Lysophosphatidylethanolamine Degradation Associated with Upregulation of Pnpla6/7 in a Murine Model of Metabolic Dysfunction-Associated Steatohepatitis

TL;DR

This study explores how LPE levels drop in a mouse model of fatty liver disease, linking it to increased Pnpla6/7 activity and liver inflammation.

Contribution

The study identifies Pnpla6/7-driven LPE catabolism as a novel mechanism contributing to LPE depletion in MASH.

Findings

HFCC/CDX mice showed reduced LPE levels in plasma and liver, correlating with liver inflammation.

LPE degradation was linked to upregulated Pnpla6/7 in the liver of MASH mice.

LPE depletion was associated with cholesterol and oxidized triglyceride accumulation in the liver.

Abstract

Metabolic dysfunction-associated steatohepatitis (MASH) is a form of fatty liver disease characterized by fat accumulation, hepatic inflammation, and fibrosis. Lysophosphatidylethanolamine (LPE), a partially deacylated product of phosphatidylethanolamine, plays significant roles in anti-inflammatory responses and mitochondrial homeostasis. Although serum LPE levels are reduced in patients with MASH, the underlying mechanisms remain unclear. In this study, we investigated LPE metabolism using liquid chromatography–tandem mass spectrometry and protein expressions in MASH mice. Male C57BL/6J mice were fed a high-fat, high-cholesterol, and cholic acid diet, along with 2% hydroxypropyl-β-cyclodextrin in drinking water (HFCC/CDX) for three weeks to induce MASH. LPE was primarily distributed in the liver and kidneys, with lower levels in the white adipose tissue. HFCC/CDX mice exhibited accumulation of cholesterols and oxidized triglycerides, accompanied by inflammation and fibrosis in the liver. In the plasma and liver of HFCC/CDX mice, most LPE species were decreased and showed negative correlations with hepatic inflammation, with the exception of LPE 18:1. Mechanistically, enhanced degradation of LPE to glycerophosphorylethanolamine was associated with upregulation of Pnpla6/7 in the liver. These findings suggest that Pnpla6/7-driven LPE catabolism is contributing to LPE depletion. This study provides a new perspective to understand the association between disrupted phospholipid metabolism and MASH pathogenesis.

Figures

Full text

1. Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a metabolic liver disorder characterized by the accumulation of ectopic fat within hepatocytes. MASLD encompasses a range of liver conditions, from simple steatosis to more severe metabolic dysfunction-associated steatohepatitis (MASH) [1]. A recent meta-analysis revealed that 38% of adults and 7–14% of children and adolescents globally are affected by MASLD [2,3]. The prevalence of MASLD is reported to be highest in Latin America (44.4%) and lowest in Western Europe (25.1%), according to population-based studies conducted between 1990 and 2019 [3]. By 2040, the prevalence of MASLD among adults is projected to exceed 55% globally [4]. However, no clinically effective treatment for MASLD is currently available owing to the limited understanding of its pathology. Given the association between MASLD and cirrhosis, hepatocellular carcinoma, and liver-related mortality, the development of therapeutic interventions is imperative.

Lysophospholipids, deacylated phospholipid products, have recently garnered attention, particularly for their roles as second messengers [5]. Furthermore, lysophospholipids are involved in the onset and progression of chronic liver disease [6]. They are mainly categorized into six types based on the differences in their head groups: lysophosphatidic acid (LPA), lysophosphatidylcholine (LPC), lysophosphatidylethanolamine (LPE), lysophosphatidylglycerol, lysophosphatidylinositol, and lysophosphatidylserine. A reduction in LPE levels has been observed in the plasma/serum of patients with MASLD and MASH [7,8]. Moreover, LPE species were decreased in the livers of MASH mice induced by a combination of a high-fat diet and injection of oxidized low-density lipoprotein [9]. There have been several reports on the bioactivity of LPE. LPE increases intracellular Ca^2+^ levels and neurite outgrowth via mitogen-activated protein kinase signaling [10,11,12]. In vivo studies suggest an anti-inflammatory role for LPE by inhibiting M1 macrophage polarization [13,14]. Furthermore, LPE has been suggested to ameliorate mitochondrial injury in vitro and in vivo [15,16]. Given the protective effects of LPEs, their reduction may contribute to the progression of MASLD and MASH. However, the mechanisms underlying LPE reduction in MASLD and MASH remain unknown.

This study aimed to investigate the mechanism underlying LPE alterations in the context of MASH at the in vivo level and the association between LPEs and the pathology of MASH. To this end, we examined lipid metabolism related to LPEs in an experimental murine model of MASH using targeted and non-targeted liquid chromatography–tandem mass spectrometry (LC-MS/MS). Preclinical MASH models include diet-, deficiency-, toxin-, or genetically induced types, or combinations thereof. An ideal translational MASH model should be established on a nutritional basis to mimic human overnutrition-driven MASH [17]. Methionine/choline-deficient diets reproduce inflammation, fibrosis, and steatosis but disrupt phosphatidylcholine (PC) and phosphatidylethanolamine (PE) synthesis artificially [18], making it unsuitable for LPE metabolism studies. Cholesterol drives the progression of MASH from simple steatosis via inflammation and fibrogenesis [19]. Western diets that contain high cholesterol better reflect obesity and insulin resistance [17] but require prolonged feeding and increase risks of mortality and age-related complications [20]. Thus, we employed a combination of diet with high-fat, high-cholesterol, and cholic acid and 2% hydroxypropyl-β-cyclodextrin in drinking water (HFCC/CDX) to rapidly induce human-like steatosis, inflammation, fibrosis, and insulin resistance in three weeks via enhanced cholesterol uptake [21,22]. We further examined genes/proteins involved in LPE synthesis and catabolism to reveal dysregulation that illuminates MASH pathogenesis.

2. Results

2.1. Distribution of LPE Species in Healthy Mice

To date, there have been no reports on the distribution of LPE species in vivo. Therefore, we initially determined LPE levels using targeted LC-MS/MS analysis across nine organs, including the brain, colon, heart, intestine, kidney, liver, pancreas, spleen, and white adipose tissue (WAT) in healthy mice. The levels of total LPE, combined major seven species (LPE 16:0, 18:0, 18:1, 18:2, 20:4, 20:5, and 22:6), exhibited the following descending order: kidney, liver, brain, colon, intestine, spleen, pancreas, heart, and WAT, with mean ± standard deviation (SD) levels of 451.62 ± 54.66, 432.35 ± 39.50, 111.45 ± 21.86, 94.43 ± 15.98, 79.75 ± 19.31, 66.58 ± 5.47, 48.47 ± 3.11, 39.43 ± 2.25, and 9.58 ± 1.45 pmol/mg, respectively (Figure 1).

The compositions of individual LPE species (16:0, 18:0, 18:1, 18:2, 20:4, 20:5, and 22:6) across these organs and plasma were as follows: brain (5.1%, 4.1%, 6.4%, 0.5%, 57.3%, 0.2%, 26.3%), colon (22.0%, 30.6%, 18.0%, 4.3%, 21.5%, 0.5%, 3.1%), heart (9.8%, 17.5%, 6.1%, 4.3%, 30.4%, 0.2%, 31.7%), intestine (15.8%, 19.5%, 7.8%, 7.9%, 42.1%, 1.0%, 5.9%), kidney (5.6%, 9.4%, 6.1%, 2.4%, 66.6%, 1.4%, 8.6%), liver (13.0%, 6.2%, 6.6%, 4.5%, 55.0%, 1.4%, 13.4%), pancreas (12.8%, 23.7%, 10.1%, 5.0%, 42.8%, 0.8%, 4.7%), spleen (13.4%, 15.9%, 4.8%, 3.0%, 52.7%, 0.6%, 9.6%), WAT (15.4%, 20.9%, 11.4%, 11.1%, 34.2%, 0.0%, 7.0%), plasma (11.5%, 9.2%, 5.1%, 7.8%, 48.2%, 1.7%, 16.5%), respectively (Figure 2). In the brain, kidney, liver, and plasma, polyunsaturated fatty acid (PUFA)-containing LPEs constitute over 70% of the total LPEs, indicating their significant prevalence. Conversely, in the colon, intestine, pancreas, and WAT, saturated fatty acid (SFA)-containing LPEs were predominant, comprising more than 30% of the total.

2.2. Characteristics of HFCC/CDX Mice

We investigated the characteristics of the experimental model (Table 1). Due to the intake of HFCC/CDX, body weight was decreased at the end of three weeks of dietary induction, accompanied by adipose tissue involution. In contrast, liver weight was increased in HFCC/CDX mice, with a significantly higher liver weight/body weight ratio than that of the control group. Plasma alanine aminotransferase (ALT) levels, indicating liver injury, were significantly increased, and aspartate aminotransferase (AST) levels, indicating systemic inflammation, tended to increase in HFCC/CDX mice (p = 0.0649). The HFCC/CDX mice showed no significant changes in fasting blood glucose (FBG) levels, insulin secretion, and homeostasis model assessment of insulin resistance (HOMA-IR), contrary to the previous reports [21,22]. Hepatomegaly and enlarged gallbladders were observed in the HFCC/CDX livers (Figure 3A). Hematoxylin-Eosin (HE) staining revealed ballooning cells, a hallmark of MASH, and immune cell infiltration in the livers of HFCC/CDX mice (Figure 3B). Oil Red O and Masson trichrome (MT) staining revealed hepatic lipid accumulation and fibrosis (Figure 3C,D). The HFCC/CDX mice exhibited steatosis (Oil Red O-positive area: 20.9%), scattered hepatocyte ballooning, lobular inflammation, and pericellular or perivenular fibrosis, consistent with acute MASH pathology.

2.3. Alteration in Lipidomic Signatures of HFCC/CDX Mice

Non-targeted LC-MS/MS revealed the alterations in lipidomic signatures in the plasma and liver of HFCC/CDX mice (Figure 4A,B). The intake of HFCC/CDX led to increased levels of circulating free cholesterol (FC) and cholesteryl ester (CE), and facilitated the accumulation of FC and CE in the liver. The total levels of bile acid (BA) were significantly elevated in both plasma and liver of HFCC/CDX mice, corresponding to the observed gallbladder enlargement. In accordance with the reduction in WAT, which is the primary site for triacylglycerol (TG) synthesis, a decrease in TG was observed in both the plasma and liver of HFCC/CDX mice. Moreover, the significant increase in hepatic oxidized triglyceride (oxTG) in HFCC/CDX mice partially accounted for the decline in intact TG levels. Diacylglycerol (DG), a precursor to both TG and phospholipids, showed a decrease in the plasma but a significant increase in the liver of HFCC/CDX mice. Notably, most hepatic phospholipids, including cytidine diphosphate-ethanolamine (CDP-Etn), phosphatidylserine (PS), PC, LPC, LPE, and LPA, were significantly decreased in HFCC/CDX mice, whereas glycerophosphorylethanolamine (GPE), a deacylated product of LPE, was increased markedly. There was no significant change in hepatic PE levels. Although a reduction in phospholipids, including PS, PC, LPC, PE, and LPE, was observed in the plasma of HFCC/CDX mice, an increase in GPE was not observed. In addition, hepatic monounsaturated fatty acid (MUFA) was increased in HFCC/CDX mice, while PUFA was significantly decreased in the plasma and liver. Collectively, disruptions in lipid metabolisms, including cholesterols, TGs, and phospholipids, were induced in the plasma and liver of HFCC/CDX mice.

2.4. Reduction in LPE Species in HFCC/CDX Mice

For a more in-depth analysis of LPE, we examined LPE alterations using targeted LC-MS/MS. Except for LPE 18:1, all LPE species exhibited a reduction in the plasma of HFCC/CDX mice (Figure 5A). The total plasma LPE levels were 78.37 ± 6.07 pmol/μL in the control and 48.18 ± 11.03 pmol/μL in HFCC/CDX mice, expressed as mean ± SD. In the liver of HFCC/CDX mice, only LPE 18:1 was increased, while all other LPE species demonstrated a significant decrease compared with the control (Figure 5B). The total liver LPE levels were also significantly reduced in the HFCC/CDX mice (control: 432.35 ± 39.50 pmol/mg; HFCC/CDX: 236.97 ± 40.55 pmol/mg). In both plasma and liver samples, LPE 20:5 was detectable in the control, whereas its levels were below the limit of quantification in the HFCC/CDX mice. The detailed LPE levels in the samples are listed in Tables S1 and S2. Receiver operating characteristic (ROC) curve analysis demonstrated that most plasma LPE species were capable of distinguishing the HFCC/CDX group from the control, with an area under the curve (AUC) exceeding 0.9 (Table 2). Furthermore, all liver LPE species could differentiate the HFCC/CDX group from the control, achieving an AUC of 1.00.

2.5. Association Between LPE Species and Hepatic Inflammation

Correlation analysis revealed that all LPE species, except for LPE 18:1, exhibited strong negative correlations with ALT levels in both plasma and liver (Figure 6A,B). Conversely, hepatic LPE 18:1 demonstrated a positive correlation with ALT levels. No significant correlations were observed between LPE levels and indicators of glucose and insulin resistance in the liver. Apart from LPE 18:1, most plasma LPE species showed positive correlations with hepatic LPE species (Figure 6C). Furthermore, there was a significant positive correlation between total plasma and total liver LPE levels (r = 0.664, p = 0.022) (Figure 6D). Overall, these results indicate that the reduction in LPE species in both plasma and liver is associated with hepatic inflammation. Moreover, our findings indicate the possibility that a reduction in hepatic LPE results in a corresponding decrease in plasma LPE levels.

2.6. Dysregulation of Lysophospholipases in the Liver of HFCC/CDX Mice

To reveal the mechanism responsible for LPE reductions in HFCC/CDX mice, the protein expression of lysophospholipases was investigated. Patatin like phospholipase domain containing protein 6/7 (PNPLA6/7) plays a role in the degradation of membrane phospholipids, such as the conversion of LPE to GPE [23]. In the liver of HFCC/CDX mice, upregulation of Pnpla6 and Pnpla7 was observed compared with the control (Figure 7A). While the levels of total liver LPE were negatively correlated with expression levels of Pnpla6 (r = −0.741, p = 0.008) and Pnpla7 (r = −0.874, p < 0.001), liver GPE levels were positively correlated with those of Pnpla6 (r = 0.762, p = 0.006) and Pnpla7 (r = 0.776, p = 0.004). Moreover, ALT levels were positively correlated with the expression levels of Pnpla6 (r = 0.860, p < 0.001) and Pnpla7 (r = 0.734, p = 0.009), whereas the other biological indices involved systemic inflammation, glucose intolerance, and insulin resistance did not show significant correlations with the expression of those enzymes, indicating that dysregulation of Pnpla6/7 is primarily associated with hepatic inflammation. We also investigated gene expression to explore the dysregulation of LPE synthesis and catabolism in alternative pathways (Figure 7B). As a result, the expression of Pnpla8, ectonucleotide pyrophosphatase/phosphodiesterase 2 (Enpp2), and ethanolamine phosphotransferase 1 (Ept1) was downregulated in the liver of HFCC/CDX mice compared with the control. There were no significant differences in the expressions of Pnpla9, lysophosphatidylcholine acyltransferase 3 (Lpcat3), choline/ethanolamine phosphotransferase 1 (Cept1), and phosphatidylserine decarboxylase (Pisd). These findings suggest that the reduction in hepatic LPE is primarily attributed to the upregulation of Pnpla6 and Pnpla7 in MASH mice.

3. Discussion

In this study, plasma LPE was reduced in the HFCC/CDX mice, consistent with the previous report on MASH patients [8]. Additionally, we observed enhanced degradation of LPE to GPE associated with upregulation of Pnpla6 and Pnpla7 in the livers of HFCC/CDX mice, indicating the possibility of accelerated LPE catabolism under MASH-like conditions. The potential degradation of LPE by phospholipases and lysopholipases on circulating lipoproteins cannot be excluded and should be clarified in future studies. Despite this, the strong positive correlation observed between hepatic and plasma LPEs—combined with evidence that LPE is a minor constituent of lipoproteins [24] and that lysophospholipids are secreted from the liver via extracellular vesicles [25,26]—suggests that changes in hepatic LPE likely contribute to alterations in plasma LPE levels. Given that LPE exhibits several protective effects against inflammation and mitochondrial injury [13,14,15,16], it is hypothesized that a reduction in LPE may be a contributing factor to the progression of MASH. However, this study provides only associative evidence, lacking direct functional validation. It is anticipated that further investigations will elucidate the effects of LPE on the pathology of MASH.

The PNPLA family is a Ca^2+^-independent phospholipase A (PLA) and is involved in maintaining homeostasis and integrity of organelle membranes by regulating lipid metabolism [27]. Pnpla6, also known as neuropathy target esterase, is broadly expressed in murine tissues, including liver, predominantly in the central nervous system [28]. PNPLA6 mutation causes a broad spectrum of neurological disorders, including gait disturbance, visual impairment, anterior hypopituitarism, and hair anomalies. Moreover, PNPLA6 ablation via knockdown or knockout reveals its essential function in neural development and embryogenesis [29]. To date, PNPLA6 dysregulation in liver diseases has not been reported. Pnpla7 is primarily expressed in skeletal muscle, testis, and liver [28]. It is reported that Pnpla7 influences hepatic lipid metabolism by regulating substrate hydrolysis, endoplasmic reticulum targeting, and interactions with lipid droplets [28,30]. Moreover, Pnpla7 is upregulated in the livers of other murine MASLD models, including the high-fat diet model and db/db model, and is associated with hepatic very-low-density lipoprotein secretion [31]. In patients with MASLD, hepatic PNPLA7 mRNA expression was upregulated and was positively correlated with plasma lipid levels [31]. Previous in vitro studies demonstrated that knockdown of PNPLA6 results in increased intracellular LPE, whereas overexpression leads to decreased LPE using human retinal pigment epithelial cells [32]. Additionally, studies utilizing COS-7 cells stably expressing PNPLA7 fused to enhanced green fluorescent protein have shown that PNPLA7 functions as an LPE-degrading enzyme [30].

The HFCC/CDX mice in this study effectively represent the acute phase of the MASH condition, characterized by hepatic steatosis, inflammation, and fibrosis, while lacking insulin resistance. The HFCC/CDX model is appropriate for assessing human MASH pathology, as it seeks to induce fibrosis by activating hepatic stellate cells with FC, thereby promoting their differentiation into myofibroblasts [19]. Moreover, it possesses the potential to replicate immune tolerance mediated by regulatory T cells, as observed in the chronic phase [22]. Consequently, the evaluation of MASH pathology using the HFCC/CDX model is translational and could be deemed valuable for predicting LPE metabolic dysfunction in the chronic phase. Although insulin resistance is a significant characteristic of MASLD and MASH progression, it is not universally present in all patients with MASH [33]. Based on our histological and biological evidence, dysregulation of LPE metabolism and Pnpla6/7 may induce features of human MASH in HFCC/CDX mice, even in the absence of insulin resistance.

LPE metabolism is maintained through the coordinated activity of PLA1/2 and lysophospholipid acyltransferases. In the Kennedy pathway, PE is synthesized when CDP-Etn is transferred to DG by EPT1 and CEPT1 [34]. PLA1/2 enzymes such as PNPLA8 and PNPLA9 hydrolyze PE to generate LPE, which can be further deacylated to GPE by lysophospholipases such as PNPLA6 and PNPLA7 [27]. Conversely, lysophospholipid acyltransferases such as LPCAT3 acylate LPE to regenerate PE via the Lands’ cycle. This process is essential for maintaining membrane composition and organelle homeostasis [35]. Our findings suggest that both Pnpla6 and Pnpla7 play a cooperative role in promoting LPE turnover during hepatic inflammation. Additional disruption of phospholipid metabolism was evident, including decreased CDP-Etn, PS, PC, LPC, and LPA levels and reduced expression of Pnpla8 and Ept1 (Figure 8). Although LPE can also be converted to LPA by ENPP2 [36], this pathway does not appear to account for the LPE reduction observed in this model. Taken together, our data indicate the association between LPE reduction and dysregulation of Pnpla6/7, providing a novel insight into the mechanism linking altered phospholipid metabolism to MASH pathogenesis.

Non-targeted lipidomic analysis showed that hepatic steatosis in HFCC/CDX mice was marked by the accumulation of oxTG rather than intact TG. Previous studies suggest that TG accumulation itself is not directly toxic and is an epiphenomenon that occurs alongside the generation of harmful lipid metabolites, lipotoxicity, and hepatocellular injury [37]. In contrast, oxidized lipids can impair protein and membrane function, promoting hepatic inflammation and injury [38]. Because oxidative stress is a key driver of MASH progression, the presence of oxTG rather than TG may be closely associated with liver damage in HFCC/CDX mice.

In the liver of HFCC/CDX mice, most LPE species exhibited a decrease and demonstrated a negative correlation with hepatic inflammation, while LPE 18:1 displayed contrary behavior, which is hypothesized to result from an increase in MUFA within the liver. In the context of MASLD, the unsaturation of SFAs and incorporation of MUFA into phospholipids serve as a protective mechanism against lipotoxicity [39,40]. Consequently, it is proposed that LPE 18:1 may have exhibited behavior distinct from other LPE species.

The endoplasmic reticulum and Golgi apparatus serve as the principal sites for the synthesis of most lipids, including PE and LPE [35]. Lipid metabolism in these organelles is particularly active in the liver [41,42] and kidney [43], consistent with the observed high abundance of LPEs in these tissues. Notably, substantial levels of LPEs were also detected in the brain, intestine, and colon. Given the abundance of PE in the brain [44], it is reasonable to infer that its degradation product, LPE, is also enriched. Moreover, as lipid uptake by intestinal mucosal cells is a highly active process [45], the intestine and colon are likely to contain abundant LPE species. In contrast, lower LPE levels were detected in WAT. WAT primarily functions in energy storage and lipid redistribution, with polar lipids representing only 0.29% of the total lipids, among which PE represents 3.64% [46]. This composition indicates that PE is a minor component in WAT, which likely accounts for the limited abundance of LPEs. We further examined the relative abundance of LPE species across organs and found that the brain, kidney, and liver—organs rich in PUFA-containing PE [47]—also exhibited elevated levels of PUFA-containing LPEs. Conversely, PUFA-containing LPEs were lower in the colon, plausibly due to degradation mediated by group X PLA2, which preferentially cleaves PUFAs from lysophospholipids and is predominantly expressed in the colon [48]. To the best of our knowledge, this is the first report on the absolute LPE distribution across multiple organs.

This study has several limitations. (1) The extraction method yielded a recovery rate of less than 50% for LPE 20:5 in WAT (Table S3), indicating the need for further optimization of the extraction procedure to improve detectability in this tissue. Nevertheless, because the method effectively extracted LPE species from multiple other tissues, it remains appropriate for comparative screening across diverse biological matrices. (2) While the HFCC/CDX mice accurately represented the characteristics of the acute phase of MASH, they did not capture the chronic phase. Consequently, the relationship between LPE dysregulation by Pnpla6/7 and chronic manifestations, particularly severe fibrosis, remains unclear. (3) The HFCC diet contains supraphysiological levels of cholesterol and cholic acid that exceed typical human dietary intake. This pharmacological component may exaggerate hepatic cholesterol accumulation and accelerate disease progression beyond that observed in human MASH, potentially limiting direct translatability to patients consuming standard Western diets. (4) This study did not consider the potential impact of sex on LPE dysregulation in the context of MASH. (5) The sample size in this study is relatively small (n = 6 per group). Although significant effects were observed, future studies are needed to validate the robustness of these findings in larger cohorts. (6) While this study provides valuable in vivo findings, it does not elucidate the direct mechanisms underlying LPE dysregulation. Further in vitro investigations and sex-stratified translational studies are required to elucidate the precise mechanisms of LPE dysregulation, its sex-specific effects in MASH, the influence of hepatic LPE alterations on circulating LPE levels, and the protective effects of LPE against MASH pathology.

4. Materials and Methods

4.1. Materials

LC-MS-grade methanol (MeOH) was purchased from Kanto Chemical Co. Inc. (Tokyo, Japan). We purchased LC-MS-grade isopropanol and chloroform from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan) and Nacalai Tesque Inc. (Kyoto, Japan), respectively. Ammonium acetate (1 mol/L solution), ethylenediaminetetraacetic acid, and Tween 20 were purchased from Sigma-Aldrich (St. Louis, MO, USA). 2,6-di-tert-butyl-p-cresol was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). EquiSPLASH and oleic acid-d9 were used as internal standards (IS) for non-targeted LC-MS/MS analysis and were purchased from Avanti Polar Lipids (Alabaster, AL, USA). For the targeted LC-MS/MS analysis, 1-palmitoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 16:0), 1-stearoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 18:0), and 1-oleoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 18:1) were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL, USA), while 1-linoleoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 18:2), 1-arachidonyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 20:4), 1-eicosapentaenoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 20:5), and 1-docosahexaenoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE 22:6) were synthesized in-house as described previously [8].

4.2. Animal Experiments

The diet-induced MASH model was prepared according to previous studies [21,22]. Seven-week-old male C57BL/6J mice were obtained from CLEA Japan, Inc. (Tokyo, Japan). Mice were housed at 22 ± 2 °C under a 12/12 h light/dark cycle and allowed ad libitum feeding and drinking throughout the experiment. Prior to the start of the experiments, mice were divided into two groups (n = 6 in each group) randomly. The control group was fed normal chow (CE-2, CLEA Japan, Inc., Tokyo, Japan), while the HFCC/CDX group was fed a diet high in fat (60%), cholesterol (1.25%), and cholic acid (0.5%), along with 2% hydroxypropyl-β-cyclodextrin in drinking water for three weeks to induce MASH (Figure S1). The individual mouse was considered the experimental unit in this study. No a priori inclusion or exclusion criteria were established. All animals allocated to each group (n = 6) were included in the final analysis with no exclusions. To minimize pain and distress, mice in both the control and HFCC/CDX groups were anesthetized with 2% isoflurane (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Blood samples were collected by cardiac puncture and transferred into tubes containing heparin sodium (Mochida Pharmaceutical Co., Ltd., Tokyo, Japan). The samples were then centrifuged to obtain plasma and stored at −80 °C. Organ samples were freshly snap-frozen in dry ice and stored at −80 °C. The composition of fatty acids in the diet is described in Table S4.

4.3. Measurement of Biological Indices

All biological indices of the fasting plasma samples collected after anesthesia at the end of three weeks were measured using commercial kits, according to the manufacturer’s instructions. The levels of ALT and AST were measured using an Alanine Transaminase Activity Assay Kit (Cayman Chemical, Ann Arbor, MI, USA) and an Aspartate Aminotransferase Activity Assay Kit (Cayman Chemical, Ann Arbor, MI, USA), respectively. FBG levels were determined using LabAssay™ Glucose (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Plasma insulin levels were measured using a Mouse/Rat Insulin ELISA Kit M1108 (Morinaga BioScience, Inc., Kanagawa, Japan). To confirm insulin sensitivity, HOMA-IR was calculated as follows: HOMA-IR = fasting insulin (μU/mL) × FBG (mg/dL)/405.

4.4. Histology Analysis

Liver sections were fixed with 10% formalin and embedded in paraffin. HE, Oil Red O, and MT staining were performed at the Sapporo General Pathology Laboratory (Sapporo, Japan). The positive areas of Oil Red O and MT staining were quantified using ImageJ software 1.54d (National Institutes of Health, Bethesda, MD, USA).

4.5. Lipid Extraction

Lipids were extracted from the samples as previously described, with some modifications [49]. Briefly, 100 mg of organ samples were weighed, followed by the addition of 1 mL of cold MeOH containing 0.01% (w/v) 2,6-di-tert-butyl-p-cresol and homogenized using a Bead Mill 4 Homogenizer (Fisherbrand, Pittsburgh, PA, USA). Supernatant (100 μL) was added to 100 μL of IS solution (1 μg/mL of EquiSPLASH lipidomix and 10 μg/mL of oleic acid-d9) in MeOH. For plasma samples, 30 μL of plasma was added to 100 μL of IS solution and 100 μL of MeOH. Subsequently, 100 μL of chloroform and 20 μL of Milli-Q water were added to the mixture. The single-phase centrifuge was dried under vacuum at 4 °C and reconstituted in 100 μL of MeOH. The prepared samples were subjected to targeted and non-targeted LC-MS/MS analysis. The recovery rates of each LPE species in the organ samples are listed in Table S3. The recovery rates of the plasma samples have been reported in a previous study [49].

4.6. Targeted LC-MS/MS Analysis of LPE Species

Targeted LC-MS/MS analysis of LPE species was performed as described previously [49]. The LC-MS/MS system consisted of a high-performance liquid chromatography system (Shimadzu, Kyoto, Japan) and a TSQ Quantum Access mass spectrometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). LC separation was conducted using a Hypersil GOLD column (50 × 2.1 mm, 5.0 μm, Thermo Fisher Scientific Inc., Waltham, MA, USA). The column and sample tray were maintained at 45 °C and 4 °C, respectively. 5 μL of the sample was injected into the system. The mobile phase consisted of 60% Milli-Q water, 20% MeOH, and 20% acetonitrile with 5 mmol/L aqueous ammonium acetate containing 500 nmol/L ethylenediaminetetraacetic acid (A) and isopropanol (B). The gradient at a flow rate of 0.4 mL/min was applied as follows: 0–1 min (80% A, 20% B); 1–3 min (40% A, 60% B); 3–5 min (40% A, 60% B); 5–6 min (0% A, 100% B); 6–7 min (0% A, 100% B); and 7–10 min (80% A, 20% B). Each LPE species was detected in the electrospray ionization negative ion mode using single reaction monitoring channels by collision-induced dissociation (Table S5). The optimized ion source parameters were set as: spray voltage, 3000 V; vaporizer temperature, 300 °C; capillary temperature, 200 °C; sheath gas (nitrogen) pressure, 30 psi; ion sweep gas (nitrogen) pressure, 4 psi; and auxiliary gas (nitrogen) pressure, 35 psi. The sensitivities of the method, including limits of detection and limits of quantification, are described in Table S6.

4.7. Non-Targeted LC-MS/MS Analysis

Non-targeted LC-MS/MS analysis was performed as described previously [50,51]. The LC-MS/MS system consisted of an HPLC system (Shimadzu, Kyoto, Japan) integrated with an LTQ-Orbitrap mass spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) operating in both negative and positive electrospray ionization modes. A 10 μL sample was injected into the instrument for analysis. LC separation was conducted on an Atlantis T3 C18 column (2.1 mm × 150 mm, 3 μm, Waters, Milford, MA, USA) at a flow rate of 0.2 mL/min at 40 °C, utilizing a mobile phase consisting of 10 mM ammonium acetate aqueous solution (A), isopropanol (B), and MeOH (C). The gradient elution program for the negative ion mode was as follows: 0–1 min (30% B and 35% C), 1–9 min (75% B and 15% C), 9–21 min (82.5% B and 15% C), 21–25 min (95% B and 5% C), 25–26 min (30% B and 35% C), and maintained at this ratio until 30 min. For the positive ion mode, the gradient elution was as follows: 0–1 min (30% B and 35% C), 1–9 min (82.5% B and 15% C), 9–15 min (95% B and 5% C), 15–25 min (95% B and 5% C), 25–26 min (30% B and 35% C), and this ratio was maintained for 30 min. MS^1^ analysis was conducted with a resolving power of 60,000 at a collision energy of 35 V (scan range: m/z 160−1900 and 150−1950 in negative and positive ion modes, respectively). MS^2^ and MS^3^ analyses were performed in ion trap mode with an isolation width of 3 and collision energies of 40 and 45 V, respectively. The capillary voltage was set at 3 kV, the capillary temperature was 330 °C, the nitrogen-sheath gas flow was set to 50 units, and the nitrogen auxiliary gas was 5 units. The raw data were processed with a mass tolerance of 5.0 ppm using MS-DIAL 4.9 [52] and Xcalibur 2.2 (Thermo Fisher Scientific, Waltham, MA, USA). The analyte amounts in the sample were determined by calculating the peak area ratios of the analyte to the IS and multiplying them by the amount of added IS. The analyte amounts were corrected based on organ weight or plasma volume. The detailed lipidomics data are provided in Table S7.

4.8. Gene Expression

Total RNA was isolated from liver samples using a NucleoSpin kit (TaKaRa Bio Inc., Shiga, Japan) in accordance with the manufacturer’s protocol. RNA concentrations were measured using NanoDrop One (Thermo Fisher Scientific Inc., Waltham, MA, USA) and subsequently reverse-transcribed into cDNA using a ReverTra Ace qPCR RT Master Mix with gDNA remover (TOYOBO, Osaka, Japan). A quantitative polymerase chain reaction was conducted using a THUNDERBIRD SYBR qPCR Mix (TOYOBO, Co., Ltd., Osaka, Japan) and a CFX Connect Real-Time PCR Analysis System (Bio-Rad Laboratories, Hercules, CA, USA). The primer sequences of β-actin, Cept1, Enpp2, Ept1, Lpcat3, Pisd, Pnpla8, and Pnpla9 are listed in Table S8. The expression levels of the target genes were quantified using the 2^−ΔΔCt^ method and normalized to β-actin expression levels.

4.9. Protein Expression

Protein extraction and Western blotting were performed using commercial kits. Following lysis of liver samples, protein levels were determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). Equal amounts of protein (30 μg per lane) were separated by a 5–20% SDS-PAGE gel (ATTO Corporation, Tokyo, Japan) and transferred onto a 0.2 μm polyvinylidene difluoride membrane (Millipore, Bedford, MA, USA). The membranes were blocked with 5% skim milk in Tris-buffered saline with 0.1% Tween 20 to prevent nonspecific binding for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies against neuropathy target esterase, also known as Pnpla6 (1:1000, sc-271049, Santa Cruz Biotechnology, Dallas, TX, USA), Pnpla7 (1:1000, FLJ00415, ProteinExpress Co., Ltd., Chiba, Japan), and Gapdh (1:1000, sc-25778, Santa Cruz Biotechnology, Dallas, TX, USA). After washing with Tris-buffered saline with 0.1% Tween 20, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies against mouse IgG or rabbit IgG (1:10,000, #330 and #458; MBL, Nagoya, Japan) for 1 h at room temperature. Protein bands were visualized using horseradish peroxidase substrate (WSE-7120S/7120L EzWestLumi Plus, ATTO Corporation, Tokyo, Japan) and imaged with a CCD system (ChemiDoc MP Imaging System, Bio-Rad, Hercules, CA, USA). The intensities were quantified using the Image Lab 6.0.1 software (Bio-Rad, Hercules, CA, USA). The expression levels were normalized to those of Gapdh.

4.10. Statistics

GraphPad Prism 8.0.1 (GraphPad Software, La Jolla, CA, USA) was used for all statistical analyses. The non-parametric Mann–Whitney U test was applied for comparisons of two groups (p < 0.05). All results are depicted as the mean ± SD. ROC curve analysis was used to determine the AUC and 95% CIs. Spearman correlation coefficients were calculated to assess the association between LPE species and biological indices or between plasma and hepatic LPE species.

5. Conclusions

In summary, this study suggests that reduced LPE levels in both plasma and liver are associated with hepatic inflammation in a murine model of MASH. This reduction appears to be associated with the upregulation of hepatic Pnpla6 and Pnpla7. Particularly, to the best of our knowledge, this study is the first report on Pnpla6 dysregulation in liver diseases. It is expected that further investigations will elucidate the mechanisms underlying dysregulation of Pnpla6, in addition to Pnpla7, in liver diseases. Our findings provide new insight into the mechanisms underlying lipid metabolic imbalance in MASH and may inform the development of novel therapeutic strategies targeting phospholipid metabolism.