2023 年 71 巻 1 号 p. 24-30
We have developed a fluorescence detection-liquid chromatography (HPLC-FL) method that involves sample pretreatment by solid-phase dispersive extraction (SPDE) and solid-phase fluorescence derivatization for the simple and rapid analysis of methamphetamine (MA) in urine. This method uses a reversed-phase polymeric solid-phase gel to clean up analytes in SPDE, followed by fluorescence derivatization with 9-fluorenylmethyl chloroformate (FMOC) in the solid-phase. The optimal conditions for SPDE and solid-phase fluorescence derivatization were obtained when J-SPEC PEP was used as the solid-phase gel and 0.5 mmol/L FMOC in 50 mmol/L borate buffer solution (pH 10) was used as the fluorescence derivatization reagent. The recovery experiment of MA in urine yielded a clean chromatogram with no interfering peaks, demonstrating the validity of our method; the recoveries were 83.6% when spiked at a low concentration level (100 ng/mL) and 80.7% when spiked at a high concentration level (1000 ng/mL). Compared with the conventional liquid-phase method, the reaction product (FMOC-MA) of solid-phase fluorescence derivatization had higher stability. Reaction rate constants were calculated by changing the temperature conditions, and physicochemical parameters, including activation energy and activation entropy involved in the degradation reaction, were obtained from the Arrhenius plot and analyzed thermodynamically. Taken together, our results suggest that the HPLC-FL method with SPDE and solid-phase fluorescence derivatization for sample pretreatment provides a simple and rapid means of analyzing MA in urine samples.
Methamphetamine (MA) is a stimulant with central nervous system stimulant effects that may temporarily induce euphoria, increase self-esteem, and enhance mental and physical performance.1) The abusive use of MA, however, can lead to immunodeficiency and neuropsychiatric disorders such as hallucinations and delusions.1) Moreover, MA leads to drug resistance, and psychological dependence may develop with repeated use, making MA abuse semi-permanent. In order to address issues surrounding MA, the Japanese government enacted the “Stimulant Control Law” in 1951, which stipulates procedures to be followed when handling stimulants, as well as penalties for their distribution and use. However, the number of arrests increased even after the law came into force and has recently reached over 10000 cases.2)
Against this backdrop, the need for a simple and rapid analytical method for MA in forensic science-related organizations, such as forensic science laboratories and the Ministry of Health, Labour and Welfare, has become apparent. Traditionally, MA analysis has been carried out using various biological samples such as blood,3–5) urine,4–10) saliva,11) sweat,12) nails,13) and hair.14) Roughly half of unchanged MA and its metabolites are excreted in urine within 24 h after administration, and approximately 90% is excreted within 4 d.15) Because the proportion of MA excreted in urine as unchanged drug and metabolites is the highest, accounting for about 40% of the administered dose,16) urine is a suitable biological sample for evaluating MA use, its benefits being a long window of detection and non-invasiveness.
In forensic science, the final conclusion of MA detection is drawn on the basis of assessment using multiple analytical methods with different principles. These methods include TLC,17) GC/MS,3,4,12,14) HPLC,5–8) LC/tandem mass spectrometry (LC/MS/MS),9) LC/time-of-flight mass spectrometry (LC/TOFMS),10) and radioimmunoassay (RIA).11) Although conventionally used for MA analysis, these methods have some shortcomings. For example, TLC is inexpensive but has low sensitivity and poor reproducibility, whereas mass spectroscopic analysis, which is highly sensitive, has issues such as complicated pretreatment procedures and high costs of mass spectrometers required for the analysis of biological samples. Although RIA also has high sensitivity and selectivity, it has some problems of limited handling due to the use of radioisotopes and the need for special dedicated measuring equipment.
Fluorescence detection is generally considered one of the most suitable methods for analyzing trace components in biological samples due to its high sensitivity, and fluorescence derivatization is widely used to increase detection sensitivity in HPLC. Regarding MA fluorescence derivatization, Farrell and Jefferies5) used sodium β-naphthoquinone-4-sulfonate (NQS), o-phthalaldehyde (OPA), and 4-chloro-7-nitro-2,1,3-benzoxadiazole (NBD-Cl); however, the stability of their reaction products was low. Hayakawa et al.6) described a fluorescence derivatization of MA that uses dansyl chloride (Dns-Cl) and naphthalene-2,3-dicarboxaldehyde (NDA) as the derivatization reagents for chemiluminescence detection. These reagents, however, require a long derivatization time and a post-column reaction system after pre-column derivatization, thereby increasing the complexity of the operation procedure. Another method reported by Falcó et al.7) also requires a cumbersome liquid-liquid extraction step following fluorescence derivatization with NQS. To address these issues, Herráez-Hernández et al.8) developed a fluorescence derivatization method for MA using 9-fluorenylmethyl chloroformate (FMOC). However, this method too needs a special device (i.e., a column-switching HPLC system) to remove excess reaction reagents. In addition, the pretreatment column used in this system needs to be replaced every 40 to 50 sample injections, making the method impractical, and furthermore, the repeated use of the same pretreatment column is a potential cause of carryover—a problem that remains to be solved in forensic appraisal. Despite these issues, fluorescence detection HPLC (HPLC-FL) is considered an excellent tool for MA analysis because the instruments used are highly versatile and have relatively high sensitivity. We have recently developed a novel extraction method called “solid-phase dispersive extraction (SPDE)”10,19–24) to overcome some of the drawbacks of the conventional cartridge-type solid-phase extraction (SPE) method. SPDE is performed in a closed system and can be used for infectious and/or chemically hazardous samples.
The solid-phase derivatization method, which is a developmental application of SPDE, is a method of reacting the target substance with a derivatization reagent while it is retained in the solid-phase, and has demonstrated usefulness in the analysis of mycotoxins in food.25) In the present study, we adopted SPDE for the analysis of MA in urine, and investigated a solid-phase fluorescence derivatization method using FMOC for the fluorescence derivatization of MA in the solid phase to achieve high sensitivity and selectivity. Here we report the development of this simple and rapid HPLC-FL method for the determination of MA in urine.
MA hydrochloride was purchased from Sumitomo Dainippon Pharma Co., Ltd. (Tokyo, Japan). MA was dissolved in water to make 1000 µg/mL standard stock solution. Working standard solutions were then prepared from the standard stock solution by dilution with water at appropriate concentrations. FMOC-Cl (special grade: >98%) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Half a millimolar per liter FMOC reagent was prepared by dissolving FMOC into water/acetonitrile mixture (19 : 1). Acetonitrile and methanol (both HPLC grade); sodium hydroxide, boric acid, and sodium tetraborate, 10-hydrate (borax; all special grade); were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Buffer solution (pH 10) was prepared by adding 50 mmol/L sodium hydroxide aqueous solution to 50 mmol/L sodium tetraborate aqueous solution until the pH reached 10.0. Water was purified with a Milli-Q Gradient A 10 system equipped with an EDS-Pak® Polisher (Merck KGaA, Darmstadt, Germany).
As for the reversed-phase (RP) polymeric solid-phase gels, J-SPEC PEP and J-SPEC PEP2 (JASCO Engineering Co., Ltd., Tokyo, Japan), Oasis HLB, MCX, WCX, and Porapak Rxn RP (Waters Corp., Milford, MA, U.S.A.), SupelMIP Amphetamines (Merck KGaA), and Strata-X 33 µm Polymeric Reversed Phase (Phenomenex Inc., CA, U.S.A.) were used. Each solid-phase gel was obtained by removing the gel from the corresponding SPE cartridge. Before use, each solid-phase gel was first conditioned with methanol and purified water sequentially, and then suspended in purified water at the concentration of 100 mg/mL. As the device for SPDE, @Roka and Captube centrifugation filter units were purchased from Frontier Science Co., Ltd. (Hokkaido, Japan). Blank human urine sample was purchased from KAC Co., Ltd. (Kyoto, Japan).
Apparatus and Operating ConditionsAn L-2100 Intelligent pump (Hitachi, Tokyo, Japan) system equipped with an L-2485 fluorescence detector (Hitachi) was used. LC separation was performed with an Inertsil® ODS-3 column (150 × 4.6 mm I.D., 5 µm; GL Science, Tokyo, Japan). Column temperature was maintained at 40 °C. The mobile phase was a mixture of acetonitrile and water in the ratio of 70 : 30 (v/v), and was delivered in the isocratic elution mode at the flow rate of 0.5 mL/min. The excitation and emission wavelength were set at 264 and 313 nm, respectively.
Pretreatment of Urine Samples by SPDESPDE was performed according to our previous method.23) Briefly, a Captube was set on top of @Roka, and a conventional 2.0 mL micro test tube was attached to the bottom of @Roka (Fig. 1). Urine sample (100 µL), water (300 µL), 50 mmol/L borate buffer solution (pH 10; 100 µL), and PEP gel suspension (100 µL; corresponding to 10 mg of PEP) were added into the Captube. The solid-phase gel suspension was immediately agitated for 10 s with a vortex mixer to sufficiently disperse the solid-phase gel in the sample solution. The centrifugal filter unit was centrifuged (2580 × g, 30 s) and the filtrate in the micro test tube attached to the bottom of @Roka was discarded together with the micro test tube. To wash the solid-phase gel, purified water (1000 µL) was added into the Captube, and the solid-phase gel was dispersed once again. Subsequently, the solvent phase was discarded as described above.
In the solid-phase fluorescence derivatization method, to the Captube that obtained above SPDE treatment, 50 mmol/L borate buffer solution (200 µL) and 0.5 mM FMOC reagent (100 µL) were added and the centrifugation filter unit for the SPDE was stirred with a vortex mixer for 1 min. After that, the SPDE device was centrifuged (2580 × g, 1 min). To wash the solid-phase gel, purified water (1000 µL) was added into the Captube and the solid-phase gel was dispersed once again. Subsequently, the solvent phase was discarded as described above. Finally, 1 mL of acetonitrile was added into the Captube and the solid-phase gel was dispersed again and then centrifuged (2580 × g, 1 min) once again to elute the reaction product (FMOC-MA). The elution operation was repeated two times, and the eluates were combined. A 10 µL aliquot of the combined eluates was injected into the HPLC-FL instrument. Figure 2 shows the fluorescence derivatization reaction by FMOC.
FMOC-MA was prepared in the solid- or liquid-phase and stored at 4, 23, and 72 °C, respectively. For the solid-phase reaction, 100 µL of MA standard solution (1000 ng/mL) prepared according to the procedure described above (sample preparation) (final MA concentration: 50 ng/mL) was used. On the other hand, for the liquid-phase reaction, 100 µL of MA standard solution (125 ng/mL) was added to 100 µL of borate buffer solution and 50 µL of FMOC, and the reaction was stirred with a vortex mixer for 60 s (final MA concentration: 50 ng/mL).
The time course after each reaction was measured by HPLC-FL, and the reaction rate constant was calculated. In addition, an Arrhenius plot was created to calculate activation energy (Ea) and activation entropy (ΔS).
Nine types of solid-phase gels were tested to determine which would contribute to the optimization of the SPDE conditions: HLB (30 and 60 µm particle size), PEP, PEP2, Strata-X, PoraPak Rxn, WCX, MCX, and SupelMIP. When WCX, MCX, and SupelMIP were used, the reactivity of fluorescence derivatization was low and an unknown peak derived from the FMOC reagent overlapped with the peak derived from MA. With regard to WCX and MCX, derivatization with the FMOC reagent was difficult because the amino group of MA is bound by WCX or MCX via an ion exchange interaction. As regards SupelMIP, derivatization was likewise difficult because it was hard for the FMOC reagent to penetrate into the site where MA and SupelMIP are bound. PEP2, Strata-X, PoraPak Rxn, and HLB reacted readily with the FMOC reagent; however, the influence of the FMOC-derived peaks was observed. On the other hand, PEP showed the best reactivity, and the effect of the FMOC-derived peak was hardly observed. From these results, it was determined that PEP was the best solid-phase gel for SPDE.
Optimum Conditions for Solid-Phase Fluorescence DerivatizationTo determine the optimum conditions for the solid-phase fluorescence derivatization, the concentration of the FMOC reagent and the pH of borate buffer solution as the reaction solvent were investigated. The concentration of the FMOC reagent was examined in the range of 0.01 to 1.5 mmol/L. As a result, the peak area of FMOC-MA increased with increasing FMOC concentration, and the peak area reached a maximum at 1.5 mmol/L (Supplementary Fig. S1). Because there was a tendency for the peak attributed to excess FMOC to increase when FMOC concentration exceeded 1.0, 0.5 mmol/L was selected as the concentration with relatively high sensitivity and minimal influence of interfering peaks.
The pH of borate buffer solution was compared in the range of 9.2 to 11, and the peak area of FMOC-MA showed a plateau in the pH range of 9.5 to 10.5. Therefore, pH 10 was adopted (Supplementary Fig. S2).
Validation of the Proposed HPLC-FL MethodThe limit of detection (LOD, S/N = 3) and the limit of quantification (LOQ, S/N > 10) of MA were 15 and 50 ng/mL, respectively. The calibration curve showed good linearity (r = 0.9972) in the range of 50 to 1000 ng/mL (Table 1). Accuracy, intra-day, and inter-day precision tests were performed using urine samples spiked with MA at a low concentration level (100 ng/g) and a high concentration level (1000 ng/g). Two replicate determinations at each concentration were carried out for five days each. Statistical analyses were performed using one-way ANOVA. Accuracy (average recovery) at the low concentration level was 83.6%, and repeatability and intermediate precision values were 8.7 and 9.3%, respectively. On the other hand, the accuracy at the high concentration level was 80.7%, and repeatability and intermediate precision values were 3.2 and 4.7%, respectively (Table 2). The representative chromatograms of MA standard solution, urine blank and urine sample to which MA at the high concentration level was added are shown in Fig. 3.
| LOD (ng/mL) | LOQ (ng/mL) | Linear range (ng/mL) | Correlation coefficient (r) | |
|---|---|---|---|---|
| MA | 15 | 50 | 50–1000 | 0.9972 |
LOD: Limit of detection (S/N = 3). LOQ: Limit of quantification (S/N > 10).
| MA | Spiked (ng/mL) | Accuracy mean (%) | Repeatability RSD (%) | Intermediate precision RSD (%) |
|---|---|---|---|---|
| Low conc. | 100 | 83.6 | 8.7 | 9.3 |
| High conc. | 1000 | 80.7 | 3.2 | 4.7 |
(n = 2 × 5 test).
Regarding the relationship between MA intake and urinary concentration, Ishimaru et al.26) reported that in 7 subjects with MA intake ranging from 0.02–0.30 g (average 0.07 g), urinary excretion within 0–24 h after intake (intravenous injection) was 1030 µM (153 µg/mL), only one sample was detected on the third day, and all samples on the fourth day and thereafter were below 200 µM (29.8 µg/mL). In addition, as a detection sensitivity for practical use, Amemiya and Nagai27) analyzed urine samples from MA abusers (68 subjects) by GC and found that the lowest concentration of MA was 1 µg/mL, with a cutoff value of 1 µg/mL. Since the LOD of the developed method is 15 ng/mL (Table 1), this method is convinced to be sufficiently practical.
Comparison of Reactivity between Solid-Phase and Liquid-Phase Reaction Methods in Fluorescence DerivatizationIn our previous work, we devised an on-column fluorescence derivatization HPLC method that uses OPA reagents, and found that the stability of OPA derivatives, which are considered to be less stable when derivatized in the liquid-phase, was significantly enhanced by the derivatization in the solid-phase.28) In this study, we investigated whether the stability of the derivatized product (FMOC-MA) would be similarly enhanced in the solid-phase fluorescence derivatization method using FMOC reagent. The proposed solid-phase fluorescence derivatization method and a conventionally used in vitro liquid-phase reaction that employs FMOC reagent were compared and examined. The liquid-phase reaction had low reactivity and a large variation of peak area (RSD: 30%), so that sufficient reproducibility could not be obtained. On the other hand, the solid-phase fluorescence derivatization method had high reactivity and its MA peak area was two to three times higher than that of the liquid-phase reaction. Moreover, its peak area showed good reproducibility with an RSD of 6.3%. The sample prepared by the solid-phase fluorescence derivatization method was stored at 4 °C and the time course was measured over 10 d. The MA peak area and the chromatogram were stable, exhibiting almost no change. These results indicate that the solid-phase fluorescence derivatization method has high reactivity and is able to store FMOC-MA stably for a long time.
Effect of Reagent on the Stability of FMOC-MAThe low stability of FMOC-MA prepared by the liquid-phase reaction was thought to be caused by the coexistence of excess (unreacted) FMOC reagent after the derivatization reaction, solvent polarity, and so on. The effects of the excess (unreacted) FMOC reagent and other solvents on FMOC-MA were investigated, as follows: an experimental method, an acetonitrile eluate (2 mL) of FMOC-MA produced by the solid-phase fluorescence derivatization method was nitrogen-purged and redissolved in 1 mL each of four solutions: 1) borate buffer + FMOC reagent, 2) borate buffer, 3) purified water, and 4) acetonitrile, and the time course was measured. In the case of borate buffer + FMOC reagent, the FMOC-MA peak overlapped with the peak derived from the excess FMOC reagent, making measurement difficult. When borate buffer and purified water were used, the FMOC-MA peak area gradually decreased with time (Figs. 4A, B) although it could be measured after redissolving in both solvents. When the polynomial approximations of these degradation processes were addressed as pseudo curves, both of them (Figs. 4A, B) showed almost the same degree of cubic function. On the other hand, FMOC-MA redissolved in acetonitrile was stable for 6 d without showing any degradation (Fig. 4C). These results suggest that the degradation of FMOC-MA proceeds in an aqueous solvent regardless of whether the solvent is basic or neutral, and that FMOC-MA is extremely stable in non-aqueous organic solvents such as acetonitrile. Together, the results suggest that the degradation of FMOC-MA may proceed via hydrolysis. The possibility that the reaction between FMOC and MA is a Schotten-Baumann reaction and the product is an easily hydrolyzable amide is supported by the results of this experiment.
Because the degradation of FMOC-MA obtained in the liquid-phase reaction was considered to be due to hydrolysis, it was feared that the coexistence of aqueous FMOC reagent after the reaction would affect the stability of FMOC-MA in the solid-phase fluorescence derivatization when FMOC-MA was exposed to the FMOC reagent for a long time. Therefore, we investigated the effects of the reaction time and the excess FMOC reagent after the solid-phase fluorescence derivatization. After the solid-phase fluorescence derivatization, the reaction mixture was not eluted immediately but left for 1 to 24 h, after which the excess FMOC reagent was removed by centrifugation. This was followed by washing and elution of the solid-phase, and measurement of the eluates by HPLC-FL. As a result, the peak area of MA varied more with time, and the peaks derived from impurities were larger, suggesting that the coexisting FMOC reagent after the solid-phase fluorescence derivatization affected the peak area. We confirmed that centrifugation should be performed immediately after vortexing to quickly remove excess (unreacted) FMOC reagent.
Thermodynamic Analysis of FMOC-MA Degradation Reaction in Liquid-Phase ReactionIn our previous study of the on-column fluorescence derivatization HPLC method using OPA reagents,28) we conducted a thermodynamic analysis of the degradation reaction of OPA derivatives and found that the solid-phase reaction enhanced the stability by decreasing ΔS compared with the liquid-phase reaction. In the solid-phase fluorescence derivatization method using the FMOC reagent developed in this study, the degradation rate of FMOC-MA could not be measured due to its extremely high stability, so we attempted to conduct a thermodynamic analysis of the degradation reaction of FMOC-MA in the liquid-phase. FMOC-MA derived from the liquid-phase reaction was stored at 4, 23, and 70 °C, respectively, and the time courses were measured to calculate the reaction rate constant (k). The results indicate that the degradation of FMOC-MA is a pseudo first-order reaction, as shown in Fig. 5. An Arrhenius plot was then created by calculating k from the slope of each regression line (Fig. 6), and Ea and ΔS calculated from the Arrhenius plot were 27.6 (KJ/mol) and −247.2 (J/K·mol), respectively. In general, the reaction is likely to proceed when Ea is significantly low and ΔS is high. The relatively low Ea value obtained in the experiment suggests that this degradation reaction proceeds relatively quickly. On the other hand, from the definition of Gibbs free energy (ΔG), ΔG = ΔH − T ΔS under constant temperature and pressure conditions. Because k increases as the reaction temperature increases, the degradation reaction is an endothermic reaction, and the activation enthalpy (ΔH) is inferred to be positive (>0). In fact, substituting the experimentally obtained Ea into the equation ΔH = Ea − RT confirmed that ΔH is positive (>0). In addition, because temperature T is positive (>0) and ΔS is negative (<0), −TΔS is positive (>0), and therefore, ΔG is always positive (>0).
a: Peak area immediately after reaction, b: peak area after each time.
Although it is expected that the degradation reaction will not occur spontaneously because ΔG is positive, the degradation reaction of FMOC-MA has a low energy barrier at the start of the reaction, suggesting that the degradation reaction usually proceeds easily under ambient temperature.
By using the SPDE method and the solid-phase fluorescence derivatization method, we were able to design a simple, rapid, and accurate HPLC-FL method for the analysis of MA in urine samples containing a number of impurities. Comparison of the solid-phase fluorescence derivatization method with the conventional liquid-phase reaction method revealed that the former is more suitable for FMOC-MA, and that FMOC-MA can be measured for 10 d after sample preparation and refrigeration storage (4 °C). This is because the solid-phase fluorescence derivatization method can remove excess FMOC reagent and aqueous solvent during the derivatization process, and the eluent is a non-aqueous organic solvent such as acetonitrile, which keeps FMOC-MA stable. That is, it is clarified that the solid-phase derivatization method developed in this study is superior to a conventional liquid-phase reaction method in terms of sensitivity and accuracy.
Taken together, the results suggest that the MA analytical method using SPDE and solid-phase fluorescence derivatization developed in this study may be applicable to the analysis of biological samples other than urine. We expect that this analytical method will be of great utility in forensic science.
This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (Grant No. 18K06609).
The authors declare no conflict of interest.
This article contains supplementary materials.
