Imaging bioactive lipid isomers in acetaminophen-induced liver injury using nano-DESI tandem MS

TL;DR

This study uses a new imaging technique to map bioactive lipids in liver tissue affected by acetaminophen overdose and shows how a treatment may help.

Contribution

A novel method for spatially mapping low-abundance isomeric bioactive lipids in liver tissues using nano-DESI tandem MS is presented.

Findings

Eicosanoids and SPMs localize to centrilobular hepatocytes after acetaminophen overdose.

4-MP treatment restores the spatial distribution of these lipids.

Nano-DESI tandem MS effectively detects low-abundance isomeric species.

Abstract

Acetaminophen (APAP) overdose is a leading cause of acute liver failure, resulting from the production of a reactive metabolite that induces hepatocyte necrosis. Current clinical treatments for APAP overdose offer limited therapeutic efficacy, highlighting the need for alternative strategies. 4-Methylpyrazole (4-MP, fomepizole) has emerged as a potential intervention to mitigate APAP toxicity in both mouse models and humans. Bioactive lipids, including eicosanoids and specialized proresolving mediators (SPMs), play essential roles in the inflammatory and resolution phases of APAP-induced liver injury. However, the impact of APAP overdose and 4-MP intervention on their distribution in liver tissue is poorly understood. Their low abundance and structural isomerism present challenges for MS imaging (MSI). In this study, we use nanospray desorption electrospray ionization MSI in tandem MS mode for the spatial mapping of eicosanoids and SPMs in liver tissues of mice subjected to moderate APAP overdose with and without 4-MP treatment. Using ammonium fluoride as a solvent dopant, known to enhance analyte signals, and leveraging MS/MS mode for isomer-specific analysis, we effectively detected low-abundance isomeric bioactive species. Our results reveal the localization of eicosanoids and SPMs in centrilobular hepatocytes following APAP overdose, correlating with APAP metabolism and hepatocyte necrosis. Notably, 4-MP treatment restores the spatial distributions of these lipids, supporting its therapeutic potential in modulating lipid-mediated inflammatory processes in APAP overdose. This study provides new insights into the localization of bioactive lipids in APAP-induced liver injury and highlights the power of nanospray desorption electrospray ionization MSI for investigating lipid-driven pathology.

Figures

Images of eicosanoids in mouse liver tissue sections obtained in nano-DESI MSI experiments in MS^2^ mode generated by plotting the signal of a selected fragment of a certain precursor ion. A: Double fluorescence images are of adjacent mouse liver tissue sections for the control, at 24 h and 48 h post-APAP overdose, and with 4-MP treatment at 24 h and 48 h post-APAP overdose. Cyp2E1, a marker of centrilobular cells (outlined in red), and Cyp2F2, a marker of periportal cells (green). B: Optical images of the mouse liver tissue sections imaging in negative ion mode. C: MS/MS ion images obtained f

Full text

Acetaminophen (APAP) overdose is a leading cause of drug-induced liver injury and acute liver failure in the Western world (1, 2). Because it is efficiently metabolized in the liver at therapeutic doses (3), APAP is considered a safe analgesic and antipyretic drug. However, APAP overdose leads to the overproduction of a reactive metabolite, N-acetyl-p-benzoquinone imine (NAPQI), which depletes GSH levels, causes protein adduct formation and mitochondrial dysfunction, ultimately leading to necrosis of centrilobular hepatocytes (4, 5, 6). Hepatocyte necrosis initiates an innate immune response, which is essential for liver recovery and regeneration after a moderate APAP overdose (5). Challenges associated with the clinical management of APAP-induced liver injury are driving the development of new drugs. Recently, 4-methylpyrazole (4-MP, fomepizole) has been identified as a strong therapeutic option for protecting against the cytotoxic effects of APAP overdose in both mouse models and humans (7, 8). In addition, drugs targeting mainly the regeneration phase after the injury show promise as a potential adjunct therapy (9).

MS has been used to gain insights into the chemical complexity of APAP metabolism after an overdose. LC-MS studies have examined APAP drug metabolism in biofluids, such as plasma, urine, or tissue homogenates (10, 11, 12, 13, 14). Significant changes in concentrations of phosphatidylcholines and other phospholipids were observed in APAP overdose cases in children’s serum (13) and mouse tissue (14). Other LC-MS studies have focused on the quantitative analysis of APAP and its metabolites in different overdose scenarios (11, 12, 15). MS imaging (MSI) is a powerful technique that enables label-free spatial mapping of endogenous molecules present in biological samples (16, 17, 18, 19, 20). MSI has been employed to understand the spatial localization of APAP and its metabolites within biological tissues (15, 21, 22, 23, 24, 25). For example, MALDI MSI has been used to study APAP metabolism in various mouse and rat tissue sections (21, 22, 24, 25). One MALDI-MSI study also found that phospholipids and other metabolites were altered in APAP overdose (24). Specifically, the concentrations of phosphatidylethanolamine species were decreased, whereas the concentrations of phosphocreatine and ceramides were increased in the damaged regions of rat liver tissues (24). Desorption electrospray ionization MSI experiments have been used to study the distribution and the dynamics of APAP metabolism, revealing zonated patterns of APAP’s metabolites in mouse liver and kidney tissue (15, 23). Despite these efforts, little is known about the spatial localization of bioactive lipids involved in the innate immune response and their resolution after hepatocyte regeneration in APAP-induced liver injury.

Bioactive lipids, such as eicosanoids and specialized proresolving mediators (SPMs), play an important role in both the innate immune response and resolution of inflammation during APAP hepatotoxicity. Eicosanoids are a diverse group of oxidized unsaturated fatty acids derived from arachidonic acid (AA) through three main biosynthetic pathways: cyclooxygenase, lipoxygenase (LOX), and cytochrome P450 (CYP). Proinflammatory eicosanoids, including prostaglandins (PGs), prostacyclins, thromboxanes, leukotrienes, and oxidized fatty acids, contribute to inflammation, oxidative stress, and tissue damage (26, 27). In contrast, SPMs are bioactive lipids that participate in the resolution of inflammation (28, 29). SPMs are dihydroxylated and trihydroxylated fatty acids derived from AA, EPA, docosapentaenoic acid, and DHA through the same biosynthetic pathways as eicosanoids. Examples of SPMs include resolvins, maresins, protectins, and lipoxins (26, 28). SPMs have specific anti-inflammatory roles in neutrophils, macrophages, platelets, and endothelial cells (28).

Despite the important roles of eicosanoids and SPMs in biological systems, MSI of these molecules in tissues has been hindered by their low abundance, poor ionization efficiency, and presence of isomers (30, 31, 32, 33). Several MSI approaches have been developed to address these challenges and facilitate imaging eicosanoids in tissue samples. For example, selected ion monitoring mode has been effective at enhancing signals of eicosanoids in nanospray desorption electrospray ionization (nano-DESI) MSI experiments (34). Nano-DESI is a liquid extraction-based ambient ionization technique (35, 36), in which analytes are extracted into a dynamic liquid bridge formed between a specially designed nano-DESI probe and sample surface (37, 38). The extracted analytes are transferred through the probe to a mass spectrometer inlet and ionized by electrospray ionization. Similar to other liquid extraction-based ambient ionization techniques, nano-DESI allows tailoring solvent composition to improve the extraction and ionization efficiency of analytes (39, 40, 41, 42, 43). This capability is particularly advantageous for imaging of eicosanoids and SPMs in biological tissues. For example, the addition of silver chloride to the extraction solvent has been used to promote Ag^+^ cationization, which substantially improves the ionization efficiency of PGs in positive mode (44). Meanwhile, the addition of ammonium fluoride (NH_4_F) to the solvent has been used to enhance signals of all the extracted lipids, including eicosanoids, in negative ionization mode nano-DESI MSI (43).

The presence of numerous isomers further complicates studies of the localization of eicosanoids and SPMs in tissues using MSI. MSI-MS/MS has been used to separate isomeric species based on their diagnostic fragments and examine their spatial distribution in tissue samples (45, 46, 47, 48, 49). MSI experiments in both MS^2^ and MS^3^ modes have been used for examining the localization of isomeric metabolites, phospholipids, and PGs in biological tissues (44, 46, 47, 49, 50). In these experiments, MS/MS spectra are acquired in each pixel of the image, and images are generated by plotting the abundance of a characteristic fragment ion to determine the spatial distribution of a target molecule. The first nano-DESI MSI-MS/MS experiment was performed on an Orbitrap instrument, which enabled imaging of approximately 300 metabolites and lipids in a tissue section in a single experiment (47). Additionally, we have developed an MSI-MS/MS approach on a triple quadrupole mass spectrometer for the spatial mapping of isomeric and isobaric lipids (45). In that study, we demonstrated the spatial mapping of two eicosanoid isomers in rat kidney tissue.

Herein, we use nano-DESI MSI-MS/MS to examine the spatial distribution of isomeric eicosanoids and SPMs in liver tissues of a mouse model of APAP hepatotoxicity. Our findings reveal the unique localization of these bioactive lipids to centrilobular hepatocytes in mouse liver tissues 24 h and 48 h after a moderate APAP overdose (300 mg/kg). Additionally, we demonstrate that 4-MP treatment after APAP overdose restores the spatial distribution of these species to resemble that of the control group. This effect suggests that 4-MP plays a role in modulating inflammatory resolution and mitigating lipid dysregulation. This comprehensive analysis provides molecular insights into the role of lipid mediators in the inflammatory and resolution phases of APAP-induced liver injury, both with and without 4-MP treatment.

Materials and Methods

Chemicals and reagents

NH_4_F was purchased from Sigma-Aldrich (St Louis, MO). Omnisolv LC-MS grade water and methanol (MeOH) were purchased from Millipore Sigma (Burlington, MA). Lysophosphatidylethanolamine (LPE) 17:1 and AA-d8 lipid standards were purchased from Avanti Polar Lipids (Alabaster, AL). Solvent mixtures used in nano-DESI MSI experiments were composed of 9:1 MeOH:H_2_O (v/v) containing 1 μM lysophosphatidylethanolamine 17:1 and AA-d8 and 500 μM NH_4_F.

Animal experiments and tissue preparation

All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Kansas Medical Center. In this study, 8- to 10-week-old male C57BL/6J mice were purchased from Jackson Laboratories (Jackson Lab, Harbor, ME). Upon receipt, the mice were maintained in a temperature-controlled room (14-h light/10-h dark cycle) where they were provided ad libitum access to both food and water. For in vivo experiments, mice (average body weight of 20–25 g) were fasted for 15 h, before intraperitoneal injections of 300 mg/kg APAP with or without 50 mg/kg 4-MP (both dissolved in warm saline) (Sigma-Aldrich, St Louis, MO) or saline vehicle. Animals were refed in all experiments at 6 h after APAP administration. All animals were euthanized by cervical dislocation under isoflurane anesthesia at 24 and 48 h post-APAP treatment. Blood was drawn from the caudal vena cava into heparinized syringes and was centrifuged at 18,000 g for 3 min to obtain plasma. The liver tissue was snap-frozen in liquid nitrogen and stored in the −80 °C freezer before further analysis. Experiments were performed following the National Research Council for the Care and Use of Laboratory Animals Guidelines.

All sections were stored in a −80 °C freezer prior to nano-DESI MSI analysis. Optical images of tissue sections were acquired using a PathScan Enabler IV pathology slide scanner (Meyer Instruments, Houston, TX). Tissue sections were allowed to thaw at room temperature prior to nano-DESI MSI experiments.

Immunostaining of frozen tissue sections

Frozen liver tissue samples were sectioned at 12 μm thickness using a Leica CM 1950 Cryostat. Sections were first deparaffinized and dehydrated before a blocking step in 5% normal goat serum (Vector Laboratories) was performed (23). This was followed by an overnight incubation step with the primary anti-Cyp2E1 and Cyp2F2 antibodies (catalog no.: ab28146, Abcam, Boston, MA; catalog no.: SC374540, Santa Cruz Biotechnology, respectively). The next day, sections were washed in PBS to apply the secondary antibodies, Alexa Fluor 594-conjugated goat anti-rabbit (catalog no.: A11037) and Alexa Fluor 488 goat anti-mouse (catalog no.: A28175) (ThermoFisher Scientific, Waltham, MA). Slides were imaged on a Nikon Eclipse Ti2 inverted fluorescence microscope (after nuclei staining with a DAPI-containing mounting medium (catalog no.: D9542) (Sigma-Aldrich).

Nano-DESI MSI

Nano-DESI MSI experiments were performed on a Q-Exactive HF-X Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA) using a custom-designed nano-DESI source (51). The high-resolution nano-DESI probe was assembled in front of the mass spectrometer inlet as described in prior studies (52, 53). Briefly, the pulled primary and nanospray capillaries with a tip size of 15–25 μm are aligned to generate a nano-DESI probe. Analytes are extracted into the liquid bridge formed by the nano-DESI probe on a sample surface and transferred through the nanospray capillary to the mass spectrometer for detection. A third capillary, which serves as a shear force probe, is positioned next to the nano-DESI probe to maintain a constant distance to the sample. We used 9:1 MeOH:H_2_O (v/v) containing lipid standards and 500 μM NH_4_F as a working solvent to enhance the ionization of analytes in negative mode. The solvent was propelled through the primary capillary using a syringe pump at 0.5 μl/min. Negative mode ionization was achieved by applying a −3.2 kV potential to the syringe needle. The heated inlet was held at 300°C. For imaging of mouse liver sections, the sample was scanned under the nano-DESI probe in lines at a rate of 8 μm/s and stepping between the lines by 25 μm. In each line scan, we acquired both full MS and MS/MS spectra using seven isolation windows. Full MS scans were acquired with a mass resolving power of 30,000 at m/z 200 for the m/z range of 140–900, with an automatic gain control target of 5 × 10^5^ ions and a maximum integration time of 75 ms. Data-independent MS/MS spectra scans were acquired using a list of six endogenous m/z of interest provided in Table 1 and the internal standard, AA-d8, at m/z 311.282 with a mass resolving power of 60,000 at m/z 200. Higher energy collision-induced dissociation was performed using a mass isolation window of 0.4 m/z and stepped collision energies of 18, 24, and 30 arbitrary units specific to the instrument. The automatic gain control target was set to 2 × 10^5^ and the integration time to 200 ms. The resulting ion images have an average pixel size of 20 μm × 25 μm.Table 1. Characteristic fragments of lipids targeted in nano-DESI MS/MSI experimentsPrecursor m/zMolecular formulaNameFragment m/zReference319.227C_20_H_32_O_3_12-HETEa179.108(56, 57, 58, 59, 60, 61)333.207C_20_H_30_O_4_Dihydro-keto PGA_2_a113.097(34)RvE2 and RvE4b115.040(29, 62, 63)PGJ_2_a171.103(57)349.202C_20_H_30_O_5_8-Iso-15-keto-PGE_2_a113.097(57)LXA_5_b115.040(61, 65)351.216C_20_H_32_O_5_LXA4b115.039(29, 57, 58, 60, 61, 65)PGE_2_, PGD_2_a189.129(29, 56, 60, 61)LXB4b221.116(29, 60, 61, 62)375.217C_22_H_32_O_5_RvD4b131.035(64)aEicosanoids.bSPMs.

Data analysis

Each line scan was acquired as an individual file (.RAW file format) using Xcalibur software (Thermo Electron, Bremen, Germany). Data processing was performed using a custom-designed Python code developed by our group (https://github.com/LabLaskin/MSIGen). Ion images were generated by plotting the ion abundance of each m/z as a function of the location on the tissue. Normalization to the internal standard was performed by dividing the signals of endogenous species by the signal of the internal standard. Replicate ion images from nano-DESI MSI experiments are provided in Supplemental Figures S1 and S2.

Region of interest (ROI) analysis was performed by comparing signals in red regions in fluorescence images, highlighting the centrilobular cells with the remaining tissue (RT). The pericentral ROI mask was derived from the red channel, whereas the RT mask was generated from the grayscale fluorescence image. ROI masks were generated in MSIGen by applying a Gaussian blur to remove image artifacts (e.g., dust on the objective lens), downscaling the image, and applying a binary threshold. Additional processing was performed to eliminate fragmented regions and to reduce contributions from intermediate regions. Fluorescence images were manually registered to optical images using an affine transformation; optical images were subsequently registered to the ion images. ROI masks were used to calculate means and standard deviations of analyte signals across each region.

Results

In this study, we examine the spatial distributions of eicosanoids and SPMs in mouse liver tissues after an APAP overdose using nano-DESI MSI. APAP hepatotoxicity was induced using a moderate overdose of 300 mg/kg, and tissues were collected at 24 and 48 h post-APAP administration. These time points were chosen to capture the expected peak of injury at 24 h and the peak of injury resolution at 48 h (54). Additionally, we explore the effect of 4-MP administration on the distribution of lipid mediators after APAP overdose at the same time points. Since APAP hepatotoxicity characteristically affects centrilobular hepatocytes, adjacent mouse liver tissue sections were stained for immunofluorescence (IF) imaging of Cyp2E1, a marker of centrilobular hepatocytes, and Cyp2F2, a marker of periportal hepatocytes (55). The IF images shown in Figures. 1A, 2A, and 3A provide a reference for the observed localizations of eicosanoids and SPMs within the mouse liver tissues using nano-DESI MSI.Fig. 1. Images of eicosanoids in mouse liver tissue sections obtained in nano-DESI MSI experiments in MS^2^ mode generated by plotting the signal of a selected fragment of a certain precursor ion. A: Double fluorescence images are of adjacent mouse liver tissue sections for the control, at 24 h and 48 h post-APAP overdose, and with 4-MP treatment at 24 h and 48 h post-APAP overdose. Cyp2E1, a marker of centrilobular cells (outlined in red), and Cyp2F2, a marker of periportal cells (green). B: Optical images of the mouse liver tissue sections imaging in negative ion mode. C: MS/MS ion images obtained for 12-HETE (319.227→179.108), PGJ_2_ (333.207→171.103), PGA_2_ (333.207→113.097), 8-iso-15-keto-PGE_2_ (349.202→113.097), and PGE_2_ (351.216→189.129) in negative ionization mode. The scale bar represents 500 μm. The normalized ion abundance scale is shown at 99.7%. D: ROI analysis showing the ratio of the average signal of the pericentral ROI (I_ROI_) to that of the RT (I_RT_) for 12-HETE, PGJ_2_, PGA_2_, 8-iso-15-keto-PGE_2,_ and PGE_2_ across treatment groups.Fig. 2. Images of SPMs in mouse tissue sections from nano-DESI MSI experiments. A: Double fluorescence images are of adjacent mouse liver tissue sections for the control, at 24 h and 48 h post-APAP overdose, and with 4-MP treatment at 24 h and 48 h post-APAP overdose. Cyp2E1, a marker of centrilobular cells (red), and Cyp2F2, a marker of periportal cells (green). B: Optical images of the mouse liver tissue sections imaging in negative ion mode. C: Ion images of [M-H]^-^ ions for RvE2/RvE4 (333.207→115.040), LXA5 (349.202→115.040), LXB4 (351.216→221.116), and RvD4 (375.217→131.035). The scale bar represents 500 μm. The ion abundance scale is shown at 99.7%. D: ROI analysis showing the ratio of the average signal of the pericentral ROI (I_ROI_) to that of the RT (I_RT_) for RvE2/RvE4, LXA5, LXB4, and RvD4 across treatment groups.Fig. 3. Images of lipids and metabolites in mouse tissue sections from nano-DESI MSI experiments. A: Double fluorescence images are of adjacent mouse liver tissue sections for the control, at 24 h and 48 h post-APAP overdose, and with 4-MP treatment at 24 h and 48 h post-APAP overdose. Cyp2E1, a marker of centrilobular cells (red), and Cyp2F2, a marker of periportal cells (green). B: Optical images of the mouse liver tissue sections were obtained in negative ion mode. C: Ion images from the MS1 scan of [M-H]^-^ ions for GSSG (m/z 305.069, [M-2H]2-), GSH (m/z 306.076), and LPA 18:3 (m/z 431.229). The scale bar represents 500 μm. The ion abundance scale is shown at 99.7%. D: ROI analysis showing the ratio of the average signal of the pericentral ROI (I_ROI_) to that of the RT (I_RT_) for GSSG, GSH, and LPA 18:3 across treatment groups.

Nano-DESI MSI experiments were performed on a Q-Exactive HF-X instrument operated in MS^2^ mode. For these experiments, we selected seven m/z windows to incorporate several key eicosanoids and SPMs known to play an important role in both the inflammatory and resolution responses in liver failure (26, 27). A list of m/z values of the precursor and product ions used to create ion images is provided in Table 1. The names assigned to each compound correspond to their most likely structural identity, guided by fragmentation spectra, biological relevance, and previous literature documenting their presence in inflammatory pathways. Aside from m/z 319.227, for which only one diagnostic fragment was included to target 12-HETE, multiple fragments were selected for other species, targeting different isomeric species. While multiple isomers are present for each m/z found in Supplemental Table S1, we selected those with well-characterized roles in inflammation and resolution pathways. All the species were identified based on the accurate m/z and MS/MS data. On-tissue MS/MS spectra are provided in Supplemental Figures S3 and S4. The characteristic product ions of different isomeric species were identified by comparing MS/MS spectra obtained in this study with those reported in the literature (29, 33, 53, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65). For each endogenous analyte, ion images were generated using characteristic product ions from the corresponding MS/MS spectra. Due to the low abundance of eicosanoids within the tissue samples, some product ions were of low abundance.

Imaging of proinflammatory lipids in mouse liver

To understand the effect of APAP-induced hepatotoxicity on the spatial distributions of proinflammatory eicosanoids, we used nano-DESI MSI-MS/MS. The results obtained for proinflammatory eicosanoids, including 12-HETE, dihydro-keto PGA_2_, PGJ_2_, PGA_2_, 8-iso-keto PGE_2_, and PGE_2_ at m/z 319.227, 333.207, 349.202, and 351.216, respectively, are shown in Figure 1. IF images of sequential mouse liver tissue sections to those imaged by nano-DESI, shown in Figure 1A, highlight the centrilobular cells stained for Cyp2E1 in red and periportal regions stained for Cyp2f2 in green (55). Optical images of the tissue sections analyzed using nano-DESI MSI are shown in Figure 1B.

MS/MS mode ion images of eicosanoids, including 12-HETE (319.227→179.108), PGJ_2_ (333.207→171.103), PGA_2_ (333.207→113.097), 8-iso-15-keto-PGE_2_ (349.202→113.097), and PGE_2_ (351.216→189.129), are shown in Figure 1C. In the control tissue, all the eicosanoids are distributed across the entire tissue, and no significant zonation between the centrilobular or periportal cells is observed. Twenty-four hours post APAP, there is a loss of centrilobular staining corresponding to mitochondrial damage and hepatocyte necrosis caused by the APAP overdose. This is consistent with a previous time-course study of APAP-induced hepatic injury that at the moderate overdose (300 mg/kg) has shown that injury peaks at 24 h (66). This is accompanied by a discrete change in the spatial distribution of 12-HETE with some spatial consolidation of the ion signal for other PGs (Fig. 1C). By 48 h post-APAP overdose, when liver resolution is initiated (3, 23, 55), there is an obvious zonation of eicosanoid signal with increased abundance in pericentral regions. This signal enhancement is evident in the ROI analysis shown in Figure 1D, where the signal in the pericentral ROI (red regions in fluorescence images marking centrilobular cells) is normalized to the signal averaged over the RT. For all eicosanoids, this I_ROI_/I_RT_ ratio is increased at 48 h post-APAP overdose. Upon intervention with 4-MP treatment after APAP overdose, the spatial distribution of eicosanoids reverts to that observed in the control tissue by 24 h (Fig. 1C), and the I_ROI_/I_RT_ ratios are restored nearly to control levels (Fig. 1D). These results establish the temporal and spatial redistribution of proinflammatory eicosanoids during APAP injury and recovery, providing a framework for comparison with other lipid mediator classes.

Imaging of SPMs

We also examined the spatial distribution of SPMs, including resolving E-series (RvE), lipoxins A and B (LXA and LXB), and resolvin D-series (RvD), within the same liver tissues as the proinflammatory eicosanoids, as shown in Figure 2. Similar to the results obtained for eicosanoids, we observe a fairly uniform distribution of the SPMs in the control tissue displayed in Figure 2C. While at 24 h post-APAP overdose, all SPMs remain broadly distributed across the tissue. By 48 h post-APAP overdose, all SPMs are localized to the centrilobular regions, except for RvE2/RvE4, which exhibits a more uniform distribution. Post-APAP overdose, 4-MP treatment restores the spatial localization of the SPMs to that observed in the control tissue. These spatial trends are supported by the ROI analysis, in which SPM signals within the centrilobular ROI are normalized to the RT. At 24 h post-APAP overdose, the I_ROI_/I_RT_ ratios are comparable to control, whereas at 48 h post-APAP overdose, the ratios are increased, indicating the enhanced abundance of SPMs in centrilobular regions. Following 4-MP treatment, the I_ROI_/I_RT_ ratios return to control levels, matching the restored spatial distribution. Of note, RvE2/RvE4 had a broader spatial distribution at 48 h post-APAP overdose compared with other SPMS. The colocalization of eicosanoids and SPMs in the centrilobular regions suggests that inflammation and resolution occur within the same liver microenvironment.

Metabolic changes in APAP tissues

Additionally, we obtained ion images of other important lipids and metabolites in liver tissue using nano-DESI MSI in the traditional MS^1^ mode. The results obtained for species from the GSH metabolism pathway are shown in Figure 3, in which IF images are shown in Figure 3A and optical images in Figure 3B. Ion images of GSSG (m/z 305.069, [M-2H]^2-^), GSH (m/z 306.076), and lysophosphatidic acid (LPA) 18:3 (m/z 431.229) are shown in Figure 3C.

Similarly to the results obtained for eicosanoids and resolvins, we observe a uniform distribution of all three species in the control liver tissue and upon 4-MP intervention in liver tissues. However, in the 24-h APAP overdose sample, GSH and GSSG are depleted in the centrilobular region. The depletion of GSH and GSSG correlates with the highest expression of CYP2E1 in the centrilobular cells, the enzyme responsible for generating NAPQI, APAP’s toxic metabolite. Although this depletion is evident in ion images, it is not detected by the ROI ratio analysis (Fig. 3D) due to the small size of the affected region relative to the predefined ROI and the substantial GSH and GSSG signal outside the ROI.

In contrast, LPA 18:3, a bioactive lysophospholipid involved in cell survival and inflammation, has similar spatial distributions to those observed for the SPMs. In the 48-h APAP overdose tissue, LPA 18:3 becomes concentrated in centrilobular regions as shown in Figure 3C and D, potentially indicating a protective or reparative response to necrotic signaling in hepatocytes. LPA is a key lipid mediator in Kupffer cells and has been shown to stimulate hepatocyte proliferation in vitro (67, 68).

Discussion

This study provides spatially resolved insight into the lipid mediators involved in inflammation signaling and resolution upon APAP-induced liver injury. Eicosanoids have known roles in inflammatory signaling pathways, and our imaging results reveal their dynamic redistribution during both injury and recovery phases. Specifically, 12-HETE, a product of the LOX pathway, has been implicated in various conditions, including hepatic injury, ischemia, hypoxia, and cancer, causing mitochondrial dysfunction and oxidative stress (27, 69, 70). Neutrophils are major producers of 12-HETE, which are recruited by Kupffer cells and contribute to the inflammatory response. Other proinflammatory eicosanoids derived in the cyclooxygenase pathway, such as PGE_2_, PGJ_2_, and PGA_2_, can exacerbate liver inflammatory responses by activating Kupffer cells, which in turn release cytokines and recruit macrophages and neutrophils to the centrilobular cells (26, 27). PGE_2_ is synthesized by Kupffer cells and is known to regulate inflammatory responses and hepatocyte regeneration. The predominant colocalization of eicosanoids to the centrilobular region at the time of active liver recovery (Fig. 1) indicates a key relationship of these lipid mediators to hepatocyte regeneration and liver recovery after APAP-induced centrilobular necrosis (71).

For SPMs, the localization of LXA5, LXB4, and RvD4 to the centrilobular regions at 48 h post-APAP overdose (Fig. 2) aligns with the known role of Kupffer cells, which become highly activated during APAP hepatotoxicity. These cells are major producers of lipoxins, which are biosynthesized from AA by the LOX pathway (27, 28). Kupffer cells recruit macrophages to the centrilobular regions by producing both proinflammatory and anti-inflammatory cytokines, indicating that the centrilobular region is a hub for both inflammation and resolution (72). The macrophages facilitate an anti-inflammatory response through the LOX-mediated biosynthesis of RvD from DHA (27, 28). Resolvins are known to prevent neutrophil infiltration, promote apoptotic cell clearance by macrophages, and facilitate inflammation resolution (73, 74, 75).

The broad distribution of RvE2/RvE4 (Fig. 2C) aligns with their known role in modulating systematic inflammatory responses and promoting the resolution of inflammation throughout the liver tissue. This is likely due to their biosynthesis in the endothelial cells and peripheral immune cells, rather than in Kupffer cells, and therefore, they are not confined to the centrilobular cells. The resolving E-series mediators are derived from EPA through oxidation by CYP450 enzymes, followed by conversion via the LOX pathway (27, 28). The colocalization of eicosanoids and SPMs in the centrilobular regions suggests that inflammation and resolution occur within the same liver microenvironment.

Redox metabolism also exhibits pronounced spatial regulation during APAP-induced injury. We observed depletion of GSH and GSSG in centrilobular regions (Fig. 3) coinciding with the highest expression of CYP2E1, the enzyme responsible for generating NAPQI, APAP’s toxic metabolite. NAPQI is a highly reactive metabolite and depletes intracellular GSH, forming APAP-GSH conjugates and leading to hepatocellular damage. The depletion of GSH in the centrilobular cells is caused by the excessive accumulation of NAPQI, which leads to mitochondrial dysfunction and the formation of reactive oxygen species, triggering cell necrosis (10). While its overall levels are depleted in the 24-h APAP sample, the localization of GSSG at the interface between centrilobular and periportal regions suggests that GSH oxidation occurs in this transition zone, where hepatocytes counteract oxidative stress. By 48 h post-APAP overdose, GSH and GSSG are both uniformly disturbed across the tissues, indicating that NAPQI accumulation has been resolved. While in the 24-h period with 4-MP intervention, GSH and GSSG are evenly distributed across the tissues, indicating that 4-MP treatment inhibits the accumulation and therefore depletion of GSH in centrilobular cells. Similar observations in APAP-induced liver injury for GSH and APAP’s metabolites were reported in other MSI studies for APAP overdose in mouse liver tissue (23).

These spatial metabolomic insights provide a more comprehensive understanding of redox imbalances, lipid signaling, and hepatocellular responses in APAP-induced liver injury. The distinct localization patterns of GSH, GSSG, and LPA 18:3 emphasize the progression of oxidative damage and the compensatory mechanisms activated during APAP toxicity.

Conclusions

We used nano-DESI MSI-MS/MS to examine the spatial distributions of proinflammatory eicosanoids and anti-inflammatory SPMs in liver tissues of APAP-treated mice. This approach enabled the sensitive, isomer-selective imaging of low-abundance bioactive lipids in biological tissues, revealing their distinct yet overlapping localizations. Notably, we observed colocalization of proinflammatory and anti-inflammatory lipids in centrilobular regions, indicating the spatial coupling of inflammatory and resolution processes during APAP-induced liver injury. Additionally, the observed localization of GSH and GSSG to centrilobular hepatocytes highlights their role in oxidative stress and redox regulation in APAP-induced injury. These results emphasize the role of lipid mediators in cellular homeostasis during the transition from inflammation to resolution. This workflow was specifically developed for spatial mapping of isomeric eicosanoids and SPMs using MS/MS acquisition with nano-DESI MSI, targeting an important class of lipid species that is particularly challenging to study due to their low abundance and isomeric complexity. This label-free strategy is well-suited for resolving lipid isomers that are present in low concentrations in biological samples. Our results establish nano-DESI MSI-MS/MS as a powerful technique for spatially resolving isomeric lipid species in complex biological tissues, offering new insights into the biochemical mechanisms underlying APAP-induced hepatotoxicity and its modulation by therapeutic interventions. More broadly, this study demonstrates the utility of nano-DESI MSI-MS/MS for unraveling the spatial complexity of lipid-mediated diseases at high chemical specificity.

Data Availability

All data supporting the findings of this study are available within this article and its supplemental data. Raw MSI data generated in this study are available from the corresponding author on reasonable request.

Supplemental Data

This article contains supplemental data (29, 34, 56, 57, 58, 59, 60, 62, 63, 65).

Conflict of Interest

The authors declare that they have no conflicts of interest with the contents of this article.