Metabolic profiling of lysophosphatidylcholines in chlorpromazine hydrochloride- and N -acetyl- p -amino-phenoltriptolide-induced liver injured rats based on ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry

Abstract

Chlorpromazine hydrochloride (CH) and N-acetyl-p-amino-phenoltriptolide (APAP) are typical acentral dopamine receptor antagonists and antipyretic analgesics in clinical applications, respectively. However, it has been reported that these 2 drugs could cause liver damage. Lysophosphatidylcholines (LPCs) have multiple physiological functions and are metabolized primarily in the liver, where it undergoes significant changes when the liver is damaged. In the study, 15 LPCs in the rat serum with CH- and APAP-induced liver injury were quantified based on ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry, and multivariate statistical analyses including principal component analysis (PCA) and orthogonal partial least squares discriminate analysis (OPLS-DA) were combined to understand CH- and APAP-induced liver injury from the perspective of LPC metabolic profiling. The quantitative results showed that there were significant changes in 10 LPCs and 5 LPCs after CH- and APAP-administration, separately. The results of PCA and OPLS-DA indicated that CH- and APAP-induced liver injury could be well distinguished by the LPC metabolic profiling, and 7 LPCs and 1 LPC biomarkers that could characterize CH- and APAP-induced liver damage in turn had been screened. This study will not only provide a new perspective for the clinical diagnosis of CH- and APAP-induced liver injury, but also offer a reference for further study of their hepatotoxicity mechanisms.

Keywords

Lysophosphatidylcholine liver injury metabolic profiling chlorpromazine hydrochloride ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry

Introduction

Chlorpromazine hydrochloride (CH) is a phenothiazine drug, which functions as a central dopamine receptor antagonist; it is widely used for treating psychosis.¹ N-acetyl-p-amino-phenol (APAP) is a widely used over-the-counter pain reliever and antipyretic available worldwide. However, both drugs can cause liver injury.^2,3 The mechanisms of CH-induced liver injury may be related to the interference of bile secretion and hypersensitivity reaction.⁴ The metabolites of CH can covalently bind to proteins in liver cells to form new neoantigens, inducing an immune response and mediating the apoptosis of liver cells through Fas.⁵ After APAP overdose, the production of N-acetyl-p-benzoquinone-imine (NAPQI), a toxic free radical metabolite, exceeded the detoxification ability of glutathione (GSH), resulting in the covalent binding of NAPQI with sulphur-containing groups in proteins resulting in protein adducts.⁶ At present, CH- and APAP-induced liver injury are clinically diagnosed using glutamatergic transaminase (ALT) and aspartate transaminase (AST) assays, but the specificity and sensitivity of such techniques are poor. Metabolomics-based techniques were first proposed by Nicholson et al. in 1999.⁷ Using high-throughput instruments, endogenous metabolites can be dynamically analysed to study the overall biological status and functional regulation of the organism. It has been widely used in disease diagnosis, mechanism exploration, and drug development.⁸ Using metabolomics to investigate the metabolite changes associated with CH- and APAP-induced liver injury will provide a basis for the mechanism exploration and early biomarker identification of such liver injury.

Lysophosphatidylcholines (LPCs) are bioactive lipids that regulate various cellular functions. Their general structure is shown in Figure 1. As the most important carrier of fatty acids in serum, LPCs can participate in many physiological and pathological processes through the G protein-coupled receptor p38 MAPK and NF-κB pathways; they also affect various biological functions mediated by Toll-like receptors, such as inducing the migration of lymphocytes and macrophages, increasing the production and release of pro-inflammatory cytokines, exacerbating oxidative stress, and promoting apoptosis.^9,10 Moreover, they contribute to various diseases, including cardiovascular disease,¹¹ Alzheimer’s disease,¹² atherosclerosis,¹³ and diabetes.¹⁴ The liver is the primary organ for LPC synthesis and catabolism. Lysophosphatidylcholines are a death effector of hepatocyte apoptosis, leading to apoptosis mediated by GSK/JNK and caspase activation in hepatocytes; their levels closely correlate with liver injury status and liver reserve function. LPCs are the most abundant class of phospholipids in serum. Considerable changes in serum LPC levels can occur due to liver disease and drug-induced liver injury (DILI). The serum levels of LPC(16:0), LPC(18:0), LPC(18:2), LPC(18:3), LPC(20:3), and LPC(20:5) in patients with liver cirrhosis caused by hepatitis B virus decreased significantly.¹⁵ The concentrations of LPC(16:0), LPC(18:0), and LPC(18:1) in the serum of mice with non-alcoholic steatohepatitis deviated from normal values.¹⁶ LPCs may be key biomarkers for liver disease and DILI.

Figure 1.

General formula of LPC structure. LPC(A:B), A: The number of carbon in “-R=O-”, B: The number of double bonds in “-R”.

Here, liquid chromatography-triple quadrupole tandem mass spectrometry (LC-QqQ-MS/MS) was used for quantitatively analysing LPCs in biological samples. QqQ-MS/MS cannot effectively exclude interferences in complex matrices. TOF-MRM acquisition mode in ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS/MS) has a target gain function to selectively increase the duty ratio of target fragments, improving the response and selectivity for determining LPC level in serum.¹⁷ In this study, based on the pharmacological experiments on rats with CH- and APAP-induced liver injury, the targeted metabolomics method based on TOF-MRM was used to analyse the changes in serum LPC metabolic profile after CH and APAP administration to explore the correlation between the hepatotoxicity of CH and APAP and metabolic profile of LPCs and to screen biomarkers characterizing CH- and APAP-induced liver injury, providing a new method for clinically diagnosing such liver injury.

Materials and methods

Chemicals and reagents

The reference standards of LPC(13:0) (internal standard, IS), LPC(14:0), LPC(16:0), LPC(17:0), LPC(18:1) and LPC(20:0) were purchased from Avanti Polar Lipids Inc. (Alabaster, AL, USA). Mass spectrometry grade-methanol and acetonitrile were obtained from Merck Inc. (Darmstadt, HE, Germany). Analytical reagent-chloroform and HPLC grade-formic acid were purchased from Xilong Technology Co., Ltd. (Shantou, China). Phosphate buffer saline (PBS) was supplied by Servicebio Technology Co., Ltd. (Wuhan, China). Bovine serum albumin (BSA) was provided by Energy Chemical Technology Co., Ltd. (Shanghai, China). CH and APAP were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Isoflurane was supplied by RWD Life Science CO., Ltd. (Shenzhen, China)

Animal treatment

32 Sprague-Dawley male rats (SPF grade, Certificate number: SCXK2020-0002) weighting 200 ± 20 g at 7 to 9-week-old were provided by Hunan Shrek Jingda Experimental Animal Technology Co. Ltd. (Changsha, China). Before starting the experiment, the animals were asked to acclimatize in ordinary cages for 7 days under normal laboratory conditions of 12 h-light/dark cycle, relative humidity of 60 ± 5% and room temperature of 24 ± 2°C, and with free access to commercial pellet diet and tap water. All the animals were handled in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The study protocols were approved by Animal Ethics Committee of Jiangxi Science and Technology Normal University. After one-week of adaptive feeding, the rats were left unfed overnight and divided randomly into four groups. The CH-treated group (n = 8) was administered by intraperitoneal injection with CH (90 mg·kg⁻¹, 25 mg·mL⁻¹ in normal saline), while CH vehicle group (n = 8) treated with an equal volume of the vehicle. The APAP-treated group (n = 8) were given orally with 180 mg·mL⁻¹APAP (1500 mg·kg⁻¹) dissolved in 0.5% sodium carboxymethyl cellulose (CMC-Na), while APAP vehicle group (n = 8) dosed with oral gavage of 0.5% CMC-Na. After CH and APAP treatment 24 h, blood samples were collected and then centrifuged for 5 min (4°C, 3500 × g) to obtain serum for biochemical indicators assay and LPC metabolic profile analysis. All the rats were sacrificed by lasting and high-density isoflurane, and then livers were subsequently collected for histopathological study.

Biochemical assays and histopathology examination

Serum biomedical indicators containing alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were estimated by automatic biochemical analyser (Chemray 240, Rayto, China) according to the according to the instructions of the commercial analyte-specific kit (Rongsheng Biotech, China). Hematoxylin and eosin (H&E) sections of liver were prepared for histopathology examination by optical microscope (Eclipse E100, Nikon, Japan).

Preparation of stock solutions

2 mg∙mL⁻¹ analyte stock solution was prepared with LPC(14:0), LPC(16:0), LPC(17:0), LPC(18:1) and LPC(20:0) standards dissolved in methanol-chloroform (1:1, v/v). The same method was utilized to produce 2 mg∙mL⁻¹ LPC(13:0) as the IS stock solution. All the stock solutions were stored at −20°C and diluted during use.

Sample preparation

Serum samples (20 μL) spiked with 20 μL IS (100 μg·mL⁻¹) were vortex-mixed with 160 μL methanol-chloroform (1:1, v/v) for 5 min. After centrifuging for 10 min (4°C, 12,000 × g), 40 μL supernatant was taken out and then diluted with 760 μL methanol-acetonitrile (3:1, v/v). 2 μL aliquots were injected into the UPLC-Q-TOF-MS/MS for analysis.

UPLC-Q-TOF-MS/MS analysis

Targeted LPCs analysis was accomplished with the UPLC-Q-TOF-MS/MS system (Xevo G2-XS QTOF, Waters, USA). Separation was performed on a ACQUITY UPLC HSS T3 column (100 × 2.1 mm, 1.8 μm) with the mobile phase (A) 0.1% formic acid in water, and the mobile phase (B) methanol. The gradient elution system was employed as follows: 0–1.0 min, 40% B; 1.0–2.0 min, 40%–85% B; 2.0–5.0 min, 85%–95% B; 5.0–7.0 min, 95% B. The flow rate was set at 0.3 mL·min⁻¹. LPCs were determined under MRM in positive ion mode, and the TOF-MRM parameters were shown in Table 1. Ion source temperature was 100°C. The capillary voltage and the cone voltage were set to 3 kV and 35 V, respectively. The desolvation gas and cone gas are nitrogen whose flow rates were 600 L·h⁻¹ and 50 L·h⁻¹ in turn.

Table 1.

TOF-MRM parameters of LPCs.

LPC	Retention time (min)	Molecular mass (Da)	Molecular formula	Qualitative ions (m/z)	Transition ions (m/z)
LPC(13:0)	5.30	453.2855	C₂₁H₄₄NO₇P	104.1079, 184.0735	454.2914 > 436.2815
LPC(14:0)	5.68	467.3012	C₂₂H₄₆NO₇P	450.3066, 184.0735	468.3073 > 104.1079
LPC(16:0)	6.57	495.3325	C₂₄H₅₀NO₇P	478.3279, 184.0735	496.3413 > 104.1079
LPC(16:1)	5.91	493.3168	C₂₄H₄₈NO₇P	476.3206, 184.0735	494.3228 > 104.1079
LPC(17:0)	7.06	509.3481	C₂₅H₅₂NO₇P	492.3469, 184.0735	510.3472 > 104.1079
LPC(17:1)	6.31	507.3325	C₂₅H₅₀NO₇P	490.3286, 184.0735	508.3403> 104.1079
LPC(18:0)	7.60	523.3638	C₂₆H₅₄NO₇P	506.3546, 184.0735	524.3723 > 104.1079
LPC(18:1)	6.83	521.3481	C₂₆H₅₂NO₇P	104.1079, 184.0735	522.3570 > 504.3464
LPC(18:2)	6.19	519.3352	C₂₆H₅₀NO₇P	104.1079, 184.0735	520.3412 > 502.3265
LPC(18:3)	6.02	517.3168	C₂₆H₄₈NO₇P	104.1079, 184.0735	518.3247 > 550.3107
LPC(20:0)	8.62	551.3951	C₃₀H₅₂NO₇P	534.3566, 104.1079	552.4047 > 184.0735
LPC(20:1)	7.83	549.3794	C₂₈H₅₆NO₇P	104.1079, 184.0735	550.3771 > 532.3798
LPC(20:3)	7.63	545.3481	C₂₈H₅₂NO₇P	104.1079, 184.0735	546.3555 > 528.3461
LPC(20:4)	6.23	543.3325	C₂₈H₅₀NO₇P	104.1079, 184.0735	544.3408 > 526.3350
LPC(20:5)	5.77	541.3168	C₂₈H₄₈NO₇P	524.3205, 184.0735	542.3263 > 104.1079
LPC(22:5)	6.55	569.3481	C₃₀H₅₂NO₇P	552.2512, 184.0735	570.2625 > 104.1079
LPC(22:6)	6.23	567.3325	C₃₀H₅₀NO₇P	550.3303, 184.0735	568.3397 > 104.1079

Method validation

Both quality control (QC) samples at three different concentrations (low, 0.050 μg/mL; medium, 0.500 μg/mL; high, 7.500 μg/mL) and 10 calibration samples (0.025, 0.050, 0.125, 0.250, 0.500, 1.000, 2.500, 5.000, 7.500 and 10.000 μg/mL) were prepared with simulated serum composing 1% BSA in PBS, in view of the endogenous level of LPCs in serum.¹⁸ The method was validated by testing parameters concerning linearity, precision, accuracy, extraction recovery and matrix effect with reference to U.S. FDA and EMEA guidance for bioanalytical method validation.^19,20 The calibration curves were established by plotting the peak area ratio (y) of LPC to the IS versus the nominal concentration (x) of LPC standards, calculated with weighted least square linear regression. The constructed calibration curves will be used for the relative quantitative analysis of the LPCs in the serum. Intra- and inter-day precision and accuracy were evaluated with QC samples and lower limit of quantification (LLOQ, 0.025 μg/mL) samples at the 3 consecutive days. Extraction recovery and matrix effect were examined with QC samples.

Data analysis

Concentration data was acquired and processed using the MassLynx 4.1 (Waters, Milford, MA, USA). Principal component analysis (PCA) and orthogonal partial least squares-discriminate analysis (OPLS-DA) were performed by SIMCA-P 14.1. Two-tailed student’ s t-test and Pearson’s correlation analysis were conducted through SPSS 25.0 (SPSS Inc., Chicago, IL, USA).

Results

Biochemical indicators and histopathological observations

The serum levels of ALT and AST were assayed at 24 h after CH and APAP administration. In the CH-treated group, the values of ALT and AST were 379.75 ± 66.24 U/L and 447.89 ± 124.01 U/L, respectively, which were markedly higher than the values of CH vehicle group, 226.54 ± 23.25 U/L (p < 0.001) and 282.11 ± 23.30 U/L (p < 0.001). Besides that, compared with the APAP vehicle group, the ALT and AST values in the APAP-treated group showed a significant increase, where the ALT and AST values were 144.77 ± 24.36 U/L and 192.57 ± 47.94 U/L in APAP-treated group in turn, while the those of APAP vehicle group were 108.47 ± 12.91 U/L (p < 0.001) and 133.58 ± 27.08 U/L (p < 0.01), respectively.

As shown in Figure 2, histological evaluation showed that all the vehicle groups displayed normal liver histology whereas CH- and APAP-treated groups showed several pathological damages that are irregular arrangement, membrane rupture, nuclear shrinkage for liver cells. The changed levels of biochemical indicators and histopathological alterations all indicated that the rats suffered from liver injury 24 h post-CH and APAP treatment.

Figure 2.

Histopathology of liver tissue 24 h after CH and APAP administration. Representative slides were shown at 400× magnifications. A/B, CH- and APAP-vehicle group; C/D, CH- and APAP-treated group.

Method validation

The TOF-MRM chromatograms of LPCs obtained from standards were shown in Figure 3. The 5 LPC calibration curves had good linearity in the range of 0.025–10.000 μg·mL⁻¹ with linear regression coefficients (r²) greater than 0.99, and the relevant parameters were listed in Table 2. Accuracy was expressed as the ratios (calculated concentration/nominal concentration). Precision was displayed as the relative standard deviation (RSD%). The accuracies were in the range of 88.80%–104.33% and precisions were less than or equal to 14.31% for LLOQ. The accuracy and precision results of QCs were summarized in Table 3. In QCs, the accuracies were between 80.21% and 113.35% with precisions less than or equal to 9.83%. The results of recovery and matrix at three different levels were displayed in Table 3. The recoveries, except for 77.03% in medium QC sample of LPC(20:0), were in the range of 85.53%–104.41% with RSD% less than 8.35%. The matrix effects were between 87.65% to 115.4% with RSD% less than 9.23%. Taken together, the method proved to be reliable and applicable for LPCs analysis in serum.

Figure 3.

TOF-MRM chromatograms of LPC standards. 5 analyte standards including LPC(14:0), LPC(16:0), LPC(17:0), LPC(18:1) and LPC(20:0), and IS standard LPC(13:0).

Table 2.

The parameters of 5 LPC calibration curves.

LPC	Slope	Intercept	r ²	LLOQ (μg·mL⁻¹)	Liner range (μg·mL⁻¹)
LPC(14:0)	0.0000704012	0.000768212	0.994784	0.025	0.025–10.000
LPC(16:0)	0.0003531050	0.001913920	0.992299	0.025	0.025–10.000
LPC(17:0)	0.0004816090	−0.001680780	0.993482	0.025	0.025–10.000
LPC(18:1)	0.0028875600	0.365455000	0.999368	0.025	0.025–10.000
LPC(20:0)	0.0032786300	−0.002950660	0.995002	0.025	0.025–10.000

Linear regression model: y = kx + b, x: LPC nominal concentration, y: the peak area ratio of LPC to the IS, k: slope, b: intercept; weighting factor: 1/x²; r²: regression coefficient.

Table 3.

The results of accuracy, precision, extraction recovery and matrix effect.

LPC	Concentration (μg/mL)	Intra-day accuracy (%)	Intra-day precision (RSD %)	Inter-day accuracy (%)	Inter-day precision (RSD %)	Extraction recovery (%)	Matrix effect (%)
LPC(14:0)	0.05	110.37	5.87	95.51	8.53	86.69 ± 6.79	110.47 ± 9.22
	0.50	94.20	6.38	89.69	6.91	89.77 ± 2.97	94.30 ± 7.94
	7.50	96.26	8.80	92.82	9.39	94.68 ± 8.16	102.90 ± 4.40
LPC(16:0)	0.05	99.13	6.79	92.93	9.44	87.66 ± 7.80	102.50 ± 6.34
	0.50	99.42	3.03	95.59	6.05	95.93 ± 4.62	95.08 ± 5.21
	7.50	88.46	9.32	88.77	9.83	99.34 ± 5.85	102.44 ± 2.03
LPC(17:0)	0.05	101.70	4.77	96.35	6.85	99.00 ± 4.4	107.83 ± 6.43
	0.50	101.59	4.70	94.57	7.56	104.41 ± 4.64	98.87 ± 5.03
	7.50	81.23	2.99	80.21	5.35	98.10 ± 3.63	93.63 ± 4.38
LPC(18:1)	0.05	97.53	2.55	100.01	5.46	98.05 ± 8.35	115.40 ± 7.35
	0.50	99.27	3.72	101.74	5.69	99.41 ± 3.41	105.02 ± 5.10
	7.50	101.08	2.78	94.12	8.81	97.75 ± 4.98	102.31 ± 2.10
LPC(20:0)	0.05	102.97	5.16	91.00	8.49	85.53 ± 8.13	105.75 ± 7.66
	0.50	87.06	2.16	87.31	4.98	77.03 ± 3.53	87.65 ± 1.02
	7.50	111.35	4.44	87.56	6.10	87.07 ± 2.61	94.62 ± 2.37

Accuracy was expressed as calculated concentration/nominal concentration, and precision was represented as RSD%; extraction recovery and matrix effect denoted as mean ± SD.

UPLC-Q-TOF-MS/MS analysis of LPCs in rat serum

The validated analysis method was applied to determine the concentration of LPCs in rat serum. The TOF-MRM chromatograms of LPCs in rat serum were shown in Figure 4. LPCs could be separated well without interference from other endogenous metabolites in serum. Currently, due to a shortage of commercial LPC standards, serum LPCs without corresponding standards were quantified relatively using existing LPC curves with similar carbon atomic numbers. (Table 4).

Figure 4.

TOF-MRM chromatograms of LPCs from rat serum. 15 LPCs in the rat serum were well separated and detected by the method.

Table 4.

The concentration results of serum LPCs in CH vehicle rats and CH -treated rats, as well as APAP vehicle rats and APAP-treated rats ( $\bar{x}$ ± SD, n = 8).

Analytes	Calibration curve	CH vehicle (μg∙mL⁻¹)	CH-treated (μg∙mL-1)	APAP vehicle (μg∙mL-1)	APAP-treated (μg∙mL-1)
LPC(14:0)	LPC(14:0)	165.04 ± 24.62	89.48 ± 23.74***	222.12 ± 31.64	156.18 ± 25.06^***
LPC(16:0)	LPC(16:0)	4150.41 ± 581.23	3102.81 ± 214.33***	3732.74 ± 1302.75	3765.87 ± 507.44
LPC(16:1)	LPC(16:0)	50.55 ± 7.90	17.33 ± 4.95***	65.17 ± 39.33	50.40 ± 8.71
LPC(17:0)	LPC(17:0)	54.78 ± 7.36	39.61 ± 4.01^***	71.10 ± 27.69	56.47 ± 7.43
LPC(17:1)	LPC(17:0)	8.01 ± 0.93	5.60 ± 0.80^***	10.50 ± 3.43	7.02 ± 0.66^*
LPC(18:0)	LPC(18:1)	182.6 ± 18.05	163.07 ± 23.69	315.60 ± 217.46	259.84 ± 135.02
LPC(18:1)	LPC(18:1)	127.71 ± 14.07	65.80 ± 8.89^***	139.15 ± 74.06	111.43 ± 16.27
LPC(18:2)	LPC(18:1)	288.62 ± 30.62	110.44 ± 19.28^***	238.63 ± 39.69	241.14 ± 29.03
LPC(18:3)	LPC(18:1)	124.08 ± 20.25	120.92 ± 27.74	112.57 ± 70.68	105.54 ± 49.27
LPC(20:1)	LPC(20:0)	49.16 ± 2.32	31.84 ± 4.85^***	23.82 ± 6.00	30.06 ± 6.21
LPC(20:3)	LPC(20:0)	37.41 ± 6.83	34.42 ± 8.19	3.86 ± 0.72	5.03 ± 1.58
LPC(20:4)	LPC(20:0)	186.68 ± 17.86	121.45 ± 6.77***	172.13 ± 58.52	159.18 ± 24.14
LPC(20:5)	LPC(20:0)	1.93 ± 0.36	1.63 ± 0.14*	1.65 ± 0.29	59.67 ± 7.10^***
LPC(22:5)	LPC(20:0)	1.35 ± 0.21	1.20 ± 0.15	2.90 ± 0.69	2.27 ± 0.24^*
LPC(22:6)	LPC(20:0)	125.43 ± 18.41	110.90 ± 17.48	163.15 ± 35.24	108.62 ± 28.74^**

Concentration data were represented as mean ± SD, ^*p < 0.05, ^**p < 0.01 and ^***p < 0.001 versus corresponding vehicle groups.

The concentrations of 15 LPCs were simultaneously quantified in the rat serum. The concentration results were presented in Table 4, and the concentration ratio results of the administered group over the vehicle group were shown in Figure 5. Most LPC levels showed a downward trend, after APAP and CH treatment. 10 LPCs were significantly lower in the CH-treated group than the CH vehicle group, with a 3-fold reduction in LPC(16:1). 5 LPCs showed remarkable changes between the APAP-treated group and the APAP vehicle group, where LPC(20:5) was marvellous with a 36-fold increase in the APAP treatment group over the corresponding vehicle group.

Figure 5.

The ratio of serum LPCs concentration in the CH-treated group (a) and APAP-treated group (b) compared to their corresponding vehicle group. Statistical tests were performed by using the two-tailed Student’s t test (*p < 0.050, **p < 0.010, ***p < 0.001).

Multivariate statistics analysis

To further investigate the role of metabolic profiling of LPC in the discrimination of hepatotoxicity, multivariate statistics analyses including PCA, OPLS-DA and correlation analysis were performed.

After data standardization, 15 variables were analysed by PCA with two principal components (PC1 and PC2). The two explained 78.3% and 93.2% of the systematic variation in CH and APAP, respectively. As shown in the two-dimension score plot (Figure 6), the APAP-treated group and CH-treated group discriminated from the corresponding vehicle groups, suggesting that the APAP- and CH-induced liver injury was closely related to the serum LPC metabolic profile. The two-dimensional score plot of the OPLS-DA model (Figure 7) displayed that the CH-treated and APAP-treated groups were clearly distinguished and both deviated from the corresponding vehicle groups, indicating that the types of CH- and APAP-induced liver injury could be discriminated through the LPC metabolic profile.

Figure 6.

The PCA score plot of serum LPCs in CH vehicle rats and CH-treated rats (a), as well as APAP vehicle rats and APAP-treated rats (b). Two-dimensional visualization of the two PC scores for samples. The score plots indicated a clear separation between the CH/APAP-treated and vehicle groups.

Figure 7.

The OPLS-DA score plot of serum LPCs from vehicle groups and APAP- and CH- treated rat groups. The score plots showed that the two vehicle groups gathered, and the CH and APAP-treated groups were obviously separated and all deviated from the vehicle groups.

The OPLS-DA models of CH and APAP were constructed, separately. The R²X, R²Y and Q²Y in the former were 73%, 97.6% and 94.9%, respectively, and those of the latter were 86.5%, 98.1% and 95.8% in turn, which indicated that the OPLS-DA models had good explanatory ability and high prediction accuracy for the original data.²¹ Furthermore, the OPLS-DA models for CH (Figure 8(a)) and APAP (Figure 9(a)) were validated by permutation (200 times). The values of R² and Q² in the two OPLS-DA models were both lower than the original values and the intercept values of the Q² regression lines were less than zero, which suggested that both OPLS-DA models were not overfitted and could be used for discovering potential biomarkers.²²

Figure 8.

Analysis of potential biomarkers of CH-treated groups using OPLS-DA. The OPLS-DA model (R2X = 73%, R2Y = 97.6%, and Q2 = 94.9%) was constructed based on the metabolic profile of LPCs from vehicle and CH-treated groups. (a) The result of permutation test (n = 200) indicated the original OPLS-DA model was valid. (b) VIP plot confirmed the significance of these potential biomarkers in classification (VIP > 1.0). (c) The OPLS loading plot showed the correlation between individual LPC and CH-induced liver injury.

Figure 9.

Analysis of potential biomarkers of APAP-treated groups using OPLS-DA. The OPLS-DA model (R2X = 86.5%, R2Y = 98.1%, and Q2 = 95.8%) was constructed based on the metabolic profile of LPCs from vehicle and APAP-treated groups. (a) The result of permutation test (n = 200) indicated the original OPLS-DA model was valid. (b) VIP plot confirmed the significance of these potential biomarkers in classification (VIP > 1.0). (c) The OPLS loading plot showed the correlation between individual LPC and APAP-induced liver injury.

The variable importance in projection (VIP) plots and loading plots based on the OPLS-DA model for CH were shown in Figures 8(b) and (c), respectively, and those for APAP were presented in Figures 9(b) and (c) in turn. The VIP values of LPC(18:2), LPC(14:0), LPC(20:4), LPC(18:1), LPC(16:1), LPC(17:1) and LPC(20:1) in the CH group as well as LPC(20:5) in APAP group were all greater than 1.0, and these LPCs underwent significant changes following CH or APAP administration, so it could be considered that they might be served as potential biomarkers to characterize CH or APAP-induced liver injury. Moreover, the correlation analyses were conducted using Pearson’s correlation analysis to demonstrate the relationship between the LPC biomarkers and the activities of ALT and AST (Table 5). The results indicated that the 2 biochemical indexes were positively correlated with LPC(20:5) in APAP, while negatively correlated with all the other biomarkers in CH.

Table 5.

The results of Pearson’s correlation analysis of serum biomarkers and biochemical indexes in rats with liver injury induced by CH and APAP.

Liver injury	LPC	ALT		AST
Liver injury	LPC	r	p	r	p
CH	LPC(18:2)	−0.703	0.002	−0.580	0.019
	LPC(14:0)	−0.742	0.001	−0.513	0.042
	LPC(20:4)	−0.682	0.004	−0.520	0.039
	LPC(18:1)	−0.685	0.003	−0.564	0.023
	LPC(16:1)	−0.668	0.005	−0.444	0.085
	LPC(17:1)	−0.492	0.053	−0.349	0.186
	LPC(20:1)	−0.740	0.001	−0.429	0.097
APAP	LPC(20:5)	0.587	0.007	0.527	0.017

r, correlation coefficient. |r| > 0.5 and p < 0.05 indicated good correlation between two variables.

Discussion

The liver, the main site of biotransformation and detoxification in vivo, is one of the most vulnerable organs to drug-induced damage. Drug-induced liver injury (DILI) is a common clinical adverse reaction, which is the main developmental obstacle (termination, approval denial, use restriction, and withdrawal from the market) for multiple drugs.²³ Traditional methods of hepatotoxicity studies usually rely on histopathological findings and biochemical indicators to detect toxicity, and their changes mostly occur after substantial damage to the liver, with low sensitivity and specificity.²⁴ For a long time, researchers mainly focused on various genes or proteins related to liver injury, such as miR-122, miR-155, glutathione-S-transferase, and paraoxonase, to study DILI.²⁵ However, the development of DILI involves many physiological and biochemical processes, including mitochondrial damage, oxidative stress, and immune response.^26–28 In addition to genes, proteins, and other macromolecules, the levels of endogenous small molecules in the body are also altered.

Metabolomics has been developed for over 20 years, improving various fields of medicine and life sciences, especially in disease diagnosis, individualized treatment, and drug toxicity evaluation, showing good prospects for application.²⁹ In DILI studies, metabolomics provides rapid screening of toxicity-related biomarkers by comparing the metabolic profiles of liver tissue extracts, plasma, serum, urine, and other biological samples before and after drug intervention, providing comprehensive information for evaluating drug hepatotoxicity and elucidating mechanisms.³⁰

APAP is currently present in over 600 different prescription and over-the-counter medications.³¹ Due to the widespread use of APAP-containing drugs, patients can intentionally or unintentionally overdose on them, and overdoses of APAP can lead to varying degrees of liver damage and even acute liver failure.³² CH is a representative classic antipsychotic drug widely used in primary hospitals because of its low price; however, it has many adverse reactions, especially liver damage.³³

In this study, the results of the pharmacological experiment showed that the serum ALT and AST levels and liver pathology after CH/APAP administration were significantly different from those in the normal group 24 h, indicating that an overdose of APAP and CH may cause severe hepatotoxicity at 24 h. LPCs are produced via lecithin cholesterol acyltransferase from phosphatidylcholines (PCs). In contrast, under the action of metabolic enzymes, such as lysophosphatidylcholine acyltransferase, lysophospholipase A1, and intra-nucleotide pyrophosphatase/phosphodiesterase, LPCs can be catabolized and converted into PCs. The LPC metabolism occurs mainly in the liver, and liver damage is reflected as fluctuations in serum LPC levels. Therefore, it is crucial to study serum LPC level for diagnosing and preventing CH- and APAP-induced liver injury. In this study, TOF-MRM combined with a multivariate statistical analysis method was used to analyse the serum LPC metabolism profile of rats with APAP- and CH-induced liver injury. This study quantitatively showed that the overall level of LPCs decreased after CH administration, among which 10 LPC levels were significantly lower than the normal group; 5 LPCs were significantly different from the normal group after APAP administration. Multivariate statistical analysis revealed that the metabolic profile of serum LPCs was closely related to CH and APAP hepatotoxicity, identifying biomarkers that can characterize CH and APAP-induced liver injury. Relevant CH biomarkers were LPC(18:2), LPC(14:0), LPC(20:4), LPC(18:1), LPC(16:1), LPC(17:1), and LPC(20:1); the APAP biomarker was LPC(20:1). These results suggest that it is feasible to evaluate the hepatotoxicity of APAP and CH from the metabolic profile of serum LPC.

LPCAT, a key enzyme in the Lands cycle, is located in the endoplasmic reticulum and involved in the synthesis of LPCs.³⁴ PCs are important components of biofilm and can protect cells and their organelles from oxidative stress, lipotoxicity, and endoplasmic reticulum stress.³⁵ After liver injury, including DILI, alcoholic liver injury, and non-alcoholic steatohepatitis, the PC demand increases, and LPC content decreases, because the body’s automatic defence function induces the activation of large amounts of LPCAT to process LPC and synthesize PC in liver lesions.^36,37 This may be the mechanism by which LPC levels were reduced after CH and APAP administration in this study. In addition, the increased demand for PCs may be used to produce large amounts of arachidonic acid (AA) regulated by phospholipase A2 (PLA2).³⁸ AA, a type of ω-6 polyunsaturated fatty acid, is a synthetic precursor of inflammatory mediators such as prostaglandin, leukotriene, and thrombotin.³⁹ The main metabolic pathways of AA in the liver involve cytochrome P450 (CYP450) ω-hydroxylation and the CYP450 epoxygenase; CYP4A and CYP4F catalyse AA to synthesize 20-hydroxyeicosatetraenoic acid (20-HETE), and AA generates epoxyeicosatrienoic acids (EETs) with the help of CYP2C and CYP2J family members. When the liver is damaged by drugs, the expression levels of CYP4A and CYP4F increase,^40,41 while the expression levels of CYP2C and CYP2J decrease.^42,43 AA can activate the inflammatory-cascade-associated metabolic pathway and induce the inflammatory injury metabolism of the liver to produce several inflammatory factors.⁴⁴ The potential functional mechanism of CH- and APAP-induced liver injury is summarized in Figure 10. Oxidative stress, inflammatory reaction, and lipid metabolism disorder may mediate CH- and APAP-induced liver injury.

Figure 10.

The potential functional mechanism of CH- and APAP-induced liver injury. The up-regulated enzymes or metabolites were labeled with red and down-regulated with blue, respectively. Oxidative stress, inflammatory reaction and lipid metabolism disorder may be the mechanisms of CH- and APAP-induced liver injury.

In the liver, LPC up-regulates genes involved in cholesterol synthesis and down-regulates genes involved in hepatic fatty acid oxidation.⁴⁵ High levels of LPCs disrupt mitochondrial integrity and increase cytochrome C release in hepatocytes. LPCs have multiple biological functions in organisms, such as pro-inflammatory functions, oxidative stress, and the induction of apoptosis.⁴⁶ The pro-inflammatory role of LPC in many diseases has been widely reported.⁴⁷ LPCs can attract inflammatory cells to increase the release of inflammatory factors such as IL-1β, IL-8, IFN-γ, IL-6, and IL-5 by increasing the production of chemokines.⁴⁸ Caspase-1, a member of the pro-inflammatory caspase family, activates IL-1β and IL-18. LPCs promote caspase-1 release in a Na⁺-dependent fashion, regulating inflammation.⁴⁹ Moreover, LPCs, especially LPC(16:0), LPC(18:1), and LPC(20:4), can induce AP-1 binding to the C/EBP of IL-6 and IL-8 by activating the ERK1/2/C-JUN/PKC promoter region to increase the release of IL-6 and IL-8.⁵⁰ The mechanism of LPC-mediated inflammatory response is also associated with NF-κB, Notch, and PI3K/Akt signalling pathways.^51–53 LPCs participate in many signal transduction effects and play a key role in liver oxidative function. As major components of oxidatively damaged low-density lipoproteins, LPCs can cause reactive oxygen species (ROS) production by activating MAPK and altering cytoplasmic Ca²⁺ levels.^54,55 In hepatocytes, LPCs can activate ERK1/2 signalling by generating mitochondrial ROS, leading to decreased intracellular SOD and GSH-PX contents and increased MDA and ROS contents.⁵⁶ Many mechanisms are involved in the induction of apoptosis by LPC, the main ones being the activation of caspase, inward calcium flow, and release of cytochrome C.⁵⁷ In addition, LPCs can also increase FasL expression level by activating the NF-κB signalling pathway, thereby inducing apoptosis.⁵⁸

In the study, metabolomics based on TOF-MRM were used to study the hepatotoxicity of APAP and CH from the perspective of serum LPC metabolic profile. These results indicated that the LPC metabolic profile was closely related to CH- and APAP-induced hepatotoxicity and provided biomarkers that can characterize CH- and APAP-induced liver injury. Thus, serum LPC profile has the potential to become a new biomarker for clinically diagnosing APAP- and CH-induced liver injury.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (NSFC No, 81660692 and 81960732).

ORCID iD

Yin-Hua Xiong

References

Boyd-Kimball

Gonczy

Lewis

, et al. Classics in chemical neuroscience: chlorpromazine. ACS Chem Neurosci 2019; 10: 79–88.

James

McGill

Roberts

, et al. Advances in biomarker development in acetaminophen toxicity. Adv Clin Chem 2020; 98: 35–50.

Xiong

Yang

, et al. Gas chromatography-mass spectrometry-based profiling of serum fatty acids in acetaminophen-induced liver injured rats. J Appl Toxicol 2014; 34: 149–157.

Yang

, et al. Metabolic profiling of fifteen amino acids in serum of chemical-induced liver injured rats by hydrophilic interaction liquid chromatography coupled with tandem mass spectrometry. Metabolomics 2012; 8: 475–483.

Andrade

Lucena

Alonso

, et al. HLA class II genotype influences the type of liver injury in drug-induced idiosyncratic liver disease. Hepatology 2004; 39: 1603–1612.

Duan

Ramachandran

Akakpo

, et al. Role of extracellular vesicles in release of protein adducts after acetaminophen-induced liver injury in mice and humans. Toxicol Lett 2019; 301: 125–132.

Nicholson

Lindon

Holmes

, et al. understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data. Xenobiotica 1999; 29: 1181–1189.

Cuperlovic-Culf

Culf

. Applied metabolomics in drug discovery. Expert Opin Drug Discov 2016; 11: 759–770.

Carneiro

Iaciura

Nohara

, et al. Lysophosphatidylcholine triggers TLR2- and TLR4-mediated signaling pathways but counteracts LPS-induced NO synthesis in peritoneal macrophages by inhibiting NF-κB translocation and MAPK/ERK phosphorylation. PLoS One 2013; 8: e76233.

10.

Magalhães

Almeida

Atella

, et al. Schistosomal‐derived lysophosphatidylcholine are involved in eosinophil activation and recruitment through Toll‐like receptor-2-dependent mechanisms. J Infect Dis 2010; 202: 1369–1379.

11.

Koenen

. Lysophosphatidylcholine in platelet microvesicles: the grease for cardiovascular disease. Thromb Haemost 2019; 119: 1202–1204.

12.

Mulder

Wahlund

Teerlink

, et al. Decreased lysophosphatidylcholine/phosphatidylcholine ratio in cerebrospinal fluid in Alzheimer's disease. J Neural Transm (Vienna) 2003; 110: 949–955.

13.

Tanaka

Zaima

Sasaki

, et al. Lysophosphatidylcholine acyltransferase-3 expression is associated with atherosclerosis progression. J Vasc Res 2017; 54: 200–208.

14.

Diamanti

Cavalli

Pan

, et al. Intra- and inter-individual metabolic profiling highlights carnitine and lysophosphatidylcholine pathways as key molecular defects in type 2 diabetes. Sci Rep 2019; 9: 9653.

15.

Lian

Liu

Hao

, et al. A serum metabonomic study on the difference between alcohol- and HBV-induced liver cirrhosis by ultraperformance liquid chromatography coupled to mass spectrometry plus quadrupole time-of-flight mass spectrometry. Chin Med J (Engl) 2011; 124: 1367–1373.

16.

Tanaka

Matsubara

Krausz

, et al. Disruption of phospholipid and bile acid homeostasis in mice with nonalcoholic steatohepatitis. Hepatology 2012; 56: 118–129.

17.

Vidova

Spacil

. A review on mass spectrometry-based quantitative proteomics: targeted and data independent acquisition. Anal Chim Acta 2017; 964: 7–23.

18.

Zhang

Liu

, et al. A novel strategy for targeted lipidomics based on LC-tandem-MS parameters prediction, quantification, and multiple statistical data mining: evaluation of lysophosphatidylcholines as potential cancer biomarkers. Anal Chem 2019; 91: 3389–3396.

19.

EMEA . “Guideline on bioanalytical method validation”. European Medicines Agency, Committee for Medicinal Products for Human Use EMEA/CHMP/EWP/192217/2009. 2011.

20.

U.S. Food and Drug Administration . guidance for industry: bioanalytical method validation, May 2018.

21.

Triba

Amathieu

, et al. PLS/OPLS models in metabolomics: the impact of permutation of dataset rows on the K-fold cross-validation quality parameters. Mol Biosyst 2015; 11: 13–19.

22.

Wolahan

Hirt

Glenn

. Translational metabolomics of head injury: exploring dysfunctional cerebral metabolism with ex vivo NMR spectroscopy-based metabolite quantification. In: Kobeissy

(ed) Brain neurotrauma: molecular, neuropsychological, and rehabilitation aspects. Boca Raton, FL: CRC Press/Taylor & Francis, 2015.

23.

Garcia-Cortes

Robles-Diaz

Stephens

, et al. Drug induced liver injury: an update. Arch Toxicol 2020; 94: 3381–3407.

24.

Hassan

Fontana

. The diagnosis and management of idiosyncratic drug-induced liver injury. Liver Int 2019; 39: 31–41.

25.

McGill

Jaeschke

. Biomarkers of drug-induced liver injury. Adv Pharmacol 2019; 85: 221–239.

26.

Fromenty

. Alteration of mitochondrial DNA homeostasis in drug-induced liver injury. Food Chem Toxicol 2020; 135: 110916.

27.

Villanueva-Paz

Morán

López-Alcántara

, et al. Oxidative stress in drug-induced liver injury (DILI): from mechanisms to biomarkers for use in clinical practice. Antioxidants (Basel) 2021; 10: 390.

28.

Girish

Sanjay

. Role of immune dysfunction in drug induced liver injury. World J Hepatol 2021; 13: 1677–1687.

29.

Quintás

Martínez-Sena

Conde

, et al. Metabolomic analysis to discriminate drug-induced liver injury (DILI) phenotypes. Arch Toxicol 2021; 95: 3049–3062.

30.

Araújo

Carvalho

, et al. Metabolomic approaches in the discovery of potential urinary biomarkers of drug-induced liver injury (DILI). Crit Rev Toxicol 2017; 47(8): 633–649.

31.

Akakpo

Ramachandran

Jaeschke

. Novel strategies for the treatment of acetaminophen hepatotoxicity. Expert Opin Drug Metab Toxicol 2020; 16(11): 1039–1050.

32.

Jaeschke

Akakpo

Umbaugh

. Novel therapeutic approaches against acetaminophen-induced liver injury and acute liver failure. Toxicol Sci 2020; 174(2): 159–167.

33.

Gandhi

Guo

Shah

, et al. Chlorpromazine-induced hepatotoxicity during inflammation is mediated by TIRAP-dependent signaling pathway in mice. Toxicol Appl Pharmacol 2013; 266: 430–438.

34.

, et al. LPCAT1 overexpression promotes the progression of hepatocellular carcinoma. Cancer Cell Int 2021; 21: 442.

35.

Ishizuka

Kon

Lee-Okada

, et al. Aging exacerbates high-fat diet-induced steatohepatitis through alteration in hepatic lipid metabolism in mice. J Gastroenterol Hepatol 2020; 35: 1437–1448.

36.

Luo

Cao

Yang

, et al. PNPLA3 rs139051 is associated with phospholipid metabolite profile and hepatic inflammation in nonalcoholic fatty liver disease. World J Clin Cases 2018; 6: 355–364.

37.

Zhao

Chen

Bonacci

, et al. Identification and characterization of a major liver lysophosphatidylcholine acyltransferase. J Biol Chem 2008; 283: 8258–8265.

38.

Zou

Wei

Zhang

, et al. Different influences of trans fatty acids on the phospholipase A2 and arachidonic acid metabolic pathway in hepatocytes. J Agric Food Chem 2021; 69: 4120–4133.

39.

Xue

Deng

Zhang

, et al. Gigantol ameliorates CCl(4)-induced liver injury via preventing activation of JNK/cPLA2/12-LOX inflammatory pathway. Sci Rep 2020; 10: 22265.

40.

Feng

. Mechanism-based inhibition of CYP450: an indicator of drug-induced hepatotoxicity. Curr Drug Metab 2013; 14: 921–945.

41.

Uehara

Murayama

Nakanishi

, et al. Immunochemical quantification of cynomolgus CYP2J2, CYP4A and CYP4F enzymes in liver and small intestine. Xenobiotica 2015; 45: 124–130.

42.

Graves

Bradbury

Gruzdev

, et al. Expression of Cyp2c/Cyp2j subfamily members and oxylipin levels during LPS-induced inflammation and resolution in mice. Faseb J 2019; 33: 14784–14797.

43.

Tolosa

Jiménez

Pérez

, et al. Customised in vitro model to detect human metabolism-dependent idiosyncratic drug-induced liver injury. Arch Toxicol 2018; 92: 383–399.

44.

Konishi

Lentsch

. Changes in arachidonic acid metabolism during liver ischemia triggers induction of inflammatory injury. Hepatology 2018; 68: 1642–1643.

45.

Hollie

Cash

Matlib

, et al. Micromolar changes in lysophosphatidylcholine concentration cause minor effects on mitochondrial permeability but major alterations in function. Biochim Biophys Acta 2014; 1841(6): 888–895. DOI: 10.1016/j.bbalip.2013.11.013

46.

Law

Chan

Marathe

, et al. An updated review of lysophosphatidylcholine metabolism in human diseases. Int J Mol Sci 2019; 20: 1149.

47.

Liu

Zhu

Chen

, et al. The mechanisms of lysophosphatidylcholine in the development of diseases. Life Sci 2020; 247: 117443.

48.

Bach

Perrin-Cocon

Gerossier

, et al. Single lysophosphatidylcholine components exhibit adjuvant activities in vitro and in vivo. Clin Vaccine Immunol 2010; 17: 429–438.

49.

Schilling

Eder

. Sodium dependence of lysophosphatidylcholine-induced caspase-1 activity and reactive oxygen species generation. Immunobiology 2011; 216: 118–125.

50.

Takabe

Kanai

Chairoungdua

, et al. Lysophosphatidylcholine enhances cytokine production of endothelial cells via induction of L-type amino acid transporter 1 and cell surface antigen 4F2. Arterioscler Thromb Vasc Biol 2004; 24: 1640–1645.

51.

Song

Gao

, et al. Activation of liver X receptor attenuates lysophosphatidylcholine-induced IL-8 expression in endothelial cells via the NF-κB pathway and SUMOylation. J Cell Mol Med 2016; 20: 2249–2258.

52.

Chang

Chen

Liou

, et al. 7-Ketocholesterol induces ATM/ATR, Chk1/Chk2, PI3K/Akt signalings, cytotoxicity and IL-8 production in endothelial cells. Oncotarget 2016; 7: 74473–74483.

53.

Deatherage

Kim

, et al. Notch transmembrane domain: secondary structure and topology. Biochemistry 2015; 54: 3565–3568.

54.

Itabe

Sawada

Makiyama

, et al. Structure and dynamics of oxidized lipoproteins in vivo: roles of high-density lipoprotein. Biomedicines 2021; 9: 655.

55.

Silva-Neto

Lopes

Atella

. Here, There, and everywhere: the ubiquitous distribution of the immunosignaling molecule lysophosphatidylcholine and its role on chagas disease. Front Immunol 2016; 7: 62.

56.

Kakisaka

Cazanave

Fingas

, et al. Mechanisms of lysophosphatidylcholine-induced hepatocyte lipoapoptosis. Am J Physiol Gastrointest Liver Physiol 2012; 302: G77–G84.

57.

Song

Liu

, et al. Luteolin inhibits lysophosphatidylcholine-induced apoptosis in endothelial cells by a calcium/mitocondrion/caspases-dependent pathway. Planta Med 2010; 76(5): 433–438.

58.

Sun

Lee

Kim

, et al. Lysophosphatidylcholine-induced apoptosis in H19-7 hippocampal progenitor cells is enhanced by the upregulation of Fas Ligand. Biochim Biophys Acta 2009; 1791: 61–68.