Quantitative targeted proteomics for understanding

2 downloads 0 Views 664KB Size Report
Apr 8, 2014 - Quantitative targeted proteomics for understanding the blood–brain barrier: towards pharmacoproteomics. Expert Rev. Proteomics Early online ...
Review

Expert Review of Proteomics Downloaded from informahealthcare.com by Kinokunya Co Ltd on 04/08/14 For personal use only.

Quantitative targeted proteomics for understanding the blood–brain barrier: towards pharmacoproteomics Expert Rev. Proteomics Early online, 1–11 (2014)

Sumio Ohtsuki*1, Mio Hirayama1, Shingo Ito1, Yasuo Uchida2, Masanori Tachikawa2 and Tetsuya Terasaki2 1 Department of Pharmaceutical Microbiology, Faculty of Life Sciences, Kumamoto University, 5-1 Oe-honmachi, Chuo-ku, Kumamoto 862-0973, Japan 2 The Division of Membrane Transport and Drug Targeting, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan *Author for correspondence: Tel.: +81 963 714 323 Fax: +81 963 714 329 [email protected]

informahealthcare.com

The blood–brain barrier (BBB) is formed by brain capillary endothelial cells linked together via complex tight junctions, and serves to prevent entry of drugs into the brain. Multiple transporters are expressed at the BBB, where they control exchange of materials between the circulating blood and brain interstitial fluid, thereby supporting and protecting the CNS. An understanding of the BBB is necessary for efficient development of CNS-acting drugs and to identify potential drug targets for treatment of CNS diseases. Quantitative targeted proteomics can provide detailed information on protein expression levels at the BBB. The present review highlights the latest applications of quantitative targeted proteomics in BBB research, specifically to evaluate species and in vivo–in vitro differences, and to reconstruct in vivo transport activity. Such a BBB quantitative proteomics approach can be considered as pharmacoproteomics. KEYWORDS: absolute quantification • blood–brain barrier • brain capillary endothelial cells • multiple reaction monitoring • pharmacoproteomics • quantitative targeted proteomics • transporter

The CNS is an important but challenging drug-target organ. Candidate CNS-acting drugs have the poorest success rate among all candidate drugs in clinical development [1], and development of more than 98% of such candidates has had to be discontinued because of insufficient permeability across the blood– brain barrier (BBB) [2,3]. Therefore, it is essential to understand the molecular mechanisms underlying BBB permeability in order to improve the success rate of drug development. The BBB is formed from brain capillary endothelial cells (BCECs) [2,3], which are connected to each other by tight junctions that restrict diffusion between the cells (FIGURE 1). Therefore, drugs have to cross the BCECs to enter the brain. Initially, researchers focused on increasing the lipophilicity of drug candidates to increase the BBB permeability, because simple diffusion of small molecules across plasma membrane of BCECs is dependent on lipophilicity. However, this idea did not work well due to the presence of multiple membrane transporters at the BBB.

10.1586/14789450.2014.893830

These transporters are membrane proteins that mediate transport of compounds across biological membranes [3–5]. At the BBB, various transporters are expressed at the blood side and the brain side of the plasma membrane. FIGURE 1C illustrates three modes of BBB transport. One is blood-to-brain influx transport, which supplies nutrients to the brain. This is mediated by solute carrier transporters, such as glucose transporter 1 (GLUT1), monocarboxylate transporter 1 (MCT1), creatine transporter and LAT, which transport glucose, ketone bodies, creatine and large neutral amino acids, respectively, into the brain. The second is the drug efflux pump, which serves to prevent entry of xenobiotics into the brain by actively pumping them out into the circulating blood. This efflux is mediated by members of the ATP-binding cassette (ABC) transporter family, which are expressed on the blood-side (luminal) membrane of BCECs. The third is brain-to-blood efflux transport, which acts to eliminate metabolites and neurotoxic compounds from brain interstitial fluid. This involves solute carrier transporters expressed on

 2014 Informa UK Ltd

ISSN 1478-9450

1

Review

Ohtsuki, Hirayama, Ito, Uchida, Tachikawa & Terasaki

B

Neuron

Astrocyte

A Brain capillary endothelial cell

Expert Review of Proteomics Downloaded from informahealthcare.com by Kinokunya Co Ltd on 04/08/14 For personal use only.

Blood

Tight junction

Pericyte

C

SLC transporters Blood-to-brain influx

Blood

Efflux pump ABC transporters Receptor-mediated transcytosis

Luminal membrane

Tight junction

ATP

ADP

The BBB acts as a barrier to exclude many drugs from the brain, and recognition by BBB transport systems greatly influences drug distribution into the brain. In addition, changes in transport function at the BBB affect homeostasis in the brain. For example, patients having loss-of-function mutation in GLUT1 or creatine transporter exhibit clinical features associated with CNS dysfunction such as mental retardation and seizures [6,7]. Therefore, an understanding of the BBB is critical for efficient development of CNS-acting drugs and drug delivery to the brain. As shown in TABLE 1, several proteomic analyses have been reported in recent years, and the findings have had a great impact on BBB research. In this review, we focus on recent quantitative targeted proteomic approaches for understanding the BBB from the viewpoint of drug development (pharmacoproteomics) including technical issues and analytical strategies to understand the human BBB. Aims of quantitative targeted proteomics in BBB research

Comprehensive global proteomics is an important approach in BBB research, as well as other research fields, to identify unknown proteins (TABLE 1A). Standard relAbluminal ative quantitative proteomics, including membrane 2D electrophoresis-based proteomics, also provide important information about Brain pathological changes in protein expression Brain-to-blood efflux in various diseases (TABLE 1B). From the viewpoint of drug development, it is necFigure 1. Structure and transport modes of the blood–brain barrier (BBB). essary to understand and predict BBB (A) Capillaries in rat brain. Brain capillaries were visualized in cerebral cortex by permeability of drugs in humans, espeendothelial-specific expression of green fluorescent protein under Tie2-promoter in cially patients [8]. Therefore, the initial Tie2/green fluorescent protein transgenic rat [62]. Scale bar: 80 mm. (B) Schematic aim of BBB pharmacoproteomics was to representation of the brain capillaries and surrounding neural cells. The BBB is formed by establish what kinds of molecules, especomplex tight junctions connecting the brain capillary endothelial cells. Tight junctions prevent paracellular diffusion between blood and brain. (C) Transport modes across the cially transporters and receptors, are BBB. SLC transport mediates blood-to-brain influx transport and brain-to-blood efflux expressed in the human BBB, because transport. ABC transporter is an efflux pump, which pumps out drugs from brain most of the early work at the molecular capillary endothelial cells to blood. Receptor-mediated transcytosis mediates transport level had been done in animals. The next of macromolecules across the BBB. aim was to clarify the amounts of moleABC: ATP-binding cassette; SLC: Solute carrier. cules expressed at the BBB of human and animals, as well as in cell lines, in order to the brain-side (abluminal) membrane of BCECs. The BCECs elucidate species and in vivo–in vitro differences. This is imporalso exhibit receptor-mediated transcytosis, which mediates trans- tant because BBB permeability is usually examined in animals port of proteins, such as influx transport of insulin and transferrin, and BBB model cell lines, since human studies are ethically and efflux transport of amyloid b-peptide [5]. These systems all problematic. The third aim was to understand molecular funcfunction cooperatively to support and protect the CNS. tions at the human BBB. The strategy to achieve this last goal doi: 10.1586/14789450.2014.893830

Expert Rev. Proteomics

informahealthcare.com

Primary cultured cells

Isolated brain capillaries

Primary cultured cells

Isolated brain capillaries

Primary cultured cells

Deracinois et al. (2012)

Chun et al. (2011)

Pottiez et al. (2010)

Lu et al. (2008)

Lu et al. (2007)

Primary cultured cells

Primary cultured cells

Isolated brain capillaries

Primary cultured cells

Primary cultured cells

Primary cultured cells

Immortalized cells

Primary cultured cells

Primary cultured cells

Onodera et al. (2013)

Pottiez et al. (2011)

Bergerat et al. (2011)

Ning et al. (2011)

Minagar et al. (2009)

Pottiez et al. (2009)

Haqqani et al. (2007)

Haseloff et al. (2006)

Franze´n et al. (2003)

Isolated brain capillaries

Isolated brain capillaries

Isolated brain capillaries

Isolated brain capillaries

Uchida et al. (2011)

Shawahna et al. (2011)

Uchida et al. (2011)

Kamiie et al. (2008)

Mouse

Human

Human

Mouse

Monkey

Mouse

Human

Rat, marmoset

Mouse

Human

Rat

Rat

Bovine

Human

Human

Rat

Bovine

Human

Bovine

Rat

Mouse

Bovine

Mouse

Bovine

Human

Species

MRM

MRM

MRM

MRM

MRM

MRM

MRM

MRM

MRM

2DE

2DE

2DE/ICAT

2DE

2DE

1D SDS–PAGE

1D PAGE, LC shotgun

2D DIGE

2D DIGE

2D DIGE

2DE

1D–PAGE, LC-shotgun

2DE, LC shotgun

LC shotgun

LC shotgun, off-line LC–MALDI–TOF

1D SDS–PAGE, LC-shotgun

Method

[13] [22]

Comparisons with liver and kidney

[12]

[9]

[21]

[35]

[38]

[20]

[10]

[61]

[60]

[59]

[58]

[57]

[56]

[55]

[54]

[53]

[52]

[51]

[16]

[50]

[49]

[48]

[47]

Ref.

Human BBB (membrane protein)

Human BBB (membrane protein and metabolizing enzymes)

Reconstruction of P-gp in vivo function

Species differences and developmental changes

Changes in KO mouse

In vitro–in vivo differences

Species differences

Mouse strain (ddY, FVB and C57BL)

Effect of TNF-a

Effect of hypoxia and reoxygenation

Effect of ischemic conditions

Effect of coculture with glial cells

Effect of glutamate

Effect of oxidative stress

Prestroke changes

Effect of coculture with glial cells

Effect of edaravone

Effect of coculture with glial cells

Isolated by immunolaser capture microdissection

Co-culture with glial cells

Co-culture with glial cells

Exosomes (extracellular microvesicles)

Note

2D DIGE: 2D fluorescence difference gel electrophoresis; 2DE: 2D electrophoresis; BBB: Blood–brain barrier; LC, Liquid chromatography; MRM: Multiple reaction monitoring.

Isolated brain capillaries

Ito et al. (2011)

Immortalized cells

Ohtsuki et al. (2013)

Isolated brain capillaries

Isolated brain capillaries

Hoshi et al. (2013)

Agarwal et al. (2012)

Isolated brain capillaries

Uchida et al. (2013)

C. Quantitative targeted proteomics

Primary cultured cells

Deracinois et al. (2013)

B. Quantitative global proteomics

Immortalized cells

Sample

Haqqani et al. (2013)

A. Global proteomics

Study (year)

Table 1. Recent proteomic analyses of the blood–brain barrier.

Expert Review of Proteomics Downloaded from informahealthcare.com by Kinokunya Co Ltd on 04/08/14 For personal use only.

Quantitative targeted proteomics of the blood–brain barrier

Review

doi: 10.1586/14789450.2014.893830

Review

Ohtsuki, Hirayama, Ito, Uchida, Tachikawa & Terasaki

Frozen human brain

Mince

Expert Review of Proteomics Downloaded from informahealthcare.com by Kinokunya Co Ltd on 04/08/14 For personal use only.

Homogenize with Potter-Elvehjem homogenizer (20 up-and-down unrotated strokes by hand)

Centrifuge the homogenate containing 17.5% dextran for 15 min at 4500 g (enrichment of vessels) Pellets Remove large vessels

Vessels

Φ100–210 μm nylon mesh

Φ20–45 μm nylon mesh

Microvessels

100 μm

Figure 2. Isolation procedure of brain capillaries from frozen human brain. The procedure was described in detail by Uchida et al. [13].

has been reconstruction of in vivo function based on expression amount and intrinsic activity as described later [8,9]. For the second and third aims, quantitative targeted proteomics using multiple reaction monitoring (MRM) is a key methodology not only because of its high sensitivity, but also because it can provide accurate absolute protein expression information [10]. Brain capillaries as materials for BBB proteomics

The cellular volume of brain capillaries is only 0.1–0.2% of the entire brain volume [11], and the brain capillaries are ramified forming a network in the cerebrum at intervals of about 40 mm. Isolated brain capillaries (microvessels) are the primary target for BBB proteomics (TABLE 1). Human brain capillaries are generally isolated from frozen brain. For example, brain capillaries were isolated from 1 to 3 g of frozen human brain by a nylon mesh method (FIGURE 2), and recovery of capillaries was about 200–300 mg protein/g brain [12,13]. The isolated samples predominantly contained microvessels as shown in FIGURE 2. Nevertheless, it is important to consider contamination by neural cells doi: 10.1586/14789450.2014.893830

because the expression of marker proteins of neurons (synaptophysin), astrocytes (GFAP), oligodendrocytes (CNPase) and pericytes (NG2 proteoglycan) was detected in isolated capillaries by means of quantitative targeted proteomics [13]. To obtain highly purified BCECs, we have isolated the cells by means of magnetic cell sorting with antibody against BCEC-specific antigen, CD31/PECAM-1 [14,15]. However, the purified BCECs were considered unsuitable for membrane proteomic analysis, since the isolated brain capillaries were treated with protease mixture to disperse the cells and membrane proteins might be truncated by protease. Lu et al. reported immunolaser capture microdissection (LCM) of mouse brain capillaries [16]. They visualized brain capillaries with anti-CD31 antibody and collected capillaries with 50,000 LCM shots. Finally, 881 proteins were identified in the collected brain capillaries. LCM is a useful methodology to obtain brain capillaries with high purity from small amounts of brain and also to analyze heterogeneous expression of proteins at the BBB in different brain regions. Interspecies comparison of BBB constituents

Species differences, especially between human and animals, are of critical importance in preclinical and clinical studies during drug development. To date, comparisons have mainly been done at the mRNA level [17,18]. However, protein level is a better parameter than mRNA level to clarify interspecies differences as it should better reflect functional differences [19]. Furthermore, quantitative targeted proteomics can determine absolute amounts of molecules so that protein expression data from different sources can be compared. Therefore, quantitative targeted proteomics provides rich information about interspecies differences and in vivo–in vitro differences. We have already reported quantitative targeted proteomics of brain capillaries of human, monkey, marmoset, rat and several different strains of mouse [12,13,20–22]. In this review, we focus on comparisons of transporter expression at the BBB among human, monkey and ddY mouse (FIGURE 3). MDR1/mdr1a/ABCB1 is a major efflux pump for multiple hydrophobic drugs and functions to pump drugs out of BCECs into the circulating blood, attenuating drug distribution to the brain (FIGURE 1C). As shown in FIGURE 3A, the protein expression level of MDR1 was not significantly different between brain capillaries of human and monkey. In contrast, in human, it was 43% of that in mouse (FIGURE 3B). The lower protein expression of MDR1 in human and monkey brain capillaries would lead us to predict higher brain distribution of MDR1 substrates in human and monkey than in mice. This prediction agrees with the results of the previous PET study, which showed that brain penetration of [18F]altanserin and [11C]GR205171, substrates of MDR1, in human was 4.5- and 8.6-fold greater than in rodents, respectively [23]. The brain penetration of [11C]verapamil and [11C]GR205171 in monkey was 4.1- and 2.8-fold greater than in rodents, respectively. In contrast, the protein expression of breast cancer resistance protein (bcrp)/ABCG2, which is also a drug efflux transporter at Expert Rev. Proteomics

Review

Quantitative targeted proteomics of the blood–brain barrier

1:3 Protein expression level in mouse brain capillaries (fmol/mg protein)

3:1 GLUT1

10 MDR1 INSR OATP-A ENT1

1

LRP1 GLUT3, 14 MCT1

MRP4 LAT1 0.1 1

10

100

Na+/K+ -ATPase

CD147 10

BCRP

1

MRP4 ENT1

LAT1 INSR LRP1

ABCA8/abca8a ABCA8/abca8b ABCA8/abca9

ABCA1 ABCA2 GLUT3, 14

0.1 1

10

100

1000

Protein expression level in human brain capillaries (fmol/mg protein)

Ratio of protein expression levels (mouse/human)

ULOQ in monkey

3:1

GLUT1

10

1

0.1

1:3

3:1

O A O TP AB O ATP-A/ CA AT - oa 1 P- B/o tpF/ at 2 oa ptp 2 O 14 AS AT C 3 TA T2 C UT D 1 M 47 R P LA 4 AB ENT1 ABCA I T1 C 8/a NS AB A8 bc R C /ab a8b A8 ca /a 8a bc LR a9 M P1 C T T 1 G AB fR LU C 1 T A M 3,1 4F2 2 D 4 h R /g c 1/ lu N a + md t3 /K + B r1 a AT CR Pa P G s LU e T1

ULOQ in human

ULOQ in human

ULOQ in human

ULOQ in human

ULOQ in monkey ULOQ in monkey ULOQ in monkey

ULOQ in human

AB O CA AT 1 O PAT A O PAT B P O -F AT M 3 LA R T1 P4 /la EN t1 T IN 1 S LR R P G MC 1 LU T T3 1 , M 14 N D a+ R /K + BC 1 AT RP P G ase LU T1

0.1

ULOQ in monkey

1

1:3

3:1

MCT1 4F2hc MDR1 TfR1 γ-GTP

0.1

1000

Protein expression level in human brain capillaries (fmol/mg protein)

10

100

1:1

ULOQ in mouse

BCRP

Na+/K+ -ATPase

TAUT OATP-F OATP-A/oatp-2 OATP-B/oatp-2 OAT3 ASCT2

ULOQ in mouse

100

1000

ULOQ in mouse ULOQ in mouse ULOQ in mouse

1000

ULOQ in human ULOQ in human ULOQ in human ULOQ in human ULOQ in human ULOQ in human ULOQ in human ULOQ in human

Protein expression level in monkey brain capillaries (fmol/mg protein)

B 1:3 1: 1

0.1

Ratio of protein expression levels (monkey/human)

Expert Review of Proteomics Downloaded from informahealthcare.com by Kinokunya Co Ltd on 04/08/14 For personal use only.

A

Figure 3. Comparison of protein expression levels in brain capillaries between human and monkey and human and mouse. (A) human – monkey and (B) human – mouse. Upper panels: comparison of absolute protein expression levels of membrane proteins between humans and animals. The solid line passing through the origin represents the line of identity, and the broken lines represent threefold differences. Each point represents the mean ± SD. The molecules on the horizontal (mouse, monkey) or vertical (human) axis are below the limits of quantification. Lower panels: ratio of protein expression levels of membrane proteins in animals to those in humans. The broken lines represent threefold differences. Each bar represents the mean ± SD. The molecules were ordered according to their expression levels. ULOQ means that the expression was under the limit of quantification in the indicated brain capillaries. INSR: Insulin receptor; TfR: Transferrin receptor. The plots are modified from [8].

the BBB, was 1.74-fold greater in monkey, but 1.85-fold lower in mouse, compared with human brain capillaries. The transporters located on the X- and Y-axis in FIGURE 3B are detected only in mouse and human brain microvessels, respectively. In mouse brain microvessels, organic anion transporters OAT3, oatp1a4/oatp2 and oatp1c1/oatp14 (1.97, 2.11 and 2.41 fmol/mg protein, respectively) were quantified [22]. In contrast, these organic anion transporters were below the limits of quantification in human (