Soluble epoxide hydrolase in the liver orchestrates abdominal aortic aneurysm formation in mice

Abstract

The liver plays an important role in cardiovascular disease by amplifying systemic inflammation, while the underlying mechanisms remain to be defined. Soluble epoxide hydrolase (sEH) is a pro-inflammatory enzyme, and pharmacological inhibition of sEH was shown to protect against various inflammatory diseases. In this study, we have identified a novel role of the liver, through expression of sEH, in the pathogenesis of abdominal aortic aneurysm (AAA). sEH expression and activity were markedly higher in mouse liver compared with aorta and further increased in the context of AAA. Pharmacological inhibition or hepatocyte-specific disruption of sEH prevented AAA formation in two animal models of AAA (angiotensin II infusion and aortic calcium chloride application in male mice), concomitant with reduced expression of complement C3 and serum amyloid A, liver-derived inflammatory factors causally linked to AAA formation. Interestingly, data from co-incubation of liver ex vivo with aorta identified galectin-3 secreted from the aneurysm-prone aorta that activates sEH in the liver. We also determined 12,13-dihydroxyoctadecenoic acid (DiHOME) and various circulating pro-inflammatory cytokines as a downstream mechanism potentially associated with hepatic sEH in the context of AAA. These novel findings provide direct evidence that bidirectional crosstalk between aorta and liver contributes to AAA via hepatic sEH.

Soluble epoxide hydrolase, a pro-inflammatory enzyme predominantly expressed in the liver, serves as a key mediator of liver-aorta crosstalk, amplifying systemic inflammation and promoting the development of abdominal aortic aneurysms.

Introduction

Aortic aneurysms are the 15th leading cause of death overall and the 10th in men older than age 55 in the United States, causing approximately 15,000 deaths per year in the US and more than 175,000 deaths globally¹. Only half of patients with aortic rupture survive to surgery^2,3, so there is a pressing need to develop new approaches that can reduce abdominal aortic aneurysm (AAA) growth and rupture risk. Smoking, age, male sex, hypertension, and hypercholesterolemia are important risk factors for AAA. Abdominal aortic aneurysms exhibit pronounced inflammation throughout the vascular wall, leading to tissue degeneration and progressive weakening of the aorta. Key pathways that orchestrate inflammation in AAA, however, remain to be determined.

Liver plays a vital role in amplifying chronic systemic inflammatory diseases via robust production of acute-phase reactant proteins⁴. Studies demonstrating a role for complement factors and serum amyloid A (SAA), inflammatory factors abundantly secreted by the liver, in the pathogenesis of AAA point to the liver’s involvement in AAA^5–9. The liver also plays a crucial role in atherosclerosis¹⁰, and liver-targeted therapeutics (e.g., statins and PCSK9 inhibitors) are highly effective at preventing and treating atherosclerosis^11,12. While much is known about how the liver contributes to atherosclerosis, the mechanisms and biochemical pathways whereby the liver contributes to AAA remain largely unknown.

Soluble epoxide hydrolase (sEH), encoded by the Ephx2 gene, is a widely distributed enzyme that catalyzes the hydrolysis of epoxy fatty acids (EpFA), such as epoxyeicosatrienoic acids [EETs, metabolized from arachidonic acid (AA)] produced by cytochrome P450 enzymes to their corresponding diols [i.e., dihydroxyeicosatrienoic acids (DHETs)]. These fatty acid-derived epoxides and diols are important lipid signaling molecules that regulate vascular inflammation. Pharmacological inhibition of sEH was shown to protect against hypertension and atherosclerosis in mice^13–15, and limited data also suggested protection against AAA formation¹⁵. However, neither the target cells nor the mechanisms of action of sEH inhibition were defined in AAA. Studies in rodents and humans have determined that sEH, along with cytochrome P450 enzymes, is expressed highly in the liver as compared to other tissues^16,17. Preferential expression of the sEH and the distantly regulated microsomal epoxide hydrolase (mEH) in the liver is consistent with the teleological role of epoxide hydrolase enzymes in xenobiotic metabolism¹⁸. The very different substrate preference for these enzymes, however, suggests that the mEH is largely responsible for metabolism of reactive epoxides on cyclic backbones, with the sEH showing a high preference for EpFA. These findings raise the possibility that sEH expressed in liver, rather than other tissues, could play a vital role in the pathogenesis of AAA. Here, using a pharmacological antagonist of sEH and hepatocyte-specific Ephx2 (which encodes sEH) knockout (KO) mice, and two distinct models of AAA formation, we investigated the role of hepatic sEH in the pathogenesis of AAA. Our data show that either pharmacological sEH inhibition or hepatocyte-specific sEH gene disruption ameliorated the production of liver-derived inflammatory mediators linked to AAA and prevented AAA formation. In addition, we provide novel evidence that 12,13-DiHOME (dihydroxyoctadecenoic acid, a linoleic acid diol) is a major downstream metabolite produced by hepatic sEH in context of AAA.

Results

sEH expression is significantly higher in the liver compared to aorta and further increased by AngII infusion

We first searched the available database (GeneAtlas U133A, gcrma) to check the expression levels of sEH in tissues and organs, and found that sEH is most highly expressed in liver compared to other tissues (Supplementary Fig. S1A). Our data likewise showed that sEH protein levels were > 10-fold higher in mouse liver compared to aorta as detected by Western blot (Fig. 1A). Furthermore, sEH mRNA expression was significantly increased in livers, but not aortas, of AngII-infused mice as compared to saline controls (Fig. 1B). Interestingly, GTEx data showed that AT1 receptor (AngII receptor type 1, Agtr1) expression is also markedly higher in mouse liver than aorta (Supplementary Fig. S1B). sEH protein was also detected in the liver by immunofluorescence, which co-localized with the hepatocyte marker albumin and was increased by AngII infusion (Fig. 1C). These findings are consistent with prior reports comparing cytochrome P450 enzyme and sEH expression in human and rodent tissues ^16,17.

Fig. 1. sEH expression was increased in the liver, but not aorta, of AngII-infused mice in conjunction with upregulated liver inflammatory mediators, which were blunted by sEH inhibitor.

AngII was infused for 3 weeks via osmotic minipump in LDLR KO mice with or without TPPU treatment. A Comparison of sEH protein expression between liver and aorta as determined by Western blot. n = 4/group. Student’s t-test. B sEH mRNA expression in liver and aorta of saline controls (n = 3) and AngII-infused mice (n = 5/liver, n = 4/aorta). One-way ANOVA. C Immunofluorescence analysis of sEH in liver tissues of AngII-infused mice. Albumin was used as a marker for hepatocytes. Scale bar = 20 µm. D Depiction of the Sp1 promoter binding site in GC-rich region between nucleotides -374 and +28 with respect to the putative sEH gene transcription start site. mRNA expression of Sp1 (E, n = 7/group, Student’s t-test), sEH (F, n = 4/group, Student’s t-test), and Jarid1b as determined by qPCR (G, n = 4/group, Student’s t-test) and quantification of H3K4me3 levels detected by Western blot (H, n = 3/group, Student’s t-test). I Reduced sEH enzymatic activity in liver and aorta by TPPU. n = 3/aorta. n = 4/saline-treated aorta, n = 6/AngII-treated aorta, n = 4/AngII + TPPU (3 mg/kg)-treated aorta, n = 4/AngII + TPPU (10 mg/kg)-treated aorta. Two-way ANOVA. (J-L) mRNA expression of C3 (J, n = 6/saline, n = 9/AngII, n = 6/AngII + TPPU (3 mg/kg), n = 5/AngII + TPPU (10 mg/kg). One-way ANOVA), IL-6 (K, n = 6/group. One-way ANOVA) and SAA (L, n = 6/saline, n = 6/AngII, n = 3/AngII + TPPU (3 mg/kg), n = 6/AngII + TPPU (10 mg/kg). One-way ANOVA) in liver tissues. mRNA expression of ATF (M, n = 6/saline, n = 6/AngII, n = 3/AngII + TPPU (3 mg/kg), n = 6/AngII + TPPU (10 mg/kg). One-way ANOVA), BiP (N, n = 6/saline, n = 6/AngII, n = 3/AngII + TPPU (3 mg/kg), n = 6/AngII + TPPU (10 mg/kg). One-way ANOVA) and CHOP (O, n = 6/saline, n = 6/AngII, n = 3/AngII + TPPU (3 mg/kg), n = 6/AngII + TPPU (10 mg/kg). One-way ANOVA) in liver tissues. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

We subsequently investigated the molecular mechanisms of hepatic sEH gene induction during AngII-induced AAA. Sp1 was previously reported as a major transcription factor for sEH gene expression in various cell lines, including human liver cancer-derived HepG2 cells¹⁹. The minimal promoter was identified as a GC-rich region between nucleotides −374 and +28 with respect to the putative transcription start site (Fig. 1D). Sp1 expression was increased in livers of AngII-infused mice compared to saline control (Fig. 1E), in conjunction with elevated sEH expression (Fig. 1F). Histone H3K4 demethylase Jarid1b was also reported to positively regulate sEH expression in response to AngII²⁰, and our data confirmed upregulated Jarid1b expression (Fig. 1G), along with a trend toward reduced H3K4me3 (indicator of Jarid1b activity) (Fig. 1H), in livers of mice infused with AngII. These findings suggest that Sp1 and/or Jarid1b may mediate induction of hepatic sEH in the context of AngII-induced AAA formation.

Smoking, male sex and aging are major risk factors for AAA, and hepatic sEH expression was reported to be positively correlated with age and male sex^21,22. Moreover, smoke exposure has been shown to stimulate pulmonary sEH activity and inflammation in mice²³. To investigate whether smoke exposure may be associated with induction of sEH and inflammatory gene expression in liver tissue, we treated mouse liver slices with 1% cigarette smoke extract (CSE) for 6 hr. Interestingly, CSE induced hepatic sEH, and inflammatory mediators SAA and C3 (both of which have been linked to AAA)^5–9, in mouse liver (Supplementary Fig. S2).

sEH inhibitor blunted expression of inflammatory mediators, ER stress genes, and macrophage infiltration in the liver of AngII-infused mice

Next, we quantified sEH activity in the liver and aorta in the murine AngII-induced AAA model. Infusion of AngII into low density lipoprotein receptor (LDLR) KO mice to induce AAA resulted in increased sEH activity in the liver, which was reduced by concurrent oral treatment with a highly selective and potent sEH inhibitor, TPPU [1-(1-Propionylpiperidin-4-yl)-3-(4-(trifluoromethoxy)phenyl)urea, 3–10 mg/kg body weight in drinking water]²⁴. In contrast, aortic sEH activity was very low in the presence or absence of TPPU (Fig. 1I). To investigate the impact of sEH inhibition on inflammatory mediator production in the liver, we quantified the expression of complement factors and SAA, previously reported to contribute to AAA formation^5,6,9, in the livers of mice infused with AngII. Interestingly, expression of C3 was significantly increased by AngII infusion and reduced by TPPU treatment (Fig. 1J). IL-6 expression, reportedly associated with liver inflammation²⁵, was also markedly increased in the livers of AngII-infused mice and reduced by TPPU treatment (Fig. 1K). Furthermore, TPPU reduced SAA mRNA expression in liver (Fig. 1L), and SAA protein content in liver, plasma and aorta (Supplementary Fig. S3). Given that SAA mRNA expression was very low in aorta compared to liver (Supplementary Fig. S4), SAA protein that accumulated in the aorta was likely derived from circulating SAA secreted from the liver. There was no change in hepatic expression of fibrinogen α, β or γ, which have also been reported to be associated with AAA formation⁷, in response to AngII infusion or TPPU treatment (Supplementary Fig. S5).

Endoplasmic reticulum (ER) stress plays an important role in liver inflammation and stress signaling²⁶. Expression of hepatic ER stress gene markers, ATF6, BiP, and CHOP, was significantly increased in the livers of AngII-infused mice and blunted by TPPU treatment (Fig. 1M–O). Taken together, these results support the hypothesis that liver inflammatory responses temporally correlate with AAA formation in mice and are mitigated by sEH inhibition.

Pharmacological inhibition of sEH protected against AngII-induced AAA formation

To investigate whether sEH inhibition could mitigate AAA formation, we administered TPPU orally to LDLR KO mice, followed by AngII infusion. TPPU treatment (3 mg or 10 mg/kg body weight) significantly reduced AAA incidence (Fig. 2A) and aortic diameter (Fig. 2B and Supplementary Fig. S6) without affecting systolic blood pressure or cholesterol/triglycerides levels (Supplementary Fig. S7). Representative in situ abdominal aortic images of each group are shown in Fig. 2C. Prominent elastin fragmentation (H&E and VVG staining, Fig. 2D) and macrophage infiltration (Mac-3 staining, Supplementary Fig. S8A) were observed in aortas of AngII-infused mice, which were markedly inhibited by TPPU treatment. Furthermore, TPPU markedly decreased expression of matrix metalloproteinase (MMP)-2/-9 in the aorta (Supplementary Fig. S8A). We observed aortic rupture in a subset of mice (approximately 20% of AngII-infused animals), which were removed from the analysis. There was no rupture in TPPU treatment groups (Supplementary Fig. S9A).

Fig. 2. Pharmacological inhibition or hepatocyte-specific sEH gene disruption protected against AngII-induced AAA formation.

A–D AngII was infused for 3 weeks via osmotic minipump in LDLR KO mice with or without TPPU treatment. A AAA incidence. B Aortic diameter. n = 4/saline, n = 8/AngII, n = 5/AngII + TPPU (3 mg/kg), n = 6/AngII + TPPU (10 mg/kg). One-way ANOVA. C Representative in situ images of aorta in AngII-infused mice with or without TPPU treatment; dashed lines delineate aorta. Scale bar = 2 mm. D Representative histology images of H&E and VVG (elastin degradation, scale bar = 50 µm). E–J AngII was infused for 3 weeks via osmotic minipump in littermate WT (sEH^WT) or hepatocyte-specific Ephx2 KO mice (sEH^ΔHepatocyte). E AAA incidence. F Aortic diameter. n = 4/saline, n = 7/sEH^WT, n = 7/sEH^ΔHepatocyte. One-way ANOVA. G SAA mRNA expression. n = 3/group. One-way ANOVA. H C3 mRNA expression. n = 3/group. One-way ANOVA. I Representative histology images of H&E and VVG (elastin degradation). J Immunohistochemical analysis for SAA and C3 protein accumulation in aortic wall of AngII-infused mice. Scale bar = 20 µm. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.

Hepatocyte-specific sEH gene disruption protected against AngII-induced AAA formation in conjunction with reduced expression of liver-derived inflammatory mediators

Although TPPU is a highly selective and potent inhibitor of sEH, we cannot exclude the possibility of off-target effects, and it does not selectively target sEH in the liver. Thus, we created hepatocyte-specific sEH deficient mice (sEH^ΔHepatocyte in low density lipoprotein receptor (LDLR) knockout (KO) background) by crossing Ephx2^flox/flox and alb-Cre mice. The presence of WT/flox Ephx2 allele and/or alb-Cre allele in mice was validated via PCR-based genotyping (Supplementary Fig. S10A). We confirmed that Ephx2 gene expression is diminished in liver, but not aorta, in Ephx2^flox/flox/alb-Cre mice (Supplementary Fig. S10B). Notably, hepatocyte-specific sEH disruption significantly reduced AAA incidence (Fig. 2E) and aortic diameter (Fig. 2F) in AngII-infused mice, without affecting blood pressure (Supplementary Fig. S11), in conjunction with reduced SAA and C3 expression in liver (Fig. 2G, H). sEH^ΔHepatocyte mice also exhibited reduced aortic elastin degradation (Fig. 2I), macrophage infiltration, and MMP-2/-9 expression (Supplementary Fig. S8B). Furthermore, SAA and C3 protein accumulation in the aortic wall was ameliorated by hepatocyte-specific sEH disruption as evaluated by immunohistochemical analysis (Fig. 2J) and Western blot (Supplementary Fig. S12). We observed aortic rupture in a subset of mice (approximately 10% of AngII-infused animals), which were removed from the analysis. No rupture was observed in sEH^ΔHepatocyte mice (Supplementary Fig. S9B).

CaCl₂-induced AAA model: sEH expression and effects of TPPU

Next, we investigated whether hepatic sEH expression is also induced by aortic CaCl₂ application, which is a local acute aortic injury model with a well-defined time course²⁷. Consistent with data in the AngII infusion model, sEH protein expression was markedly increased in the livers of CaCl₂-treated mice compared to control (Fig. 3A). In addition, there was a trend toward increased expression of SAA in liver, plasma, and aorta, and reduction by TPPU treatment, following aortic CaCl₂ application (Supplementary Fig. S13). Notably, TPPU treatment likewise protected against AAA formation induced by CaCl₂ application (Fig. 3B, C), in conjunction with decreased elastin degradation (Fig. 3D), macrophage infiltration, and MMP-2/-9 expression (Supplementary Fig. S14A).

Fig. 3. sEH expression was increased in the liver by CaCl₂ application to the aorta, and pharmacological inhibition or hepatocyte-specifc sEH gene disruption protected against AAA formation induced by CaCl₂.

A–D CaCl₂ or saline (sham) was applied to the infrarenal aorta of C57Bl/6 mice with or without TPPU treatment. A sEH protein expression in liver as measured by Western blot. n = 5/group. Student’s t-test. B Aortic diameter. n = 4/sham, n = 8/CaCl₂, n = 7/CaCl₂ + TPPU (3 mg/kg), n = 7/CaCl₂ + TPPU (10 mg/kg). One-way ANOVA. C Representative images of aorta in mice with or without oral TPPU treatment. Scale bar = 1 mm. D Representative histology; H&E and VVG (elastin degradation, scale bar = 50 µm). E–J CaCl₂ or saline (sham) was applied to the infrarenal aorta of sEH^WT or sEH^ΔHepatocyte. E Aortic diameter. n = 4/saline, n = 4/sEH^WT, n = 3/sEH^ΔHepatocyte. One-way ANOVA. F Representative aortic images of aorta in sEH^WT and sEH^ΔHepatocyte mice. Scale bar = 1 mm. G SAA mRNA expression. n = 3/group. One-way ANOVA. H C3 mRNA expression. n = 3/group. One-way ANOVA. I Representative images of H&E and VVG (elastin degradation). Black scale bar = 100 µm, white scale bar = 20 µm. F Immunohistochemical analysis of SAA and C3 protein accumulation in the aortic wall. Scale bar = 50 µm. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001.

Hepatocyte-specific sEH gene disruption protected against CaCl₂-induced AAA formation in conjunction with reduced expression of liver inflammatory mediators

Next, we employed sEH^ΔHepatocyte mice in the CaCl₂ application model. Hepatocyte-specific disruption of sEH strongly protected against CaCl₂-induced AAA formation (Fig. 3E, F), in association with reduced hepatic expression of SAA and C3 (Fig. 3G, H). Hepatocyte-specific sEH disruption also ameliorated major pathological features of AAA, such as elastin degradation (Fig. 3I), macrophage infiltration, MMP-2/-9 expression (Supplementary Fig. S14B). SAA and C3 protein accumulation in the aorta was also significantly reduced in or sEH^ΔHepatocyte mice compared to WT mice (Fig. 3J).

Co-incubation of CaCl₂-treated aorta with liver augmented hepatic sEH gene transcription and inflammatory gene expression

Our data in the CaCl₂ model suggested that local aortic injury might trigger secretion of molecules that can induce hepatic sEH expression and inflammatory mediator production. To test this hypothesis, we co-incubated sham- or CaCl₂-treated, aneurysm-prone aorta (harvested 7 days after laparotomy, prior to AAA formation)²⁸ with liver harvested from control mice for 12 h (Fig. 4A). Compared to sham aorta, CaCl₂-treated aorta significantly increased sEH gene expression and activity in the liver (Fig. 4B, C), in conjunction with increased C3 and SAA expression (Fig. 4D, E). We identified Sp1 and Jarid1b as potential upstream regulators of sEH induction in AngII-infused mice (Fig. 1D–H)^19,20. Likewise, we observed that compared with sham aorta, co-incubation with CaCl₂-treated aorta significantly upregulated the expression of Sp1 and Jarid1b in liver tissue (Fig. 4F, G).

Fig. 4. Co-incubation of aneurysm-prone aorta with liver augmented hepatic sEH gene transcription and inflammatory gene expression via aorta-derived transferable factors.

A–G Fresh liver slices from control mice were co-incubated for 12 h with aorta harvested from sham- or CaCl₂-treated mice 7 days after laparotomy. A Experimental scheme. B sEH mRNA expression. n = 3/group. Student’s t-test. C sEH enzymatic activity. n = 3/group. Student’s t-test. D mRNA expression of C3. n = 3/group. Student’s t-test., (E) SAA. n = 3/group. Student’s t-test. F Sp1. n = 4/group. Student’s t-test. G Jarid1b. n = 4/group. Student’s t-test. H–K Whole aortas from sham- or CaCl₂-treated mice harvested 7 days after laparotomy were incubated for 12 h, and conditioned media were then collected for proteomic analysis. H Sample preparation scheme. I Volcano plot of all differentially-expressed proteins between sham and CaCl₂ treatment. Top 20 most upregulated J and downregulated (K) proteins as predicted by gene ontology (GO) enriched analysis. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01.

Profiling of candidate molecules secreted from the aneurysm-prone aorta

To begin to identify signaling mediators released from aneurysm-prone aorta that might induce hepatic sEH, we performed LC-MS/MS-based unbiased proteomic analysis of incubation medium collected from sham- or CaCl₂-treated aorta (Fig. 4H, I). We identified a number of upregulated proteins [e.g., plastin-2, kininogen-1, prosaposin, periostin, calreticulin, galectin-3 (gal-3), cathepsin S and xanthine dehydrogenase/oxidase] (Fig. 4J) and downregulated proteins (e.g., fatty acid synthase, glycerol-3-phosphate dehydrogenase, transketolase, troponin C, glutamine synthase and citrate synthase) (Fig. 4K) in incubation media from CaCl₂-treated aorta compared to sham. Interestingly, GO enriched analysis predicted the upregulation in cell spreading, exocytosis, and apoptotic signaling pathways (Supplementary Fig. S15), and downregulation in NAD⁺/nicotinamide metabolism and glycolytic process pathways (Supplementary Fig. S16) in CaCl₂-treated compared to control aorta.

Gal-3 as a candidate mediator of hepatic sEH induction

Damaged tissues release various damage-associated molecule patterns (DAMPs) such as high-mobility group box 1 (HMGB-1), osteopontin (Spp1), and gal-3, which have been causally linked to AAA. Interestingly, proteomics data showed that gal-3 was highly upregulated in the incubation medium collected from CaCl₂-treated aorta (Figs. 4J and 5A) while neither HMGB-1 nor osteopontin was found to be differentially expressed. Gal-3 protein was markedly increased in CaCl₂-treated aorta (Fig. 5B), and the immunostaining pattern suggests that gal-3 is likely expressed in inflammatory cells and vascular smooth muscle cells (VSMC), as has been reported in the elastase-induced AAA model. Furthermore, gal-3 expression was highly increased in plasma collected from mice with CaCl₂-induced AAA, compared to sham (Fig. 5C, D).

Fig. 5. Identification of gal-3 as a candidate mediator of hepatic sEH induction.

A Quantification of gal-3 protein levels (LC/MS) in conditioned media from whole aortas (sham vs CaCl₂, n = 3). Gal-3 protein expression in aortas (B, immunostaining) and plasma (C: ELISA, D: Western blot). mRNA expression of sEH (E), Sp1 (F) and Jarid1b (G) in liver slices treated with or without recombinant gal-3 protein (10 µg in 500 µl DMEM medium for 24 h, n = 3). mRNA expression of sEH (H), C3 (I), and Jarid1b (J) in liver slices (sEH^WT vs sEH^ΔHepatocyte) treated with or without recombinant gal-3 protein (n = 3).

Next, we treated liver ex vivo with recombinant gal-3 protein (10 µg in 500 µl DMEM medium for 24 h). Gal-3 strongly induced hepatic sEH (Fig. 5E, H), Sp1 (Fig. 5F), and Jarid1b (Fig. 5G). Moreover, C3 (Fig. 5I) and SAA (Fig. 5J) levels were also significantly increased by gal-3 treatment. Interestingly, the effects of gal-3 on the liver were abolished in hepatocyte-specific sEH KO mice (Fig. 5H–G). These results suggest that gal-3 produced by the injured aorta could be a potential candidate inducer of sEH expression in the liver.

12,13-DiHOME, a linoleic acid diol produced by sEH, is associated with AAA formation

Epoxides of arachidonic (AA) and linoleic acid (LA) are the primary substrates of sEH, which converts them to their corresponding diols. To begin to address which sEH metabolites may contribute to AAA formation, we analyzed selected lipid metabolites produced by sEH in liver and plasma of mice with AngII-induced AAA. Interestingly, serum levels of 12,13-DiHOME, a LA diol, were significantly increased in plasma and in liver-conditioned medium from these mice, and were reduced by the sEH inhibitor, TPPU, as measured by ELISA (Fig. 6A, B). Moreover, in patients with AAA vs. age-, sex- and race-matched controls, 12,13-DiHOME levels in blood were increased in AAA patients (Fig. 6D). In contrast, levels of AA-derived 14,15-DHET were comparatively low, not different between AAA and controls in mice or humans, and minimally affected by TPPU (Fig. 6C, E).

Fig. 6. 12,13-DiHOME, a linoleic acid diol produced by sEH, is associated with AAA formation.

Levels of 12,13-DiHOME (A, n = 6/saline, n = 7/AngII, n = 8/AngII + TPPU. One-way ANOVA) and 14,15-DHET (C, n = 3/group. One-way ANOVA) in plasma. Levels of 12,13-DiHOME in liver-conditioned medium (B, n = 3/saline, n = 3/AngII, n = 5/AngII + TPPU. One-way ANOVA). Levels of circulating 12,13-DiHOME (D, n = 8/control, n = 7/AAA. Student’s t-test), but not 14,15-DHET (E, n = 5/group. Student’s t-test), are increased in patients with AAA. Quantification of EpOMES and DiHOMEs in liver and plasma of liver-specific sEH KO mice (CaCl₂ model). 12,13-DiHOME in liver (F, n = 3/group, Student’s t-test), 12,13-DiHOME in plasma (G, n = 3/group, Student’s t-test), 12,13-EpOME in liver (H, n = 3/group, Student’s t-test), 12,13-EpOME in plasma (I, n = 3/group, Student’s t-test), 9,10-DiHOME in liver (J, n = 3/group, Student’s t-test), 9,10-DiHOME in plasma (K, n = 3/group, Student’s t-test), 9,10-EpOME in liver (L, n = 3/group, Student’s t-test), 9,10-EpOME in plasma (M, n = 3/group, Student’s t-test). Ex vivo treatment with 12,13-DiHOME, but not 14,15-DHET, increased expression of inflammatory mediators (C3 and SAA) in liver (N, O, n = 5/control, n = 5/12.13-DiHOME, n = 3/9,10-DiHOME, n = 3/14,15-DHET. One-way ANOVA) and conditioned medium (P, Q, n = 3/group. One-way ANOVA). Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

9,10-DiHOME and 12,13-DiHOME are two main metabolites converted from 9,10-EpOME and 12,13-EpOME, respectively, by sEH. To confirm that 12,13-DiHOME is responsible for inflammatory mediator production in the liver, we next quantified levels of EpOMEs and DiHOMEs in liver and plasma of mice with CaCl₂-induced AAA mice by LC-MS/MS. Notably, significant reductions in 12,13-DiHOME were detected in both liver homogenates (Fig. 6F) and plasma (Fig. 6G) from the hepatocyte-specific sEH KO mice, with reciprocal increases in 12,13-EpOME (Fig. 6H, I). In contrast, levels of 9,10-DiHOME (Fig. 6J, K) and 9,10-EpOME (Fig. 6L, M) were much lower in liver and plasma and not consistently affected by hepatocyte-specific sEH gene disruption.

To test whether 12,13-DiHOME increases C3 and SAA production in the liver, we treated liver slices with 12,13-DiHOME, 9,10-DiHOME, or 14,15-DHET (100 nM) ex vivo for 12 h. Notably, only 12,13-DiHOME strongly induced C3 and SAA gene expression (Fig. 6M, O) and protein release (Fig. 6P, Q) from the liver, while 9,10-DiHOME and 14,15-DHET, respectively, had minimal or no effects, suggesting a possible role for LA-derived, sEH metabolite 12,13-DiHOME in promoting inflammatory mediator production in the liver in the context of AAA.

sEH inhibition reduces levels of circulating pro-inflammatory cytokines

To investigate whether sEH inhibition can reduce systemic inflammation during AAA formation, we profiled pro-inflammatory cytokine expression in plasma of mice with AngII-induced AAA, in the absence or presence of TPPU treatment. There was a strong trend towards increased pro-inflammatory cytokines, particularly IL-1β (Fig. 7A) and MCP-1 (Fig. 7F), in plasma of mice with AngII-induced AAA, which tended to be reduced by TPPU treatment (Fig. 7A–J). These results raise the possibility that sEH in the liver can regulate systemic inflammatory responses, which in turn may result in protection against AAA.

Fig. 7. sEH inhibition reduces levels of circulating pro-inflammatory cytokines.

A–J Plasma levels of cytokines were analyzed in mice with (1) saline infusion, (2) AngII infusion, and (3) AngII infusion + TPPU (10 mg/kg) treatment groups. A IL-1β (n = 3/saline, n = 4/AngII, n = 4/AngII + TPPU, One-way ANOVA), (B) IL-12p70 (n = 3/saline, n = 5/AngII, n = 4/AngII + TPPU, One-way ANOVA), (C) IL-23 (n = 3/saline, n = 5/AngII, n = 5/AngII + TPPU, One-way ANOVA), (D) INFγ (n = 3/saline, n = 5/AngII, n = 5/AngII + TPPU, One-way ANOVA), (E) TNFα (n = 3/saline, n = 5/AngII, n = 6/AngII + TPPU, One-way ANOVA), (F) MCP-1 (n = 3/saline, n = 4/AngII, n = 3/AngII + TPPU, One-way ANOVA), (G) IL-27 (n = 3/saline, n = 5/AngII, n = 4/AngII + TPPU, One-way ANOVA), (H) IL-17A (n = 3/saline, n = 3/AngII, n = 6/AngII + TPPU, One-way ANOVA), (I) IFNβ (n = 3/saline, n = 4/AngII, n = 4/AngII + TPPU, One-way ANOVA), (J) GM-CSF (n = 3/saline, n = 4/AngII, n = 4/AngII + TPPU, One-way ANOVA). Data are represented as mean ± SEM. *p < 0.05, **p < 0.01.

Discussion

The liver is widely accepted to play a crucial role in atherosclerosis, which, like AAA is a vascular inflammatory disease¹⁰. Moreover, liver-targeted therapeutics (e.g., statins and PCSK9 inhibitors) are highly effective at preventing and treating atherosclerosis^11,12. Here, using a pharmacological inhibitor and hepatocyte-specific gene disruption approach in two distinct animal models of AAA, we tested the hypothesis that the specific activity of sEH in the liver plays a crucial role in AAA. Major findings of this study are: (1) sEH expression and activity in the liver, but not aorta, is upregulated in the context of AAA formation, potentially through induction of the transcription factor Sp1 and/or the histone demethylase Jarid1b; (2) treatment with a pharmacological sEH inhibitor prevented AAA formation in conjunction with diminished expression of liver-derived inflammatory mediators, (3) accumulation of liver-derived pro-inflammatory mediators in the aorta is prevented by pharmacological sEH inhibition or hepatocyte-specific sEH gene disruption, (4) hepatocyte-specific sEH KO mice are protected against AAA formation; (5) aneurysm-prone aorta releases transferrable factors that trigger induction of sEH in the liver; and (6) levels of the sEH-derived lipid diol 12,13-DiHOME are increased in mice and humans with AAA and may contribute to induction of inflammatory gene expression in the liver. These findings provide novel insight into the liver’s role in orchestrating AAA via sEH and may have important implications with regard to developing effective medical therapy for this disorder.

sEH is widely expressed in various cell types¹⁶, and pharmacological inhibition of sEH has been shown to improve vascular function and diminish inflammation. sEH inhibitors were also reported to reduce blood pressure in rodent models of hypertension, including AngII infusion^29–32, though we were unable to reproduce that finding in our LDLR KO mice for reasons which are unclear. Effects of sEH inhibitors on the vasculature are generally presumed to be due to local inhibition of vascular EET metabolism, as EETs can be produced and metabolized by blood vessels. However, sEH expression is by far highest in the liver compared to other tissues and organs^16,17, and hepatic sEH expression is upregulated with aging, smoking, and male sex^19–21, all of which are important risk factors for AAA. The liver also has high specific activities of the cytochrome P450 enzymes, which, among other functions, generate the inflammation-resolving epoxy fatty acids. The liver is a very large tissue and heavily vascularized by the sinusoid system. Thus, it should dominate both the production of epoxy fatty acids and their hydrolysis to diols at the whole-body level. A dominant role of the liver in producing epoxy fatty acids and their hydrolysis products was recently reported in the context of therapy for traumatic brain injury³³ and Alzheimer’s disease³⁴. Nevertheless, we cannot exclude a role for sEH expression in aortic or inflammatory cells in AAA pathogenesis, as we did not specifically knock out sEH in these cell types.

The liver can amplify systemic inflammation via a number of biochemical pathways. Complement proteins and SAA, synthesized by the liver and circulating in the blood, are well-defined acute phase proteins associated with systemic inflammation. In both the AngII-infusion and CaCl₂ models of AAA formation, C3 and SAA gene expression was significantly upregulated in the liver and abrogated by hepatocyte-specific sEH disruption. Furthermore, C3 and SAA protein accumulation in the aortic wall was markedly attenuated by hepatocyte-specific sEH gene disruption, suggesting that sEH in the liver dominates the production of key inflammatory mediators that circulate from the liver and contribute to the pathogenesis of AAA. However, the complete repertoire of mediators regulated by sEH in the liver, and how these mediators (individually or collectively) promote AAA, remains to be defined.

We employed two complementary murine models of AAA in this study. The AngII-infused model represents a dynamic and heterogeneous process characterized by rapid macrophage infiltration and thrombus formation, followed by medial rupture, immune-mediated remodeling, and eventual extracellular matrix accumulation and wall thinning³⁵. Recognition of these spatial and temporal variations is critical for accurate histopathological interpretation. In contrast, the CaCl₂-induced model involves topical application of CaCl₂ to the infrarenal aorta, leading to localized vascular injury, inflammation, and calcification that collectively promote aortic dilation. However, this model lacks key pathological features such as frequent rupture or atherothrombosis observed in human AAAs. Our findings demonstrate that pharmacological inhibition of sEH or hepatocyte-specific sEH KO effectively prevented AAA development in both models. Notably, DAMPs released from the injured aortic tissue in the CaCl₂ model were identified as putative mediators of hepatic sEH induction, highlighting a bidirectional pathological interaction between the aorta and the liver. DAMPs such as gal-3^36,37, HMGB-1³⁸, and Spp1³⁹ have been reported to contribute to AAA formation. Gal-3, HMGB-1 and Spp1 were also reported to be associated with liver dysfunction^29–31. Gal-3 is highly expressed in various cell types pertinent to AAA, including VSMC, endothelial cells (EC), and inflammatory cells, but not in hepatocytes [GTEx (Gencode ID: ENSG00000131981.15) and GeneAtlas dataset (U133A)], suggesting that gal-3 could participate in the limb of the axis connecting the aorta to sEH induction in the liver. Importantly, levels of gal-3 were reported to be elevated in AAA tissue and serum of humans with AAA³². In this study, we identified gal-3 produced and released by CaCl₂-treated aorta could be a potential mediator for hepatic sEH induction in CaCl₂-induced AAA model. Further studies are required to determine the precise mechanisms of aorta-derived gal-3 and sEH induction in the liver.

sEH functions to hydrolyze lipid peroxide species, with epoxides of AA and LA being its primary substrates. While AA epoxides and diols (e.g., EETs and DHETs) have been widely studied in cardiovascular pathophysiology⁴⁰, LA metabolites (e.g., EpOMES and DiHOMEs) have received far less attention, despite the fact that LA is the most abundant fatty acid in Western diets, and increased levels of LA diols are associated with various inflammatory diseases⁴¹. A previous study also showed that both diglyceride and triglyceride species enriched in LA were specifically associated with AAA⁴², suggesting a link between LA consumption and AAA in patients. Interestingly, we found that serum levels of 12,13-DiHOME (a LA diol), but not 14,15-DHET (an AA diol), are elevated in mice and humans with AAA (Fig. 5A, D). Unfortunately, we were not able to obtain liver samples from humans with AAA to confirm induction of sEH or increased levels of 12,13-DiHOME. Nevertheless, our observations are in line with the above-mentioned study in humans showing an association between LA-containing lipid species and AAA⁴². Notably, 12,13-DiHOME has been reported to promote inflammation in prior studies^43,44 and was identified as a novel biomarker for left atrial remodeling in patients with atrial fibrillation⁴⁵. These results suggest that sEH-derived 12,13-DiHOME may play a broad role in promoting inflammation in cardiovascular disease

To assess whether variants in the sEH gene are associated with cardiovascular disease, we examined genome-wide association studies (GWAS) for all annotated and unannotated SNPs within and around the EPHX2 gene locus. We found that rs751141 in the EPHX2 gene locus is associated with familial hypercholesterolemia (Clinvar), recurrence of atrial fibrillation after catheter ablation⁴⁶, and increased risk for ischemic stroke and cerebral infarction in Chinese populations^47,48. Furthermore, genetic variation in EPHX2 was reported to be associated with risk of incident of coronary heart disease⁴⁹. Notably, there was no significant difference in EPHX2 expression between control aorta and AAA tissue (GSE183464, Supplementary Fig. S17). Likewise, aortic smooth muscle cells isolated from humans with AAA exhibited no difference in EPHX2 expression compared to controls (GSE237229, Supplementary Fig. S18). These findings are consistent with our data showing that Ephx2 expressed outside of the aorta plays a role in AAA formation.

One limitation of this study is the lack of analysis of human liver tissues, which were unavailable from AAA patients. Consequently, further investigation is required to elucidate the mechanisms linking hepatic sEH to AAA in humans. In addition, the substantial heterogeneity among AAA patients, driven by genetic and environmental factors such as sex, race, diet, smoking status, and comorbidities, poses a major challenge to therapeutic development⁵⁰. The extent to which these environmental factors correspond to those modeled in experimental AAA remains unclear. Nonetheless, several pharmacological sEH inhibitors are currently under clinical investigation for other human diseases. Three sEH transition state inhibitors, similar to TPPU, including AR9281⁵¹, GSK2256294⁵², and EC5026⁵³ have been used in several clinical trials. In particular, EC5026 has completed human phase 1a and 1b clinical safety trials with no adverse effects and is entering an early efficacy trial with FDA approval in 2024. Our findings suggest that sEH inhibitors currently being evaluated in several clinical trials potentially could be repurposed to treat AAA. Further investigations in humans are required to use the pharmacological sEH inhibitors for AAA treatment.

Several substrates involved in our proposed mechanism, including arachidonic acid-derived EETs and linoleic acid–derived EpOMEs, are indeed cytochrome P450 (CYP450)-sensitive. CYP450 enzymes are abundantly expressed in the liver, and commonly prescribed medications such as statins, beta-blockers, and certain antihypertensive agents can modulate CYP450 activity in humans. Such modulation may alter the formation of EETs and EpOMEs, as well as their downstream metabolites (DHETs and DiHOMEs), potentially influencing inflammatory responses implicated in AAA pathogenesis. These interactions should be carefully considered when translating our experimental findings to clinical settings in the future.

In conclusion, using two distinct animal models, we provide evidence that sEH expressed in the liver orchestrates AAA formation. We also provide evidence of an aorta-liver axis, wherein the aneurysm-prone aorta secretes factors, including gal-3, that might induce sEH in the liver, which in turn secretes inflammatory mediators that promote AAA. Pharmacological inhibitors of sEH, stable mimics of epoxy fatty acids, or gene therapy targeting sEH in the liver, may represent attractive strategies to prevent or treat AAA.

Methods

Ethical statement

The work was conducted in accordance with ARRIVE guidelines. Animal experimental protocols were approved by the Institutional Animal Care and Use Committee at the Medical College of Georgia at Augusta University (protocol approval #: 2013-0528) and complied with National Institute of Health guidelines. Collection of the human blood samples and related experiments were designated as exempted from human subjects research by the Institutional Review Board at Augusta University and performed in compliance with the Declaration of Helsinki. All ethical regulations relevant to human research participants were followed. All experiments were carried out in accordance with institutional biosafety and chemical safety guidelines and regulations.

Mice

Ephx2 floxed mice (Ephx2^flox/flox) were obtained from NIEHS/NIH colony and bred with albumin (alb)-Cre mice to create hepatocyte-specific Ephx2 KO mice. For angiotensin II (AngII) infusion model, Ephx2^flox/flox mice and alb-Cre mice were bred with LDLR KO mice, respectively, and intercrossed to generate Ephx2^flox/flox/alb-Cre mice in the LDLR KO background. Only male mice were used in this study because, in humans, AAA are more common in males compared with females. There is also a significant sex difference of AAA incidence in males versus females in the AngII-induced animal model of AAA (80–100% vs ~20%, respectively). 10–12 weeks old male mice were anesthetized by isoflurane vaporizer (0.5–1.0 L/min for oxygen flowmeter, 4–5% for induction and 1–3% for maintenance, EZ Anesthesia Systems) and euthanized with intraperitoneal pentobarbital 150 mg/kg, inhaled anesthesia (isoflurane) followed by carbon dioxide (CO₂) narcosis and cervical dislocation or bilateral thoracotomy, in accordance with AVMA Panel 2007 recommendations and institutional IACUC guidelines. All mice were randomized to different treatment groups to minimize experimental variability, and surgeons were blinded to group allocation to prevent accidental or selection bias. We have complied with all relevant ethical regulations for animal use.

AngII-induced AAA model

AngII (1000 ng/kg/min, Enzo Life Sciences) was infused into 10–12 week old male mice via osmotic minipumps (ALZET Model 2004) as described previously^54–56. In some experiments, mice were co-treated orally with an sEH inhibitor, TPPU [1-(1-Propionylpiperidin-4-yl)-3-(4-(trifluoromethoxy)phenyl)urea, 3–10 mg/kg body weight in drinking water]. Abdominal aortic luminal diameter was measured using high frequency ultrasound. Mice were euthanized at three weeks after minipump implantation, and abdominal aortic outer diameter was also measured via microscopy. Aortic tissues, liver, and blood were collected from mice for further analysis. AAA was defined as local dilation of the abdominal aorta that is at least 1.5 times the size of the reference aortic diameter.

CaCl₂-induced AAA model

CaCl₂ application model of AAA (using 0.5 mol/L of CaCl₂, Sigma-Aldrich) was performed as described previously^54–56. In brief, following laparotomy, saline (sham control) or 0.5 mol/L of CaCl₂ was applied to the infrarenal aortic adventitial surface for 15 min, followed by rinsing with 0.9% sterile saline and surgical closure. In some experiments, mice were co-treated orally with TPPU (3–10 mg/kg of body weight in drinking water) to inhibit sEH activity. Abdominal aortic luminal diameter was measured using high frequency ultrasound. After three weeks, mice were also anesthetized, abdominal aortic outer diameter was measured via microscopy, and tissues were collected for further analysis.

sEH activity assay

Soluble epoxide hydrolase activity was determined using sEH activity assay kit (BioVision) according to the manufacturer’s protocol. Briefly, liver and aorta tissues were homogenized with sEH assay buffer included in the kit, and total protein was extracted. Protein concentration was determined with Pierce BCA Protein Assay Kit (ThermoFisher). sEH activity was calculated by measuring fluorescence (Ex/Em 362/460 nm) in kinetic mode using the standard curve and equation provided by manufacturer. A potential limitation of commercially available sEH activity assay kits is that they may detect other enzymes with hydrolase activities, such as glutathione S transferases, which can open epoxide groups and release the fluorophore detected by these kits.

Blood pressure measurement

Blood pressure was measured using a previously validated tail cuff method (Coda 6, Kent Scientific, Torrington, CT). Mice were conditioned to the instrument and procedure for 5 consecutive days prior to pump implantation. To ensure a more robust estimation of systolic blood pressure (SBP), we used the interquartile mean of SBP measurements achieved through 30 measurement cycles every other day.

Immunohistochemistry

Paraffin-embedded aortic tissue sections were stained with hematoxylin and eosin (H&E), Verhoeff van Gieson (VVG), Mac-3 (1:100 dilution, BD Pharmingen 550592), MMP-2 (1:100 dilution, Bioss, bs4599R), MMP-9 (1:100 dilution, Bioss, bs0397R), C3 (1:500 dilution, Novus Biologicals, NB200-540), SAA (1:100 dilution, R&D Systems, AF2948)), gal-3 (1:500 dilution, Cell Signaling, 89572S) and HistoMouse-SP (Invitrogen) or DAB Substrate (Vector Labs) kits were used for visualization.

Western blotting

Protein extraction and Western blotting were performed as described previously^54–56. Antibodies used were: sEH (1:1000 dilution, Cayman Chemicals, 10010146), SAA (1:1000 dilution, R&D Systems, AF2948), transferrin (1:500 dilution, Abcam, Ab82411), H3K4me3 (1:500 dilution, Cell Signaling Technology, 9751), gal-3 (1:1000 dilution, Santa Cruz, sc-32790) and GAPDH (1:1000 dilution, Invitrogen, AM4300).

Quantitative PCR

Total RNA was extracted from aortic and liver tissues with QIAzol Lysis Reagent, and purified with RNeasy Lipid Tissue Mini Kit (Qiagen). Real time quantification of mRNA levels of the genes of interest was performed using Brilliant II SYBR Green QPCR Master Mix (Agilent Technologies) per manufacturer’s instructions. Normalized Ct values were subjected to statistical analysis and fold difference was calculated by ∆∆ Ct method as described previously^54–56. Primer sequences are shown in Supplementary Table.

Enzyme-linked immunosorbent assay (ELISA)

Levels of 12,13-DiHOME (dihydroxyoctadecenoic acid) and 14,15-DHET in conditioned medium and plasma of mouse and human blood were also measured by ELISA kits (Cayman Chemical) per manufacturer’s instructions. Plasma levels of gal-3 were also measured using ELISA kits (R&D Systems). Collection of human bloods were designated as exempted from human subjects research by the Institutional Review Board (IRB) at Augusta University.

Ex vivo co-incubation of liver with cigarette smoke extract

Fresh liver tissues harvested from 8 to 12 week old C57Bl/6 mice were cut into 3–5 mm slices (approximately 5 mg in weight) and co-incubated with 1% cigarette smoke extract (University of Kentucky Reference Cigarette 3R4F) in 500 µl of DMEM medium without FBS or antibiotics for 6 h at 37°. Then, liver tissues were harvested for gene expression quantification as described above.

Ex vivo co-incubation of liver with aorta

Following laparotomy in C57Bl/6 mice, saline (sham control) or 0.5 mol/L of CaCl₂ was applied to the infrarenal aortic adventitial surface for 15 min, followed by rinsing with 0.9% sterile saline and surgical closure. After 7 days [a time point prior to AAA formation], aortas were harvested. Liver tissues were freshly collected from control mice and cut into 3–5 mm slices, as described above, and co-incubated with two aortas harvested from sham- or CaCl₂-treated mice in 500 µl of DMEM medium without FBS and antibiotics at 37°. After 12 h of incubation, liver tissues were harvested to measure sEH activity and gene expression as described above.

Ex vivo co-incubation of liver with recombinant gal-3 protein

Fresh liver tissues were cut into 3–5 mm slices (approximately 5 mg in weight) and co-incubated with recombinant mouse gal-3 protein (10 µg in 500 µl DMEM medium without FBS or antibiotics for 24 h at 37°). Then, liver tissues were harvested for protein and gene expression quantification as described above.

Liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis, data processing and bioinformatics

Whole aortas from either sham- or CaCl₂-treated mice were harvested 7 days after laparotomy and transferred to a 48-well plate with 500 µl of serum-free DMEM media. The conditioned media were collected after 12 h incubation, centrifuged at 1000 × g for 10 min, filtered through a Nalgene 0.2 µm pore vacuum filter, and concentrated using a 10 kDa MWCO Amicon Ultra-15 centrifugal filter (MiliporeSigma) operating at 1500 × g for 30 min with 50 mM ammonium bicarbonate buffer until the final concentration was 2 mg/ml. Protein concentration was determined with Pierce BCA Protein Assay Kit (ThermoFisher). Protein digestion and mass spectrometry were performed as previously described⁵. Briefly, extracted proteins were separated using an Ultimate 3000 nano-UPLC system (Thermo Scientific) and run on an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). Raw data were analyzed using Proteome Discoverer (v1.4, Thermo Scientific) and searched against the UniProt mouse protein database. Peptide-spectrum match (PSM) values, a total number of identified peptide spectra matched for the protein, were log-transformed to normalize the dataset. LIMMA package on R was used to compare protein levels in conditioned media of the aorta obtained from sham- versus CaCl₂-treated mice, and p-values were adjusted using FDR (false discovery rate), with significance set at an adjusted p-value < 0.05. The annotated genes were used as input for functional enrichment analysis, which was performed using the Gene Ontology (GO) terms.

Lipid measurements

EETs/DHETs and EpoMEs (epoxyoctadecamonoenoic acids)/DiHOMEs were analyzed using mass spectrometry as previously described⁴³. Briefly, isotope-labeled standards were added to the samples in methanol as internal controls and extracted using solid phase extraction. The extracted samples were analyzed using the Ultimate 3000 nano-UPLC system (Thermo Scientific) and Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). In some experiments, levels of 12,13-DiHOME and 14,15-DHET in plasma and conditioned medium were also measured by ELISA kits as described above.

Profiling of pro-inflammatory cytokines

Mouse plasma cytokine profiling was performed at the Georgia Cancer Center Immune Monitoring Shared Resources at Augusta University following the manufacturer’s instructions. Biolegend Mouse Inflammation 13-Plex Panel FbBA228ser was used to quantify key pro-inflammatory and anti-inflammatory cytokines. Briefly, plasma samples were thawed on ice, and 20 µL of each sample was used for the assay. Samples were incubated with magnetic capture beads coated with specific antibodies against each target cytokine in a 96-well plate. The beads were then washed using a magnetic plate washer to remove unbound proteins. Detection antibodies were added to each well and incubated for 1 h at room temperature, followed by a secondary incubation with streptavidin-phycoerythrin for signal amplification. Plates were washed again, and beads were resuspended in assay buffer for reading. Fluorescence intensity was measured using a Luminex 200 instrument, and cytokine concentrations were calculated based on standard curves generated with recombinant protein standards.

Statistics and reproducibility

All statistical analyses were performed using Graphpad Prism version 10.2.2 Software (GraphPad Software, Inc., USA). Results are expressed as mean ± SEM. Differences between two groups were analyzed by student’s t-test. Multiple group datasets were evaluated for normality, and differences were analyzed by two-way ANOVA followed by Bonferroni post-hoc analysis. p values less than 0.05 were considered to be significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

H.W.K. and N.L.W. conceived and designed research, D.S.K., T.H., M.O., B.G., H.S., P.V., R.C., N.C., M.M., R.P., S.C., T.W.B., R.H., G.A., H.G.S., M.S., S.H. W.Z., and P.C. performed experiments, D.S.K., T.H., M.O, B.G., P.V., R.C., N.C., N.L.W., and H.W.K. analyzed data, D.J.F, G.C., Y.H., X.L., R.L., A.G., L.Y., D.C.Z., S.H.H., B.D.H., N.L.W., and H.W.K. interpreted results of experiments, N.L.W. and H.W.K. wrote manuscript, D.J.F., R.L., S.H.H., B.D.H., N.L.W., and H.W.K. edit and revised manuscript, N.L.W. and H.W.K. approved final version of manuscript.

Peer review

Peer review information

Communications Biology thanks María Galán and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Christopher Hine and Dario Ummarino. [A peer review file is available].

Data availability

Datasets analyzed in this study are available from the corresponding authors upon reasonable request. Specific data underlying each Figure is noted in Figure Legends and Supplementary Data file. The LC-MS/MS proteomics data have been deposited to the ProteomeXchange Consortium via the Proteomics Identification database (PRIDE) partner repository with the dataset identifier PXD073197.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-09765-x.