Article pubs.acs.org/molecularpharmaceutics
Quantitative Targeted Absolute Proteomic Analysis of Transporters, Receptors and Junction Proteins for Validation of Human Cerebral Microvascular Endothelial Cell Line hCMEC/D3 as a Human Blood− Brain Barrier Model
Downloaded via OPEN UNIV OF HONG KONG on January 23, 2019 at 12:42:27 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Sumio Ohtsuki,†,‡ Chiemi Ikeda,† Yasuo Uchida,† Yumi Sakamoto,† Florence Miller,§,∥,⊥ Fabienne Glacial,§,∥,⊥ Xavier Decleves,#,○ Jean-Michel Scherrmann,#,○ Pierre-Olivier Couraud,§,∥,⊥ Yoshiyuki Kubo,† Masanori Tachikawa,† and Tetsuya Terasaki*,† †
Division of Membrane Transport and Drug Targeting, Graduate School of Pharmaceutical Sciences, Tohoku University, Sendai, Japan ‡ Department of Pharmaceutical Microbiology, Faculty of Life Sciences, Kumamoto University, Kumamoto, Japan § INSERM, U1016, Institut Cochin, Paris, France ∥ CNRS, UMR8104, Paris, France ⊥ Université Paris Descartes, Paris, France # Neuropsychopharmacologie des addictions (CNRS UMR 8206), Université Paris Descartes, Faculté de Pharmacie, Paris, France ○ INSERM U705, Neuropsychopharmacologie des addictions, Paris, France ABSTRACT: Human cerebral microvascular endothelial cell line hCMEC/D3 is an established model of the human blood− brain barrier (BBB). The purpose of the present study was to determine, by means of quantitative targeted absolute proteomics, the protein expression levels in hCMEC/D3 cells of multiple transporters, receptors and junction proteins for comparison with our previously reported findings in isolated human brain microvessels. Among 91 target molecules, 12 transporters, 2 receptors, 1 junction protein and 1 membrane marker were present at quantifiable levels in plasma membrane fraction of hCMEC/D3 cells. ABCA2, MDR1, MRP4, BCRP, GLUT1, 4F2hc, MCT1, ENT1, transferrin and insulin receptors and claudin-5 were detected in both hCMEC/D3 cells and human brain microvessels. After normalization based on Na+/K+ ATPase expression, the differences in protein expression levels between hCMEC/D3 cells and human brain microvessels were within 4-fold for these proteins, with the exceptions of ENT1, transferrin receptor and claudin-5. ABCA8, LAT1, LRP1 and γ-GTP were below the limit of quantification in the cells, but were found in human brain microvessels. ABCA3, ABCA6, MRP1 and ATA1 were found only in hCMEC/D3 cells. Furthermore, compared with human umbilical vein endothelial cells (HUVECs) as reference nonbrain endothelial cells, MDR1 was found only in hCMEC/D3 cells, and GLUT1 expression was 15-fold higher in hCMEC/D3 cells than in HUVECs. In conclusion, this is the first study to examine the suitability and limitations of the hCMEC/D3 cell line as a BBB functional model in terms of quantitative expression levels of transporters, receptors and tight junction proteins. KEYWORDS: human blood−brain barrier, transporter, protein quantification, quantitative targeted absolute proteomics, multidrug resistant protein 1, breast cancer resistant protein, ATP-binding cassette transporter, receptor, junction protein, claudin-5
■
INTRODUCTION The blood−brain barrier (BBB) is localized at the interface between blood and cerebral tissue and is formed by the endothelial cells of brain microvessels, which display a unique phenotype characterized by the presence of intercellular tight junctions and the polarized expression of numerous transport systems. ATP binding cassette (ABC) transporters, such as MDR1, BCRP and MRP4, limit entry of drugs to the brain across the BBB.1 Among solute carrier (SLC) transporters, © 2012 American Chemical Society
OAT3, Oatp1a4/Oatp-2 and Oatp1C1/Oatp-14 were reported to be involved in permeability of anionic drugs, opioids and/or thyroid hormones across the BBB.2−4 The BBB also expresses several SLC transporters for nutrients: GLUT1 for glucose, Received: Revised: Accepted: Published: 289
August 6, 2012 October 28, 2012 November 8, 2012 November 8, 2012 dx.doi.org/10.1021/mp3004308 | Mol. Pharmaceutics 2013, 10, 289−296
Molecular Pharmaceutics
Article
employ our recently developed liquid chromatography−tandem mass spectrometry (LC−MS/MS)-based absolute quantification by multiplexed multiple/selective reaction monitoring (multiplexed MRM/SRM) method, which we have used to determine the protein expression levels of 114 membrane proteins in human brain microvessels.13,14,26 The purpose of the present study was, therefore, to comprehensively compare the protein expression levels of multiple key membrane proteins, including transporters, receptors and junction proteins, in hCMEC/D3 cells and human brain microvessels, as well as in human umbilical vein endothelial cells (HUVECs), as prototypical nonbrain human endothelial cells.
MCT1 for lactate and ketone bodies and LAT1 for amino acids and L-dopa. The expressions of these transporters change depending upon nutrient availability and disease status.5−7 Furthermore, the BBB possesses receptor-mediated transporters, including insulin receptor, transferrin receptor and LRP1, which also play important roles in delivering macromolecules into the brain.8,9 Our recent report demonstrated that the BBB expresses NPR-C, which is involved in elimination of atrial natriuretic peptide from the brain.10 A recent study documented marked species differences in mRNA expression of transporters in the brain microvessels of various species, including mouse, rat, porcine, bovine and human.11 Our recent quantitative targeted absolute proteomics (QTAP) study confirmed species differences in the expression of transporters at the protein level in human and mouse brain microvessels.12−14 Functions and protein expressions of OAT3, Oatp1a4/Oatp-2 and Oatp1C1/Oatp-14 were detected in mouse and rat BBB, but the protein expression levels of their homologues were below the limit of quantification in isolated human microvessels.12,14 These observations highlighted the importance of using human in vitro BBB models for analysis of the molecular functions and regulation of human BBB. The human brain endothelial cell line hCMEC/D3 retains most of the morphological and functional characteristics of brain endothelial cells (i.e., expression of tight junction proteins and multiple active transporters and receptors), even in monoculture without glial cells, and is considered to be a unique in vitro model of the human BBB.15−18 It has been widely used by many laboratories to analyze the functions of the human BBB and their modulation by pharmacological agents, xenobiotics, pathogens and other stimuli.19−23 Since comprehensive measurement of transport activities at the in vivo human BBB would be difficult, e.g., requiring the use of multiple appropriately labeled selective substrates, we considered that quantitative analysis and comparison of the expression levels of key transporters and other molecules would be an efficient alternative approach to validation of hCMEC/ D3 cells as a BBB model, at least in terms of potential activities. We initially reported that hCMEC/D3 cells express mRNAs of MDR1, MRP1, MRP5 and BCRP, as well as the corresponding proteins.18 Subsequently, Carl et al. performed a comprehensive quantitative PCR study of transporters in hCMEC/D3 cells, confirming these initial observations and reporting higher expression of MDR1, MRP1 and MRP4 mRNAs as compared with MRP5 and BCRP mRNAs.15 Dauchy et al. reported a similar expression pattern of ATP-binding cassette (ABC) transporter mRNAs in freshly isolated human brain microvessels, although BCRP mRNA was more highly expressed than MDR1, MRP1, MRP4 and MRP5 mRNAs.16 These reports suggest that the pattern of expression of ABC transporter mRNAs in hCMEC/D3 cells is at least qualitatively similar to that of human brain microvessels, although the expression levels differ. However, mRNA expression does not always correlate with protein expression and molecular function.24,25 Because (i) membrane proteins, such as transporters, receptors and junction proteins, play important roles in BBB function and (ii) protein expression levels, rather than mRNA expression levels, are more closely related to protein function, we considered that quantitative analysis of key proteins in hCMEC/D3 cells and comparison of the results with those for freshly isolated human brain microvessels would be a rational strategy for further testing the validity of hCMEC/D3 cells as an in vitro human BBB model. For this purpose, we aimed to
■
MATERIALS AND METHODS Cell Culture. hCMEC/D3 cells were seeded at 1.5 × 106 cells on collagen type I coated 10 cm dishes. The cells were cultured in EBM-2 medium (Takara Bio, Shiga, Japan) supplemented with 2.5% fetal bovine serum, 0.025% VEGF, 0.025% R3-IGF, 0.025% hEGF, 0.01% hydrocortisone, 5 μg/ mL bFGF, 10 mM HEPES and 1% penicillin−streptomycin in an atmosphere of 95% air and 5% CO2 at 37 °C for 3−4 days for routine culture. The culture period was extended to 6 days for cells used to prepare whole cell and plasma membrane fraction, and the passage number was kept below 35.18 HUVECs (BD Biosciences, Bedford, MA) were seeded at 1.0−2.0 × 105 cells on 10 cm tissue culture dish (BD Biosciences), and cultured in HuMedia-EG2 supplemented with HuMedia-EG supplement set (Kurabo, Osaka, Japan) in an atmosphere of 95% air and 5% CO2 at 37 °C for 4 days. The cells were collected for membrane preparation after 3 passages. Preparation of Membrane Fraction. Cells were scraped from culture dishes and collected by centrifugation at 230g for 5 min at 4 °C. The cells were suspended in suspension buffer (10 mM Tris-HCl (pH 7.4) containing 250 mM sucrose, 1 mM EGTA) with protease cocktail (SIGMA, St. Louis, MO) and then lysed by nitrogen cavitation at 450 psi for 15 min at 4 °C in a pressure vessel (Parr, Moline, IL). The homogenate was centrifuged at 10000g for 10 min at 4 °C. The supernatant was ultracentrifuged at 100000g for 40 min at 4 °C. The resulting pellet was suspended in suspension buffer. The suspension was layered on top of 38% (w/v) sucrose solution and centrifuged at 100000g for 40 min at 4 °C. The turbid layer at the interface was recovered, suspended in suspension buffer, and centrifuged at 100000g for 40 min at 4 °C to obtain the plasma membrane fraction. Protein concentrations were measured by the Lowry method using the DC protein assay reagent (Bio-Rad, Hercules, CA). LC−MS/MS-Based Protein Quantification Analysis. Protein quantitation of the target molecules was simultaneously performed by LC−MS/MS using our previously described multiplexed MRM/SRM method.14 Protein expression levels were determined by quantifying specific target peptides produced by trypsin digestion. The absolute amount of each target peptide was determined by using an internal standard peptide, which is a stable isotope-labeled peptide with identical amino acid sequence to that of the corresponding target peptide. The target peptide for quantification was selected based on in silico selection criteria as described previously.26 Quantification of human transporters and other membrane proteins was carried out under the MRM/SRM conditions reported previously.12,14,26 290
dx.doi.org/10.1021/mp3004308 | Mol. Pharmaceutics 2013, 10, 289−296
Molecular Pharmaceutics
Article
Protein samples (100 μg) were suspended in suspension buffer containing 7 M guanidium hydrochloride and 10 mM EDTA. Samples were reduced with dithiothreitol at room temperature for 60 min under a nitrogen atmosphere, and Scarbamoylmethylated with iodoacetamide at room temperature for 60 min. The alkylated proteins were precipitated with a mixture of methanol and chloroform. The precipitates were dissolved in 6 M urea, diluted with 100 mM Tris-HCl (pH 8.0) and digested with tosylphenylalanyl chloromethyl ketonetreated trypsin at an enzyme/substrate ratio of 1:100 at 37 °C for 16 h. The tryptic digests were spiked with stable isotope-labeled internal standard peptides and acidified with formic acid for analysis with the HPLC system, which was connected to an electrospray ionization triple quadrupole mass spectrometer (API5000, AB Sciex, Foster City, CA, USA) operated in positive ionization mode. LC was performed with C18 columns. Linear gradients of 1−50% acetonitrile in 0.1% formic acid were applied to elute the peptides at a flow rate of 50 μL/min. The mass spectrometer was set up to run a multiplexed MRM/SRM experiment for peptides. The ion counts in the chromatograms were determined by using the quantitation procedures in Analyst software (AB Sciex). One specific peptide was selected for quantification of each target protein, and measured at 4 different MRM/SRM transitions. The amount of each peptide was determined as the average of 3 or 4 MRM/SRM transitions from one sample. In cases where signal peaks were obtained at only 2 or fewer transitions, the amount of peptide in the sample was defined as under the limit of quantification, unless otherwise noted in the tables. The value of the limit of quantification in each transition was defined as the value giving a peak area of 5000 counts for each protein sample, calculated by using the peak area of the isotope-labeled internal standard peptide in the corresponding transition of the same sample.
Table 1. Protein Expression Levels of Membrane Proteins in hCMEC/D3 Cellsa protein expression level (fmol/μg protein) transporters ABCA2 ABCA3c ABCA6c ABCB1/MDR1 ABCC1/MRP1c ABCC4/MRP4 ABCG2/BCRP SLC2A1/GLUT1 SLC3A2/4F2hc SLC16A1/MCT1 SLC29A1/ENT1 SLC38A1/ATA1c receptors transferrin receptor insulin receptor junction protein claudin-5 plasma membrane marker Na+/K+ ATPase
plasma membrane fraction
whole cell fraction
7.16 ± 0.63b 0.742 ± 0.046 1.62 ± 0.07 3.87 ± 0.39 1.65 ± 0.23 0.311 ± 0.041 2.18 ± 0.06 74.4 ± 4.4 1.90 ± 0.23 1.87 ± 0.22b 5.94 ± 0.35 1.57 ± 0.06b
ULQ (
