Confirmatory Analysis of Continuous Erythropoietin Receptor Activator

Oct 14, 2010 - Continuous erythropoietin receptor activator (CERA) is the third generation of recombinant human erythropoietin (rhEPO) medication that...
0 downloads 0 Views 249KB Size
Download PDF
Anal. Chem. 2010, 82, 9074–9081

Confirmatory Analysis of Continuous Erythropoietin Receptor Activator and Erythropoietin Analogues in Equine Plasma by LC-MS for Doping Control Fuyu Guan,† Cornelius E. Uboh,*,†,‡ Lawrence R. Soma,† George Maylin,§ Zibin Jiang,† and Jinwen Chen† School of Veterinary Medicine, University of Pennsylvania, New Bolton Center Campus, 382 West Street Road, Kennett Square, Pennsylvania 19348, United States, Pennsylvania Equine Toxicology and Research Center, West Chester University, West Chester, Pennsylvania 19382, United States, and Equine Drug Testing and Research Program, College of Veterinary Medicine, Cornell University, 925 Warren Drive, Ithaca, New York 14850, United States Continuous erythropoietin receptor activator (CERA) is the third generation of recombinant human erythropoietin (rhEPO) medication that retains the effect of promoting red blood cell production but has longer duration of action in the body. CERA, rhEPO, and darbepoetin alpha (DPO) can be misused to enhance performance in both human and equine athletes. To deter such misuse, a very selective and sensitive liquid chromatography-tandem mass spectrometric (LC-MS/MS) method has now been developed for identification of CERA, rhEPO, and DPO in equine plasma. The method employs a new signature tryptic peptide, T8 (54MEVGQQAVEVWQGLALLSEAVLR76, common to the three proteins), and improved immunoaffinity extraction. The analytes were extracted by anti-rhEPO antibodies from plasma samples that were pretreated with polyethylene glycol (PEG) 6000. The extracted analytes were digested by trypsin and analyzed by LC-MS/MS. The limit of identification was 0.5 ng/mL for CERA, 0.2 ng/mL for rhEPO, and 0.1 ng/ mL for DPO in equine plasma; the limit of detection was 0.3 ng/mL for CERA, 0.1 ng/mL for rhEPO, and 0.05 ng/mL for DPO. Specificity of the method was assessed via BLAST and SEQUEST protein database searches, and the T8 is extremely specific at both peptide and product ion levels for the identification of CERA, rhEPO, and DPO. This method was successful in identifying CERA and DPO in plasma samples collected from research horses post the drug administrations. It provides a useful tool in the fight against blood doping with CERA, rhEPO, and DPO in racehorses. Additionally, the following two technical approaches adopted in this study may also be helpful in protein identifications and biomarker discoveries in a broad scope: precipitating plasma proteins with PEG 6000 to improve immunoaffinity extraction efficiency * To whom correspondence should be addressed. Address: PA Equine Toxicology and Research Center, 220 East Rosedale Avenue, West Chester, PA 19382. E-mail: [email protected]. Phone: +1-610-436-3501. Fax: +1-610436-3504. † University of Pennsylvania. ‡ West Chester University. § Cornell University.

9074

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

of the target proteins and making a large and more lipophilic peptide detectable at low concentrations by increasing its solubility in the sample solvent. Erythropoietin (EPO) is an important protein hormone that stimulates red blood cell production.1 For utilization of the medical benefit of EPO, recombinant human EPO (rhEPO) was genetically engineered and approved for treatment for anemia.2 An analogue of rhEPO, darbepoetin alfa (DPO), has the same pharmacological effect as rhEPO but longer duration of action.3 Continuous erythropoietin receptor activator (CERA) is the third generation of rhEPO medication, which has the same pharmacological effect of promoting erythrocyte production. Superior to its predecessors, rhEPO and DPO, CERA has an even longer duration of action in the body.4,5 However, the medical benefits of CERA, DPO, and rhEPO may be misused in human and equine sports. Stimulation of erythrocyte production by CERA, DPO, or rhEPO can translate into increased delivery of oxygen to tissues of muscles,6 which would enhance stamina and performance of an athlete who uses one of the EPOs for blood doping. There are reports on the abuse of CERA in human sports7,8 and misuse of rhEPO and DPO in horse racing.9,10 The abuse of these protein-based drugs is of great concern to the sports community and horse racing industry as well, and thus, they are banned and placed on the 2010 Prohibited

(1) (2) (3) (4) (5) (6) (7) (8)

(9) (10)

Jelkmann, W. Eur. J. Haematol. 2007, 78, 183–205. Winearls, C. G. Nephrol., Dial., Transplant. 1998, 13 (Suppl 2), 3–8. Egrie, J. C.; Browne, J. K. Br. J. Cancer 2001, 84, 3–10. Besarab, A.; Salifu, M. O.; Lunde, N. M.; Bansal, V.; Fishbane, S.; Dougherty, F. C.; Beyer, U. Clin. Ther. 2007, 29, 626–639. Fishbane, S.; Pannier, A.; Liogier, X.; Jordan, P.; Dougherty, F. C.; Reigner, B. J. Clin. Pharmacol. 2007, 47, 1390–1397. Borrione, P.; Mastrone, A.; Salvo, R. A.; Spaccamiglio, A.; Grasso, L.; Angeli, A. J. Endocrinol. Invest. 2008, 31, 185–192. Lasne, F.; Martin, L.; Martin, J. A.; de Ceaurriz, J. Haematologica 2009, 94, 888–890. Lamon, S.; Giraud, S.; Egli, L.; Smolander, J.; Jarsch, M.; Stubenrauch, K.G.; Hellwig, A.; Saugy, M.; Robinson, N. J. Pharm. Biomed. Anal. 2009, 50, 954–958. Guan, F.; Uboh, C.; Soma, L.; Birks, E.; Chen, J.; Mitchell, J.; You, Y.; Rudy, J.; Xu, F.; Li, X.; Mbuy, G. Anal. Chem. 2007, 79, 4627–4635. Guan, F.; Uboh, C. E.; Soma, L. R.; Birks, E.; Chen, J.; You, Y.; Rudy, J.; Li, X. Anal. Chem. 2008, 80, 3811–3817. 10.1021/ac102031w  2010 American Chemical Society Published on Web 10/14/2010

List of the World Anti-Doping Agency11 and are also banned by the Association of Racing Commissioners International.12 To deter the abuse of these protein-based drugs and enforce the prohibition, reliable analytical methods for detection and identification of CERA, DPO, and rhEPO in biological samples are necessary. Three types of analytical techniques are of practical use in doping analysis of EPO drugs. Enzyme immunoassay (EIA) or enzyme-linked immunosorbent assay (ELISA) has been reported for screening analysis of human EPO,13-15 rhEPO,16,17 and CERA.8,18 In addition, a sensitive immunochromatographic assay was proposed for detection of human EPO and its analogues.19 These reported EIA methods can only be used for the detection of the EPO drugs in screening analyses but not for confirmatory analysis because of false positive results due to cross reaction of the antibodies used in EIA with other proteins present in plasma or blood. Isoelectric focusing (IEF) with immunoblotting has been reported for the detection of rhEPO in human urine,20 DPO in human and equine urine,21,22 and CERA in human urine and plasma.7 These methods are employed by all doping analysis laboratories in human sports. The IEF-based methods can detect intact target proteins themselves but cannot provide “fingerprints” of the analytes that are of significant importance in doping confirmatory analyses. Given the disadvantages of the EIA and IEF techniques mentioned above, LC-MS methods are always preferred and have recently been developed for the confirmatory identification of rhEPO and DPO in equine9,10 and human23 plasma. Aided by proteomic techniques, these methods use unique tryptic peptides to provide “fingerprints” for the identification of the protein-based drugs. rhEPO and DPO in postrace equine samples can be differentiated.10 An LC-MS method modified from the previously published method9 was recently reported for identification of CERA, rhEPO, and DPO in equine plasma,24 using one of the two tryptic peptides previously reported.9 The method uses nano LC to increase sensitivity for the analytes, but the analysis time is long (e.g., 58 min per sample). We evaluated our previous methods9,10 for identification of CERA in equine plasma, (11) World Anti-Doping Agency. The 2010 Prohibited List. http://www.wadaama.org/Documents/World_Anti-Doping_Program/WADP-Prohibited-list/ WADA_Prohibited_List_2010_EN.pdf. (Accessed July 20, 2010). (12) Association of Racing Commissioners International. Uniform Classification Guidelines for Foreign Substances and Recommended Penalties and Model Rule. http://www.arci.com/druglisting.pdf. (Accessed July 20, 2010). (13) Wognum, A. W.; Lansdorp, P. M.; Eaves, A. C.; Krystal, G. Blood 1989, 74, 622–628. (14) Ma, D. D.; Wei, A. Q.; Dowton, L. A.; Lau, K. S.; Wu, Z. H.; Ueda, M. Br. J. Hamaetol. 1992, 80, 431–436. (15) Yan, J.; Mi, J. B.; Chang, W. B. Chin. Chem. Lett. 2004, 15, 939–942. (16) Gimenez, E.; de Bolos, C.; Belalcazar, V.; Andreu, D.; Borras, E.; De la Torre, B. G.; Barbosa, J.; Segura, J.; Pascual, J. A. Anal. Bioanal. Chem. 2007, 388, 1531–1538. (17) Yanagihara, S.; Kori, Y.; Ishikawa, R.; Kutsukake, K. J. Immunoassay Immunochem. 2008, 29, 181–196. (18) Van Maerken, T.; Dhondt, A.; Delanghe, J. R. J. Appl. Physiol. 2010, 108, 800–803. (19) Loennberg, M.; Drevin, M.; Carlsson, J. J. Immunol. Methods 2008, 339, 236–244. (20) Lasne, F.; de Ceaurriz, J. Nature 2000, 405, 635. (21) Catlin, D. H.; Breidbach, A.; Elliott, S.; Glaspy, J. Clin. Chem. 2002, 48, 2057–2059. (22) Lasne, F.; Popot, M.-A.; Varlet-Marie, E.; Martin, L.; Martin, J.-A.; Bonnaire, Y.; Audran, M.; de Ceaurriz, J. J. Anal. Toxicol. 2005, 29, 835–837. (23) Guan, F.; Uboh, C. E.; Soma, L. R.; Birks, E.; Chen, J. Int. J. Sports Med. 2009, 30, 80–86. (24) Yu, N. H.; Ho, E. N. M.; Wan, T. S. M.; Wong, A. S. Y. Anal. Bioanal. Chem. 2010, 396, 2513–2521.

but the preliminary results indicated that the sensitivity for CERA was lower than that for rhEPO and DPO. The limit of confirmation for CERA (3 ng/mL) was far higher than that for rhEPO and DPO (0.4 and 0.2 ng/mL, respectively), which might prevent the applicability of those methods to identification of CERA at low concentrations in postrace samples. A new method with better sensitivity for CERA had to be developed. Thus, the purpose of the present study was to develop a sensitive and reliable LC-MS method for confirmatory analysis of CERA in equine plasma. In the present study, we use a new and specific tryptic peptide, which is common to CERA, rhEPO, and DPO but different from those used in previous publications,9,10,24 for identification of the analytes. MATERIALS AND METHODS EPO Reference Standards. Mircera containing methoxy polyethylene glycol-epoetin beta (CERA) at 50 µg/0.3 mL was purchased from Roche Registration Limited (Welwyn Garden City, UK), stored at 4 °C according to the manufacturer’s recommendation, and used as reference standard. DPO solution at 500 µg/ mL (albumin free, in polysorbate solution) was kindly donated by Amgen (Thousand Oaks, CA) and stored at 4 °C as recommended by the manufacturer. rhEPO reference material was obtained from MyBioSource (San Diego, CA), dissolved in water (HPLC grade) resulting in a concentration of 1.0 mg/mL, and stored at -70 °C in 50 µL aliquots in plastic Eppendorf vials. Intermediate CERA solution at 10 µg/mL, rhEPO solution at 100 µg/mL, and DPO solution at 20 µg/mL in water were freshly prepared by dilution of the relevant stock solution with water. Working standard solutions of CERA, rhEPO, or DPO at lower concentrations in water were freshly prepared by consecutive onetenth dilutions of the respective intermediate standard solutions. Human EPO (hEPO) T8 fragment (purity >95% by HPLC), 54 MEVGQQAVEVWQGLALLSEAVLR76, was purchased from AnaSpec (Fremont, CA). It was dissolved in acetonitrile/water/ formic acid (50/50/0.1, v/v/v) resulting in a concentration of 1.0 mg/mL. Intermediate and working solutions of hEPO T8 standard were prepared by diluting the stock solution. Chemicals and Reagents. The chemicals and reagents used were described in detail elsewhere9 and are briefly listed here: polyclonal anti-rhEPO antibodies (purified rabbit IgG) from R&D Systems (Minneapolis, MN); Dynabeads M-280 tosylactivated magnetic beads (2 × 109 beads/mL or 30 mg/mL) from Invitrogen (Carlsbad, CA); trypsin (sequencing grade modified) from Promega (Madison, WI); Igepal CA-630, polyethylene glycol 6000 (PEG 6000), and bradykinin fragment 2-9 from Sigma (St. Louis, MO). PEG 6000 solution at 50% (w/v) was prepared by dissolving 100 g of PEG 6000 dry powder in 200 mL of phosphate-buffered saline (PBS) solution (pH 7.4). Sample Preparation. Samples were prepared using a procedure modified from that detailed in the previous publication.9 The procedure included the following steps: linkage of anti-rhEPO antibodies to magnetic beads; immunoaffinity separation of CERA, rhEPO, or DPO from equine plasma by anti-rhEPO antibodies; buffer exchange of CERA, rhEPO, or DPO eluate in preparation for enzymatic digestion. The modification is briefly described below: plasma samples were pretreated with PEG 6000 to precipitate proteins prior to immunoaffinity separation, and a Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

9075

different format of the molecular weight cutoff filters and shorter centrifugation time were used in the buffer exchange step. CERA calibrators at 0.3, 0.5, 1.0, 2.0, and 3.0 ng/mL in plasma were prepared by adding 6, 10, 20, and 40 µL of CERA solution at 0.1 ng/µL and 6 µL of CERA solution at 1 ng/µL to different 2.0 mL aliquots of blank plasma in 5 mL plastic microtubes (Argos Technologies, Elgin, IL), respectively. To each spiked plasma sample, 2.0 mL of 50% PEG 6000 in PBS was added. The mixture was incubated in an Isotemp incubator (Fisher Scientific, Dubuque, IA) at 37 °C for 15 min while the microtube was continuously and gently shaken by a Labquake tube rotator (Thermo Fisher Scientific, Pittsburgh, PA). The microtubes were centrifuged at 3800g for 7 min using swinging bucket rotors in a Sorvall Legend MACH 1.6/R centrifuge at 20 °C (Kendro Laboratory Products Inc., Asheville, NC). The supernatant in each microtube was transferred to a fresh 5 mL plastic microtube, and 1.0 mL of the magnetic beads coated with anti-rhEPO antibodies was added. The plasma sample with the antibodies was incubated at 37 °C for 16-22 h while it was continuously and gently shaken by the tube rotator. Following the incubation, the magnetic beads were separated from the plasma sample by an MPC-L magnetic particle concentrator (Invitrogen), and the antibodies to which CERA bounded were rinsed four times with 1% Igepal CA-630 in PBS at ambient temperature, as described in the previous publication.9 CERA was eluted from the antibodies by incubating them in 1.0 mL of 0.1% PEG 6000 in PBS at pH 2, at ambient temperature for 30 min. CERA eluate was filtered through a 0.22 µm membrane (Ultrafree-CL centrifugal filter devices, Millipore Corporation, Billerica, MA) by centrifugation at 2500g for 5 min. The filtrate was transferred into an Amicon Ultra filter with molecular weight cutoff of 30 kDa (Millipore Corporation), and the latter was centrifuged at 3800g for 7 min as described above. To the retentate in the filter, 1.0 mL of NH4HCO3 (50 mM, pH 7.8) was added, and the filter was centrifuged at 3800g for 7 min. This process was repeated five more times. The retentate in the filter was transferred to a fresh 1.5 mL plastic microvial (Thermo Fisher Scientific), and its volume (∼50 µL) was estimated using a pipet. The volume of the retentate was increased to 86 µL by adding the bicarbonate buffer. CERA extracts were either immediately enzyme digested and analyzed or stored at -70 °C pending analysis. DPO calibrators at 0.05, 0.1, 0.2, 0.5, 1, and 2.5 ng/mL and rhEPO calibrators at 0.1, 0.2, 0.5, 1, and 2.5 ng/mL in plasma were prepared and processed in the same way as were CERA calibrators. CERA and DPO administration samples (details on the administrations of CERA and DPO are described in the Supporting Information) were similarly processed. Tryptic Digestion. The above CERA, rhEPO, or DPO extract was heated in a water bath at 80 °C for 7 min to denature proteins. After the vial was cooled to ambient temperature, 10 µL of trypsin (20 µg/100 µL in the bicarbonate buffer) was added. The vial was briefly shaken by vortex. The mixture was incubated at 37 °C for 3 h.9 After cooling of the digestion solution to ambient temperature, 4 µL of 10% formic acid was added to quench the digestion. Fifty microliters of acetonitrile was added to the digest to aid dissolution of the tryptic peptide, T8. The digest was subjected to liquid chromatography-tandem mass spectrometric (LC-MS/ MS) analysis. 9076

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

LC-MS/MS Analyses. All LC-MS/MS analyses were conducted on a Thermo Scientific LTQ XL linear ion trap mass spectrometer with an IonMax electrospray ionization (ESI) source coupled to a Surveyor Plus liquid chromatograph with an online degasser and a Surveyor Plus autosampler (Thermo Fisher Scientific, San Jose, CA). The mass spectrometer was calibrated monthly using the calibration standard as per instructions in the instrument manual. LC separations were carried out on a wide-pore Zorbax 300SBC18 column (50 × 1.0 mm I.D., 3.5 µm) with a Zorbax StableBond guard column (17 × 1.0 mm I.D., 5 µm) (Agilent, Wilmington, DE) maintained at 26 °C. Mobile phase A (H2O/ acetonitrile/formic acid, 99/1/0.1, v/v/v) and B (H2O/acetonitrile/formic acid, 5/95/0.1, v/v/v) were employed for elution of the target peptide from the column. The mobile phase gradient was programmed as follows: 25% B (0-1 min) was increased to 49% B (1-9 min), to 61% B (9-17 min), and to 80% B (17.0-17.5 min), held at 80% B (17.5-22.0 min), decreased to 25% B (22.0-22.5 min), and held at 25% B (22.5-28.0 min). The mobile phase flow rate was also programmed: 50 µL/min (0-17.5 min) was increased to 100 µL/ min (17.5-18.0 min), held at 100 µL/min (18.0-27.0 min), decreased to 50 µL/min (27.0-27.5 min), and held at 50 µL/ min (27.5-28.0 min). A 20 µL aliquot of tryptic digest of one of the analytes was injected for analysis. Optimization of ESI source parameters of the LTQ XL instrument was described in detail elsewhere.9 Briefly, the sheath gas flow was 30 (arbitrary units), auxiliary gas flow was 15 (arbitrary units), sweep gas flow rate was 1 (arbitrary unit), and the ion transfer capillary temperature was 325 °C. A signature tryptic peptide, 54MEVGQQAVEVWQGLALLSEAVLR76 (T8), was employed for the confirmatory analysis of CERA, rhEPO, or DPO in equine plasma, and the triply charged peptide at m/z 843.4 was monitored in MS/MS. The scan range for product ions was m/z 230-2000. The parameters for MS/MS experiments were as follows: wideband activation enabled, 25% normalized collision energy, 0.25 of the Q value, 30 ms of activation time, and m/z 1.5 of isolation width. The maximum ion injection time was set at 50 ms. Two microscans were summed to a scan. SEQUEST Searches. To examine the specificity of product ions of the tryptic peptide T8 for identification of the analytes, SEQUEST searches were conducted with Bioworks (V 3.3.1 SP1, Thermo Fisher Scientific, San Jose, CA). Raw data files from the targeted MS/MS experiments on the triply charged T8 from CERA, rhEPO, or DPO were used in searches against comprehensive protein databases, and SEQUEST search was performed only on those scans within the chromatographic peak of the T8. The search parameters chosen were as follows: mass type, monoisotopic precursor and fragments; enzyme, trypsin; missed cleavage site, 0; peptide mass tolerance, 3.0 AMU; fragment ion tolerance, 0.25 AMU; ions and ion series calculated, b ions and y ions. The protein databases searched were as follows: uniref100 fasta database (UniProt Release 2010_07, June 15, 2010; the UniProt Consortium of European Bioinformatics Institute, Swiss Institute of Bioinformatics and Protein Information Resource), and an equine protein FASTA database (April 2004) in Bioworks.

RESULTS AND DISCUSSION Tryptic Digestion of CERA. CERA is PEGylated epoetin beta with a PEG polymer of ∼30 kDa attached to the protein via an amide bound with the amino group of either the alanine in position 1 or one of the lysines in positions 45 and 52 of EPO.25 The PEGylation may affect both immunoaffinity extraction of CERA from plasma and subsequent tryptic digestion for LC-MS analysis, because it usually results in increased resistance of proteins to proteolytic digestion and decreased immunogenicity.26 On the basis of this consideration, the time for tryptic digestion of CERA neat standard was prolonged from 3 to 22 h, but it did not lead to any increase in the yield of the T6 (46VNFYAWK52) fragment. Thus, an incubation time of 3 h was used in this study. Utilization of the Peptide, T8, for Identification of CERA, rhEPO, and DPO. The T8 of rhEPO, 54MEVGQQAVEVWQGLALLSEAVLR76, is common to CERA and DPO; it can be employed for identification of the three proteins because of its long amino acid sequence. The longer the sequence of a peptide, the more specific it is for identification of the protein from which it derives.27 In the initial phase of development of an LC-MS method for the identification of rhEPO and DPO, the T8 was chosen as a specific-peptide candidate. However, it was disqualified because it was not detectable when the concentration of its precursor protein was below 100 ng/mL.9 It was not realized until recently that the absence of the T8 from detection at low concentrations might be due to its low solubility in the sample solvent (0.1% formic acid in water). Loss of peptides due to adsorption to vials during sample preparations is a common problem in peptide analyses.28,29 One of the approaches to addressing this problem is the addition of an organic solvent to the samples.10,28 In the present study, acetonitrile was added to the sample solvent to increase solubility of the T8, and its effect on detectability or recovery of the latter was experimentally assessed (S-Table 1 in the Supporting Information). The results in S-Table 1 indicate that the T8 chromatographic peak area was the largest when 70 µL of acetonitrile (37%) was added to the DPO digest. When 80 or 90 µL of acetonitrile was added, the T8 chromatographic peak was broadened and split, which is undesired. Conversely, the T8 was almost undetectable when less than 30 µL of acetonitrile (16%) was added to the DPO digest. From these results and the consideration for peak sharpness, it was decided that 50 µL of acetonitrile (33%) be added to 100 µL of DPO, rhEPO, or CERA digest to increase solubility of the T8. To verify the identity of the T8 from CERA, hEPO T8 reference standard was obtained and compared with the former. As expected, CERA T8 had the same product ion profile (S-Figure 1 in the Supporting Information) and LC retention time (tR ) 13.50 min) as hEPO T8 (tR ) 13.55 min). These results confirmed the identity of CERA T8 fragment. The identity of CERA T8 was further confirmed by comparison of its accurate-mass product ions with those of hEPO T8 (S-Table Jelkmann, W. Nephrol., Dial., Transplant. 2007, 22, 2749–2753. Jain, A.; Jain, S. K. Crit. Rev. Ther. Drug Carrier Syst. 2008, 25, 403–447. Gowd, K. H.; Krishnan, K. S.; Balaram, P. Mol. BioSyst. 2007, 3, 554–566. John, H.; Walden, M.; Schafer, S.; Genz, S.; Forssmann, W. G. Anal. Bioanal. Chem. 2004, 378, 883–897. (29) Pezeshki, A.; Vergote, V.; Van Dorpe, S.; Baert, B.; Burvenich, C.; Popkov, A.; De Spiegeleer, B. J. Pharm. Biomed. Anal. 2009, 49, 607–612. (25) (26) (27) (28)

2 in the Supporting Information). Additionally, the charge state of doubly charged b product ions of CERA T8, b16-b18, (b19-H2O), b20-b22, was verified by isotopic peak distributions from accurate mass measurements (data not shown). Although both doubly and triply charged forms of CERA T8 were observed under the present experimental condition, the triply charged peptide was predominant while the doubly charged form was less than 30% of the former. Thus, the triply charged T8 was employed for the identification of CERA, rhEPO, and DPO. In the product ion spectrum of the triply charged T8 (S-Figure 1 in the Supporting Information), there were singly charged y-ion series and doubly charged b-ion series. It should be noted that doubly charged b-ion series are rare in tryptic peptides though singly charged b ions are common, and thus, the former would be uniquely specific for the identification of CERA, rhEPO, and DPO. Additionally, doubly charged ions can be detected with higher sensitivity than the singly charged by a mass spectrometer.30 Furthermore, the longer the sequence of a peptide such as the T8, the more hydrophobic and retainable by an LC column it is, and the less chromatographic interferences it would suffer in identification of the protein from which it derives. Finally, the T8 would ionize better in the ESI source than peptides with shorter sequences such as the T6, because the former elutes at a higher organic percentage of the mobile phase from the LC column. In short, the T8 is better in all aspects than the T6 or T17 (144VYSNFLR150) for the identification of CERA, rhEPO, and DPO. Precipitation of Plasma Proteins by PEG 6000 to Aid in Extraction of CERA. Polyethylene glycol (PEG) is widely used as a fractional precipitating agent for the purification of proteins from a variety of sources including plasma, because of its nondenaturing characteristics.31,32 PEG size and concentration are two important parameters in protein precipitations.32 The research results from the Equine Drug Testing and Research Program at the Cornell University indicated that pretreatment of racehorse plasma samples by adding equal volume of 50% PEG 6000 in water could significantly reduce false-positive results by the ELISA kits from StemCell Technologies (Vancouver, Canada). It was reported that protein precipitation by 50% PEG 6000 in 0.15 M saline was employed prior to EIA in the detection of CERA in human plasma.18 In the present study, precipitation of plasma proteins by PEG 6000 was evaluated regarding its effect on immunoaffinity extraction of CERA from equine plasma. The addition of 2 mL of 50% PEG 6000 in H2O to 2 mL of equine plasma did not affect the extraction efficiency of CERA by anti-rhEPO antibodies. The effect of 0.5, 1, and 2 mL of 50% PEG 6000 in PBS (pH 7.4) added to 2 mL of equine plasma was compared; the addition of 2 mL of 50% PEG 6000 resulted in increased extraction efficiency for CERA, rhEPO, and DPO (Table 1) while that of 0.5 or 1 mL of 50% PEG 6000 did not lead to any increase in the extraction efficiency (data not shown). These results indicate that PBS used as a solvent for PEG 6000 plays a role in increasing the extraction efficiency of CERA, which can be explained in view of the fact that PBS is a necessary media for (30) l’Huillier, A.; Lompre, L. A.; Mainfray, G.; Manus, C. Phys. Rev. A 1983, 27, 2503. (31) Atha, D. H.; Ingham, K. C. J. Biol. Chem. 1981, 256, 2108–2117. (32) Tscheliessnig, A.; Ong, D.; Lee, J.; Pan, S. Q.; Satianegara, G.; Schriebl, K.; Choo, A.; Jungbauer, A. J. Chromatogr. 2009, 1216, 7851–7864.

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

9077

Table 1. Effect of PEG 6000 on Immunoaffinity Extraction Efficiency of CERA, rhEPO, and DPOa

CERA rhEPO DPO

T8 peak area (no PEG 6000)

T8 peak area (PEG 6000 added)

percent increase in peak area

1735 2875 2937

4313 5916 5802

249 206 198

a CERA, rhEPO, or DPO (2 ng each) spiked to 2 mL of blank equine plasma was extracted without or with addition of 2 mL of 50% PEG 6000 in PBS (pH 7.4). Each peak area value listed is the average from measurements of duplicate samples.

proteins such as anti-rhEPO antibodies to function normally. The increased extraction efficiency resulting from the addition of 2 mL of 50% PEG 6000 was beneficial to developing a sensitive method for identification of CERA in plasma. Thus, plasma samples (2 mL each) were pretreated with 50% PEG 6000 in PBS. Proteins precipitated from 2 mL of equine plasma by PEG 6000 were visible and examined by both the bottom-up proteomic technique33 and microfluidics-based electrophoresis. LC-MS/MS analysis (details on the sample preparation, mobile phase gradient, and data-dependent MS/MS are described in the Supporting Information) indicated that the precipitated proteins contained equine serum albumin, immunoglobulin, and apolipoprotein A-I precursor (S-Table 3 in the Supporting Information). The results from electrophoretic analysis of the precipitated proteins by microfluidic chips (S-Figure 2 in the Supporting Information) are presented in S-Table 4 (Supporting Information). The protein with molecular weight of 68.9 kDa in S-Table 4 (Supporting Information) might be equine serum albumin (68.6 kDa from UniProtKB, www.ebi.ac.uk/uniprot) and that with molecular weight of 174.1 kDa might be equine immunoglobulin. It was reported that large proteins in human plasma such as fibrinogen, fibronectin, von Willebrand factor, and coagulation factor XIII were precipitated by 2% PEG 4000 while complement C1, C4b-binding protein and its complex with complement C4, low-density lipoproteins, IgM, and some small proteins, which interacted with the high-molecular-mass proteins and coprecipitated with them, were precipitated by 4% PEG 4000.34 It was also reported that human serum albumin was partially precipitated by 16% PEG 4000.35 In consideration of the results obtained on the PEG 6000 precipitate and those reported, we conclude that the partial removal of abundant plasma proteins such as serum albumin and immunoglobins by PEG 6000 precipitation contributed to the increased extraction efficiency of CERA from equine plasma. Method Evaluation. Identification of CERA, rhEPO, or DPO in equine plasma was carried out using the unique tryptic peptide T8, via retention time (tR) and major product ions. As shown in Figure 1, CERA spiked to blank plasma and recovered by immunoaffinity separation resulted in a chromatographic peak of the T8 at tR of 13.48 min. The chromatographic peak was verified to be the T8 by the MS/MS spectrum (S-Figure 3 in the Supporting Information). In contrast, blank plasma did not give rise to a chromatographic peak at or around the tR of the T8 and, thus, did not cause any interference to the identification of (33) Han, X. M.; Aslanian, A.; Yates, J. R. Curr. Opin. Chem. Biol. 2008, 12, 483–490. (34) Jin, Y.; Manabe, T. Electrophoresis 2009, 30, 3613–3621. (35) Shimazaki, Y.; Kodama, A. Anal. Chim. Acta 2009, 643, 61–66.

9078

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

Figure 1. LC-MS/MS chromatograms of blank equine plasma (2 mL, bottom panel) and CERA (1 ng) spiked to it (top panel) indicating the chromatographic peak of the T8 (m/z 843.4 for triply charged species) at tR of 13.48 min. The product ions used in reconstructing the chromatograms were m/z 674.4 (y6), 870.6 (b162+), 927.1(b172+), 1070.8(b202+), 1120.2(b212+), and 1176.8(b222+).

CERA (Figure 1). Analyte carryover from an analysis to the next is sometimes an issue in peptide analyses and has to be addressed before a reliable analytical method can be developed.10,36 In the present study, the possibility of analyte carryover was carefully examined, and no carryover was observed in the identification of CERA, rhEPO, and DPO. The retention time reproducibility was 13.45 ± 0.05 min (mean ± SD, n ) 21) with a maximum range of 13.35-13.53 min, based on the retention time of the T8 from CERA, rhEPO, and DPO spiked to blank plasma and measured on three separate days in a period of 3 months. It was noted that the area of the T8 chromatographic peak from 5 ng of CERA, rhEPO, or DPO spiked to 2 mL of plasma was remarkably higher than that from the same quantity of the respective neat standards. The ratio of the peak area from the analyte spiked to plasma to that from the neat standard was 3.8, 1.9, and 1.7 for CERA, rhEPO, and DPO, respectively. These significant differences in signal intensity of the T8 from the neat standards and those spiked to plasma might result from such possible cause: ion enhancement of the matrix effect, partial loss of the T8 in digests of the neat standards due to lipophilic adsorption to plastic vials, or both. The ion enhancement, if present, may be beneficial to the qualitative analysis of CERA, rhEPO, and DPO even though the matrix effect is usually harmful to quantifications of analytes.37 The present method is primarily for qualitative analysis of CERA, rhEPO, and DPO; presence of one of the analytes in equine (36) Oe, T.; Ackermann, B. L.; Inoue, K.; Berna, M. J.; Garner, C. O.; Gelfanova, V.; Dean, R. A.; Siemers, E. R.; Holtzman, D. M.; Farlow, M. R.; Blair, I. A. Rapid Commun. Mass Spectrom. 2006, 20, 3723–3735. (37) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 1998, 70, 882–889.

Table 2. SEQUEST Search Results for CERA or DPO Spiked Plasma Samples and a DPO Administration Sample Demonstrating Specificity of the T8 for Identification of CERA and DPOa EPO

concentration

CERA

1 ng/mL

DPO

0.2 ng/mL

DPO

9 days post administration

peptide matched

P (pep) -8

protein matched

species Homo sapiens Congregibacteria litoralis KT71 Homo sapiens Pectobacterium carotovorum subsp. carotovorum PC1 Homo sapiens Tribolium castaneum

MEVGQQAVEVWQGLALLSEAVLR TQVALDDLDASELQTLIAELVLR MEVGQQAVEVWQGLALLSEAVLR VNDTYGHNIGDDVIR

8.90 × 10 2.70 × 10-2 5.80 × 10-8 1.20 × 10-3

EPO_HUMAN A4AD59_9GAMM EPO_HUMAN C6DCR6_PECCP

MEVGQQAVEVWQGLALLSEAVLR ELVLVDEITAGNVCGIGGLESAIVR

1.30 × 10-6 1.10 × 10-2

EPO_HUMAN UPI0000D55A65

a The FASTA protein database searched was the comprehensive UniProt uniref100 (Release 2010_07, June 15, 2010) that contains varieties of proteins from many different species. The first and second matches to the product ion spectrum of the T8 are listed for each sample; the first match is far more reliable than the second one in view of the probability. P (pep) value indicates the probability of finding a match as good as or better than the observed peptide match by chance.

plasma is deemed confirmed if all the criteria proposed below are met. The criteria proposed in this study for the identification of CERA, rhEPO, or DPO are the presence of the major product ions of the T8: m/z 1176.9 (b222+), 1120.4 (b212+), 1071.0 (b202+), 970.4 (b182+), 674.4 (y6), and 288.3 (y2) with signal-to-noise (S/ N) > 3 in the product ion spectrum obtained from the LC-MS/ MS chromatogram. The criteria also include a retention time match (±0.2 min window) of the chromatographic peak from a suspect sample to that of the T8 from reference standard of CERA, rhEPO, or DPO spiked to blank equine plasma. Under these criteria, the limit of confirmation (the lowest concentration at which the presence of an analyte can be confirmed) was 0.5 ng/mL for CERA (Figure 1, and S-Figure 3 in the Supporting Information), 0.2 ng/mL for rhEPO (S-Figure 4 in the Supporting Information), and 0.1 ng/mL for DPO in equine plasma (S-Figure 5 in the Supporting Information). The limit of detection was 0.3, 0.1, and 0.05 ng/mL for CERA, rhEPO, and DPO, respectively (S/N > 3 for the chromatographic peak of the T8 in the LC-MS/MS chromatogram). CERA in equine plasma cannot be quantified by the present method because the calibration curve was not linear in the concentration range of 0.5-3 ng/mL. However, rhEPO and DPO can be quantified with external calibrations. The calibration curves were linear in the range of 0.1-1 ng/mL for rhEPO and 0.05-2.5 ng/mL for DPO, with coefficients of determination (r2) > 0.98 (S-Figure 6 in the Supporting Information). The sensitivity of the present method would be further improved if a more sensitive ion trap instrument such as the LTQ Velos (Thermo Fisher Scientific) or a triple quadruple instrument is employed. Specificity for Identification of CERA and EPO Analogues in Equine Plasma. The T8 contains 23 amino acid residues covering 14% of the total sequence of CERA and is longer in sequence than a combination of both the T6 (7 amino acid residues) and T17 (7 amino acid residues) that were used for the identification of rhEPO and DPO.9 Thus, the T8 would be more specific in theory for the identification of CERA, rhEPO, and DPO than either the T6 or T17. To verify this prediction and demonstrate specificity of the T8 for the identification of CERA, rhEPO, and DPO, a basic local alignment search tool (BLAST) search was carried out against the comprehensive protein database of the European Bioinformatics Institute. The search result (S-Table 5 in the Supporting Information) indicates that the full T8 sequence is found only in the EPO molecule of a few species such as human, chimpanzee, macaca sp., and crab-eating macaque. It should be noted that the T8 (23 amino

acid residues) of CERA, rhEPO, and DPO is significantly different in sequence length from the respective T8 (39 amino acid residues) of equine EPO.9 Thus, it is concluded that the former is uniquely specific at amino acid sequence level for the identification of CERA, rhEPO, and DPO. The T8 of CERA generates more product ions (10 major product ions) than the T6 (six major product ions) or T17 (six major product ions). For CERA spiked at low concentrations to equine plasma, the T8 is a better peptide for identification than the T6 or T17. As shown in S-Figure 3 (Supporting Information), there are less background noise peaks, and the product ions have better signal-to-noise ratio in the MS/MS spectrum of the T8 from 1 ng of CERA spiked to 2 mL of plasma than in those of the T6 and T17 from 2 ng of CERA added to 2 mL of plasma. To assess the possibility that the product ion spectrum of the T8 can be derived from irrelevant peptide of other proteins, SEQUEST38 search was conducted against a comprehensive protein FASTA39 database that contains varieties of proteins from many different species. The search result (Table 2) indicates that the experimental product ion spectrum of the T8 from either CERA or DPO spiked at very low concentrations to plasma is matched primarily to hEPO, which is correct. In Table 2, the first match was always hEPO and was far more reliable than the second match in view of the probability. Manual examination of the first-match results reveals that all the major experimental product ions of the T8 from CERA or DPO spiked to plasma are matched to the predicted ones of the T8 from hEPO (S-Figures 7 and 8 in the Supporting Information). In the second-match results, however, some of the major experimental product ions are not matched (data not shown). It should be noted that, even at the concentration of CERA as low as the limit of confirmation (0.5 ng/mL), the product ion spectrum of the T8 obtained was still matched to hEPO (S-Figure 7 in the Supporting Information). The product ion spectrum of the T8 from the plasma sample collected 9 days post a DPO administration was also matched to hEPO (Table 2, S-Figure 9 in the Supporting Information). In conclusion, the T8 is extremely specific at both amino acid sequence and product ion spectrum levels for the identification of CERA, rhEPO, and DPO. (38) Sadygov, R. G.; Cociorva, D.; Yates, J. R. Nat. Methods 2004, 1, 195–202. (39) Pearson, W. R.; Lipman, D. J. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 2444– 2448.

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

9079

Figure 2. LC-MS/MS chromatogram of the T8 from CERA in a plasma sample (1.5 mL) pooled from equal volume of the samples collected at 192 and 240 h post administration, showing the chromatographic peak at tR of 13.50 min (top panel) and the product ions (y2-y6, b162+, b172+, b202+-b222+ in bottom panel) for confirmation.

Figure 3. LC-MS/MS chromatogram of T8 from DPO in the plasma sample collected 12 days post administration showing the chromatographic peak at tR of 13.43 min and the product ions (y2, y4, y6, b172+, b202+, b212+, and b222+) for confirmation of DPO.

Preliminary Application. The present method was successful in analyzing CERA and DPO administration samples. CERA was identified in plasma samples up to 192 h post the administration (IM dose of 100 µg to a horse, Figure 2). DPO was identified in plasma samples up to 12 days post the administration (IM dose of 100 µg to a horse, Figure 3), and its concentration was estimated to be 0.08 ng/mL. The product ion spectrum of the T8 obtained from the sample collected 9 days post the administration of DPO was matched to hEPO by SEQUEST search (Table 2, 9080

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

S-Figure 9 in the Supporting Information), which was correct. These results demonstrate the applicability and validity of the current method for analyses of postrace equine plasma samples. The present method is more sensitive (0.5 vs 1 ng/mL in LOC) and uses less volume of plasma (2 vs 5 mL) for the identification of CERA than a reported method.24 The limitation of this method is that it cannot differentiate among CERA, rhEPO, and DPO. However, this limitation does not weaken the capability and usefulness of the method for the identification of CERA, rhEPO,

and DPO because none of them is endogenously produced by the horse. CONCLUSION A very selective and sensitive LC-MS/MS method has been developed for the identification of CERA, rhEPO, and DPO in equine plasma. The identification was achieved via the LC retention time and major product ions of the new signature tryptic peptide, T8. The T8 is better than the T6 and T17 employed in previous publications, in view of sensitivity. The limit of identification was 0.5 ng/mL for CERA, 0.2 ng/mL for rhEPO, and 0.1 ng/mL for DPO in equine plasma; the limit of detection was 0.3 ng/mL for CERA, 0.1 ng/mL for rhEPO, and 0.05 ng/ mL for DPO. The specificity of the method was assessed via BLAST and SEQUEST protein database searches, and the T8 is extremely specific at both amino acid sequence and product ion spectrum levels for the identification of CERA, rhEPO, and DPO. To the authors’ knowledge, it is the first time in the field of doping analysis that the specificity of product ions of a unique peptide from a protein at practically low concentrations in plasma for identification of the latter was assessed and verified by a bioinformatics approach such as SEQUEST search. The LC-MS method was successful in identifying CERA and DPO in plasma samples collected from research horses post the drug administrations. It provides a useful tool in the fight against blood doping with CERA, rhEPO, and DPO in racehorses. Additionally, precipitation of plasma proteins with PEG 6000 was conducted to improve immunoaffinity extraction efficiency

of the target proteins. The large and lipophilic peptide, T8, was made detectable at low concentrations by increasing its solubility in the sample solvent. ACKNOWLEDGMENT Financial support was provided by The Pennsylvania Racing Commissions, and financial contributions were made by the Pennsylvania Harness Horsemen Association at Pocono Downs and Chester Downs, Meadows Standardbred Owners Association, Horsemen Benevolent and Protective Association at Penn National and Presque Isles Downs. The authors are grateful to Amgen Inc. (Thousand Oaks, CA) for kind donation of DPO reference standard. The authors thank Mrs. Pamela Brown and HFL Sport Science (Fordham, Cambridgeshire, UK) for their help with the purchase of CERA reference standard. The ProteinProspector software developed by Dr. Alma Burlinggame and his group at the University of California, San Francisco, California, was used in interpreting product ions of the triply charged T8 of rhEPO. BLAST search was performed using the ExPASy Proteomics Server (the Swiss Institute of Bioinformatics). SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review August 12, 2010. Accepted September 28, 2010. AC102031W

Analytical Chemistry, Vol. 82, No. 21, November 1, 2010

9081