Article pubs.acs.org/ac
Water-Soluble Dried Blood Spot in Protein Analysis: A Proof-ofConcept Study Cecilie Rosting, Astrid Gjelstad, and Trine Grønhaug Halvorsen* Department of Pharmaceutical Chemistry, School of Pharmacy, University of Oslo, 0316 Oslo, Norway ABSTRACT: In the present work human chorionic gonadotropin (hCG) was used as a model protein in a proof-ofconcept study combining water-soluble dried blood spot (DBS) material in liquid chromatography−tandem mass spectrometry (LC−MS/MS)-based protein analysis. A watersoluble material consisting of commercially available carboxymethyl cellulose (CMC) was evaluated as sampling material for this purpose. The material dissolved readily at physiological pH. Different sample preparation methods were evaluated, and in the final method, 15 μL of whole blood was deposited and dried on CMC before the whole spot was dissolved prior to cleanup by immunoaffinity extraction, tryptic digest, and preconcentration by solid-phase extraction (SPE). The results indicated complete dissolution of hCG from the spots, acceptable limit of detection (LOD) (0.1 IU/mL), linearity (R2 = 0.959), accuracy (16%), and precision (≤22%). Long-term stability (45 days) of hCG in dried spots at reduced temperatures (≤8 °C) was also demonstrated. The analyte recovery was comparable to the commercially available nonsolvable cellulose material (FTA DMPK-C card).
D
The small sample volumes of DBS are advantageous when collecting blood from children or small animals. However, small sample volumes can also be a limitation for some analytes, e.g., low-abundant analytes, due to low analyte concentration in biological samples. High recovery from the elution step will consequently be crucial when DBS is used as sampling format for these analytes. The elution of the analytes from the commercial sampling cards is often incomplete and will together with the small sample volumes influence on the assay sensitivity. The elution of analytes from the commercial sampling card is also time-consuming and frequently referred to as the bottleneck of the DBS format.17 To circumvent these challenges a water-soluble material as sampling medium in DBS was introduced by our group in 2013.18,19 The sampling materials evaluated in those works were noncommercially available alginate and chitosan foam, and the analytes were small drug molecules. In the work presented in this article, the analyte of interest is a protein and the analysis is performed by LC−MS/MS using the bottom-up principle. Human chorionic gonadotropin (hCG) was chosen as model protein. hCG is a low-abundant endogenous protein used in therapy, for athlete doping, and as a biomarker in a number of disease states.20 This protein is therefore a suitable, but challenging, analyte to detect and quantify from small volumes of biological samples. A new feature of water-soluble DBS is therefore investigated in the present paper: the combination of water-soluble DBS
ried blood spot (DBS), introduced by Guthri and Susi in 1963, has lately gained considerable interest as a sampling and storage format for samples of whole blood and other matrixes [known as dried matrix spots (DMS)].1 Whole blood is obtained from a finger or a heel prick and then applied to the sample paper where the blood spots are dried before storage. Prior to analysis the spots are punched out, and the analytes are often eluted from the sampling card using water and organic solvents. Small sampling volumes, easy transport, improved analyte stability and reduced biohazard risk are some of the advantages when using DBS as sampling and storage format. Due to these advantages the DBS technique is useful in, i.e., diagnosis and mapping of diseases, doping tests in sport, and in sampling from animals for research purposes. Proteins are important target molecules in this context, and detection and quantifications of proteins are consequently an interesting area in analysis.2−4 Immunoassays have been used for this purpose for decades, but liquid-phase chromatography−mass spectrometry (LC−MS) has lately become an important tool in detecting and quantifying these macromolecules.5 Although more costly in operation, LC−MS has advantages such as less false positive rate and possibilities to differentiate between similar forms of the protein in one single analysis.6 DBS has been performed on a wide range of small molecules, but even though many proteins are interesting molecules to analyze from DBS samples, only a few publications have focused on DBS in LC−MS/MS-based protein analysis.7−16 This is probably due to the large dynamic range of proteins in blood, making analysis of especially low-abundance biomarkers a challenge.4 © 2015 American Chemical Society
Received: May 7, 2015 Accepted: June 23, 2015 Published: June 23, 2015 7918
DOI: 10.1021/acs.analchem.5b01735 Anal. Chem. 2015, 87, 7918−7924
Article
Analytical Chemistry
out the whole sample and transferring it to a 1.5 mL Eppendorf protein LoBind vial for dissolution in 1 mL of phosphate buffer saline (PBS, pH 7.4). Immunoaffinity Extraction. Immunoaffinity extraction was performed using magnetic beads coated with antibody E27. Coupling of antibodies to magnetizable particles was performed as described previously.21,22 Prior to immunoaffinity extraction the required volume of 10 μg/mL E27 antibodycoated magnetic beads was added to an Eppendorf vial and washed with 1 mL of PBS with 0.05% (v/v) Tween 20 and 2 × 1 mL with PBS before being redissolved in the same volume as had been pipetted out in the Eppendorf vial. An amount of 20 μL of this solution was transferred to each of the vials containing the dissolved DBS samples. Immunoextraction was performed for 1 h using a Hula mixer (Invitrogen, Carlsbad, CA, U.S.A.). The beads were subsequently washed with 1 mL of PBS with 0.05% (v/v) Tween 20, 1 mL of PBS, and with 1 mL of 10 mM Tris−HCl buffer (pH 7.4). For each wash step the samples were mixed and centrifuged before placing them in the magnetic rack (DynaMag-2 from Invitrogen) for removal of the wash solution. Afterward, the samples were redispersed in 80 μL of 50 mM ABC buffer to prepare for the reduction, alkylation, and tryptic digest. Tryptic Digest. The digest was performed by adding 10 μL of 50 μg/mL freshly made trypsin in freshly made ABC buffer, using overnight digestion at 37 °C. The proteins were reduced and alkylated prior to digestion: 4 μL of 50 mM DTT was added, and the samples were incubated at 60 °C for 15 min and then alkylated with 6 μL of 200 mM IAA. The samples were stored in dark conditions for at least 15 min during the alkylation. Solid-Phase Extraction (SPE) Using In-House Made SPE Columns. The samples were additionally cleaned up and enriched after the tryptic digest using SPE. The SPE tips were made in house by punching out six discs of C8 3M Empore material (Teknolab AS, Kolbotn, Norway) with a diameter of approximately 1 mm using a Pasteur pipet as described elsewhere.21 The discs were transferred from the Pasteur pipet and into the lower part of a 300 μL bevel point pipet tip (VWR, Hanover, Germany) using a metal wire. The SPE material was activated using 100 μL of MeCN followed by 100 μL of 20 mM FA. A volume of 10 μL of internal standard (IS) in the concentration 2.5 pmol/mL was added to the digested sample before the whole sample (∼110 μL) was transferred to the SPE tip. The tips were then washed with 100 μL of 20 mM FA before being eluted with 60 μL of MeCN/0.01% (v/v) TFA in the ratio 80:20. The elute was evaporated to dryness at 60 °C under N2 gas and reconstituted in 30 μL of 20 mM FA containing 0.001% (w/v) PEG.23 SPE Using Conventional SPE Columns (1 mL/100 mg). SPE on conventional SPE columns was performed on Chromabond C18 1 mL/100 mg columns (Macherey-Nagel, Oensingen, Switzerland) as follows: conditioned with 1 mL of MeCN and washed with 1 mL 20 mM FA. The tryptic-digested DBS samples were centrifuged (15 000 rcf, 15 min) before 0.7 mL of the supernatant was applied onto the SPE column. The column was then washed with 1 mL of 20 mM FA before elution with 1 mL MeCN/0.01% (v/v) TFA (ratio 80:20). Ultrafiltration Prior to SPE. Digested samples were added to Vivaspin filters with cutoff of 5000 kDa (GE Healthcare, Freiburg, Germany) and centrifuged (15 000 rcf, 15 min). The filtrate was transferred to an in-house made SPE column and SPE was performed as described above.
and LC−MS/MS-based protein analysis. In addition, aspects of using commercially available carboxymethyl cellulose (CMC) as the water-soluble sampling material is examined. The evaluation is based on suitability of the material in combination with protein analysis focusing on dissolution of polymer, cleanup of blood sample, and overall method performance.
■
EXPERIMENTAL SECTION Chemicals. Pregnyl (hCG) was distributed from Organon (Oss, The Netherlands). Tosyl phenylalanyl chloromethyl ketone (TPCK) treated trypsin, iodoacetic acid (IAA), 1,4dithiothretiol (DTT), formic acid (FA), trifluoroacetic acid (TFA), poly(ethylene glycol) (PEG) 20000 20% (w/v), dimethyl sulfoxide (DMSO), Tris, ammomium bicarbonate (ABC), Tween 20, and the internal standard AQUA peptide (amino acid sequence: VLQGVLPALPQVVCNY[R_13C6_15 N4]) were all purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). The international reference standard for intact hCG (standard code 99/688) was obtained from National Institute of Biological Standards. Acetonitrile (MeCN), sodium chloride, potassium chloride, disodium phosphate, monopotassium phosphate, and hydrochloric acid 37% were obtained from Merck (Darmstadt, Germany). Human serum was obtained from healthy volunteers from Oslo University Hospital, Ullevaal (Oslo, Norway), and whole blood was donated by two persons in BD vacutainer K2EDTA tubes. For the experiments in the section “Matrix Effects”, whole blood evacuated in BD vacutainer lithium heparin tubes was obtained from additionally four persons. Anti-hCG (monoclonal antibody E27) was donated by the Central Laboratory, Norwegian Radium Hospital, Oslo University Hospital (Oslo, Norway). Solutions. Preparation of hCG Stock Solution. The Pregnyl-hCG standard solutions were obtained by dissolving one ampule of Pregnyl (5000 IU) in 1 mL of ion-exchange water. This stock solution was stored at −25 °C and thawed before dilution in serum, whole blood, or 50 mM freshly made ABC buffer, respectively. All samples were spiked on the day of experiment. Internal Standard Solution. An amount of 1 nmol of dried AQUA peptide was dissolved in 20 μL of 10% (v/v) FA solution. A volume of 180 μL of 0.1% (v/v) FA was then added and the solution was vortexed. This solution served as a stock solution (5 nmol/mL). The cysteine residues on the AQUA peptide were reduced and alkylated before use in order to have the same physical− chemical properties as the signature peptide (βT5). This was done as previously described by Lund et al.:21 A volume of 40 μL of stock solution (5 nmol/mL) was diluted with 940 μL of freshly prepared 200 mM ABC buffer. An amount of 8 μL of freshly prepared 200 mM DTT was then added and the solution was under vibration heated at 95 °C for 20 min. The solution was then cooled to room temperature before a volume of 12 μL of 800 mM IAA was added. The solution was placed in the dark for 15 min. The final solution was diluted to 50 pmol/mL in 20 mM FA, frozen in −25 °C, and thawed on the day of use. Preparation of DBS. Whole blood was thawed on the day of experiment. The whole blood was spiked (