2023 Volume 48 Issue 1 Pages 15-24
We developed a derivatization technique that involves microwave heating to reduce the overall forensic analysis time of phosphorus-containing amino acid herbicides (PAAHs). Combined with an extraction method that uses titanium (IV) oxide (TiO2), we were able to obtain a practical analytical method for PAAHs and their metabolites in samples intended for poisoning cases. The optimized derivatization conditions were 700 W power and 5-min irradiation time, which is a significant time-saving. The plasma samples extracted using TiO2-packed Tip columns and derivatized under the optimized conditions had an intra-day accuracy and precision within 9.3% and 9.0%, respectively. The intermediate accuracy and precision were within 8.8% and 8.5%, respectively, and the recoveries were more than 91.2%. Similarly, for urine samples, the intra-day accuracy and precision were within 13.3% and 9.1%, respectively. The intermediate accuracy and precision were within 13.6% and 10.3%, respectively, and finally, the recoveries were more than 88.2%. In addition to reducing the pretreatment time, this method was suitable for reducing the overall labor burden on laboratories responsible for routine analysis because of its stable validation data.
Glyphosate (GLYP) and glufosinate (GLUF), which are classified as phosphorus-containing amino acid herbicides (PAAHs), are used on agricultural land and in no-crop areas, such as roads and gardens, because they are less toxic to humans and non-selectively kill unwanted plants and weeds. They are both used worldwide, and several formulations containing GLYP and GLUF have been launched in Japan. Because these formulations are readily available, PAAHs are subject to analysis in cases of poisoning by suicide or accidental ingestion as well as in cases of plant mortality. According to the most recent statistics from the Japan Poison Information Center, in terms of the number of telephone consultations received regarding acute human toxicity, more than half of the herbicide-related calls involved GLYP and GLUF (Japan Poison Information Center, 2020). There have also been many reported cases of suicide and poisoning due to ingestion of these herbicides (Zouaoui et al., 2013). Therefore, PAAHs and their metabolites, such as aminomethylphosphonic acid (AMPA) and 3-methylphosphinicopropionic acid (MPPA), are important analysis targets in forensic science to solve poisoning cases.
Because PAAHs and their metabolites are extremely polar (Fig. 1), analytical methods different from those used for other drugs are required. Also, since forensic samples are expected to be in a variety of matrices, a method specific to PAAHs and their metabolites is desirable for efficient sample extraction. As a selective extraction method, solid-phase extraction (SPE), which uses titanium (IV) oxide (TiO2) and zirconium (IV) oxide (ZrO2), can be applied to specifically adsorb the phosphonic and phosphinic acid groups of PAAHs for effective extraction (Ishiwata et al., 2007; Saito et al., 2020; Watanabe et al., 2014a).
N-Acetyl-O-methyl derivatizations of glyphosate (GLYP) and glufosinate (GLUF), which are classified as phosphorus-containing amino acid herbicides (PAAHs), their metabolites, aminomethylphosphonic acid (AMPA) and 3-methylphosphonicopropionic acid (MPPA), and the internal standard (IS).
Because these components have multiple polar functional groups, derivatization is generally required for instrumental analysis performed after extraction. Post-extraction analytical methods include tert-butyldimethylsilyl-derivatized gas chromatography-mass spectrometry (GC-MS) (Hori et al., 2003; Motojyuku et al., 2008; Tsunoda, 1993), N-labeled high-performance liquid chromatography (HPLC) (Akuzawa and Akaiwa, 1997; Hori et al., 2002), and N-acetyl-O-methyl-derivatized liquid chromatography-mass spectrometry (LC-MS) (Sato et al., 2009; Watanabe et al., 2014b). Recently, a direct detection method that does not require derivatization has been reported. The method uses LC analytical columns with amphoteric ions (Saito et al., 2020; Yoshioka et al., 2011). Utilizing some analytical methods for PAAHs that are similar to those used in routine qualitative and quantitative analyses contributes to overall labor reduction.
Among these detection methods, N-acetyl-O-methyl derivatization LC-MS can be used with reversed-phase analytical columns, which are frequently used in routine analyses, and the metabolites to be analyzed can also be derivatized. Additionally, because excess derivatization reagent can be removed, there is little damage to the analytical instrumentation. The derivatized products can also be analyzed using GC-MS (Stalikas and Pilidis, 2000), making them versatile in the field of forensic science. However, since this derivatization process requires a long time and high-temperature heating, the development of rapid derivatization technology is desired.
Recently, derivatization involving microwave heating has been investigated for various substances to accelerate the derivatization process (Deng et al., 2005; Luca et al., 2003; Xu et al., 2011). In forensic science, it has been applied to drugs of abuse to shorten their reaction time (De Brabanter et al., 2013; Margalho et al., 2020; Chung et al., 2008). In some of these reports, experiments were conducted using microwave irradiation in domestic microwave ovens, which also reduces the economic burden on laboratories. If microwave heating can be applied to the N-acetyl-O-methyl derivatization of PAAHs, it is expected to enable the rapid analysis of PAAHs and help reduce labor.
Therefore, the purpose of this study was to develop an analytical method for identifying and quantifying PAAHs in biological samples, involving extraction with TiO2 followed by derivatization under microwave heating, to significantly reduce the overall analysis time through rapid derivatization. Microwave irradiation was performed using a domestic microwave oven and other equipment commonly used in laboratories. This method enabled the identification and quantification of PAAHs, including metabolites, from plasma and urine in a short time, thereby reducing the overall burden on routine analysis. It was also useful for screening because of its rapid processing.
GLYP, GLUF ammonium salt, and MPPA were purchased from FUJIFILM Wako Pure Chemicals Corporation (Osaka, Japan). AMPA and the internal standard (IS), DL-2-amino-3-phosphonopropionic acid, were purchased from Sigma-Aldrich (St. Louis, MO, USA). Trimethylorthoacetate (TMOA) was purchased from Sigma-Aldrich (St. Louis, MO, USA), whereas acetic acid and acetic anhydride were purchased from FUJIFILM Wako Pure Chemicals Corporation (Osaka, Japan). A Titansphere® Phos-TiO Spin Tip column (5010-21308) packed with TiO2 was purchased from GL Science (Tokyo, Japan) and used for SPE. Plasma was purchased from BioIVT (Westbury, NY, USA) and urine was obtained from Lee Biosolutions (Maryland Heights, MO, USA). Ultrapure water was obtained using Milli-Q Integral 3 (Merck Millipore, Billerica, MA, USA). All other chemicals were purchased from FUJIFILM Wako Pure Chemicals Corporation. The standard solution to be analyzed in this study was prepared as a 1 mg/mL stock solution by dissolving it in ultrapure water. This stock solution was added to blank plasma and urine samples to achieve the appropriate concentrations and was used as the validation sample.
SPE using Titansphere® Phos-TiO Spin Tip columnSPE was performed using Titansphere® Phos-TiO Spin Tip columns via centrifugation at 500 × g. Prior to sample loading, the Tip column was conditioned with 100 µL of acetonitrile followed by 100 µL of a 0.1% trifluoroacetic acid (TFA)-50% acetonitrile solution (solution A). Prior to loading, 20 µL of plasma or urine samples were deproteinized by adding 40 µL of 12.5% TFA, and the supernatant was used as the loading sample. The loading sample (50 µL) was mixed with 100 µL of solution A and loaded onto the Tip column after conditioning. The column was washed twice with solution A and eluted with 50 µL of 5% aqueous ammonia. The eluate was evaporated to dryness in a 1.5-mL screw-cap glass vial and used for subsequent derivatization.
Microwave derivatizationDerivatization was performed in a domestic microwave oven EMM2300JW6 (Electrolux, Stockholm, Sweden). The stability of the microwave irradiation was assessed because it would affect the reproducibility of the experiments. The temperature change in 1 L of ultrapure water was measured after microwave irradiation at 700 W (nominal) for 5 min, which were the optimum conditions for the microwave irradiation. The measurements were performed every weekday for 2 months to ensure consistent performance of the domestic microwave system that is commercially available for general food processing. The effective absorbed power used to increase the temperature (ΔT) of ultrapure water was calculated using the following formula (Houšová and Hoke, 2002):
where P is the effective absorbed power (W), ΔT is the temperature increase (K), and t is the irradiation time (s).
The derivatization reagent was prepared by mixing equal amounts of TMOA and acetic acid with 10% acetic anhydride. The resultant derivatization reagent mixture (30 µL) was added to the extraction residue in a glass vial, the lid was immediately closed, and the vial was placed at the center of the microwave oven. The microwave power for derivatization was 700 W (nominal) and the irradiation time was 5 min. The reaction solution was evaporated to dryness under reduced pressure and dissolved in 100 µL of LC-MS eluent A for analysis.
LC/ESI-TOF-MS conditionsSamples were analyzed using a micrOTOFII TOF-MS system (Bruker Daltonics, MA, USA) coupled to a Thermo Scientific UltiMate 3000 HPLC system (Thermo Scientific, Waltham, MA, USA). An Inertsil C30 S-Select LC column (GL Science, Tokyo, Japan) with a particle size of 3 µm and column dimensions of 2.0 × 150 mm was used. The column oven temperature was maintained at 30°C, 10 mM ammonium formate-0.1% formic acid was used as eluent A, and acetonitrile was used as eluent B. For isocratic analysis with 15% eluent B at a 0.2 mL/min flow rate, 10 µL were injected to separate the analytes.
The eluent that passed through the column was pumped directly to the electrospray ion (ESI) source of the TOF-MS and measured in positive ion mode. The parameters of the MS were as follows: scan range: m/z: 50–300, capillary voltage: 4.5 kV, capillary exit voltage: 100 V, hexapole RF voltage:100 Vpp, dry gas (N2): 8.0 L/min, dry temperature: 180°C, nebulizer (N2): 1.6 bar. Each N-acetyl-O-methyl-derivatized component was analyzed using the Compass DataAnalysis software (Version 4.0, Bruker Daltonics) at m/z: 254.0793 (GLYP and IS), m/z: 182.0582 (AMPA), m/z: 252.1001 (GLUF), and m/z 181.0630 (MPPA), and was detected in the extracted-ion chromatogram.
Validation dataIn this study, validation data were obtained for the limit of detection (LOD), limit of quantification (LOQ), precision and accuracy of intra-day and intermediate quantification, and extraction recovery to verify the optimized method. Standard samples for the calibration curve were prepared by spiking blank plasma and urine samples with the stock solution to achieve concentrations of 0.025, 0.05, 0.1, 0.25, 0.5, and 1 µg/mL for each component. Quality control (QC) solutions were prepared from blank plasma and urine samples spiked at concentrations of 0.05, 0.5, and 1 µg/mL. A volume of 10 µL of 1 µg/mL IS solution prepared in ultrapure water was added to the mixture of the loading sample and solution A for subsequent processing. The calibration curve was plotted with the peak area ratio of the analyte and IS on the y-axis and the known concentrations of the calibration standard samples on the x-axis, and a linear equation was obtained using the least-squares method.
The LOD was determined as the lowest concentration that would give a response greater than three times the baseline noise, and the LOQ was defined as the concentration at which the accuracy of quantitation was less than 20%.
Intra-day accuracy was determined by analyzing five replicates of QC samples for each concentration in one day. The intermediate accuracy was determined from the average of three repeated analyses of five replicates of the QC samples performed in one day.
Extraction recoveries were determined by preparing two sets of samples: One set was a Tip-column-extracted sample of each concentration of QC sample (i.e., pre-spiked sample), and the other was a sample of blank plasma and urine Tip-column eluate spiked with the same amount of components as the pre-spiked sample (i.e., post-spiked sample). Samples were prepared by spiking 10 µL of 1 µg/mL IS into all pre- and post-spiked samples. After the LC/ESI-TOF-MS analysis, the peak area ratio of the pre-spiked sample/IS was divided by the peak area ratio of the post-spiked sample/IS and expressed as a percentage, and the recovery was calculated as the average of five repetitions.
Because microwave output affects the reproducibility of the derivatization, it is important to confirm the microwave output over time and understand the stability of the equipment (De Brabanter et al., 2013; Margalho et al., 2016; Luca et al., 2003). In this study, the temperature rise of ultrapure water caused by microwave radiation was monitored every weekday for two months, and the effective absorbed power was calculated using Eq. 1. The effective absorbed power of this device when ultrapure water was heated with 700 W microwave radiation for 5 min was 539 ± 7.9 W.
Optimization of the microwave derivatization conditionsBecause microwave heating has not been investigated for the derivatization of PAAHs using TMOA, we needed to optimize the microwave irradiation conditions. During microwave-induced derivatization, the time, power, and solvent are important parameters. In this study, reagent addition and amount were examined in addition to time and power. The solvent of choice was acetic acid, which is used also in conventional methods. Alcohols with high microwave heating efficiencies were not used, considering the interference to the experiment caused by excessive heating and the complexity of operation. Other solvents were not considered because of the moderate heating efficiency of acetic acid and ease of distillation of the derivatization reagent (Tokuyama and Nakamura, 2005). An amount of 10 ng of each PAAH and their metabolites used for optimization were derivatized. The derivatized products were analyzed using LC/ESI-TOF-MS, and the reaction efficiency was evaluated using the peak area values obtained. In the subsequent verification of microwave power, irradiation time, and added reagents, the derivatization reagents used were 50 µL of TMOA and 50 µL of acetic acid.
Effect of microwave powerTo confirm the effect of microwave radiation on the derivatization of PAAHs and their metabolites, the derivatization reaction was performed at microwave powers of 500 and 700 W (the latter being the maximum power) and irradiation times of 1 and 3 min. As shown in Fig. 2, GLYP is detected with a high intensity under all conditions. Other components are detected with high intensities at 700 W for 3 min but AMPA reaction efficiency is lower under all conditions. Sato et al. (2009) noted that the reaction efficiency of AMPA was not as good as that of other components and that higher reaction temperatures or longer reaction times were required. Notably, our current results support previous findings. Because the reaction efficiency of GLYP is good, AMPA, which has the same phosphonic acid structure as GLYP, is considered to have low reaction efficiency for N-acetylation. The microwave power that could efficiently derivatize each component in this experiment was found to be 700 W and was used in subsequent experiments.
Effect of microwave heating with different microwave powers using TMOA and acetic acid as derivatization reagents. TMOA (50 µL) and acetic acid (50 µL) were added to 10 ng of each component, and derivatization was performed using microwave heating. After derivatization via microwave irradiation at 500 and 700 W for 1 and 3 min, each component corresponding to 1 ng was analyzed using LC/ESI-TOF-MS, and the mean ± S.D. of the obtained peak areas is shown. N. D., not detectable.
To increase the reaction efficiency of the N-acetylation of AMPA, acetic anhydride, which is an acetyl group source, was added to the derivatization reagent, and the production efficiency of each derivatization product was verified. Acetic anhydride was added at 10%, 50%, and 90% to 50 µL of the derivatization reagent acetic acid and irradiated with 700 W microwave radiation for 1 and 3 min. The results are shown in Fig. 3. The detection intensity of AMPA increased with the addition of acetic anhydride. AMPA is highly effective after 3 min of irradiation at all three acetic anhydride concentrations tested. For the other components, a decrease in derivatization products is observed with increasing concentration of acetic anhydride. This decrease in derivatization products occurs under insufficient amounts of acetic acid, suggesting that the amount of acetic acid is also important for adequate derivatization of GLYP, GLUF, and MPPA. Based on the verification results, we report that the acetic anhydride concentration that efficiently derivatizes all components is 10%.
Effect of adding acetic anhydride to the derivatization reagent. Derivatization was performed by adding 50 µL of TMOA and acetic acid, respectively, supplemented with 10%, 50%, and 90% acetic anhydride, to 10 ng of each component followed by microwave heating. After derivatization via microwave irradiation at 500 and 700 W for 1 and 3 min, each component corresponding to 1 ng was analyzed using LC/ESI-TOF-MS, and the mean ± S.D. of the obtained peak areas is shown.
The reaction efficiency was verified by setting the total amount of derivatization reagent to 10, 30, 50, 70, and 100 µL; the results are shown in Fig. 4. GLYP and GLUF exhibit maximum peak areas at 50 µL, which decrease at larger reagent volumes. AMPA and MPPA display maximum peak areas at 30 µL, which decrease at larger reagent volumes. The observed decrease in the respective reaction efficiencies can be attributed to the diffusion of irradiated power in the reaction solution as the liquid volume increases. Because the actual biological samples contain trace amounts of PAAH metabolites relative to the parent compounds (Hori et al., 2003; Motojyuku et al., 2008; Yoshioka et al., 2011; Zouaoui et al., 2013), the optimum total volume of the derivatization reagent is 30 µL, prioritizing the reaction efficiency of AMPA and MPPA.
Optimization of derivatization reagent amounts. TMOA and acetic acid with 10% acetic anhydride (1:1, v/v) was added to 10 ng of each component and the total amount was added at 10 to 100 µL, followed by derivatization under microwave heating. After derivatization using microwave irradiation at 700 W for 3 min, each component corresponding to 1 ng was analyzed using LC/ESI-TOF-MS conditions, and the mean ± S.D. of the obtained peak areas is shown. Each component in the figure is indicated as follows (□): GLYP, (❖): AMPA, (●): GLUF, and (▲): MPPA.
PAAHs and their metabolites with optimized derivatization reagents were derivatized within microwave irradiation times ranging from 1 to 6 min to verify the amount of each derivatization product. As shown in Fig. 5, all the components display an increase in reaction products with increasing irradiation time. Over the time range measured, there is no excessive reduction of product amounts, and all components reach equilibrium at approximately 5 min. Therefore, the optimum irradiation time is 5 min because excessive heating could lead to the decomposition of reaction products and prolong the analysis time.
Optimization of the microwave irradiation time by varying the heating time. A total of 30 µL of TMOA and acetic acid with 10% acetic anhydride (1:1, v/v) was added to 10 ng of each component, and the derivatization was performed using microwave heating at 700 W for an irradiation time of 1 to 6 min. After derivatization, each component corresponding to 1 ng was analyzed using LC/ESI-TOF-MS. The mean ± S.D. of the obtained peak areas is shown. Each component in the figure is indicated as follows (□): GLYP, (❖): AMPA, (●): GLUF, and (▲): MPPA.
At the end of the optimization experiment, the performance of the microwave heating method was verified by comparing the detection intensity of each component derivatized using the conventional heating method and microwave heating method optimized in this study. In the conventional heating method, 50 µL of TMOA and 50 μL of acetic acid were heated using a sand bath heater. The results are shown in Fig. 6. For all components, the detection performance of the microwave heating method is comparable with that of the conventional heating method. The time required for the derivatization of PAAHs and their metabolites with TMOA and acetic acid is significantly reduced when using microwave irradiation. Reducing the derivatization time from at least 30 min to 5 min is beneficial for forensic analysis.
Comparison of detection intensities between the conventional and microwave heating methods. Each component (10 ng) was derivatized via the conventional heating method with TMOA and acetic acid, and microwave heating method with TMOA and acetic acid with 10% acetic anhydride at 700 W for 5 min. After analyzing each component corresponding to 1 ng using LC/ESI-TOF-MS, the mean ± S.D. of the obtained peak areas is shown.
Samples of plasma and urine containing PAAHs and their metabolites were used for poisoning cases. The prepared calibration and QC samples were extracted in TiO2-packed columns, and the resulting PAAHs and their metabolites were derivatized under the optimized microwave irradiation conditions. After analyzing the samples using LC/ESI-TOF-MS, the validation data were calculated from the obtained peak area values of each component. DL-2-amino-3-phosphonopropionic acid was used as the internal standard to calculate these data. Recently, the technique of using a compound labeled with a stable isotope as an internal standard has been widely used in quantitative analysis because of its high accuracy. However, stable isotope-labeled internal standards are commercially unavailable for all analytes of interest, and these commercial products are generally expensive. This method uses internal standards that are less expensive than stable isotope-labeled compounds and are structural isomers of GLYP, thereby reducing the economic burden and facilitating their introduction into routine work. DL-2-amino-3-phosphonopropionic acid has previously been used to quantitatively analyze PAAHs in blood and urine samples, achieving good results (Hori et al., 2003; Motojyuku et al., 2008; Saito et al., 2020; Sato et al., 2009). DL-2-amino-3-phosphonopropionic acid and GLYP did not interact under the separation conditions used in this method. The calibration curve data for each component are shown in Table 1. The accuracy, precision, and extraction recoveries for plasma- and urine-spiked samples are shown in Tables 2 and 3, respectively.
| Compounds | Matrix | Linear range (µg/mL) | R2 | LOD (µg/mL) |
|---|---|---|---|---|
| GLYP | Plasma | 0.025-1 | 0.9997 | 0.010 |
| Urine | 0.025-1 | 0.9997 | 0.025 | |
| AMPA | Plasma | 0.025-1 | 0.9999 | 0.010 |
| Urine | 0.025-1 | 0.9995 | 0.010 | |
| GLUF | Plasma | 0.025-1 | 0.9999 | 0.005 |
| Urine | 0.025-1 | 0.9990 | 0.010 | |
| MPPA | Plasma | 0.025-1 | 0.9996 | 0.010 |
| Urine | 0.025-1 | 0.9999 | 0.025 |
| Compounds | Spiked Level (µg/mL) |
Intra-day (n = 5) | Intermediate (n = 3) | Recovery (%) | |||
|---|---|---|---|---|---|---|---|
| Accuracy (%RE) | Precision (%RSD) | Accuracy (%RE) | Precision (%RSD) | ||||
| GLYP | 0.05 | −7.0 | 5.6 | 2.6 | 8.5 | 101.1 | |
| 0.5 | −3.7 | 2.7 | −3.1 | 4.1 | 97.7 | ||
| 1 | −3.6 | 2.8 | −2.0 | 1.9 | 93.1 | ||
| AMPA | 0.05 | 3.0 | 9.0 | 8.8 | 5.0 | 96.5 | |
| 0.5 | 4.8 | 0.9 | 5.8 | 4.7 | 98.4 | ||
| 1 | 6.8 | 2.9 | 7.5 | 0.6 | 91.2 | ||
| GLUF | 0.05 | −4.6 | 5.7 | −0.9 | 8.0 | 97.2 | |
| 0.5 | 0.1 | 4.6 | −3.9 | 4.1 | 97.4 | ||
| 1 | −1.3 | 1.5 | −0.6 | 3.0 | 94.8 | ||
| MPPA | 0.05 | −9.3 | 3.2 | −2.8 | 7.5 | 96.6 | |
| 0.5 | −2.7 | 4.5 | 4.7 | 6.2 | 102.5 | ||
| 1 | −3.7 | 6.5 | 3.3 | 8.3 | 97.9 | ||
| Compounds | Spiked Level (µg/mL) |
Intra-day (n = 5) | Intermediate (n = 3) | Recovery (%) | |||
|---|---|---|---|---|---|---|---|
| Accuracy (%RE) | Precision (%RSD) | Accuracy (%RE) | Precision (%RSD) | ||||
| GLYP | 0.05 | 2.5 | 6.1 | 5.4 | 2.7 | 94.5 | |
| 0.5 | 0.8 | 5.1 | −0.2 | 1.0 | 91.6 | ||
| 1 | 8.0 | 4.2 | 7.4 | 5.8 | 103.1 | ||
| AMPA | 0.05 | 7.1 | 2.2 | 9.1 | 3.0 | 92.3 | |
| 0.5 | 2.2 | 5.6 | 2.1 | 6.7 | 91.9 | ||
| 1 | 12.6 | 3.3 | 11.6 | 2.6 | 102.1 | ||
| GLUF | 0.05 | 6.9 | 5.4 | 4.9 | 10.1 | 91.7 | |
| 0.5 | 7.5 | 6.5 | 2.5 | 4.7 | 88.2 | ||
| 1 | 7.5 | 9.1 | 7.3 | 0.2 | 94.8 | ||
| MPPA | 0.05 | 9.2 | 2.2 | −2.3 | 10.3 | 95.8 | |
| 0.5 | 5.9 | 5.9 | 4.3 | 2.6 | 89.6 | ||
| 1 | 13.3 | 7.1 | 13.6 | 0.8 | 100.8 | ||
All components show linearity in the range of 0.025–1 µg/mL. The LOD for GLYP, AMPA, and MPPA in the plasma-spiked samples is 0.010 µg/mL, whereas the LOD for GLUF is 0.005 µg/mL. In the urine-spiked samples, the LOD is 0.025 µg/mL for GLYP and MPPA, and 0.010 µg/mL for GLUF and AMPA. The LOD does not necessarily match the tendency of the study results shown in Figs. 4 and 5. These figures show the results of the derivatization reagent amounts and microwave irradiation time, indicating that the detection intensity tended to be higher for GLUF, GLYP, MPPA, and AMPA under optimal conditions. This trend indicates that the LOD of GLUF is lower than that of each component in plasma- and urine-spiked samples. However, the LODs for GLYP, AMPA, and MPPA did not follow this trend. These differences can be attributed to the influence of each matrix. In addition, the differences in the LOD of GLYP, GLUF, and MPPA among the matrices may also indicate that these components are more strongly influenced by the matrix and their own structural properties compared to the other components. The LOD of each component had the same capability level as that reported in previous studies on the analysis of PAAHs (Guo et al., 2018; Sato et al., 2009; Watanabe et al., 2014a; Yoshioka et al., 2011). The concentration of GLYP in blood samples from poisoning cases is at mild levels of several micrograms per milliliter. In addition, it has been reported that the concentration of AMPA in blood samples is lower than that of unmetabolized GLYP (Guo et al., 2018; Hori et al., 2003; Yoshioka et al., 2011; Zouaoui et al., 2013). The concentration of GLUF in blood samples have also been reported to range from a few to several hundred micrograms per milliliter, depending on the degree of intoxication (Akuzawa and Akaiwa, 1997; Hirose et al., 1999; Hori et al., 2002; Saito et al., 2020), and the concentration of MPPA in serum samples have been reported to be lower than those of parent unmetabolized GLUF (Hori et al., 2001). In the case of urine samples, concentrations of PAAHs are often higher after ingestion owing to their high polarity, with urinary GLYP concentrations exceeding several hundred micrograms per milliliter at the mildly impaired levels (Zouaoui et al., 2013). Additionally, urinary AMPA levels are lower than those of the parent unmetabolized GLYP (Hori et al., 2003; Zouaoui et al., 2013). Similar to GLYP, urinary GLUF has been reported to be present at high concentrations (Akuzawa and Akaiwa, 1997; Hirose et al., 1996; Hirose et al., 1999), and urinary MPPA is thought to occur at low concentrations. Fig. 7 shows chromatograms obtained by extracting and analyzing blank plasma and urine samples spiked with PAAHs and their metabolites at 0.5 µg/mL (0.5 µg/mL QC sample). Because each component can be clearly detected, even at concentrations lower than the mild level, this method is applicable to actual blood and urine samples. The ability to detect low levels of metabolites is extremely useful for determining poison ingestion and estimating the sample itself.
Extraction ion chromatograms of PAAHs and their metabolites in samples extracted from plasma and urine. Plasma and urine extract samples at 0.5 µg/mL were labeled A and B, respectively. The chromatograms of each a, b, c, and d were plotted in a software as follows; a: m/z 254.0793 ± 0.005, b: m/z 182.0582 ± 0.005, c: m/z 252.1001 ± 0.005, and d: m/z 181.0630 ± 0.005. Numbers in the chromatogram are as follows; 1: IS (DL-2-amino-3-phosphonopropionic acid), 2: GLYP, 3: AMPA, 4: GLUF, and 5: MPPA.
The accuracy and precision of the plasma and urine samples were validated using three spiked concentrations (0.05, 0.5, and 1 µg/mL). The recoveries at these concentrations were also evaluated. The intra-day accuracy (%RE, relative error) of plasma samples (n = 5) ranges from 0.1 to 9.3% across all components. The precision (%RSD, relative standard deviation) ranges from 0.9 to 9.0%. The intermediate accuracy (%RE) of plasma samples (n = 3) ranges from 0.6 to 8.8% across all components. The precision (%RSD) ranges from 0.6 to 8.5%. Furthermore, the recoveries at the three spiked concentrations ranges from 91.2 to 102.5% across all components. The accuracy and precision of the plasma samples converge to low values of less than 15% for all intra-day and intermediate data, and the recoveries for these concentrations are high.
Subsequently, the intra-day accuracy (%RE) for urine samples (n = 5) ranges from 0.8 to 13.3% across all components, whereas the precision (%RSD) ranges from 2.2 to 9.1%. The intermediate accuracy (%RE) for the urine samples (n = 3) ranges from 0.2 to 13.6% across all components, while the precision (%RSD) ranges from 0.2 to 10.3%. Furthermore, the recoveries at the three spiked concentrations ranges from 88.2 to 103.1% across all components. The accuracy and precision of the urine samples also converge to low values of less than 15% for all intra-day and intermediate data, and the recoveries for these concentrations are high. Validation data for plasma and urine samples were obtained well throughout this study, indicating that the developed method, which combines extraction with TiO2 and derivatization with microwave heating, is useful for proving cases of PAAH poisoning.
When crimes related to poisons occur, the results of the estimation of the poisoning agent at the crime scene provide extremely important investigative information. However, there are no means for identifying PAAHs in biological samples from crime scenes. Analytical methods, such as GC-MS and LC-MS, are essential for obtaining rapid and accurate poisoning information to substantiate a crime. PAAHs that are rapidly derivatized using the method developed in this study do not require special analytical methods and can be analyzed using general GC-MS or LC-MS. The fact that laboratories engaged in analysis can analyze the derivatized components obtained via this method using the analytical techniques that they routinely use will contribute to the further reduction of the analysis time.
However, the analysis of PAAHs and their metabolites using TiO2 often shows poor reproducibility of MPPA. If MPPA is the subject of analysis, fresh biological samples may be necessary. Because some sample matrices are difficult to apply, such as hemolyzed blood, continued improvement is desired in the application of such high matrix effects.
The purpose of this study was to analyze PAAHs and their metabolites in biological samples using rapid derivatization and to shorten the overall analysis time. Based on the above results, by optimizing the microwave heating conditions, the derivatization treatment time, which is normally at least 30 min, can be reduced to 5 min using this method. A stable analytical method was developed in this study using a TiO2-packed column for pre-extraction. This study is the first to verify the derivatization of PAAHs and their metabolites with microwave heating, and we believe that it will greatly contribute to speeding up the analysis of real samples.
Conflict of interestThe authors declare that there is no conflict of interest.
