Determination of glutamine and serine in rat cerebrospinal fluid using capillary electrochromatography with a modified photopolymerized sol–gel monolithic column

https://doi.org/10.1016/S0021-9673(03)00451-5Get rights and content

Abstract

Capillary electrochromatographic separations of amino acid mixtures were studied using two modified porous photopolymerized sol–gel monolithic columns. One was modified with dimethyloctadecylchlorosilane (DMOS), and the other was modified with DMOS, followed by chlorotrimethylsilane to end-cap residual silanol groups. Prior to separation, amino acids were derivatized with 4-fluoro-7-nitro-2,1,3-benzoxadiazole using as a mobile phase 50 mM phosphate (pH 2.5), water, and acetonitrile in the ratio of 1:1:8. Five derivatized amino acids (Asn, Phe, Ala, Ile, and Leu) were separated within 7 min. Theoretical plate numbers varied between 58 700 and 105 000/m. This separation method with the end-capped monolithic column was applied to rat cerebrospinal fluid. The dominant amino acid found was Gln at a concentration of 420 μM along with small quantities of Ser (54 μM).

Introduction

Capillary electrochromatography (CEC), which is a hybrid separation technique of HPLC and CE [1], [2], enables the analysis of small volumes of sample with short run times because of the high separation efficiencies that can be achieved. A variety of analytes have been separated by CEC. They include biologically important molecules such as proteins, peptides, and amino acids. CEC is suitable for the analysis of real samples whose limitations include small sample volumes, high sample complexity, and low analyte concentrations. But to date, only a few analyses of real samples by CEC have been reported [3], [4], [5].

Because capillaries have a very narrow diameter (typically less than 100 μm), elaborate work including the fabrication of on-column frits is required to prepare a stationary phase with good reproducibility and homogeneity. Bubble formation or deterioration of the separation efficiency has been observed in frit sections [6], [7], [8]. To overcome these difficulties and problems of packed CEC column preparation, monolithic CEC columns were first introduced by Hjertén and co-workers [9] and developed further by Svec and Fréchet [10]. Monolithic CEC columns are easily prepared and show good chromatographic performances [11]. Many monolithic columns for CEC have been fabricated using a variety of monomers, such as acrylamide [12], [13], [14], [15], [16], [17], [18], methacrylate [19], [20], [21], [22], and alkoxysilane [23], [24], [25], [26], [27], [28], [29].

Recently at Stanford University, we prepared monolithic columns from methacryloxypropyltrimethoxysilane (MPTMS) [30], [31]. MPTMS, which contains both methacrylate and alkoxysilane groups, was used to create a photopolymerized sol–gel (PSG) monolith in a single-step reaction. The PSG monolith showed good separation of some neutral compounds, such as alkylbenzenes and polycyclic aromatic hydrocarbons. The separation of charged compounds, however, was not satisfactory when using this PSG monolithic column because charged compounds were adsorbed onto the residual silanol groups of the PSG monolith surface. Consequently, we modified the PSG monolith with silane-coupling reagents. The modified PSG was effective for the separation of both charged and uncharged compounds [32].

In this paper, dimethyloctadecylchlorosilane (DMOS), which is a typical coupling reagent to prepare ODS packing materials for HPLC, was chosen for use as a silane-coupling reagent. An end-capping reaction with chlorotrimethylsilane (CTMS) was also tried after the modification of PSG monolith with DMOS. The end-capping reaction is often carried out for many ODS packings that are used in HPLC columns to prevent adsorption of analytes by the residual silanol groups. The separation properties of an octadecyl PSG monolithic column (non-end-capped) and an end-capped PSG monolithic column were compared by CEC.

Amino acids were chosen as charged samples, because these are very important components in many biological and food samples. Rat cerebrospinal fluid (CSF) is an example of a real biological sample. Techniques such as HPLC, CE, and CE/microdialysis have been used for the analysis of amino acids in rat CSF [33], [34], [35]. In fact, the measurement of the Gln concentration in CSF has been used to diagnose encephalopathies of hepatic origin [35]. All amino acids are derivatized with a fluorogenic reagent, 4-fluoro-7-nitro-2,1,3-benzoxadiazole (NBD-F), to obtain highly sensitive detection. The separation properties of derivatized amino acids on these monolithic columns were examined and the modified PSG monolithic column was applied to the determination of amino acids in rat CSF.

Section snippets

Materials and chemicals

UV-transparent coated fused-silica capillaries (75 μm I.D.×375 μm O.D.) were purchased from Polymicro Technologies (Phoenix, AZ, USA). MPTMS, DMOS, NBD-F, and CTMS were purchased from Tokyo Kasei (Tokyo, Japan). Toluene, acetonitrile, thiourea, ammonium acetate, and trifluoroacetic acid were from Kanto Kagaku (Tokyo, Japan). Amino acids were purchased from Sigma–Aldrich (Milwaukee, WI, USA). Diethylamine, phosphoric acid, sodium dihydrogenphosphate dehydrate, and boric acid were from Wako

Results and discussion

In the analysis of charged analytes by CEC, two major driving forces work to cause migration. One is electroosmotic flow (EOF) and the other is the electrophoretic mobility of the analyte. In a fused-silica capillary, EOF is generated in neutral and basic conditions, but is negligible below pH 3. We examined elution time of thiourea, which was often used as an EOF marker [30], [31]. Thiourea was eluted on both the non-end-capped and the end-capped monolithic column when using a mixture of

Conclusion

A non-end-capped and an end-capped ODS-modified PSG monolithic column were prepared for the separation of NBD-amino acids. The end-capped PSG monolithic column was shown to be superior to the non-end-capped monolithic column for the separation of NBD-amino acids. CEC with LIF detection and an end-capped monolithic column allowed for simpler sample pretreatment of low volumes of rat CSF. Rapid separation of the amino acids in CSF was possible.

Acknowledgments

The work at the University of Shizuoka has been supported by a grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and that at Stanford University by Beckman-Coulter, Inc. We also gratefully thank Ciba for donating the Irgacure 1800, Professor Masayuki Sato for the loan of a photochemical reactor and Dr Takashi Okura for donation of rat CSF.

References (39)

  • V. Pretorius et al.

    J. Chromatogr.

    (1974)
  • D.A. Stead et al.

    J. Chromatogr. A

    (1998)
  • S. Fanali et al.

    J. Chromatogr. A

    (2001)
  • B. Behnke et al.

    J. Chromatogr. A

    (1995)
  • H. Zou et al.

    J. Chromatogr. A

    (2002)
  • C. Ericson et al.

    J. Chromatogr. A

    (1997)
  • H. Minakuchi et al.

    J. Chromatogr. A

    (1997)
  • H. Minakuchi et al.

    J. Chromatogr. A

    (1998)
  • N. Ishizuka et al.

    J. Chromatogr. A

    (1998)
  • M. Kato et al.

    J. Chromatogr. A

    (2002)
  • D.J. Begley et al.

    J. Chromatogr. B

    (1994)
  • G. Alfredsson et al.

    J. Chromatogr.

    (1983)
  • S. Tucci et al.

    Clin. Biochem.

    (1998)
  • M. Kato et al.

    J. Chromatogr. A

    (2001)
  • H. Yamamoto et al.

    J. Chromatogr.

    (1992)
  • J.H. Knox et al.

    Chromatographia

    (1991)
  • S. Thiam et al.

    Anal. Chem.

    (2000)
  • B. Behnke et al.

    Electrophoresis

    (2000)
  • H. Rebscher et al.

    Chromatographia

    (1994)
  • Cited by (0)

    View full text