- Email: [email protected]
A recombinant polypeptide of the megakaryocyte potentiating factor is a potential biomarker in plasma for the detection of mesothelioma
Biochemical and Biophysical Research Communications 486 (2017) 526e532
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
A recombinant polypeptide of the megakaryocyte potentiating factor is a potential biomarker in plasma for the detection of mesothelioma Irina Raiko a, 1, Hans-Peter Rihs a, *, 1, Jan Gleichenhagen a, Ingrid Sander a, Jens Kollmeier b, Martin Lehnert a, Thomas Brüning a, Georg Johnen a a b
IPA - Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Institute of the Ruhr-University Bochum, Germany Lungenklinik Heckeshorn, HELIOS Clinic Emil von Behring, Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 February 2017 Accepted 16 March 2017 Available online 19 March 2017
Malignant mesothelioma (MM) is a fatal disease mostly associated with asbestos exposure and difficult to detect by non-invasive methods. This study aimed to use recombinant fragments of the megakaryocyte potentiating factor (MPF) for the development of cost-effective MPF ELISAs. Three polypeptides spanning the MPF region (MPF1-148, MPF 34-288, MPF/MSLN254-400) were produced in E.coli as maltosebinding protein hybrids. After isolation, Factor Xa digest, and purification, the polypeptides were used for the generation of rabbit antibodies and development of ELISAs. Forty-one MM patients with known histological subtype before tumor-specific treatment and 70 asbestos-exposed individuals free of any cancer were matched according to age, gender, and smoking. Plasma of all subjects was tested with the three newly developed polyclonal antibody-based ELISAs and a commercial mesothelin assay (MESOMARK™). The latter differentiated patients (median concentration 1.95 nM) from controls (median 1.07 nM, p < 0.0001) and showed an area under curve (AUC) of 0.77 in receiver operating characteristics (ROC) analysis. Of the MPF variants, exclusively the ELISA based on antibodies against the MPF34-288 fragment displayed significantly (p ¼ 0.0002) higher values in patients than in controls (median 1.61 nM versus 0.88 nM; AUC ¼ 0.72). The combination of the MPF34-288 and mesothelin displayed an improved ROC performance (AUC ¼ 0.80). The MPF34-288 ELISA could be a cost-effective and minimal-invasive contribution to support a diagnosis of mesothelioma, especially in regions with a limited medical care. © 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Biomarker Diagnostic Mesothelioma Megakaryocyte potentiating factor
1. Introduction Malignant mesothelioma (MM) is a fatal disease associated with asbestos exposure as key risk factor [1,2]. Because the treatment options are particularly limited in advanced disease, an early diagnosis combined with a multimodal therapeutic approach may be helpful to improve the survival rate as reported for MM of the epithelioid subtype [3,4]. Although there is an urgent need for noninvasive detection methods, the availability of such markers is scarce [5]. One of the best available markers are the soluble mesothelin-related peptides (SMRP), also known as mesothelin [1,6,7]. Several efforts were made to investigate other proteins of
€vention und Arbeitsmedizin * Corresponding author. Institut für Pra €t der Deutschen Gesetzlichen Unfallversicherung, Institut der Ruhr-Universita Bochum (IPA), Buerkle-de-la-Camp-Platz 1, 44789 Bochum, Germany. E-mail address: [email protected] (H.-P. Rihs). 1 Equally contributed.
the mesothelin family and to study their potential as MM biomarkers. One of them is the megakaryocyte potentiating factor (MPF) also known as N-ERC/mesothelin [8]. Both, mesothelin and MPF are products of the mesothelin gene MSLN, encoding a 71-kDa precursor protein which is cleaved by furin-like proteases into two fragments. The 31-kDa soluble N-terminal MPF fragment is secreted into blood, the C-terminal 40-kDa fragment remains membrane-bound and is classified as mature mesothelin [9e11]. The first ELISA developed by Shiomi et al. [11] was based on the monoclonal antibody (mAb) 7E7 and the polyclonal antibody (pAb) 282 directed against N-ERC/mesothelin. Subsequently, they improved this assay by replacing the pAb by the newly generated mAb 16K16 [12] and finally by mAb 20A2 [13]. In addition, two other research groups developed MPF assays using other generated pairs of monoclonal anti-MPF antibodies [14,15]. Due to its solubility and physiological origin, MPF was considered to be a promising marker to detect MM. The performance of the different MPF assays was evaluated also by other groups, mostly in comparison
http://dx.doi.org/10.1016/j.bbrc.2017.03.077 0006-291X/© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
with the “gold standard” SMRP, and depended on the antibodies used [16e18]. Because the availability of cheap MPF assays is limited, we decided to generate our own pAbs as the basis for an affordable immunoassay suitable for screening of a large cohort of asbestosexposed workers. Different MPF peptides, based on three fragments spanning the entire MPF coding region of the MSLN gene, were used to obtain pAbs for the development of three different MPF ELISAs. The purpose of this study was to evaluate the ability of the three ELISAs to discriminate between MM patients and controls. 2. Materials and methods 2.1. Patients and controls Patients with diagnosed malignant mesothelioma (MM) were recruited at the “HELIOS Clinic Emil von Behring”, Berlin, Germany. None of the patients was operated or treated by chemotherapy or radiation therapy prior to blood collection. The patient group comprised 41 persons (33 males and 8 females, mean age 69 years, range: 35e85 years). According to histological analyses, 31 of them had epithelioid, six biphasic, and four sarcomatoid mesothelioma. The control group included 70 asbestos-exposed individuals with and without benign asbestos-related diseases like pleural plaques or asbestosis but without any evidence of mesothelioma or other cancers matched according to age, gender, and smoking status to the MM patients. Controls were frequency-matched to cases by age in 5-year groups, using the following intervals: 45, 46e50, 51e55, 56e60, 61e65, 66e70, 71e75, 76e80, 81e85 years. The study was approved by the ethics committee of the RuhrUniversity Bochum (reference number: 3217-08). All individuals gave written informed consent for their participation.
527
annealing at 56 C, and elongation at 72 C for 1 min each for a total of 40 cycles. After a final elongation step at 72 C for 10 min an aliquot of the PCR-reaction was analyzed by agarose gel (1.8%) electrophoresis. All three amplification products were sub-cloned into the pDrive system (Qiagen, Hilden, Germany) for sequence analysis. One DNA clone of each fragment was chosen to generate the target fragment comprising MPF amino acid residues 1e148, 34e288, and MPF/MSLN 254e400. The gel-purified fragments were cloned into the XmnI-HindIII restricted pMALc2 expression vector (New England Biolabs GmbH, Frankfurt, Germany) to express three different MPF target fragments in frame at the C-terminus of the 42.7-kDa carrier maltose-binding protein (MBP), necessary for the purification by affinity chromatography. The resulting MBP-MPF hybrids were purified with an amylose resin (New England Biolabs GmbH) as described [19]. Aliquots of the isolated fusion proteins were specifically restricted by Factor Xa to cleave the MPF target from the MBP carrier and separate the fragments on a 4e16% Novex gel. After staining the gel with 0.1 M KCl for 30 min at 4 C the target protein band was cut out under incident light on a dark underground, homogenized, and eluted using the EzWay™ PAG protein elution kit (KOMABIOTECH, Seoul, Korea). The eluate was collected and an aliquot was run with a defined amount of broad range marker (New England Biolabs) on a 10% SDS-PAGE to calculate the concentration by scanning the appropriate gel slots after Coomassie staining with a GS-800 Calibrated Densitometer (Bio-Rad Laboratories GmbH, Munich, Germany). About 100 pmol of the target protein was blotted on a nylon membrane for analysis of the first five N-terminal amino acid residues by an Applied Biosystems Procise Sequencer using the Dansyl-Edman method. The remaining protein targets were collected and used for antibody generation. 2.4. Generation of polyclonal antibodies against different MPF variants
2.2. Blood samples Peripheral blood was collected in S-Monovette EDTA gel tubes (Sarstedt, Nümbrecht, Germany). Within 30 min after blood collection, samples were centrifuged at 2,000g for 10 min and the upper plasma phase was immediately frozen and stored at 20 C. 2.3. MPF antigen preparation Total RNA (0.1 mg) from HeLa cells (cervical adenocarcinoma, ATCC entry CCL-2) were obtained from Yorkshire Bioscience Ltd, York, U.K. The cDNA synthesis was performed with 2 mg total RNA in a total volume of 20 ml using the Transcriptor High Fidelity cDNA Synthesis Kit (Roche Diagnostics, Germany). Aliquots of the cDNA were used to amplify independently three MPF fragments corresponding to amino acid residues 1e148, with primer pair MPF1_FspI_F (50 -TGCGCAATGCCAAC-GGCTCGA-30 ) and MPF148_HindIII_R (50 -CAAGCTTCAATT-GGCCTTCGTGAT-30 ), to amino acid residues 34e288 with primer pair MPF34_AfeI_F (50 AGCGCTTCGA-GGACCCTGGCTGGAGAG-30 ) and MPF288_HindIII_R (50 -TAAGCTTCAGAT-GGTCCGTTCGGA-30 ), and to amino acid residues 254e400 with primer pair MPF254_SmaI_F (50 CCCGGGCTGGGCCAG-CCCATC-30 ) and MPF400_HindIII_R (50 GAAGCTTCATTCAAGCAAAGCCTTCAG-30 ). The PCR reaction contained 3 ml cDNA, 20 pmol of each of the primers per pair, 5 ml dNTPs, 5 ml 10 PCR buffer yielding a final concentration of 1.5 mM MgCl2, and 2.5 units of Taq polymerase in a final volume of 50 ml. PCR conditions for all three constructs included an initial denaturation step at 95 C for 5 min, followed by denaturation at 95 C,
Immunization of a rabbit with purified recombinant MPF (1 mg each of MPF1-148, MPF34-288, or MPF/MSLN254-400) was conducted by Charles River (Kisslegg, Germany) exactly in accordance with the standard protocol of the company. Antisera were received 72 days after primary immunization. Sera containing anti-MPF pAbs were loaded via FPLC (Thermo Fisher Scientific, Uppsala, Sweden) on a protein G column (GE Healthcare, Munich, Germany), washed with 20 mM phosphate buffer pH 7.0 and eluted with 0.1 M glycine-HCl buffer pH 2.8. Eluted fractions were neutralized directly with 1/ 10 vol 1 M Tris-HCl pH 9.0, pooled and dialyzed in phosphatebuffered saline (PBS) pH 7.4. Protein concentration was determined by Bradford protein assay (Bio-Rad Laboratories GmbH) with bovine serum albumin (BSA) as a standard and yielded an antibody concentration of 0.841 mg/ml (anti-MPF1-148), 1.1 mg/ml (antiMPF34-288), and 1.73 mg/ml (anti-MPF/MSLN254-400). One part of each of the purified pAbs was used later as capture antibodies. Another part was biotinylated by mixing the antibodies with a 33fold molar excess of biotin-N-hydroxysuccinimide ester (Roche Diagnostics GmbH, Mannheim, Germany), dissolved in dimethylsulfoxid and incubated under continuous agitation for 4 h at room temperature (RT). The biotinylated antibodies were dialyzed extensively with PBS. Capture and biotinylated antibodies were stored in aliquots at 80 C until use. 2.5. Measurement of mesothelin Mesothelin in plasma samples was measured using the ELISA kit MESOMARK™ (Fujirebio Diagnostics, Inc., Malvern, PA, USA).
528
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
2.6. Development of the polyclonal antibody-based MPF sandwich ELISAs
intervals (CI) were calculated. P values < 0.05 were considered statistically significant.
MaxiSorp microtiter plates (Nunc, Roskilde, Denmark) were coated with purified pAbs against MPF34-288 (100 ml/well, dilution 1:750 in 100 mM carbonate-bicarbonate buffer, pH 9.6) and allowed to adhere overnight at 4 C. The plates were blocked with 200 ml/well of Universal Casein Dilution/Blocker (Fitzgerald Industries International, Acton, MA, USA) for 2 h at RT and then washed three times with PBST (PBS containing 0.05% Tween 20). Purified recombinant MPF34-288 protein was used as a standard. A stock solution (47.645 nM) was serially diluted in the blocking solution to generate a reference curve ranging from 1.832 nM to 0.007 nM. Plasma samples were diluted also in blocking solution (1:10) and analyzed in duplicate. The plates were incubated for 1 h at RT, washed three times and incubated for 1 h with biotinylated pAbs (100 ml/well, dilution 1:1500 in PBST). After a triple washing procedure, 100 ml/well of 1:10,000 diluted horseradish-peroxidasestreptavidin conjugate (Fitzgerald Industries International, Acton, MA, USA) was added. One hour later, plates were washed again (3) and 100 ml H2O2-activated ABTS [2,2'-azino-bis(3ethylbenzothiazoline-6-sulfonic acid)diammonium salt] substrate solution (Sigma, Taufkirchen, Germany) was added to each well. The enzyme reaction was stopped by addition of 100 ml 0.32% NaF. The absorbance was read at 414 nm. Dose-response curves for standards were obtained by 4-parameter curve fitting using SoftMax Pro 5.4.1 (Molecular Devices, Sunnyvale, Calif., USA). The assay's lower limit of detection was defined by adding 0.1 OD units (rounded 8-fold mean of the standard deviation of background values from eight plates) to the background value of each plate. MPF amounts above this value were considered as detectable. MPF sandwich ELISAs based on anti-MPF1-148 and anti-MPF/MSLN254-400 pAbs were performed analogously. Antibody dilutions and standard concentrations are presented in the results.
3. Results
2.7. Statistical analysis Non-parametric Mann Whitney's test for group comparisons, Spearman's correlation coefficient, and drawing of the graphs and ROC curves were performed with Prism 5 (GraphPad Software, Inc., San Diego, CA, USA). From the ROC curves, AUC with 95% confidence
3.1. The maltose-binding protein (MBP)-megakaryocyte potentiating factor (MPF) fusion proteins The three MBP-MPF hybrids comprised the maltose-binding protein (MBP) carrier, followed by a single alanine spacer and the target part of MPF comprising amino acid residues 1e148, 34e288, or MPF/MSLN 254e400 as shown in Fig. 1. The fusion protein yields after affinity chromatography were 29.4 mg (MBP-MPF1-148), 4.7 mg (MBP-MPF34-288), and 8.1 mg (MBP-MPF/MSLN254-400) per liter LB medium. Fusion proteins were cleaved by Factor Xa and cleaned by gel electrophoresis. The resulting target fragments were finally used for immunization. 3.2. Optimization of the MPF ELISAs The three recombinant human MPF fragments were separately used to raise pAbs in rabbit. Resulting affinity-purified pAbs were utilized for MPF capturing. A part of the purified antibodies was biotinylated and used for MPF detection. Various concentrations of antibodies were tested to yield an optimized assay protocol. In case of the MPF34-288 ELISA optimized dilutions for capture and detection antibodies were 1:750 and 1:1,500, respectively. In this combination the assay had a mean detection range of 0.05 nM ± 0.004e1.81 nM ± 0.01 (n ¼ 13). For MPF1-148 antibody dilutions were 1:1000 and 1:1,300, and the mean detection range was 0.010 nM ± 0.001e1.067 nM ± 0.25 (n ¼ 15). For MPF/MSLN254400 optimal dilutions for capture and detection antibodies were 1:2500 and 1:4,000, respectively, resulting in a mean detection range of 0.026 nM ± 0.008e1.420 nM ± 0.072 (n ¼ 15). 3.3. MPF determinations in plasma samples of mesothelioma patients and in matched controls To assess the capabilities of the three developed ELISAs the MPF concentrations in plasma of 41 patients with diagnosed and histologically characterized malignant mesothelioma (MM) and of 70
Fig. 1. Structure of the Mesothelin gene, its gene products, and the recombinant fragments used for immunization. After factor Xa digest from the maltose-binding protein (MBP)-carrier (42.7 kDa), polyclonal antibodies were raised in rabbits by individual immunization with each of the three targets MPF1-148 (~16.4 kDa), MPF34-288 (~28.2 kDa), and MPF/MSLN254-400 (~16.3 kDa).
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
529
Fig. 2. Concentrations of MPF and Mesothelin in plasma of mesothelioma patients and matched controls. All MPF measurements are given in nM. A median value is shown for MM patients and the matched controls. A) MPF11-148, B) MPF/MSLN254-400, C) MPF34-288, D) Mesothelin ELISA (MESOMARK™).
asbestos-exposed matched controls were measured. As shown in Fig. 2A and B, when using MPF1-148 and MPF/MSLN254-400 ELISAs, measured MPF target concentrations were all below 2 nM without significant differences between MM patients and matched controls. In contrast, MPF values measured with the MPF34-288 ELISA system in the same plasma samples displayed a significantly (p ¼ 0.0002) higher level (median: 1.61 nM, range 0.31e39.8 nM) in MM patients when compared to matched controls (median: 0.88 nM, range 0.38e6.11 nM) (Fig. 2C). 3.4. Comparison of the MPF34-288 ELISA with the mesothelin assay Next, we investigated the performance of the MPF34-288 ELISA in comparison with the mAb-based MESOMARK™ mesothelin assay, to measure mesothelin concentrations in the same patient and control samples. As shown in Fig. 2D, the mesothelin assay differentiated between MM patients and controls (p < 0.0001) with a median mesothelin concentration of 1.95 nM in MM patients and of 1.07 nM in the asbestos-exposed control group. 3.5. ROC curve analyses of the MPF34-288 ELISA and the mesothelin assay There was a high correlation between MPF34-288 and mesothelin
Fig. 3. ROC curves for differentiating mesothelioma patients from asbestosexposed controls. Shown are the results for the MPF34-288 ELISA (blue), the Mesothelin ELISA (red), and the combination of both (black).
530
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
levels in MM patients (r ¼ 0.7983; p < 0.0001) as well as in the whole study group (patients and controls; r ¼ 0.6979, p < 0.0001). To assess the ability of our assay to distinguish between MM patients and asbestos-exposed controls, ROC curves were generated for the MPF34-288 and the mesothelin ELISA (Fig. 3). The AUC for MPF was 0.72 (95% CI 0.61e0.82) and for mesothelin 0.77 (95% CI 0.67e0.86), while for the combination of MPF and mesothelin the AUC was 0.80 (95% CI 0.71e0.89). 3.6. Performance of the MPF34-288 ELISA and the mesothelin assay in mesothelioma patients with different histological subtypes Next, we analyzed our MM study group with regard to histological subtypes. As shown in Fig. 4A, MPF values in the epithelioid subtype (n ¼ 31, median 1.76 nM) were significantly higher (p < 0.0001) than in controls (n ¼ 70, median 0.88 nM). In contrast, MPF values in the sarcomatoid subgroup (n ¼ 4, median 0.65 nM)
were similar to those in controls and significantly lower than in epithelioid MM (p ¼ 0.014) but not in biphasic MM (p ¼ 0.067). MPF concentrations in the biphasic subtype (n ¼ 6, median 1.50 nM) were similar to the epithelioid subtype but the difference to the controls did not reach significance (p ¼ 0.083). Comparable results were obtained with the mAb-based mesothelin assay (Fig. 4B). However, not only the epithelioid (n ¼ 31, median 2.01 nM), but also the biphasic subgroup (n ¼ 6, median 2.93 nM) showed a significant difference (p < 0.001 and p ¼ 0.0068) compared to controls (median 1.07 nM). Additionally, mesothelin values in the biphasic subgroup were significantly higher (p ¼ 0.038) in comparison to the sarcomatoid subgroup (n ¼ 4, median 1.03 nM). 4. Discussion Several studies [1,6,20e22] have shown that increased amounts
Fig. 4. MPF and mesothelin concentrations in plasma of mesothelioma patients according to histological subtypes. A) Results for MPF34-288, B) Results for Mesothelin for matched controls (filled squares), epithelioid (open bullets), biphasic (filled triangles), and sarcomatoid MM (filled diamonds). Concentrations are given in nM. A median value is also shown for the matched controls as well as for the subtypes.
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
of the 40-kDa C-terminal mature mesothelin (MSLN) and the 31kDa N-terminal mature MPF [12,13,15,17,23] could be promising biomarkers for an earlier detection and the monitoring of malignant mesothelioma (MM). Both, MPF and mesothelin originate from a common 71-kDa precursor protein which harbors a furin cleavage site between amino acid residues 288e293, which is obviously responsible for separation of mesothelin from MSLN, followed by the release of MPF into blood. Sapede et al. [24] revealed evidence that the secretion of soluble mesothelin variant 3 into blood could be the result of an abnormal splicing event. We have constructed three MPF variants spanning amino acid residues 1e400 of the 71-kDa precursor protein to generate three overlapping MPF-targets: 1e148, 34e288, and 254e400. All three yielded a suitable antiserum after immunization in rabbits and allowed the development of three ELISAs. The antibodies developed against the MPF34-288 target measured concentrations of 0.31e39.8 nM in human plasma and showed significant concentration differences in the plasma of 41 MM patients when compared with 70 matched controls. In contrast, no differences were observed for antibodies developed against MPF1-148 and MPF/ MSLN254-400. Recently, Cerciello et al. [25] could demonstrate that the region between residues 386e396 carries an important signature for the detection of MM in serum when glycosylated. Because our E.coli constructs are non-glycosylated this fact might be the reason that the ELISA for this construct was not able to detect differences between MM patients and controls. Another explanation may lay in an un-physiological conformation of this fragment that still includes the furin cleavage site. Furthermore, the work of Onda et al. [14] showed some evidence that the region between residues 1e291 might be more important because seven of their mAbs bound to MPF and did not bind to a recombinant hybrid consisting of mesothelin amino acid residues 296e599 and the Fc fragment of IgG1, which served as a negative control. Shiomi et al. [12] could show by epitope mapping that one of their two mAbs bound between amino acid residues 68e73 of N-ERC/mesothelin and the second one between residues 134e139. Whether a possible glycosylation or an altered confirmation due to the included signal peptide (residues 1e33) are responsible for the fact that the MPF1148 ELISA detected only low concentrations in plasma and did not display differences between patients and controls needs further investigations. Regarding the sequence of MPF comprising residues 34e288 it becomes clear that this stretch contains an important region for the detection of MPF that apparently is antigenic independently of glycosylation. There is still a lack of validated biomarkers for the early detection of mesothelioma. The most promising marker so far is mesothelin but a single marker is not sufficient to reach adequate sensitivity and specificity in the challenging setting of a prospective study with pre-diagnostic samples [26]. Consequently, a panel of several biomarkers will be necessary to achieve adequate sensitivity for screening in a high-risk population like former asbestos workers. Many different marker assays in turn will increase the cost of screening. Thus, available assays should be cheap or should have the potential to be developed into an affordable format. The intention here was to develop an affordable immunoassay for MPF and to test its performance. The performance of MPF34-288 (AUC ¼ 0.72) was comparable to that of mesothelin (AUC ¼ 0.77) and is in accordance with results published for other MPF and mesothelin assays (MPF AUC range: 0.61e0.88; mesothelin AUC range: 0.71e0.92) [15e18]. Furthermore, a combination of MPF34288 and mesothelin showed a benefit regarding the performance by increasing the AUC to 0.80. This is in contrast to the observation of Hollevoet et al. [17] and Creaney et al. [16,18] who did not see a significant increase by combining both markers. Nevertheless, a high correlation would be expected because both proteins are
531
derived from the same gene. However, there are differences regarding the release of the proteins into the bloodstream. In theory, MPF should always be released whenever the MSLN gene is expressed and the resulting precursor protein processed. In contrast, mesothelin remains anchored to the cell membrane unless the protein is shed enzymatically from the membrane or the (tumor) cell is degraded, or e less frequently e an aberrant form of soluble mesothelin is expressed [20,22,24]. Again, differences between the assays as well as the size and composition of the study populations may have contributed to the different results. An analysis of histological subtypes can reveal subgroups of patients that might not benefit from a biomarker test. Levels of mesothelin and MPF have been described as being markedly lower in samples from patients with sarcomatoid compared to epithelioid MM [12,13,17,21,27]. This is in accordance with the results obtained for the MPF34-288 ELISA. The small number of the sarcomatoid (N ¼ 4) and biphasic (N ¼ 6) subtypes, however, limits the interpretation of our results. All results will have to be verified in a larger and independent group of patients and matched controls. To validate MPF and other markers for the diagnosis of early stages of mesothelioma, a longitudinal study with serial and pre-diagnostic samples of an at-risk cohort will be necessary. In summary, a new MPF ELISA based on pAbs raised against a recombinant MPF34-288 polypeptide was established and is able to distinguish between asbestos-exposed controls and MM patients, but excluding those with a sarcomatoid subtype. Its diagnostic performance tested so far was comparable with the commercial MESOMARK™ mesothelin assay. After validation in a prospective study, the MPF34-288 ELISA could be a suitable, cost-effective tool to support a minimal-invasive diagnosis of MM, especially in regions with limited medical care. Conflicts of interest No potential conflicts of interest were disclosed. Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit-sectors. Transparency document Transparency document related to this article can be found online at http://dx.doi.org/10.1016/j.bbrc.2017.03.077. References [1] B.W. Robinson, A.W. Musk, R.A. Lake, Malignant mesothelioma, Lancet 366 (2005) 397e408. [2] B.W. Robinson, R.A. Lake, Advances in malignant mesothelioma, N. Engl. J. Med. 353 (2005) 1591e1603. [3] D.J. Sugarbaker, R.M. Flores, M.T. Jaklitsch, et al., Region margins, extrapleural nodus status, and cell type determine postoperative long-term survival in trimodality therapy of malignant pleural mesothelioma: results in 183 patients, J. Thorac. Cardiovasc Surg. 117 (1999) 54e63 discussion 63e65. [4] I. Pastan, R. Hassan, Discovery of mesothelin and exploiting it as a target for immunotherapy, Cancer Res. 74 (2014) 2907e2912. [5] S. van der Bij, E. Schaake, J.A. Burgers, et al., Markers for the non-invasive diagnosis of mesothelioma: a systematic review, Br. J. Cancer 104 (2011) 1325e1333. [6] R. Hassan, A.T. Remaley, M.L. Sampson, et al., Detection and quantification of serum mesothelin, a tumor marker for patients with mesothelioma and ovarian cancer, Clin. Cancer Res. 12 (2006) 447e453. [7] E.K. Park, A. Sandrini, D.H. Yates, et al., Soluble mesothelin-related protein in an asbestos-exposed population: the dust diseases board cohort study, Am. J. Respir. Crit. Care Med. 178 (2008) 832e837. [8] N. Yamaguchi, K. Hattori, M. Oheda, et al., A novel cytokine exhibiting megakaryocyte potentiating activity from a human pancreatic tumor cell line
532
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
HPC-Y5, J. Biol. Chem. 269 (1994) 805e808. [9] K. Chang, I. Pastan, Molecular cloning and expression of a cDNA encoding a protein detected by the K1 antibody from an ovarian carcinoma (OVCAR-3) cell line, Int. J. Cancer 57 (1994) 90e97. [10] R. Hassan, T. Bera, I. Pastan, Mesothelin: a new target for immunotherapy, Clin. Cancer Res. 10 (2004) 3937e3942. [11] K. Shiomi, H. Miyamoto, T. Segawa, et al., Novel ELISA system for detection of N-ERC/mesothelin in the sera of mesothelioma patients, Cancer Sci. 97 (2006) 928e932. [12] K. Shiomi, Y. Hagiwara, K. Sonoue, et al., Sensitive and specific new enzymelinked immunosorbent assay for N-ERC/mesothelin increases its potential as a useful serum tumor marker for mesothelioma, Clin. Cancer Res. 14 (2008) 1431e1437. [13] T. Sato, Y. Suzuki, T. Mo, et al., Newly established ELISA for N-ERC/mesothelin improves diagnostic accuracy in patients with suspected pleural mesothelioma, Cancer Med. 3 (2014) 1377e1384. [14] M. Onda, S. Nagata, M. Ho, et al., Megakaryocyte potentiation factor cleaved from mesothelin precursor is a useful tumor marker in the serum of patients with mesothelioma, Clin. Cancer Res. 12 (2006) 4225e4231. [15] K. Iwahori, T. Osaki, S. Serada, et al., Megakaryocyte potentiating factor as a tumor marker of malignant pleural mesothelioma: evaluation in comparison with mesothelin, Lung Cancer 62 (2008) 45e54. [16] J. Creaney, D. Yeoman, Y. Demelker, et al., Comparison of osteopontin, megakaryocyte potentiating factor, and mesothelin proteins as markers in the serum of patients with malignant mesothelioma, J. Thorac. Oncol. 3 (2008) 851e857. [17] K. Hollevoet, K. Nackaerts, J. Thimpont, et al., Diagnostic performance of soluble mesothelin and megakaryocyte potentiating factor in mesothelioma, Am. J. Respir. Crit. Care Med. 181 (2010) 620e625.
[18] J. Creaney, S. Sneddon, I.M. Dick, et al., Comparison of the diagnostic accuracy of the MSLN gene products, mesothelin and megakaryocyte potentiating factor, as biomarkers for mesothelioma in pleural effusions and serum, Dis. Markers 35 (2013) 119e127. [19] H.P. Rihs, Z. Chen, S. Schumacher, et al., Recombinant Hev b 1: large scale production and immunological characterization, Clin. Exp. Allergy 30 (2000) 1285e1292. [20] I. Hellstrom, J. Raycraft, S. Kanan, et al., Mesothelin variant 1 is released from tumor cells as a diagnostic marker, Cancer Epidemiol. Biomark. Prev. 15 (2006) 1014e1020. [21] A. Scherpereel, B. Grigoriu, M. Conti, et al., Soluble mesothelin related peptides in the diagnosis of malignant pleural mesothelioma, Am. J. Respir. Crit. Care Med. 173 (2006) 1155e1160. [22] H.L. Beyer, R.D. Geschwindt, C.L. Glover, et al., MESOMARK: a potential test for malignant pleural mesothelioma, Clin. Chem. 53 (2007) 666e672. [23] M. Maeda, O. Hino, Blood tests for asbestos-related mesothelioma, Oncology 71 (2006) 26e31. [24] C. Sapede, A. Gauvrit, I. Barbieux, et al., Aberant splicing and protease involvement in mesothelin release from epthelioid mesothelioma cells, Cancer Sci. 99 (2008) 590e594. [25] F. Cerciello, M. Choi, A. Nicastri, et al., Identification of a seven glycopeptide signature for malignant pleural mesothelioma in human serum reaction monitoring, Clin. Proteomics 10 (2013) 10e16. [26] J. Creaney, N.J. Olsen, F. Brims, et al., Serum mesothelin for early detection of asbestos-induced cancer malignant mesothelioma, Cancer Epidemiol. Biomark. Prev. 19 (2010) 2238e2246. [27] B.W. Robinson, J. Creaney, R. Lake, et al., Mesothelin-family proteins and diagnosis of mesothelioma, Lancet 362 (2003) 1612e1616.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
A recombinant polypeptide of the megakaryocyte potentiating factor is a potential biomarker in plasma for the detection of mesothelioma Irina Raiko a, 1, Hans-Peter Rihs a, *, 1, Jan Gleichenhagen a, Ingrid Sander a, Jens Kollmeier b, Martin Lehnert a, Thomas Brüning a, Georg Johnen a a b
IPA - Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Institute of the Ruhr-University Bochum, Germany Lungenklinik Heckeshorn, HELIOS Clinic Emil von Behring, Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 February 2017 Accepted 16 March 2017 Available online 19 March 2017
Malignant mesothelioma (MM) is a fatal disease mostly associated with asbestos exposure and difficult to detect by non-invasive methods. This study aimed to use recombinant fragments of the megakaryocyte potentiating factor (MPF) for the development of cost-effective MPF ELISAs. Three polypeptides spanning the MPF region (MPF1-148, MPF 34-288, MPF/MSLN254-400) were produced in E.coli as maltosebinding protein hybrids. After isolation, Factor Xa digest, and purification, the polypeptides were used for the generation of rabbit antibodies and development of ELISAs. Forty-one MM patients with known histological subtype before tumor-specific treatment and 70 asbestos-exposed individuals free of any cancer were matched according to age, gender, and smoking. Plasma of all subjects was tested with the three newly developed polyclonal antibody-based ELISAs and a commercial mesothelin assay (MESOMARK™). The latter differentiated patients (median concentration 1.95 nM) from controls (median 1.07 nM, p < 0.0001) and showed an area under curve (AUC) of 0.77 in receiver operating characteristics (ROC) analysis. Of the MPF variants, exclusively the ELISA based on antibodies against the MPF34-288 fragment displayed significantly (p ¼ 0.0002) higher values in patients than in controls (median 1.61 nM versus 0.88 nM; AUC ¼ 0.72). The combination of the MPF34-288 and mesothelin displayed an improved ROC performance (AUC ¼ 0.80). The MPF34-288 ELISA could be a cost-effective and minimal-invasive contribution to support a diagnosis of mesothelioma, especially in regions with a limited medical care. © 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Biomarker Diagnostic Mesothelioma Megakaryocyte potentiating factor
1. Introduction Malignant mesothelioma (MM) is a fatal disease associated with asbestos exposure as key risk factor [1,2]. Because the treatment options are particularly limited in advanced disease, an early diagnosis combined with a multimodal therapeutic approach may be helpful to improve the survival rate as reported for MM of the epithelioid subtype [3,4]. Although there is an urgent need for noninvasive detection methods, the availability of such markers is scarce [5]. One of the best available markers are the soluble mesothelin-related peptides (SMRP), also known as mesothelin [1,6,7]. Several efforts were made to investigate other proteins of
€vention und Arbeitsmedizin * Corresponding author. Institut für Pra €t der Deutschen Gesetzlichen Unfallversicherung, Institut der Ruhr-Universita Bochum (IPA), Buerkle-de-la-Camp-Platz 1, 44789 Bochum, Germany. E-mail address: [email protected] (H.-P. Rihs). 1 Equally contributed.
the mesothelin family and to study their potential as MM biomarkers. One of them is the megakaryocyte potentiating factor (MPF) also known as N-ERC/mesothelin [8]. Both, mesothelin and MPF are products of the mesothelin gene MSLN, encoding a 71-kDa precursor protein which is cleaved by furin-like proteases into two fragments. The 31-kDa soluble N-terminal MPF fragment is secreted into blood, the C-terminal 40-kDa fragment remains membrane-bound and is classified as mature mesothelin [9e11]. The first ELISA developed by Shiomi et al. [11] was based on the monoclonal antibody (mAb) 7E7 and the polyclonal antibody (pAb) 282 directed against N-ERC/mesothelin. Subsequently, they improved this assay by replacing the pAb by the newly generated mAb 16K16 [12] and finally by mAb 20A2 [13]. In addition, two other research groups developed MPF assays using other generated pairs of monoclonal anti-MPF antibodies [14,15]. Due to its solubility and physiological origin, MPF was considered to be a promising marker to detect MM. The performance of the different MPF assays was evaluated also by other groups, mostly in comparison
http://dx.doi.org/10.1016/j.bbrc.2017.03.077 0006-291X/© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
with the “gold standard” SMRP, and depended on the antibodies used [16e18]. Because the availability of cheap MPF assays is limited, we decided to generate our own pAbs as the basis for an affordable immunoassay suitable for screening of a large cohort of asbestosexposed workers. Different MPF peptides, based on three fragments spanning the entire MPF coding region of the MSLN gene, were used to obtain pAbs for the development of three different MPF ELISAs. The purpose of this study was to evaluate the ability of the three ELISAs to discriminate between MM patients and controls. 2. Materials and methods 2.1. Patients and controls Patients with diagnosed malignant mesothelioma (MM) were recruited at the “HELIOS Clinic Emil von Behring”, Berlin, Germany. None of the patients was operated or treated by chemotherapy or radiation therapy prior to blood collection. The patient group comprised 41 persons (33 males and 8 females, mean age 69 years, range: 35e85 years). According to histological analyses, 31 of them had epithelioid, six biphasic, and four sarcomatoid mesothelioma. The control group included 70 asbestos-exposed individuals with and without benign asbestos-related diseases like pleural plaques or asbestosis but without any evidence of mesothelioma or other cancers matched according to age, gender, and smoking status to the MM patients. Controls were frequency-matched to cases by age in 5-year groups, using the following intervals: 45, 46e50, 51e55, 56e60, 61e65, 66e70, 71e75, 76e80, 81e85 years. The study was approved by the ethics committee of the RuhrUniversity Bochum (reference number: 3217-08). All individuals gave written informed consent for their participation.
527
annealing at 56 C, and elongation at 72 C for 1 min each for a total of 40 cycles. After a final elongation step at 72 C for 10 min an aliquot of the PCR-reaction was analyzed by agarose gel (1.8%) electrophoresis. All three amplification products were sub-cloned into the pDrive system (Qiagen, Hilden, Germany) for sequence analysis. One DNA clone of each fragment was chosen to generate the target fragment comprising MPF amino acid residues 1e148, 34e288, and MPF/MSLN 254e400. The gel-purified fragments were cloned into the XmnI-HindIII restricted pMALc2 expression vector (New England Biolabs GmbH, Frankfurt, Germany) to express three different MPF target fragments in frame at the C-terminus of the 42.7-kDa carrier maltose-binding protein (MBP), necessary for the purification by affinity chromatography. The resulting MBP-MPF hybrids were purified with an amylose resin (New England Biolabs GmbH) as described [19]. Aliquots of the isolated fusion proteins were specifically restricted by Factor Xa to cleave the MPF target from the MBP carrier and separate the fragments on a 4e16% Novex gel. After staining the gel with 0.1 M KCl for 30 min at 4 C the target protein band was cut out under incident light on a dark underground, homogenized, and eluted using the EzWay™ PAG protein elution kit (KOMABIOTECH, Seoul, Korea). The eluate was collected and an aliquot was run with a defined amount of broad range marker (New England Biolabs) on a 10% SDS-PAGE to calculate the concentration by scanning the appropriate gel slots after Coomassie staining with a GS-800 Calibrated Densitometer (Bio-Rad Laboratories GmbH, Munich, Germany). About 100 pmol of the target protein was blotted on a nylon membrane for analysis of the first five N-terminal amino acid residues by an Applied Biosystems Procise Sequencer using the Dansyl-Edman method. The remaining protein targets were collected and used for antibody generation. 2.4. Generation of polyclonal antibodies against different MPF variants
2.2. Blood samples Peripheral blood was collected in S-Monovette EDTA gel tubes (Sarstedt, Nümbrecht, Germany). Within 30 min after blood collection, samples were centrifuged at 2,000g for 10 min and the upper plasma phase was immediately frozen and stored at 20 C. 2.3. MPF antigen preparation Total RNA (0.1 mg) from HeLa cells (cervical adenocarcinoma, ATCC entry CCL-2) were obtained from Yorkshire Bioscience Ltd, York, U.K. The cDNA synthesis was performed with 2 mg total RNA in a total volume of 20 ml using the Transcriptor High Fidelity cDNA Synthesis Kit (Roche Diagnostics, Germany). Aliquots of the cDNA were used to amplify independently three MPF fragments corresponding to amino acid residues 1e148, with primer pair MPF1_FspI_F (50 -TGCGCAATGCCAAC-GGCTCGA-30 ) and MPF148_HindIII_R (50 -CAAGCTTCAATT-GGCCTTCGTGAT-30 ), to amino acid residues 34e288 with primer pair MPF34_AfeI_F (50 AGCGCTTCGA-GGACCCTGGCTGGAGAG-30 ) and MPF288_HindIII_R (50 -TAAGCTTCAGAT-GGTCCGTTCGGA-30 ), and to amino acid residues 254e400 with primer pair MPF254_SmaI_F (50 CCCGGGCTGGGCCAG-CCCATC-30 ) and MPF400_HindIII_R (50 GAAGCTTCATTCAAGCAAAGCCTTCAG-30 ). The PCR reaction contained 3 ml cDNA, 20 pmol of each of the primers per pair, 5 ml dNTPs, 5 ml 10 PCR buffer yielding a final concentration of 1.5 mM MgCl2, and 2.5 units of Taq polymerase in a final volume of 50 ml. PCR conditions for all three constructs included an initial denaturation step at 95 C for 5 min, followed by denaturation at 95 C,
Immunization of a rabbit with purified recombinant MPF (1 mg each of MPF1-148, MPF34-288, or MPF/MSLN254-400) was conducted by Charles River (Kisslegg, Germany) exactly in accordance with the standard protocol of the company. Antisera were received 72 days after primary immunization. Sera containing anti-MPF pAbs were loaded via FPLC (Thermo Fisher Scientific, Uppsala, Sweden) on a protein G column (GE Healthcare, Munich, Germany), washed with 20 mM phosphate buffer pH 7.0 and eluted with 0.1 M glycine-HCl buffer pH 2.8. Eluted fractions were neutralized directly with 1/ 10 vol 1 M Tris-HCl pH 9.0, pooled and dialyzed in phosphatebuffered saline (PBS) pH 7.4. Protein concentration was determined by Bradford protein assay (Bio-Rad Laboratories GmbH) with bovine serum albumin (BSA) as a standard and yielded an antibody concentration of 0.841 mg/ml (anti-MPF1-148), 1.1 mg/ml (antiMPF34-288), and 1.73 mg/ml (anti-MPF/MSLN254-400). One part of each of the purified pAbs was used later as capture antibodies. Another part was biotinylated by mixing the antibodies with a 33fold molar excess of biotin-N-hydroxysuccinimide ester (Roche Diagnostics GmbH, Mannheim, Germany), dissolved in dimethylsulfoxid and incubated under continuous agitation for 4 h at room temperature (RT). The biotinylated antibodies were dialyzed extensively with PBS. Capture and biotinylated antibodies were stored in aliquots at 80 C until use. 2.5. Measurement of mesothelin Mesothelin in plasma samples was measured using the ELISA kit MESOMARK™ (Fujirebio Diagnostics, Inc., Malvern, PA, USA).
528
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
2.6. Development of the polyclonal antibody-based MPF sandwich ELISAs
intervals (CI) were calculated. P values < 0.05 were considered statistically significant.
MaxiSorp microtiter plates (Nunc, Roskilde, Denmark) were coated with purified pAbs against MPF34-288 (100 ml/well, dilution 1:750 in 100 mM carbonate-bicarbonate buffer, pH 9.6) and allowed to adhere overnight at 4 C. The plates were blocked with 200 ml/well of Universal Casein Dilution/Blocker (Fitzgerald Industries International, Acton, MA, USA) for 2 h at RT and then washed three times with PBST (PBS containing 0.05% Tween 20). Purified recombinant MPF34-288 protein was used as a standard. A stock solution (47.645 nM) was serially diluted in the blocking solution to generate a reference curve ranging from 1.832 nM to 0.007 nM. Plasma samples were diluted also in blocking solution (1:10) and analyzed in duplicate. The plates were incubated for 1 h at RT, washed three times and incubated for 1 h with biotinylated pAbs (100 ml/well, dilution 1:1500 in PBST). After a triple washing procedure, 100 ml/well of 1:10,000 diluted horseradish-peroxidasestreptavidin conjugate (Fitzgerald Industries International, Acton, MA, USA) was added. One hour later, plates were washed again (3) and 100 ml H2O2-activated ABTS [2,2'-azino-bis(3ethylbenzothiazoline-6-sulfonic acid)diammonium salt] substrate solution (Sigma, Taufkirchen, Germany) was added to each well. The enzyme reaction was stopped by addition of 100 ml 0.32% NaF. The absorbance was read at 414 nm. Dose-response curves for standards were obtained by 4-parameter curve fitting using SoftMax Pro 5.4.1 (Molecular Devices, Sunnyvale, Calif., USA). The assay's lower limit of detection was defined by adding 0.1 OD units (rounded 8-fold mean of the standard deviation of background values from eight plates) to the background value of each plate. MPF amounts above this value were considered as detectable. MPF sandwich ELISAs based on anti-MPF1-148 and anti-MPF/MSLN254-400 pAbs were performed analogously. Antibody dilutions and standard concentrations are presented in the results.
3. Results
2.7. Statistical analysis Non-parametric Mann Whitney's test for group comparisons, Spearman's correlation coefficient, and drawing of the graphs and ROC curves were performed with Prism 5 (GraphPad Software, Inc., San Diego, CA, USA). From the ROC curves, AUC with 95% confidence
3.1. The maltose-binding protein (MBP)-megakaryocyte potentiating factor (MPF) fusion proteins The three MBP-MPF hybrids comprised the maltose-binding protein (MBP) carrier, followed by a single alanine spacer and the target part of MPF comprising amino acid residues 1e148, 34e288, or MPF/MSLN 254e400 as shown in Fig. 1. The fusion protein yields after affinity chromatography were 29.4 mg (MBP-MPF1-148), 4.7 mg (MBP-MPF34-288), and 8.1 mg (MBP-MPF/MSLN254-400) per liter LB medium. Fusion proteins were cleaved by Factor Xa and cleaned by gel electrophoresis. The resulting target fragments were finally used for immunization. 3.2. Optimization of the MPF ELISAs The three recombinant human MPF fragments were separately used to raise pAbs in rabbit. Resulting affinity-purified pAbs were utilized for MPF capturing. A part of the purified antibodies was biotinylated and used for MPF detection. Various concentrations of antibodies were tested to yield an optimized assay protocol. In case of the MPF34-288 ELISA optimized dilutions for capture and detection antibodies were 1:750 and 1:1,500, respectively. In this combination the assay had a mean detection range of 0.05 nM ± 0.004e1.81 nM ± 0.01 (n ¼ 13). For MPF1-148 antibody dilutions were 1:1000 and 1:1,300, and the mean detection range was 0.010 nM ± 0.001e1.067 nM ± 0.25 (n ¼ 15). For MPF/MSLN254400 optimal dilutions for capture and detection antibodies were 1:2500 and 1:4,000, respectively, resulting in a mean detection range of 0.026 nM ± 0.008e1.420 nM ± 0.072 (n ¼ 15). 3.3. MPF determinations in plasma samples of mesothelioma patients and in matched controls To assess the capabilities of the three developed ELISAs the MPF concentrations in plasma of 41 patients with diagnosed and histologically characterized malignant mesothelioma (MM) and of 70
Fig. 1. Structure of the Mesothelin gene, its gene products, and the recombinant fragments used for immunization. After factor Xa digest from the maltose-binding protein (MBP)-carrier (42.7 kDa), polyclonal antibodies were raised in rabbits by individual immunization with each of the three targets MPF1-148 (~16.4 kDa), MPF34-288 (~28.2 kDa), and MPF/MSLN254-400 (~16.3 kDa).
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
529
Fig. 2. Concentrations of MPF and Mesothelin in plasma of mesothelioma patients and matched controls. All MPF measurements are given in nM. A median value is shown for MM patients and the matched controls. A) MPF11-148, B) MPF/MSLN254-400, C) MPF34-288, D) Mesothelin ELISA (MESOMARK™).
asbestos-exposed matched controls were measured. As shown in Fig. 2A and B, when using MPF1-148 and MPF/MSLN254-400 ELISAs, measured MPF target concentrations were all below 2 nM without significant differences between MM patients and matched controls. In contrast, MPF values measured with the MPF34-288 ELISA system in the same plasma samples displayed a significantly (p ¼ 0.0002) higher level (median: 1.61 nM, range 0.31e39.8 nM) in MM patients when compared to matched controls (median: 0.88 nM, range 0.38e6.11 nM) (Fig. 2C). 3.4. Comparison of the MPF34-288 ELISA with the mesothelin assay Next, we investigated the performance of the MPF34-288 ELISA in comparison with the mAb-based MESOMARK™ mesothelin assay, to measure mesothelin concentrations in the same patient and control samples. As shown in Fig. 2D, the mesothelin assay differentiated between MM patients and controls (p < 0.0001) with a median mesothelin concentration of 1.95 nM in MM patients and of 1.07 nM in the asbestos-exposed control group. 3.5. ROC curve analyses of the MPF34-288 ELISA and the mesothelin assay There was a high correlation between MPF34-288 and mesothelin
Fig. 3. ROC curves for differentiating mesothelioma patients from asbestosexposed controls. Shown are the results for the MPF34-288 ELISA (blue), the Mesothelin ELISA (red), and the combination of both (black).
530
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
levels in MM patients (r ¼ 0.7983; p < 0.0001) as well as in the whole study group (patients and controls; r ¼ 0.6979, p < 0.0001). To assess the ability of our assay to distinguish between MM patients and asbestos-exposed controls, ROC curves were generated for the MPF34-288 and the mesothelin ELISA (Fig. 3). The AUC for MPF was 0.72 (95% CI 0.61e0.82) and for mesothelin 0.77 (95% CI 0.67e0.86), while for the combination of MPF and mesothelin the AUC was 0.80 (95% CI 0.71e0.89). 3.6. Performance of the MPF34-288 ELISA and the mesothelin assay in mesothelioma patients with different histological subtypes Next, we analyzed our MM study group with regard to histological subtypes. As shown in Fig. 4A, MPF values in the epithelioid subtype (n ¼ 31, median 1.76 nM) were significantly higher (p < 0.0001) than in controls (n ¼ 70, median 0.88 nM). In contrast, MPF values in the sarcomatoid subgroup (n ¼ 4, median 0.65 nM)
were similar to those in controls and significantly lower than in epithelioid MM (p ¼ 0.014) but not in biphasic MM (p ¼ 0.067). MPF concentrations in the biphasic subtype (n ¼ 6, median 1.50 nM) were similar to the epithelioid subtype but the difference to the controls did not reach significance (p ¼ 0.083). Comparable results were obtained with the mAb-based mesothelin assay (Fig. 4B). However, not only the epithelioid (n ¼ 31, median 2.01 nM), but also the biphasic subgroup (n ¼ 6, median 2.93 nM) showed a significant difference (p < 0.001 and p ¼ 0.0068) compared to controls (median 1.07 nM). Additionally, mesothelin values in the biphasic subgroup were significantly higher (p ¼ 0.038) in comparison to the sarcomatoid subgroup (n ¼ 4, median 1.03 nM). 4. Discussion Several studies [1,6,20e22] have shown that increased amounts
Fig. 4. MPF and mesothelin concentrations in plasma of mesothelioma patients according to histological subtypes. A) Results for MPF34-288, B) Results for Mesothelin for matched controls (filled squares), epithelioid (open bullets), biphasic (filled triangles), and sarcomatoid MM (filled diamonds). Concentrations are given in nM. A median value is also shown for the matched controls as well as for the subtypes.
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
of the 40-kDa C-terminal mature mesothelin (MSLN) and the 31kDa N-terminal mature MPF [12,13,15,17,23] could be promising biomarkers for an earlier detection and the monitoring of malignant mesothelioma (MM). Both, MPF and mesothelin originate from a common 71-kDa precursor protein which harbors a furin cleavage site between amino acid residues 288e293, which is obviously responsible for separation of mesothelin from MSLN, followed by the release of MPF into blood. Sapede et al. [24] revealed evidence that the secretion of soluble mesothelin variant 3 into blood could be the result of an abnormal splicing event. We have constructed three MPF variants spanning amino acid residues 1e400 of the 71-kDa precursor protein to generate three overlapping MPF-targets: 1e148, 34e288, and 254e400. All three yielded a suitable antiserum after immunization in rabbits and allowed the development of three ELISAs. The antibodies developed against the MPF34-288 target measured concentrations of 0.31e39.8 nM in human plasma and showed significant concentration differences in the plasma of 41 MM patients when compared with 70 matched controls. In contrast, no differences were observed for antibodies developed against MPF1-148 and MPF/ MSLN254-400. Recently, Cerciello et al. [25] could demonstrate that the region between residues 386e396 carries an important signature for the detection of MM in serum when glycosylated. Because our E.coli constructs are non-glycosylated this fact might be the reason that the ELISA for this construct was not able to detect differences between MM patients and controls. Another explanation may lay in an un-physiological conformation of this fragment that still includes the furin cleavage site. Furthermore, the work of Onda et al. [14] showed some evidence that the region between residues 1e291 might be more important because seven of their mAbs bound to MPF and did not bind to a recombinant hybrid consisting of mesothelin amino acid residues 296e599 and the Fc fragment of IgG1, which served as a negative control. Shiomi et al. [12] could show by epitope mapping that one of their two mAbs bound between amino acid residues 68e73 of N-ERC/mesothelin and the second one between residues 134e139. Whether a possible glycosylation or an altered confirmation due to the included signal peptide (residues 1e33) are responsible for the fact that the MPF1148 ELISA detected only low concentrations in plasma and did not display differences between patients and controls needs further investigations. Regarding the sequence of MPF comprising residues 34e288 it becomes clear that this stretch contains an important region for the detection of MPF that apparently is antigenic independently of glycosylation. There is still a lack of validated biomarkers for the early detection of mesothelioma. The most promising marker so far is mesothelin but a single marker is not sufficient to reach adequate sensitivity and specificity in the challenging setting of a prospective study with pre-diagnostic samples [26]. Consequently, a panel of several biomarkers will be necessary to achieve adequate sensitivity for screening in a high-risk population like former asbestos workers. Many different marker assays in turn will increase the cost of screening. Thus, available assays should be cheap or should have the potential to be developed into an affordable format. The intention here was to develop an affordable immunoassay for MPF and to test its performance. The performance of MPF34-288 (AUC ¼ 0.72) was comparable to that of mesothelin (AUC ¼ 0.77) and is in accordance with results published for other MPF and mesothelin assays (MPF AUC range: 0.61e0.88; mesothelin AUC range: 0.71e0.92) [15e18]. Furthermore, a combination of MPF34288 and mesothelin showed a benefit regarding the performance by increasing the AUC to 0.80. This is in contrast to the observation of Hollevoet et al. [17] and Creaney et al. [16,18] who did not see a significant increase by combining both markers. Nevertheless, a high correlation would be expected because both proteins are
531
derived from the same gene. However, there are differences regarding the release of the proteins into the bloodstream. In theory, MPF should always be released whenever the MSLN gene is expressed and the resulting precursor protein processed. In contrast, mesothelin remains anchored to the cell membrane unless the protein is shed enzymatically from the membrane or the (tumor) cell is degraded, or e less frequently e an aberrant form of soluble mesothelin is expressed [20,22,24]. Again, differences between the assays as well as the size and composition of the study populations may have contributed to the different results. An analysis of histological subtypes can reveal subgroups of patients that might not benefit from a biomarker test. Levels of mesothelin and MPF have been described as being markedly lower in samples from patients with sarcomatoid compared to epithelioid MM [12,13,17,21,27]. This is in accordance with the results obtained for the MPF34-288 ELISA. The small number of the sarcomatoid (N ¼ 4) and biphasic (N ¼ 6) subtypes, however, limits the interpretation of our results. All results will have to be verified in a larger and independent group of patients and matched controls. To validate MPF and other markers for the diagnosis of early stages of mesothelioma, a longitudinal study with serial and pre-diagnostic samples of an at-risk cohort will be necessary. In summary, a new MPF ELISA based on pAbs raised against a recombinant MPF34-288 polypeptide was established and is able to distinguish between asbestos-exposed controls and MM patients, but excluding those with a sarcomatoid subtype. Its diagnostic performance tested so far was comparable with the commercial MESOMARK™ mesothelin assay. After validation in a prospective study, the MPF34-288 ELISA could be a suitable, cost-effective tool to support a minimal-invasive diagnosis of MM, especially in regions with limited medical care. Conflicts of interest No potential conflicts of interest were disclosed. Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit-sectors. Transparency document Transparency document related to this article can be found online at http://dx.doi.org/10.1016/j.bbrc.2017.03.077. References [1] B.W. Robinson, A.W. Musk, R.A. Lake, Malignant mesothelioma, Lancet 366 (2005) 397e408. [2] B.W. Robinson, R.A. Lake, Advances in malignant mesothelioma, N. Engl. J. Med. 353 (2005) 1591e1603. [3] D.J. Sugarbaker, R.M. Flores, M.T. Jaklitsch, et al., Region margins, extrapleural nodus status, and cell type determine postoperative long-term survival in trimodality therapy of malignant pleural mesothelioma: results in 183 patients, J. Thorac. Cardiovasc Surg. 117 (1999) 54e63 discussion 63e65. [4] I. Pastan, R. Hassan, Discovery of mesothelin and exploiting it as a target for immunotherapy, Cancer Res. 74 (2014) 2907e2912. [5] S. van der Bij, E. Schaake, J.A. Burgers, et al., Markers for the non-invasive diagnosis of mesothelioma: a systematic review, Br. J. Cancer 104 (2011) 1325e1333. [6] R. Hassan, A.T. Remaley, M.L. Sampson, et al., Detection and quantification of serum mesothelin, a tumor marker for patients with mesothelioma and ovarian cancer, Clin. Cancer Res. 12 (2006) 447e453. [7] E.K. Park, A. Sandrini, D.H. Yates, et al., Soluble mesothelin-related protein in an asbestos-exposed population: the dust diseases board cohort study, Am. J. Respir. Crit. Care Med. 178 (2008) 832e837. [8] N. Yamaguchi, K. Hattori, M. Oheda, et al., A novel cytokine exhibiting megakaryocyte potentiating activity from a human pancreatic tumor cell line
532
I. Raiko et al. / Biochemical and Biophysical Research Communications 486 (2017) 526e532
HPC-Y5, J. Biol. Chem. 269 (1994) 805e808. [9] K. Chang, I. Pastan, Molecular cloning and expression of a cDNA encoding a protein detected by the K1 antibody from an ovarian carcinoma (OVCAR-3) cell line, Int. J. Cancer 57 (1994) 90e97. [10] R. Hassan, T. Bera, I. Pastan, Mesothelin: a new target for immunotherapy, Clin. Cancer Res. 10 (2004) 3937e3942. [11] K. Shiomi, H. Miyamoto, T. Segawa, et al., Novel ELISA system for detection of N-ERC/mesothelin in the sera of mesothelioma patients, Cancer Sci. 97 (2006) 928e932. [12] K. Shiomi, Y. Hagiwara, K. Sonoue, et al., Sensitive and specific new enzymelinked immunosorbent assay for N-ERC/mesothelin increases its potential as a useful serum tumor marker for mesothelioma, Clin. Cancer Res. 14 (2008) 1431e1437. [13] T. Sato, Y. Suzuki, T. Mo, et al., Newly established ELISA for N-ERC/mesothelin improves diagnostic accuracy in patients with suspected pleural mesothelioma, Cancer Med. 3 (2014) 1377e1384. [14] M. Onda, S. Nagata, M. Ho, et al., Megakaryocyte potentiation factor cleaved from mesothelin precursor is a useful tumor marker in the serum of patients with mesothelioma, Clin. Cancer Res. 12 (2006) 4225e4231. [15] K. Iwahori, T. Osaki, S. Serada, et al., Megakaryocyte potentiating factor as a tumor marker of malignant pleural mesothelioma: evaluation in comparison with mesothelin, Lung Cancer 62 (2008) 45e54. [16] J. Creaney, D. Yeoman, Y. Demelker, et al., Comparison of osteopontin, megakaryocyte potentiating factor, and mesothelin proteins as markers in the serum of patients with malignant mesothelioma, J. Thorac. Oncol. 3 (2008) 851e857. [17] K. Hollevoet, K. Nackaerts, J. Thimpont, et al., Diagnostic performance of soluble mesothelin and megakaryocyte potentiating factor in mesothelioma, Am. J. Respir. Crit. Care Med. 181 (2010) 620e625.
[18] J. Creaney, S. Sneddon, I.M. Dick, et al., Comparison of the diagnostic accuracy of the MSLN gene products, mesothelin and megakaryocyte potentiating factor, as biomarkers for mesothelioma in pleural effusions and serum, Dis. Markers 35 (2013) 119e127. [19] H.P. Rihs, Z. Chen, S. Schumacher, et al., Recombinant Hev b 1: large scale production and immunological characterization, Clin. Exp. Allergy 30 (2000) 1285e1292. [20] I. Hellstrom, J. Raycraft, S. Kanan, et al., Mesothelin variant 1 is released from tumor cells as a diagnostic marker, Cancer Epidemiol. Biomark. Prev. 15 (2006) 1014e1020. [21] A. Scherpereel, B. Grigoriu, M. Conti, et al., Soluble mesothelin related peptides in the diagnosis of malignant pleural mesothelioma, Am. J. Respir. Crit. Care Med. 173 (2006) 1155e1160. [22] H.L. Beyer, R.D. Geschwindt, C.L. Glover, et al., MESOMARK: a potential test for malignant pleural mesothelioma, Clin. Chem. 53 (2007) 666e672. [23] M. Maeda, O. Hino, Blood tests for asbestos-related mesothelioma, Oncology 71 (2006) 26e31. [24] C. Sapede, A. Gauvrit, I. Barbieux, et al., Aberant splicing and protease involvement in mesothelin release from epthelioid mesothelioma cells, Cancer Sci. 99 (2008) 590e594. [25] F. Cerciello, M. Choi, A. Nicastri, et al., Identification of a seven glycopeptide signature for malignant pleural mesothelioma in human serum reaction monitoring, Clin. Proteomics 10 (2013) 10e16. [26] J. Creaney, N.J. Olsen, F. Brims, et al., Serum mesothelin for early detection of asbestos-induced cancer malignant mesothelioma, Cancer Epidemiol. Biomark. Prev. 19 (2010) 2238e2246. [27] B.W. Robinson, J. Creaney, R. Lake, et al., Mesothelin-family proteins and diagnosis of mesothelioma, Lancet 362 (2003) 1612e1616.