MS

Accepted Manuscript Title: Optimization of sample preparation for transporter protein quantification in tissues by LC-MS/MS Authors: Skaidre Jankovskaja, Junichi Kamiie, Melinda Rezeli, Lena Gustavsson, Yutaka Sugihara, Tasso Milliotis, Tautgirdas Ruzgas, Gy¨orgy Marko-Varga PII: DOI: Reference:

S0731-7085(18)31665-0 https://doi.org/10.1016/j.jpba.2018.10.013 PBA 12262

To appear in:

Journal of Pharmaceutical and Biomedical Analysis

Received date: Revised date: Accepted date:

27-7-2018 3-10-2018 5-10-2018

Please cite this article as: Jankovskaja S, Kamiie J, Rezeli M, Gustavsson L, Sugihara Y, Milliotis T, Ruzgas T, Marko-Varga G, Optimization of sample preparation for transporter protein quantification in tissues by LC-MS/MS, Journal of Pharmaceutical and Biomedical Analysis (2018), https://doi.org/10.1016/j.jpba.2018.10.013 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

SHORT COMMUNICATION Optimization of sample preparation for transporter protein quantification in tissues by LC-MS/MS

IP T

Skaidre Jankovskaja1, 2, 3*, Junichi Kamiie1, 5*, Melinda Rezeli1*, Lena Gustavsson7, Yutaka Sugihara1,4, Tasso Milliotis6, Tautgirdas Ruzgas2, 3 and György Marko-Varga1,8

*

Equally contributed

SC R

1

ED

M

A

N

U

Clinical Protein Science and Imaging, Department of Biomedical Engineering, Lund University, Lund, Sweden; 2 Department of Biomedical Science, Faculty of Health and Society, Malmö University, Malmö, Sweden; 3 Biofilms –Research Center for Biointerfaces, Malmö University, Malmö, Sweden 4 Division of Oncology and Pathology, Dept. of Clinical Sciences, Lund University, Lund, Sweden; 5 Laboratory of Veterinary Pathology, School of Veterinary Medicine, Azabu University, Sagamihara, Kanagawa, Japan; 6 AstraZeneca R&D, Innovative Medicines, Mölndal, Sweden ; 7 Department of Drug Metabolism, H. Lundbeck A/S, Valby, Denmark; 8 Centre of Excellence in Biological and Medical Mass Spectrometry "CEBMMS", Biomedical Centre D13, Lund University, Lund, Sweden; *

Corresponding authors (name, address, telephone and fax numbers and email address)

A

CC E

PT

György Marko-Varga, Biomedical Center, BMC D13, SE-221 84 Lund, Sweden, Tel: +46-46222 37 21, Fax: + 46-46-222 45 27, [email protected]

1

Highlights Evaluation of drug transporter quantity is needed for building physiologically-based pharmacokinetic models (PBPK)



Transporter protein quantification methodology was improved by optimizing plasma membrane isolation and protein digestion procedures Established method shows good reproducibility (CV>10%) Our optimized method resulted in 2-3 times higher Bcrp and Na+/K+ATPase quantities in mouse liver and kidney’s cortex comparing to the previously reported amounts

 

IP T



SC R

ABSTRACT

U

Background. Reproducible quantification of drug transporter protein expression in tissues is important for predicting transporter mediated drug disposition. Many mass-spectrometry based transporter protein quantification methods result in high variability of the estimated transporter quantities. Therefore, we aimed to evaluate and optimize mass spectrometry-based quantification method for drug transporter proteins in tissues.

M

A

N

Materials and methods. Plasma membrane (PM) proteins from mouse tissues were isolated by applying three extraction protocols: commercial plasma membrane extraction kit, tissue homogenization by Potter-Elvehjem homogenizer in combination with sucrose-cushion ultracentrifugation, and PM enrichment with Tween 40. Moreover, five different protein digestion protocols were applied on the same PM fraction. PM isolation and digestion protocols were evaluated by measuring the amount of transporter proteins by liquid chromatographytandem mass spectrometry in selected reaction monitoring mode.

CC E

PT

ED

Results. Mouse liver homogenization by Potter-Elvehjem homogenizer in combination with sucrose-cushion ultracentrifugation and PM enrichment with Tween 40 resulted in two time higher transporter protein quantity (Breast cancer resistance protein (Bcrp) 18.0 fmol/µg protein) in comparison with the PM samples isolated by extraction kit (Bcrp 9.8 fmol/µg protein). The evaluation of protein digestion protocol revealed that the most optimal protocol for PM protein digestion is with Lys-C and trypsin, in combination with trypsin enhancer and heat denaturation. Overall, quantities of Bcrp and Na+/K+ATPase proteins evaluated in mouse liver and kidney cortex by using our optimized PM isolation method, as well as, established digestion protocol were two to three times higher than previously reported and coefficient of variation

(CV) for technical replicates was below 10 %.

A

Conclusion. We have established an improved transporter protein quantification methodology by optimizing PM isolation and protein digestion procedures. The optimized procedure resulted in a higher transporter protein yield and improved precision. Key words: transporter proteins, quantification, optimization, reproducibility, LC-MS/MS, Bcrp.

2

IP T SC R

1. INTRODUCTION

M

A

N

U

The concept that membrane transporter proteins facilitate the flux of molecules across eukaryotic cell membranes is widely accepted. Following this, it is recognised that transporter proteins play an important role in absorption, distribution, metabolism and excretion (ADME) of drugs [1]. Previous in vitro and clinical research have indicated that some drugs as well as disease states modify the expression of drug transporters in humans, which leads to altered pharmacokinetics and subsequently, changed pharmacology/toxicology [2,3]. Knowing the amount of drug transporter proteins in tissues is necessary for building physiologically-based pharmacokinetic models (PBPK) used in the prediction of pharmacokinetic profiles including drug-drug interaction [4–6]. Subsequently, reliable PBPK would improve justification of drug dosing. Thus, accurate quantitative methods of transporter protein quantification are required.

A

CC E

PT

ED

Quantitative proteomics using liquid chromatography-tandem mass spectrometry (LC-MS/MS) in selected reaction monitoring (SRM) mode has been increasingly applied for hepatic and intestinal transporter protein quantification [7]. However, up to date there is no standardized methodology for transporter quantification [8]. Although digested peptides from nearly any protein can be measured precisely by SRM, design of the overall method that gives an accurate transporter quantification is not guaranteed [9]. As in any bioanalytical assay, efficient sample preparation is very important for SRM analysis. It should be emphasised that drug transporter proteins are localised in the plasma membrane (PM) of the cell. Thus, isolated and purified PM fraction of the cells is theoretically the ideal material for transporter quantification. However, it has been shown that currently available methods tend to have poor performance in separating PM fraction from other cellular membranes [6]. Additionally, isolation of membrane proteins is hampered by their relative low abundance in total cell lysates, their frequently large size and their hydrophobic properties. Generally, different technologies focused on cell disruption, membrane solubilisation/precipitation, and enrichment need to be considered in order to obtain pure PM protein fraction [10]. Another key point for reliable drug transporter protein quantification is efficiency of enzymatic digestion. Many reports have found drastic differences in quantification when using different peptides, which is possibly due to different digestion efficiencies [11,12]. This problem arises from the difference in amount of solvent accessible protein regions which are more likely digested to completion [11,13]. Therefore, digestion efficiency issue could be solved by at least two ways: selection of the best target peptides or enzymatic digestion improvement. Both ways would result in better transporter protein quantitation accuracy and reproducibility, which relies 3

IP T

on degree of protein digestion. However, the best target peptide selection for more than 360 different drug transporter proteins [14] might be labour intense. As an alternative, a considerable effort has been devoted to increse the efficiency of enzymatic digestion for reducing the level of missed cleavage sites and thus the reproducibility of mass spectrometric measurements [15–17]. Typical digestion process involves (i) denaturation of proteins by denaturant like urea and/or heating, (ii) reduction of cysteine residues by dithiothreitol (DTT), (iii) alkylation of thiols by iodoacetamide (IAA) and (iv) protein digestion by addition of trypsin [13,18,19]. Thus, the main focus for efficient transporter protein digestion is balancing between facilitating proteolytic accessibility (e.g., partial protein digestion by endopeptidase Lys-C; denaturation conditions) without using conditions that interfere with subsequent MS analysis [12].

N

U

SC R

PM purity and digestion efficiency are important factors for accuracy of quantitation of transporter proteins, although evaluation of those are difficult. Theoretically it is obvious that low purity of PM and low digestion efficiency result in low quantitative values of determined expression levels of transporter proteins. So far, several studies of transporter protein quantification using LC-MS/MS have been reported, however, they have not stated the PM purity and digestion efficiency. Higher PM purity and digestion efficiency should increase the measured transporter protein quantity and reproducibility of the measurement, so, the method that show higher quantitative values of transporter proteins should be more accurate method for transporter quantitation. The method with pure PM and high efficiency of enzymatic digestion is necessary for accurate quantitation of transporter proteins.

PT

ED

M

A

In this work we have optimized a method for drug transporter protein quantification in tissues by targeting two issues: PM fraction purity and digestion efficiency, which are the main sources of variability in transporter protein quantification. We have managed to enrich the PM fraction and establish reproducible protein digestion protocol. Improved PM isolation and optimized digestion procedure resulted in high precision between replicates as well as higher quantitative values of transporter proteins when compared with previously reported values [14]. This paper summarises the optimized methodology for quantification of plasma membrane proteins in detail.

CC E

2. MATERIALS AND METHODS 2.1 Animals

A

Frozen mouse (C57BL/6; ~8 weeks) liver and kidney tissue was provided by Kenichi Miharada, Ph.D (Dept. Molecular Medicine and Gene Therapy, Lund Stem Cell Center, Lund University). Animals were housed and bred according to regulations for the protection of laboratory animals. The experimental procedures have been approved by the local ethical committee (M32-14). 2.2 Reagents and solutions The reagents, such as 2-Amino-2-(hydroxymethyl)-1,3-propanediol (Tris), NaCl, MgCl2, sucrose, ethylenediaminetetraacetic acid (EDTA), polysorbate 40 (Tween 40), ammonium bicarbonate (AMBIC), dithiothreitol (DTT), iodoacetamide (IAA) and formic acid, were purchased from Sigma Aldrich (St. Lois, MO, USA) if nothing else is indicated. Organic solvents were of LC-MS quality and delivered by Merck (Darmstadt, Germany). 4

2.3 Plasma membrane isolation Since liver is a major excretory organ that contributes to the elimination of drugs and metabolites [12], mouse liver was chosen for the development of drug transporter protein quantification method and its evaluation. Mouse kidney (cortex) was chosen for optimized method validity demonstration.

PT

ED

M

A

N

U

SC R

IP T

Plasma membrane isolation was performed by using three extraction protocols. In the first protocol, plasma membrane proteins were isolated according to the ProteoExtract Native Membrane Extraction Kit (Merck KGaA, Darmstadt, Germany) user protocol without any modifications. In the second method, PM isolation protocol was adopted from Kamiie and coworkers (2008) [14]. According to the protocol [14] mouse tissues (liver, kidney cortex) were dissected into 1mm pieces and homogenized in Homogenization buffer ((10mM Tris-HCl, 10mM NaCl, 1.5mM MgCl2, pH 7.4, 38% (w/v) sucrose solution) containing MS-SAFE Protease and Phosphatase Inhibitor Cocktail (Sigma Aldrich, St. Lois, MO, USA) (10 mL to 1 g tissue) using Potter-Elvehjem homogenizer. Homogenization step was performed on ice. All reagents and buffers were cooled down by placing it on ice prior the usage. The homogenate was centrifuged at 8,000 x g for 10 min at 4°C. Then, the supernatant was ultra-centrifuged at 100,000 x g for 60 min at 4°C. After the ultra-centrifugation, the pellet was suspended in Suspension buffer (10mM Tris-HCl, 250mM sucrose, 1mM EDTA, pH 7.4) and layered on top of 2 mL of Homogenization buffer. Then, the sample was ultra-centrifuged at 100,000 x g for 40 min at 4°C. The turbid layer at the interface was collected, suspended in the Suspension buffer and ultra-centrifuged at 100,000 x g for 40 min at 4°C. The precipitate (PM fraction) was re-suspended in 10mM Tris-HCl, pH 8.0. In the third protocol, the above described PM isolation procedure was improved by addition of 5 % w/w of Tween 40 into the Homogenization buffer, and then, agitation of the sample at 4°C for 1 hour. After that, the homogenate was centrifuged at 10,000 x g for 10 min at 4 °C, the supernatant was layered on top of Homogenization buffer and centrifuged at 100, 000 x g for 40 min at 4 °C. Then the third protocol was continued as the second protocol. For all protocols final protein concentration in the isolated PM fractions was measured by using PierceTM BCA Protein Assay Kit (Thermo Fisher Scientific, Rockford, USA). Detailed protocol for PM isolation by using sucrose cushion method can be found in the Supplementary material, S-1.

CC E

2.4 Protein digestion

A

Five protocols were tested and evaluated by using the same PM isolate (PM was isolated by using the second protocol as described above) from mouse liver. Several parameters were varied between the protocols: e.g. solubilisation agents, enzymes, protein reduction and alkylation time. Here we describe the optimized method only, because it performs better than the other tested protocols. Detailed description of all five tested protocols can be found in the Supplementary material, S-2. Aliquots containing 20 µg of PM proteins were diluted to 200 µL volume by adding 10 mM Tris-HCl (pH 8.0). In order to get rid of remaining Tween 40 in the sample, methanolchloroform precipitation (on ice) was performed. Precipitated pellet was solubilized with 20 µL of 0.2 % ProteaseMAX (Pmax) surfactant (Promega, Madison, WI) dissolved in 50 mM AMBIC, pH~8.0 by vortexing and agitating the sample for 30 min at room temperature. 5

Proteins were reduced by adding DTT to the sample (1µL of 6 mM DTT/1 µg of protein) and incubating it for 20 min at 56 °C. Disulphide bond reforming was prevented by alkylation with IAA (1 µL of 15 mM IAA/1 µg of protein) for 20 min at room temperature in the dark. Prior to protein digestion, samples were diluted 4 times with 25 mM of AMBIC, pH~8.0. Two steps proteins digestion was performed with mass spectrometry grade LysC (Wako Chemicals, Richmond, VA) and sequencing grade modified Trypsin (Promega, Madison, WI) at an enzyme/protein ratio of 1:20. Samples were incubated with LysC for 2 hours and 30 min at room temperature prior to trypsin digestion for over 12 hours at 37 °C. Digestion with trypsin was facilitated with addition of 1 µL of 1% Pmax.

IP T

2.5 Sample preparation for mass spectrometric analysis

U

SC R

Protease Max was degraded by incubating the sample for 5 min at 95 °C. The tryptic digests were acidified with 2 µL of 100% formic acid and the mixture (230 fmol) of high purity reference peptides (JPT Peptide Technologies, GmbH, Berlin, Germany) was simultaneously added to the sample. The solution was centrifugated at 15,000 x g for 15 min at 20 °C in order to remove non-solubilized material. The supernatant was desalted by Micro SpinColumnsTM Silica C18 (The Nest group Inc., South Borough) according to the manufacturer´s instructions. After off-line desalting the sample was dried in centrifugal evaporator and then re-suspended in 5% ACN/0.1% formic acid in water (aiming to have 1 µg/µL sample concentration).

N

2.6 Mass spectrometric analysis

CC E

PT

ED

M

A

The LC-MS/MS analysis in SRM analytical mode was carried out on an EASY n-LC II pump connected to TSQ Vantage mass spectrometer equipped with a Nanospray Flex ion source. The method was previously described by Rezeli and co-workers [20]. Shortly, prior to peptide separation, their pre-concentration was performed on an Easy C18-A1 pre-column (Thermo Scientific, Waltham, MA). Then, samples were loaded on an in-house packed analytical column (75 μm x 200 mm fused silica column packed with ReproSil C18, 3 μm, 120 Å from Dr. Maisch GmbH, Germany) and separated by using linear gradient made from solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). Elution profile at the flow rate of 300 nL/min was performed as follows: 10 % B was increased to 35 % B in 30 min and then to 90 % B during 5 min. The 90% B solvent flow was then kept for additional 5 min. For SRM acquisition Q1 and Q3 were operated at unit resolution (0.7 FWHM) and the collision gas pressure in Q2 was set to 1.2 mTorr. The cycle time was 3 s, and tuned S-lens value was used during data acquisition.

A

Quantification of peptides was based on the corresponding light and heavy peptide peak area ratios. Peak integration was performed automatically by using Skyline v3.5 software (MacCoss Lab Software, Seattle, WA). All data were manually checked to confirm correct peak assignment by the software. 2.7 Evaluation of PM isolation and digestion efficiency The overall strategy to quantify drug transporter proteins was based on nanoLC-MS/MS in SRM mode. To compare the quantitative values between this study and previous study, we selected breast cancer resistance protein (Bcrp) which is highly expressed in mouse liver [14] and Na+/K+ATPase as a plasma membrane marker, as target proteins. One unique target peptide that had been reported previously [14] was selected for each protein. 6

For quantification stable isotope dilution strategy was used; the samples were spiked with a known amount of heavy isotope labelled standard peptides. The quantity of Bcrp and Na+/K+ATPase were estimated by the normalized peak area ratios of their endogenous target peptides to isotope-labelled peptides (Figure 1) by using at least three transitions listed in Table 1. The heavy and endogenous peptide peak areas were measured, and the amount of target proteins were calculated based on the relationships presented below:

IP T

Amount of target protein (fmol/µg total membrane protein) = (endogenous/heavy peptide peak area ratio*reference peptide amount (fmol))/total amount of membrane proteins (µg)

SC R

where the amount of spiked isotope-labelled reference peptide and the total amount of protein in the membrane fraction were used to calculate the amount of target protein in the sample.

3. RESULTS and DISCUSSION

U

3.1 Plasma membrane isolation strategies

PT

ED

M

A

N

Since drug transporter proteins are located in cell PM, one of the most crucial parts in the sample preparation for transporter quantification is the isolation of pure PM fraction. Therefore, we investigated how drug transporter protein quantities varied between different plasma membrane isolation methods. As it is shown in Figure 1 and described in section 2.7, different PM isolation protocols were evaluated based on comparison of PM marker Na+K+ATPase and Bcrp transporter quantitative values. Three plasma membrane isolation methods were investigated: 1) commercially available plasma membrane extraction kit; 2) sucrose cushion method, where the sample was homogenized with Potter-Elvehjem homogenizer and then PM was isolated by ultracentrifugation in combination with sucrose cushion; and 3) sucrose cushion method supplemented with the use of surfactant, where the PM fraction was enriched by usage of the surfactant Tween 40. Obtained results are summarized and presented in Table 2.

A

CC E

The highest total amount of isolated PM proteins was achieved using commercially available PM extraction kit, where 2.7 mg of PM proteins were estimated (Table 2) by BCA assay. This is 25 times more of PM proteins in comparison with sucrose cushion method (0.1 mg of PM proteins). However, as it was previously reported extraction kits (for example, Abcam Phase Separation kit) claiming of capability to isolate pure plasma membrane may result in PM contaminated with other intracellular membranes [21]. Based on our results (Table 2) by using another PM isolation kit (ProteoExtract Native Membrane Extraction Kit) same problem seemed to occur. Two times higher quantitative values for the PM marker Na+/K+ATPase and Bcrp transporter were observed when PM proteins were isolated by sucrose cushion method in comparison with extraction kit (Table 2). As well as, coefficient of variation (CV (%)) for technical replicates was lower for PM proteins isolated with sucrose cushion (4.6 % for Na+/K+ATPase and 10.9 % for Bcrp) in comparison with extraction kit (6.4 % for Na+/K+ATPase and 23.7 % for Bcrp). Based on the results, it can be speculated that commercial ProteoExtract Native Membrane Extraction Kit isolates PM fraction contaminated with other proteins, for example, cytosolic proteins, membranes from intracellular organelles 7

SC R

IP T

(etc., mitochondria). Therefore, it seems reasonable to choose sucrose cushion as the method for transporter quantification, since this strategy results in a purer plasma membrane fraction and subsequently, higher transporter protein quantitative values and lower variability between replicates. However, a disadvantage with the sucrose cushion method is the relatively low yield, which may be a problem when limited amount of sample is available. To overcome this issue, we decided to enrich the PM fraction by introduction of surfactant into the PM isolation procedure. Herein, mouse liver homogenate was treated with mild surfactant (Tween 40) prior to ultra-centrifugation in combination with sucrose cushion. The results (Table 2) revealed that Tween 40 application can successfully increase PM fraction up to 11 times (1.1 mg of PM proteins) in comparison with simple sucrose cushion application (0.1 mg of PM proteins). Plasma membrane proteins obtained by sucrose cushion strategy supplemented with surfactant, shows similar quantitative values to the method without addition of Tween 40 (Table 2), which could be a proof of weight, that sucrose cushion method results in relatively pure PM fraction. In conclusion, the combination of sucrose cushion and PM enrichment with Tween 40 (Table 2, marked grey) was found to be the most efficient extraction protocol (from the three investigated protocols) for PM isolation for drug transporter protein quantification. 3.2 Digestion of plasma membrane proteins under various conditions

A

CC E

PT

ED

M

A

N

U

In addition to the isolation of a pure plasma membrane fraction, protein digestion efficiency is another key point for reproducible and efficient peptide-based quantification of transporter proteins [22]. In our study, we have focused on the improvement of protein digestion efficiency by facilitating protease access to the proteins during enzymatic digestion. The motivation to improve transporter protein digestion efficiency was based on previously conducted studies in our laboratory where one step digestion with trypsin (Tryp) was compared with two steps digestion with Lys-C and trypsin. To evaluate the efficiency of one step trypsin digestion and to compare it with the combined Lys-C/Tryp digestion, PM fractions isolated from mouse liver and digested with Tryp only or with Lys-C/Tryp were analysed using shotgun proteomic approach (non-targeted LC-MS/MS analysis). After running database search to identify peptides and proteins in the samples, the number of missed cleavage sites was also investigated and the two digestion methods were compared. Correctly cleaved and miss cleaved peptides was calculated and their amount was expressed in percentage. In samples digested with trypsin alone more than 50% of the identified peptides contained 1 or 2 missed cleavage sites, in contrary in sample digested with Lys-C and trypsin the proportion of peptides with missed cleavages was below 20 %. In addition, the amount of Na+/K+ATPase measured in these samples followed the trend that samples with less missed cleavages showed higher amount of Na+/K+ATPase (unpublished results). These results suggested that, Lys-C pre-digestion improves subsequent tryptic digestion efficiency by cleaving on the C-terminal side of lysine residues and generating smaller protein fragments, which are easier accessible by trypsin [23]. Even though, in the above described pilot study only the effect of Lys-C addition on conventional trypsin digestion was investigated, we believed that non-complete protein denaturation prior to enzymatic digestion may lead to tight folding of proteins and therefore limited susceptibility to trypsin and subsequently higher amount of peptides with missed cleavages, lower measured transporter protein quantities and lower reproducibility of technical replicates. Following these initial observation, five different digestion protocols were investigated (Table 3), where digestion efficiency was evaluated based on Na+/K+ATPase and Bcrp plasma 8

membrane protein quantitative values. In order to compare the reproducibility of different protocols, each protocol was carried out in six technical replicates (Table 3). All different digestion procedures were applied for the same mouse liver PM fraction isolated by using sucrose cushion method in combination with surfactant Tween 40. Not surprisingly, overall assessment of obtained results (Table 3) indicated that digestion conditions considerably influence peptide-based PM protein quantification results.

SC R

IP T

First, it was observed (Table 3, protocol 1) that the usage of a widely accepted protein digestion protocol with trypsin resulted in high variability (CV) between replicates for Na+/K+ATPase and Bcrp, 47 % and 11 %, respectively. In addition to decreasing the variability, the yield of the selected proteins was also significantly improved by the method optimization. Taking into consideration the quantity of Bcrp transporter, the observed value was two-fold higher (20.5±5 fmol/µg protein) with protocol 4 (Table 3) than with the traditional trypsin digestion (9.1±1 fmol/µg protein) (Table 3, protocol 1). Moreover, almost three times higher quantitative values for the PM marker Na+/K+ATPase (98.9±11 as compared to 33.3±16 fmol/µg protein) were observed after trypsin method supplementation with protein denaturation induction by heat and addition of Pmax surfactant (Table 3, protocol 4).

ED

M

A

N

U

Exploring the results more closely, it shows (Table 3, protocol 4) that the highest Na+/K+ATPase quantitaties were achieved by performing protein digestion with trypsin, together with the enhancer Pmax and protein denaturation by heat. For this reason, protocol 4 could be considered as the optimal method. However, for choosing the optimal protocol the assay reproducibility should also be considered. Therefore, we suggest the two step digestion with Lys-C and trypsin, as well as the usage of Pmax and heat denaturation as a procedure for transporter quantification in tissues (Table 3, protocol 5-grey colour). This method, shows CV (%) values lower than 10 % for both Na+/K+ATPase and Bcrp, 3.7 % and 7.3 %, respectively, which is an acceptable variation. In addition, the measured value for the same target peptide of Bcrp transporter protein (Table 1) is almost two times higher (17.3 fmol/µg protein; 7.3 CV(%) (Table 3, protocol 5)) than the previously reported (8.84 fmol/µg protein; 10.6 CV(%)) [14].

A

CC E

PT

In summary, the obtained results are in agreement with recently published paper, where authors have speculated that differences in the sample processing protocols, including the use of sequential enzymatic digestion with Lys-C and trypsin compared with digestion with trypsin alone, could add the value to the quantitative differences observed [24]. However, in the reported case, biological variability could also play a role on quantitative values of membrane proteins. In our studies, biological variability was eliminated, since all five digestion protocols were applied on the same PM isolate. Therefore, we are more confident in stating that observed variability in our studies is due to sample preparation procedures and not because of the biological variation. 3.3 Application of the optimized sample preparation protocol After the thorough examination of various sample preparation conditions, we aimed to show that our optimized preparation protocol is valid for other tissue samples as well. Therefore, PM fraction of mouse kidney cortex was enzymatically digested by using our optimized protocol (Table 3, protocol 5). The results showed that the quantity of Bcrp transporter in PM fraction isolated from renal cortex was 158.9±10 fmol/µg protein, which is almost 3 times higher by measuring the same target peptide, than it was previously reported (53.4±1.62 fmol/µg protein) 9

[14]. Additionally, a higher quantitative value was obtained for Na+/K+ATPase (907.3 ±21 fmol/µg protein) compared to 254±7.55 fmol/µg protein reported by [14]. CV (%) values for both two proteins were fairly low, 10.4% and 4.1%, respectively (n=3 for each determination). Even though we compare our obtained quantitative values with previously reported ones, we are aware of the fact that biological sample variability might have some impact on transporter protein quantification. 4. CONCLUSIONS

SC R

IP T

The scope of this study was to evaluate different sample preparation protocols for drug transporter protein quantification and establish an optimized method for transporter quantification in mouse tissues. We have managed to establish a reproducible methodology by optimizing PM isolation as well as protein digestion procedures. Since, absolute values for Bcrp or Na+/K+ATPase proteins are not known, in our current study we have established a method with very good precision with coefficient of variation (CV) for technical replicates below 10 %.

N

U

Lastly, we would like to point out that there is no single method that will work for every problem or for every sample type [25]. It is important to be aware of different sample preparation procedures and analyse sample preparation protocol thoroughly in order to achieve better results as it was attempted in this work.

A

Acknowledgement

ED

M

We all would like to Thank You Dr. Kenichi Miharada from Lund University for kindly providing mouse tissues.

J. Keogh, B. Hagenbuch, C. Rynn, B. Stieger, G. Nicholls, Membrane Transporters: Dundamentals, Function and Their Role in ADME, RSC drug Discovery Series No. 54, Drug Transporters: Role and Importance In ADME and Drug Development. 1 (2016) 3-45.

CC E

[1]

PT

References

A

[2]

I. Ali, J.R. Slizgi, J.D. Kaullen, M. Ivanovic, M. Niemi, P.W. Stewart, A.S. Barritt, K.L.R. Brouwer, Transporter-Mediated Alterations in Patients With NASH Increase Systemic and Hepatic Exposure to an OATP and MRP2 Substrate, Clin. Pharmacol. Ther. 0 (2018) 1–8. doi:10.1002/cpt.997.

[3]

J.D. Lutz, B.J. Kirby, L. Wang, Q. Song, J. Ling, B. Massetto, A. Worth, B.P. Kearney, A. Mathias, Cytochrome P450 3A Induction Predicts P-glycoprotein Induction; Part 2: Prediction of Decreased Substrate Exposure After Rifabutin or Carbamazepine, Clin. Pharmacol. Ther. 0 (2018) 1–9. doi:10.1002/cpt.1072.

[4]

S. Bosgra, E. Van De Steeg, M.L. Vlaming, K.C. Verhoeckx, M.T. Huisman, M. Verwei, H.M. Wortelboer, Predicting carrier-mediated hepatic disposition of rosuvastatin in man by scaling from individual transfected cell-lines in vitro using 10

absolute transporter protein quantification and PBPK modeling, Eur. J. Pharm. Sci. 65 (2014) 156–166. doi:10.1016/j.ejps.2014.09.007. A.T. Heikkinen, F. Lignet, P. Cutler, N. Parrott, The role of quantitative ADME proteomics to support construction of physiologically based pharmacokinetic models for use in small molecule drug development, Proteomics - Clin. Appl. 9 (2015) 732– 744. doi:10.1002/prca.201400147.

[6]

A.T. Heikkinen, N. Parrott, T. Dunkley, P. Cutler, Targeted Proteomics to Support Transporter IVIVE and PBPK. RSC Drug Siscovery Series No. 55, Drug Transporters: recent Advances and Emerging Technologies. 2 (2016) 44-72.

[7]

L. Wang, K.J. Rubadue, J. Alberts, D.W. Bedwell, K.J. Ruterbories, Development of a rapid and sensitive multiple reaction monitoring proteomic approach for quantification of transporters in human liver tissue, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 1061–1062 (2017) 356–363. doi:10.1016/j.jchromb.2017.07.051.

[8]

C. Wegler, F.Z. Gaugaz, T.B. Andersson, J.R. Wiśniewski, D. Busch, C. Gröer, S. Oswald, A. Norén, F. Weiss, H.S. Hammer, T.O. Joos, O. Poetz, B. Achour, A. Rostami-Hodjegan, E. Van De Steeg, H.M. Wortelboer, P. Artursson, Variability in Mass Spectrometry-based Quantification of Clinically Relevant Drug Transporters and Drug Metabolizing Enzymes, Mol. Pharm. 14 (2017) 3142–3151. doi:10.1021/acs.molpharmaceut.7b00364.

[9]

M.A. Gillette, S.A. Carr, Quantitative Analysis of Peptides and Proteins in Biomedicine by Targeted Mass Spectrometry, Nat Methods. 10 (2013) 28–34. doi:10.1038/nmeth.2309.

A

N

U

SC R

IP T

[5]

M

[10] S.M. Moore, S.M. Hess, J.W. Jorgenson, Extraction, Enrichment, Solubilization, and Digestion Techniques for Membrane Proteomics, J. Proteome Res. 15 (2016) 12431252. doi: 10.1021/acs.jproteome.5b01122. Epub 2016 Mar 24.

PT

ED

[11] B. Chen, L. Liu, H. Ho, Y. Chen, Z. Yang, X. Liang, J. Payandeh, B. Dean, C.E.C.A. Hop, Y. Deng, Strategies of Drug Transporter Quantitation by LC-MS: Importance of Peptide Selection and Digestion Efficiency, AAPS J. 19 (2017) 1469–1478. doi:10.1208/s12248-017-0106-4.

CC E

[12] L.M.B. Emi Kimoto, Membrane Protein Quantification by Peptide-Based Mass Spectrometry Approaches: Studies on the Organic Anion-Transporting Polypeptide Family, J. Proteomics Bioinform. 6 (2013) 229–236. doi:10.4172/jpb.S4-003.

A

[13] J.L. Proc, M.A. Kuzyk, D.B. Hardie, J. Yang, D.S. Smith, A.M. Jackson, C.E. Parker, C.H. Borchers, A quantitative study of the effects of chaotropic agents, surfactants, and solvents on the digestion efficiency of human plasma proteins by trypsin, J. Proteome Res. 9 (2010) 5422–5437. doi:10.1021/pr100656u. [14] J. Kamiie, S. Ohtsuki, R. Iwase, K. Ohmine, Y. Katsukura, K. Yanai, Y. Sekine, Y. Uchida, S. Ito, T. Terasaki, Quantitative atlas of membrane transporter proteins: Development and application of a highly sensitive simultaneous LC/MS/MS method combined with novel in-silico peptide selection criteria, Pharm. Res. 25 (2008) 1469– 1483. doi:10.1007/s11095-008-9532-4. [15] L.H. Betancourt, A. Sanchez, I. Pla, M. Kuras, Q. Zhou, R. Andersson, G. MarkoVarga, Quantitative assessment of urea in-solution Lys-C/trypsin digestions reveals superior performance at room temperature over traditional proteolysis at 37°C., J. 11

Proteome Res. (2018) acs.jproteome.8b00228. doi:10.1021/acs.jproteome.8b00228. [16] T. Glatter, C. Ludwig, E. Ahrné, R. Aebersold, A.J.R. Heck, A. Schmidt, Large-scale quantitative assessment of different in-solution protein digestion protocols reveals superior cleavage efficiency of tandem Lys-C/trypsin proteolysis over trypsin digestion, J. Proteome Res. 11 (2012) 5145–5156. doi:10.1021/pr300273g. [17] J.R. Wiśniewski, M. Mann, Consecutive proteolytic digestion in an enzyme reactor increases depth of proteomic and phosphoproteomic analysis, Anal. Chem. 84 (2012) 2631–2637. doi:10.1021/ac300006b.

IP T

[18] B. Prasad, J.D. Unadkat, Optimized Approaches for Quantification of Drug Transporters in Tissues and Cells by MRM Proteomics, AAPS J. 16 (2014) 634–648. doi:10.1208/s12248-014-9602-y.

SC R

[19] J.M. Burkhart, C. Schumbrutzki, S. Wortelkamp, A. Sickmann, R.P. Zahedi, Systematic and quantitative comparison of digest efficiency and specificity reveals the impact of trypsin quality on MS-based proteomics, J. Proteomics. 75 (2012) 1454– 1462. doi:10.1016/j.jprot.2011.11.016.

N

U

[20] M. Rezeli, K. Sjödin, H. Lindberg, O. Gidlöf, B. Lindahl, T. Jernberg, J. Spaak, D. Erlinge, G. Marko-Varga, Quantitation of 87 Proteins by nLC-MRM/MS in Human Plasma: Workflow for Large-Scale Analysis of Biobank Samples, J. Proteome Res. 16 (2017) 3242–3254. doi:10.1021/acs.jproteome.7b00235.

M

A

[21] V. Kumar, B. Prasad, G. Patilea, A. Gupta, L. Salphati, R. Evers, C.E.C.A. Hop, J.D. Unadkat, Quantitative transporter proteomics by liquid chromatography with tandem mass spectrometry: Addressing methodologic issues of plasma membrane isolation and expression-activity relationship, Drug Metab. Dispos. 43 (2015) 284–288. doi:10.1124/dmd.114.061614.

PT

ED

[22] Y. Uchida, M. Tachikawa, W. Obuchi, Y. Hoshi, Y. Tomioka, S. Ohtsuki, T. Terasaki, A study protocol for quantitative targeted absolute proteomics (QTAP) by LC-MS/MS: Application for inter-strain differences in protein expression levels of transporters, receptors, claudin-5, and marker proteins at the blood-brain barrier in ddY, FVB, and, Fluids Barriers CNS. 10 (2013) 1. doi:10.1186/2045-8118-10-21.

CC E

[23] S.J. Cordwell, T.E. Thingholm, Technologies for plasma membrane proteomics, Proteomics. 10 (2010) 611–627. doi:10.1002/pmic.200900521. [24] E. Therapeutics, Special Section on Transporters in Drug Disposition and Pharmacokinetic Prediction — Short Communication Comparison of Proteomic Quantification Approaches for Hepatic Drug Transporters : Multiplexed Global Quantitation Correlates with Targeted Proteomic Q, (2018) 692–696.

A

[25] M.H. Elliott, D.S. Smith, C.E. Parker, C. Borchers, Current trends in quantitative proteomics, J. Mass Spectrom. 44 (2009) 1637–1660. doi:10.1002/jms.1692.

12

Figure captions

(A)

IP T

(B)

U

SC R

(C)

A

CC E

PT

ED

M

A

N

Figure 1. Example of endogenous Bcrp protein quantification. A-C are SRM chromatograms of labelled (blue) and non-labelled (red) peptide signals obtained from plasma membrane (PM) fractions isolated from mouse liver by using: A- commercially available extraction kit; B-sucrose cushion method; C-sucrose cushion method supplemented with Tween 40. PM proteins were digested with two steps digestion protocol with Lys-C and trypsin, in combination with Protease Max surfactant (Protocol 3, Table 3).

13

Table

Table 1. Targeted peptide sequences and selected product ions for quantification of each membrane protein with nLC-MRM/MS. Membrane protein (Gene name)

Parent ion (++)

Peptide sequence

Bcrp (Abcg2)

Q64436

Na+/K+ATPase (Atp4a)

VDNSSLTGESEPQTR VDNSSLTGESEPQTR*

522.8 527.8

1 757.5 767.5

2 644.4 654.4

810.4 815.4

1004.5 1014.5

903.416 913.4

y series

3 430.3 440.3

4

717.4 727.4

501.3 511.3

y4, y5, y6

y4, y5, y6, y7

SC R

Q7TMS5

SSLLDVLAAR SSLLDVLAAR*

Product ions

IP T

Uni-Prot Accession No.

*Labelled amino acid

U

Table 2. Comparison of the plasma membrane isolation strategies from mouse liver tissue. Quantitative values of two membrane proteins.

2.7 0.1

Na+/K+ATPase

1.1

Bcrp Mean±SD (fmol/µg protein)

CV (%)

46.9±3 96.6±4

6.4 4.6

9.8±2 16.3±2

23.7 10.9

92.5±4

4.7

18.0±2

10.4

A

CV (%)

Mean±SD fmol/µg protein)

M

1. Extraction kit 2. Sucrose cushion 3. Sucrose cushion+Tween40

Total amount of isolated PM proteins (mg)

ED

Plasma membrane (PM) isolation strategy

N

Membrane protein amounts

PT

The amount of liver for each method was the same – 0.35 g. 10-20 µg of PM proteins were digested with two steps digestion protocol with Lys-C and trypsin, in combination with Protease Max surfactant (Protocol 3, Table 3). The amount of each protein was determined as average of three technical replicates. Each value presents mean ± standard deviation (n=3).

CC E

Table 3. Comparison of the digestion strategies used for membrane proteins isolated from mouse liver tissue.

A

Membrane protein digestion strategy (protocol)

1. Trypsin 2. Trypsin+Lys-C 3. Trypsin+Lys-C+Pmax 4. Trypsin+Pmax_heat 5. Trypsin+Pmax+Lys-C_heat

Membrane protein amounts Na+/K+ATPase Bcrp Mean±SD (fmol/µg protein)

CV (%)

Mean±SD (fmol/µg protein)

CV (%)

33.3±16 49.6±21 83.8±03 98.9±11 90.7±3

46.9 56.1 3.5 11.4 3.7

9.1±1 9.8±2 12.1±1 20.5±5 17.3±1

11.1 48.3 7.7 26.3 7.3

Samples were proceeded from the same initial PM fraction. Each value presents mean ± standard deviation (n=6; except Protocol 2 for both proteins, and Protocol 3 for Bcrp protein n=5). (Statistical significance evaluation can be found in supplementary material, S-3).

14