Identification of postmortem product of amlodipine decomposition by hemoglobin with LC-Q-Orbitrap-MS

Abstract

Amlodipine (AM), a dihydropyridine calcium channel blocker, is frequently prescribed for hypertension in the clinical setting. Because AM has been detected in various lethal poisoning and suicide cases, it is important to determine its precise concentration in postmortem blood to serve as definitive evidence of death by intoxication. However, blood AM concentration at autopsy frequently differs from that at the time of death. In this study, we found that AM undergoes dehydrogenation by H₂O₂ at temperatures ranging from 4 to 45ºC. Mass spectra measured by quadrupole-Orbitrap mass spectrometry hyphenated with liquid chromatography showed the generation of 3-ethyl 5-methyl 2-((2-aminoethoxy)methyl)-4-(2-chlorophenyl)-6-methylpyridine-3,5-dicarboxylate (AM-PDP-1) in H₂O₂ and Hb/H₂O₂ reaction solutions incubated with AM and in postmortem blood of persons who died of drowning, fire, disease, drug poisoning, CO poisoning, traumatic shock, falling, or choking, after intentional ingestion of AM. AM-PDP-1 is the novel postmortem degradation compound of AM in blood. This compound in the Hb/H₂O₂ reaction solution was more stable than AM at 4–45ºC. These results show that AM-PDP-1, formed by H₂O₂-mediated postmortem AM decomposition, is a potential biomarker to correct for AM concentration in postmortem blood.

INTRODUCTION

Amlodipine (hereinafter “AM”), 3-O-ethyl 5-O-methyl 2-(2-aminoethoxymethyl)-4-(2-chlorophenyl)-6-methyl-1,4-dihydropyridine-3,5-dicarboxylate, is a dihydropyridine calcium channel blocker that was approved for use as an antihypertensive agent by the USFDA in 1987. Because of its selectivity for peripheral blood vessels, AM is less associated with myocardial depression and cardiac conduction abnormalities than other calcium channel blockers (Meredith and Elliott, 1992). In Japan, AM is widely used to treat as many as 43 million people with hypertension. In addition, AM should be used continuously because hypertension is a chronic condition (Hirawa et al., 2019). On the other hand, AM is one of the primary causes of death by suicide- or accident-related intoxication globally (Alvarez et al., 2020). AM overdose can result in hypotension, reflex tachycardia, coma, and death. The lethal blood AM concentration is 0.1−0.2 μg/mL (Schulz et al., 2020). Determination of the precise AM concentration in postmortem blood is essential in forensic toxicology to offer definitive evidence of death by intoxication, but AM concentrations in blood at autopsy may not reflect those at the time of death (Alvarez et al., 2020; Peters and Steuer, 2019). Currently, there is no definitive cause for the changes in AM concentration in postmortem blood.

Hemolysis is one of typical postmortem changes (Hirabayashi, 1953). Hemoglobin (Hb) is released from erythrocytes by hemolysis after death, and Hb is important contributor to postmortem changes in blood chemical concentration (Yamagishi et al., 2021a, 2021b, 2022, 2023, 2024). Hb-mediated postmortem changes of ingested chemicals are characterized by two key mechanisms. The first one is the oxidative decomposition of chemical compounds such as zolpidem, paliperidone, and bromazepam by hydroxyl radicals generated from the Fenton reaction between hydrogen peroxide (H₂O₂) and Hb (Yamagishi et al., 2021b, 2023, 2024). H₂O₂ is generated through the dismutation of the superoxide anion (O₂^•–) from oxyhemoglobin (HbO₂) autoxidation (Shikama 1998), and Hb has four divalent iron ions in its molecule. In addition, decomposition products by hydroxyl radicals generated from the Fenton reaction are the potential biomarker to correct for chemical compound concentration in postmortem blood. The second mechanism is the covalent binding of carbamate pesticides (e.g., aldicarb, methomyl, and oxamyl) to Hb (Yamagishi et al., 2021a, 2022). As far as we know, there are no published reports on the postmortem interaction between AM and Hb. Thus, in this study, we hypothesize that Hb influences postmortem changes in blood AM concentration and examine the relationship between them.

We measured AM and its decomposition product in each reaction solution for the Fenton reaction (e.g., H₂O₂, Hb, or Hb/H₂O₂). We used liquid chromatography hyphenated with quadrupole-Orbitrap mass spectrometry (LC-Q-Orbitrap-MS) to detect and identify AM and its decomposition product. Because Orbitrap-MS yields accurate molecular mass values and tandem MS such as Q-Orbitrap-MS provides chemical structure information (Yamagishi et al., 2023, 2024), LC-Q-Orbitrap-MS can detect untargeted small molecules such as unknown AM derivatives produced from postmortem changes. The derivatives produced from postmortem changes is the potential biomarker to correct for AM concentration in postmortem blood.

MATERIALS AND METHODS

Chemicals

AM (CAS No.: 88150-42-9), 0.1 M phosphate buffer solution (pH 7.4), 1 M ammonium formate solution (HPLC grade), H₂O₂ (30%, guaranteed grade), and formic acid (FA, 98–100%, LC–MS grade) were supplied from Fujifilm Wako Pure Chemical (Osaka, Japan). Diazepam-d5 was supplied from Hayashi Pure Chemical Ind., Ltd. (Osaka, Japan). Hb (Human, H7379) and 1 M MgCl₂ were obtained from Sigma-Aldrich (St. Louis, MO, USA). Acetonitrile (ACN, LC–MS grade) and water (LC–MS grade) were obtained from Kanto Chemical (Tokyo, Japan). NADPH Regeneration System was purchased from Promega (Madison, WI, USA). Pooled human liver microsomes from 200 donors of mixed-gender (20 mg/mL) was purchased from Xenotech (Kansas City, USA).

Instrumentation

LC-Q-Orbitrap-MS was used to detect and identify AM and its decomposition product. The liquid chromatography system is a Vanquish Flex Binary LC system (Thermo Fisher Scientific, Waltham, MA, USA). The Q-Orbitrap-MS instrument is a Q-Exactive plus Orbitrap mass spectrometer (Thermo Fisher Scientific). Table S1 shows the instrumentation and operational settings for liquid chromatography and mass spectrometry.

Recovery of AM from the reaction solutions containing chemicals for the Fenton reaction

The reported concentration and reaction time conditions of each chemical such as Hb and H₂O₂ in the Fenton reaction were used (Yamagishi et al., 2021b, 2023, 2024). To evaluate the recovery of AM from the reaction solution containing chemicals for the Fenton reaction, a 100 µL aliquot of a reaction solution containing Hb (final concentration, 0.1 g/mL), AM (final concentration, 100 ng/mL), and H₂O₂ (final concentration, 0.3 mg/mL) was used. The control samples contained 0.1 M phosphate buffer (pH 7.4) and AM, but not Hb or H₂O₂. The reaction solutions were incubated in an incubator at 37ºC for 0 and 24 hr. After incubation, 1,900 µL of ACN containing diazepam-d5 (25 ng/mL) as the internal standard (IS) was added. The solutions containing ACN were sonicated, vortexed, and centrifuged at 10,000 x g for 10 min, and the supernatant was subjected to LC-Q-Orbitrap-MS to determine AM concentration. The extracted ion of AM was detected at m/z 409.1525 with an m/z tolerance of 5 ppm.

To evaluate the temperature-dependent recovery of AM from the reaction solution containing chemicals for the Fenton reaction, temperatures were set at 4, 20, 37, and 45ºC. The pretreatment conditions were the same as those for the evaluation of AM recovery from the reaction solution containing chemicals for the Fenton reaction mentioned above.

Detection and identification of unknown AM decomposition product in the presence of H₂O₂ and Hb/H₂O₂

We used the same reaction and pretreatment conditions as those for the evaluation of AM recovery from the reaction solution containing chemicals for the Fenton reaction, with minor modifications. The final concentration of AM in the reaction solution was 10 µg/mL. Temperature was maintained at 37ºC.The reaction solution after pretreatment was subjected to LC-Q-Orbitrap-MS to detect and identify the unknown AM decomposition product in the sample. Compound Discoverer version 3.3 software (Thermo Fisher Scientific) was employed to detect the unknown AM decomposition product by difference analysis.

For determination of the unknown AM decomposition product by in vitro metabolism of AM with human liver microsomes, a 100 µL reaction solution consisted of 0.1 µg/mL AM, 1 mg/mL microsomes and NADPH Regeneration System solution in 0.1 mM phosphate buffer (pH 7.4). The reaction solution was incubated at 37°C for 2 hr. After incubation, 1,900 µL of ACN was added, and the mixture was stirred, centrifuged at 10,000 x g for 10 min, and the supernatant was subjected to LC-Q-Orbitrap-MS.

Stability of unknown AM decomposition product in the reaction solution containing chemicals for the Fenton reaction

We used the reaction and pretreatment conditions in the previous section “Detection and identification of unknown AM decomposition product in the presence of H₂O₂ and Hb/H₂O₂” with minor modifications. Incubation temperatures were set at 4, 20, 37, and 45ºC and sampling times were 2, 4, 8, and 24 hr. The samples after pretreatment were subjected to LC-Q-Orbitrap-MS to determine the concentrations of AM and the unknown AM decomposition product in the supernatant. The extracted ions of AM and the unknown AM decomposition product were detected at m/z 409.1525 and 407.1368 with an m/z tolerance of 5 ppm.

Detection of postmortem AM decomposition product in the blood of persons who died after intentional ingestion of AM

Autopsy samples were analyzed following the World Medical Association's Code of Ethics (Declaration of Helsinki). The ethics committees of Chiba University (approval no. 2819, Graduate School of Medicine; R001, Graduate School of Pharmaceutical Sciences) approved the present study. Forensic autopsies were conducted at the behest of death investigation agencies for criminal investigations, without interaction with families before or after the procedures. On our official Web site, we state that we collect samples from autopsies for research purposes. The families can choose to opt out by facsimile if they do not consent.

In the present study, postmortem blood samples were collected from ninety-seventh autopsy cases in which AM was not detected and seventeen autopsy cases in which AM was detected by a drug screening test performed in our laboratory. Blood samples were collected for 14 days postmortem. Causes of death included drowning, drug poisoning, fire, disease, CO poisoning, traumatic shock, falling, and choking. Autopsies were performed in January, March, April, May, June, July, August, September, November, and December.

We used the conditions for drug screening in our laboratory to detect the postmortem AM decomposition product. To a 100 µL aliquot of each blood sample, 400 µL of ACN containing IS (250 ng/mL) was added, and the mixture containing ACN was vortexed and centrifuged at 10,000 x g for 10 min. The supernatant was subjected to LC-Q-Orbitrap-MS. The extracted ion of the postmortem AM decomposition product was detected at m/z 407.1368 with an m/z tolerance of 5 ppm.

Statistics

The Student’s t-test was performed for comparisons of two groups. Pearson’s correlation analysis was performed between time after death and ratio of metabolite A to AM. Double (**) asterisks indicate level of significance at p < 0.01. Data are expressed as means ± standard deviation (SD).

RESULTS

Changes in AM concentration in the reaction solution containing chemicals for the Fenton reaction

The limit of detection for AM in the reaction solution by LC-Q-Orbitrap-MS was 1 ng/mL. AM concentrations in control and Hb solution at 37ºC decreased after 24 hr compared with that at 0 hr (Fig. 1). The percentages of undecomposed AM in control and Hb solution at 24 hr relative to those at 0 hr were 69.7 ± 3.8%, and 74.8 ± 1.2%, respectively. AM concentration in the Hb solution at 37°C was not significantly decreased after 24 hr compared with that in control. On the other hand, no AM was detected in H₂O₂ and Hb/H₂O₂ solutions at 37ºC after 24 hr.

Fig. 1

Effects of H₂O₂, Hb, and Hb/H₂O₂ on amlodipine concentration. A reaction solution containing 100 ng/mL amlodipine and H₂O₂, Hb, or Hb/H₂O₂ was incubated for 0 and 24 hr. The control sample was composed of 0.1 M phosphate buffer (pH 7.4) and amlodipine. Temperatures were maintained at 37ºC. The limit of detection for amlodipine in the reaction solution by LC-Q-Orbitrap-MS was 1 ng/mL. Data are expressed as means ± standard deviation (SD) (n = 4).

Effects of temperature on AM concentration in H₂O₂ and Hb/H₂O₂ solutions

Here as well, the limit of detection for AM in the reaction solution by LC-Q-Orbitrap-MS was 1 ng/mL. AM concentration in the H₂O₂ solution recorded a significant decrease at 4ºC after 24 hr compared with that at 0 hr and was no longer detected at 20, 37 or 45ºC after 24 hr (Fig. 2A). The percentages of undecomposed AM in H₂O₂ solution at 4, 20, 37, and 45ºC after 24 hr relative to that at 0 hr were 28.7 ± 1.6%, 0%, 0%, and 0%, respectively. The concentration decrease of AM in H₂O₂ solution at each temperature followed the order 20, 37, and 45ºC > 4ºC.

Fig. 2

Effect of temperature on the concentrations of amlodipine in H₂O₂ (A) and Hb/H₂O₂ (B). A reaction solution containing 100 ng/mL amlodipine and H₂O₂ or Hb/H₂O₂ was incubated for 0 and 24 hr. Temperatures were maintained at 4, 20, 37, and 45ºC. The limit of detection for amlodipine in the reaction solution by LC-Q-Orbitrap-MS was 1 ng/mL. The Student’s t-test was performed for comparisons of two groups. Significant differences are indicated by ** (p < 0.01). Data are expressed as means ± standard deviation (SD) (n = 4).

AM was not detected in Hb/H₂O₂ solution at all temperatures after 24 hr (Fig. 2B).

Detection and identification of unknown AM decomposition product in the H₂O₂ and Hb/H₂O₂ solution

Utilizing Compound Discoverer version 3.3 software, we performed a difference analysis of the mass spectra of control, H₂O₂- and Hb/H₂O₂-treated samples at 24 hr. An ion visible only in the H₂O₂- and Hb/H₂O₂-treated samples was extracted at m/z 407.1368 (Figs. 3B and 3C). This ion corresponds to the unknown AM decomposition product that was eluted at 6.9 min. We named the decomposition product Uk-1.

Fig. 3

Elution profiles of Uk-1 at m/z 407.1368 in control (A), H₂O₂ (B), and Hb/H₂O₂ (C) at 24 hr. A reaction solution containing 10 µg/mL amlodipine and H₂O₂ or Hb/H₂O₂ was incubated for 24 hr. Temperature was maintained at 37ºC.

Figure 4A shows the results of PRM (MS/MS analyses) of Uk-1 in the positive ion mode. One precursor ion (Uk-1-1) and six product ions (Uk-1-2–Uk-1-7) with m/z values of 407.1364, 364.0940, 346.0837, 318.0523, 314.0571, 286.0262, and 258.0313, respectively, were detected by Orbitrap-MS. Figure 4B and Table S2 show the assignment of Uk-1 precursor and product ions. Notably, the Δm/z values for the precursor and product ions of Uk-1 were less than ± 2.2 ppm compared with the theoretical m/z. In addition, this fragment pattern was in good agreement with that of 3-ethyl 5-methyl 2-((2-aminoethoxy)methyl)-4-(2-chlorophenyl)-6-methylpyridine-3,5-dicarboxylate produced by human microsome (Figs. 4A and 4C). Uk-1 was reasonably assigned to 3-ethyl 5-methyl 2-((2-aminoethoxy)methyl)-4-(2-chlorophenyl)-6-methylpyridine-3,5-dicarboxylate (hereinafter “AM-PDP-1”).

Fig. 4

MS/MS spectrum of Uk-1 detected in the positive ion mode (A), the assignment of Uk-1 precursor and product ions (B), and MS/MS spectrum of 3-ethyl 5-methyl 2-((2-aminoethoxy)methyl)-4-(2-chlorophenyl)-6-methylpyridine-3,5-dicarboxylate produced by human microsome (C). The retention time of Uk-1 was 6.9 min. The UK-1 peak in Fig. 3C was subjected to further LC-Q-Orbitrap-MS analysis (MS/MS analyses). The assignment of Uk-1 precursor and product ions is summarized in Figure 4B and Table S2.

Changes in AM-PDP-1 concentration in the reaction solutions for the Fenton reaction

No AM-PDP-1 was detected in all reaction solutions at 0 hr. AM-PDP-1 peak areas for the reaction solutions at 4 and 20ºC were not changed after 4, 8, and 24 hr compared with that at 2 hr. AM-PDP-1 peak areas for the reaction solution at 37ºC were not changed after 4 and 8 hr compared with that at 2 hr. On the other hand, the AM-PDP-1 peak area for the reaction solution at 37ºC decreased after 24 hr compared with that at 2 hr. AM-PDP-1 peak areas for the reaction solution at 45ºC showed a time-dependent decrease and a significant difference from that at 2 hr (Fig. 5). The stability of AM-PDP-1 in the reaction solution at each temperature decreased in the following order: 4 and 20ºC > 37 ºC > 45ºC.

Fig. 5

Changes in AM-PDP-1 peak areas with time after treatment. A reaction solution containing 10 µg/mL amlodipine and Hb/H₂O₂ was incubated for 2, 4, 8, and 24 hr. Temperatures were maintained at 4, 20, 37, and 45ºC. The Student’s t-test was performed for comparisons between 2 hr and other sampling times. Significant differences are indicated by ** (p < 0.01). Data are expressed as means ± standard deviation (SD) (n= 4).

Detection of AM decomposition product (AM-PDP-1) in postmortem blood of persons who died after intentional ingestion of AM

Figure 6A shows the elution profile of AM-PDP-1 at m/z 407.1368 in the postmortem blood collected from autopsy cases in which AM was not detected. AM-PDP-1 was not detected in all postmortem blood samples collected from autopsy cases in which AM was not detected. Figure 6B shows the elution profile of AM-PDP-1 at m/z 407.1368 in the postmortem blood of person who died after intentional ingestion of AM. AM-PDP-1 was detected in all postmortem blood samples collected from persons who died after intentional ingestion of AM (Fig. S1). The ratio of AM-PDP-1 peak area to AM peak area increased with time after death in the case of same death causes (Figs. 7A and 7B). The ratio of AM-PDP-1 to AM was calculated on the basis of their peak areas in the chromatograms, and was plotted versus the estimated time after death. A positive linear relationship was noted between the time after death and the ratio of AM-PDP-1 to AM (Figs. 7A and 7B), and the correlation coefficient (r2) was 0.55 and 0.57.

Fig. 6

Elution profiles of AM-PDP-1 at m/z 407.1368 detected in the blood collected from autopsy cases in which AM was not detected (A) and the blood of persons who died after intentional ingestion of amlodipine (B).

Fig. 7

Effect of time after death on ratio of AM-PDP-1 to AM peak area in blood collected from autopsy cases in which causes of death were fire (A) (n = 5) and drowning (B) (n = 4). x-axis was time after death (Days) and y-axis was ratio of metabolite A to AM (%). Pearson’s　correlation analysis was performed.

DISCUSSION

A novel mechanism for Hb-mediated postmortem changes of chemical compounds such as drugs and pesticides was reported recently (Yamagishi et al., 2021a, 2021b, 2022, 2023, 2024). Because the chemical structures of pesticides and drugs decomposed by Hb are altered after death, Hb seems to be able to decompose various drugs. Figure 1 shows that AM is decomposed by H₂O₂, indicating that AM is the target of dehydrogenation by H₂O₂ (Fig. 4B). H₂O₂ is generated by the dismutation of O₂^•– from HbO₂ autoxidation (Shikama 1998). HbO₂, a major protein found in erythrocytes, is released from erythrocytes by hemolysis after death (Hirabayashi 1953).

In forensic toxicology, drug concentrations in blood need to be accurately measured to prove death by drug intoxication. It has been reported that postmortem blood AM concentration of persons who died of AM intoxication is changed in various environments after death (Alvarez et al., 2020). Depending on the cause of death, dead bodies are exposed to various temperatures (e.g., 4–45ºC in Japan) (Lifschultz and Donoghue, 1998). We found that AM in H₂O₂ and Hb/H₂O₂ solutions was hardly detected at all temperatures examined for 24 hr (Figs. 2A and 2B). Thus, temperature is not the main factor influencing postmortem changes in AM concentration by Hb. In addition, AM shows high reactivity to H₂O₂. To summarize, our findings indicate that AM is decomposed in dead bodies exposed to various temperatures, and the rapid decomposition of AM takes place regardless of the temperature.

In the present study, AM-PDP-1 was detected in all postmortem blood samples collected from persons who died of drowning, drug poisoning, fire, disease, CO poisoning, traumatic shock, falling, and choking in all sessions such as January, March, April, May, June, July, August, September, November, and December after intentional ingestion of AM (Fig. 6B). AM in postmortem blood is decomposed at different temperatures and circumstances to which dead bodies are exposed after death, but postmortem AM-PDP-1 production is not affected by the cause of death (Fig. 2). In addition, AM-PDP-1 in the reaction solution is stable at 4 and 20 ºC and more stable than AM at 4–45ºC (Figs. 2 and 6). AM-PDP-1 concentration in post-mortem blood showed a time-dependent increase after death in the case of same death cause (Figs. 7A and 7B). This indicates that AM-PDP-1 is useful as a biomarker of AM poisoning in clinical practice. On the other hand, AM-PDP-1 is the major metabolite of AM in humans, and is detected in blood, feces, and urine (Stopher et al., 1988). AM is converted into AM-PDP-1 by cytochrome P450 (CYP) such as CYP3A4 in humans (Stopher et al., 1988). However, the activities of CYP enzymes, such as CYP3A4, are lost or severely reduced after death (Hansen et al., 2019). Because the amount of AM-PDP-1 increased in vivo with time after death, we thought that the increase was driven by H₂O₂ derived from HbO₂ rather than CYP in post-mortem blood. The issue of how to distinguish AM-PDP-1 produced before and after death remains unresolved.

In conclusions, AM was decomposed in dead bodies exposed to 4–45ºC temperatures in Japan. AM-PDP-1 formed by the dehydrogenation of its free form by H₂O₂ derived from HbO₂ is the postmortem decomposition product of AM. The accurate determination of AM concentration at the time of death on the basis of AM-PDP-1 levels remains challenging. The relationship between the concentrations of AM and AM-PDP-1 in postmortem blood requires further investigation.

Funding information

The present study was supported by a grant from the Asahi Glass Foundation, the Research Foundation for Pharmaceutical Sciences, and JSPS KAKENHI Grants Numbers 20K18978, 22K19664, and 25K02898.

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

•  Alvarez,  J.C.,  Mayer-Duverneuil,  C.,  Cappy,  J.,  Lorin de la Grandamison,  G. and  Knapp-Gisclon,  A. (2020): Postmortem fatal and non-fatal concentrations of amlodipine. Forensic Sci. Int., 316, 110555.
•  Hirawa,  N.,  Umemura,  S. and  Ito,  S. (2019): Viewpoint on Guidelines for Treatment of Hypertension in Japan. Circ. Res., 124, 981-983.
•  Hansen,  J.,  Palmfeldt,  J.,  Pedersen,  K.W.,  Funder,  A.D.,  Frost,  L.,  Hasselstrøm,  J.B. and  Jornil,  J.R. (2019): Postmortem protein stability investigations of the human hepatic drug-metabolizing cytochrome P450 enzymes CYP1A2 and CYP3A4 using mass spectrometry. J. Proteomics, 194, 125-131.
•  Hirabayashi,  A. (1953): Experimente über die hämolyse des blutes aus leichen in verschiedener zeit nach dem Tode. Arch. Histol. Jpn., 5, 565-571.
•  Lifschultz,  B.D. and  Donoghue,  E.R. (1998): Forensic pathology of heat-and cold-related injuries. Clin. Lab. Med., 18, 77-90.
•  Meredith,  P.A. and  Elliott,  H.L. (1992): Clinical pharmacokinetics of amlodipine. Clin. Pharmacokinet., 22, 22-31.
•  Peters,  F.T. and  Steuer,  A.E. (2019): Antemortem and postmortem influences on drug concentrations and metabolite patterns in postmortem specimens. WIREs Forensic Sci., 1, e1297.
•  Schulz,  M.,  Schmoldt,  A.,  Andresen-Streichert,  H. and  Iwersen-Bergmann,  S. (2020): Revisited: therapeutic and toxic blood concentrations of more than 1100 drugs and other xenobiotics. Crit. Care (Fullerton), 24, 195.
•  Shikama,  K. (1998): The molecular mechanism of autoxidation for myoglobin and hemoglobin. a venerable puzzle. Chem. Rev., 98, 1357-1374.
•  Stopher,  D.A.,  Beresford,  A.P.,  Macrae,  P.V. and  Humphrey,  M.J. (1988): The metabolism and pharmacokinetics of amlodipine in humans and animals. J. Cardiovasc. Pharmacol., 12 (Suppl 7), S55-S59.
•  Yamagishi,  Y.,  Iwase,  H. and  Ogra,  Y. (2021a): Post-mortem changes of methomyl in blood with hemoglobin. Chem. Res. Toxicol., 34, 161-168.
•  Yamagishi,  Y.,  Iwase,  H. and  Ogra,  Y. (2021b): Effects of hemoglobin on post-mortem oxidation of bromazepam. Fundam. Toxicol. Sci., 8, 61-67.
•  Yamagishi,  Y.,  Nagasawa,  S.,  Iwase,  H. and  Ogra,  Y. (2022): Evaluation of post-Mortem interaction between hemoglobin and oxime-type carbamate pesticides. Chem. Res. Toxicol., 35, 1110-1116.
•  Yamagishi,  Y.,  Inokuchi,  G.,  Hoshioka,  Y.,  Nagasawa,  S.,  Iwase,  H. and  Ogra,  Y. (2023): Identification of postmortem paliperidone metabolite in human blood by LC-Q-Orbitrap-MS. J. Anal. Toxicol., 47, 517-522.
•  Yamagishi,  Y.,  Nagasawa,  S.,  Iwase,  H. and  Ogra,  Y. (2024): Identification of post-mortem product of zolpidem degradation by hemoglobin via the Fenton reaction. J. Toxicol. Sci., 49, 261-268.