Subscriber access provided by Kaohsiung Medical University
A Peptide-Based Sandwich Immunoassay for the Quantification of the Membrane Transporter Multidrug Resistance Protein 1 Oliver Poetz, Theresa Dieze, Helen S. Hammer, Frederik Weiss, Cornelia Sommersdorf, Markus F Templin, Christina Esdar, Astrid Zimmermann, Stefan Stevanovi#, Jens Bedke, Arnulf Stenzl, and Thomas O. Joos Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00152 • Publication Date (Web): 23 Mar 2018 Downloaded from http://pubs.acs.org on March 26, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
A PeptidePeptide-Based Sandwich Immunoassay for the Quantification of the Membrane Transporter Multidrug Resistance Protein 1 Oliver Poetz1,2, Theresa Dieze1, Helen Hammer1,2, Frederik Weiß1,2, Cornelia Sommersdorf1,2, Markus F. Templin1, Christina Esdar3, Astrid Zimmermann3, Stefan Stevanovic4, Jens Bedke5, Arnulf Stenzl5 and Thomas O. Joos1,2 1
NMI Natural and Medical Sciences Institute at the University of Tuebingen, Markwiesenstr. 55, Reutlingen, Germany SIGNATOPE GmbH Markwiesenstr. 55, Reutlingen, Germany 3 Merck KGaA, Frankfurter Str. 250, 64293 Darmstadt, Germany 4 Eberhard Karls University, Department of Immunology, Tuebingen, Germany 5 Eberhard Karls University, Department of Urology, Tuebingen, Germany 2
ABSTRACT: Multi-transmembrane proteins are notoriously difficult to analyze. To date, rapid and cost-efficient detection methods are lacking and only mass spectrometry-based systems allow reliable quantification of these proteins. Here, we present a novel type of sandwich immunoassay that is capable of sensitively detecting multidrug resistance protein 1 (MDR1), a prototypic 12transmembrane-domains transporter. In a first assay step, complex samples are enzymatically fragmented into peptides as routinely done for mass spectrometry. A proteotypic peptide derived from MDR1 was chosen and antibodies targeting this peptide were used to build a sandwich immunoassay. Validation of the optimized assay showed good sensitivity, reproducibility and it allowed reliable quantification of MDR1; cross-validation by mass spectrometry demonstrated the applicability for routine analyses in clinical and pharmaceutical research. MDR1 was quantified in primary human renal cell carcinoma and corresponding normal tissue, and down-regulation or expression loss was found in tumor tissue corroborating its importance in drug resistance and efficacy.
The most widely applied approaches for detecting transporter proteins are Western blotting 1 and immunohistochemistry. 2 In recent years mass spectrometry-based assays have been developed that use tryptic peptides as surrogates for membrane transporters. 3-6 These assays show significant advantages over Western blotting, since mass spectrometry-based approaches provide unambiguous specificity and allow accurate quantification by using isotope labeled standards. Recently, we developed such an MSbased assay that employs immunoaffinity enrichment of peptide fragments derived from cytochrome P450 3A4, 3A5, 3A7, and multidrug resistance protein 1 (MDR1). 7 In this multiplex assay, the immunoaffinity step is introduced into sample preparation and thereby sample throughput and sensitivity is increased. Nevertheless, this advanced assay system does not match the speed and low cost of a classical sandwich immunoassay. Usually, in the sandwich configuration antibodies are used to capture and detect an intact protein of interest. However, the generation of specific antibodies is not easily accomplished for many membrane proteins and multi-transmembrane proteins like transporters present especially problematic targets. Therefore, this assay concept is hardly applicable for the detection and quantification of these proteins. Rauh-Adelman and co-workers employed a modified concept for setting up sandwich immunoassays. 8 In their approach, the sandwich antibodies are not generated against the whole protein, but against one peptide that is released form the target protein by enzymatic digestion. They established several sandwich immunoassays for different proteins in the EGFR/MAPkinase pathway by capturing and detecting a proteotypic peptide that originated from the proteins of interest using a standard sandwich assay configuration. This allows the determination of the total amount of a kinase and when a suitable phospho-specific antibody is available, the degree of phosphorylation of a protein can be determined. Here, we employed a variation of this concept and show that peptide-centric sandwich immunoassays are capable of detecting and quantifying notoriously difficult protein such as the 12-
transmembrane domain protein MDR1 (also termed permeability glycoprotein 1, P-gp 1). MDR1 is a key transporter in drug disposition and an ATP-dependent export pump. Drugs like doxorubicin, daunorubicin, paclitaxel, colchicine, and imatinib are substrates of this transporter protein. MDR1 is expressed in many organs sustaining the barrier function of cells and tissues. Moreover, mechanisms of acquired tumor drug resistance based on the up-regulation of MDR1 have been described. 9,10 As a consequence, MDR1 is a pharmaceutical target and combined treatments of MDR1 inhibitors with chemotherapeutic agents shall overcome respective drug resistant tumor variants. 11 In a recent finding MDR1 expression has been associated with crizotinib and ceritinib resistance 12, two novel anaplastic lymphoma kinase (ALK) inhibitors approved for treatment of non-small cell lung cancer (NSCLC). 13,14 Thus, the transporter bears the potential of a clinical decision biomarker for non-small lung cancer. Consequently, an easy, fast and reliable method for the quantification of MDR1 would be of merit in pharmaceutical and clinical settings. Hence, we developed a capture antibody towards the Nterminal end of the MDR1-specific peptide EANIHAFIESLPNK and used a motif-specific antibody towards LPNK as detector for establishing a MDR1 sandwich immunoassay. Finally, we used the test to investigate the presence of MDR1 in xenograft and kidney tumor tissue.
EXPERIMENTAL SECTION Generation of xenografts. The in vivo experiment was performed in accordance with the German animal welfare regulations and approved by the Regierungspräsidium Darmstadt, Hessen, Germany (protocol registration number DA 4/211). The human ovarian cancer cell line A2780 (ECACC # 93112519) and the Adriamycin (doxorubicin) resistant cell line A2780ADR (ECACC #93112520) were obtained from the European Collection of Authenticated Cell Cultures and cultivated according to the supplier. 5x106 cells were injected s.c. in the right flank of HSD athymic mice. 15 days after injection tumors were harvested at a size of
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
averagely 500 mm3. Tumor pieces of approximately 100 mg were shock frozen in liquid nitrogen. Renal cell carcinomas. Kidney tumors and surrounding normal tissues were obtained from the Clinic for Urology, University Hospital Tuebingen. The local ethical committee approved the study (No. 438/2004V), and informed consent was obtained from the patients. Sample preparation. Extraction of proteins from tissue samples was performed in a ball mill (Sartorius) as described earlier. 15 Enzymatic sample fragmentation. Enzymatic fragmentation was performed as follows. 60 µg extracted protein was reduced in 5 mM tris(2-carboxyethyl)phosphine (SIGMA) in 50 mM triethanolamine (SIGMA), pH 8.5, 0.5% n-octyl-beta-Dglucopyranoside (SIGMA). Samples were denatured for 5 min at 99°C and cooled down to room temperature. Iodoacetamide (SIGMA) was added for alkylation to give a final concentration of 10 mM. After an incubation for further 20 min at RT trypsin (Promega) was added in a protein: enzyme ratio of 1:20. Proteolysis was performed for 1h with continuous mixing (650 rpm) in a temperature-controlled mixer (Eppendorf). Enzymatic reaction was stopped by a 5 min heating step at 99°C and adding PMSF to a final concentration of 1 mM. Antibodies and Peptides. Rabbit polyclonal antibodies were generated towards the peptide epitopes NH2-EANIHA and LPNK-COOH as described previously. 16 All synthetic peptides were purchased from INTAVIS. Peptide-epitope characterization. Antibody binding and characterization of the recognized epitopes were analyzed by liquid chromatography coupled to high-resolution mass spectrometry as described previously. 7 The analysis was performed using 20 µg lysate from cultured HEPG2 cells. Mass spectrometry-based immunoassay. MDR1 was quantified by a mass spectrometry-based immunoassay as described previously. 7 Microsphere-based sandwich immunoassay for MDR1. The MDR1 specific immunoassay was established on a bead-based assay platform (Luminex xMAP, Luminex Corp.) using standard protocols. Polyclonal antibodies recognizing the peptide epitope NH3-EANIHA were covalently immobilized onto color-coded magnetic polystyrene microspheres as described previously (Carson and Vignali, 1999). Proteolytically fragmented samples were incubated with 2000 capture microspheres in a 96-well low binding plate (Nunc) in a temperature-controlled plate shaker (Eppendorf) for 2 h at 25 °C at 650rpm. Two wash steps with 100 µl PBS 0.1% Tween each followed. 30 µL biotinylated anti-LPNK antibody was used for the detection step at a concentration of 1 µg/ml for another 60 min. After two further washing steps sandwich complexes were visualized by incubation with phycoerythrinconjugated streptavidin (2.5 µg/ml, Prozyme) for 45 min at 25 °C; data acquisition in a FlexMap3D instrument was performed according to the manufacturer’s instructions. The purified synthetic peptide EANIHAFIESLPNK was used as standard for quantification. MDR1 concentration was calculated according to a 4parametric logarithmic fit using the fluorescence data generated with a 7-point calibration curve ranging from 60 nM down to 90 pM. Material for assay validation. Xenograft tumors, A2780 and A2780ADR were lysed as described previously 17 and digested as described above. Characterization of the A2780-derived tumors by the mass-spectrometry based assay system showed no detectable signal for the surrogate peptide indicating that this complex sample is suitable for use in assay validation.
RESULTS
Page 2 of 8
Selection of Surrogate Peptide. The peptide EANIHAFIESLPNK was identified as a proteotypic peptide for the MDR1 protein and has been used as a surrogate in a previously described mass spectrometry-based immunoassay 7; Weiß et al. used this peptide for monitoring MDR1 induction in an experiment with statin-treated human hepatocytes. The 14-mer peptide is released from MDR1 after enzymatic fragmentation with trypsin and spans the amino acid sequence of the protein from position 1151 to 1164 (Fig. 1A). This area is located in the cytosolic domain of the transporter and neither a post-translation modification nor a single nucleotide polymorphism are known for this sequence part. 18 The peptide is unique and thereby proteotypic; this is a prerequisite for peptides employed as protein surrogates after tryptic fragmentation. 19
A
B
MDR1 sequence aa 1000-1280
Tryptic fragments area aa 1094-1188
... AKISAAHIIMIIEKTPLIDSY STEGLMPNTLEGNVTFGEVVF NYPTRPDIPVLQGLSLEVKKG QTLALVGSSGCGKSTVVQLLE RFYDPLAGKVLLDGKEIKRLN VQWLRAHLGIVSQEPILFDCS IAENIAYGDNSRVVSQEEIVR AAKEANIHAFIESLPNKYSTK VGDKGTQLSGGQKQRIAIARA LVRQPHILLLDEATSALDTES EKVVQEALDKAREGRTCIVIA HRLSTIQNADLIVVFQNGRVK EHGTHQQLLAQKGIYFSMVSV QAGTKRQ
... VLLDGK EIK R LNVQWLR AHLGIVSQEPILFDCSIAENIAYGDNSR VVSQEEIVR AAK EANIHAFIESLPNK YSTK VGDK GTQLSGGQK QR IAIAR ...
Capture antibody anti-EANIHA
Detector antibody anti-LPNK
EANIHAFIESLPNK
Fig. 1. Epitope selection for the peptide-based sandwich assay for detection MDR1. (A) The proteotypic peptide EANIHAFIESLPNK is used as surrogate peptide for the protein MDR1. It originates from MDR1 and is formed upon enzymatic fragmentation using trypsin. The sequence area aa 1000-1280 and tryptic peptide sequences surrounding the surrogate peptide are shown. (B) Anti-NH2-EANIHA antibody is immobilized on color-coded magnetic beads and used for the immunoprecipitation. Anti-LPNK-COOH is used as the detector in the sandwich assay configuration.
For the establishment of a peptide-centric sandwich immunoassay we generated an antibody towards NH2-EANIHA. For antibody characterization proteolytically fragmented proteins from HEPG2 hepatoma cells were used. After immunoprecipitation captured peptides were analyzed by LC-MSMS and a total of 10 peptides were found to be enriched in the immunoprecipitate. The peptide with the highest peptide spectra matches (PSMs) observed was the MDR1 peptide. Five peptides showed no sequence similarities and four shared the three-amino acid motif EAN. As a second antibody for the proteotypic peptide an existing antibody that targets the C-terminal motif LPNK-COOH was employed. This antibody is capable of enriching 88 different peptides from a complex sample including the MDR1 peptide. 7 The fact that only the chosen MDR1 peptide is recognized by both antibodies indicates that these could be used as a sandwich immunoassay pair. Since the anti EANIHA antibody showed a higher selectivity we decided to use this N-terminus-specific antibody as capture and the antibody towards the C-terminus LPNK as detector (Fig. 1B).
ACS Paragon Plus Environment
2
Characteristics of the Microsphere-Based MDR1 Sandwich Immunoassay. For the initial estimation of the detection and quantification limits the synthetic EANIHAFIESLPNK peptide was spiked into standard immunoassay buffer and the blank-based method was employed. 20 Lower limit of detection (background signal + 3 standard deviations) was estimated to be 74 pM and the lower limit of quantification (background signal + 10 standard deviations) was found to be 80 pM when using standard immunoassay buffer (Candor Bioscience). For mimicking a complex biological sample, we took advantage of the availability of a well characterized tumor cell line; A2780 is an ovarian cancer cell line that lacks expression of MDR1. Therefore, this sample was used as matrix, as it presents a suitable background control for the absence of MDR1. Hence, the standard peptide was spiked into enzymatically fragmented A2780 lysate. Final concentration of spiked peptide was adjusted from 60 nM down to 90 pM in 10 µg of the matrix. Results correlated well between buffer and cell line - derived matrix, but lower limit of detection was estimated to be 110 pM and the lower limit of quantification (background signal + 10 standard deviations) was calculated to be 120 pM (Fig. 2A). Sensitivity in artificial matrix was slightly better than in the biological matrix. Precision analysis of Microsphere-Based MDR1 Sandwich Immunoassay. For the analysis of assay precision, we mixed and analyzed tissue material from an animal experiment: We used the derivative of the A2780 cell line. A2780ADR is stably adreomycin-resistant and known to express high levels of MDR1. Both cell-lines – A2780 and A2780ADR - were subcutaneously injected in athymic mice to generate tumor xenografts and protein lysates obtained from such tumor were analyzed with the new assay system. Tumors derived from the parental line cell line (A2780) show an expected expression of MDR1 whereas no signal was obtained from A2870ADR-derived tumors (see also Fig. 3). For the determination of assay precision analysis we mixed protein extracts prepared from tumors derived from A2780 (Group 1 Animal 3 tumor A) and A2780ADR (Group 2 Animal 2 tumor A) in the ratios 1:20, 1:1, and pure A2780ADR lysate, mimicking low, medium, and high MDR1 expression. We processed each sample 5 times on the same day and on 5 different days to determine intra-assay and inter-assay precision. Intraassay CVs did not exceed 5%, and inter-assay CVs did not exceed 13% (Tab. 1 and Fig. 2B). Therefore, the overall inter-assay variance including the tryptic digest procedure were not higher than 13%. Hence, precision of the assay is comparable with good sandwich immunoassays targeting intact proteins even if a tryptic digest is performed upstream in the assay procedure. Recovery analysis of Microsphere-Based MDR1 Sandwich Immunoassay. Values for assay recovery could not be determined since no certified standard or qualified sample for MDR1 was available. However, for cross-validation we re-analyzed the same samples using an established immunoaffinity-MS assay. 7 Results between the two methods differed 17 % for the low, 9 % for the medium, and 11% for the high-level MDR1 sample (Fig. 2C). These results show an excellent degree of correlation given the fact that very different assay read-out systems were employed.
Buffer digested A2780
A Median fluorescence intensities [AU]
The combination of the epitopes which are bound by the capture and detector antibodies is unique in the human proteome a prerequisite for any method using peptides as surrogates for protein quantification.
10000 8000 6000 4000 2000 0 0,1
1
10
MDR1 Peptide [nM] Intra-assay precision
B 5 MDR1 per total protein [fmol µg-1]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
Inter-assay precision
4 3 2 1 0 high
medium
low MS-based immunoassay
C
5 MDR1 per total protein [fmol µg-1]
Page 3 of 8
Sandwich immunoassay
4 3 2 1 0 low
medium
high
Fig. 2. MDR1 Assay Validation. A Standard curve for MDR1 quantification in assay buffer and proteolyzed sample. Serial dilutions of the synthetic surrogate peptide were prepared in commercially available assay buffer (black line) or proteolyzed A2780 cells (red line) and analyzed (n=3). Results were fitted to a 4-parametric logarithmic function. B Intra-day and inter day precision analysis. Protein extracts prepared from tumors derived from A2780 (Group 1 Animal 3 Tumour A) and A2780ADR (Group 2 Animal 2 Tumour A) were mixed in the ratios 1:20, 1:2, and pure A2780ADR lysate, mimicking low, medium, and high-level MDR1 abundance. Samples were processed including the digestion procedure 5 times on the same day or on 5 different days. Intra-assay CVs did not exceed 5%, and inter-assay CVs did not exceed 14%. C For cross-validation samples were processed by an immunoaffinity-mass spectrometry assay (grey bars) and by the peptide-centric sandwich immunoassay (dark bars).
ACS Paragon Plus Environment
3
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Tab. 1: Intra- and inter-assay variance in samples expressing low, medium and high amounts of MDR1. Replicate [fmol /µg protein]
Mean (n=5)
Intra-day precision [%]
1
2
3
4
5
low
0.30
0.30
0.30
0.30
0.31
0.30
2.26
medium
1.75
1.84
1.76
1.80
1.82
1.79
1.95
high
4.28
4.37
4.06
4.14
3.85
4.14
4.34
Replicate [fmol /µg protein]
low
1
2
3
4
5
0.30
0.27
0.27
0.29
0.29
Mean (n=5) 0.28
Page 4 of 8 A2780 1/1 A 1/1 B 1/2 A 1/2 B 1/3 A 1/3 B 1/4 A 1/4 B 1/5 A 1/5 B 1/6 A 1/6 B
A2780-derived tumours *
*
A2780 ADR 2/1 A 2/1 B 2/2 A 2/2 B 2/3 A 2/3 B 2/4 A 2/4 B 2/5 A 2/5 B 2/6 A 2/6 B
Inter-day precision [%] 4.43
*
A2780ADR-derived tumours *
0
medium
1.75
1.36
1.32
1.43
1.25
1.42
12.37
high
4.28
3.56
3.41
3.58
3.19
3.60
10.09
Quantification of MDR1 in A2780 and A2780ADR derived xenografts. MDR1 levels have been associated with elevated drug-resistance in different tumors. Its overexpression can result in tolerance to chemotherapeutics by causing an increased export rate of xenobiotica. 18 In pharmaceutical research the generation of tumor xenografts in immune-deficient mice is a wellestablished model to test the efficacy of drug candidates in vivo 21-23 . Knowledge on MDR1 overexpression in these vivo models or in standard cell culture systems has to be seen as valuable information on the characteristics of the animal model and on the acquisition of multi drug resistance. We tested the sandwich immunoassay in such a setting and generated xenograft tumors in twelve immunodeficient mice using human A2780 and A2780ADR cell-lines. After 15 days, the mice were sacrificed and the tumor resected. Two tumors per animal were analyzed for MDR1 using the newly developed assay system; as control the parental cell lines were included in the measurements. For A2780 no MDR1 was detected, for A2780ADR a concentration of 8.61 fmol MDR1 per µg protein was measured (Fig. 3). In A2780-derived xenograft tumors in most cases no expression of MDR1 was observed; Interestingly one tumor (1/6B) showed a low expression at 0.52 fmol per µg protein extract. The analysis of 2780ADR-derived tumors revealed MDR1 expression ranging from 0.87 to 4.63 fmol per µg protein extract. It is of interest to note that the detected signal for MDR1 is specific for the human protein, since the employed antibody pair does not recognize the proteotypic peptide from murine MDR1.
1
2
3
4
5
6
MDR1 per total protein [fmol µg-1]
Fig. 3. MDR1 quantification in A2780 and A2780ADR-derived xenograft tumors. MDR1 was quantified in proteolytically digested protein extracts from xenograft tumors using a peptide-centric sandwich immunoassay. Two tumors per animal were analyzed. Protein extraction and measurements have been performed once. Tumors which were not available are marked with an asterisk.
Quantification of MDR1 in renal cell carcinoma. Surgical resection is the first line of treatment for primary renal cell carcinomas (RCC). Renal carcinomas generally have a poor prognosis and minimal response to classical radiotherapy or chemotherapy. In metastatic RCC patients are treated with tyrosine kinase inhibitors like Pazopanib or Sunitinib in the first-line setting. Since both drugs are substrate and inhibitor of the MDR1 transporter knowledge on protein expression could be relevant for tumor treatment. 24-27 Depending on the RCC type MDR1 mRNA overexpression and down-regulation has been reported. 28,29 In a feasibility study, we analyzed normal and tumor tissue pairs from 9 patients diagnosed with primary kidney tumors categorized according the American Joint Committee on Cancer (Tab. 2). (29) Results revealed down-regulation in two tumors when compared to normal kidney tissue and total loss of the MDR1 transporter in the remaining seven tumor tissues (Fig. 4). Tab. 2. Clinical data summary of RCC patients including TNM Classification (Edge and Compton, 2010). Age pT N M [years]
G
Subtype
Diameter [cm]
0
2
ccRCC
13
0
1
2-3
ccRCC
7
0
0
2
ccRCC
5
3a
0
0
2
ccRCC
4
72
3b
0
0
2
ccRCC
6,5
w
59
3b
0
0
2
ccRCC
3
7
m
58
1
0
0
1
ccRCC
2,7
8
m
61
2
0
1
2
ccRCC
5,2
9
f
51
3a
0
0
1-2
ccRCC
6
Patient
Gender
1
m
61
3b
2
2
w
57
3a
3
w
82
1
4
m
61
5
m
6
ACS Paragon Plus Environment
4
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry The peptide-centric MDR1 sandwich immunoassay presents a refined approach that builds on the strengths of the assay system described by Rauh-Adelmann and colleagues. 8 For their approach the two epitopes that are required to build the sandwich assay are found within the proteolytically released analyte peptide. Thereby, the simultaneous binding of two different antibodies is only possible on longer peptides and proteases that are releasing longer fragments (e.g. LysC) are advantageous for sample preparation. 31 Here, we employed N-and C-terminus-specific antibodies which allowed to reduce the length of the targeted peptide down to 13 amino acids, a third of the length of the peptides which were detected by the method described by Rauh-Adelman. Shorter target peptides facilitate an easier provision of synthetic peptide standards and circumvent solubility problems since larger peptides tend to precipitate. In addition, the use of trypsin allowed us to reduce the sample preparation time down to one hour and the use of terminal peptide epitopes allowed the set-up of a costefficient antibody generation and characterization pipeline (data not shown).
MDR1 per total protein [fmol µg -1]
Page 5 of 8
Patient Fig. 4. MDR1 quantification in renal cancer and healthy tissue. MDR1 was quantified in proteolytically digested protein extracts from renal cancer tissue and healthy tissue using a peptide-centric sandwich immunoassay. Results for healthy tissue are presented in dark grey bars and for tumors in light grey bars. Due to limited tissue amounts, sample were processed and analyzed once.
DISCUSSION In the present study, we describe the development of a sandwich immunoassay that uses a proteotypic peptide to detect and quantify the 12-transmembrane-domain transporter protein MDR1. To our knowledge this is the first example that a drug transporter is reliably quantified in a sandwich immunoassay format. Currently about 900 antibodies towards MDR1 are listed in the public antibody database Antibodypedia 30, nevertheless no information on a MDR1-specific sandwich assay has been published so far. This is probably not only caused by the difficulty of finding matching antibodies, but also due to the fact that the quantitative extraction and solubilisation of a multitransmembrane protein is difficult and may result in protein denaturation, which in turn may cause problems for antibody binding. In the assay described here, MDR1 quantification is possible, since the sample is totally digested into peptides as the first step of the assay procedure. This kind of sample preparation is commonly used for mass-spectrometry and here we show that it can be useful to analyze difficult proteins such as multitransmembrane proteins in a sandwich assay format. The enzymatic release of a proteotypic peptide from the protein of interest resulted in a surrogate peptide that could be captured by an Nterminus-specific antibody and detected by a C-terminus-specific antibody. Referencing to an external peptide, standard curves were generated and allowed quantification of the transporter protein via the surrogate peptide. The peptide detection strategy circumvents the two major problems that frequently hamper the analysis of complex membrane proteins; (i) solubilization and stabilization of the protein for the immunoassay is not required since the sample preparation is performed under denaturing conditions for the enzymatic release of the surrogate peptide, (ii) the required peptide specific antibodies are easily produced and controlled for specificity as established methods can be employed to generate these assay reagents. Moreover, mass spectrometric analysis of the immunoprecipitates using the capture and detector antibodies individually allowed a detailed analysis of crossreactivity and specificity. Based on the these results we could exclude potential cross-reactivities of the sandwich assay, since no other epitope combination is present in a digested human proteome comprising both antibody epitopes.
For assay validation purposes A2780 and A2780ADR-derived xenograft tumor material was an ideal material, since MDR1 is only expressed in A2780ADR cell line. Therefore, complex control samples expressing no analyte and samples expressing the protein of interest in a higher amount were available for determination of assay accuracy and precision. By mixing A2780 and A2780ADR-derived tumor extracts in defined ratios, low, medium and high control samples could be generated. The assay characterization revealed good performance in terms of sensitivity, accuracy and precision. Lower limit of quantification was calculated to 90 pM. Accuracy was within a range of 70-110% when peptide spiked into a proteolytically fragmented A2780-xenograft tumor sample compared to spike into an artificial assay buffer. Intra- and interday assay precision of three generated QC samples (high, medium and low MDR1 content) was below 13 % deviation after quantification and thereby meet the FDA recommendations for bioanalytical assays. Therefore, the overall precision of the assay including the digest procedure were excellent. The determination of analyte recovery was not feasible, since a reference standard is not available to date. Such a reference standard would require the generation of purified recombinant protein. This is a general issue for all multi-transmembrane proteins, since this would also require solubilizing of these proteins. As a final step of the assay characterization we cross-validated the sandwich immunoassay with a previously established MS-based immunoassay as recommended in the FDA guidelines. Results were in good concordance over the different assay platforms. In a next step, we demonstrated that the assay is suitable to support xenograft studies. MDR1 was quantified in xenograft tumors derived from A2780 and A2780ADR cell lines. We showed that acquired multi-drug resistance based on MDR1 up-regulation could be easily detected and the transporter was quantified from 10 µg protein extract. The MDR1 is detected only derived from the human tumor xenografts. Endogenous mouse MDR1 does not contribute to the measured signal, since the detected proteotypic peptide is not found in mice. Therefore, the assay is ready to address this pharmacological question. Finally, the analysis of MDR1 in 9 clinically derived ccRCCs tumor tissue samples revealed down-regulation of MDR1 on protein level. The results confirm data observed from large-scale mRNA analyses of clear cell renal carcinomas. 28 On protein level an immunohistochemistry-focused analysis of RCCs Haenisch and colleagues analyzed 22 normal tissues, 19 ccRCC and 4 nonccRCCs and reported that the MDR1 content in the analyzed ccRCCs was about 70 times lower than in normal tissue samples 29 . Therefore, our results confirm the finding that MDR1 is downregulated in ccRCCs directly on the protein level. Since the drugs - Sunitinib, and Pazopanib - used for the follow up-treatment after
ACS Paragon Plus Environment
5
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
surgery, at the time of metastasis, are MDR1 substrates and inhibitors these tumors should be sensitive to these compounds. This MDR1 assay and probably other transporter assays using the peptide-centric sandwich assay concept probably allow the analysis of the transporter expression in cancer resections. Results might explain drug resistance and sensitivity in patients and could improve the treatment regimen. Larger sample collections of ccRCCs should be analyzed to corroborate this finding since the number of specimen in our study and in the two other studies was low. 28,29
CONCLUSION In summary, the proposed method of a peptide-centric sandwich immunoassay enables fast and quantitative MDR1 detection in wide variety of sample types. The developed proteotypic sandwich immunoassay allows the quantification of one 12transmembrane domain protein. The approach is generally applicable to other difficult proteins and it has the big advantage that generation of the required antibodies and of standard peptides is easier as it is for standard immunoassays. Compared to MS-based or Western-blot read outs the workflow is simple and hundreds of samples per day can be analyzed on a standard assay platform. Moreover, compared to immunohistochemistry the results are quantitative since a peptide standard is used as reference. We have demonstrated, that the assay is ready to support drugefficacy studies in mouse xenograft models and is suitable for the analysis of clinical specimen to investigate potential acquired MDR1-based drug resistance. Future transporter assay developments on a multiplex capable platform could support and improve cancer treatment options by analyzing the tumor-specific transporter protein profile.
AUTHOR INFORMATION Corresponding Author * E-mail
[email protected]; phone +49 (0)7121 744086-1.
Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.
ACKNOWLEDGMENT This work was funded by the German Federal Ministry of Education and Research (FKZ031A142).
REFERENCES (1) Tucker, T. G.; Milne, A. M.; Fournel-Gigleux, S.; Fenner, K. S.; Coughtrie, M. W. Biochemical pharmacology 2012, 83, 279-285. (2) Mignogna, C.; Staibano, S.; Altieri, V.; De Rosa, G.; Pannone, G.; Santoro, A.; Zamparese, R.; D'Armiento, M.; Rocchetti, R.; Mezza, E.; Nasti, M.; Strazzullo, V.; Montanaro, V.; Mascolo, M.; Bufo, P. BMC cancer 2006, 6, 293. (3) Groer, C.; Bruck, S.; Lai, Y.; Paulick, A.; Busemann, A.; Heidecke, C. D.; Siegmund, W.; Oswald, S. Journal of pharmaceutical and biomedical analysis 2013, 85, 253-261. (4) Kamiie, J.; Ohtsuki, S.; Iwase, R.; Ohmine, K.; Katsukura, Y.; Yanai, K.; Sekine, Y.; Uchida, Y.; Ito, S.; Terasaki, T. Pharmaceutical research 2008, 25, 1469-1483. (5) Miliotis, T.; Ali, L.; Palm, J. E.; Lundqvist, A. J.; Ahnoff, M.; Andersson, T. B.; Hilgendorf, C. Drug Metab Dispos 2011, 39, 24402449. (6) Schaefer, O.; Ohtsuki, S.; Kawakami, H.; Inoue, T.; Liehner, S.; Saito, A.; Sakamoto, A.; Ishiguro, N.; Matsumaru, T.; Terasaki, T.; Ebner, T. Drug Metab Dispos 2012, 40, 93-103. (7) Weiss, F.; Schnabel, A.; Planatscher, H.; van den Berg, B. H.; Serschnitzki, B.; Nuessler, A. K.; Thasler, W. E.; Weiss, T. S.; Reuss,
Page 6 of 8
M.; Stoll, D.; Templin, M. F.; Joos, T. O.; Marcus, K.; Poetz, O. Scientific reports 2015, 5, 8759. (8) Rauh-Adelmann, C.; Moskow, J. M.; Graham, J. R.; Yen, L. G.; Boucher, J. I.; Murphy, C. E.; Nadler, T. K.; Gordon, N. F.; Radding, J. A. Anal Biochem 2008, 375, 255-264. (9) Shen, D. W.; Fojo, A.; Chin, J. E.; Roninson, I. B.; Richert, N.; Pastan, I.; Gottesman, M. M. Science 1986, 232, 643-645. (10) Gottesman, M. M.; Fojo, T.; Bates, S. E. Nature reviews. Cancer 2002, 2, 48-58. (11) Ma, W.; Feng, S.; Yao, X.; Yuan, Z.; Liu, L.; Xie, Y. Scientific reports 2015, 5, 18789. (12) Katayama, R.; Sakashita, T.; Yanagitani, N.; Ninomiya, H.; Horiike, A.; Friboulet, L.; Gainor, J. F.; Motoi, N.; Dobashi, A.; Sakata, S.; Tambo, Y.; Kitazono, S.; Sato, S.; Koike, S.; John Iafrate, A.; Mino-Kenudson, M.; Ishikawa, Y.; Shaw, A. T.; Engelman, J. A.; Takeuchi, K., et al. EBioMedicine 2016, 3, 54-66. (13) Shaw, A. T.; Kim, D. W.; Mehra, R.; Tan, D. S.; Felip, E.; Chow, L. Q.; Camidge, D. R.; Vansteenkiste, J.; Sharma, S.; De Pas, T.; Riely, G. J.; Solomon, B. J.; Wolf, J.; Thomas, M.; Schuler, M.; Liu, G.; Santoro, A.; Lau, Y. Y.; Goldwasser, M.; Boral, A. L., et al. The New England journal of medicine 2014, 370, 1189-1197. (14) Kim, D. W.; Mehra, R.; Tan, D. S.; Felip, E.; Chow, L. Q.; Camidge, D. R.; Vansteenkiste, J.; Sharma, S.; De Pas, T.; Riely, G. J.; Solomon, B. J.; Wolf, J.; Thomas, M.; Schuler, M.; Liu, G.; Santoro, A.; Sutradhar, S.; Li, S.; Szczudlo, T.; Yovine, A., et al. Lancet Oncol 2016. (15) Eisen, D.; Planatscher, H.; Hardie, D. B.; Kraushaar, U.; Pynn, C. J.; Stoll, D.; Borchers, C.; Joos, T. O.; Poetz, O. J Proteomics 2013. (16) Hoeppe, S.; Schreiber, T. D.; Planatscher, H.; Zell, A.; Templin, M. F.; Stoll, D.; Joos, T. O.; Poetz, O. Mol Cell Proteomics 2010. (17) Eisen, D.; Planatscher, H.; Hardie, D. B.; Kraushaar, U.; Pynn, C. J.; Stoll, D.; Borchers, C.; Joos, T. O.; Poetz, O. J Proteomics 2013, 90, 85-95. (18) Hodges, L. M.; Markova, S. M.; Chinn, L. W.; Gow, J. M.; Kroetz, D. L.; Klein, T. E.; Altman, R. B. Pharmacogenetics and genomics 2011, 21, 152-161. (19) Gerber, S. A.; Rush, J.; Stemman, O.; Kirschner, M. W.; Gygi, S. P. Proc Natl Acad Sci U S A 2003, 100, 6940-6945. (20) ICH. In International Conference on Harmonization (ICH) of Technical Requirements for the Registration of Pharmaceuticals for Human Use: Geneva, 1996. (21) Siolas, D.; Hannon, G. J. Cancer research 2013, 73, 53155319. (22) Morton, C. L.; Houghton, P. J. Nature protocols 2007, 2, 247250. (23) Gao, H.; Korn, J. M.; Ferretti, S.; Monahan, J. E.; Wang, Y.; Singh, M.; Zhang, C.; Schnell, C.; Yang, G.; Zhang, Y.; Balbin, O. A.; Barbe, S.; Cai, H.; Casey, F.; Chatterjee, S.; Chiang, D. Y.; Chuai, S.; Cogan, S. M.; Collins, S. D.; Dammassa, E., et al. Nature medicine 2015, 21, 1318-1325. (24) van Leeuwen, R. W.; van Gelder, T.; Mathijssen, R. H.; Jansman, F. G. Lancet Oncol 2014, 15, e315-326. (25) Tang, S. C.; Lankheet, N. A.; Poller, B.; Wagenaar, E.; Beijnen, J. H.; Schinkel, A. H. J Pharmacol Exp Ther 2012, 341, 164173. (26) Tang, S. C.; Lagas, J. S.; Lankheet, N. A.; Poller, B.; Hillebrand, M. J.; Rosing, H.; Beijnen, J. H.; Schinkel, A. H. Int J Cancer 2012, 130, 223-233. (27) Shukla, S.; Robey, R. W.; Bates, S. E.; Ambudkar, S. V. Drug Metab Dispos 2009, 37, 359-365. (28) Sultmann, H.; von Heydebreck, A.; Huber, W.; Kuner, R.; Buness, A.; Vogt, M.; Gunawan, B.; Vingron, M.; Fuzesi, L.; Poustka, A. Clinical cancer research : an official journal of the American Association for Cancer Research 2005, 11, 646-655. (29) Haenisch, S.; Zimmermann, U.; Dazert, E.; Wruck, C. J.; Dazert, P.; Siegmund, W.; Kroemer, H. K.; Warzok, R. W.; Cascorbi, I. The pharmacogenomics journal 2007, 7, 56-65. (30) Bjorling, E.; Uhlen, M. Mol Cell Proteomics 2008, 7, 20282037.
ACS Paragon Plus Environment
6
Page 7 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry (31) Burkhart, J. M.; Schumbrutzki, C.; Wortelkamp, S.; Sickmann, A.; Zahedi, R. P. J Proteomics 2012, 75, 1454-1462.
ACS Paragon Plus Environment
7
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 8
for TOC only
ACS Paragon Plus Environment
8