Quantification of Human Growth Hormone in Serum with a Labeled

To manage and inform diagnostic or therapeutic decisions, measurement results which are accurate, specific, and comparable between laboratories are ...
0 downloads 0 Views 349KB Size
Article pubs.acs.org/ac

Quantification of Human Growth Hormone in Serum with a Labeled Protein as an Internal Standard: Essential Considerations Caroline Pritchard,*,†,‡ Kate J. Groves,† Sabine Biesenbruch,† Gavin O’Connor,† Alison E. Ashcroft,‡ Cristian Arsene,§ Dirk Schulze,§ and Milena Quaglia† †

LGC, Queens Road, Teddington, London TW11 0LY, United Kingdom Astbury Centre for Structural Molecular Biology, Faculty of Biological Sciences, University of Leeds, Leeds, LS2 9JT, United Kingdom § Physikalisch-Technische Bundesanstalt, Bundesallee 100, 38116 Braunschweig, Germany ‡

ABSTRACT: To manage and inform diagnostic or therapeutic decisions, measurement results which are accurate, specific, and comparable between laboratories are required. Two challenges associated with this are the definition of the measurand and the commutability of the reference standard used. Once the measurand is defined, the next step in improving standardization is developing traceable quantification methods for proteins in biological fluids. A novel reference method for the quantification of recombinant human growth hormone (rhGH) in serum has been developed using multistep sample cleanup at the protein level, tryptic digestion, and isotope dilution mass spectrometry (IDMS). Critical considerations for using isotopically labeled rhGH as the internal standard are described. A bulk serum sample was prepared at the clinically relevant level of 10 ng/g and quantified using the method described to give results traceable to the International System of Units (SI) with a total measurement uncertainty of > > > >

531.3 551.2 530.3 585.4 875.5

label transition (m/z) 392.2 337.7 427.8 470.8 686.3

> > > > >

541.3 561.2 540.3 595.4 885.5

fragment

collision energy (V)

window start time (min)

y5 y3 y4 y6 y8

9 7 10 12 20

10 15.5 23.5 32 39

Each dried aliquot was dissolved in 500 μL of low pH buffer (1 M urea, 5% formic acid, 0.1% BSA). Low pH sample cleanup was performed by combining all four aliquots per sample onto one SPE cartridge. SPE was performed as follows with 3 mL volumes used unless stated. Elution buffers were (A) 1% formic acid in H2O and (B) 1% formic acid in acetonitrile. Conditioning: MeOH, H2O; load sample (4× aliquots, 2 mL total); wash: 0% B, 5% B, 30% B. Samples were eluted using 2× 1.5 mL of 70% B into 2× 2 mL of LoBind Eppendorfs. This gives a total of two aliquots per sample. Samples were frozen at −80 °C and dried at 30 °C, 99% (Grenoble, France). To ensure no bias occurred due to the contribution of unlabeled protein, a high concentration (40 μg/g) digestion of the labeled protein was performed and the percentage isotopic labeling determined for each quantification peptide: T12 LEDGSPR 99.9%, T19 IVQCR 99.8%, T8 SNLELLR 99.9%, T1 FPTIPLSR 99.9%, and T11 DLEEGIQTLMGR 99.8%. For further details, see Pritchard et al.15 The serum used was a growth hormone depleted off the clot human serum (SciPac, Sittingbourne, UK), hereafter referred to as blank serum. Other materials were ultrapure water (18 MΩ cm−1), formic acid (Fisher Scientific, Loughborough, UK), urea, bovine serum albumin (BSA), TPCK treated trypsin, triethylammonium bicarbonate pH 8.5 (TEAB), tris(2carboxyethyl)phosphine (TCEP), sodium dodecyl sulfate (SDS), methylmethanethiosulfate (MMTS) (all SigmaAldrich, Gillingham, UK), and acetonitrile and methanol (both Optigrade HPLC Special grade, LGC Standards). Standard Preparation. A stock standard of the WHO material was prepared gravimetrically at ∼1.1 mg/g in water, with subsequent dilutions prepared in 0.1% BSA in water. The stock standard and dilutions of the IS were prepared in a similar manner. Quantification of the stock standard was performed as described by Pritchard et al.2 A bulk mix WHO-spiked serum was prepared gravimetrically at 10.50 ± 0.53 ng/g in blank serum. The spiked serum was mixed gently for 1 h, before being stored in ∼1050 μL aliquots at −20 °C to prevent repeat freeze−thaw cycles. Blend Preparation. For EM-IDMS quantification experiments, a pair of blends was prepared gravimetrically for each quantification repeat. The sample blend contained the WHOspiked serum (800 μL, 10.5 ng/g), and the ILP and the calibration blend contained 800 μL of blank serum with both the WHO and ILP. The mass fraction of the labeled protein was equivalent in both blends and added such that equal instrument response for the natural and label signals was observed. Each sample was then split volumetrically into two aliquots, and 400 μL of high pH buffer (4 M urea in 100 mM ammonium acetate, pH 10) was added to each aliquot. Samples were denatured under shaking overnight at 25 °C. Sample Cleanup. Sample cleanup was performed using a two-dimensional SPE separation at high pH and low pH using Strata-XL C18 (3 mL, 200 mg, Phenomenex, Macclesfied, UK) cartridges for both steps. High pH sample cleanup was performed using 400 μL aliquots of denatured serum samples (two aliquots per sample) as follows with 3 mL volumes used unless stated. Elution buffers were (A) 1% formic acid in H2O and (B) 1% formic acid in acetonitrile. Conditioning: MeOH, H2O; load sample; wash: 5% B, 30% B. Samples were eluted using 2× 1.5 mL of 70% B into 2× 2 mL LoBind Eppendorf tubes (FisherScientific). This gives a total of four aliquots per sample. Samples were frozen at −80 °C and dried at 30 °C,