Article Cite This: J. Proteome Res. XXXX, XXX, XXX-XXX
pubs.acs.org/jpr
Quantification of Total Vitamin-D-Binding Protein and the Glycosylated Isoforms by Liquid Chromatography−Isotope Dilution Mass Spectrometry Lisa E. Kilpatrick* and Karen W. Phinney National Institute of Standards and Technology, Material Measurement Laboratory, Biomolecular Measurement Division, 100 Bureau Drive, Stop 8314, Gaithersburg, Maryland 20899, United States S Supporting Information *
ABSTRACT: Vitamin-D-binding protein (VDBP), a transporter of 25-hydroxyvitamin D metabolites, has three common isoforms. The relationship of the isoforms and their glycosylation state with various diseases has been under recent examination. In this work, liquid chromatography coupled to isotope dilution mass spectrometry was evaluated for quantification of VDBP, the three common isoforms, and total glycosylation. Protocols using guanidine, urea, RapiGest, trifluoroethanol, or tris buffer were also evaluated for optimal tryptic digestion. Differences in peptide release were detected between purified and plasma VDBP; however, for both protein sources, ELPEHTVK, TSALSAK, and VLEPTLK concentrations were reproducible between most protocols tested. The isoform-specific peptides, LPDATPK, LPDATPTELAK, and LPEATPTELAK, were optimally released when TFE was added to plasma. The total VDBP concentration calculated from the three shared peptides resulted in 97.6% accuracy compared with the concentration from amino acid analysis. Glycosylation of VDBP was also calculated for purified protein and donor samples using the ratio of the isoform-specific peptide(s) to the total protein concentration. Glycosylation of purified VDBP was found to be 99.5−111.1% the value determined by semiquantitative analysis of the intact protein by LC−MS. This approach may be used to quantify other samples containing a mixture of isoforms and post-translational modifications. KEYWORDS: vitamin-D-binding protein, GC-globulin, quantification, isoforms, glycosylation
■
INTRODUCTION
Vitamin-D-binding protein (VDBP) or group-specific component of serum (GC-globulin), a member of the albumin family, is expressed by hepatocytes in the liver and other tissues and is found in plasma and other locations in the body. In addition to transporting 25-hydroxyvitamin D [25(OH)D] metabolites, VDBP has several functions including binding fatty acids, acting as an actin scavenger, and is involved in osteoclast and macrophage activation.1,2 There are three common singlenucleotide polymorphisms for VDBP, GC-1f, GC-1s, and GC2, resulting in amino acid changes at two different positions in the sequence (Figure 1), but >120 rare isoforms have been identified.3 Depending on the isoform, one or two sites of Olinked glycosylation may be located on a threonine near the amino acid sequence unique to each isoform (Figure 1). The type, site and occupancy of glycosylation have been found to vary, but typically a di- or trisaccharide is linked to one or two threonine residue(s) at positions 434 or 436 in 98% purity determined by LC with detection at 220 nm and were purchased from GenScript (Piscataway, NJ), Biomatik (Wilmington, DE), or EZBiolab (Carmel, IN). A Discover BIO Wide C18 column (2.1 × 150 mm, 3 μm, Sigma-Aldrich, St. Louis, MO) and guard column of same packing material were used for LC−IDMS and LC−MS/ MS. A ProSwift RP-4H column (1.0 × 250 mm, ThermoFisher Scientific) was used for LC−MS analyses. Separation of amino acids was performed using a Primesep 100 column (2.1 × 250 mm, 5 μm, SIELC Technologies, Wheeling, IL). LC−MS-grade solvents were purchased from Fisher Scientific or B&J. Sample Preparation
All calibrants and samples were prepared gravimetrically. Stock solutions of each unlabeled and labeled peptide were prepared at 10 pmol/μL in 50 mmol/L Tris buffer (pH 8.0). Each peptide was prepared below the solubility limit in aqueous solution, as determined by GenScript. Labeled peptides were analyzed by multiple reaction monitoring (MRM) to check for the presence of unlabeled peptides. The labeled peptide transitions used were >99% of the total peak area measured. Calibrants were prepared with a final concentration of 100 fmol/μL of each labeled peptide and either (0.8, 4, 20, 100, or 500 or 2500) fmol/μL of each unlabeled peptide. Two mixtures B
DOI: 10.1021/acs.jproteome.7b00560 J. Proteome Res. XXXX, XXX, XXX−XXX
Journal of Proteome Research
Article
of the labeled peptides were also prepared. The first mixture, used with the individual donor samples, had a final concentration of 500 fmol/μL for each peptide. The second mixture, used with samples from pooled sources, contained the shared peptides at 500 fmol/μL each and the isoform-specific peptides at 50 fmol/μL. The ratio of unlabeled to labeled peptide were generally within a factor of 3; exceptions were observed for some digestions that did not have good release of all tryptic peptides. Pooled plasma (SRM 1950) and plasma from donors were collected with lithium heparin added as an anticoagulant and were stored at −80 °C until use. Purified VDBP or rVDBP was buffer-exchanged with 25 mmol/L ammonium bicarbonate using Zeba desalting columns and dried in a centrifugal vacuum (Labconco, Kansas City, MO). Stock solutions of the dried protein were then prepared in 50 mmol/L Tris buffer at concentrations of 0.5 g/L. The digestion procedure used was adapted from the protocol provided with Promega trypsin and is described below. Samples were prepared with 10 μL of plasma (or 10 μL of purified VDBP) in low binding tubes (Eppendorf, Hauppauge, NY), followed by the addition of 10 μL of one of the following denaturants: 10 mol/L urea, 8 mol/ L guanidine HCl, 1.25% RapiGest (by volume), or TFE. Control samples were also prepared where Tris buffer alone or containing HSA, at a 1:10 mass ratio to VDBP, was added. A final concentration of 5 mmol/L TCEP was used to reduce disulfide bonds with incubation at room temperature or 60 °C for 1 h. Similar to previous work,66 carbamylation of the lysine and arginine residues in the protein was avoided by adding urea to samples following the heat denaturing step. Iodoacetamide (25 mmol/L final concentration) was added and the sample was incubated in the dark for 45 min. Samples were diluted so that final volumes were the same and concentrations of denaturants were
