Quantitative - UPLC - MS - MS - analysis - of - underivatised - amino - acids - in - body - fluids - is - a-reliable - tool - for - the - diagnosis - and - follow - up

Quantitative - UPLC - MS - MS - analysis - of - underivatised - amino - acids - in...

(Parte 1 de 3)

Quantitative UPLC-MS/MS analysis of underivatised amino acids in body fluids is a reliable tool for the diagnosis and follow-up of patients with inborn errors of metabolism

W.A. Huub Waterval 1, Jean L.J.M. Scheijen 1, Marjon M.J.C. Ortmans-Ploemen, Catharina D. Habets-van der Poel, Jörgen Bierau ⁎

Laboratory of Biochemical Genetics, Department of Clinical Genetics, Maastricht University Medical Centre, PO Box 5800, 6202 AZ, Maastricht, The Netherlands abstractarticle i nfo

Keywords: Amino acids Tandem mass spectrometry LC-MS/MS Inborn errors of metabolism Validation Ninhydrin detection

Background: An electro-spray ionisation ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) application for the quantitative analysis of amino acids was developed. The suitability for the detection and follow-up of patients suffering from inborn errors of metabolism (IEM) was assessed by extensive cross-validation with ion-exchange liquid chromatography (IEX-LC) with post-column ninhydrin derivatisation, participation in external quality control (ERNDIM) and analysis of samples of patients with confirmed IEM. Methods: Prior to analysis plasma and urine samples were merely diluted 150-fold in mobile phase. Amino acids were detected in the multiple reaction monitoring mode (MRM) in the ESI-positive mode. The analytical results were compared with IEX-LC. External quality control scheme performance is presented. Results: Comprehensive analysis of amino acids in plasma and urine was achieved with a run-to-run time of 30 min.ValidationresultsweresatisfactoryandtherewasaverygoodcorrelationbetweenUPLC-MS/MSandIEXLC. Analytical results obtained in the external quality control scheme were essentially the same as those of the other participants. Patients suffering from IEM were readily identified. Conclusion:UPLC-MS/MSanalysisofaminoacidsinbodyfluidsisrapid,reliableandsuitableforthediagnosisand follow-up patients with IEM. © 2009 Elsevier B.V. All rights reserved.

1. Introduction

Quantitative analysis of amino acids (A) in body fluids is the basis for the diagnosis of inborn errors of metabolism affecting amino acid metabolism. The most frequently applied method for the quantitative analysis of physiological A in body fluids is ion-exchange chromatography (IEX-LC) with post-column ninhydrin derivatisation [1]. This technique is referred to as classic amino acid analysis and is performed using dedicated equipment having excellent reproducibility. Moreover, there is a vast world wide experience with this technique. Quantitative analysis of all physiological amino acids is time-consuming, often requiring 2–3 h per sample, and is not suited for high sample throughput. Moreover, numerous artefacts hamper the method and certain compounds cannot be separated using this technique.

Tandemmassspectrometry(MS/MS)isapowerfultechniqueforthe quantitative analysis of small molecules. The technique is based on physico-chemical characteristics of analytes in an electrical field at near vacuum and subsequent collision induced fragmentation [2]. In general practice, a tandem mass spectrometer is operated in combination with high performance liquid chromatography (HPLC), which greatly increases its analytical performance [3,4]. Ions and other components from the sample matrix suppressing the signal output are removed and potential isobars are separated. HPLC-MS/MS facilitates rapid and specific analysis. Ultra-performance liquid chromatography (UPLC) has evolved from HPLC. Due to its greater analytical power, UPLC allows down scaling of sample volume and high throughput sample handling. UPLC-MS/MSsupersedes HPLC-MS/MS in analyticalperformance [5,6].

In recent years the use of HPLC-MS/MS has become more popular in hospital laboratories and is applied to the quantitative analysis of small molecules such as metabolites and pharmaceuticals [2].I ti s a powerful tool in the diagnosis of inborn errors of metabolism [7] and in newborn screening programs [8]. Most HPLC-MS/MS methods are dedicated applications for the detection or therapeutic follow-up for one specific disease or group ofdiseases.MS/MS applications covering

Abbreviations: A, Amino acid(s); IEX-LC, Ion-exchange liquid chromatography;

MS/MS, Tandem mass spectrometry; HPLC, High performance liquid chromatography; UPLC, Ultra-performance liquid chromatography; IS, Stable isotope-labelled internal standard(s); SSA, Sulphosalicylic acid; IEM, Inborn errors of metabolism; ERNDIM, European Research Network for evaluation and improvement of screening, Diagnosis and treatment of Inherited disorders of Metabolism; TDHFA, Tridecafluoroheptanoic acid; MRM, Multiple reaction monitoring; LOD, Limit of detection; LOQ, Lower limit of quantification; ULQ, Upper limit of quantification; NTBC, 2-(2-Nitro-4-trifluoromethylbenzoyl)-1,3-cyclohexanedione. ⁎ Corresponding author. Tel.: +31 43 387 78 35; fax: +31 43 387 79 01.

E-mail address: jorgen.bierau@gen.unimaas.nl (J. Bierau).These authors contributed equally to the work presented.

Contents lists available at ScienceDirect

Clinica Chimica Acta jou rnal homepage: w.else vier .com/locate/clinchim a wider range of disorders are the (semi-)quantitative analysis of acylcarnitines for the detection of defects in fatty acid oxidation and organic acidaemias [7,9] and the analysis of purines and pyrimidines in urine [10,1]. It has been demonstrated in non-human samples that analysis of underivatised A by tandem mass spectrometry is possible [12,13]. Piraud and co-workers demonstrated that HPLC-MS/MS analysis of A and related compounds for the detection of inborn errors is in theory feasible [14–16]. Extensive data on the practicality and validity of LC-MS/MS analysis of A in comparison with the established techniques have hitherto not been reported.

In our laboratory a project was initiated to replace the classic amino acid analyser with a UPLC-MS/MS application. A reliable, robust and simplified method was developed, validated and implemented for the comprehensive quantitative analysis of underivatised A in body fluids using UPLC-MS/MS requiring minimal sample volumes and manipulations. The analytical results were compared with those of the classic amino acid analyser and of other laboratories by means of participation in the ERNDIM external quality control scheme. The suitabilityof the method todetect patients suffering from inborn errors of metabolism is demonstrated.

2. Materials and methods 2.1. Chemicals

A standards and a prefabricated A standard solution (A9906) were obtained from Sigma-Aldrich (Zwijndrecht, The Netherlands). Stable isotope-labelled (IS) A were obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA). All solutions were prepared using LC-MS Ultra High Purity water and LC-MS-grade acetonitrile (J.T. Baker, Deventer, The Netherlands). Tridecafluoroheptanoic acid (TDFHA) 9% was obtained from Sigma-Aldrich (Zwijndrecht, The Netherlands). Sulphosalicylic acid (SSA) was obtained from Acros Organics (Geel, Belgium).

2.2. Patient samples

In 2008, all patient samples were routinely analysed using the method presented, in total 517 plasma samples and 365 urine samples were analysed. Samples of patients with known IEM were re-analysed. Urine samples from the patients with FIGLUria and prolidase deficiency were provided by Dr. L. Dorland, Department of Metabolic and Endocrine Diseases, University Medical Center Utrecht, The Netherlands. The urine samples from the patients with N-Aspartylglucosaminuria and Ornithine δ-aminotransferase deficiency were from the ERNDIM diagnostic proficiency testing scheme. Samples were used according to the “Code for proper use of human tissue” as formulated by the Dutch Federation of Medical Scientific Societies.

2.3. Standards

The purchased standard A solution contained 500 µM of each A. A stock solution containing 500 µM of L-S-Cys, PEA, L-Asn, L-Gln, L-Hci, DALA, Trp, L-Hcar, L-Sac, L-Aile, Asa was prepared in water. Prior to analysis this stock solution was mixed 1:1 (v/v%) with the purchased A standard solution (A9906) resulting in final concentrations of 250 µM for each A. A mixture of stable isotope-labelled A was prepared in 0.1 M HCl and was used as internalstandard for UPLC-MS/MS analysis.Theconcentration of the IS A was between 70 and 800 µM. The stable isotope-labelled A used are listed in Table 1.

2.4. Standard sample preparation for UPLC-MS/MS analysis

Patient samples were stored at −20°C and thawed in a water bath at 37°C prior to analysis and homogenated by vortex mixing. Ten µl of plasma or urine was mixed with 10 µlof the internal standard solution and 1500 µlof 0.5 mM TDHFA inwater. The sample was then ready for analysis.

2.5. Ion-exchange liquid chromatography (IEX-LC)

IEX-LC with post-column ninhydrin-derivatisation [1] was performed using a

Biochrom-20 A analyzer (Cambridge, UK) using the separation programme for physiological fluids supplied by the manufacturer. Patient samples were stored at −20°C and thawed in a water bath at 37°C prior to analysis and homogenated by vortex mixing. To 150 µl of plasma or urine 150 µl of 5% (m/v) SSA containing 300 µM aminoethylcysteine as the internal standard was added. The mixture was mixed thoroughly and incubated for 15 min on ice and subsequently centrifuged for 10 min at 25,000×g at 4 °C. The supernatant was then filtered through a 0.2 µm nylon filter (Costar, Corning, NY) for 1 min at 25,000×g at 4 °C.170 µl of the filtratewas transferred to a clean injection vial. The injected volume was 50 µl.

2.6. Ultra-performance liquid chromatography conditions

Liquid chromatography was performed at 30°C using a Acquity UPLC BEH C18, 1.7 µm, 2.1×100 m column (Waters, Milford MA) and a gradient system with the mobile phase consisting of buffer A; 0.5 mM TDFHA and buffer B; 0.5 mM TFHA in acetonitrile (100%) at a flow rate of 650 µl/min (split less). The gradient programme used was: initial 9.5% A and 0.5% B; linear gradient to 70% A and 30% B in 14 min; hold for 3.5 min, returntoinitial conditions in 1 min at a flow rate of 700 µl/min, followed by equilibration for 10 min. One minute prior to the next sample injection the flow was set to 650 µl/min. Run-to-run time was 30 min. The injected volume was 5 µl.

2.7. Tandem mass spectrometry conditions

Mass spectrometry experiments and optimisation of the method were performed usingaMicromassQuattroPremierXETandemMassSpectrometer(Waters,Milford,MA).

Themassspectrometerwasusedinthemultiplereactionmonitoringmode(MRM)in the ESI-positive mode. The desolvatation temperature was 450 °C, and the source temperature was 130 °C. The capillary voltage was set at 0.5 kV and the cone voltage was set at 25 V. Nitrogen gas was used as desolvatation gas and as cone gas. Nitrogen gas was producedusingaNM30Lnitrogengenerator(PeakScientific,Renfrewshire,Scotland).The cone gas flow was 50 l/h and the desolvatation gas flow was 800 l/h. Optimal detection conditionsweredeterminedbyconstantinfusionof standardsolutions(50 µM)insolvent A using a split system. MRM and daughter-ion scans were performed using argon as the collision gas at a pressure of 3.8×10 mbarand a flow of 0.2 ml/min. Apparatus settings for detection and internal standards used are shown in Table 2.

3. Validation procedure 3.1. Linearity, accuracy and detection limits

Because of the number of analytes and complexity of the sample matrices,linearityandaccuracy(recovery)weredeterminedbyspiking3 independent plasma,and urinesamplesand water with A in dissolved water. Between 0 and 1000 µmol/l per A was added to the sample matrix.Therecoverywasdeducedfromtheslopeofthecalibrationcurve and corrected forthe recoveryinwater to minimisedifferences between thedifferentstandardsused.Thelimitofdetection(LOD)wasdetermined inwateratasignal-to-noiseratioof3,lowerlimitsofquantification(LOQ) were determinedin plasmaand urineat a signal-to-noiseratio of 5.

3.2. Precision (intra- and inter-assay variation)

Theintra-assayvariationwasdeterminedby10consecutiveanalyses of basal and spiked plasma and urine samples. In order to obtain a samplewithhighconcentrationsofAAthatnormallyarepresentintrace amounts80–150 µmol/loftheseAAwereadded.Thiswasalsodoneina

Table 1 Twenty isotope-labelled A used as internal standards with the specificm ass transitions (MRM), declustering potentials (DP in V), collision energies (CE in eV), dwell times (DT in ms) and retention times (RT in min).

Abbreviation Monitoredtransition DP CE DT RT

urine sample to which 80–540 µmol/l of selected A were added. This ensuredthatthevariationcouldbeassessedforallAAatahighandalow level. The inter-assay variation was determined by analysis of basal and spiked plasma and urine samples on, respectively,14 and 18 days.

3.3. Comparison with Biochrom-20 analysis

In comparative studies IEX-LC was considered as the reference method. Plasma and urine samples were simultaneously prepared for analysis using both techniques. For both matrices correlation curves wereconstructedforeachAAphysiologicallypresent(x-axis:Biochrom- 20, y-axis UPLC-MS/MS). Only data points within the reference range were used. Bland–Altman analyses were performed for all A in both matrices [17]. The difference between the results obtained with the

Biochrom-20 and UPLC-MS/MS (Biochrom-20 minus UPLC-MS/MS) was plotted against the mean value of both methods. Results are presentedasthemeandeviation(%)ofallmeasurementsperAAineach matrix and the mean difference in absolute measured value (µmol/l), between 21 and 50 data points were obtained.

3.4. External quality control

Inter-laboratory variation was assessed by participation in the

ERDNIM quantitative external quality control scheme [18]. This scheme has more than 100 participants. The measured value was expressed relative to the median of all submissions by all participants ((our value/median of participants)*100%). The mean of these results and the standard deviation were calculated for each series. The 2008 series were initially analysed with the method presented. The 2005 to 2007 series were re-analysed, taking into account artefacts of storage and freeze-thawing.

4. Results 4.1. Linearity and recovery

Linear calibration curves (R2N0.9) were obtained in water, plasma and urine for all A. The recovery was 100±10% for the vast majority of the compounds in plasma and urine. In plasma, R2b0.9 was observed for S-Cys, PSer, PEA, Gln, Sac, Ans and Car. Recoveries N110% inplasmawereobservedfor S-Cys, PSer, Ala, Asa, Hisand Trp. In urine, R2b0.9 was observed for S-Cys, PSer, PEA, (Cys)2, Hcy(ala), Abu, Hci and Sac. Recoveries N110% in urine were observed for PSer and Trp and recoveries b90% were observed for PEA, DALA and Car.

A with highly polar or charged side chains, i.e. S-Cys, PSer, PEA and Tau showed minimal retention and co-eluted with (inorganic) ionsandpolarcompoundsfromthesamplematrix.Thiscausedmatrix effects in the form of ion suppression and hence less than satisfactory results. However, the quantification of Tau was nonproblematic because the IS adequately corrected for ion suppression.

(Parte 1 de 3)