| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Anaesthesia, National University of Singapore, 5 Lower Kent Ridge Road, Singapore 119074
aauthor for correspondence: e-mail anast{at}nus.edu.sg
Because the cerebral spinal fluid (CSF) bathes the whole central nervous system, its contents reflect changes occurring in the central neurons. Quantitative analysis of amino acids present in CSF has been helpful in improving our understanding of disease conditions associated with pain (1)(2)(3). Amino acids in CSF can be determined either by postcolumn derivatization with ninhydrin or o-phthalaldehyde (OPA) by commercially available amino acid analyzers or by precolumn derivatization with different reagents, such as dansyl chloride, phenylisothiocyanate, fluorenylmethyl chloroformate, dabsyl chloride, and OPA, followed by HPLC separation (4)(5)(6). Precolumn derivatization with OPA has been used predominantly for analysis of amino acids in CSF (1)(2)(7). However, OPA reacts only with primary amines, and the OPA adducts are unstable (4)(8). The preanalytical processes, including sample storage conditions and the pretreatment used for amino acid analysis in physiologic fluids, have not been standardized, making it difficult to compare results among laboratories (5)(9)(10)(11)(12). Pretreatment with different deproteination methods, including strong acids (3)(4)(9)(10), organic solvents (13)(14)(15), or ultrafiltration (2)(16)(17), has been shown to adversely affect the quantitative results for amino acids (16)(18).
Dabsyl chloride has been used to convert primary and secondary amines to their colored derivatives with subsequent separation by HPLC (19). This method has been used for analysis of amino acids and other compounds from physiologic fluid and tissue extracts (8)(15)(16)(20). However, use of this method for CSF samples has not been reported. We report an improved method using dabsyl chloride in the presence of the nonionic neutral surfactants Triton X-100 (Sigma) or Tween 20 (Bio-Rad), which allowed us to analyze amino acids and other physiologic compounds in CSF without any pretreatment.
Dabsyl derivatization was carried out in the dabsylation buffer consisting of 0.15 mol/L NaHCO3 buffer (pH 9.0) and 0.5 mL/L Triton X-100 adjusted to pH 9.0 with 1 mol/L NaOH. Dabsyl chloride (Pierce) was dissolved in acetone (12.35 mmol/L). The 23-compound calibration mixture (Std AA) used for this experiment was prepared by adding glutamine, asparagine, taurine, citrulline, and 
-aminobutyric acid (GABA) to a commercial 18-amino acid calibration mixture (Sigma) and diluted with the dabsylation buffer to achieve a final concentration of 1 nmol per 20 µL for each of the amino acids. Dabsylation buffer (20 µL) and dabsyl chloride reagent (40 µL) were added to 20 µL of Std AA or CSF sample (all CSF samples used for this study were obtained from patients with chronic knee osteoarthritis; informed consent from patients and Institutional Review Board approval were obtained) in 1-mL vials. The vials were capped and incubated at 65–70 °C for 20 min. The samples were cooled to room temperature, and the pH was adjusted to 6.5 with 0.33 mol/L phosphoric acid. To stabilize the dabsyl derivatives, we added 60 µL of the dilution buffer, consisting of a mixture of 50 mL of acetonitrile, 25 mL of ethanol, and 25 mL of mobile phase A used in the HPLC process (16), to the dabsylated samples. A Gynkotek HPLC system with a P 580A HPG pump, autosampler, and ultraviolet-visible photodiode array detector (UVD 340S) was used. The dabsyl derivatives of amino acids were separated on a LiChrosphere 100 RP-18 column [250 x 2 mm (i.d.); 5 µm particle size; Merck] packed by ChromatoResearch (Japan). The column was maintained at 50 °C, and 20 µL of the solution was loaded on the column. Mobile phase A consisted of 9 mmol/L lithium phosphate, 40 mL/L dimethylformamide, 3 g/L guanidine thiocyanate, and 2 g/L potassium perchlorate, with the pH adjusted to 6.5 with 0.33 mol/L phosphoric acid. Mobile phase B was 800 mL/L acetonitrile in water. The elution was performed at a flow rate of 200 µL/min using a gradient system (described in the caption for Fig. 1
), with absorbance monitored at 438 nm.
|
During the dabsylation experiments, the reaction mixtures frequently showed some precipitates or turbidity after dabsylation. We found that the surfactants Triton X-100 (0.25, 0.5, or 1 mL/L) or Tween (0.5 mL/L) when added to the dabsylation buffer cleared the reaction mixture and also increased the recovery ratios of the amino acids, especially the hydrophilic amino acids, such as aspartate, glutamate, asparagine, serine, threonine, GABA, and glycine (Table 1
). Triton X-100 (0.5 mL/L) gave the best separation results. We used nine 20-µL aliquots prepared from three CSF samples for recovery studies and other statistical analyses. For recovery studies, we added 2 nmol of glutamine (CSF contains glutamine in concentrations 10-fold higher than those of other amino acids) and 400 pmol of each of the other 22 amino acids to each of these nine CSF aliquots. These tubes were dabsylated individually and subsequently analyzed separately by HPLC. The results showed a mean recovery of 101.6% for all compounds (Table 1
).
|
The use of acidic pretreatment conditions for deproteination has been reported to cause an increase in glutamate and aspartate concentrations in CSF samples, as a result of hydrolysis of their corresponding amide forms (7)(10)(18)(21). In addition, the dabsyl derivatization yield is also significantly decreased (16). Krause et al.(16) recommended ultrafiltration for pretreatment, but when we used Amicon filters (Millipore) with molecular weight cutoffs of 3000 or 30 000, the results indicated that ultrafiltration decreased the derivatization yield by at least 20%. The loss could be attributable to interaction of some amino acids and certain high-molecular-weight components in the CSF, which were retained in the filter during ultrafiltration. Although addition of organic solvents such as methanol (13), ethanol (14), and acetonitrile(15) could precipitate high-molecular-weight proteins, the process could also remove some free amino acids mechanically during centrifugation.
The recovery of hydrophilic amino acids, especially glycine and GABA, which are known to be inhibitory neurotransmitters, was increased by 50–60%. Moreover, Triton also improved the detection of citrulline. The recovery ratio of citrulline in one CSF sample (CSF B) was doubled, and in another sample (CSF E), citrulline could be detected only when dabsylation was carried out in the presence of Triton (Table 1
). Because citrulline is a resulting metabolite when NO is produced from arginine, citrulline concentrations can be used as an indicator of NO activity. The ability to detect these amino acids is important because they have been shown to be involved in pain transmission mechanisms. It is known that many amino acids are present as free and bound forms in CSF (22)(23). However, Triton and Tween cannot release free amino acids from their conjugated or chemically bound forms, and they affect only the hydrophobic and/or hydrogen bonding interactions between free amino acids and certain compounds present in CSF. To date, no specific binding protein for amino acids has been identified in CSF. Therefore, the improved recovery of hydrophilic amino acids achieved with Triton and Tween suggests the presence of a specific binding protein in CSF. Further experiments are needed to confirm whether this is the case.
Chromatograms of the Std AA mixture and a CSF sample are shown in panels A and B of Fig. 1
. Although seldom present in CSF, cystine was also detected. Because previously reported chromatographic conditions (8)(15)(16) could not be applied directly to CSF analysis, we evaluated the optimum conditions by studying the effects of different salts, such as lithium, sodium, and potassium phosphate buffers; amines such as triethylamine (TEA), trimethylamine, morpholine and guanidine thiocyanate; and the chaotropic agent potassium perchlorate in different concentrations in mobile phase A. The use of 9 mmol/L lithium phosphate buffer in place of the commonly used sodium buffer not only improved the homogeneity of peak separation but also enhanced the sensitivity. There was a mean increase of 15% in the peak heights when lithium buffer was used. The use of TEA in mobile phase A, as suggested by Krause et al. (16), enabled separation of most of the compounds except two pairs, glutamine–citrulline and glycine–arginine. Substitution of TEA with 3 g/L guanidine thiocyanate in mobile phase A facilitated the separation of these two pairs of amino acids, but this particular mobile phase was unable to separate alanine and arginine clearly. This problem was solved by the addition of 2 g/L potassium perchlorate as a chaotropic agent.
The detection limits for all of the dabsyl derivatives, the data for regression analysis, and other statistics are listed in Table 2 in the Data Supplement that accompanies the online version of this Technical Brief at http://www. clinchem.org/content/vol50/issue2/. The mean within-day reproducibility was 7.4%, and the mean between-day reproducibility was 5.4%. The dabsylated samples in the dilution buffer were stable for up to 48 h when stored at room temperature protected from light and remained stable for up to 1 week when stored at 4 °C.
In conclusion, by adding a nonionic surfactant we were able to use precolumn derivatization of amino acids with dabsyl chloride to aid in the analysis of various amino acids and other compounds of interest in a small amount of CSF without the need for deproteination. Because our method is able to detect all of the amino acids, including the well-known NO-related compound citrulline, we believe that our method could be very useful for studying different types of pain conditions and other central nervous system diseases. Because the analysis of amino acids in CSF is not yet standardized, especially the sample collection, pretreatment, storage, and analysis steps (11)(12)(18), we hope that our report helps to improve and simplify the sample pretreatment protocols.
Acknowledgments
This research was supported by a grant from the National Medical Research Council of Singapore.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  R. Sethuraman, M. G. Krishnamoorthy, T.-L. Lee, E. H. C. Liu, S. Chiang, W. Nishimura, M. Sakai, T. Minami, and S. Tachibana Simultaneous Analysis of D- and L-Serine in Cerebrospinal Fluid by Use of HPLC Clin. Chem., August 1, 2007; 53(8): 1489 - 1494. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
