Inhibition of intracellular versus extracellular cathepsin D differentially alters the liver lipidome of mice with metabolic dysfunction‐associated steatohepatitis

Isabeau Vermeulen; Mengying Li; Hester van Mourik; Tulasi Yadati; Gert Eijkel; Benjamin Balluff; Roger Godschalk; Lieve Temmerman; Erik A L Biessen; Aditya Kulkarni; Jan Theys; Tom Houben; Berta Cillero‐Pastor; Ronit Shiri‐Sverdlov

doi:10.1111/febs.17358

. 2024 Dec 26;292(7):1781–1797. doi: 10.1111/febs.17358

Inhibition of intracellular versus extracellular cathepsin D differentially alters the liver lipidome of mice with metabolic dysfunction‐associated steatohepatitis

Isabeau Vermeulen ¹, Mengying Li ², Hester van Mourik ^2,³, Tulasi Yadati ², Gert Eijkel ¹, Benjamin Balluff ¹, Roger Godschalk ⁴, Lieve Temmerman ⁵, Erik A L Biessen ^5,⁶, Aditya Kulkarni ⁷, Jan Theys ³, Tom Houben ², Berta Cillero‐Pastor ^1,⁸, Ronit Shiri‐Sverdlov ^2,^✉

PMCID: PMC11970712 PMID: 39726152

Abstract

The prevalence of metabolic dysfunction‐associated steatotic liver disease (MASLD) progressing to metabolic dysfunction‐associated steatohepatitis (MASH), characterized by hepatic inflammation, has significantly increased in recent years due to unhealthy dietary practices and sedentary lifestyles. Cathepsin D (CTSD), a lysosomal protease involved in lipid homeostasis, is linked to abnormal lipid metabolism and inflammation in MASH. Although primarily intracellular, CTSD can be secreted extracellularly. Our previous proteomics research has shown that inhibition of extracellular CTSD results in more anti‐inflammatory effects and fewer potential side effects compared to intracellular CTSD inhibition. However, the correlation between reduced side effects and alterations in the hepatic lipid composition remains unknown. This study aims to investigate the correlation between intra‐ and extracellular CTSD inhibition and potential alterations in the hepatic lipid composition in MASH. Low‐density lipoprotein receptor knockout (Ldlr ^−/−) mice were fed a high‐fat diet for 10 weeks and received subcutaneous injections every 2 days of vehicle, intracellular CTSD inhibitor (GA‐12), or extracellular CTSD inhibitor (CTD‐002). Matrix‐assisted laser desorption/ionization mass spectrometry imaging (MALDI‐MSI) was used to visualize and compare the lipid composition in liver tissues. Hepatic phosphatidylcholine remodeling was observed with both inhibitors, suggesting their therapeutic potential in treating MASH. Treatment with an intracellular CTSD inhibitor resulted in elevated levels of cardiolipin, reactive oxygen species, phosphatidylinositol, phosphatidylethanolamine, and lipids that are linked to mitochondrial dysfunction and inflammation, and induced more oxidative stress. The observed modifications in lipid composition demonstrate the clinical advantages of extracellular CTSD inhibition as a potentially beneficial therapeutic approach for MASH.

Keywords: cathepsin, lipidomics, MALDI, small‐compound inhibitors, steatotic liver disease

Cathepsin D (CTSD), primarily intracellular, is crucial for normal cellular processes but can be secreted extracellularly in diseases like metabolic dysfunction‐associated steatohepatitis (MASH), associated with liver fat accumulation. Our study investigated how inhibiting CTSD inside and outside of cells influences lipid levels in a mouse model that closely resembles human MASH. Our findings revealed that inhibiting the external activity of CTSD might offer a more effective and protective approach for MASH.

graphic file with name FEBS-292-1781-g001.jpg

Abbreviations

CL: cardiolipin
CTSD: cathepsin D
LDA: linear discriminant analysis
Ldlr^−/−: low‐density lipoprotein receptor knockout
MALDI‐MSI: matrix‐assisted laser desorption/ionization mass spectrometry imaging
MASH: metabolic dysfunction‐associated steatohepatitis
MASLD: metabolic dysfunction‐associated steatotic liver disease
PA: phosphatic acid
PBCM: peripheral blood mononuclear cells
PC: phosphatidylcholine
PCA: principal component analysis
PE: phosphatidylethanolamines
PG: phosphatidylglycerol
PI: phosphatidylinositol
ROS: reactive oxygen species

Introduction

The etiological connection between an unhealthy lifestyle and the emergence of metabolic dysfunction‐associated steatotic liver disease (MASLD) (previously known as nonalcoholic fatty liver disease (NAFLD)) is well‐established [1, 2]. In the progression of MASLD, a range of liver conditions can occur due to the buildup of lipids in the liver. This spectrum typically begins with simple steatosis, known as fatty liver, and can progress to a more severe form of metabolic dysfunction‐associated steatohepatitis (MASH) (previously known as nonalcoholic steatohepatitis (NASH)), which is characterized by inflammation [3]. Over the past few decades, there has been a substantial increase in the prevalence of MASH from 25.3% to 38.0% [1, 4]. Recently, resmetirom, an oral thyroid hormone receptor‐β (THR‐β) agonist, became the first drug approved by the FDA for the treatment of MASH and is currently under regulatory review in the EU [5]. However, not all patients show liver histology improvement during clinical trials. Therefore, studying additional therapeutic targets remains important to improve patient outcomes [5].

In recent years, numerous studies have focused on identifying novel biological targets for the treatment of MASH. Lysosomal enzymes, particularly cathepsins, have emerged as promising options given their involvement in various cellular processes during the development of MASH [6, 7]. Cathepsins, divided into several different protease families, are located within endo/lysosomal compartments and extra‐lysosomal locations, such as the cytosol, nucleus, and mitochondria. These intracellular cathepsins are involved in maintaining cellular homeostasis, including the immune response and energy regulation [8, 9]. Cathepsin D (CTSD), one of the most abundant proteases, is typically located intracellularly and primarily contributes to cellular homeostasis. However, in various pathological conditions, such as MASH, there is an increased secretion of cathepsin D outside the cell [9, 10]. Plasma CTSD levels have emerged as a potential biomarker for early MASH stages [11]. Subsequent studies by our research group have elucidated the role of CTSD in type 2 diabetes and MASLD and demonstrated that inhibiting extracellular CTSD reduced hepatic steatosis in MASLD rats [12, 13]. Building on these findings, further investigations were conducted on the effects of extra‐ and intracellular CTSD inhibition in a humanized MASH mouse model, more specifically low‐density lipoprotein receptor knockout (Ldlr ^−/−) mice [14]. In this study, the mice clearly exhibited key pathological characteristics of MASH. These characteristics included significant hepatic steatosis, disrupted lipid metabolism, and signs of metabolic inflammation. The observed alterations in lipid metabolism pathways and evidence of lysosomal dysfunction, coupled with changes in lipid biosynthesis, strongly indicated that the model mimics MASH pathology [14]. The results of this study further showed that extracellular CTSD inhibition reduced hepatic steatosis and inflammatory status compared to intracellular CTSD inhibition in MASH mice [14]. Additionally, proteomics data revealed that extracellular CTSD inhibition promoted an anti‐inflammatory hepatic protein profile, while intracellular inhibition downregulated proteins associated with mitochondrial oxidative phosphorylation [13, 14]. However, while proteomic analysis demonstrated the differences in lipid metabolism pathways and mitochondrial function between extra‐ and intracellular CTSD inhibitors treated mice, the exact hepatic lipid composition and modifications upon both inhibitors remained unexplored.

In this study, we utilized matrix‐assisted laser desorption–ionization mass spectrometry (MALDI‐MSI) to investigate alterations in hepatic lipid composition and distribution in high‐fat, high‐cholesterol fed Ldlr ^−/− mice following inhibition of intra‐and extracellular CTSD, using GA‐12 and CTD‐002 inhibitors, respectively. MALDI‐MSI is a label‐free analytical technique capable of visualizing the spatial distribution of various biomolecules, including lipids [15]. Coating liver sections with a thin layer of matrix aids in the ionization and desorption of lipids from the liver. Next, the molecular profile was recorded by systematically directing a laser across the tissue section in a pixelated movement. This molecular profile consisted of the mass‐to‐charge ratio and relative intensity. Subsequently, the x‐ and y‐coordinates of the pixels, along with the relative intensities of the molecules, were used to generate molecular heatmaps [16]. MALDI‐MSI has been applied to diverse preclinical models to explore lipid alterations after drug administration. For instance, one study examined amitriptyline‐treated rat livers and revealed changes in lipid profiles indicative of drug‐induced steatosis [17]. Another study focused on mapping lipid changes in the gastrointestinal tract after oral drug delivery [18]. These examples highlight the capability of MALDI‐MSI to investigate drug‐induced lipid modifications in different biological systems.

Through this approach, we effectively investigated the potential side effects of intracellular and extracellular CTSD inhibition on the liver lipidome, revealing lipid classes associated with metabolic inflammation and mitochondrial function. Additional experiments were conducted to assess oxidative stress‐induced lipotoxicity and further substantiate our conclusions drawn from the lipid data. This detailed analysis provides a comprehensive understanding of the lipidomic consequences associated with CTSD inhibition.

Results

CTSD inhibition results in different hepatic phosphatidylcholine profiles compared to vehicle‐treated Ldlr ^−/− mice

Ldlr ^−/− mice treated with intracellular and extracellular CTSD inhibitors showed differences in lipid levels, and distinct lipid pathways were observed in previous proteomics results, such as fatty acid biosynthesis and steroid hormone biosynthesis [14]. However, it was not yet clear which lipid classes were different and which lipids were affected or involved in these pathways. To compare lipid profiles between the groups, we performed MALDI‐MSI.

Using positive‐mode MALDI‐MSI, we observed distinct lipid profiles in liver tissues treated with intracellular or extracellular CTSD inhibitors compared to vehicle‐treated livers. This mode enables the detection of various lipid classes, including phosphatidylcholines (PCs) and sphingomyelin (SM). Employing PCA‐LDA, we successfully differentiated both the intracellular and extracellular inhibitor‐treated groups from the vehicle‐treated group (Fig. S1). The score projection of discriminant function 1 (DF1) further highlighted the distinct lipid profiles between the intracellular and extracellular inhibitor‐treated groups compared to the vehicle‐treated group (Fig. 1A,B). From this score projection, it was observed that the lipids exhibited a homogenous distribution, characterized by a lack of specific regions. This observation aligns with the H&E staining of liver sections, further confirming the homogeneity within the liver tissue samples (Fig. S2). The scaled loading projection of DF1 provided a spectrum of differential profiles of the treated groups and vehicle‐treated groups (Fig. 1A,B).

Fig. 1 — Score projection of discriminant function 1 (DF1) in positive mode The comparisons show a different lipid profile between (A) the vehicle‐treated group (n = 3) and the extracellular inhibitor‐treated group (n = 3) with their corresponding scaled loading score spectrum (B) between the vehicle‐treated group and the intracellular inhibitor‐treated group (n = 3) with their corresponding scaled loading score spectrum and (C) between the extracellular inhibitor‐treated group and the intracellular inhibitor‐treated group with their corresponding scaled loading score spectrum. In the score projection images shown on the left, each color represents different levels or intensities of specific molecular patterns found in the tissues. The images compare the extracellular inhibitor‐treated group to the intracellular inhibitor‐treated group, and each of these to the vehicle‐treated group. These comparisons reveal the sections where molecular patterns differ between each pair of treatment conditions revealing to which group they belong. On the right, it shows the scaled loading spectrum for each comparison. A higher score is attributed to a lipid when it contributes more to the difference between the conditions. In each panel, the lipids with the highest scores and thus distinctive for each condition are listed (Scale bar = 3 mm).

In the comparative analysis between distinct groups of mice, we identified distinct PCs (Tables S1 and S2). Notably, PCA‐LDA enabled differentiation of the intracellular inhibitor‐treated group from the extracellular inhibitor‐treated group, revealing distinct lipid profiles (Fig. 1C). Analysis of the specific PCs associated with each inhibitor revealed that the intracellular inhibitor‐treated group exhibited an abundance of potassium (K) adducts, whereas the extracellular inhibitor‐treated group displayed a higher abundance of protonated (H) and sodium (Na) adducts (Table S3).

Phospholipid dynamics between inhibitors‐ and vehicle‐treated Ldlr ^−/− mice

Using the negative‐mode MALDI‐MSI approach, our study also showed substantial differences between the lipid profiles of the three groups of mice. Diverse lipid classes, including phosphatic acids (PAs), phosphatidylethanolamines (PEs), phosphatidylinositol's (PIs), cardiolipins (CLs), and phosphatidylglycerols (PGs), were observed and identified (Fig. 2).

Fig. 2 — Score projection of discriminant function 1 (DF1) in negative mode The comparisons show a different lipid profile between (A) the vehicle‐treated group (n = 3) and the extracellular inhibitor‐treated group (n = 3) with their corresponding scaled loading score spectrum (B) between the vehicle‐treated group and the intracellular inhibitor‐treated group (n = 3) with their corresponding scaled loading score spectrum (C) between the extracellular inhibitor‐treated group and the intracellular inhibitor‐treated group with their corresponding scaled loading score spectrum. In the score projection images shown on the left, each color represents different levels or intensities of specific molecular patterns found in the tissues. The images compare the extracellular inhibitor‐treated group to the intracellular inhibitor‐treated group and each of these to the vehicle‐treated group. These comparisons reveal the sections where molecular patterns differ between each pair of treatment conditions revealing to which group they belong. On the right, it shows the scaled loading spectrum for each comparison. A higher score is attributed to a lipid when it contributes more to the difference between the conditions. In each panel, the lipids with the highest scores and thus distinctive for each condition are listed (Scale bar = 3 mm).

Several notable differences were found based on the PCA‐LDA analysis between different conditions. In both the intracellular inhibitor‐treated and extracellular inhibitor‐treated groups, an enrichment of PIs was found in comparison to the vehicle‐treated group (Fig. 2A,B).

Moreover, lipid PA 38:4 was relatively increased in both inhibitor‐treated groups compared to the vehicle‐treated group. Additionally, a set of lipids, namely cholesterol sulfate and taurocholic acid, a secondary bile acid, were specifically identified in the vehicle‐treated group compared to the inhibitor‐treated groups (Tables S4–S6).

In the comparative analysis between the intracellular inhibitor‐treated group and the extracellular inhibitor‐treated group, a notable observation was the enrichment of PEs, which were more distinctive in the intracellular inhibitor‐treated group, while PGs were more abundant in the extracellular inhibitor‐treated group (Fig. 2C; Table S6).

Cardiolipins are enriched in the livers of intracellular CTSD inhibitor‐treated Ldlr ^−/− mice

Moving beyond the first PCA‐LDA analysis, which focused on the predominant lipid classes driving inter‐group variance based on relative intensity, we then focused on the specific examination of cardiolipins. In contrast to the lipid classes that are prominent in the PCA‐LDA analysis due to their higher relative intensities, such as PIs, PEs, and PGs, CLs are situated in a mass range above m/z 1400 and exhibit comparatively lower intensities [19]. Consequently, they did not emerge with the highest scaled loading in the PCA‐LDA analysis. However, exploration of CLs is particularly relevant because of the distinctive nature of cardiolipins and their potential role in the observed variations among the groups.

Focusing on the mass range above 1400 m/z in the PCA‐LDA analysis, the lipid profiles revealed a downregulation of cardiolipins in the extracellular inhibitor‐treated group compared to both the intracellular inhibitor‐treated and vehicle‐treated groups. Remarkably, the identified cardiolipins seemed to be distinctive for the intracellular inhibitor‐treated group when compared to the extracellular inhibitor‐treated group (Fig. 3A; Table S7). The visual representation of CLs 72:6 and CLs 72:7 showed a relatively lower intensity of these cardiolipin species in the extracellular inhibitor‐treated group than in the other two groups (Fig. 3B).

Fig. 3 — Elevated cardiolipins in the intracellular inhibitor‐treated group. (A) Scaled loading spectrum of the PCA‐LDA comparing the extracellular inhibitor‐treated group (n = 3) with the intracellular inhibitor‐treated group (n = 3) in negative mode shows several cardiolipins (CL) to be specific for the intracellular inhibitor‐treated group. (B) MALDI‐MSI images of 1452.00 *m/z* (CL 72:6) and 1449.98 (CL 72:7) show the difference in relative intensity between the vehicle‐treated group, the intracellular inhibitor‐treated group, and the extracellular inhibitor‐treated group (Scale bar = 6 mm).

Lipotoxicity upon treatment with CTSD inhibitors

Cardiolipins, located in the inner mitochondrial membrane, are susceptible to attack by reactive oxygen species (ROS), which are overproduced during oxidative stress [20]. Given the increased abundance of cardiolipins in the intracellular CTSD inhibitor‐treated group, we investigated the potential generation of oxidative stress by increased ROS production induced by both intracellular and extracellular inhibitors of CTSD. We performed an ROS detection cell‐based assay in PBMCs derived macrophages, which were treated with 100 μm of the intracellular CTSD inhibitor, 100 μm of the extracellular CTSD inhibitor, or negative control (0.1% DMSO). ROS levels in human macrophages treated with the intracellular CTSD inhibitor increased continuously compared to those in the extracellular inhibitor and negative control (Fig. 4A). Likewise, as hepatocytes are the main affected hepatic cell type in MASH, ROS levels were measured in AML‐12 cells in response to the CTSD inhibitors. Remarkably, no significant differences were observed in ROS production upon incubation with either intra‐ or extracellular CTSD inhibitors in AML‐12 cells (Fig. 4B).

Fig. 4 — Reactive oxygen species (ROS) measurements in response to treatment with CTSD inhibitors. (A) Human peripheral blood mononuclear cells (PBMCs) derived macrophages and (B) the murine hepatocyte cell line AML‐12 isolated from the mouse liver were treated with 100 μm of the intracellular CTSD inhibitor (GA‐12), 100 μm of the extracellular CTSD inhibitor (CTD‐002), or negative control (0.1% DMSO) for 55 min. ROS levels were measured through a 2′,7′‐Dichlorodihydrofluorescein diacetate (DCFDA) assay.

The effects of the CTSD inhibitors on oxidative stress‐induced lipotoxicity were evaluated by measuring malondialdehyde (MDA) levels and gene expression of Blvrb and Srxn1. These experiments were conducted in intact liver tissue from HFC‐fed mice treated with vehicle, intra‐ or extracellular CTSD inhibitors, as well as AML‐12 cells. MDA, a marker of lipid peroxidation, was measured to assess oxidative stress. The results showed no significant differences in MDA levels between the vehicle/control‐treated group and CTSD inhibitor‐treated groups in both liver tissue and AML‐12 cells, respectively (Fig. 5A,B). In addition, gene expression of oxidative stress‐induced genes Blvrb and Srxn1 was measured. While higher gene expression of Blvrb was observed in the intracellular CTSD‐treated cells compared to both control (P < 0.05) and extracellular CTSD inhibitor (P = 0.07) in AML‐12 cells, no significant differences were found in the treated liver tissue (Fig. 5C,D). A similar trend was observed for Srxn1, where gene expression was higher in the intracellular CTSD inhibitor‐treated cells compared to the extracellular CTSD inhibitor group in AML‐12 cells (P = 0.07), yet no differences were observed in the treated liver tissue (Fig. 5E,F).

Fig. 5 — Oxidative stress measurements in response to treatment with CTSD inhibitors. A malondialdehyde (MDA) assay was performed in both (A) intact liver tissue from the mice fed with HFC diet treated with vehicle, intracellular CTSD inhibitor, or extracellular CTSD inhibition, and (B) AML‐12 cells treated with 0.1% DMSO (Control), 100 μm of the intracellular CTSD inhibitor or, 100 μm of the extracellular CTSD inhibitor; gene expression of *Blvrb* was measured using RT‐qPCR in both (C) intact liver tissue treated with the CTSD inhibitors and (D) AML‐12 cells; gene expression of *Srxn1* was measured using RT‐qPCR in both (E) intact liver tissue treated with the CTSD inhibitors and (F) AML‐12 cells. Statistical differences were tested with a one‐way ANOVA with Tukey's *post hoc* test. * P < 0.05. Data are displayed as mean ± SEM.

Lastly, to determine the effect of the CTSD inhibitors on liver cell viability, we assessed the viability of hepatocyte cells in vitro using the AML‐12 cell line in response to the extra‐ and intracellular CTSD inhibitors. The extracellular CTSD inhibitor significantly reduced the cell viability compared to the control (P < 0.0001) and the intracellular CTSD inhibitor (P < 0.01) (Fig. S4).

Discussion

In this study, we aimed to understand the lipid profile upon the treatment of intracellular and extracellular CTSD inhibition. Our previous proteomics study revealed that intracellular CTSD inhibition downregulated proteins involved in mitochondrial oxidative phosphorylation and the electron transport chain. Furthermore, we observed an enrichment of hepatic lipid metabolic pathways in extracellular CTSD inhibitor‐treated mice, including linoleic acid, steroid hormone biosynthesis, and fatty acid [14]. Building upon these results, the current study focused on an in‐depth analysis of hepatic lipid composition in Ldlr ^−/− mice fed a high‐fat diet.

Our results demonstrated that mice treated with both intracellular and extracellular CTSD inhibitors induced remodeling of similar phosphatidylcholines (PCs) compared to vehicle‐treated mice. The anti‐inflammatory properties of PCs have been investigated as potential targets for managing inflammation in various chronic conditions, such as inflammatory bowel disease, Alzheimer's disease, and liver diseases, such as MASLD [21, 22, 23, 24]. Several studies have revealed a link between modulation of nuclear factor kappa B (NF‐κB), a crucial immune response regulator, and the anti‐inflammatory effects of PCs. More specifically, PCs play a role in preventing the translocation of NF‐κB to the nucleus. This inhibition prevents the activation of numerous genetic responses, such as cytokines, which are responsible for the secretion of proinflammatory mediators [22, 25]. In addition, PCs also have anti‐steatotic properties and are currently investigated as a therapeutic target in the context of MASH [26]. In line with these data, we also found an enrichment of phosphatidylinositol's (PIs) in both inhibitor‐treated groups compared to the vehicle‐treated group. PIs engage in signaling pathways related to liver function, including the PI3K/AKT phosphorylation pathway and the NF‐κB pathway [27]. PIs play crucial roles in modulating the activity of macrophages and influencing the inflammatory response within the liver [28]. These regulatory effects extend to immune cell activation, cytokine production, and the broader inflammatory environment associated with liver disease. Several studies have shown that PIs have anti‐inflammatory effects [29]. The observed results imply that the enrichment of PCs and PIs may contribute to the potential anti‐inflammatory effects of intracellular and extracellular CTSD inhibition. This observation is in line with a previous study showing that both inhibitors triggered anti‐inflammatory effects on monocytes, whereas the intracellular CTSD inhibitor showed a mild effect compared with the extracellular CTSD inhibitor [14].

The vehicle‐treated group showed a specific increase in only one secondary bile acid, taurallocholic acid, derived from cholesterol, compared with the inhibitor‐treated groups. Secondary bile acids, which form microbial action on primary bile acids, play a crucial role in emulsifying fats, aiding their digestion and absorption [30, 31]. However, dysregulation of bile acid metabolism has been linked to increased risks of MASH [32]. The accumulation of bile acids in the liver can lead to excessive production of toxic metabolites and cytokines that play roles in hepatic inflammation and fibrosis, which exacerbates the progression of MASLD/MASH [32]. Based on our data, CTSD inhibitors may potentially reduce hepatic injury by regulating bile acids in the liver. However, further investigation of the broader spectrum of bile acids is needed to fully understand their role in this process and to enhance our understanding of CTSD inhibition.

Next, we detected lipid changes between the intracellular and extracellular CTSD inhibitor‐treated groups. We demonstrated that intracellular CTSD inhibition induced more potential side effects than extracellular CTSD inhibition through alterations in different lipid classes that are linked to mitochondrial dysfunction or negatively linked to inflammation, such as cardiolipins (CLs), phosphatidylethanolamines (PEs), phosphatidic acids (PAs), and phosphatidylglycerols (PGs). Interestingly, no differences were found in the spatial distributions of different lipid classes. Although previous research has shown that Ldlr deficiency plays a significant role in lipid metabolism, the correlation is multifaceted [33]. Further research into the mechanism underlying lipid droplet formation in the Ldlr ^−/− mouse model, particularly in the context of MASH, would be valuable for investigating this complex process.

In mice treated with an extracellular CTSD inhibitor, reduced levels of cardiolipins (CLs) were observed compared to those treated with an intracellular CTSD inhibitor. CLs are a group of functional lipids that are found exclusively in the inner mitochondrial membrane, where they play essential roles in various mitochondrial processes, including respiration and energy production [34]. CLs dysfunction has been linked to the onset of mitochondrial disorders and diseases related to oxidative stress and apoptosis, including obesity and MASLD [20, 35]. Oxidation and peroxidation of CLs contribute to mitochondrial dysfunction, leading to the overproduction of reactive oxygen species (ROS) and disruption of oxidative phosphorylation [20].

Furthermore, enrichment of phosphatidylethanolamines (PEs) in intracellular CTSD inhibitor‐treated mice compared to extracellular CTSD inhibitor‐treated mice was also observed. Similar to CLs, PEs are notably present in the inner membrane of mitochondria to maintain mitochondrial function [36]. Excessive levels of certain phospholipids, including PEs, may affect the mitochondrial dynamics, respiratory chain activity, and glucose metabolism. These changes can affect mitochondrial function, including energy production and increased susceptibility to oxidative stress [37].

Besides affecting the mitochondria, lipid dysfunction can also induce endoplasmic reticulum (ER) stress within the hepatocytes [38]. Several studies have shown that an imbalance of specifically the PC/PE ratio could lead to ER stress, which induced the unfolded protein response (UPR) in hepatocytes [38, 39, 40, 41]. The UPR can result in inflammation, disturbed lipid metabolism, and hepatocyte apoptosis [42]. Moreover, studies have shown that decreased hepatic PC/PE ratio was strongly associated with the progress of MASLD. In MASLD/MASH patients, it has also been demonstrated that they have a lower PC/PE ratio compared to healthy controls [23, 43]. In the current study, there was an imbalance of PC/PE ratio due to enrichment of PE levels in the intracellular CTSD inhibitor‐treated group, suggesting that besides mitochondrial stress, ER stress could be induced by this inhibitor as well. We also identified a higher abundance of potassium adducts of phosphatidylcholine lipids in the intracellular inhibitor‐treated group than in the extracellular inhibitor‐treated group. Imbalance of potassium can contribute to oxidative stress and tissue damage in various pathophysiological diseases. Oxidative stress has a notable impact on ion channels, specifically voltage‐gated potassium channels, leading to modifications in their gating properties and ion selectivity [44, 45].

In human macrophages, treatment with the intracellular CTSD inhibitor induced overproduction of ROS levels compared to both the extracellular CTSD inhibitor group and control group. Mechanistically, inhibition of intracellular CTSD, which disturbs the physiological environment, may lead to increased ROS production and subsequent mitochondrial dysfunction [9]. In addition, targeting extracellular CTSD, which is typically associated with pathological conditions [9], implies a lower extent of cellular damage. In contrast to macrophages, ROS levels were not significantly changed in murine hepatocytes upon incubation with any of the CTSD inhibitors. These data demonstrate that the toxic effects of intracellular CTSD inhibition are primarily mediated by macrophages, although it remains possible that hepatocytes may possess a capacity to counteract the ROS‐induced effects of both CTSD inhibitors. This is in line with other studies demonstrating that some drugs specifically lead to toxicity by inducing ROS in macrophages [46, 47], which causes oxidative stress and triggers inflammatory pathways [48]. While macrophages secrete ROS as a natural immune response, excessive secretion can lead to oxidative stress and inflammation [49]. Altogether, our results are consistent with those of a previous study showing that intracellular CTSD inhibitor regulated the proteins involved in mitochondrial function, which indicates that intracellular CTSD inhibitor may disrupt the function of mitochondria that are susceptible to ROS attack, thereby leading to toxicity [14].

Moreover, our findings revealed a notable increase in phosphatidylglycerols (PGs) and phosphatidic acids (PAs) in the group treated with the extracellular CTSD inhibitor compared to the group treated with the intracellular CTSD inhibitor. Previous studies explored the inhibitory effects of PGs on inflammation. PGs can effectively inhibit the production of inflammatory mediators mediated by toll‐like receptors (TLR2 and TLR4) [50, 51]. Additionally, both PGs and PAs can integrate into CLs, leading to enhanced mitochondrial activity and inflammation inhibition [50, 52]. Collectively, these data suggest that the extracellular CTSD inhibitor has more beneficial effects than the intracellular CTSD inhibitor by protecting mitochondrial function and reducing inflammation in MASH Ldlr ^−/− mice by modulating specific classes of lipids. The presence of both anti‐inflammatory and inflammatory lipids in the intracellular inhibitor suggests a complex regulatory interplay within the liver, reflecting the dynamic balance between the pro‐ and anti‐inflammatory processes. In summary, despite the identification of anti‐inflammatory lipids in the intracellular CTSD inhibitor, the complex regulatory dynamics within the liver suggest that its efficacy in attenuating inflammation may be limited compared with that of extracellular inhibitor. Further research is needed to elucidate the potential failure of intracellular CTSD inhibitor to manage inflammatory responses.

Furthermore, the effects of the CTSD inhibitors on lipotoxicity were measured. The intracellular CTSD inhibitor upregulated Srxn1 and Blvrb, genes that are known to be upregulated by oxidative stress, compared to extracellular CTSD inhibitor, suggesting that the intracellular CTSD inhibitor induced more oxidative stress. Additionally, the extracellular CTSD inhibitor significantly reduced cell viability in AML‐12 cells compared to the control and the intracellular CTSD inhibitor‐treated cells. These data are contradictory to earlier findings that indicated that the extracellular CTSD inhibitor induced less toxicity compared to the intracellular CTSD inhibitor. This decrease in cell viability is likely not related to toxicity but rather to the inhibitory effect of extracellular CTSD on proliferation in immortalized cell lines [8]. This pathological role of extracellular cathepsins in cancer is in line with other findings describing that extracellular cathepsins play a significant role in liver cancer and regulate numerous physiological processes, such as cell proliferation, migration, and angiogenesis [8].

In the current study, the identification of CTSD as a metabolic regulator has potential implications for enhancing MASH treatment. Given the role of extracellular CTSD in MASH, strategies centered around inhibiting extracellular CTSD may offer a promising alternative to existing anti‐steatosis and anti‐inflammatory agents with fewer side effects. To fully establish extracellular CTSD inhibition as a viable robust therapeutic option for MASH, it is imperative to validate these results in cohorts and to conduct more comprehensive investigations into the mechanisms of CTSD. Nevertheless, our findings suggest that extracellular CTSD holds promising clinical value, potentially paving the way for innovative approaches for MASH treatment.

The changes in lipid composition observed through MALDI‐MSI and lipotoxicity effects upon CTSD inhibitors suggest the potential clinical advantages of extracellular CTSD inhibition as a novel approach to prevent MASH development. Further investigations focused on evaluating the safety in a broader range of preclinical models are warranted to confirm our findings.

Materials and methods

Chemicals and reagents

Norharmane, methanol, chloroform, Gill's hematoxylin, and xylene were obtained from Sigma‐Aldrich (Zwijndrecht, the Netherlands). The ITO glass slides were obtained from Delta Technologies (Loveland, CO, USA). Hydrogen peroxide (H₂O₂) was obtained from Merck KgAA (Darmstadt, Germany). The intracellular CTSD (GA‐12) and extracellular CTSD (CTD‐002) inhibitors were designed by Aten Porus Lifesciences Pvt Ltd., Bengaluru, India [14].

Mice, diet, and intervention

As described previously, female Ldlr ^−/− mice (8–14 weeks of age), on a C57BL/6 background, sourced from Maastricht University and bred in‐house were fed a high‐fat, high‐cholesterol (HFC) diet for 10 weeks and were divided into three groups (n = 3 per group) [14]. HFC‐fed mice were subcutaneously injected with intracellular CTSD inhibitor (GA‐12, 50 mg·kg⁻¹ body weight), extracellular CTSD inhibitor (CTD‐002, 50 mg·kg⁻¹ body weight), or vehicle treatment (5% DMSO, 40% PEG400, 10% Ethanol, 45% Saline) every 2 days for 10 weeks. Animals were housed socially in groups of three, with standard cage enrichment and food and water ad libitum. All the mice in the same cage were allocated to the same experimental group. Persons who were performing the injections were blinded. The well‐being of the mice was assessed daily using standard scoring. Within the group of mice presented in this article, no animals had to be excluded based on humane endpoints and no adverse events were observed. Collection of tissue specimens was performed as described previously [7, 14]. The animal study was reviewed according to Dutch regulations and approved by the Committee for Animal Welfare of Maastricht University (project license number: AVD107002016743; working protocol: 2016‐003‐003).

Sample preparation for MALDI‐MSI

Cryosectioning

Tissue samples were cut using a cryostat (Leica Biosystems, Nußloch, Germany) into 10 μm thick sections at −20 °C. Liver sections were mounted onto an indium tin oxide (ITO) glass slide (4–8 Ω resistance; Delta Technologies, Loveland, CO, USA). All slides were stored at −80 °C until further analysis. Liver sections from each condition were randomly placed on the ITO slides.

Matrix application

Slides were transported from −80 °C to a silica carrier box and then vacuum‐dried for 15 min to ensure that the tissue was dry before matrix application. An HTX‐TM sprayer (HTX Technologies, LLC, Carrboro, NC, USA) was used to spray 15 layers of norhamane (7 mg·mL⁻¹) in methanol/chloroform (1 : 2, v/v) onto the tissue sections. The following parameters were used: temperature of 30 °C, 0.12 mL·min⁻¹ flow rate, drying time of 30s, between each layer, 3 mm tracking space, and 1200 mm·s⁻¹ velocity. The gas pressure was set at 10 psi with a flow rate of N₂ of 3 L·min⁻¹.

MALDI‐MSI

Imaging of lipids

Lipid mass spectra were acquired on a tims‐TOF fleX (Bruker Daltonics Inc., Bremen, Germany) in both positive and negative polarity. For each condition (n = 9, three mice per condition), a pixel size of 20 μm × 20 μm was employed. Following the first analysis in the negative mode, the identical sample section was subsequently re‐analyzed in the positive mode, with an offset of 10 μm in the x‐direction to avoid sampling the same location.

Identification of lipids

Additional MALDI‐MSI MS/MS studies were performed on consecutive sections to structurally identify the lipids of interest using an Orbitrap‐Elite hybrid ion trap MS instrument (Thermo Fisher Scientific GmbH, Bremen, Germany) in data‐dependent acquisition (DDA) mode. These experiments were also performed for both positive and negative polarities. MS1 data were collected at a resolution of 240 000 at 400 m/z over a mass range of m/z 200–2000. The MS2 data were collected in an ion trap with an isolation width of 0.7 m/z. In the negative mode, a collision energy of 38 eV was used, whereas in the positive mode, a collision energy of 30 eV was used. The stage step size was set to 25 μm (horizontal) × 50 μm (vertical).

Histological staining

Consecutive slides of liver tissue used for MALDI‐MSI of each group (vehicle‐treated, intracellular CTSD inhibitor (GA‐12), extracellular CTSD inhibitor (CTD‐002)‐treated) were immersed in distilled tap water for 3 min, after which they were subjected to hematoxylin staining (0.1% Gill's solution) for 3 min. The slides were a 3‐min rinsing step in running tap water, followed by a brief rinse in distilled water. Eosin staining (0.2%) was applied for 30 s, followed by a short rinse in 70% ethanol to remove excess eosin. The next step involved two rounds of dehydration in 100% ethanol, each lasting 2 min, followed by equilibration in xylene for 5 min (2×). To preserve the stained sections, they were mounted using Entellan, covered with a glass coverslip, and left to dry at room temperature. Finally, stained sections were scanned at 20× magnification using a digital scanner (Aperio CS2; Leica Biosystems, Nußloch, Germany).

Cell culture

Peripheral blood mononuclear cells (PBMCs) were isolated from leukocyte reduction system cones, a by‐product of thrombopheresis of routine blood donations from healthy volunteers collected from the Blood Bank (RWTH University Hospital Aachen, Germany), by density centrifugation with Ficoll. The leukocyte reduction system cones were collected from healthy volunteers according to the local regulations of the RWTH Aachen. Experiments were conducted according to the Declaration of Helsinki and each healthy volunteer provided written consent. Monocytes were subsequently isolated from PBMCs, as described previously [53]. The samples were collected in November 2023. Cells were cultured in RPMI‐1640 medium (GIBCO, Invitrogen, Breda, the Netherlands), supplemented with 10% fetal bovine serum (FBS) (Bodinco B.V., Alkmaar, the Netherlands), Glutamax™ 2 mm (GIBCO, Life Technologies Limited, Paisley, UK), sodium pyruvate 1 mm (GIBCO, Life Technologies Corporation, New York, NY, USA), and gentamycin 50 μg·mL⁻¹ (GIBCO, Life Technologies Corporation), for 8–9 days to generate macrophages.

The murine AML‐12 (alpha mouse liver 12; immortalized cell line derived from 3‐month‐old mouse liver) (RRID: CVCL_0140) cell line (ATCC CRL‐2254) were routinely cultured in 75 cm² plastic tissue culture flasks at 37 °C with 5% CO₂, using DMEM‐F12 (GIBCO, Invitrogen) cell culture medium supplemented with 10 μg·mL⁻¹ Human Recombinant Insulin (Sigma‐Aldrich), 5.5 μg·mL⁻¹ transferrin (Sigma‐Aldrich), 5 ng·mL⁻¹ sodium selenite (Sigma‐Aldrich), 40 ng·mL⁻¹ dexamethasone (Sigma‐Aldrich), and 10% fetal calf serum (Bodinco B.V.). All experiments were performed with mycoplasma‐free cells.

ROS assay

Macrophages and AML‐12 cells were seeded at 100 000 cells/well and 6000 cells/well, respectively, in 96 well plates and starved overnight in complete medium without serum. Cells were subsequently treated with 100 μm of the intracellular CTSD inhibitor (GA‐12), 100 μm of the extracellular CTSD inhibitor (CTD‐002) (Aten Porus Lifesciences Pvt Ltd.), or negative control (HBBS^+/+; GIBCO, Invitrogen with 0.1% DMSO) and subjected to a 2′,7′‐Dichlorofluorescin Diacetate (DCFDA) assay according to the manufacturer's protocol (Sigma‐Aldrich). The readings were recorded every minute using a fluorescent plate reader (Spark 10M plate reader; Tecan, Männedorf, Switzerland) for 55 min.

MDA (malondialdehyde) measurement

Homogenized liver tissue and cell lysate were used to assess malondialdehyde (MDA). AML‐12 cells were seeded at 200 000 cells/well in six well plates. The liver tissue weight of the HFC mice ranged from 55 to 120 mg, and the MDA data were ultimately adjusted based on the total protein content. The MDA assay was performed as previously described [54]. In short, cell lysate or MDA standard was diluted 10× with a reagent mix. This mix consisted of 10 parts reagent A (0.012 m 2‐thiobarbituric acid (TBA), 0.32 m H₃PO₄, and 0.01% EDTA) and 1 part reagent B (butylated hydroxytoluene in ethanol, at 1.5 mg·mL⁻¹). The MDA standards were prepared using 0–10 μm MDA solutions in PBS and treated identically to the samples. The mixtures were heated at 99 °C for 1 h. After cooling, the product was extracted into 500 μL butanol by vigorous shaking, followed by centrifugation at 16 260 g for 3 min. Subsequently, 100 μL of the butanol extract was transferred to a black 96‐well plate for fluorescence measurement, with excitation at 530 nm and emission at 560 nm, using the Spectramax ID3 (Molecular Devices, San jose, CA, USA).

Quantitative real‐time PCR analysis

RNA isolation, cDNA synthesis, and the real‐time quantitative PCR (qPCR) were determined as described previously [55]. In short, RNA of cell pellet and homogenized tissue sample were isolated by Tri Reagent (Sigma‐Aldrich). cDNA was synthesized from 3 to 20 ng RNA·μL⁻¹. After preparing the cDNA, it was diluted 10‐fold. For each gene, a qPCR mastermix was prepared by mixing the forward (FW) and reverse (RV) primers (Eurofins, Heerenveen, the Netherlands) with Sensimix (Bioline, Deventer, the Netherlands). Each well was filled with 5 μm FW primer, 5 μm RV primer, and Sensimix. Subsequently, diluted cDNA and mastermix were pipetted into each well. The plate was then covered with a plastic seal and mixed gently and quickly spun down to ensure all liquids were collected at the bottom of the wells.

The thermal cycling program using Bio‐Rad Opus CFX384 (Bio‐Rad, Hercules, CA, USA) for the reaction included the following steps: an initial enzyme activation at 95 °C for 10 min, followed by 45 cycles of denaturation at 95 °C for 10 s, and annealing/elongation at 60 °C for 20 s. After these cycles, a melting curve was performed from 60 °C to 95 °C at a rate of 0.11 °C·s⁻¹ with five acquisitions per second. Finally, the reaction mixture was cooled from 95 °C to 40 °C. The primer sequences used for Srxn1‐F, Blvrb, B2M, and CycloA are described in Table S8. The data were analyzed using Bio‐Rad CFX Maestro. Expression of the gene Srxn1 and Blvrb were normalized to B2M and CycloA.

Cell viability assay

AML‐12 cells were plated in 96‐well plates at a density of 6000 cells/well in supplemented medium with insulin, transferrin, sodium selenite, and dexamethasone. The cells were treated with 100 μm of the intracellular CTSD inhibitor (GA‐12), 100 μm of the extracellular CTSD inhibitor (CTD‐002), or negative control (0.1% DMSO) for 72 h. Cell viability was tested by a Sulforhodamine B (SRB) assay. First, cells were fixed by trichloroacetic acid (TCA; supplemented with 5% FBS medium; Sigma‐Aldrich) for 1 h at 4 °C. Subsequently, 0.4% SRB in 1% acetic acid was added to the fixed cells and was incubated at room temperature for 30 min in the dark. Next, the plate was rinsed under running tap water and subsequently, in 1% acetic acid. Lastly, the SRB dye was solubilized in unbuffered 10 mm trishydroxymethyl aminomethane (Sigma‐Aldrich) on a shaker for 1 h in the dark. Absorbance was measured at 490 nm using a benchmark microplate reader (Bio‐Rad, Veenendaal, the Netherlands).

Data analysis

Peak picking was performed as a preprocessing step on both datasets separately (positive and negative polarity). Principal component analysis (PCA) and linear discriminant analysis (LDA) were performed on both datasets using the in‐house built ChemomeTricks toolbox for matlab version 2014a (The MathWorks, Natick, MA, USA).

The regions of interest (ROI) were determined by only including on‐tissue pixels and categorized according to each experimental condition, namely vehicle, intracellular inhibitor treatment, and extracellular inhibitor treatment (n = 9, 3 mice per condition). Next, PCA‐LDA was performed on both the positive and negative full datasets, where PCA was performed to reduce the dimensionality of the data, followed by LDA to determine the differences between the groups [56]. Discriminant functions (DF) are composite variables designed to maximize the variance between groups, while minimizing the variance within each group. The scores corresponding to distinct discriminant functions were calculated for each spectrum (pixel), and histograms were constructed for each group by plotting them along the DF1 axis (Figs S1 and S3). We selected the top 20 scaled loadings for each condition, where the scaled loading was multiplied by the variance of the respective m/z value. This enabled us to understand how each m/z value or lipid contributed to the separation of conditions across DF1 considering their relative intensities. We identified distinct lipids by further focusing on those with a non‐scaled loading above 0.1 or below −0.1, as this criterion ensured that the lipids selected not only had high intensity but also made significant contributions to the differences observed between the conditions.

To identify specific lipids, the parent ion masses obtained from the raw full MSI1 scans were paired with their corresponding MS2 scans. All MS/MS files were then converted into the imzML format and imported into LipostarMSI for subsequent analysis. The Lipid Maps database (edition July 2020) served as a reference database for the lipid search [57]. [M‐H]⁻ or [M‐H]⁺, including Na⁺ and K⁺, were selected as the basis for database queries, with a mass tolerance of 0.0 Da ± 3 p.p.m. For MS/MS analyses, the precursor [M‐H]⁻ ion or [M‐H]⁺ ion was subjected to fragment matching with an m/z tolerance of 0.25 Da ± 0 p.p.m. Only lipids with a minimum carbon chain length of 12 were considered for the identification.

To test statistical differences in ROS levels, cell viability, MDA, and gene expression, a one‐way ANOVA was performed in graphpad prism 9.0, San Diego, CA, USA. The data were expressed as mean ± SEM and considered significant at P < 0.05.

Conflict of interest

AK is a co‐founder of Avaliv Therapeutics, which has filed intellectual property to protect inhibitors. The terms of this arrangement were reviewed and approved by Maastricht University, aligning with its policy on research objectivity. AK is an employee of Aten Porus Lifesciences and is responsible for chemistry‐related work for inhibitor development. The authors have no conflicts of interest to disclose.

Author contributions

The study was conceived and designed by IV, TY, BC‐P, and RS‐S. IV conducted MALDI‐MSI experiments and performed statistical analyses of the lipid data. ML, HM, TY, and TH were involved in the sample collection, ROS experiments, cell viability assay, and subsequent analyses. RG provided oxidative stress‐induced lipotoxicity data. LT and EALB facilitated the availability of human PBMCs. AK provided inhibitors for this study. AK and JT assisted in the review of the manuscript. GE and BB aided in the analysis of lipid data. IV, ML, and HM were responsible for drafting and writing the manuscript, with input from all co‐authors. A critical final review of the study was performed by JT, TH, BC‐P, and RS‐S. All the authors approved the final version of the manuscript.

Peer review

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/febs.17358.

Supporting information

Fig. S1. Score distribution plot (histogram) of discriminant function 1 (DF1) in negative mode.

Fig. S2. Hematoxylin and eosin staining.

Fig. S3. Score distribution plot (histogram) of discriminant function (DF1) in positive mode.

Fig. S4. Cell viability analysis using Sulforhodamine B (SRB) assay.

Table S1. Overview of the differential lipids found in positive mode MALDI‐MSI between the extracellular inhibitor‐treated group and the vehicle‐treated group.

Table S2. Overview of the differential lipids found in positive mode MALDI‐MSI between the intracellular inhibitor‐treated group and the vehicle‐treated group.

Table S3. Overview of the differential lipids found in positive mode MALDI‐MSI between the intracellular inhibitor‐treated group and the extracellular inhibitor‐treated group.

Table S4. Overview of the differential lipids found in negative mode MALDI‐MSI between the extracellular inhibitor‐treated group and the vehicle‐treated group.

Table S5. Overview of the differential lipids found in negative mode MALDI‐MSI between the intracellular inhibitor‐treated group and the vehicle‐treated group.

Table S6. Overview of the differential lipids found in negative mode MALDI‐MSI between the intracellular inhibitor‐treated group and the extracellular inhibitor‐treated group.

Table S7. Overview of the differential cardiolipins found in negative mode MALDI‐MSI between the intracellular inhibitor‐treated group and the extracellular inhibitor‐treated group.

Table S8. Forward and reverse primer sequences of genes used for RT‐qPCR.

FEBS-292-1781-s001.pdf^{(672.8KB, pdf)}

Acknowledgements

The authors would like to thank Dennis Meesters for his guidance in finalizing the ARRIVE guidelines. We would like to thank Phyllis Jessen for her contribution to performing oxidative stress experiments. Financial support was provided by the Dutch Province of Limburg under the LINK program as part of the M4i research program and by LINK 2.0, as part of the MERLN research program. ML was supported by the Chinese Scholarship Council (file number CSC202007550013). The authors used Large Language Model (ChatGPT) to refine the textual content. The authors reviewed and edited the ChatGPT‐generated output and take full responsibility for the content of the publication.

Isabeau Vermeulen and Mengying Li contributed equally to this work and shared first authorship.

Berta Cillero‐Pastor and Ronit Shiri‐Sverdlov contributed equally to this work and shared their last authorship.

Data availability statement

All data are included in this article/Supporting Information or uploaded in DataHub. Further inquiries can be directed to the corresponding authors.

References

1. Rinella ME, Lazarus JV, Ratziu V, Francque SM, Sanyal AJ, Kanwal F, Romero D, Abdelmalek MF, Anstee QM, Arab JP et al. (9900) A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology 78, 1966–1986. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Zhu Y, Yang H, Liang S, Zhang H, Mo Y, Rao S, Zhang Y, Zhang Z, Wang W & Yang W (2022) Higher adherence to healthy lifestyle score is associated with lower odds of non‐alcoholic fatty liver disease. Nutrients 14, 4462. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Hamid O, Eltelbany A, Mohammed A, Alsabbagh Alchirazi K, Trakroo S & Asaad I (2022) The epidemiology of non‐alcoholic steatohepatitis (NASH) in the United States between 2010–2020: a population‐based study. Ann Hepatol 27, 100727. [DOI] [PubMed] [Google Scholar]
4. Wong VW‐S, Ekstedt M, Wong GL‐H & Hagström H (2023) Changing epidemiology, global trends and implications for outcomes of NAFLD. J Hepatol 79, 842–852. [DOI] [PubMed] [Google Scholar]
5. Keam SJ (2024) Resmetirom: first approval. Drugs 84, 1–7. [DOI] [PubMed] [Google Scholar]
6. Ruiz‐Blazquez P, Pistorio V, Fernandez‐Fernandez M & Moles A (2021) The multifaceted role of cathepsins in liver disease. J Hepatol 75, 1192–1202. [DOI] [PubMed] [Google Scholar]
7. Houben T, Oligschlaeger Y, Hendrikx T, Bitorina AV, Walenbergh SM, van Gorp PJ, Gijbels MJ, Friedrichs S, Plat J & Schaap FG (2017) Cathepsin D regulates lipid metabolism in murine steatohepatitis. Sci Rep 7, 3494. [DOI] [PMC free article] [PubMed] [Google Scholar]
8. van Mourik H, Li M, Baumgartner S, Theys J & Shiri‐Sverdlov R (2022) All roads lead to cathepsins: the role of cathepsins in non‐alcoholic steatohepatitis‐induced hepatocellular carcinoma. Biomedicine 10, 2351. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Yadati T, Houben T, Bitorina A & Shiri‐Sverdlov R (2020) The ins and outs of cathepsins: physiological function and role in disease management. Cells 9, 1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Vidak E, Javoršek U, Vizovišek M & Turk B (2019) Cysteine cathepsins and their extracellular roles: shaping the microenvironment. Cells 8, 264. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Walenbergh SM, Houben T, Hendrikx T, Jeurissen ML, van Gorp PJ, Vreugdenhil AC, Adriaanse MP, Buurman WA, Hofker MH, Mosca A et al. (2015) Plasma cathepsin D levels: a novel tool to predict pediatric hepatic inflammation. Am J Gastroenterol 110, 462–470. [DOI] [PubMed] [Google Scholar]
12. Ding L, Houben T, Oligschlaeger Y, Bitorina AV, Verwer BJ, Tushuizen ME & Shiri‐Sverdlov R (2020) Plasma cathepsin D activity rather than levels correlates with metabolic parameters of type 2 diabetes in male individuals. Front Endocrinol (Lausanne) 11, 575070. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Khurana P, Yadati T, Goyal S, Dolas A, Houben T, Oligschlaeger Y, Agarwal AK, Kulkarni A & Shiri‐Sverdlov R (2019) Inhibiting extracellular cathepsin D reduces hepatic steatosis in Sprague–Dawley rats. Biomolecules 9, 171. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Yadati T, Houben T, Bitorina A, Oligschlaeger Y, Gijbels MJ, Mohren R, Lutjohann D, Khurana P, Goyal S, Kulkarni A et al. (2021) Inhibition of extracellular cathepsin D reduces hepatic lipid accumulation and leads to mild changes in inflammationin NASH mice. Front Immunol 12, 675535. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Van Hese L, De Vleeschouwer S, Theys T, Rex S, Heeren RMA & Cuypers E (2022) The diagnostic accuracy of intraoperative differentiation and delineation techniques in brain tumours. Discov Oncol 13, 123. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Barre FPY, Claes BSR, Dewez F, Peutz‐Kootstra C, Munch‐Petersen HF, Gronbaek K, Lund AH, Heeren RMA, Come C & Cillero‐Pastor B (2018) Specific lipid and metabolic profiles of R‐CHOP‐resistant diffuse large B‐cell lymphoma elucidated by matrix‐assisted laser desorption ionization mass spectrometry imaging and in vivo imaging. Anal Chem 90, 14198–14206. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Kampa JM, Sahin M, Slopianka M, Giampa M, Bednarz H, Ernst R, Riefke B, Niehaus K & Fatangare A (2020) Mass spectrometry imaging reveals lipid upregulation and bile acid changes indicating amitriptyline induced steatosis in a rat model. Toxicol Lett 325, 43–50. [DOI] [PubMed] [Google Scholar]
18. Strindberg S, Plum J, Bagger C, Janfelt C & Mullertz A (2021) Visualizing the journey of fenofibrate through the rat gastrointestinal tract by matrix‐assisted laser desorption/ionization‐mass spectrometry imaging. Mol Pharm 18, 2189–2197. [DOI] [PubMed] [Google Scholar]
19. Amoscato AA, Sparvero LJ, He RR, Watkins S, Bayir H & Kagan VE (2014) Imaging mass spectrometry of diversified cardiolipin molecular species in the brain. Anal Chem 86, 6587–6595. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Paradies G, Paradies V, Ruggiero FM & Petrosillo G (2019) Role of cardiolipin in mitochondrial function and dynamics in health and disease: molecular and pharmacological aspects. Cells 8, 728. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Ai R, Xu J, Ji G & Cui B (2022) Exploring the phosphatidylcholine in inflammatory bowel disease: potential mechanisms and therapeutic interventions. Curr Pharm Des 28, 3486–3491. [DOI] [PubMed] [Google Scholar]
22. Treede I, Braun A, Sparla R, Kühnel M, Giese T, Turner JR, Anes E, Kulaksiz H, Füllekrug J, Stremmel W et al. (2007) Anti‐inflammatory effects of phosphatidylcholine. J Biol Chem 282, 27155–27164. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Arendt BM, Ma DW, Simons B, Noureldin SA, Therapondos G, Guindi M, Sherman M & Allard JP (2013) Nonalcoholic fatty liver disease is associated with lower hepatic and erythrocyte ratios of phosphatidylcholine to phosphatidylethanolamine. Appl Physiol Nutr Metab 38, 334–340. [DOI] [PubMed] [Google Scholar]
24. Whiley L, Sen A, Heaton J, Proitsi P, García‐Gómez D, Leung R, Smith N, Thambisetty M, Kloszewska I & Mecocci P (2014) Evidence of altered phosphatidylcholine metabolism in Alzheimer's disease. Neurobiol Aging 35, 271–278. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Cauvi DM, Hawisher D, Derunes J & De Maio A (2023) Phosphatidylcholine liposomes reprogram macrophages toward an inflammatory phenotype. Membranes (Basel) 13, 141. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Osipova D, Kokoreva K, Lazebnik L, Golovanova E, Pavlov C, Dukhanin A, Orlova S & Starostin K (2022) Regression of liver steatosis following phosphatidylcholine administration: a review of molecular and metabolic pathways involved. Front Pharmacol 13, 797923. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Wang M, Zhang J & Gong N (2022) Role of the PI3K/Akt signaling pathway in liver ischemia reperfusion injury: a narrative review. Ann Palliat Med 11, 806–817. [DOI] [PubMed] [Google Scholar]
28. Wassef NM, Roerdink F, Richardson EC & Alving CR (1984) Suppression of phagocytic function and phospholipid metabolism in macrophages by phosphatidylinositol liposomes. Proc Natl Acad Sci USA 81, 2655–2659. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. van Dieren JM, Simons‐Oosterhuis Y, Raatgeep HC, Lindenbergh‐Kortleve DJ, Lambers ME, van der Woude CJ, Kuipers EJ, Snoek GT, Potman R, Hammad H et al. (2011) Anti‐inflammatory actions of phosphatidylinositol. Eur J Immunol 41, 1047–1057. [DOI] [PubMed] [Google Scholar]
30. Alberto González‐Regueiro J, Moreno‐Castañeda L, Uribe M & Carlos Chávez‐Tapia N (2017) The role of bile acids in glucose metabolism and their relation with diabetes. Ann Hepatol 16, S15–S20. [DOI] [PubMed] [Google Scholar]
31. Jiao TY, Ma YD, Guo XZ, Ye YF & Xie C (2022) Bile acid and receptors: biology and drug discovery for nonalcoholic fatty liver disease. Acta Pharmacol Sin 43, 1103–1119. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Yu Q, Jiang Z & Zhang L (2018) Bile acid regulation: a novel therapeutic strategy in non‐alcoholic fatty liver disease. Pharmacol Ther 190, 81–90. [DOI] [PubMed] [Google Scholar]
33. Bieghs V, Van Gorp PJ, Wouters K, Hendrikx T, Gijbels MJ, van Bilsen M, Bakker J, Binder CJ, Lutjohann D, Staels B et al. (2012) LDL receptor knock‐out mice are a physiological model particularly vulnerable to study the onset of inflammation in non‐alcoholic fatty liver disease. PLoS One 7, e30668. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Dudek J (2017) Role of cardiolipin in mitochondrial signaling pathways. Front Cell Dev Biol 5, 90. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. El‐Hafidi M, Correa F & Zazueta C (2020) Mitochondrial dysfunction in metabolic and cardiovascular diseases associated with cardiolipin remodeling. Biochim Biophys Acta Mol Basis Dis 1866, 165744. [DOI] [PubMed] [Google Scholar]
36. Calzada E, Onguka O & Claypool SM (2016) Phosphatidylethanolamine metabolism in health and disease. Int Rev Cell Mol Biol 321, 29–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. van der Veen JN, Lingrell S, da Silva RP, Jacobs RL & Vance DE (2014) The concentration of phosphatidylethanolamine in mitochondria can modulate ATP production and glucose metabolism in mice. Diabetes 63, 2620–2630. [DOI] [PubMed] [Google Scholar]
38. Halbleib K, Pesek K, Covino R, Hofbauer HF, Wunnicke D, Hänelt I, Hummer G & Ernst R (2017) Activation of the unfolded protein response by lipid bilayer stress. Mol Cell 67, 673–684.e8. [DOI] [PubMed] [Google Scholar]
39. Lebeaupin C, Vallée D, Hazari Y, Hetz C, Chevet E & Bailly‐Maitre B (2018) Endoplasmic reticulum stress signalling and the pathogenesis of non‐alcoholic fatty liver disease. J Hepatol 69, 927–947. [DOI] [PubMed] [Google Scholar]
40. Trentzsch M, Nyamugenda E, Miles TK, Griffin H, Russell S, Koss B, Cooney KA, Phelan KD, Tackett AJ & Iyer S (2020) Delivery of phosphatidylethanolamine blunts stress in hepatoma cells exposed to elevated palmitate by targeting the endoplasmic reticulum. Cell Death Discovery 6, 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Gao X, van der Veen JN, Vance JE, Thiesen A, Vance DE & Jacobs RL (2015) Lack of phosphatidylethanolamine N‐methyltransferase alters hepatic phospholipid composition and induces endoplasmic reticulum stress. Biochim Biophys Acta 1852, 2689–2699. [DOI] [PubMed] [Google Scholar]
42. Elvas F, Stroobants S & Wyffels L (2017) Phosphatidylethanolamine targeting for cell death imaging in early treatment response evaluation and disease diagnosis. Apoptosis 22, 971–987. [DOI] [PubMed] [Google Scholar]
43. Ling J, Chaba T, Zhu LF, Jacobs RL & Vance DE (2012) Hepatic ratio of phosphatidylcholine to phosphatidylethanolamine predicts survival after partial hepatectomy in mice. Hepatology 55, 1094–1102. [DOI] [PubMed] [Google Scholar]
44. Udensi UK & Tchounwou PB (2017) Potassium homeostasis, oxidative stress, and human disease. Int J Clin Exp Physiol 4, 111–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Miranda MR, Vestuto V, Moltedo O, Manfra M, Campiglia P & Pepe G (2023) The ion channels involved in oxidative stress‐related gastrointestinal diseases. Oxygen 3, 336–365. [Google Scholar]
46. Keeney JTR, Miriyala S, Noel T, Moscow JA, St. Clair DK & Butterfield DA (2015) Superoxide induces protein oxidation in plasma and TNF‐α elevation in macrophage culture: insights into mechanisms of neurotoxicity following doxorubicin chemotherapy. Cancer Lett 367, 157–161. [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Li Y, Lin J, Wang P, Luo Q, Zhu F, Zhang Y, Hou Z, Liu X & Liu J (2020) Tumor microenvironment cascade‐responsive nanodrug with self‐targeting activation and ROS regeneration for synergistic oxidation‐chemotherapy. Nanomicro Lett 12, 182. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Laskin DL (2009) Macrophages and inflammatory mediators in chemical toxicity: a battle of forces. Chem Res Toxicol 22, 1376–1385. [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Xu GX, Wei S, Yu C, Zhao SQ, Yang WJ, Feng YH, Pan C, Yang KX & Ma Y (2023) Activation of Kupffer cells in NAFLD and NASH: mechanisms and therapeutic interventions. Front Cell Dev Biol 11, 1199519. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Chen WW, Chao YJ, Chang WH, Chan JF & Hsu YH (2018) Phosphatidylglycerol incorporates into cardiolipin to improve mitochondrial activity and inhibits inflammation. Sci Rep 8, 4919. [DOI] [PMC free article] [PubMed] [Google Scholar]
51. Choudhary V, Uaratanawong R, Patel RR, Patel H, Bao W, Hartney B, Cohen E, Chen X, Zhong Q, Isales CM et al. (2019) Phosphatidylglycerol inhibits toll‐like receptor‐mediated inflammation by danger‐associated molecular patterns. J Invest Dermatol 139, 868–877. [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Martensson CU, Doan KN & Becker T (2017) Effects of lipids on mitochondrial functions. Biochim Biophys Acta Mol Cell Biol Lipids 1862, 102–113. [DOI] [PubMed] [Google Scholar]
53. Domínguez‐Andrés J, Arts RJW, Bekkering S, Bahrar H, Blok BA, de Bree LCJ, Bruno M, Bulut Ö, Debisarun PA, Dijkstra H et al. (2021) In vitro induction of trained immunity in adherent human monocytes. STAR Protoc 2, 100365. [DOI] [PMC free article] [PubMed] [Google Scholar]
54. Güngör N, Knaapen AM, Munnia A, Peluso M, Haenen GR, Chiu RK, Godschalk RWL & van Schooten FJ (2009) Genotoxic effects of neutrophils and hypochlorous acid. Mutagenesis 25, 149–154. [DOI] [PubMed] [Google Scholar]
55. Freitas FCP, Depintor TS, Agostini LT, Luna‐Lucena D, Nunes FMF, Bitondi MMG, Simões ZLP & Lourenço AP (2019) Evaluation of reference genes for gene expression analysis by real‐time quantitative PCR (qPCR) in three stingless bee species (Hymenoptera: Apidae: Meliponini). Sci Rep 9, 17692. [DOI] [PMC free article] [PubMed] [Google Scholar]
56. Hoogerbrugge R, Willig SJ & Kistemaker PG (1983) Discriminant analysis by double stage principal component analysis. Anal Chem 55, 1710–1712. [Google Scholar]
57. Tortorella S, Tiberi P, Bowman AP, Claes BSR, Scupakova K, Heeren RMA, Ellis SR & Cruciani G (2020) LipostarMSI: comprehensive, vendor‐neutral software for visualization, data analysis, and automated molecular identification in mass spectrometry imaging. J Am Soc Mass Spectrom 31, 155–163. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials