B3
- cancer-type organic anion transporting polypeptide 1B3
- E217βG
- estradiol-17β-glucuronide
- HEK293
- human embryonic kidney 293
- LAMP1
- lysosomal-associated membrane protein 1
- Lt-OATP1B3
- liver-type organic anion transporting polypeptide 1B3
- OATP1B3
- organic anion transporting polypeptide 1B3
- OG
- Oregon Green 488 carboxylic acid, succinimidyl ester
- OICE
- Optical Imaging Center Erlangen
- SLC
- solute carrier; VC, vector control
- Copyright © 2022 by The American Society for Pharmacology and Experimental Therapeutics