Capillary electrophoresis with contactless conductometric detection for rapid screening of formate in blood serum after methanol intoxication

Journal of Chromatography A, 1281 (2013) 142–147

Contents lists available at SciVerse ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

Capillary electrophoresis with contactless conductometric detection for rapid screening of formate in blood serum after methanol intoxication ˇ a,∗ , Frantiˇsek Foret a , Robert Bocek b Petr Kubán a b

Bioanalytical Instrumentation, CEITEC MU, Veveri 97, 602 00 Brno, Czech Republic Department of Anaesthesiology, Resuscitation and Intensive Care, Hospital and Policlinics Havíˇrov Dˇelnická 1132/24, 73601 Havíˇrov, Czech Republic

a r t i c l e

i n f o

Article history: Received 26 November 2012 Received in revised form 4 January 2013 Accepted 7 January 2013 Available online 15 January 2013 Keywords: Capillary electrophoresis Contactless conductivity detection Formate Methanol intoxication

a b s t r a c t A new method for rapid, direct determination of formate in blood serum samples by capillary electrophoresis with contactless conductometric detection is presented. A selective separation of formate was achieved in approximately 1 min using an electrolyte system comprising 10 mM l-histidine, 15 mM glutamic acid and 30 ␮M cetyltrimethylammonium bromide at pH 4.56. The only sample preparation was dilution (1:100) with deionized water. The limit of detection and limit of quantitation was 2.2 ␮M and 7.3 ␮M, respectively, which corresponds to 0.22 mM and 0.73 mM in undiluted blood serum. The method provides a simple and rapid diagnostic test in suspected methanol intoxication cases. The method has been successfully tested on determination of formate in blood of a patient admitted to the hospital under acute methanol intoxication. The peak concentration of formate detected in the patient blood serum was 12.4 mM, which is 10- to 100-fold higher than the normal values in healthy population. The developed method presents the fastest test currently available to detect formate in blood samples. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Intoxication by industrial chemicals such as methanol, formaldehyde or ethylene glycol is occurring regularly worldwide. Sometimes, the scale of such a poisoning can be similar to common acts of terrorism and have thus a significant impact on society. For instance, in documented cases of methanol poisoning in Kenya 2000 [1], Tunisia 2003 [2], India 2009 [3], and even in Europe: Estonia 2001 [4], Norway 2002 [5] and most recently the Czech Republic (September 2012), the number of fatalities was comparable to Matsumoto city incident 1994 [6], Tokio Sarin attack 1995 [7] or Moscow theater hostage crisis 2002, in all of which, highly toxic nerve gases have been used. Rapid analysis and identification of suspected toxic chemical compounds in body fluids during acute intoxication is important, especially with respect to the administration of an efficient antidote and initiation of a subsequent medical treatment. Often, the toxicity of the metabolites may be higher than the toxicity of the originally ingested compound. This is for instance true for methanol poisoning: the toxicity of methanol is much lower than the toxicity of its metabolic products, formaldehyde and formate. Initial symptoms of methanol intoxication are difficult to diagnose and are generally non-specific. At a later stage of intoxication, the concentration

∗ Corresponding author. Tel.: +420 532290201; fax: +420 541212113. ˇ E-mail address: [email protected] (P. Kubán). 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2013.01.035

of methanol in blood may already be significantly lowered, as the major part is converted to formate. This occurs typically between 12 and 24 h after methanol ingestion. The symptoms specific to elevated levels of formate include abdominal discomfort, nausea, headache, dizziness, blurring or complete loss of vision, large anion gap and deep metabolic acidosis leading to hyperventilation [8]. If untreated, the high concentration of formate can cause serious damage to the optical nerve [9,10], respiratory failure [11], renal failure [12], coma, cerebral edema, seizures and eventual death from cardiorespiratory arrest [13]. In healthy individuals, typical concentration levels of formate in blood range from 0.07 to 1.2 mM [14,15]. Formate concentration can be significantly elevated during acute methanol intoxication. From the available literature data, the concentration of formate in identified methanol poisoning cases was found to be between 0.2 and 32 mM [15–17]. Blood formate concentrations above 10 mM (0.5 g/L) have been associated with severe toxicity due to methanol ingestion, permanent tissue damage, or fatality [18,19]. Therefore, the determination of formate rather than methanol should be used as a marker for acute methanol intoxication. A simple, but fast, screening method for formate able to quantify its concentrations above the maximum normal levels (1.2 mM) would be of great importance in clinical practice. Further, monitoring of levels of formate in blood with high temporal resolution during subsequent medical treatment (hemodialysis, antidote administration [20]) is also important as it may be used to identify the time point when patient’s condition has been stabilized.

P. Kubánˇ et al. / J. Chromatogr. A 1281 (2013) 142–147

143

Currently, the methods for formate determination in blood samples used in clinical laboratories are either time consuming, require difficult sample preparation/derivatization and/or may not be sufficiently sensitive. Typically, headspace GC-FID [21–26] or GC–MS [27–29] is used after conversion of formate into its volatile esters. The described procedures often involve several pretreatment steps and take between 30 and 60 min. HPLC has also been used for formate analysis [30] but there is no significant time saving during sample preparation compared to GC. Enzymatic essays based on conversion of formate to carbon dioxide by a specific enzyme and subsequent spectrophotometric detection of the reaction byproduct have been used as well [31–34]. Unfortunately, none of the conventional methods offers the required simplicity and rapidity needed for fast screening purposes. Capillary electrophoresis (CE) is becoming popular in analysis of samples of biological origin, mostly because these samples are often unavailable in large quantities and have rather complicated matrix composition. Another major advantage of CE is the analysis speed that could be important in the fast screening, such as the one for formate. The analysis of biological samples by CE has been extensive, as documented by several review articles [35–37] and the variety of analyzed biological fluids is not limited to blood samples but includes also cerebrospinal fluid, tear fluid, saliva and urine. The biological samples are predominantly analyzed for their protein and biomolecular content, however, analysis of small organic molecules, including organic acids is becoming also important [38]. The analysis of formate by CE has however been rare [39,40] and its determination has not been associated with methanol intoxication. The introduction of capacitively coupled contactless conductivity detection (C4D) [41,42] has opened another dimension for the analysis of low molecular weight compounds because their relatively high equivalent conductance provides a sensitive response in C4D. It is thus not surprising that the analysis of these compounds in biological samples by CE-C4D has recently obtained significant attention. The samples analyzed include saliva [43], blood plasma [44–49] and urine [50–52]. None of the presented methods described the analysis of formate, except a recent study by ˚ Tuma et al. [52]. In their work urine has been analyzed for a variety of organic acids in metabolic disorder screening, including formate. Unfortunately, formate has not been separated from fumarate and tartrate as it has not been the target analyte. In this work we present a new method and background electrolyte (BGE) system in which fast and selective separation of formate is achieved in blood serum samples. The method is the fastest available, selective screening for formate, which can readily be applied for analysis of formate during the suspected methanol intoxication. The method was applied for the screening of formate during the recent methanol intoxication cases in the Czech Republic in September 2012.

flushed with BGE solution for 1 min. At the end of a working day, the capillaries were washed with DI water for 10 min, followed by applying a vacuum for 5 min to remove any liquid from inside and stored dry overnight. All CE experiments were performed at ambient temperature.

2. Experimental

3. Results and discussion

2.1. Material and methods

3.1. Optimization of the electrolyte system

2.1.1. Electrophoretic system A purpose-built CE instrument was employed for all electrophoretic separations. The separation voltage of −18 kV was provided by a high voltage power supply unit (Spellman CZE2000R, Spellman, Pulborough, UK). The separation capillaries used were fused silica (FS) capillaries (50 ␮m I.D., 375 ␮m O.D., 40 cm total length, 20 cm effective length, Polymicro Technologies, Phoenix, AZ, USA). Prior to the first use, the separation capillary was preconditioned by flushing with 0.1 M NaOH for 30 min, deionized (DI) water for 10 min and background electrolyte (BGE) solution for 10 min. Between two successive injections, the capillary was

Low molecular weight organic acids were previously separated using a MES/HIS electrolyte with pH 6 by Law et al. [54]. At this pH, comigration of several acids exists due to their similar electrophoretic mobilities. To separate the comigrating acids, MES/HIS electrolyte was modified with 0.025% HP-␤-CD and 10% methanol. 16 organic acids were separated in 14 min, but the separation sys˚ tem was not applied to the analysis of biological samples. Tuma et al. [52] have separated 29 organic acids in urine, however the critical analytes for this study (formate, fumarate and tartarate) were not resolved. In here, we have optimized the background electrolyte (BGE) composition to achieve fastest possible separation of

2.1.2. Injection Injection of standard solutions and blood serum samples was carried out hydrodynamically. The injection capillary end was immersed in a sample vial and elevated to a height of 10 cm for 20 s. 2.1.3. Detection system A C4D was used for the detection of the separated analytes. It consisted of an external function generator (GW Instek GFG-8219A, New Taipei City, Taiwan) providing a sinusoidal excitation signal (frequency: 290 kHz, amplitude: 20 V peak-to-peak) to an in-house built detector cell [53] with a pre-amplifier (OPA655, Burr Brown, TX, USA). The amplified cell current was led to an external detector circuitry for further processing. Data were collected using Panther 1000 AD convertor. 2.2. Chemicals 2.2.1. Reagents, standards, electrolytes All chemicals were of reagent grade and DI water (Purite, Neptune, Watrex, Prague, CR) was used for stock solution preparation and dilutions. 10 mM stock solutions of inorganic anions were prepared from their sodium salts (chloride, nitrate, sulfate, phosphate all from Pliva-Lachema, Brno, Czech Republic). Organic acids (formic, fumaric, tartaric, malonic, maleic, malic, succinic, acetic) were prepared from reagent grade chemicals (Sigma–Aldrich, Steinheim, Germany, Pliva-Lachema, Brno, Czech Republic). Lithium lactate was from Pliva-Lachema, Brno, Czech Republic. BGE for CE measurements was prepared daily by diluting 100 mM stock solutions of l-histidine (HIS, Sigma–Aldrich) and 2-(N-morpholino)ethanesulfonic acid (MES, Sigma–Aldrich) or 50 mM glutamic acid (GLU, Sigma–Aldrich) to the required concentration. Cetyltrimethylammonium bromide (CTAB, Sigma–Aldrich) was prepared as 10 mM stock solution in 5% acetonitrile and was added to the BGE to yield the final concentration of 30 ␮M. 2.2.2. Sample preparation Blood serum samples were obtained from Department of Anaesthesiology, Resuscitation and Intensive Care, Hospital and Policlinics Havíˇrov. Before injection into the CE instrument, the serum samples were diluted 1:100 with DI water. No other treatment was necessary. The lyophilized serum samples were purchased as lyophilized powders from Sigma and prepared according to supplier’s instructions. All samples were stored at −20 ◦ C.

144

P. Kubánˇ et al. / J. Chromatogr. A 1281 (2013) 142–147

Table 1 Ionic mobilities and pKa values of organic acids in the study.

2.8

pKa1

pKa2 /pKa3

Formic Fumaric Tartaric Malonic Maleic Malic Citric Phosphoric Succinic Lactic Acetic

56 31, 61.2 32.6, 60.7 40.7, 67 41.3, 62.4 32.6, 59 28.7, 54.7, 74.7 35.1, 61.5, 71.5 33, 60.9 36.5 42.4

3.75 3.02 3.04 2.85 1.92 3.46 3.13 2.12 4.20 3.86 4.76

– 4.38 4.37 5.69 6.23 5.05 4.76/6.39 7.47/12.36 5.64 – –

formate from the other possible interferents in blood serum samples using a simple electrolyte system. CTAB was added to the BGE solution to reverse the electroosmotic flow and to modify the capillary wall to avoid the adsorption of proteins from the serum sample. The major task of this study was the optimization of the separation of formate from the rest of possible interfering organic acids without the necessity for these acids to be fully resolved from each other. The separation selectivity can be modified by adjusting the pH of the BGE, because formate is rather strong acid (pKa = 3.75) and has the smallest molecule, compared to the other possible comigrating species. For comparison, Table 1 lists a compilation of pKa s and ionic mobilities of the possible interfering organic acids [55]. By decreasing the pH of the BGE, most of the acids will be slowed down and may be separated from formate. Initially, we have modified the composition of MES/HIS BGE by increasing the concentration of MES from 20 mM to 60 mM. The pH of the BGE changes from 6 to pH 5.8 at 40 mM MES and to pH 5.6 at 60 mM MES concentrations, respectively. However, as the pH decreases only slightly, a full baseline separation of formate/fumarate/tartrate was not achieved. Moreover, using 60/20 mM MES/HIS BGE resulted in large currents in the separation capillary and drifting baseline. An alternative electrolyte, where MES was replaced by glutamic acid (GLU) was tested. Glutamic acid has significantly lower pKa s (GLU, pKa1 = 2.19, pKa2 = 4.25) than MES (pKa = 6.1 [56]) and could be used to decrease the pH of the BGE more significantly than MES. The concentration of HIS was decreased to 10 mM and GLU was added at 5, 10, 15 and 20 mM concentrations. The corresponding pH of the BGEs was: 6.09, 5.19, 4.56, 4.29. Fig. 1 shows the dependence of migration times of 14 analytes (inorganic anions and organic acids) on the concentration of GLU added to the 10 mM HIS. Three inorganic anions that are present in blood serum (chloride, nitrate and sulfate) were added to the model mixture. The migration velocity and separation selectivity of most analytes changes significantly with decrease of pH. At GLU concentrations of 15 and 20 mM full separation of formate from all interfering acids can be achieved. The baseline separation of formate from other interferents is equally as important as is its detection sensitivity. Fig. 2 shows a plot of peak areas, peak height of formate and resolution (formate/fumarate) vs. increasing GLU concentration in the studied range. Additionally the insert of the figure contains a detailed set of electropherograms showing an increase in resolution between formate and fumarate peaks. As the sensitivity and resolution does not change significantly from 15 to 20 mM GLU, the separation electrolyte composed of 10 mM HIS, 15 mM GLU and 30 ␮M CTAB was chosen and used in all subsequent experiments. Using the optimized BGE, a separation of model mixture of inorganic anions and organic acids is shown in Fig. 3. Note that there is no peak of tartrate. In this electrolyte, tartaric acid gave a very broad and dispersed peak only at the concentration above 500 ␮M. The reason for this specific behavior is not currently understood and would require a more detailed study.

chloride nitrate sulfate

2.4

migration time (min)

Ionic mobility [cm /Vs × 10 ] 5

formate fumarate

maleate malate citrate phosphate

malonate

lactate succinate

tartarate

acetate

2

1.6

1.2

0.8

0.4 0

5

10

15

20

25

concentration of GLU in 10 mM HIS buffer (mM) Fig. 1. Dependence of migration times of selected analytes on the concentration of glutamic acid in the separation electrolyte composed of 10 mM l-histidine and 30 ␮M CTAB. Separation voltage: −18 kV. Hydrodynamic injection: 10 cm/20 s. C4D detection.

3.2. Analytical parameters of the method The analytical parameters of the developed CE method were investigated. The injection was hydrodynamic from 10 cm for 20 s. The calibration graphs were constructed by preparing the calibration solutions of formate in DI water (calibration set 1), and also by spiking the concentrated stock solution of formate into diluted blood serum sample (1:100, v/v) in which the natural concentration of formate was below the limit of detection (calibration set 2). Table 2 lists the corresponding calibration parameters for

100

5 20 mM 15 mM 10 mM

4

80 5mM

PA,PH ( mV.s, mV)

Acid

2

0.9 1 1.1 1.2 1.3 1.4

3

time (min)

60

Rs 2

40 1

20

0 0

10 20 concetration of GLU (mM)

30

Fig. 2. The dependence of peak area (䊉), peak height () of formate and resolution between formate/fumarate peaks () on the concentration of glutamic acid. The insert in the figure shows the actual separation of formate and fumarate peaks at 100 ␮M concentration. CE conditions: Separation electrolyte: 10 mM HIS, 30 ␮M CTAB and varying concentrations of GLU. Separation voltage: −18 kV. Hydrodynamic injection: 10 cm/20 s. Anion concentrations: 100 ␮M. C4D detection.

P. Kubánˇ et al. / J. Chromatogr. A 1281 (2013) 142–147

10 mV

nitrate

formate

145

lactate

A

formate

maleate/malonate

B detector signal (mV)

fumarate phosphate lactate succinate

chloride

20 mV

0

detector signal (mV)

malate/citrate

sulfate

acetate

0.5

1

1.5

2

2.5

time (min) Fig. 3. Separation of model mixture of 13 anions and organic acids. CE conditions: separation electrolyte: 10 mM HIS/15 mM GLU, 30 ␮M CTAB, pH 4.68. Separation voltage: −18 kV. Hydrodynamic injection: 10 cm/20 s. Anion concentrations: 50–200 ␮M. C4D detection.

calibration set 1 and 2 in the concentration range between 5 and 200 ␮M with the R2 varying from 0.9976 to 0.9985. Note that although there was no formate detected in the original diluted blood serum, and the calibration curves have identical slope, the intercept in calibration set 2 is slightly higher. The two calibration sets differ by approximately 3% and for exact quantitation the calibration set 2 should be used as it more realistically reflects the composition of the analyzed samples. The limit of detection, limit of quantitation and repeatability of migration times and peak areas of formate for repeated injection of diluted blood serum sample (n = 10) are also shown in Table 2. The LOD for formate is 2.2 ␮M, which corresponds to 0.22 mM of formate in undiluted blood serum sample. The LOD given here was calculated from the calibration curve with hydrodynamic injection at 10 cm for 20 s, with the criterion of 3 S/N (peak to peak). The LOQ of 7.3 ␮M is slightly below the upper concentration limit of formate concentration in normal blood serum samples [15]. Therefore, if the concentration of formate falls within the normal range, there should not be any peak or the peak should be only very small. However, if the normal range is significantly exceeded, such as in the case of methanol intoxication, the peak of formate will be clearly detectable. To increase the sensitivity the injection time could be increased up to 60 s without deterioration of the separation efficiency or the dilution factor of the blood serum could be decreased. We have analyzed a wide range of blood serum samples which were Table 2 Analytical parameters of the developed CE-C4D method for determination of formate in blood serum. Calibration set

Range (␮M)

Equation

R2

Set 1 Set 2

5–200 5–200

y = 0.0093x + 0.0137 y = 0.0093x + 0.0305

0.9976 0.9985

Analyte

LOD (␮M)

LOQ (␮M)

RSD, tM

RSD, PA

Formate

2.2

7.3

0.78%

3.41%

tM, migration time; PA, peak areas; n = 10.

C D

0

0.5

1

1.5

2

2.5

time (min) Fig. 4. The electropherograms of diluted blood serum samples (1:100 v/v with DI water). Normal blood serum of a healthy volunteer (trace A), lyophilized blood serum sample (trace B), blood serum sample of the patient with methanol intoxication after 25 h of hemodialysis (trace C), blood serum sample of the same patient 1 week after hospitalization (trace D). CE conditions as in Fig. 3.

both normal human serum with no intoxication, lyophilized serum sample (Sigma–Aldrich) and blood serum sample of a patient subjected to hemodialysis after methanol intoxication. In all of the analyzed samples the concentration of formate was within the physiological range. Fig. 4. shows a zoomed series of electropherograms of these serum samples. In all samples, except the lyophilized serum (trace B) no peak of formate was detected. The concentration of formate in the lyophilized serum sample was estimated to be 1.3 ␮M. 3.3. Robustness toward repeated injection of serum samples The injection of various blood samples (whole blood, serum, plasma) often results in capillary fouling and change in EOF and corresponding migration times. The changes are attributed mainly to the adsorption of proteins from the serum to the capillary wall. The problem becomes apparent when analyzing positively charged compounds from the serum in an uncoated fused silica capillary, as the major serum protein, HSA (pI values given in the literature are between 4.7 and 5.3) [57–59] becomes positively charged and attaches to the negatively charged wall of the fused silica capillary. In our previous work [46], this phenomenon was observed and even if the serum sample was diluted 1:80, significant change in migration times was observed for serum and whole blood samples (about 20% increase over 15 consecutive runs). There are various pretreatment methods that can be used for real blood or serum samples, for instance deproteinization with acetonitrile [50], use of electrodialysis [46] or electromembrane extraction [45]. However all of these sample treatments introduce additional complexity and increase the analysis time. We have also demonstrated several years ago that a suitable selection of BGE with dynamic modification of capillary surface using 7 ␮M CTAB and electrokinetic injection can to some extent eliminate protein adsorption in CE of cationic analytes even at low pH of the BGE [60]. In this work, a higher concentration of CTAB (30 ␮M) was used compared to our previous work. The EOF was reversed by the attachment of CTAB to the capillary wall

146

P. Kubánˇ et al. / J. Chromatogr. A 1281 (2013) 142–147

2

100 mV formate lactate

1.6

detector signal (mV)

migration time (min)

E lactate

1.2 formate nitrate 0.8

D C formate

lactate

B

chloride formate lactate

A

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

run # 0 Fig. 5. Repeatability of migration times of repeatedly injected serum sample 1:10 (open symbols) and 1:100 (full symbols). Chloride (, ♦), Nitrate (), formate(䊉, ) lactate (, ).

resulting in positively charged coating and repulsion of the proteins from the blood serum samples. This is demonstrated in Fig. 5 that shows repeatability of migration times of the detected anions in the blood serum sample. Two dilution factors were tested, 1:10 and 1:100. The diluted serum sample was injected repeatedly 20 times onto the separation capillary. While with 1:10 dilution, there is some small, observable increase in migration times, especially for the later migrating anions, there is no detectable deterioration of performance for the 1:100 dilution. This is apparent from the slopes of the data in Fig. 5. In fact one and the same capillary was used for the whole set of experiments presented in this article without deterioration of the performance, when serum samples diluted 1:100 were injected. The total number of injections of diluted serum samples was estimated to be more than 100. It can be thus safely stated that the current BGE composition and CE separation conditions are suitable for analysis of large number of serum samples, provided the appropriate dilution factor 1:100 is maintained. 3.4. Analysis of blood serum samples The developed CE method was applied for detection of formate in real blood serum sample of a patient admitted to the Department of Anaesthesiology, Resuscitation and Intensive Care, Hospital and Policlinics in Havíˇrov with serious methanol intoxication. Fig. 6 shows an analysis of diluted blood serum sample (1:100) from the patient in the beginning phase of hemodialytical treatment (trace A), and samples taken at time interval of 1 h (trace B), 25 h (trace C) and 1 week (trace D) after the first sample. For comparison, a model solution with chloride concentration of 1 mM and other ions at 100 ␮M is shown in trace E. In traces A and B, significantly elevated concentration of formate was detected. The concentration was evaluated using the calibration curve set 2 and the quantitative data from all samples are summarized in Table 3. The concentration of formate in trace A was 12.4 mM which is more than 10 times higher than the maximum physiological concentration of formate in normal serum. In trace B, which was taken 1 h after the sample in trace A, the concentration of formate decreased to 9.4 mM demonstrating the successful progress of the hemodialytical removal of

1

2

3

time (min) Fig. 6. CE analysis of diluted serum samples and a standard solution. Diluted serum sample of a patient in the beginning phase of hemodialytical treatment (trace A), 1 h after the after the first sample (trace B), 25 h thereafter (trace C) and 1 week after (trace D). Trace E is a model solution containing 1 mM chloride and 0.1 mM concentrations of other analytes. CE conditions as in Fig. 3. Table 3 Quantitation of formate in blood serum. Sample

Formate concentration

#1, 25.9.2012, 11:00 AM #2, 25.9.2012, 12:00 AM #3, 26.9.2012, 8:50 AM #4, 01.10.2012, 6:30 AM #5, Lyophilized serum

12.4 mM 9.4 mM ND ND 1.3 ␮M

this toxic metabolite. After 25 h there was no detectable concentration of formate. 4. Conclusions A simple and rapid method for determination of formate in blood serum was presented. This method allows fast (<2 min) screening of patient blood serum for formate that is a metabolic product during methanol intoxication. The only sample preparation is dilution with DI water. This newly developed method provides a very fast response, which is important in case of acute intoxication as immediate treatment and antidote administration can be initiated. Further, a high temporal resolution of the method allows simple monitoring of the levels of formate in blood during hemodialytical treatment and can significantly save on the cost of the dialysis by identifying the time point at which the treatment is no more necessary as the toxic metabolite is completely removed from the blood stream. The method may find wide application in the clinical and toxicological practice. Acknowledgments The work was realized in CEITEC – Central European Institute of Technology with research infrastructure supported by the project CZ.1.05/1.1.00/02.0068 financed from European Regional Development Fund.

P. Kubánˇ et al. / J. Chromatogr. A 1281 (2013) 142–147

References [1] K. Ahmad, Lancet 356 (2000) 1911. [2] N. Brahmi, Y. Blel, N. Abidi, N. Kouraichi, H. Thabet, A. Hedhili, M. Amamou, Clin. Toxicol. 45 (2007) 717–720. [3] S. Shah, V. Pandey, N. Thakore, I. Mehta, J. Assoc. Physicians India 60 (2012) 34–36. [4] R. R Paasma, K.E. KE Hovda, A. Tikkerberi, D. Jacobsen, Clin. Toxicol. 45 (2007) 152–157. [5] K.E. Hovda, O.H. Hunderi, A.B. Tafjord, O. Dunlop, N. Rudberg, D. Jacobsen, J. Intern. Med. 258 (2005) 181–190. [6] T. Nakajima, S. Sato, H. Morita, N. Yanagisawa, Occup. Environ. Med. 54 (1997) 697–701. [7] H. Murakami, Underground: The Tokyo Gas Attack and the Japanese Psyche, Vintage International, New York, USA, 2000. [8] J.A. Kruse, Intensive Care Med. 18 (1992) 391–397. [9] G. Martin-Amat, K.E. McMartin, M.S. Hayreth, T.R. Tephly, Toxicol. Appl. Pharmacol. 45 (1978) 201–205. [10] N.J. Reddy, M. Sudini, L.D. Lewis, Clin. Toxicol. 48 (2010) 967–973. [11] R.B. Naik, W.P. Stephens, D.J. Wilson, A. Walker, H.A. Lee, J. Postgrad. Med. 56 (1980) 451–456. [12] D. Verhelst, P. Moulin, V. Haufroid, X. Wittebole, M. Jadoul, P. Hantson, Int. J. Toxicol. 23 (2004) 267–273. [13] W.R. Henderson, J. Brubacher, Can. J. Emerg. Med. 4 (2001) 34–40. [14] K.E. Hovda, K.S. Andersson, P. Urdal, D. Jacobsen, Clin. Toxicol. 43 (2005) 221–227. [15] F. Westphal, G. Rochholz, S. Ritz-Timme, N. Bilzer, H.W. Schutz, H.-J. Kattsch, Int. J. Legal Med. 114 (2001) 181–185. [16] J. Viinamaki, I. Rasanen, E. Vuori, I. Ojanpera, Forensic Sci. Int. 208 (2011) 42–46. [17] L.A. Ferrari, M.G. Arado, C.A. Nardo, L. Giannuzzi, Forensic Sci. Int. 133 (2003) 152–158. [18] P. Mahieu, A. Hassoun, R. Lauwerys, Hum. Toxicol. 8 (1990) 936– 938. [19] E. Tanaka, K. Honda, H. Horiguchi, S. Misawa, J. Forensic Sci. 36 (1991) 936–938. [20] J. Brent, K. McMartin, S. Phillips, C. Aaron, K. Kulig, N. Engl. J. Med. 344 (2001) 424–429. [21] C. Abolin, J.D. McRae, T.N. Tozer, S. Takki, Biochem. Med. 23 (1980) 209–218. [22] A.D. Fraser, W. MacNeil, J. Anal. Toxicol. 13 (1989) 73–76. [23] G.R. Jones, P.P. Singer, K. Rittenbach, J. Forensic Sci. 52 (2007) 1376–1382. [24] A. Sokoro, D. Lehotay, J. Eichhorst, R. Treble, J. Anal. Toxicol. 31 (2007) 342–346. [25] H.R. Wallage, J.H. Watterson, J. Anal. Toxicol. 32 (2008) 241–247. [26] X.-P. Lee, T. Kumazawa, K. Kondo, K. Sato, O. Suzuki, J. Chromatogr. B 734 (1999) 155–162. [27] S. Kage, K. Kudo, H. Ikeda, N. Ikeda, J. Chromatogr. B 805 (2004) 113–117. [28] I. Rasanen, J. Viinamaki, E. Vuori, I. Ojanpera, J. Anal. Toxicol. 34 (2010) 113–121.

147

[29] J.K. Kim, T. Shiraishi, E.-I. Fukusaki, A. Kobayashi, J. Chromatogr. A 986 (2003) 313–317. [30] S. Ohmori, I. Sumii, Y. Toyonaga, K. Nakata, M. Kawase, J. Chromatogr. 426 (1988) 15–24. [31] J.R. Quayle, Methods Enzymol. 9 (1966) 360–364. [32] S. Grady, J. Osterloh, J. Anal. Toxicol. 10 (1986) 1–5. [33] J.E. Buttery, B. Chamberlain, J. Anal. Toxicol. 12 (1988) 292–294. [34] M. Ogata, T. Iwamoto, Int Arch. Occup. Environ. Health 62 (1990) 227–232. [35] B. Baena, A. Cifuentes, C. Barbas, Electrophoresis 26 (2005) 2622–2636. [36] C. Klampfl, Electrophoresis 28 (2007) 3362–3378. [37] V. Galli, A. Garcia, L. Saavedra, C. Barbas, Electrophoresis 24 (2003) 1951–1981. [38] A.R. Timerbaev, J. Sep. Sci. 31 (2008) 2012–2021. [39] M. Shirao, R. Furuta, S. Suzuki, H. Nakazawa, S. Fujita, T. Maruyama, J. Chromatogr. A 680 (1994) 247–251. [40] J. Romano, P. Jandik, W.R. Jones, P.E. Jackson, J. Chromatogr. 546 (1991) 411–421. [41] A. Zemann, E. Schnell, D. Volgger, G.K. Bonn, Anal. Chem. 70 (1998) 563–567. [42] J.A. Fracassi da Silva, C.L. do Lago, Anal. Chem. 70 (1999) 4339–4343. [43] L. Vitali, V.T. Favere, G.A. Micke, J. Chromatogr. A 1218 (2011) 2327–2333. ˇ P.P. Boˇcek, Electrophoresis 32 (2011) 1182–1189. [44] L. Strieglerová, P. Kubán, ˇ P. Boˇcek, J. Chromatogr. A 1218 (2011) 6248–6255. [45] L. Strieglerová, P. Kubán, ˇ P. Kubán, ˇ I.K. Kiplagat, P. Boˇcek, Electrophoresis 32 (2011) [46] T.K.O. Doan, P. Kubán, 464–471. [47] W. Pormsila, R. Morand, S. Krähenbühl, P.C. Hauser, Electrophoresis 32 (2011) 884–889. [48] D.T. Rajh Vidal, M.A. Augelli, G.M. Hotta, F.S. Lopes, C.L. do Lago, Electrophoresis 32 (2011) 896–899. ˇ ˚ [49] P. Tuma, K. Málková, E. Samcová, K. Stulík, J. Sep. Sci. 33 (2010) 2394–2401. [50] W. Pormsila, R. Morand, S. Krähenbühl, P.C. Hauser, J. Chromatogr. B 879 (2011) 921–926. [51] T. Mantim, D. Nacapricha, P. Wilairat, P.C. Hauser, Electrophoresis 33 (2012) 388–394. ˇ ˚ [52] P. Tuma, E. Samcová, K. Stulík, Anal. Chim. Acta 685 (2011) 84–90. ˇ P.C. Hauser, Meas. Sci. Technol. 17 (2006) [53] L. Zhang, S.S. Khaloo, P. Kubán, 3317–3322. [54] W.S. Law, J.H. Zhao, P.C. Hauser, S.F.Y. Li, J. Sep. Sci. 30 (2007) 3247–3254. [55] T. Hirokawa, M. Nishino, N. Aoki, Y. Kiso, J. Chromatogr. 271 (1983) D1–D106. [56] http://www.sigmaaldrich.com/life-science/core-bioreagents/biologicalbuffers/mes-specification.html (last accessed 22.11.12). [57] M. Ching-Lam Yeung, K. Man-Chung Wong, Y. Kai Tiu Tsang, V. Wing-Wah Yam, Chem. Commun. 46 (2010) 7709–7711. [58] K. Langera, S. Balthasara, V. Vogelb, N. Dinauerc, H. von Briesenc, D. Schubertb, Int. J. Pharm. 257 (2003) 169–180. [59] V.A. VA Gaevskaia, G. Azhitskii, Ukr. Biokhim. Zh. 50 (1978) 735–738. ˇ O. Oldhoff, B. Karlberg, J. Chromatogr. A 857 (1999) 321–326. [60] P. Kubán,