Carboxylesterase 1-mediated endocannabinoid metabolism in skin: role in melanoma progression in BRafV600E/Pten-/- mice.

Abstract

Background

Melanoma is a highly aggressive skin cancer with a poor prognosis. The endocannabinoids 2-arachidonoylgylcerol (2-AG) and anandamide have been linked to melanoma progression, though their roles remain unclear. We hypothesized that the 2-AG-arachidonate-prostaglandin axis could drive aggressive melanoma progression.

Methods

The genetically engineered melanoma mouse model B6-Tyr::CreER^T2; BRaf^CA; Pten^loxP was characterized by targeted metabolomics. Functionally expressed serine hydrolases in the tumor tissue were identified by chemoproteomics. Pharmacological inhibition of carboxylesterase 1 (CES1) was achieved through chronic in vivo i.p. treatment with JZL184 (10 mg/kg daily), confirmed by activity-based protein profiling (ABPP) and targeted lipidomics. CES1-mediated 2-AG hydrolysis was further confirmed in radiotracer-based assays using CES1-transfected cell lines.

Results

The diacylglycerol and protein kinase C activator 1-stearoyl-2-arachidonoyl-sn-glycerol (SAG) was significantly elevated in the nodular-like melanoma tissues, along with 2-AG and arachidonic acid (ARA), compared to normal skin. AEA and other N-acylethanolamines were decreased, while, notably, prostaglandin levels remained unchanged. Significant changes in the levels of neuromodulators and neurotransmitters, including serotonin and adenosine, were observed. Pronounced differences between serine hydrolase activity in normal skin and melanoma tissue were identified by ABPP. Intriguingly, CES1 was identified as the only 2-AG-hydrolyzing enzyme in this melanoma tissue, as MAGL and ABHD6/12 were not expressed. The MAGL inhibitor JZL184 also efficiently inhibited CES1 in vitro and in vivo, increasing glycerol esters and reducing tumor progression. Additionally, scRNA-seq data from previous studies revealed divergent MAGL/CES1 expression patterns across different human melanoma subtypes.

Conclusions

A role of CES1 expression in skin is demonstrated for the first time. Our study suggests that 2-AG degradation to arachidonate favors melanoma progression, either reflecting the carcinogenic role of ARA or that monoacylglycerols like 2-AG and/or other CES1 substrates may exert antitumor effects, indicating that CES1 could be a potential therapeutic target. CES1 expression and high SAG, 2-AG, and ARA levels may be a signature of specific BRAF-driven malignant melanoma subtypes which are associated with discrete metabolic adaptations.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40170-025-00378-2.

Untitled section

Keywords: CES1, Melanoma, 2-AG, Lipidomics, Targeted Metabolomics, Chemoproteomics, Cancer, Biomarkers

Article notes

Untitled section

Received 2024 Oct 22; Accepted 2025 Feb 3; Collection date 2025.

Author notes

Introduction

Malignant melanoma, the most aggressive form of skin cancer, poses a significant challenge for therapy due to its heterogeneity, high risk for metastasis, and resistance to conventional treatments [1]. Melanomas are clinically categorized by their growth pattern (such as superficial spreading, lentigo, or nodular melanoma), anatomical location (e.g., cutaneous, mucosal, or acral melanoma), or pigmentation status (e.g., melanotic versus amelanotic melanoma) [2]. Approximately 50% of melanomas carry activating BRAF mutations, with over 90% being the V600E variant. BRAF^V600E has been linked to various mechanisms driving melanomagenesis, primarily through the dysregulated activation of downstream MEK/ERK effectors [3]. Recent advancements in treatment, particularly with targeted therapies, such as BRAF, MEK, and immune checkpoint inhibitors, have significantly improved the prognosis for patients with common subtypes of cutaneous melanoma (CM) [4, 5]. However, research and clinical trials predominantly focus on the common subtypes, resulting in a significant gap in our understanding of the molecular characteristics and effective treatment strategies for rarer melanoma variants from patients with CM who present increased metastasis and resistance to treatments [4]. This highlights the urgent need to identify molecular markers to better characterize the heterogeneity and subtypes of melanoma, which is crucial for understanding both the mechanisms behind individual treatment responses and possible resistance to targeted therapy.

To date, few studies have explored the role of the endocannabinoid system (ECS), particularly the endocannabinoid-prostaglandin axis, in melanoma development and progression [6, 7]. The roles of endocannabinoids in melanoma thus remain unclear, as both pro- and anti-tumorigenic effects have been described [8]. For example, high levels of arachidonic acid (ARA), the major tissue n-6 polyunsaturated fatty acid, have been implicated in promoting tumor progression through the activation of pro-inflammatory pathways, modulation of the tumor microenvironment, and enhancement of metastatic potential [9, 10]. ARA is a precursor of eicosanoids, including prostaglandins (Fig. 1) and leukotrienes, which are generated via cyclooxygenases and can contribute to cancer cell proliferation, angiogenesis, and resistance to apoptosis [11, 12]. In addition to the classical PLA₂ pathway leading to ARA from plasma membranes, the major tissue endocannabinoid 2-arachidonoylglycerol ( 2-AG) may serve as a precursor for ARA and potentially eicosanoids (Fig. 1) [13]. Elevated levels of 2-AG have also been shown to exert antitumor effects in pancreatic cancer [14] and non-small cell lung cancer [15]. The primary tissue enzyme responsible for 2-AG degradation is monoacylglycerol lipase (MAGL) (Fig. 1) [16–19].

Fig. 1Schematic representation of the 2-AG-ARA-PGs axis. Abbreviations: 2-AG, 2-arachidonoylglycerol; ABHD6, alpha/beta-hydrolase domain-containing protein 6; ABHD12, alpha/beta-hydrolase domain-containing protein 12; ARA, arachidonic acid; CES1, carboxylesterase 1; CES2, carboxylesterase 2; COX1, Cyclooxygenase 1 COX2, Cyclooxygenase 2; DAGL α, Diacylglycerol lipase α; DAGL β, Diacylglycerol lipase β; MAGL, monoacylglycerol lipase; PGD2, prostaglandin D2; PGE2, prostaglandin E2; SAG, 1-stearoyl-2-arachidonoyl-sn-glycerol

Notably, several other serine hydrolases can hydrolyze 2-AG in a cell- and tissue-specific manner, including alpha/beta-hydrolase domain-containing proteins 6 and 12 (ABHD6, ABHD12) [19], and carboxylesterases 1 and 2 (CES1, CES2) [20, 21]. Among these, carboxylesterases (CES), which are members of the serine hydrolase family, play a significant role by hydrolyzing carboxylic esters and amides [22, 23]. The role of CES1 in the ECS remains poorly understood.

In humans, CES1 is the predominant isoform, primarily found in the liver but also present in other tissues, including adipose tissue, lungs, and intestine [24–26]. CES1 plays an essential chemoprotective role, through the detoxification of a wide range of molecules, including narcotics (e.g., cocaine and heroin), pesticides, and other xenobiotics [27–30]. It also participates in the metabolism of several therapeutic drugs, such as clopidogrel [31] and oseltamivir [32], and promotes the conversion of different prodrugs to their active forms, including imidapril and temocapril [33]. In mice, eight Ces1 isoforms have been reported—Ces1a, Ces1b, Ces1c, Ces1d, Ces1e, Ces1f, Ces1g, and Ces1h [34], each showing a distinctive tissue distribution and substrate affinities compared to the human CES1 [22].

Importantly, CES1 plays a vital role in endogenous lipid metabolism by hydrolyzing triglycerides and cholesterol esters, regulating intracellular levels of free fatty acids and cholesterol, and maintaining lipid homeostasis [30]. Dysregulation of CES1 activity can lead to altered lipid metabolism [22, 35], contributing to pathologies such as obesity, diabetes, and atherosclerosis [36–38]. Interestingly, some studies have implicated CES1 expression in cancer pathophysiology as a biomarker and a potential therapeutic target for hepatocellular carcinoma [39–41] and colorectal cancer [42]. However, to date, the role of CES1 in skin or melanoma progression has not been reported.

In this study, we used a genetically engineered mouse model of human malignant melanoma based on the findings of Dankort et al., who in 2009 demonstrated that the BRaf^V600E mutation, combined with the loss of the tumor suppressor gene Pten, leads to the development of a highly aggressive form of nodular-like melanoma in mice, characterized by 100% penetrance, short latency and metastases in lymph nodes and lungs [43]. The model is based on the melanocyte-specific expression of BRaf^V600E and Pten deletion mediated by Cre-lox recombination under the control of the Tyrosinase-CreER^T2 promoter. Cre recombinase is activated by the active metabolite of the pro-drug tamoxifen, 4-hydroxytamoxifen (4-OHT), which allows the Cre-lox recombination only in melanocytes and leads to the melanoma formation within 5–7 weeks (Fig. 2A) [43]. This model shows a translationally relevant genetic profile and mimics the molecular pathology of human malignant melanoma, providing a basis for studying the molecular mechanisms underlying melanoma progression and for the pre-clinical evaluation of targeted therapies.

Melanoma induction in the B6-Tyr::CreERT2; BRafCA; Pten.loxP mouse model. A Schematic representation of the timeline for tumor induction and formation. B Representative images of a mouse used in this study showing tumor formation and progression after induction with 4-OHT. Green arrow indicates an induced nodular-like tumor, red arrow indicates a spontaneous tumor formed due to Cre-recombinase “leakage” — Fig. 2Melanoma induction in the B6-Tyr::CreER^T2; BRaf^CA; Pten.^loxP mouse model. A Schematic representation of the timeline for tumor induction and formation. B Representative images of a mouse used in this study showing tumor formation and progression after induction with 4-OHT. Green arrow indicates an induced nodular-like tumor, red arrow indicates a spontaneous tumor formed due to Cre-recombinase “leakage”

Despite these advantages, the model presents challenges due to its aggressiveness and a high rate of spontaneous tumor formation—60 to 90% of mice develop tumors even without 4-OHT induction. This "leakiness" in Cre recombinase expression and activity leads to spontaneous activation of Braf^V600E and Pten deletion, resulting in tumorigenesis without the controlled chemical induction [44, 45]. Because the spontaneous tumors vary in consistency and cannot be reliably used for controlled experiments, they were omitted from this study (Fig. 2B). A precise induction was critical for obtaining accurate melanoma formation. The induction over a large skin area resulted in multiple tumors, complicating volume assessment during treatment. Therefore, it was crucial to minimize the induction area to avoid widespread melanoma formation. Precise application of 4-OHT to a confined area of the skin was essential, avoiding any spillage that could cause multiple tumors. Additionally, we found that in some cases, a single application of 4-OHT was insufficient for melanoma development, and repeated inductions over three consecutive days were necessary.

The present study provides novel insights into the metabolism of bioactive lipids in the context of BRaf^V600E-driven malignant melanoma. We found that certain aggressive and metastatic nodular-like melanomas with low NAEs but high 1-stearoyl-2-arachidonoyl-sn-glycerol (SAG), 2-AG, and ARA levels can progress without MAGL and upregulate CES1 to drive the hydrolysis of 2-AG, 2-oleoylglycerol (2-OG), and other glycerol esters (Fig. 1). This finding highlights a previously underexplored compensatory mechanism that may help melanomas to sustain their growth and progression through alternative lipid metabolic pathways characterized by high SAG, 2-AG, and ARA levels and CES1/Ces1 overexpression in a subtype of BRAF^V600E-driven malignant melanoma. We show that in the absence of MAGL, CES1 significantly contributes to the hydrolysis of 2-AG and other glycerol esters in the skin, inhibiting melanoma progression. Importantly, our in vivo experiments revealed that in tumors with elevated SAG levels, inhibiting Ces1c markedly impairs melanoma tumor growth, highlighting CES1 as a potential therapeutic target for malignant melanoma.

We identified Ces1c as the key isoform present in these nodular-like melanomas. Our in vivo experiments revealed that covalent Ces1c inhibition in the absence of MAGL leads to attenuation of melanoma tumor growth. This study not only sheds light on the metabolic flexibility of melanomas but also suggests the exploration of targeted therapeutic interventions involving CES1 inhibition.

Materials and methods

Substances and reagents

JZL184 was purchased from MedChem Express (HY-15249); WWL113 (SML1179) and MAFP (M2939) were purchased from Sigma-Aldrich; URB597 (10,046) and THL (orlistat, 10,005,426) were purchased from Cayman chemical; all at HPLC purity of > 98%. ActivX™ TAMRA-FP serine hydrolase probe was purchased from Thermo-Scientific (88,318). All amino acids and metabolites and their deuterated standards were analytical grade and were purchased from Sigma-Aldrich (Switzerland). Antibodies and other reagents are described in their corresponding sections. Radiolabeled 2-oleoylglycerol ([³H]−2-OG) was obtained from the American Radiolabeled Chemicals, Inc., MO, USA (ART 0400–250 µCi).

Animal experiments

C57BL/6 J BRaf^CA, Pten^loxP/loxP, and Pten^loxP/loxP, Tyr::CreER^T2 mice were imported from McMahon’s laboratory and bred to get the Braf^CA, Pten^loxP, Tyr::CreER^T2 experimental mice. Animals were kept in a temperature-controlled and humidity-controlled room (23 ± 2ºC; 50 ± 1%) with a 12:12 h light–dark cycle, having access to food and water ad libitum. Animals were handled according to Swiss federal legislation and guidelines of the local animal facility at the University of Bern. The protocols were approved by the respective government authorities (Veterinäramt Kanton Bern, experimental license BE-79/19).

Melanoma induction

A cohort of 24 male and female mice was randomly divided into two groups to prevent bias and melanoma tumors were induced at the age of 6 weeks: mice were anesthetized by intraperitoneal (i.p.) injection of a mix of Dorbene (0.5 mg/kg), Dormicum (5 mg/kg) and Fentanyl (0.05 mg/kg) in 0.9% sterile NaCl. The back of each mouse was shaved and depilated with “Veet” cream, and a spot was tattooed with Ketchum green tattoo paste (Ketchum, QKI147325087). After, 1 µl of 1 mg/ml 4-OH-Tamoxifen in DMSO was applied right beneath the tattooed spot. The liquid was left for 30 min and removed with a humid paper towel. The anesthesia was antagonized by subcutaneous injection of Alzan (2.5 mg/kg), Flumazenil (0.5 mg/kg), and Naloxone (1.2 mg/kg) in 0.9% sterile NaCl. This procedure was repeated for three consecutive days.

JZL184 treatment in vivo and sample collection

Six weeks post-induction, mice were administered daily for three weeks with 10 mg/kg of JZL184 or vehicle (Ethanol: Kolliphor RH 40: 0.9% NaCl (1:1:8)) by intraperitoneal (i.p.) injection. Tumor volume was calculated using the formula area x height, where tumor height was measured weekly using a digital caliper, and the tumor area was calculated using ImageJ 1.53q software [46]. At the end of the 3rd week of treatment, animals were sacrificed by i.p. injection of ketamine/xylazine mix (87.5/12.5 mg/kg), tissues were collected, snap-frozen, and stored at −80 °C. The samples from the non-treated group, which were used for LC–ESI–MS/MS analysis, were collected 12 weeks post-induction.

Cell culture

HEK293 cells (human embryonic kidney; RRID:CVCL_0045) were purchased from ATCC—Global Bioresource Center (CRL-1573). For maintenance, the cells were cultured at 37 °C with 5% CO₂ humidified atmosphere in fully supplemented media: Dulbecco’s Modified Eagle Medium (DMEM) high glucose (Sigma-Aldrich, D6546), supplemented with 10% fetal bovine serum (FBS, Sigma-Aldrich).

Transfection and selection

The CES1-expressing vector pcDNA3.1( +)-P2A-eGFP was purchased at Sino Biological (OHu18153C). For the stable transfection, HEK293 cells were seeded into 6-well plates at 40%—60% confluency, allowing them to attach for 24 h in fully supplemented media. Then, the cells were transfected with 2 µg of plasmid using jetPRIME transfecting reagent (Polyplus, 114–07), following the manufacturer’s protocol and replacing the media with fresh media 4 h post-transfection. Two days post-transfection, cells were subjected to antibiotic selection with 400 µg/ml Geneticin (G418 Sulfate; Thermo Fisher, 10,131,027) for 2–3 weeks. To select the single-clone-derived colonies, the clonal selection was performed by seeding a single cell suspension in 96-well plates, followed by their expansion. The presence of CES1 expression was confirmed by the reporter GFP signal, ABPP, and Western blot.

2-Oleoylglycerol hydrolysis assay

2-Oleoylglycerol (2-OG) hydrolysis assay was performed on the stable HEK293-CES1 cells. Cells were seeded in 24-well plates, 150.000 cells/well. 24 h post-seeding, cells were pre-incubated with 1 µM of JZL184, WWL113, or DMSO for 30 min. Then, 1 µM of 2-OG mix (1000:1 ratio of 2-OG to [³H]−2-OG) was added for 30 min. To stop the reaction the cells and the supernatant were mixed with ice-cold methanol/chloroform (1:1, v/v, 1 mL). The samples were vortexed thoroughly and centrifuged at 16,000 × g, 10 min, 4 °C. The aqueous and organic phases were collected in scintillation vials, mixed with 3 ml of scintillation cocktail Ultima Gold (Revvity Health Sciences, Inc., 60,133,299), and analyzed on Packard Tri-Carb 2100 TR scintillation counter. The normalized counts are defined as % of hydrolyzed 2-OG relative to DMSO control.

Mouse tissue sample preparation

The tissues were cryopulverized with the use of a mortar and pestle. After, the samples were homogenized by bead beating with chrome steel 2.3 mm beads (BioSpec, NC0605481), using a Mini-Beadbeater-24 (Bio Spec Products Inc.) for 1 min in 2 mg/ml solution of collagenase IV (Worthington, LS004189) in PBS + CaCl²⁺ + Mg²⁺. The tumor samples were incubated at 37 °C, shaking for 2 h, and the skin samples for 3 h. After, the samples were filtered with 70 µm filters and centrifuged at 16.000 rpm, 4 °C for 25 min. The supernatant was transferred to a new tube, and the centrifugation was repeated. This yielded the membranal fraction as a pellet and the cytosolic fraction as the supernatant. Total protein was quantified by Pierce™ BCA Protein Assay Kit (Thermo-Scientific, 23,225) following the manufacturer’s protocol. The samples were diluted to 3 mg/ml and kept at −80 °C.

Activity-based protein profiling (ABPP)

ABPP experiments were performed using cytosolic protein fraction preparations obtained from mouse tumor and skin samples. To identify the serine hydrolases of interest, 30 µg of proteins were preincubated with indicated concentrations of MAFP, WWL113, JZL184, URB597, THL, or DMSO as control for 30 min at room temperature (RT), shaking. Then, the samples were labeled with a TAMRA-FP serine hydrolase probe at a final concentration of 125 nM for 30 min at RT. The reaction was stopped with 4 × Laemmli Buffer (200 mM Tris–HCl, pH 6.8, 8% (w/v) sodium dodecyl sulfate (SDS), 0.15% bromophenol blue, 40% glycerol, 5% (v/v) beta-mercaptoethanol), the samples were boiled at 70 °C for 5 min, cooled down to RT and separated by SDS-PAGE. The gels were scanned with Typhoon 9400 imager using TAMRA filter (λ_ex 550 nm and λ_em 570 nm). After, the gels were stained with coomassie to quantify the homogeneity of sample loading or instead continued with the Western blot analysis. The ABPP experiments were also performed to confirm CES1 expression. 20 µg of cell homogenates were used and the ABPP followed the same protocol described above.

Western blot analysis

Protein extracts of the cytosolic fractions were separated using 10–12% SDS-PAGE, and the proteins were transferred onto a nitrocellulose membrane in transfer buffer (18 mM Tris, 140 mM glycine; 20% v/v methanol, pH 8.0) using the Trans-Blot Turbo Transfer System (BioRad, #1,704,150). The membranes were blocked with 5% non-fatty milk in tris-buffered saline with 0.1% Tween 20 (TBST) overnight at 4 °C and then incubated with corresponding primary antibodies for 2 h at RT. Afterward, the membranes were washed with TBST three times and incubated with the corresponding secondary antibody for 2 h at RT. Membranes were washed with TBST, and the presence of antibodies was detected with Typhoon FLA 9500 (GE Healthcare). The following antibodies were used for immunoblotting analysis: Rabbit anti-CES1 (ABclonal Cat# A11478, RRID:AB_2861576) primary antibody (1:3000), Alexa Fluor 488 Goat Anti-Rabbit (Thermo Fisher Scientific Cat# A-11034, RRID:AB_2576217) secondary antibody (1:3000). Rabbit anti-GAPDH (Cell Signaling Technology Cat# 2118, RRID:AB_561053) was used as a loading control (1:3000).

Deglycosylation via PNGase F treatment

Cytosolic fraction of mouse melanoma and skin tissues (20 μg) were deglycosylated by treatment with PNGase F at 37 °C for 17 h. The samples were labeled with TAMRA-FP serine hydrolase probe (125 nM final concentration) for 30 min and separated by SDS-PAGE.

Sample preparation for LC–ESI–MS/MS assay

Mouse melanoma and skin samples were cryopulverized with the use of a mortar and pestle and transferred to 2 ml XXTuff reinforced microvials (Bio Spec, 330TX), weighted, and 4 volumes of acetonitrile (for endocannabinoids and bioactive lipids) or 1 mM EDTA in H₂O (for amino acids and neurotransmitters) were added. The homogenization was performed by bead beating with chrome steel 2.3 mm beads (BioSpec, NC0605481) using a Mini-Beadbeater-24 (Bio Spec Products Inc.). For protein precipitation, amino acids, and neurotransmitters analysis, 50 μl of the sample were mixed with 200 μl of methanol (MeOH) and centrifuged at 16.000 rpm, 4 °C for 10 min. For endocannabinoids and bioactive lipids analysis, this step was avoided. Then, 50 μl of supernatant was recovered, mixed with 5 μl of internal standard, and 10 μl were injected into the LC–ESI–MS/MS system.

LC–ESI–MS/MS method for amino acid and neurotransmitter quantification

LC–ESI–MS/MS method for endocannabinoids and bioactive lipids quantification

A hybrid triple quadrupole 4000 QTRAP mass spectrometer (AB Sciex Concord, Ontario, Canada) coupled to a Shimadzu UFLC (Shimadzu Corporation, Kyoto, Japan) with a cooled auto-sampler was used. Sample temperature was maintained at 4 °C in the auto-sampler during analysis. The chromatographic separation was carried out on the Reprosil‐PUR C18 column, 3 μm particle size; 2 × 50 mm, Dr. A. Maisch, High-Performance LC‐GMBH (Ammerbuch, Germany) maintained at 40 °C. The system was operated in positive and negative modes with two different elution mobile phases and gradient methods. In positive mode, the elution mobile phases were composed of 2 mM ammonium acetate: water: formic acid (95:5:0.1, v/v/v) (Mobile Phase A) and 2 mM ammonium acetate in methanol (Mobile Phase B). The gradient started at 15% B, increasing linearly to 70% at 3.5 min, then to 99% at 8 min, maintained until 12 min, with subsequent re‐equilibration at 15% for 2.5 min. The flow rate was 0.35 ml/min. In negative mode, the organic mobile phase (B) was substituted with acetonitrile 0.1% formic acid to minimize crosstalk between analytes with the same mass ([M‐H]‐, m/z 351 PGE2/D2). The gradient started at 5% B, increasing linearly to 40% at 3 min, then to 65% at 9 min, and linearly again to 95% at 10 min; this was maintained until 14 min, with subsequent re‐equilibration at 5% for 3 min. The flow rate was 0.3 ml/min. Peaks were integrated, and the Analyst software version 1.6.3. (AB Sciex Concord, Ontario, Canada) was used for quantification. Identification of compounds in samples was confirmed by comparison of precursor and product ion m/z values and LC retention times with standards. Further details regarding LC–ESI–MS/MS conditions are described in Supplementary Material.

Samples biotinylation

Two hundred µl of each sample (3 mg/ml) were mixed with ActivX™ Desthiobiotin-FP Serine Hydrolase Probe (Thermo-Scientific, 88,317) at a final concentration of 2 μM. The samples were vortexed and incubated at RT, shaking for 1 h. The excess of the probe was washed with PBS using Vivaspin® 2, 2.000 MWCO Hydrosart filtration columns (Sartorius, VS02H91), centrifuged at 5.000 g, 4 °C for 20 min. The supernatant was collected in a new tube and diluted with 2.5% SDS until the final concentration of 0.5% SDS. The samples were incubated at 95 °C for 5 min to denature the proteins. After cooling down to RT, the samples were diluted with PBS until a final concentration of 0.2% SDS and incubated with Pierce™ Avidin Agarose beads (Thermo-Scientific, 20,219) for 1 h while rotating. The beads were isolated by centrifugation at 1.600 g for 1 min and washed with 0.2% SDS (3 times) and PBS (5 times) using Micro Bio-Spin™ Columns (BioRad, 7,326,204). The beads were then transferred to 1.5 ml screw-cap tubes and labeled for submission.

Proteomics. Sample preparation and LC–ESI–MS/MS analysis

The proteomic analysis was carried out at the Protein and Proteomics Analyses Functional Genomics Centre, Zurich. The beads were washed twice with 100 μl of 10 mM Tris, 2 mM CaCl₂, pH 8.2, and the wash was discarded; next, the reduction and alkylation with TCEP and chloroacetamide were performed with 45 μl of 10 mM Tris, 2 mM CaCl2, pH 8.2 buffer and 5 μl of Trypsin (100 ng/μl in 10 mM HCl). The digestion was performed at 800 rpm, 37 °C for 4 h. Supernatants were collected, and the peptides were extracted from beads with 150 μl of 0.1% TFA 50% acetonitrile solution. The supernatants were combined and dried. The digested samples were dried and dissolved in 20 μl 0.1% formic acid and transferred to the autosampler vials for LC–ESI–MS/MS analysis. 2 μl were injected on a nanoAcquity UPLC coupled to a Q-Exactive mass spectrometer (Thermo-Scientific).

Statistical analyses

The acquired MS data from the proteomics analysis were converted to a Mascot Generic File format (.mgf files) and were processed for identification using the Mascot search engine (Matrix Science). The spectra were searched against the Mus musculus database. The following modifications were set—variable modifications: Oxidation (M); fixed modifications: carbamidomethylation ©. The differential expression analysis was performed by the Functional Genomics Center, Zurich. The report is stored in the LIMS system bfabric [47] under order 31,647, with the work unit name 288,771. The protein identification and quantification were performed using FragPipe. For filtering and normalizing the data, generating visualizations, and computing differential expression analysis, R statistical software package prolfqua was used [48]. To further improve the power of the differential expression test, the protein variances were moderated [49]. Gel-based ABPP band intensity was quantified using ImageJ 1.53q software, and data are shown as the mean ± standard deviation (SD). LC–ESI–MS/MS data was analyzed by two-tailed paired t-test; the multiple comparisons were analyzed by two-way analysis of variance (ANOVA) and Tukey’s or Šídák’s post hoc test, where applicable. The post hoc test was only performed when F in ANOVA achieved P < 0.05. To analyze potential associations between the analytes in skin and melanoma tissues, Spearman correlation coefficients were calculated (two-tailed, 95% confidence interval). All statistical analyses were performed using GraphPad Prism for Microsoft Windows® (Versions 9, GraphPad software®, CA, USA) where P < 0.05 was accepted as significant; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Results

Differential regulation of bioactive lipids in melanoma

To assess the metabolic profiles of malignant melanoma and normal skin tissues, we employed LC–ESI–MS/MS-based targeted metabolomics on samples from skin and melanoma tissues obtained from the induced B6-Tyr::CreER^T2; BRaf^CA; Pten^loxP mice (Fig. 3). We first investigated potential changes in the metabolic signatures obtained by targeted metabolomicsTo further investigate the metabolic differences between malignant melanoma and normal skin tissues, we compared their metabolic profiles using a radar chart, emphasizing key variations in metabolite abundances (Fig. 3C). The most striking difference was observed in serotonin (5-HT) levels, which showed a 5.3-fold increase in melanoma tissues (Fig. 3J). Given our interest in the 2-AG-arachidonate-prostaglandin axis, another notable change observed in melanoma tissues was in the increased levels of SAG 3.3-fold (Fig. 3D). Additionally, 2-AG and ARA levels were increased in melanoma by 1.3- and 0.7-fold, respectively (Fig. 3E,G). This suggested a potential increase in the enzymatic activity responsible for the hydrolysis of SAG to 2-AG and, subsequently, from 2-AG to ARA. Nevertheless, the levels of prostaglandins, downstream products of ARA metabolism (Fig. 1), remained unchanged between melanoma and normal skin tissues (Fig. 3H,I). Furthermore, we observed a significant increase in other bioactive molecules, such as adenosine (Fig. 3K) and GABA (Fig. 3L). In contrast, all long-chain fatty acid N-acylethanolamines—AEA, LEA, PEA, OEA, SEA—were significantly reduced in melanoma compared to normal skin (Fig. 3M-Q). Supplementary Figure S1 shows additional analyzed bioactive lipids and neurotransmitters.

Fig. 3Metabolic signatures reveal that SAG, 2-AG, and ARA are increased in melanoma. Spearman correlation heatmaps of quantified metabolites in (A) skin and (B) melanoma. Red color indicates positive (> 0), white color indicates zero (0), blue color indicates negative (< 0) correlation, and the color intensity represents variation in the levels across the color scale. C Radar chart illustrating the levels of amino acids, neurotransmitters, and bioactive lipids in melanoma and normal skin, quantified by LC–ESI–MS/MS. Data are shown as fold change of the mean values, melanoma levels shown in red, and skin levels in blue, using skin levels as the baseline (fold change 0). D-Q Representative charts of endogenous levels of bioactive lipids in melanoma and skin. Data are shown as violin plots (N = 14 for each group) and analyzed by two-tailed paired t-test; ***P < 0.001; ****P < 0.0001. Abbreviations: 2-AG, 2-arachidonoylglycerol; 2-OG, 2-oleoylglycerol; 20:4 PE, 1,2-diarachidonoyl-sn-glycero-3-phosphoethanolamine; 5-HT, serotonin; Ach, acetylcholine; Ado, L-adenosine; AEA, anandamide; Ala, L-alanine; ARA, arachidonic acid; Arg, L-arginine; Asn, L-asparagine; Asp, L-aspartic acid; C18-20:4 PC, 1-(1Z-octadecenyl)−2-arachidonoyl-sn-glycero-3-phosphocholine; C18-20:4 PE, 1-(1Z-octadecenyl)−2-arachidonoyl-sn-glycero-3-phosphoethanolamine; Cho, choline; CRT, corticosterone; Cys, L-cysteine; ETA, ethanolamine; GABA, gamma-aminobutyric acid; Glu, L-glutamic acid; Gln, L-glutamine; Gly, L-glycine; GSH, glutathione; His, L-histidine; Hist, histamine; LEA, linoleoyl ethanolamide; Leu, L-leucine; Lys, L-lysine; Met, L-methionine; NE, norepinephrine; OEA, oleoyl ethanolamide; PEA, palmitoyl ethanolamide; PGD2, prostaglandin D2; PGE2, prostaglandin E2; Phe, L-phenylalanine; Pro, L- proline; SAG, 1-stearoyl-2-arachidonoyl-sn-glycerol; SEA, stearoyl ethanolamide; Ser, L-serine; Tau, L-taurine; Trp, L-tryptophan; Tyr, L-tyrosine; Val, L-valine

Expression of Ces1 in melanoma and normal skin

Because of the increased SAG, 2-AG, and ARA levels in melanoma tissues and our interest in the source of ARA, we conducted a targeted proteomics analysis to identify the enzymes responsible for the hydrolysis of these bioactive lipids. It is important to note that the desthiobiotin-FP serine hydrolase probe used in this study does not detect DAGLα/β, the enzymes responsible for SAG hydrolysis. Therefore, our investigation centered on serine hydrolases capable of hydrolyzing 2-AG and leading to ARA production.

Our chemoproteomics analysis yielded 35 serine hydrolases, comprising 22 metabolic serine hydrolases and 13 serine proteases (Fig. 4). Notably, two serine proteases—plasminogen (Plg) and the CAAX prenyl protease 1 homolog (Zmpste24)—were detected only in melanoma, while two metabolic serine hydrolases—phospholipase B-like 1 (Plbd1) and carboxylesterase 1F (Ces1f)—were found only in normal skin. The remaining 31 enzymes were detected in both melanoma and normal skin tissues (Fig. 4A).

Serine hydrolases found in melanoma and normal skin tissues. A Heatmap depicting the − log10 transformed False Discovery Rate (FDR) of the differential expression analysis of serine hydrolases between mouse melanoma and skin. Red color indicates positive (> 0), white color indicates zero (0), blue color indicates negative (< 0) expression differences and the color intensity represents variation in the levels across the color scale. Pink color represents proteins present only in melanoma; green color represents proteins present only in the skin. B Volcano plot showing − log10 transformed FDR of the differential analysis between Diseased (melanoma) and Control (skin) groups. Red line indicates the − log10(FDR) of FDR = 0.1, green lines indicate the difference of −1 and + 1. Differences and FDRs estimated using missing value imputation are shown in orange. Metabolic serine hydrolases are shown in blue; serine proteases are shown in red. Isoforms of Ces1 and Ces2 are indicated on the graph. C Representative gel-based ABPP of mouse melanoma and skin samples. Samples were incubated with the activity probe TAMRA-FP (125 nM) for 30 min at RT. D Quantification of Ces1c band intensity from the gel-based ABPP. Data are shown as mean ± SD from three independent experiments. Representative gel-based ABPP of mouse skin (E) and melanoma (G) samples. Samples were incubated with the indicated concentrations of MAPF, WWL113, JZL184, URB297, and THL for 30 min at RT before incubation with the activity probe TAMRA-FP (125 nM) for 30 min at RT. Quantification of Ces1c band intensity from the gel-based ABPP of mouse skin (F) and melanoma (H) samples. Data are shown as mean ± SD from four independent experiments, performed in triplicates. Abbreviations: M – melanoma, S – skin — Fig. 4Serine hydrolases found in melanoma and normal skin tissues. A Heatmap depicting the − log₁₀ transformed False Discovery Rate (FDR) of the differential expression analysis of serine hydrolases between mouse melanoma and skin. Red color indicates positive (> 0), white color indicates zero (0), blue color indicates negative (< 0) expression differences and the color intensity represents variation in the levels across the color scale. Pink color represents proteins present only in melanoma; green color represents proteins present only in the skin. B Volcano plot showing − log₁₀ transformed FDR of the differential analysis between Diseased (melanoma) and Control (skin) groups. Red line indicates the − log₁₀(FDR) of FDR = 0.1, green lines indicate the difference of −1 and + 1. Differences and FDRs estimated using missing value imputation are shown in orange. Metabolic serine hydrolases are shown in blue; serine proteases are shown in red. Isoforms of Ces1 and Ces2 are indicated on the graph. C Representative gel-based ABPP of mouse melanoma and skin samples. Samples were incubated with the activity probe TAMRA-FP (125 nM) for 30 min at RT. D Quantification of Ces1c band intensity from the gel-based ABPP. Data are shown as mean ± SD from three independent experiments. Representative gel-based ABPP of mouse skin (E) and melanoma (G) samples. Samples were incubated with the indicated concentrations of MAPF, WWL113, JZL184, URB297, and THL for 30 min at RT before incubation with the activity probe TAMRA-FP (125 nM) for 30 min at RT. Quantification of Ces1c band intensity from the gel-based ABPP of mouse skin (F) and melanoma (H) samples. Data are shown as mean ± SD from four independent experiments, performed in triplicates. Abbreviations: M – melanoma, S – skin

Unexpectedly, the most frequently expressed 2-AG-hydrolyzing enzymes MAGL, ABHD6, and ABHD12 were absent in melanoma and skin tissues. The only identified enzymes known to hydrolyze 2-AG were the carboxylesterases Ces1 and Ces2. Three Ces1 isoforms were detected, Ces1c, Ces1d, and Ces1f, along with two Ces2 isoforms, Ces2c and Ces2e (Fig. 4A,B). Another Ces2 isoform involved in 2-AG hydrolysis, Ces2g, was not detected in melanoma or skin tissues. Interestingly, differential expression analysis showed that Ces1c expression did not significantly differ between melanoma and normal skin. In contrast, Ces1d expression was dramatically reduced in melanoma, and Ces1f was only detected in the normal skin (Fig. 4A,B). The expression profiles of Ces2 isoforms also varied, with Ces2e being elevated in melanoma, while Ces2c was more abundant in normal skin (Fig. 4A,B).

Activity-based protein profiling (ABPP) of melanoma and skin samples revealed a very prominent band around ∼70 kDa (Fig. 4C). This band was more pronounced in melanoma samples, suggesting an increased enzyme activity compared to the normal skin (Fig. 4D). In addition to MAFP, a potent non-specific inhibitor of serine hydrolases, CES1 was also inhibited by the MAGL inhibitor JZL184 and WWL113, known CES1/Ces1 inhibitors in both humans and mice [20, 50, 51] (Fig. 4E-H). CES1 was identified as the Ces1c isoform, previously described to appear at ∼70 kDa due to extensive glycosylation, thus distinguishing it from other Ces1 isoforms (Supplementary Figure S2) [52].

Ces1 inhibition impairs melanoma growth in vivo

Given the elevated levels of SAG, 2-AG, and ARA observed in mouse melanomas, along with the detection of active Ces1 isoforms in both melanoma and skin tissues, we next evaluated the effects of in vivo Ces1 inhibition on melanoma progression. Based on the in vitro results (Fig. 4E-H) and the absence of MAGL (Fig. 4A), we selected JZL184 for our in vivo study, as this MAGL-specific inhibitor also covalently inhibits CES1 [20, 50, 51] and the compound is established as for in vivo studies. Mice were treated with 10 mg/kg of JZL184 or vehicle, administered daily via intraperitoneal injection (i.p.) for three consecutive weeks (Fig. 5A). After two weeks of daily treatment, JZL184 significantly but moderately slowed melanoma progression compared to the vehicle-treated group, an effect that persisted through the third week of treatment (Fig. 5B). Since this melanoma model is very aggressive, this effect was also therapeutically significant.

Fig. 5CES1 inhibition impairs melanoma growth. A Schematic representation of the experimental design and treatment schedule. B Normalized tumor volume of JZL184- and vehicle-treated groups. The initial tumor volume was considered 100% for both groups. JZL184 and vehicle (Ethanol: Kolliphor RH 40: 0.9% NaCl (1:1:8)) were administered by i.p. injection. Data are shown as mean ( +) and minimum to maximum whiskers (N = 14 for each group) and analyzed by two-way ANOVA with multiple comparisons and Šídák’s post hoc test; *P < 0.05. C Representative gel-based ABPP of mouse melanoma and skin samples after the in vivo treatment with JZL184 or vehicle. Samples were incubated with the activity probe TAMRA-FP (125 nM) for 30 min at RT. D) Quantification of the Ces1c band intensity from the gel-based ABPP of mouse melanoma and skin samples after the in vivo treatment with JZL184 or vehicle. Data are shown as mean ± SD from two independent experiments (N = 12 for each group) and analyzed by unpaired t-test ****P < 0.0001. E Representative gel-based ABPP and immunoblot overlay of mouse melanoma and skin samples after the in vivo treatment with JZL184 or vehicle. Samples were first incubated with the activity probe TAMRA-FP (red) (125 nM) for 30 min at RT, then transferred to a nitrocellulose membrane and blotted with anti-CES1 antibody (green)

To confirm the efficacy of the in vivo Ces1 inhibition, we conducted ex vivo ABPP on both melanoma and skin samples, showing that Ces1 was effectively inhibited in the JZL184-treated group, indicated by a clear reduction in the ex vivo ABPP(Fig. 5C,D). The ABPP/WB overlay further validated the in vivo inhibition of the Ces1c isoform, confirming it as the primary target of JZL184 and reinforcing the specificity of the compound in inhibiting Ces1 activity in the treated samples (Fig. 5E). Notably, inhibition of Ces1d/f isoforms was also observed in skin samples (Fig. 5E), demonstrating that JZL184 effectively targets multiple Ces1 isoforms in vivo.

Covalent inhibition of Ces1 leads to increased 2-AG and 2-OG skin levels in vivo

We performed an in-depth targeted metabolomic analysis using LC–ESI–MS/MS to quantify the alterations in molecular signatures of the mouse melanomas and skin following in vivo administration of JZL184 (Fig. 6A,B). As expected, the most significant changes were observed in the levels of 2-AG and 2-OG, which strongly increased after in vivo JZL184 treatment. Specifically, 2-AG showed a twofold increase in melanoma and a 0.7-fold increase in the skin, while 2-OG exhibited a 1.7-fold increase in melanoma and a 1.5-fold increase in the skin (Fig. 6C,D). These findings indicate that Ces1 mediates the hydrolysis of both 2-AG and 2-OG in melanoma and normal skin tissues of the B6-Tyr::CreER^T2; BRaf^CA; Pten^loxP mice. To confirm the ability of CES1/Ces1 to hydrolyze 2-AG/2-OG, we generated a stable CES1-expressing cell line using HEK293 cells, which hydrolyzed approximately 3.5 times more 2-OG than the wild type cells, and this hydrolysis was effectively inhibited by the specific CES1 inhibitors JZL184 and WWL113 (Fig. 6I).

Fig. 6JZL184 increases 2-AG and 2-OG levels in vivo. Radar charts illustrating the levels of amino acids, neurotransmitters, and bioactive lipids in (A) skin and (B) melanoma samples, quantified by LC–ESI–MS/MS. Data are shown as mean values of bioactive molecule amounts of the JZL184-treated group (yellow) compared to the vehicle group (green) used as 100%. C-H Representative charts of endogenous levels of bioactive lipids in melanoma and skin after JZL184 treatment in vivo. Data are shown as violin plots (N = 12 for each group) and analyzed by two-way ANOVA with multiple comparisons and Tukey’s post hoc test; *P < 0.05; **P < 0.01; ****P < 0.0001. I Normalized % of 2-OG hydrolysis in HEK293 and HEK293-CES1 cells. Samples were preincubated with 1 μM JZL184 or WWL113 for 30 min, followed by 1 μM 2-OG incubation for 30 min. Data are shown as mean ± SD from three independent experiments performed in triplicates and analyzed by two-way ANOVA with multiple comparisons and Šídák's post hoc test; ****P < 0.0001. Abbreviations: 2-AG, 2-arachidonoylglycerol; 2-OG, 2-oleoylglycerol; 20:4 PE, 1,2-diarachidonoyl-sn-glycero-3-phosphoethanolamine; Ach, acetylcholine; Ado, L-adenosine; AEA, anandamide; Ala, L-alanine; ARA, arachidonic acid; Arg, L-arginine; Asn, L-asparagine; Asp, L-aspartic acid; C18-20:4 PC, 1-(1Z-octadecenyl)−2-arachidonoyl-sn-glycero-3-phosphocholine; C18-20:4 PE, 1-(1Z-octadecenyl)−2-arachidonoyl-sn-glycero-3-phosphoethanolamine; Cho, choline; CRT, corticosterone; Cys, L-cysteine; ETA, ethanolamine; GABA, gamma-aminobutyric acid; Glu, L-glutamic acid; Gln, L-glutamine; Gly, L-glycine; GSH, glutathione; His, L-histidine; LEA, linoleoyl ethanolamide; Leu, L-leucine; Lys, L-lysine; Met, L-methionine; NE, norepinephrine; OEA, oleoyl ethanolamide; PEA, palmitoyl ethanolamide; PGD2, prostaglandin D2; PGE2, prostaglandin E2; Phe, L-phenylalanine; Pro, L- proline; SAG, 1-stearoyl-2-arachidonoyl-sn-glycerol; SEA, stearoyl ethanolamide; Ser, L-serine; Tau, L-taurine; Trp, L-tryptophan; Tyr, L-tyrosine; Val, L-valine

Another interesting observation was the 0.3-fold reduction in ARA levels in melanomas following in vivo JZL184 administration (Fig. 6F), whereas the levels of SAG and prostaglandins remained consistent between the two treatment groups (Fig. 6E,G,H). Additionally, melanomas from the JZL184-treated group exhibited elevated levels of GABA and phenylalanine (Supplementary Figure S3). The long-chain fatty acid N-acylethanolamines—AEA, LEA, PEA, OEA, SEA – did not show any variations after the treatment (Supplementary Figure S4).

Distinct expression patterns of CES1 and MAGL in melanoma from RNA-seq data

Following up on our findings regarding the role of CES1/Ces1 in 2-AG/2-OG hydrolysis and melanoma progression, we analyzed the available RNA sequencing (RNA-seq) data on the expression of 2-AG-hydrolysing enzymes in both human malignant melanoma and human skin tissues. The expression of CES1 in human skin was confirmed by ABPP/WB overlay using normal human skin tissue (Supplementary Figure S5). Figure 7 shows an integrative analysis of single-cell RNA sequencing (scRNA-seq) data for CES1 and MAGL expression across 33 human melanoma tissues, as reported by Tirosh et al. and Jerby-Arnon et al. [53, 54]. This analysis is complemented by Supplementary Figure S6, which shows the expression levels of other 2-AG-hydrolyzing enzymes—ABHD6 and ABHD12.

Fig. 7CES1 and MAGL expression in human malignant melanoma. Scatter plots of the integrative analysis of scRNA-seq data of human melanoma samples with a t-stochastic neighbor embedding (t-SNE) of single-cell profiles (dots) of (A) malignant and (B) non-malignant cells, colored by (A) tumor or by (B) post hoc annotation. A Only tumors with at least 50 malignant cells are shown. Expression of CES1 in (A’) malignant and (B’) non-malignant cells. Expression of MAGL in (A’’) malignant and (B’’) non-malignant cells. Heatmaps depicting CES1 and MAGL expression in (C) malignant and (D) non-malignant cells. Red color indicates positive expression (> 0), white color indicates zero (0), and the color intensity represents variation in the levels across the color scale. Abbreviations: CAF, cancer-associated fibroblast; NK, natural killer cell [53, 54]

The scRNA-seq data showed that CES1 expression in malignant melanoma cells is generally lower compared to MAGL (Fig. 7A’,A’’). Among the 2-AG-hydrolyzing enzymes, ABHD12 was the most widely expressed (Supplementary Figure S6A’’). Non-malignant immune cells exhibited low CES1 expression, which was not associated with any specific cell type (Fig. 7B’). In contrast, MAGL was highly expressed in macrophages (Fig. 7B’’), as well as ABHD12 (Supplementary Figure S6B’).

Notably, in the malignant cells of the Mel78 sample, we found a remarkable case where CES1 expression was dramatically high (Fig. 7A’,C), while MAGL expression was nearly absent (Fig. 7A’’,C). This finding is particularly meaningful as it mirrors what we observed in the BRaf^V600E/Pten^−/− mice, where high CES1 expression coincided with the absence of MAGL (Fig. 4A). Therefore, the specific expression pattern of 2-AG-hydrolyzing enzymes suggests differential roles of CES1 and MAGL in different melanoma subtypes.

Discussion

Despite significant advances in melanoma treatment, the prognosis for metastatic BRAF^V600E-driven melanoma remains poor [55], highlighting the critical need for a deeper understanding of its underlying mechanisms. Among the numerous pathways implicated in melanoma progression, the ECS potentially plays a role in tumor development and growth [8, 56]. Beyond their critical role as neuromodulators [57], endocannabinoids are involved in numerous physiological processes, including cell proliferation, differentiation, apoptosis, and immune response modulation [58–60]. In agreement, a dysregulation of endocannabinoid metabolism and signaling has been linked to various pathologies, including cancer, where it contributes to immune evasion, tumor growth, and metastasis [56, 59, 61].

In our study, we focused on the lipid remodeling and endocannabinoid-prostaglandin axis, as well as differences in serine hydrolase expression in melanoma versus normal skin. Employing LC–ESI–MS/MS, we examined the bioactive lipid profiles in a genetically engineered BRaf^V600E/Pten^−/− mouse model. In melanoma tissue, we observed consistently reduced NAEs, which may result from decreased NAPLD activity [62, 63], but elevated levels of SAG, 2-AG, and ARA (Fig. 3). The significant increase in SAG levels observed in melanomas suggests a potential role in activating the MAPK/ERK signaling pathway. While diacylglycerols (DAGs) are predominantly recognized for their activation of various protein kinase C (PKC) isoforms [64, 65], they also mediate the activation of other proteins, including RasGRPs [66, 67]. RAS guanyl-releasing protein 1 (RasGRP1) functions as a guanine nucleotide exchange factor, and its activation by the DAGs leads to the activation of the RAS small GTPase protein—a member of the MAPK/ERK signaling pathway—through the exchange of bound GDP for GTP [68]. The presence of RasGRP1 has been confirmed in keratinocytes derived from the skin, and overexpression of this protein has been linked to the promotion of skin tumor formation [69], likely mediated through direct RAS activation [70, 71]. It has previously been shown that basal levels of total diacylglycerol were elevated fivefold in B16 melanoma cells compared to levels found in quiescent or proliferating Mel-ab melanocytes and appear to arise largely from the breakdown of phosphatidylinositol phospholipids accompanied by a significant rise in phospholipase C activity [72]. The genetically engineered mouse model used in this study carries melanocyte-specific BRaf^V600E and Pten^KO that drive the hyperactivation of the MAPK/ERK signaling pathway. The elevated levels of SAG observed in the tumors might potentiate this signaling cascade through PKC activation and enhanced RasGRP1 activation, thereby ensuring persistent and amplified MAPK/ERK pathway activation, thus promoting melanoma progression.

The hydrolysis of membranal SAG by DAG lipases -alpha o -beta leads to the formation of one of the most abundant endocannabinoids, 2-AG, and the structurally similar 2-OG [73, 74]. In our study, 2-AG levels were significantly increased in melanoma tissues compared to normal skin (Fig. 3E). While 2-AG has been implicated in both pro- and anti-tumorigenic activities, the balance of its effects may be context-dependent, depending on receptor signaling or degradation by serine hydrolases feeding the arachidonate pathways. For example, elevated 2-AG levels have been correlated with enhanced glioma tumor growth [75]. In contrast, the inhibition of the primary 2-AG-hydrolyzing enzyme, MAGL, led to elevated 2-AG levels and has been shown to impair cell migration in aggressive breast, ovarian, and melanoma cancer cell lines [76]. Moreover, MAGL inhibition has also been shown to reduce colon carcinoma progression [77] and impair the growth of non-small cell lung cancer [15]. Additionally, JZL184-mediated MAGL inhibition induced an anti-hyperalgesic effect in a mouse model of mechanical hyperalgesia [78]. The hydrolysis of 2-AG leads to the formation of ARA, elevated levels of which are known to promote melanoma progression and tumor proliferation [58, 79, 80]. ARA has also been implicated in the development of other types of cancer, including colorectal cancer [81, 82].

The chronic in vivo administration of JZL184 to the BRaf^V600E/Pten^−/− mice further increased 2-AG and 2-OG levels and decreased ARA levels in melanoma and normal skin (Fig. 6C,D,F), leading to attenuated tumor growth (Fig. 5B). Although JZL184 is a specific MAGL inhibitor, the observed effects were attributed to CES1 inhibition, as a particularly striking finding from our study was the absence of MAGL and any other canonical 2-AG-hydrolyzing enzymes, such as ABHD6 or ABHD12 (Fig. 4A,B). While CES1 has been previously described to hydrolyze 2-AG [20, 83], the role of this enzyme in the skin and malignant melanoma has remained completely unexplored until now. This study is the first to demonstrate that metabolic pathways involving elevated levels of SAG, 2-AG, and ARA may contribute to melanoma progression, with CES1 directly influencing lipid metabolism within the melanoma tumor microenvironment.

Importantly, available scRNA-seq data on human samples confirms that CES1 is predominantly expressed in the liver [24, 26, 84], with particularly high expression levels in hepatocytes, the primary cell type of the liver [85, 86]. Interestingly, in the skin, CES1 appears to be expressed in fibroblasts, where it shows substantial expression, as well as in endothelial cells, stromal cells, and melanocytes, although at lower levels [84, 87–92]. In immune cells, CES1 expression has been detected in monocytes/macrophages, dendritic cells, and NK cells [91–93]. Therefore, the metabolism of bioactive lipids, such as SAG, 2-AG, and ARA may be compartmentalized within the tumor microenvironment, with different cell types potentially contributing to the production and hydrolysis of specific lipids.

We identified the Ces1c isoform as the most active serine hydrolase in melanoma tissues from the BRaf^V600E/Pten^−/− mice (Fig. 4C,D; Supplementary Figure S2). Although the Ces1d and Ces1f isoforms were also detected, their prevalence in normal skin suggests a minor role in tumor progression (Fig. 4A,B). While our targeted proteomics data also identified Ces2c and Ces2e, which were previously reported to hydrolyze 2-AG [35], the primary Ces2 isoform Ces1g, which is generally associated with 2-AG hydrolysis [94], was not detected in our analysis. This suggests that CES1, particularly the Ces1c isoform, is the dominant enzyme responsible for 2-AG degradation in melanoma of the BRaf^V600E/Pten^−/− mouse model. Beyond 2-AG, CES1 has also been shown to hydrolyze prostaglandins [95]. However, we did not observe a decrease in PGE2/PGD2 levels following in vivo JZL184 treatment and CES1 inhibition (Fig. 6G,H), indicating that the Ces1c isoform is not involved in PGE2/PGD2 degradation in BRaf^V600E/Pten^−/− mice. Likewise, the lower anandamide and overall NAEs levels observed in the melanoma was not changed by JZL184 treatment, indicating that the biosynthesis of NAEs via N-acyl phosphatidylethanolamine phospholipase D (NAPEPLD) rather than hydrolysis by CES1 could be involved [62, 63].

Our biochemical findings align with the scRNA-seq data reported by Tirosh et al. and Jerby-Arnon et al., which showed a unique expression pattern of CES1 in malignant melanoma cells [53, 54]. While CES1 expression was generally lower than MAGL (Fig. 7A’,A’’), the notable exception of a tumor exhibiting high CES1 expression with nearly absent MAGL expression (Fig. 7A’,C) suggests a compensatory mechanism for 2-AG hydrolysis, where CES1 assumes a primary 2-AG-hydrolyzing role in the absence of MAGL. This variability in enzyme expression reflects metabolic flexibility and adaptations of melanoma tumors that could be exploited therapeutically. Furthermore, in the melanoma patient samples analyzed, CES1 expression did not associate with any specific immune cell type (Fig. 7B’). This suggests that the presence of CES1 signal in melanoma tissue may originate from various cell types, supporting the previously described scRNA-seq data [91–93]. In contrast, MAGL expression is highly associated with macrophages (Fig. 7B’’), similar to ABHD12 expression (Supplementary Figure S6B’). Future studies will have to explore the long-term effects of CES1 inhibition and the mechanism of action of 2-AG and other lipids, their roles in metastasis, and their impact on overall survival in melanoma, potentially paving the way for novel therapeutic approaches in melanoma treatment.

Conclusions

Targeted metabolomics revealed distinct metabolic signatures in skin and BRaf^V600E driven melanoma tissues, showing notable changes such as increased levels of serotonin, SAG, 2-AG, and ARA, alongside decreased levels of NAEs. Using chemoproteomics, we show the expression of multiple Ces1 isoforms—Ces1c, Ces1d, and Ces1f—in mouse melanoma and normal mouse and human (Ces1) skin. The Ces1c isoform was identified as the primary enzyme driving 2-AG degradation, apparently associated with lipid metabolism and melanoma progression. The potential translational relevance of our findings is supported by scRNA-seq data from Tirosh et al. and Jerby-Arnon et al., which reported a case with elevated CES1 expression in the absence of MAGL. We hypothesize a compensatory mechanism driven by SAG that sustains 2-AG synthesis and hydrolysis, leading to ARA production and favoring tumor growth, as 2-AG is known to slow tumor progression and ARA promotes it. Importantly, CES1 was not responsible for the significantly lower NAE levels in melanoma tissues. In vivo inhibition of Ces1c significantly increased levels of 2-AG and 2-OG without changing the extractable bulk amount of ARA. CES1 inhibition in the tumor tissues led to a marked slowdown in melanoma growth, suggesting a role of 2-AG and potentially local ARA metabolism that cannot be easily detected in the tumor tissue, independent of prostaglandins. CES1 expression could thus be a potential marker for classifying these aggressive melanoma subtypes. Overall, the identification of CES1 activity in melanoma provides new insights into the metabolic adaptations of these tumors and opens avenues for exploring diagnostic or therapeutic opportunities for melanoma characterization and treatment.

Supplementary Information

Acknowledgements

We would like to acknowledge Prof. Dr. med. Robert Hunger from the Department of Dermatology, Inselspital, Bern University Hospital for fruitful discussions.

Abbreviations

2-AG: 2-Arachidonoylglycerol
2-OG: 2-Oleoylglycerol
AAs: Amino acids
ABHD6: Alpha/beta-hydrolase domain-containing protein 6
ABHD12: Alpha/beta-hydrolase domain-containing protein 12
ACh: Acetylcholine
Ado: Adenosine
AEA: Anandamide
Ala: Alanine
ARA: Arachidonic acid
Arg: Arginine
Asn: Asparagine
Asp: Aspartic acid
C18-20:4 PC: 1-(1Z-octadecyl)-2-arachidonoyl-sn-glycero-3-phosphocholine
C18-20:4 PE: 1-(1Z-octadecyl)-2-arachidonoyl-sn-glycero-3-phosphoethanolamine
20:4 PE: 1,2-Diarachidonoyl-sn-glycero-3-phosphoethanolamine
CES: Carboxylesterases
CES1: Carboxylesterase 1
CES2: Carboxylesterase 2
Chol: Choline
CM: Cutaneous melanoma
CRT: Corticosterone
COX1: Cyclooxygenase 1
COX2: Cyclooxygenase 2
Cys: Cysteine
DAGLα: Diacylglycerol lipase α
DAGL β: Diacylglycerol lipase β
ECS: Endocannabinoid system
ETA: Ethanolamine
GABA: gamma-Aminobutyric acid
Glu: Glutamic acid
Gln: Glutamine
Gly: Glycine
GSH: Glutathione
His: Histidine
Hist: Histamine
LAT1: Large neutral amino acid transporter 1
LAT2: Large neutral amino acid transporter 2
LC–ESI–MS/MS: Liquid chromatography coupled to tandem mass electron ionization electrospray spectrometry
LEA: Linoleoyl ethanolamide
Leu: Leucine
Lys: Lysine
MAGL: Monoacylglycerol lipase
MeOH: Methanol
Met: Methionine
NAEs: N-acylethanolamines
NAPE-PLD: N-acyl phosphatidylethanolamine phospholipase D
NE: Norepinephrine
OEA: Oleoyl ethanolamide
PEA: Palmitoyl ethanolamide
PGD2: Prostaglandin D2
PGE2: Prostaglandin E2
Phe: Phenylalanine
PKC: Protein kinase C
Pro: Proline
RasGRP1: RAS guanyl-releasing protein 1
RNA-seq: RNA sequencing
RT: Room temperature
SAG: 1-Stearoyl-2-arachidonoyl-sn-glycerol
scRNA-seq: Single-cell RNA sequencing
SEA: Stearoyl ethanolamide
Ser: Serine
Tau: Taurine
Thr: Threonine
Trp: Tryptophan
Tyr: Tyrosine
Val: Valine
WB: Western blot

Authors’ contributions

V.M.: Investigation and methodology; formal analysis; visualization; manuscript writing—original draft preparation, reviewing, and editing. D.P.: LC–ESI–MS/MS methodology and measurements; formal analysis. R-P.C.: funding acquisition. J.G.: Overall conceptualization; funding acquisition; formal analysis; manuscript writing—reviewing and editing. All authors read and approved the final manuscript.

Funding

This work was supported by NCCR TransCure 51NF40-160620 and the SNFS grant 189220 (JG).

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Animals were bred and handled according to Swiss federal legislation and guidelines of the local animal facility at the University of Bern. The protocols were approved by the Veterinäramt Kanton Bern under the experimental license BE-79/19.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Footnote Group

References

Untitled section

References

1.Centeno PP, Pavet V, Marais R. The journey from melanocytes to melanoma. Nat Rev Cancer. 2023;23(6):372–90.
2.Bastian BC. The molecular pathology of melanoma: an integrated taxonomy of melanocytic neoplasia. Annu Rev Pathol. 2014;9:239–71.
3.Ascierto PA, Kirkwood JM, Grob JJ, Simeone E, Grimaldi AM, Maio M, et al. The role of BRAF V600 mutation in melanoma. J Transl Med. 2012;10(1):1–9.
4.Boutros A, Croce E, Ferrari M, Gili R, Massaro G, Marconcini R, et al. The treatment of advanced melanoma: Current approaches and new challenges. Crit Rev Oncol Hematol. 2023;2024(September 2023):104276. 10.1016/j.critrevonc.2024.104276.
5.Villani A, Potestio L, Fabbrocini G, Troncone G, Malapelle U, Scalvenzi M. The treatment of advanced melanoma: therapeutic update. Int J Mol Sci. 2022;23(12):1–17.
6.Scheau C, Draghici C, Ilie MA, Lupu M, Solomon I, Tampa M, et al. Neuroendocrine factors in melanoma pathogenesis. Cancers (Basel). 2021;13(9):1–41.
7.Ramer R, Wittig F, Hinz B. The endocannabinoid system as a pharmacological target for new cancer therapies. Cancers (Basel). 2021;13(22):5701.
8.Fraguas-Sánchez AI, Martín-Sabroso C, Torres-Suárez AI. Insights into the effects of the endocannabinoid system in cancer: a review. Br J Pharmacol. 2018;175(13):2566–80.
9.Pompeia C, Lima T, Curi R. Arachidonic acid cytotoxicity: Can arachidonic acid be a physiological mediator of cell death? Cell Biochem Funct. 2003;21(2):97–104.
10.Nakanishi M, Rosenberg DW. Roles of cPLA2α and arachidonic acid in cancer. Biochim Biophys Acta - Mol Cell Biol Lipids. 2006;1761(11):1335–43.
11.Horn L, Backlund M, Johnson DH. Targeting the eicosanoid pathway in non-small-cell lung cancer. Expert Opin Ther Targets. 2009;13(6):675–88.
12.Chen X, Sood S, Yang C, Li N, Sun Z. Five-lipoxygenase pathway of arachidonic acid metabolism in carcinogenesis and cancer chemoprevention. Curr Cancer Drug Targets. 2006;6(7):613–22.
13.Nomura DK, Morrison BE, Blankman JL, Long JZ, Lichtman AH, Conti B, et al. Endocannabinoid hydrolysis generates brain prostaglandins that promote neuroinflammation. Science (80- ). 2011;334(November):809–14.
14.Qiu C, Yang L, Wang B, Cui L, Li C, Zhuo Y, et al. The role of 2-arachidonoylglycerol in the regulation of the tumor-immune microenvironment in murine models of pancreatic cancer. Biomed Pharmacother. 2019;115(March):108952. 10.1016/j.biopha.2019.108952.
15.Kienzl M, Hasenoehrl C, Maitz K, Sarsembayeva A, Taschler U, Valadez-Cosmes P, et al. Monoacylglycerol lipase deficiency in the tumor microenvironment slows tumor growth in non-small cell lung cancer. Oncoimmunology. 2021;10(1):1965319. 10.1080/2162402X.2021.1965319.
16.Dinh TP, Kathuria S, Piomelli D. RNA interference suggests a primary role for monoacylglycerol lipase in the degradation of the endocannabinoid 2-arachidonoylglycerol. Mol Pharmacol. 2004;66(5):1260–4.
17.Dinh TP, Freund TF, Piomelli D. A role for monoglyceride lipase in 2-arachidonoylglycerol inactivation. Chem Phys Lipids. 2002;121(1–2):149–58.
18.Saario SM, Savinainen JR, Laitinen JT, Järvinen T, Niemi R. Monoglyceride lipase-like enzymatic activity is responsible for hydrolysis of 2-arachidonoylglycerol in rat cerebellar membranes. Biochem Pharmacol. 2004;67(7):1381–7.
19.Savinainen JR, Saario SM, Laitinen JT. The serine hydrolases MAGL, ABHD6 and ABHD12 as guardians of 2-arachidonoylglycerol signalling through cannabinoid receptors. Acta Physiol. 2012;204(2):267–76.
20.Xie S, Borazjani A, Hatfield MJ, Edwards CC, Potter PM, Ross MK. Inactivation of lipid glyceryl ester metabolism in human THP1 monocytes/macrophages by activated organophosphorus insecticides: Role of carboxylesterases 1 and 2. Chem Res Toxicol. 2010;23(12):1890–904.
21.Streit TM, Borazjani A, Lentz SE, Wierdl M, Potter PM, Gwaltney SR, et al. Evaluation of the “side door” in carboxylesterase-mediated catalysis and inhibition. Biol Chem. 2008;389(2):149–62.
22.Lian J, Nelson R, Lehner R. Carboxylesterases in lipid metabolism: from mouse to human. Protein Cell. 2018;9(2):178–95.
23.Long JZ, Cravatt BF. The metabolic serine hydrolases and their functions in mammalian physiology and disease. Chem Rev. 2011;111(10):6022–63.
24.Jones RD, Taylor AM, Tong EY, Repa JJ. Carboxylesterases are uniquely expressed among tissues and regulated by nuclear hormone receptors in the mouse. Drug Metab Dispos. 2013;41(1):40–9.
25.Satoh T, Taylor P, Bosron WF, Sanghani SP, Hosokawa M, La Du BN. Current progress on esterases: From molecular structure to function. Drug Metab Dispos. 2002;30(5):488–93.
26.Brzezinski MR, Abraham TL, Stone CL, Dean RA, Bosron WF. Purification and characterization of a human liver cocaine carboxylesterase that catalyzes the production of benzoylecgonine and the formation of cocaethylene from alcohol and cocaine. Biochem Pharmacol. 1994;48(9):1747–55.
27.Cashman JR, Perotti BYT, Berkman CE, Lin J. Pharmacokinetics and molecular detoxication. Environ Health Perspect. 1996;104(March):23–40.
28.Dean A, Christian D, Sample RH, Bosron WF. Human liver cocaine esterases: ethanol-mediated formation of ethylcocaine. FASEB J. 1991;5(12):2735–9.
29.Kamendulis LM, Brzezinski MR, Pindel EV, Bosron WF, Dean RA. Metabolism of cocaine and heroin is catalyzed by the same human liver carboxylesterases. J Pharmacol Exp Ther. 1996;279(2):713–7.
30.Ross MK, Streit TM, Herring KL, Xie S. Carboxylesterases: dual roles in lipid and pesticide metabolism. J Pestic Sci. 2010;35(3):257–64.
31.Lewis JP, Horenstein RB, Ryan K, O’Connell JR, Gibson Q, Mitchell BD, et al. The functional G143E variant of carboxylesterase 1 is associated with increased clopidogrel active metabolite levels and greater clopidogrel response. Pharmacogenet Genomics. 2013;23(1):1–8.
32.Tarkiainen EK, Backman JT, Neuvonen M, Neuvonen PJ, Schwab M, Niemi M. Carboxylesterase 1 polymorphism impairs oseltamivir bioactivation in humans. Clin Pharmacol Ther. 2012;92(1):68–71. 10.1038/clpt.2012.13/nature06264.
33.Takai S, Matsuda A, Usami Y, Adachi T, Sugiyama T, Katagiri Y, et al. Hydrolytic profile for ester- or amide-linkage by carboxylesterases pI 5.3 and 4.5 from human liver. Chem Pharm Bull. 1997;20(8):869–73. Available from: http://www.mendeley.com/research/geology-volcanic-history-eruptive-style-yakedake-volcano-group-central-japan/
34.Holmes RS, Wright MW, Laulederkind SJF, Cox LA, Hosokawa M, Imai T, et al. Recommended nomenclature for five mammalian carboxylesterase gene families: Human, mouse, and rat genes and proteins. Mamm Genome. 2010;21(9–10):427–41.
35.Chalhoub G, Kolleritsch S, Maresch LK, Taschler U, Pajed L, Tilp A, et al. Carboxylesterase 2 proteins are efficient diglyceride and monoglyceride lipases possibly implicated in metabolic disease. J Lipid Res. 2021;62:100075. 10.1016/j.jlr.2021.100075.
36.Quiroga AD, Li L, Trötzmüller M, Nelson R, Proctor SD, Köfeler H, et al. Deficiency of carboxylesterase 1/esterase-x results in obesity, hepatic steatosis, and hyperlipidemia. Hepatology. 2012;56(6):2188–98.
37.Bie J, Wang J, Yuan Q, Kakiyama G, Ghosh SS, Ghosh S. Liver-specific transgenic expression of cholesteryl ester hydrolase reduces atherosclerosis in LDLR -/- mice. J Lipid Res. 2014;55(4):729–38. 10.1194/jlr.M046524.
38.Xu J, Xu Y, Xu Y, Yin L, Zhang Y. Global inactivation of carboxylesterase 1 (Ces1/Ces1g) protects against atherosclerosis in Ldlr -/- mice. Sci Rep. 2017;7(1):1–12. 10.1038/s41598-017-18232-x.
39.Li G, Li X, Mahmud I, Ysaguirre J, Fekry B, Wang S, et al. Interfering with lipid metabolism through targeting CES1 sensitizes hepatocellular carcinoma for chemotherapy. JCI Insight. 2023;8(2):1–18.
40.Na K, Kim M, Kim CY, Lim JS, Cho JY, Shin H, et al. Potential Regulatory role of human-carboxylesterase-1 glycosylation in liver cancer cell growth. J Proteome Res. 2020;19(12):4867–83.
41.Na K, Lee EY, Lee HJ, Kim KY, Lee H, Jeong SK, et al. Human plasma carboxylesterase 1, a novel serologic biomarker candidate for hepatocellular carcinoma. Proteomics. 2009;9(16):3989–99.
42.Capece D, D’Andrea D, Begalli F, Goracci L, Tornatore L, Alexander JL, et al. Enhanced triacylglycerol catabolism by carboxylesterase 1 promotes aggressive colorectal carcinoma. J Clin Invest. 2021;131(11):e137845.
43.Dankort D, Curley DP, Cartlidge RA, Nelson B, Karnezis AN, Damsky WE, et al. BrafV600E cooperates with Pten loss to induce metastatic melanoma. Nat Genet. 2009;41(5):544–52.
44.Hooijkaas AI, Gadiot J, Van Der Valk M, Mooi WJ, Blank CU. Targeting BRAF V600E in an inducible murine model of melanoma. Am J Pathol. 2012;181(3):785–94. 10.1016/j.ajpath.2012.06.002.
45.Deken MA, Song JY, Gadiot J, Bins AD, Kroon P, Verbrugge I, et al. Dermal delivery of constructs encoding Cre recombinase to induce skin tumors in PtenLoxP/LoxP, BrafCA/+ mice. Int J Mol Sci. 2016;17(12):1–13.
46.Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: An open-source platform for biological-image analysis. Nat Methods. 2012;9(7):676–82.
47.Türker C, Akal F, Joho D, Panse C, Barkow-Oesterreicher S, Rehrauer H, et al. B-fabric: The Swiss army knife for life sciences. Adv Database Technol - EDBT 2010 - 13th Int Conf Extending Database Technol Proc. 2010;717–20. https://dl.acm.org/doi/10.1145/1739041.1739135.
48.Wolski WE, Nanni P, Grossmann J, d’Errico M, Schlapbach R, Panse C. prolfqua: a comprehensive R-package for proteomics differential expression analysis. J Proteome Res. 2023;22(4):1092–104.
49.Smyth GK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004;3(1):1544–6115.
50.Szafran BN, Lee JH, Borazjani A, Morrison P, Zimmerman G, Andrzejewski KL, et al. Characterization of endocannabinoid-metabolizing enzymes in human peripheral blood mononuclear cells under inflammatory conditions. Molecules. 2018;23(3167):1–13.
51.Crow JA, Bittlesa V, Borazjania A, Potterb PM, Rossa MK. Covalent inhibition of recombinant human carboxylesterase 1 and 2 and monoacylglycerol lipase by the carbamates JZL184 and URB597. Biochem Pharmacol. 2012;84(9):1215–22.
52.Szafran BN, Borazjani A, Scheaffer HL, Crow JA, McBride AM, Adekanye O, et al. Carboxylesterase 1d inactivation augments lung inflammation in mice. ACS Pharmacol Transl Sci. 2022;5(10):919–31.
53.Tirosh I, Izar B, Prakadan SM, Wadsworth MH, Treacy D, Trombetta JJ, et al. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science (80- ). 2016;352(6282):189–96.
54.Jerby-Arnon L, Shah P, Cuoco MS, Rodman C, Su MJ, Melms JC, et al. A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade. Cell. 2018;175(4):984-997.e24.
55.Bhatia P, Friedlander P, Zakaria EA, Kandil E. Impact of BRAF mutation status in the prognosis of cutaneous melanoma: an area of ongoing research. Ann Transl Med. 2015;3(2):1–7.
56.Laezza C, Pagano C, Navarra G, Pastorino O, Proto MC, Fiore D, et al. The endocannabinoid system: a target for cancer treatment. Int J Mol Sci. 2020;21(3):747.
57.Maccarrone M, Di Marzo V, Gertsch J, Grether U, Howlett AC, Hua T, et al. Goods and Bads of the Endocannabinoid System as a Therapeutic Target: Lessons Learned after 30 Years. Pharmacol Rev. 2023;75:885–958.
58.Schmitt A, Schmitz W, Hufnagel A, Schartl M, Meierjohann S. Peroxiredoxin 6 triggers melanoma cell growth by increasing arachidonic acid-dependent lipid signalling. Biochem J. 2015;471(2):267–79.
59.Pacher P, Bátkai S, Kunos G. The endocannabinoid system as an emerging target of pharmacotherapy. Pharmacol Rev. 2006;58(3):389–462.
60.Oesch S, Gertsch J. Cannabinoid receptor ligands as potential anticancer agents – high hopes for new therapies? J Pharm Pharmacol. 2009;61(7):839–53.
61.Velasco G, Sanchez C, Guzman M. Endocannabinoids and Cancer. In: Endocannabinoids Handbook of Experimental Pharmacology. Springer, Cham; 2015. p. 449–72. Available from: https://link.springer.com/chapter/10.1007/978-3-319-20825-1_16#citeas
62.Ueda N, Tsuboi K, Uyama T. Metabolism of endocannabinoids and related N-acylethanolamines: Canonical and alternative pathways. FEBS J. 2013;280(9):1874–94.
63.Leung D, Saghatelian A, Simon GM, Cravatt BF. Inactivation of N-Acyl phosphatidylethanolamine phospholipase D reveals multiple mechanisms for the biosynthesis of endocannabinoids. Biochemistry. 2006;45(15):4720–6.
64.Marquez VE, Blumberg PM. Synthetic diacylglycerols (DAG) and DAG-lactones as activators of protein kinase C (PK-C). Acc Chem Res. 2003;36(6):434–43.
65.Walker J, Sando J. Activation of protein kinase C by short chain phospholipid micelles. Adv Exp Med Biol. 1989;255:29–36.
66.Yang CF, Kazanietz MG. Divergence and complexities in DAG signaling: Looking beyond PKC. Trends Pharmacol Sci. 2003;24(11):602–8.
67.Brose N, Rosemund C. Move over protein kinase C, you’ve got company: alternative cellular effectors of diacylglycerol and phorbol esters. J Cell Sci. 2002;115(23):4399–411.
68.Ebinu JO, Bottorff DA, Chan EYW, Stang SL, Dunn RJ, Stone JC. RasGRP, a Ras guanyl nucleotide- releasing protein with calcium- and diacylglycerol-binding motifs. Science (80- ). 1998;280(5366):1082–6.
69.Oki-Idouchi CE, Lorenzo PS. Transgenic overexpression of RasGRP1 in mouse epidermis results in spontaneous tumors of the skin. Cancer Res. 2007;67(1):276–80.
70.Rambaratsingh RA, Stone JC, Blumberg PM, Lorenzo PS. RasGRP1 represents a novel non-protein kinase C phorbol ester signaling pathway in mouse epidermal keratinocytes. J Biol Chem. 2003;278(52):52792–801. 10.1074/jbc.M308240200.
71.Tuthill MC, Oki CE, Lorenzo PS. Differential effects of bryostatin 1 and 12-O-tetradecanoylphorbol-13-acetate on the regulation and activation of RasGRP1 in mouse epidermal keratinocytes. Mol Cancer Ther. 2006;5(3):602–10.
72.Brooks G, Goss MW, East JA, Hart IR. Growth of melanocytic cells is associated with down-regulation of protein kinase C α, δ, and ∈ isoforms: Possible role of diacylglycerol. J Biol Chem. 1993;268(32):23868–75. 10.1016/S0021-9258(20)80466-9.
73.Mechoulam R, Ben-Shabat S, Hanus L, Ligumsky M, Kaminski NE, Schatz AR, et al. Identification of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. Biochem Pharmacol. 1995;50(1):83–90.
74.Sugiura T, Kondo S, Sukagawa A, Nakane S, Shinoda A, Itoh K, et al. 2-Arachidonoylglycerol: A possible endogenous cannabinoid receptor ligand in brain. Biochem Biophys Res Commun. 1995;2015:89–97.
75.Wu X, Han L, Zhang X, Li L, Jiang C, Qiu Y, et al. Alteration of endocannabinoid system in human gliomas. J Neurochem. 2012;120(5):842–9.
76.Kopp F, Komatsu T, Nomura DK, Trauger SA, Thomas JR, Siuzdak G, et al. The glycerophospho metabolome and its influence on amino acid homeostasis revealed by brain metabolomics of GDE1(-/-) mice. Chem Biol. 2010;17(8):831–40. 10.1016/j.chembiol.2010.06.009.
77.Pagano E, Borrelli F, Orlando P, Romano B, Monti M, Morbidelli L, et al. Pharmacological inhibition of MAGL attenuates experimental colon carcinogenesis. Pharmacol Res. 2017;119:227–36. 10.1016/j.phrs.2017.02.002.
78.Khasabova IA, Chandiramani A, Harding-Rose C, Simone DA. Increasing 2-arachidonoyl glycerol signaling in the periphery attenuates mechanical hyperalgesia in a model of bone cancer pain. Pharmacol Res. 2011;64(1):60–7.
79.Gavrilova-Ruch O, Schönherr R, Heinemann SH. Activation of hEAG1 potassium channels by arachidonic acid. Pflugers Arch Eur J Physiol. 2006;453(6):891–903.
80.He C, Qu X, Wan J, Rong R, Huang L, Cai C, et al. Inhibiting delta-6 desaturase activity suppresses tumor growth in mice. PLoS ONE. 2012;7(10):1–8.
81.Xu C, Gu L, Hu L, Jiang C, Li Q, Sun L, et al. FADS1-arachidonic acid axis enhances arachidonic acid metabolism by altering intestinal microecology in colorectal cancer. Nat Commun. 2023;14(1):2042.
82.Nkondjock A, Shatenstein B, Maisonneuve P, Ghadirian P. Assessment of risk associated with specific fatty acids and colorectal cancer among French-Canadians in Montreal: a case-control study. Int J Epidemiol. 2003;32(2):200–9.
83.Szafran BN, Borazjani A, Seay CN, Carr RL, Lehner R, Kaplan BLF, et al. Effects of chlorpyrifos on serine hydrolase activities, lipid mediators, and immune responses in lungs of neonatal and adult mice. Chem Res Toxicol. 2021;34(6):1556–71.
84.Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, et al. Tissue-based map of the human proteome. Science (80- ). 2015;347(6220):1260419.
85.MacParland SA, Liu JC, Ma XZ, Innes BT, Bartczak AM, Gage BK, et al. Single cell RNA sequencing of human liver reveals distinct intrahepatic macrophage populations. Nat Commun. 2018;9(1):1–21.
86.Andrews TS, Nakib D, Perciani CT, Ma XZ, Liu L, Winter E, et al. Single-cell, single-nucleus, and spatial transcriptomics characterization of the immunological landscape in the healthy and PSC human liver. J Hepatol. 2024;80(5):730–43.
87.Henry GH, Malewska A, Joseph DB, Malladi VS, Lee J, Torrealba J, et al. A cellular anatomy of the normal adult human prostate and prostatic urethra. Cell Rep. 2018;25(12):3530-3542.e5. 10.1016/j.celrep.2018.11.086.
88.Solé-Boldo L, Raddatz G, Schütz S, Mallm JP, Rippe K, Lonsdorf AS, et al. Single-cell transcriptomes of the human skin reveal age-related loss of fibroblast priming. Commun Biol. 2020;3(1):1–12. 10.1038/s42003-020-0922-4.
89.Eraslan G, Drokhlyansky E, Anand S, Fiskin E, Subramanian A, Slyper M, et al. Single-nucleus cross-tissue molecular reference maps toward understanding disease gene function. Science (80- ). 2022;376(6594):eabl4290.
90.Ganier C, Mazin P, Herrera-Oropeza G, Du-Harpur X, Blakeley M, Gabriel J, et al. Multiscale spatial mapping of cell populations across anatomical sites in healthy human skin and basal cell carcinoma. Proc Natl Acad Sci. 2024;121(2):1–10. Available from: https://www.pnas.org/doi/pdf/10.1073/pnas.2313326120
91.The Tabula Sapiens Consortium*. The Tabula Sapiens: a multiple-organ, single-cell transcriptomic atlas of humans. Science (80- ). 2022;376(6594):eabl4896.
92.Reynolds G, Vegh P, Fletcher J, Poyner EFM, Stephenson E, Goh I, et al. Developmental cell programs are co-opted in inflammatory skin disease. Science. 2021;371(6527):eaba6500.
93.Suo C, Dann E, Goh I, Jardine L, Kleshchevnikov V, Park JE, et al. Mapping the developing human immune system across organs. Science (80- ). 2022;376(6597):1–15.
94.Szafran B, Borazjani A, Lee JH, Ross MK, Kaplan BLF. Lipopolysaccharide suppresses carboxylesterase 2g activity and 2-arachidonoylglycerol hydrolysis: a possible mechanism to regulate inflammation. Prostaglandins Other Lipid Mediat. 2015;121:199–206.
95.Scheaffer HL, Borazjani A, Szafran BN, Ross MK. Inactivation of CES1 blocks prostaglandin D2Glyceryl ester catabolism in monocytes/macrophages and enhances its anti-inflammatory effects, whereas the pro-inflammatory effects of prostaglandin E2Glyceryl ester are attenuated. ACS Omega. 2020;5(45):29177–88.

Associated Data

Supplementary Materials

Data Availability Statement

Data is provided within the manuscript or supplementary information files.

Studies

Conditions

Cannabinoids and active compounds

Related research

Abstract

Background

Methods

Results

Conclusions

Graphical Abstract

Supplementary Information

Untitled section

Article notes

Untitled section

Author notes

Introduction

Materials and methods

Substances and reagents

Animal experiments

Melanoma induction

JZL184 treatment in vivo and sample collection

Cell culture

Transfection and selection

2-Oleoylglycerol hydrolysis assay

Mouse tissue sample preparation

Activity-based protein profiling (ABPP)

Western blot analysis

Deglycosylation via PNGase F treatment

Sample preparation for LC–ESI–MS/MS assay

LC–ESI–MS/MS method for amino acid and neurotransmitter quantification

LC–ESI–MS/MS method for endocannabinoids and bioactive lipids quantification

Samples biotinylation

Proteomics. Sample preparation and LC–ESI–MS/MS analysis

Statistical analyses

Results

Differential regulation of bioactive lipids in melanoma

Expression of Ces1 in melanoma and normal skin

Ces1 inhibition impairs melanoma growth in vivo

Covalent inhibition of Ces1 leads to increased 2-AG and 2-OG skin levels in vivo

Distinct expression patterns of CES1 and MAGL in melanoma from RNA-seq data

Discussion

Conclusions

Supplementary Information

Acknowledgements

Abbreviations

Authors’ contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

Footnote Group

References

Untitled section

References

Associated Data

Supplementary Materials

Data Availability Statement

Additional materials