| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
(Department of Clinical Biochemistry, Vejle County Hospital, Kabbeltoft 25, Vejle, DK 7100, Denmark;
aauthor for correspondence: fax 45-79406853, e-mail anvisc{at}vgs.vejleamt.dk)
Abstract
Background: Methylmalonic acid (MMA) is a biochemical marker for cobalamin deficiency, particularly in cases where the cobalamin concentration is moderately decreased or in the low-normal range. Liquid chromatography–tandem mass spectrometry (LC-MS/MS) with electrospray ionization is a rapid, robust method that has been used in MMA analysis. We developed a simple method combining solid-phase extraction (SPE) and derivatization to prepare serum or plasma for LC-MS/MS analysis of MMA.
Methods: Deuterated internal standard d3-MMA was added to serum or plasma before SPE on strong anion-exchange (SAX) columns. After elution with HCl–butanol (10:90 by volume) and addition of 1 g/L formic acid, the samples were simultaneously derivatized and evaporated by heating to 70 °C for 15 min followed by 54 °C overnight in uncapped vials. Acetonitrile and 1 g/L formic acid were added to the samples before injection into the LC-MS/MS system. MMA and d3-MMA were quantified in the multiple-reaction monitoring mode. Calibrators were prepared in serum by the standard addition method.
Results: The MMA assay was linear up to 200 µmol/L. Interassay CVs were 6.7%, 5.0%, and 5.0% for mean concentrations of 0.15, 0.36, and 0.65 µmol/L, respectively.
Conclusions: Our simplified sample preparation and derivatization method is suitable for use in MMA analyses. MMA elutes with the derivatization reagent, and derivatization and evaporation are performed simply by leaving the uncapped vials in a heating block overnight. The method shows good linearity and precision.
The methylmalonic acid (MMA) concentration in serum or plasma is commonly used as a marker for cobalamin deficiency (1)(2)(3)(4). The preferred method for analysis has been gas chromatography–mass spectrometry (GC-MS) (5)(6)(7)(8)(9)(10), but GC-MS has the disadvantage of requiring a time-consuming sample preparation step involving either solid-phase extraction (SPE) (7) or liquid–liquid extraction (6)(8)(9)(10). Silyl derivatization (6)(8)(10) or preparation of alcohol esters (5)(7)(9) is also required, and the turnaround time for GC-MS is typically 15–25 min (5)(6)(7)(8)(9)(10).
Succinic acid (SA) is a product of methylmalonyl-CoA degradation, and serum concentrations usually range from 1 to 12 µmol/L; however, higher SA concentrations (>25 µmol/L) have been reported (11). On the other hand, serum MMA concentrations typically are <0.28 µmol/L. Because the molecules are very similar and have the same relative molecular mass, a complete lack of interference from SA must be demonstrated when developing an assay for MMA.
A liquid chromatography–tandem mass spectrometry (LC-MS/MS) method for the analysis of MMA has been published by Magera et al. (12). This method uses a sample preparation procedure very similar to those for GC-MS methods, but the esterification is performed with butanol instead of cyclohexanol, the alcohol preferred for GC-MS analyses. Sample preparation includes evaporation of the sample 3 times, with transfer to a new sample tube or vial after each evaporation step.
Kushnir et al. (11) also used LC-MS/MS, and their sample preparation before derivatization included liquid–liquid extraction. The advantage of their method is a very short chromatographic run of <1 min, with differentiation of MMA and SA based on differences in fragmentation of the 2 isomers in the collision cell. However, this differentiation could be instrument dependent.
We have developed a simple, automated SPE sample-preparation procedure that eliminates the need for repeated evaporation and transfer of serum samples for LC-MS/MS MMA analyses.
MMA was obtained from Fluka, and deuterated MMA (d3-MMA) and butanol (d9-butanol) were from Cambridge Isotopes Laboratory. SAX columns (50 mg) were purchased from Argonaut. Acetonitrile and methanol (both HPLC gradient grade) were from Baker, and butanol (HPLC grade) was from Merck. Formic acid and 37% HCl (both analytical grade) were from Baker. Autosampler vials with conical bottoms (1.1 mL) were purchased from Brown. The purified water used throughout met ASTM Type 1 specifications.
SPE was performed automatically on a Gilson Aspec XL4. The flow rates were 6 mL/min during column conditioning and 1 mL/min during sample loading, washing, and elution. The columns were purged with 1 mL of air after application of samples or solvents. After the final elution step, the needle was washed with methanol to remove any HCl–butanol before processing of the next sample.
Serum was recentrifuged before analysis to avoid clotting of the robot needle. The Gilson Aspec then performed all of the following steps: 750 µL of serum (calibrator, control, or sample) was mixed with 375 µL of d3-MMA (2 µmol/L in water). The SAX columns were conditioned with 1 mL of methanol and 1 mL of water before 750 µL of the mixed sample (corresponding to 500 µL of serum) was loaded. The columns were washed sequentially with 1 mL each of water, methanol, and butanol before elution of the MMA with 300 µL of a 10:90 mixture of concentrated (37%) HCl and butanol. Conical vials were placed directly in the collection racks of the Gilson Aspec. After sample collection, 80 µL of formic acid (1 g/L) was added to each vial.
After the eluate was vortex-mixed, MMA was derivatized to dibutyl-MMA directly in the elution solvent by incubation of the uncapped sample vials for 15 min at 70 °C in a heating block in a hood. After 15 min, the temperature was lowered to 54 °C, and the samples were left overnight for evaporation of the HCl–butanol reagent. The samples did not evaporate to dryness, and the amount of liquid left was 
100 µL. Before the vials were placed in the HPLC autosampler, 500 µL of acetonitrile–water (20:80 by volume) was added to the vials and the vials were capped.
The analysis was performed on a Waters 2795 Alliance HPLC system connected to a Micromass Quattro Micro tandem mass spectrometer. The column was a Waters Symmetry C18 cartridge [50 x 2.1 mm (i.d.); 3.5 µm bead size], and the column was eluted isocratically with 1 g/L formic acid–acetonitrile (35:65 by volume) at a flow rate of 0.2 mL/min. The injection volume was 5 µL, and the column temperature was 40 °C.
The multiple-reaction monitoring (MRM) transitions used for quantification of MMA and d3-MMA were m/z 231
119 and m/z 234122, respectively. The mass spectrometer was used in the positive-ion mode, and the collision gas was argon with a pressure of 0.3 Pascals. The desolvation temperature was 250 °C, the source temperature was 120 °C, the desolvation gas flow was 600 L/h, and the cone gas flow was 50 L/h. The capillary voltage was 3 kV, the cone voltage was 20 V, and the collision energy was 10 eV. The Q1 and Q3 mass resolutions were at unit mass (0.85 full width at half maximum). Quantitative data analysis was performed with QuanLynx software (Waters/Micromass).
Two sets of calibrators were prepared in pooled donor serum: one for routine analysis of MMA and one for checking the linearity of the method over an extended range.
For routine analysis, a stock solution of 1000 µmol/L MMA in water was added to aliquots of the pool to obtain MMA concentrations of 0.00, 0.25, 0.50, and 1.00 µmol/L. A linear least-squares regression with a weighting index of 1/x was performed on height ratios of MMA/d3-MMA vs MMA concentrations in the 4 human serum calibrators to generate a calibration curve (see Fig. 1 in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol52/issue4). The x-axis value at y = 0 was obtained from the regression line for the calibration curve; this value represents the amount of endogenous MMA in the serum pool. The calibrators were then assigned values of the added plus the endogenous amount, and a weighting index of 1/x was used throughout.
To evaluate the extended linearity, we added stock solution containing 10 000 µmol/L MMA in water to aliquots of the pool to obtain MMA concentrations of 25, 50, 75, 100, 150, and 200 µmol/L.
Control samples were prepared in the same way as described for the preparation of calibrators; a serum pool was prepared, and 0.5 µmol/L MMA was added to a portion of the pool. We also used a commercially available bovine serum control sample (K99; DEKS) with an assigned value of 0.38 µmol/L.
Chromatograms obtained with a normal serum sample containing 0.15 µmol/L MMA are shown in Fig. 1
. The MRM transitions show MMA, d3-MMA, and SA and demonstrate that there is no interference from SA because it is completely separated from MMA.
|
The MMA calibration curve was linear to 200 µmol/L (R2 = 1.00). The mean (SD) values and the relative standard deviation (RSD) for the slope (a) and intercept (b) obtained from 5 consecutive calibration curves used for routine analysis during 5 weeks are shown in Table 1
. Calculations of the endogenous amount were performed with data from the 5 calibration curves as well, and the results are summarized in Table 1
. All 5 calibration curves had a correlation (R2) >0.99. Recovery and precision data collected over a 3-month period with weekly analyses are also summarized in Table 1
.
|
We estimated the limit of quantification (LOQ) by assaying a serum sample containing 0.24 µmol/L MMA in a 5-fold dilution with water (0.048 µmol/L). The sample was analyzed on 3 different days, and the mean result (RSD) was 0.049 µmol/L (12%). This concentration met the validation requirements of a mean deviation and imprecision <20%. The estimated LOQ is well below the MMA concentration in any physiologic sample.
We assessed ion suppression by adding MMA derivatized with d9-butanol to 10 different prepared samples and comparing the signal to that for the same derivative added to acetonitrile–water (20:80 by volume). In 9 of 10 samples, the signal was decreased to 23%–65% of the value in acetonitrile–water, whereas in 1 sample, it was increased to 124%. However, because an internal standard is used and because the MMA signal generally is very high even for low MMA concentrations, this ion suppression does not compromise the analysis.
In the Scandinavian countries, a monthly external quality-control scheme for MMA has been available for many years, and until now, all participants have used GC-MS for the MMA analysis. For 12 samples in 2004, the linear regression equation for the present method (y) and the mean of results from the control scheme (x) was: y = 0.9685x – 0.0036 µmol/L (r2 = 0.990). A difference plot of the results is shown in Fig. 2 of the online Data Supplement.
It is generally recommended that esterification of carboxylic acids with alcohol should be performed in water-free solvents to avoid hydrolysis of the ester. In our experience, addition of 1 g/L formic acid before esterification and simultaneous evaporation improved the yield of dibutyl-MMA. Likewise, the yield of dibutyl-MMA was not affected by the water content in the concentrated HCl–butanol mixture.
To confirm that the lengthy evaporation overnight does not cause degradation of the MMA derivative, we compared our method with the sample preparation method published by Magera et al. (12). A serum calibrator containing 1.21 µmol/L MMA was prepared with both methods. The yield of dibutyl-MMA in the method of Magera et al. (12) was only 34% of the yield obtained in the present method [MMA peak height of 41 974 (CV = 6.4%) with the method described by Magera et al. (12); MMA peak height of 123 225 (CV = 3.5%) with our method].
Preparation of MMA calibrators in water is suggested in most publications on GC-MS methods and in the LC-MS/MS method published by Magera et al. (12). In our experience, the use of aqueous calibrators stored in aliquots at –20 °C gave rise to different derivatization yields from series to series and results that were consistently too high, possibly because of adsorption of MMA to tube walls during storing. Accordingly, in serum enriched with 0.50 µmol/L MMA, we obtained a recovery of the added amount of 120%. We therefore chose to use the standard addition method and prepare enriched serum calibrators. Another advantage of the serum-based calibrators is that the yields of derivatized MMA and d3-MMA, i.e., the peak heights, are consistently high. During method development, we chose a serum sample volume of 500 µL per SPE column, based on a previously published method (12). Preliminary experiments have shown that the quality of the MMA analysis, in terms of recovery, precision, and LOQ, would be unaffected or even improved by use of 150 µL of serum instead. With the smaller sample volume, the ion suppression almost disappears (signal, 91%–100% of the expected value).
Acknowledgments
We thank Marie Karoline Zanoni, medical laboratory technologist, for excellent technical assistance and invaluable suggestions in the method development.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  T. M. Annesley Methanol-Associated Matrix Effects in Electrospray Ionization Tandem Mass Spectrometry Clin. Chem., October 1, 2007; 53(10): 1827 - 1834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. la Marca, S. Malvagia, E. Pasquini, M. Innocenti, M. A. Donati, and E. Zammarchi Rapid 2nd-Tier Test for Measurement of 3-OH-Propionic and Methylmalonic Acids on Dried Blood Spots: Reducing the False-Positive Rate for Propionylcarnitine during Expanded Newborn Screening by Liquid Chromatography-Tandem Mass Spectrometry Clin. Chem., July 1, 2007; 53(7): 1364 - 1369. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
