Investigating microwave hydrolysis for the traceable

0 downloads 0 Views 402KB Size Report
Dec 23, 2010 - these can be reduced with the use of isotopically labeled standards ... the hydrolysis step must be fully evaluated because the internal.
Analytical Biochemistry 412 (2011) 40–46

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

Investigating microwave hydrolysis for the traceable quantification of peptide standards using gas chromatography–mass spectrometry Caroline Pritchard, Frank A. Torma, Chris Hopley, Milena Quaglia, Gavin O’Connor ⇑ LGC, Queens Road, Teddington, Middlesex, TW11 0LY, UK

a r t i c l e

i n f o

Article history: Received 29 October 2010 Received in revised form 7 December 2010 Accepted 7 December 2010 Available online 23 December 2010 Keywords: Amino acids Absolute quantification of peptides GC–MS Microwave hydrolysis IDMS

a b s t r a c t Over the past decade, a number of endogenous peptides and endogenous peptide analogs have been employed in therapeutics and as diagnostic markers. The use of peptides as standards for the absolute quantification of proteins has become commonly accepted. Consequently, the requirement for standard peptides traceable to the International System of Units with low associated measurement uncertainty, and for accurate methods of peptide quantification, has increased. Here we describe a method of peptide quantification involving microwave-assisted acid hydrolysis followed by gas chromatography–mass spectrometry that enables traceable quantification of a peptide by exact matching isotope dilution mass spectrometry where the total hydrolysis time required is only 3 h. A solution of angiotensin I was quantified using this method, and the results were in agreement with those obtained previously using an oven hydrolysis liquid chromatography–tandem mass spectrometry method. Ó 2011 LGC. Published by Elsevier Inc. All rights reserved.

The requirement for reliable methods of peptide quantification is rapidly increasing due to the use of peptides in an expanding number of fields. Analogs of endogenous peptides such as hormones and neurotransmitters [1–3], and recombinant proteins such as insulin, vaccines, and antibodies, have been used extensively as therapeutic agents. Endogenous peptides and proteins such as human growth hormone have also been applied in diagnostics to monitor physiological conditions. However, the clinical tests currently performed in laboratories require improved standardization [4–7]. With the development of novel biopharmaceuticals and new diagnostic tools, there is a growing need for accurate methods of peptide quantification and for standards traceable to the International System of Units (SI)1 to ensure both method specificity and interlaboratory comparability. SI-traceable standard peptides are, along with total tryptic digestion, one of the essential requirements for performing absolute protein quantification—a fundamental element of proteomics [8]. This typically involves tryptic digestion of the protein followed by quantification of the peptides released [9–12].

⇑ Corresponding author. Fax: +44 20 8943 2767. E-mail address: [email protected] (G. O’Connor). Abbreviations used: SI, International System of Units; UV, ultraviolet; MS, mass spectrometry; GC–MS, gas chromatography–mass spectrometry; EM–IDMS, exact matching isotope dilution mass spectrometry; LC–MS/MS, liquid chromatography– tandem mass spectrometry; NIST, National Institute of Standards and Technology; MTBSTFA, N-methyl-N-(tert-butyldimethylsilyl)-trifluoroacetamide; t-BDMCS (tertbutyldimethylcholorsilane); EI, electron ionization; SIM, selected ion monitoring; RSD, relative standard deviation. 1

Quantifying standard peptides is commonly performed by acid hydrolysis combined with analysis of the amino acids released [13]. Because complete acid hydrolysis results in molar ratios of the amino acids released relative to the parent peptide being studied, the concentration of the peptide can be determined from the amino acid concentrations, for which well-characterized pure materials, such as those provided by the National Institute of Standards and Technology or the National Metrology Institute of Japan, are readily available. This is traditionally done by ninhydrin derivatization and ultraviolet (UV) detection. However, these methods are slow, and for the improved selectivity and sensitivity required for quantification of multiple peptide standards with differing amino acid sequences, mass spectrometry (MS)-based detection techniques with or without precolumn derivatization are now often used [14–16]. Gas chromatography–mass spectrometry (GC–MS) offers high-chromatographic resolution, good sensitivity, and low variability in measurement precision, and it would appear to be the method of choice for amino acid analysis. However, sample derivatization is required for GC–MS, and consequently issues such as derivative instability, reagent interferences, and unwanted side reactions typically must be considered in detail for accurate quantification. The unwanted effects of derivatization can be reduced with the use of exact matching isotope dilution mass spectrometry (EM–IDMS) [17], in which isotopically labeled forms of the amino acids are added to the sample prior to any sample preparation and behave in an analogous manner throughout. In addition, a blend containing the natural and isotopically labeled amino acids is prepared at the same concentration as the sample such that they are indistinguishable to the mass spectrometer. This type of method,

0003-2697/$ - see front matter Ó 2011 LGC. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2010.12.015

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

when practiced correctly, relates directly to the SI with small measurement uncertainties. Reliability is ensured for peptide quantification only if complete molar release of a number of amino acids (typically P3) has been established [18]. The peptide hydrolysis step, therefore, is critical for traceable quantification to be achieved and must be validated individually for each peptide of interest. Conventional hydrolysis is performed under vacuum with hydrochloric acid at 110 °C for 24 h in either the liquid or vapor phase. However, it has been demonstrated [18] that significantly longer hydrolysis times (>50 h) are required for complete amino acid release. The hydrolysis stage, therefore, is a considerable bottleneck within this process. Microwave irradiation has been used to accelerate reactions in the biological field for some time, including applications such as peptide synthesis, enzymatic digestion, and acid hydrolysis [19,20], with high temperatures being applied for short periods of time (typically 99%). Labeled amino acids were obtained from Cambridge Isotopes Laboratory (Andover, MA, USA): L-arginine/HCl (13C6, 98%), L-isoleucine (13C6, 98%), 13 13 15 15 L-leucine ( C6 N, 98%), L-phenylalanine ( C9 N, 99%), L-proline 13 13 15 15 ( C5 N, 98%), and L-valine ( C5 N, 98%). The reference material used for the traceable mass fraction determination of each amino acid (purity) was SRM 2389 (National Institute of Standards and Technology [NIST], Gettysburg, PA, USA), which contains 17 free amino acids (Ala, Arg, Asp, Cys, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Tyr, and Val). The sample used was a bulk dilution of angiotensin I CRM 998 (sequence DRVYIHPFHL, NIST, Gaithersburg, MD, USA) with a concentration of 57.5 ± 1.73 nmol/ g. The concentration was previously measured by oven hydrolysis combined with IDMS and LC–MS/MS, as described previously by Burkitt and coworkers [18].

41

Other materials were ultra-pure water (18 MX cm), N-methyl-N(tert-butyldimethylsilyl)-trifluoroacetamide (MTBSTFA) + 1% tertbuty(dimethylcholorsilane) (t-BDMCS), hydrochloric acid (37 wt.% in water, 99.999%) (all from Sigma–Aldrich, Gillingham, UK), trifluoroacetic acid (Fluka), and acetonitrile (OptiGrade HPLC Special Grade, LGC Standards, Teddington, UK). Methods Sample preparation Amino acid standard solutions were gravimetrically prepared in 100 mmol/L HCl and stored at 5 ± 4 °C. The individual mass fractions for each gravimetrically prepared stock standard were determined by EM–IDMS and GC–MS analysis using the NIST standard reference material 2389 as the principal calibration standard, providing traceability to the SI. Two equimolar standard mixes containing the six natural and six labeled amino acids, respectively, were prepared gravimetrically at 50 nmol/g. Sample blends and calibration blends were prepared gravimetrically from the standard mixes in the following manner. The sample blend contained angiotensin I and labeled amino acids, and the calibration blend contained natural and labeled amino acids. Mass fractions of the labeled amino acids were equivalent in both blends and added such that equal instrument response for the natural and labeled signals was observed. Amino acid hydrolysis Microwave oven hydrolysis was performed in a Discover benchtop microwave coupled to a hydrolysis kit and vacuum pump (all CEM Microwave Technology, Buckingham, UK). Calibration and sample blends were dried in a vacuum centrifuge and put into the microwave hydrolysis vessel with 5 ml of 6 M HCl. The vessel was sealed and frozen in a slurry of dry ice and acetone before being flushed with nitrogen and then evacuated to 85 kPa (15 s each, performed five times). The samples were then hydrolyzed using the following program: ramp for 8 min at 200 W, hold at 165 °C for 3 h, maximum pressure set to 0.75 MPa. GC–MS analysis Blends were freeze-dried and then derivatized by adding 80 ll of MTBSTFA + 1% t-BDMCS. Samples were heated in covered vessels in an oven at 85 °C for 1 h before direct analysis by fast GC– MS. Samples were injected five times each, alternating with the respective calibration blend, and the results were determined using the EM–IDMS equation as described by Burkitt and coworkers [18]. GC–MS analysis was carried out on a Finnigan Trace DSQ mass spectrometer (ThermoFisher Scientific, Loughborough, UK) using electron ionization (EI) coupled to a GC Pal autosampler (CTC, Basingstoke, UK) and a Broad Trace Ultra GC system using splitless injection (1 ll). Separation was carried out using helium carrier gas (continuous flow, 1 ml/min) and a splitless injector with the following conditions: injector temperature 280 °C and splitless time 0.75 min. Amino acids were separated using a Zebron ZB-5HT INFERNO GC column (15 m  0.25 mm, 0.25 lm film thickness) with a Z-guard column (length 2 m, 0.53 mm i.d.) (both from Phenomenex, Macclesfield, UK). The GC temperature program used throughout was as follows: 130 °C for 1 min, 35 °C/min to 360 °C, hold 3 min, and total run time of 10.6 min (cycle time 17 min). The MS conditions were as follows: transfer line temperature 300 °C and source temperature 250 °C. Data were recorded in selected ion monitoring (SIM) mode, and the mass-to-charge ratios (m/z) monitored for the natural and labeled analogs were 442, 448 (Arg); 274, 279

42

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

(Ile); 302, 309 (Leu); 234, 243 (Phe); 258, 263 (Pro); and 260, 265 (Val). Results and discussion Derivatization and GC–MS method optimization The traditional method of amino acid analysis uses pre- or postcolumn derivatization with reagents such as ninhydrin and phenylisothiocyanate coupled with optical detection. These methods are highly reproducible but can be slow and lacking in analyte specificity. As such, they are being replaced by MS-based detection so that increased numbers of peptide standards with a variety of sequences can be measured accurately. Many LC–MS methods for amino acid analysis require derivatization, and consequently issues such as derivative instability, reagent interferences, and unwanted side reactions must typically be considered in detail for accurate quantification [21]. However, these can be reduced with the use of isotopically labeled standards and EM–IDMS [22]. Under such an internal standardized calibration regime, the results are based on the measured peak area ratio response between the analyte and its isotopically labeled form (N/L ratio). Once equilibration of the isotopic and natural forms within the sample has been achieved, both species will behave in an identical manner such that losses during sample preparation will be accounted for [17]. Consequently, extensive validation and analysis of recoveries for the derivatization step are not required. However, the hydrolysis step must be fully evaluated because the internal standards used do not account for this. GC–MS has been used for amino acid analysis for many years [14,16] and provides high sensitivity and selectivity, good chromatographic resolution, and rapid analysis times. A method for rapid GC–MS analysis of amino acids was developed here using a mix of standard amino acids (NIST SRM 2389). Because the samples to be analyzed using this method are clean standard materials, a shorter column (15 m) was chosen given that the required resolution could be achieved in shorter analysis times. Although derivatives of all the amino acids present in the sample were observed, equimolar release of multiple amino acids is required to confirm that hydrolysis has proceeded to completion and only nine amino acids (Ala, Arg, Gly, Ile, Leu, Lys, Phe, Pro, and Val) were stable under the hydrolysis conditions used. Good separation of these nine amino acids was achieved with a total run time of 10.6 min (cycle

Fig.1. TIC of SIM masses with the nine amino acids used for quantification highlighted. Baseline separation of isomers Ile and Leu is evident, with a total separation time of less than 6 min for all amino acids.

time 17 min; see Fig. 1 for separation). This is significantly quicker than conventional LC–UV amino acid analyses (>90 min) [23] and faster than many LC–MS methods (20–40 min) [15,22]. Isoleucine and leucine isomers, typically not baseline separated by LC, were separated by 0.1 min with a resolution of Rs = 5. For a method designed to assign values to standard materials for primary calibration materials, reducing the associated uncertainty is essential. To give the most well-defined peak shape, and consequently the lowest N/L ratio precision, a maximum of four SIM masses, corresponding to the natural and label of two amino acids, were monitored in each time window. Blanks were run to ensure that no interferences were present at the m/z SIM masses monitored for the natural and labeled amino acids. The instrument setup was optimized such that the precision of five repeat N/L ratio measurements was consistently less than 1.0% RSD (relative standard deviation) for all amino acids. Most amino acids converted rapidly to the tert-butyldimethysilyl derivatives, with conditions of 85 °C and 60 min providing the most intense signal for GC–MS analysis. Greater signal intensity improves the reproducibility of the measurements and reduces the uncertainty associated with the N/L peak ratio measurements. Signal intensities of the tert-butyldimethysilyl derivatives of the most basic residues, such as arginine and lysine, were typically an order of magnitude lower than those for other amino acids due to the poor formation of the derivative [14], although ratio precision was still comparable (0.5–1.0% RSD). Amino acid purity Natural amino acid standard solution concentrations were determined by comparison of solution standards prepared gravimetrically from the solid amino acid materials against the NIST standard reference material 2389 using EM–IDMS by GC–MS analysis. The purities determined for each solid material (provided with a stated purity of >99%) were as follows: arginine 98.8 + 1.2/ 2.9%, isoleucine 94.9 ± 2.5%, leucine 100.0 + 0/ 2.8%, phenylalanine 97.1 ± 2.6%, proline 98.3 + 1.7/ 2.7%, and valine 97.4 ± 2.6%. The discrepancies between the measured and stated purities of more than 99% indicate the requirement for back-calibration to a certified standard so as to provide results traceable to the SI. Hydrolysis optimization Complete peptide hydrolysis is critical for traceable quantification to be achieved because the assumption that peptide concentration can be inferred from amino acid concentration is based on the criterion of complete amino acid release [24]. This can be assessed by monitoring the ratio of natural amino acid release against the labeled amino acid spike added prior to the hydrolysis. Some amino acids are released more slowly than others, particularly Ile and Val, due to the steric effects of the hydrocarbon side chain preventing facile access to the carbonyl bond [25]. Monitoring the release of a number of amino acids (P3) per peptide, including those that are sterically hindered, and that the results are equimolar provides further confidence in the quantification method. Furthermore, the amino acids must also be stable throughout the hydrolysis conditions. Natural amino acids must first be released from the peptide before degradation can occur, whereas the labeled amino acids are free throughout the hydrolysis. Any degradation effects will influence the labeled amino acids to a greater degree, causing a positive bias in the results [11]. This immediately excludes a number of amino acids, such as methionine that is partially oxidized during hydrolysis, as potential quantification candidates. Stabilizers and protective agents (e.g., phenol) can be added to improve recovery of the more labile amino

43

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

Fig.2. Results for hydrolysis conditions over 1 h at temperatures of 150 °C (A), 160 °C (B), and 165 °C (C), showing improved equimolar release of six amino acids on increased temperature (all 200 W, two repeats). (D) Hydrolysis after 3 h at 165 °C, showing equimolar release of all six amino acids and improved repeatability of method. The lower solid line represents the average value for all amino acids, and the upper line represents the CRM value. Error bars on all graphs represent standard individual measurement uncertainties.

Table 1 Examples of peptide standards intended for protein quantification that were quantified using the microwave hydrolysis method described in this article. Peptide sequence

Protein of interest

Amino acids used for quantification

Molar concentration (nmol/g)

YSFLQNPQTSLCFSESIPTPSNREETQQKSNLELLR TGQIFK AFVFPK APLTKPLK YEVQGEVFTKPQLWP

Human growth hormone Human growth hormone C-reactive protein C-reactive protein C-reactive protein

L, I, P, R, F G, L, I, F A, V, K, P, R L, A, K, P G, L, V, K, P, F

46.73 ± 2.13 48.84 ± 2.59 66.62 ± 1.15 78.54 ± 1.60 20.32 ± 1.05

(4.6%) (5.3%) (1.7%) (2.0%) (5.2%)

Note: a variety of sequence lengths and amino acid combinations are included to demonstrate the wide applicability of the approach. Uncertainties (in parentheses) are expanded total measurement uncertainties (k = 2).

acids [13], but they were found to scavenge the derivatization agent, resulting in poor sensitivity. This resulted in a pool of nine ‘‘stable’’ amino acids (Ala, Arg, Gly, Ile, Leu, Lys, Phe, Pro, and Val) that could potentially be used for peptide quantification. From this pool, we have set the criterion that, for peptides used for further protein quantification, a minimum of three amino acids is required to quantify each peptide while ensuring that hydrolysis has proceeded to completion. Of the potential nine amino acids that can be used, six were present in the angiotensin sample and all were measured in this study. Acid hydrolysis is conventionally performed under vacuum with hydrochloric acid at 110 °C for several hours [13]. Microwave irradiation is known to accelerate reactions such as enzymatic digestion and acid hydrolysis [19,20,26]. With the advent of specifically designed benchtop systems for microwave hydrolysis that use small sample volumes (100–300 ll), uniform microwave fields, and full temperature control, improved reproducibility of the hydrolysis step is expected. Here we investigated microwave hydrolysis as a potential alternative to oven hydrolysis for the quantification of the peptide angiotensin I. Both temperature and length of hydrolysis time were investigated for the microwave hydrolysis system. Temperature is a major factor in reducing the hydrolysis time; it has been shown that

for each 10 °C increase in temperature, the rate of hydrolysis increases by a factor of 2 [13]. However, the maximum temperature that could be investigated due to the pressure limitations of the vessel was 165 °C. A combination of increased temperature (165 °C) and time (3 h) was finally used to provide the reproducibility and accuracy required for absolute quantification. Amino acid release was monitored by measuring the peak area ratio relative to the labeled amino acid spike added to the sample prior to hydrolysis. Multiple samples were run in each hydrolysis experiment to determine the variability due to the conditions to which the samples were subjected. Initially, a microwave hydrolysis program of 15 min at 150 °C (150 W) was run. It was found that this was not quantitative given that release was only 70–80% of that expected for Phe, Leu, and Pro, approximately 55% of that expected for Ile and Arg, and less than 30% of that expected for Val. The length of hydrolysis time was then increased to 1 h, and this was used as a minimum time for future experiments as amino acid release at this point was increased to more than 80% for all amino acids, including the more sterically hindered valine and isoleucine. Effects of temperature increases were compared for 1-h hydrolysis runs at 150, 160, and 165 °C using a power of 200 W in all cases (see Fig. 2). Release of four of the amino acids (Phe, Pro, Leu, and Arg) was apparently complete at all three temperatures, although

44

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

variability of these results improved from 5% RSD at 150 °C to 1.5– 2.0% RSD at the higher temperatures, suggesting a more reproducible hydrolysis. Both Ile and Val were not fully released at 150 °C (83% and 89%, respectively) or 160 °C (92% and 97%, respectively), but at 165 °C Val had been completely released after 1 h, although Ile was still incomplete (94%). Due to pressure limitations of the system, further increases in temperature could not be investigated and the hydrolysis time was subsequently extended at this temperature to achieve total equimolar release for Ile so as to perform accurate peptide quantification. Hydrolysis times of 1, 2, and 3 h were compared at 165 °C and 200 W. Increasing the hydrolysis time improved the reproducibility of the repeat hydrolysis experiments such that the variability between the samples for each amino acid was reduced to less than 2% RSD after 3 h (see Fig. 2C and D). In addition, the variability between the results across the six amino acids decreased with increased hydrolysis time from 4% RSD for 1 h to less than 2.5% RSD for 3 h, giving equimolar results. This indicates that all amino acids, including those known to have slow reaction kinetics (Val and Ile), are completely released from the peptide after 3 h and that quantification can be performed with high confidence and low uncertainty. Fig. 2D shows the results obtained by using the optimal conditions of 165 °C and 3 h. The optimized hydrolysis conditions described here were shown to give equimolar release for a number of peptides of varying amino acid sequences and sequence lengths for use in protein quantification. As such, the method shows general applicability to many peptides. Examples of results obtained from a number of peptides are shown in Table 1. However, it is worth noting that to provide accurate results, it is necessary to validate the hydrolysis conditions used for each potential peptide of interest because resistance to hydrolysis may vary to some extent depending on the amino acid sequence. Quantification and uncertainty The optimized microwave hydrolysis GC–MS method (165 °C, 200 W, 3 h) was then used to quantify an angiotensin standard sample analyzed previously by oven hydrolysis combined with

EM–IDMS by LC–MS/MS. Three repeat experiments, each with three sample and calibration blends, were run with the final microwave hydrolysis conditions, and six amino acids were monitored. Mass fractions were calculated for each individual amino acid using the EM–IDMS equation [17,18] and converted to the amount of substance content (mol/g) to determine the final concentration of the angiotensin I sample and demonstrate total molar release of all amino acids from the peptide (individual results shown in Fig. 3). Using the microwave digestion GC–MS method, a concentration of 56.6 ± 3.2 nmol/g (expanded uncertainty of 5.6%, k = 2) was obtained. Results reported here were in agreement with those obtained previously by oven hydrolysis and LC–MS/MS and agreed well with the certified value (57.5 ± 1.73 nmol/g). The amount of substance concentration (nmol/g) of the angiotensin I peptide was calculated by averaging the amount of substance concentration of the six amino acids measured in the hydrolysate. The measurement uncertainty was estimated in accordance with EURACHEM and the ISO Guide to the Expression of Uncertainty in Measurement [27] as described by Burkitt and coworkers [18]. Briefly, individual uncertainties for each amino acid result were calculated from, and further combined with, the uncertainty associated with the blend-to-blend variation (standard deviation of the mean of all 27 amino acid results). The experiments were designed in such a way as to evaluate the uncertainty of the microwave hydrolysis GC–MS method, and as such a single bulk sample was used. Consequently, neither the uncertainty associated with sample homogeneity nor gravimetric preparation was assessed, and a different uncertainty equation should be used when multiple independent samples are prepared. The total uncertainty used here contains a component associated with the between-blend hydrolyses to account for the global variability of the hydrolysis between repeat samples and a component associated with the multiple amino acids used to infer peptide concentration, ensuring that any variability associated with the extent of digestion is also encompassed. The uncertainty budget is dominated by the uncertainty associated with the purity of the amino acid standards (>70%) (Fig. 4). If the amino acid purity uncertainty component is removed from the calculation, the total combined expanded uncertainty falls to 3.0%

Fig.3. GC–MS results for angiotensin acid hydrolysis showing results for all six amino acids. Three repeat experiments (batches 1–3, with each batch containing three samples) were run, and all nine results are shown here. Error bars represent individually calculated uncertainties, and the solid line represents the average of all amino acid values.

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

45

Fig.4. Total combined uncertainty budget for microwave hydrolysis GC–MS analysis of angiotensin (standard unexpanded uncertainty of 2.8%). The most significant component is due to amino acid standard purity; with this removed, the standard uncertainty due solely to the method is determined to be 1.5%. To further reduce measurement uncertainty for peptide standards, further improvements to the purity of amino acid primary standards are necessary.

(k = 2). This clearly shows the importance of the availability of amino acid standards with well-characterized purity measurements in producing traceable peptide standards with the lowest uncertainties. Conclusions A rapid and robust method for absolute, SI-traceable peptide quantification via microwave hydrolysis and GC–MS analysis with a combined expanded standard measurement uncertainty of less than 6%, where the uncertainty component attributed to the uncertainty of the pure amino acid standards was 2.6%, has been described. Complete equimolar release of the amino acids of interest, as required for absolute peptide quantification, was demonstrated to occur in 3 h at 165 °C for microwave hydrolysis. This represents a 95% reduction in the time required for hydrolysis over the traditional oven hydrolysis method (56 h). A standard sample of angiotensin I was used to demonstrate the accuracy and precision of this method, and results were in agreement with previous oven hydrolysis LC–MS/MS results. Quantification was performed by EM–IDMS, and results between the two methods agreed within the measurement uncertainties. The method described here is currently applied in our laboratory to absolutely quantify peptide solutions that are subsequently used as standards for absolute quantification of clinically relevant proteins such as human growth hormone and C-reactive protein by EM–IDMS [5] and using stable isotopically labeled peptides as internal standards. The availability of traceable standards is fundamental to maintaining the traceability chain and to ensuring interlaboratory and batch-to-batch reproducibility. This is of particular importance when the proteins quantified are to be used as primary standards in clinical environments. Acknowledgments The research within the EURAMET joint research project leading to these results received funding from the European Community’s Seventh Framework Programme, ERA-NET Plus, under grant 217257. The work described was supported, in part, by the UK National Measurement System Chemical and Biological Metrology Programme. References [1] J. Stewart, Bradykinin antagonists as anti-cancer agents, Curr. Pharm. Des. 9 (2003) 2036–2042.

[2] B. Szende, G. Kéri, TT-232: a somatostatin structural derivative as a potent antitumor drug candidate, Anticancer Drugs 14 (2003) 585–588. [3] B. Yegen, Bombesin-like peptides: candidates as diagnostic and therapeutic tools, Curr. Pharm. Des. 9 (2003) 1013–1022. [4] H. John, M. Walden, S. Schaefer, S. Genz, W. Forssmann, Analytical procedures for quantification of peptides in pharmaceutical research by liquid chromatography–mass spectrometry, Anal. Bioanal. Chem. 378 (2004) 883– 897. [5] C. Pritchard, M. Quaglia, W.I. Burkitt, C. Mussell, G. O’Connor, H. Parkes, Fully traceable absolute protein quantification of somatropin: allowing independent comparison of somatropin standards, Clin. Chem. 55 (2009) 1984–1990. [6] E.P. Diamandis, Mass spectrometry as a diagnostic and a cancer biomarker discovery tool: opportunities and potential limitations, Mol. Cell. Proteomics 3 (2004) 367–378. [7] J. Villanueva, D.R. Shaffer, J. Philip, C.A. Chaparro, H. Erdjument-Bromage, A.B. Olshen, M. Fleisher, H. Lilja, E. Brogi, J. Boyd, M. Sanchez-Carbayo, E.C. Holland, C. Cordon-Cardo, H.I. Scher, P. Tempst, Differential exoprotease activities confer tumor-specific serum peptidome patterns, J. Clin. Invest. 116 (2006) 271–284. [8] R. Aebersold, M. Mann, Mass spectrometry-based proteomics, Nature 422 (2003) 198–207. [9] S. Ong, M. Mann, Mass spectrometry-based proteomics turns quantitative, Nat. Chem. Biol. 1 (2005) 252–262. [10] W. Yan, S.S. Chen, Mass spectrometry-based quantitative proteomic profiling, Brief Funct. Genomics Proteomics 4 (2005) 27–38. [11] C.G. Arsene, R. Ohlendorf, W. Burkitt, C. Pritchard, A. Henrion, G. O’Connor, D.M. Bunk, B. Güttler, Protein quantification by isotope dilution mass spectrometry of proteolytic fragments: cleavage rate and accuracy, Anal. Chem. 80 (2008) 4154–4160. [12] M. Quaglia, C. Pritchard, Z. Hall, G. O’Connor, Amine-reactive isobaric tagging reagents: requirements for absolute quantification of proteins and peptides, Anal. Biochem. 379 (2008) 164–169. [13] M. Fountoulakis, H.W. Lahm, Hydrolysis and amino acid composition of proteins, J. Chromatogr. A 826 (1998) 109–134. [14] H.J. Chaves das Neves, A.M. Vasconcelos, Capillary gas chromatography of amino acids, including asparagine and glutamine: sensitive gas chromatographic–mass spectrometric and selected ion monitoring gas chromatographic–mass spectrometric detection of the N, O(S)-tertbutyldimethylsilyl derivatives, J. Chromatogr. 392 (1987) 249–258. [15] M. Kato, H. Kato, S. Eyama, A. Takatsu, Application of amino acid analysis using hydrophilic interaction liquid chromatography coupled with isotope dilution mass spectrometry for peptide and protein quantification, J. Chromatogr. B 877 (2009) 3059–3064. [16] M.W. Duncan, A. Poljak, Amino acid analysis of peptides and proteins on the femtomole scale by gas chromatography/mass spectrometry, Anal. Chem. 70 (1998) 890–896. [17] A. Henrion, Reduction of systematic errors in quantitative analysis by isotope dilution mass spectrometry (IDMS): an iterative method, Fresenius J. Anal. Chem. 350 (1994) 657–658. [18] W.I. Burkitt, C. Pritchard, C. Arsene, A. Henrion, D. Bunk, G. O’Connor, Toward Système International d’Unité-traceable protein quantification: from amino acids to proteins, Anal. Biochem. 376 (2008) 242–251. [19] J.R. Lill, E.S. Ingle, P.S. Liu, V. Pham, W.N. Sandoval, Microwave-assisted proteomics, Mass Spectrom. Rev. 26 (2007) 657–671. [20] J.M. Collins, N.E. Leadbeater, Microwave energy: a versatile tool for the biosciences, Org. Biomol. Chem. 5 (2007) 1141–1150. [21] K. Petritis, P. Chaimbault, C. Elfakir, M. Dreux, Parameter optimization for the analysis of underivatized protein amino acids by liquid chromatography and ionspray tandem mass spectrometry, J. Chromatogr. A 896 (2000) 253–263.

46

Traceable quantification of peptide standards / C. Pritchard et al. / Anal. Biochem. 412 (2011) 40–46

[22] H. Lee, S. Park, G. Lee, Determination of phenylalanine in human serum by isotope dilution liquid chromatography/tandem mass spectrometry, Rapid Commun. Mass Spectrom. 20 (2006) 1913–1917. [23] H. Kaspar, K. Dettmer, W. Gronwald, P.J. Oefner, Advances in amino acid analysis, Anal. Bioanal. Chem. 393 (2009) 445–452. [24] M. Weiss, M. Manneberg, J.F. Juranville, H.W. Lahm, M. Fountoulakis, Effect of the hydrolysis method on the determination of the amino acid composition of proteins, J. Chromatogr. A 795 (1998) 263–275.

[25] D. Roach, C.W. Gehrke, The hydrolysis of proteins, J. Chromatogr. 52 (1970) 393–404. [26] P.M. Reddy, W.-Y. Hsu, J.F. Hu, Y.-P. Ho, Digestion completeness of microwaveassisted and conventional trypsin-catalyzed reactions, J. Am. Soc. Mass Spectrom. 21 (2010) 421–424. [27] International Organisation for Standardisation, Guide to the Expression of Uncertainty in Measurement, ISO, Geneva, Switzerland, 1995.