All animal use procedures were approved by the Institutional Animal Care and Use Committee (IACUC #201900248) at Michigan State University (East Lansing, Michigan) and were performed in accordance with local and federal guidelines.

Eight multiparous (2.5 ± 1.06; mean parity ± SD) lactating Holstein dairy cows from the Michigan State University Dairy Cattle Teaching and Research Center were enrolled for the in vivo part of this study. At enrollment, cows averaged (mean ± SD), 204 ± 22 d in milk and, on the scale from 1 to 5 as defined by Wildman et al., a body condition score of 3.26 ± 0.15 (25). All cows received a common diet formulated to meet the nutrient requirements of the average lactating cow (26). Cows were housed in individual tie stalls bedded with sawdust, cleaned twice daily, and equipped with automatic waterers. Cows were fed 115% expected intake at 7000 h daily. Feed access was blocked from 0500 to 0700 h for orts (uneaten feed) collection and measurements along with offering of new feed. Cows were milked three times per day at 0400, 1200 and 2000 h. Standard herd health examinations were conducted throughout the study.

Solid phase extractions and quantitative targeted lipidomic analyses of plasma and adipocyte culture medium samples were performed by the Wayne State University Lipidomics Core (Detroit, Michigan) using high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). For recovery and quantitation, samples (1 mL plasma or culture medium) were spiked and mixed thoroughly with 1 ng of each of the following internal standards: 15(S)-HETE-d₈, leukotriene B₄-d₄, resolvin D2-d₅, 14 (15)-EpETrE-d₁₁, and prostaglandin E₁-d₄. Polyunsaturated fatty acids and their metabolites were then extracted using C18 extraction columns as described previously (34–37). In brief, spiked samples were applied to conditioned C18 cartridges, washed with HPLC-grade water, hexane, and dried under vacuum. Following, 0.5 mL methanol was used to elute samples from cartridges. The eluate was then dried under nitrogen, and the residue redissolved in methanol:25 mM aqueous ammonium acetate (1:1).

Following, HPLC was performed using Luna C18 (3µm, 2.1x150 mm, StrataX SPE cartridge, 30 mg sorbent; Phenomenex, Torrance, CA) columns. The mobile phase consisted of a gradient between methanol-water-acetonitrile at 10:85:5 v/v and 90:5:5 v/v, both with 0.1% ammonium acetate. The gradient program consisted of 0–1 min, 50%; 1–8 min, 50-80%; 8–15 min, 80-95%; and 15–17 min, 95%. The flow rate was 0.2 mL/min. The HPLC eluate was then directly introduced to the ESI source of the QTRAP7500 triple quadrupole mass analyzer (SCIEX, Framingham, MA) operated in negative ion mode. The operating conditions were as follows: Curtain gas and GS1 at 40 psi, GS2 at 70 psi, temperature at 500˚C, ion spray voltage at -2500 V, collision gas at 12 psi, declustering potential at -60 V, and entrance potential at -7 V. The eluate was monitored by Multiple Reaction Monitoring (MRM) to detect unique molecular ion-daughter ion combinations for each transition to monitor a total of 6 lipids and 83 lipid mediators. The MRM monitored each transition for 120 sec around the established retention time for each lipid mediator. Optimixed collisional energies (18–35 eV) and collision cell exit potentials (7–10 V) are used for each MRM transition. Mass spectra for each detected lipid mediator were recorded using the Enhanced Product Ion (EPI) feature to verify identities of detected peaks and MRM transition and retention times matched the standards. Data were collected and MRM transition chromatograms were quantified using SCIEX OS 3.4 Software (SCIEX). In each chromatogram, internal standard signals were used for recovery normalization and quantitation of each lipid mediator. Details on MRM-internal standard mapping are included in Supplementary File S1 .

To ensure robustness and accuracy in our evaluation of treatment effects, OXL with insufficient or inconsistent detection across samples (<3 observations per group per time relative to infusion) were excluded from the statistical analysis. Health parameters and plasma OXL data were analyzed using the GLIMMIX procedure in SAS (SAS On Demand for Academics, SAS Institute, Cary, NC). For in vivo data, variables evaluated over time were analyzed with repeated measures using the random effect of animal and the fixed effects of treatment, time, and their interactions. PRE OXL concentration values were included as covariates. For in vitro data, the fixed effect of treatment was assessed. The normality of all variables was verified using the Shapiro-Wilk Test (P<0.05). Residuals of each model were checked and found to be normally distributed. Results are presented as mean ± SEM unless otherwise stated. Significance was declared at P ≤ 0.05 and tendencies at P ≤ 0.10.

The relative impacts of in vivo LPS infusion on 13 user-defined OXL biosynthetic pathways were assessed by quantitative enrichment analysis on the web-based platform MetaboAnalyst 6.0 (38) as approached in (39). This method employs the global test algorithm and uses a generalized linear model to produce a Q-statistic for each self-defined set of metabolites. Of the 75 metabolites detected, 69 were recognized across the software’s referenced databases and were included in our analyses. Pathways were considered significantly enriched with Holm P<0.00625 (=0.05/8) and false discovery rate (FDR) <5%. Principal component analyses were performed using the MetaboAnalyst 6.0 platform to determine if exposure to endotoxin (e.g., LPS versus SAL) and timing relevant to exposure (e.g., +2H, +12H) resulted in distinct plasma OXL profiles. These results, which facilitate the visualization of sample clustering across treatment and time groups, are presented as score plots.

Correlation analyses were conducted in JMP Pro (v.17, SAS Institute Inc.). To initially screen OXL for associations with physiological parameters, the Fit Y by X function was used to evaluate all OXL in relation to respiratory rate, heart rate, and rectal temperature across treatments. The top ten OXL exhibiting linear correlations were selected based on the strength (correlation coefficient) and statistical confidence (P-value) of each correlation. Top-ranking OXL were then analyzed alongside all physiological parameters and OXL using the Multivariate and Correlations function, allowing for a comprehensive assessment of correlations across health parameters and OXL profiles.

In vivo infusion of LPS induced signs and symptoms characteristic of endotoxemia. Following LPS infusion, cows exhibited tachypnea, fever, and tachycardia, whereas animals that received SAL did not ( Figure 1 ). Exposure to LPS increased respiratory rate immediately following infusion through +2H compared to SAL and gradually returned to normal (eupneic) levels by 5 h post-infusion ( Figure 1 ). LPS induced fevers in cows, with rectal temperatures peaking at approximately 4 h post-infusion, and temperatures transiently returning to normal (afebrile) temperatures by +12H ( Figure 1 ). As expected, LPS exposure induced excessive salivation, increased urination and defecation, and coldness of the extremities, particularly of the ears (data not shown). As reported in the companion paper, LPS reduced milk yield and feed intake during the first 3 d following LPS infusion (32).

Our initial analysis assessed the effect of systemic endotoxin exposure on plasma eicosanoid profile. Of the 13 pathways and corresponding metabolite sets analyzed, at +2H, LPS altered COX-derived EPA-based, CYP-derived AA-based, COX-derived AA-based, and LOX-derived AA pathways compared to SAL ( Figure 2 ). Elevations in the COX-derived EPA-based and LOX-derived AA-based OXL metabolite sets were sustained between LPS and SAL through +12H ( Figure 2 ).

Principal component analyses (PCA) revealed distinct separation of OXL profiles between LPS- and SAL treatments at +2H ( Figure 3A ); however, plasma OXL profiles trended toward convergence in LPS and SAL at +12H ( Figure 3A’ ). Of the 75 quantified, LPS significantly upregulated plasma levels of 6 and downregulated 8 OXL at +2H compared to SAL ( Figure 3B ). At +12H, however, LPS upregulated abundances of 4 and downregulated 7 OXL relative to SAL ( Figure 3B’ ). Compared to SAL, LPS increased plasma content of total FA, total n-3 FA, total n-6 FA, and OXL by species relative to SAL ( Figure 3C ). Compared to SAL, LPS increased plasma content of total FA, n-3, n-6, HODE, and TX at +2H ( Figures 3C, D ). Interestingly, elevations in n-3 FA and TX tended to be sustained in LPS through +12H ( Figures 3C, D ). In contrast, LPS reduced total HETE relative to SAL at +2H, with a similar trend observed at +12H ( Figure 3D ).

Compared to SAL, LPS increased plasma AA (2.6-fold) and ALA (2.5-fold) at +2H ( Figure 3E ). Additionally, LPS tended to increase DPA, EPA, and LA at +2H by 3-, 1.9-, and 3-fold, respectively, relative to SAL. At +12H, however, LPS increased AA and DHA 1.4- and 1.5-fold, respectively, over SAL ( Figure 3E ). Corresponding with elevations in relative precursor FA, LPS increased total AA-derived and LA-derived OXL levels in plasma at +2H compared to SAL ( Figure 3F ). There was no effect of treatment observed on plasma EPA-, DPA-, and DHA-derived OXL levels ( Figure 3F ).

Compared to SAL, LPS increased the 5-LOX AA-derived product 18-carboxy-dinor-LTB₄ in plasma +2H and +12H ( Figure 4A ). A tendency for the same pattern was observed in 5,6-DiHETE at both time points post-infusion. LPS tended to enhance 5-HETE, a product of 12-LOX, at +12H, but not at +2H ( Figure 4A ). Compared to SAL, products of the 15-LOX pathway, 15-HETE and 15-epi-LXA₄, exhibited opposing trends in LPS, with 15-HETE levels elevated and 15-epi-LXA₄ levels reduced at both +2H and +12H ( Figure 4A ). LPS tended to enhance 15-OxoETE concentrations at +12H compared to SAL ( Figure 4A ). LPS increased the COX-derived TXB₂ 17-fold over SAL at +2H; this gap lessened to 3-fold at +12H ( Figure 4A ). LPS tended to increase the prostanoid-derived 6-keto-PGF1α and the CYP metabolite 11-HETE at +2H and +12H relative to SAL at each respective time point ( Figure 4A ). Notably, compared to SAL, LPS increased adipocytes’ release of 5-HETE and 6-keto-PGF1α into the media ( Figure 4B ).

LPS augmented plasma content of LOX-derived LA metabolites 9- and 13-HODE at +2H compared with SAL ( Figure 4C ). Additionally, LPS increased plasma content of the CYP product 9,10-DiHOME at +2H compared to SAL ( Figure 4C ). LPS tended to increase 9,10-DiHOME at +12H and 12,13-DiHOME at +2H ( Figure 4C ). However, LPS increased plasma 12,13-EpOME content at +12, but not at +2H, compared to SAL ( Figure 4C ). Exposure to LPS increased adipocytes’ release of 13-HODE into the media, corresponding with the metabolite’s changes in plasma ( Figure 4D ). However, in contrast with systemic observations, LPS reduced adipocytes’ release of 9,10-diHOME and 12,13-EpOME compared to CON ( Figure 4D ).

At +2H, MaR1 and MaR2 remained unchanged in LPS, yet decreased in SAL relative to baseline ( Figure 5A ). While no differences were found at +2H, LPS increased plasma 17-HDoHE at +12H ( Figure 5A ). Compared to SAL, LPS increased plasma content of the 15-LOX product protectin D1 (PD1) at +2H, however, no difference was observed between treatments at +12H ( Figure 5A ). Compared to SAL, LPS increased the concentrations of all quantified Rv (RvD3, RvD4, RvD5, and RvD6) at +2H; however, at +12H, there was no effect of treatment ( Figure 5A ). No differences were observed between treatments in non-enzymatically derived plasma -HDoHE (4-, 7-, 10-, 11-, 13-, 14-, or 16-HDoHE) at any time point (data not shown). Contrasting with the changes observed in plasma, LPS attenuated adipocyte release of DHA ( Figure 5B ). However, LPS upregulated adipocytes release of 19,20-DiHDPA compared to CON ( Figure 5B ).

At +12H, LPS reduced plasma content of the 12-LOX EPA-based product, 12-HEPE, but was unchanged at +2H, compared to SAL ( Figure 5C ). Similarly, LPS tended to reduce plasma content of the COX-derived metabolite TXB₃ at +12H ( Figure 5C ). In contrast, LPS tended to increase plasma content of 15-HEPE, a product of 15-LOX, compared to SAL at +12H ( Figure 5C ). While no differences were observed for RvE1 or RvE3 between treatments, LPS tended to increase resolvin E2 (RvE2) at +2H and augmented RvE2 + 12H ( Figure 5C ) compared to SAL. No differences were observed in adipocyte production of EPA-based OXL upon LPS exposure (data not shown).

Although LPS increased plasma DPA compared to SAL, there was no effect of treatment on the DPA-based OXL PD1_n-3,DPA, RvD5_n-3,DPA, or AT-RvD5_n-3,DPA (data not shown).

Our multivariate analysis revealed a strong positive correlation between respiratory rate and plasma content of RvD5_{n-3, DPA} ( Figure 6 ). Respiratory rate and plasma content of 17-HDoHE levels also demonstrated a positive association ( Figure 6 ). Negative correlations were observed between respiratory rate and 5-HETE, 10-HDoHE, 9-oxoODE, and 20-HDoHE ( Figure 6 ). Positive correlations were identified between heart rate and 5,6-DiHETE, TXB₂, 18-carboxy-dinor-LTB₄, and 19,20-DiHDPA ( Figure 6 ). No significant correlations were detected between rectal temperature and plasma content of OXL ( Figure 6 ).

The clinical responses to LPS infusion observed in cows—tachycardia, fever, tachypnea, salivation, polyuria, increased defecation, and cold extremities—underscore our model’s capacity to reliably recapitulate the acute inflammatory responses observed in humans and other species following single LPS infusions (5). Our results are consistent with previous reports in cows, wherein fevers peaked at 4 h post-infusion and respiratory rates were highest at +2H (40). These well-described symptoms are considered hallmarks of an acute inflammatory response to endotoxemia, reflecting the systemic activation of immune pathways triggered by endotoxin exposure (41). Our findings further validate the use of in vivo LPS infusion as a reproducible method to study endotoxemia in vivo, providing a robust platform to investigate the downstream biochemical and metabolic changes associated with this condition. Moreover, this model enables further exploration of the role of bioactive lipid mediators, such as OXL, in modulating the inflammatory response and potentially influencing the recovery and health outcomes with translational potential across species.

We report, for the first time, that LPS infusion-induced endotoxemia triggers a dynamic and complex systemic OXL response, with distinct lipid mediator pathways activated in a time-dependent manner in dairy cows. The observed alterations in specific OXL pathways, particularly the COX-derived EPA-based and LOX-derived AA-based OXL, indicate that our in vivo model of endotoxemia recapitulates the rapid OXL biosynthetic response to endotoxin observed in vitro (42). The early and sustained changes in these key mediators may reflect an adaptive or maladaptive immune response to endotoxin, suggesting that OXL play critical roles in the pathogenesis and clinical progression of endotoxemia. Moreover, the persistent alterations in COX-derived EPA-based OXL may indicate a shift toward anti-inflammatory and pro-resolving responses, as COX-synthesized EPA-derived OXL are generally regarded as less inflammatory than their AA-derived counterparts.

In contrast, the sustained increase in LOX-derived AA-based OXL, including LT and HETE, could contribute to a dysfunctional inflammatory response during endotoxemia. Previous research highlights LT and HETE as key mediators of neutrophil chemotaxis, vascular permeability, and immune cell activation (43–45). The persistence of these pathways observed in the present study suggests that the inflammatory cascade and OXL production triggered by LPS does not resolve fully within the observed timeframe, potentially contributing to the chronic aspects of endotoxemia, such as prolonged immune activation, tissue damage, and end-stage organ failure (46, 47).

The distinct separation observed in OXL profiles between the LPS and SAL treatments at +2H underscores the rapid and profound impact of endotoxin exposure on systemic lipid mediator production. The clear distinction in metabolite profiles in vivo is in line with a robust inflammatory response initiated early after LPS exposure, as reported previously (48, 49). The trend toward similarity in OXL profiles at +12H suggests that, while the acute effect of inflammation diminishes over time, residual effects of endotoxin persist, particularly through sustained alterations in key OXL synthesizing pathways.

Endotoxin exposure increased plasma concentrations of both n-3 and n-6 FA at +2H, including AA, ALA, DPA, EPA, and LA. The heightened levels of these FA suggest that LPS triggers the hydrolysis and release of FA from cellular stores (e.g., phospholipids and neutral lipids), likely to fuel the synthesis of bioactive lipid mediators as part of the immune response (33). The sustained elevations in plasma AA and sizeable increase in DHA concentrations in both plasma and adipocyte media further support the role of FA in sustained inflammatory signaling and immune modulation during endotoxemia.

Endotoxin exposure leads to distinct and time-dependent responses in AA-derived OXL, suggesting a complex and dynamic response to systemic inflammation. The upregulation of the 5-LOX byproduct 18-carboxy-dinor-LTB₄ at both +2H and +12H highlights that stimulation of leukotriene synthesis is prolonged following LPS exposure. This molecule is a metabolite of LTB₄, a potent pro-inflammatory mediator involved in neutrophil recruitment and activation, which contributes to sustained inflammation during endotoxemia (50). In rodents and humans, the 5-LOX-derived metabolite, 5-HETE, is produced by neutrophils in response to bacterial pathogens and their remnants, including LPS (51). Notably, 5-HETE and its isomers are potent chemoattractants that induce neutrophil migration and macrophage infiltration (52, 53). In the present study, 5-HETE levels tended to be elevated in plasma at +12H, but not +2H, in cows treated with LPS, suggesting this OXL may be involved in the latter stages of the inflammatory response. However, 5-HETE release by adipocytes was dramatically elevated following 3 h of LPS exposure, suggesting that AT may play a more immediate role in 5-HETE production during endotoxemia. Moreover, this finding suggests that adipocytes could significantly contribute to the early inflammatory response, potentially serving as an initial chemoattractant source to prime the infiltration of immune cells into AT commonly observed during inflammatory disease (54, 55).

The trend toward elevated plasma 5,6-DiHETE, a dihydroxy derivative of leukotriene A₄ (LTA₄), in response to LPS at +2H and +12H further suggests that activation of the 5-LOX pathway is sustained, reinforcing its potential significance in mounting the immune response to endotoxemia. In fact, several studies have identified anti-inflammatory and pro-resolving characteristics associated with this molecule, including its capacity to reduce vascular permeability and endothelial barrier disruption (56, 57).

Several COX-derived, AA-based OXL were upregulated systemically in response to LPS, including TXB₂, which was elevated nearly 17-fold at +2H, and 3-fold at +12H. This sharp rise suggests that thromboxane-mediated pro-thrombotic and vasoconstrictive responses occur during the acute phase of endotoxemia. The attenuation of plasma TXB₂ at +12H suggests that tapering of the thromboxane surge occurs, although residual effects of COX activation persist. Notably, no differences were observed in adipocytes’ TXB₂ production following LPS exposure, suggesting that the systemic effects observed may occur independently of adipocytes, rather, through production by platelets or immune cells (58).

The COX metabolite 6-keto-PGF1α, a stable product of prostacyclin, tended to be elevated at +2H and was markedly greater at +12H in response to LPS, reflecting sustained vascular and endothelial activation. These findings are in line with the well-established role of prostacyclins in regulating blood flow, limiting platelet aggregation, and inhibiting leukocyte adhesion (59–61). Interestingly, production of 6-keto-PGF1α was drastically elevated in adipocytes following LPS exposure, suggesting that AT may contribute actively to the vascular and endothelial remodeling seen during endotoxin-induced inflammation (62).

The CYP-derived 11-HETE exhibited a pattern in plasma similar to 6-keto-PGF1α, increasing at both +2H and +12H in LPS compared to SAL. As no differences were observed in adipocyte production of 11-HETE in response to LPS exposure, these effects further emphasize that LPS persistent endothelial and vascular effects during systemic inflammation occur independently of adipocyte responses. Moreover, 11-HETE is involved in modulating vascular tone and inflammation, highlighting the multifaceted involvement of CYP pathway in regulating endotoxin-induced vascular dysfunction, particularly in the acute phase of inflammation [as reviewed in (63)].

Also important to consider is that not all AA-derived OXL exhibit pro-inflammatory actions. Of note is the CYP-derived, pro-resolving mediator 15-epi-LXA₄, which attenuates cytokine release and enhances pathogen clearance in human and murine models of pneumonia (64). We found that LPS treatment decreased plasma content of this compound, a stable and longer-acting derivative of lipoxin A₄ (65). This finding in our bovine model contrasts with those previously described in the lungs of mice, which report that 15-epi-LXA₄ increases following LPS exposure (64). These divergent responses between species may reflect differences in the regulation of lipoxin pathways between species, across tissues, or varying roles of 15-epi-LXA₄ in systemic versus local inflammatory contexts.

Several LA-derived metabolites were altered in plasma following LPS infusion, with elevated levels of the LOX-derived 9-HODE and 13-HODE observed at +2H. These metabolites, which are known markers of oxidative stress and inflammation, may contribute to the tissue damage and cellular dysfunction frequently observed during endotoxemia. Of equal importance to note, however, are the pro-lipogenic, PPARγ-mediated effects exhibited by these HODE, which may serve as a negative feedback loop to limit AT lipolysis during endotoxemia (11).

Additionally, the CYP-derived product, 9,10-DiHOME was elevated in plasma at +2H, indicating upregulated activity of the CYP-to-soluble epoxide hydrolase (sEH) pathway in the early phase of endotoxin-induced inflammation (66). A trend toward higher plasma levels of 9,10-DiHOME at +12H and 12,13-DiHOME at +2H in response to LPS suggests these metabolites may play key roles in modulating the inflammatory response over a more extended period. Moreover, LPS increased plasma 12,13-EpOME at +12H, but not +2H, pointing toward a time-dependent activation of the CYP and sEH pathways that may contribute to the progression of inflammation and vascular permeability in the later stages of endotoxemia. In contrast, both 9,10-DiHOME and 12,13-EpOME production were decreased in adipocytes exposed to LPS, suggesting that endotoxin may suppress local synthesis of these products, perhaps to limit inflammation within AT (14). The discrepancy observed between plasma and adipocyte DiHOME and EpOME levels suggests a compartmentalized regulation of CYP metabolites during endotoxemia, with circulating levels reflecting an upregulation to support systemic inflammatory responses while, within AT, downregulated production of these OXL to limit local inflammation and maintain metabolic homeostasis (67). Collectively, our observations underscore the importance of LA-derived metabolites in the inflammatory cascade, with AT potentially functioning as both a target and source of LA-based OXL during systemic inflammation.

DHA-derived OXL, including resolvins (Rv), protectins (PD), and maresins (MaR) are central to the resolution of inflammation. Interestingly, while exposure to LPS increased several D-series Rv (RvD3, RvD4, RvD5, and RvD6) at +2H, no differences were observed at +12H, suggesting a short-lived spike in the abundance of pro-resolving mediators early in the inflammatory response. These findings may highlight an attempt by the immune system to counteract the pro-inflammatory signals generated by AA-derived metabolites. However, the lack of sustained elevations in Rv beyond +2H may indicate that the resolution phase of inflammation is not adequate during endotoxemia, which may contribute to chronic inflammation and delayed recovery times. We speculate that this shift may be attributable to rapid exhaustion of DHA reserves within cell phospholipid membranes or lipid droplets during endotoxemia. This concept aligns with findings in human and murine macrophages, which demonstrate that D-series Rv activate phospholipase D, freeing lipid-containing phosphatidic acids and cholines from cell membranes (68, 69). Important to note, however, is that this response may also be driven by changes in oxidative enzymes’ activation status and preferences for specific FA substrates during inflammatory challenges.

Similarly, the 15-LOX-derived pro-resolving metabolite, PD1, was elevated at +2H, but returned to baseline levels at +12H. In LPS-stimulated macrophages of primates, PD1 exerts its anti-inflammatory effects through stimulation of peroxisome proliferator-activated receptors α and γ (PPARα, PPARγ), which inhibit NF-κB translocation and downregulate transcription of pro-inflammatory genes encoding TNF-α, IL-6, and iNOS (16, 70). Taken together, these findings suggest that, while pro-resolving mediators are initially upregulated during LPS-mediated inflammation, their effects may be transient and limited in the context of chronic endotoxin exposure. Moreover, the transient spike observed in PD1 levels implies that anti-inflammatory pathways are activated but may be insufficiently sustained to fully outweigh the pro-inflammatory cascades initiated by endotoxemia.

While the LPS treatment did not increase plasma 19,20-DiHDPA, its production was notably enhanced in adipocytes following endotoxin exposure. This finding suggests that AT may selectively increase the synthesis of 19,20-DiHDPA as part of the localized response to pathogens, independent of systemic changes detectable in plasma. Of interest, 19,20-DiHDPA is associated with anti-inflammatory effects and may act to offset the production of pro-inflammatory mediators within AT. The upregulation of this metabolite’s synthesis in adipocytes, specifically, suggests a potential protective mechanism within AT aimed at mitigating the negative consequences of prolonged inflammation and preserving adipocyte functionality during endotoxemia.

The differential regulation of EPA-derived OXL during endotoxin exposure reflects a complex interplay between pro-inflammatory and pro-resolving pathways. The suppression of specific LOX and COX metabolites, such as 12-HEPE and TXB₃, suggests a potential shift away from certain inflammatory mediators in response to endotoxemia. This reduction could indicate an adaptive response aimed at limiting excessive inflammation in the latter stages of endotoxin exposure, however, this response may impair the immune system’s capacity to effectively clear pathogens.

Conversely, the elevation of 15-LOX products and specific E-series Rvs, such as RvE2, points toward a heightened activation of pathways involved in inflammatory resolution. In fact, RvE2 serves as a competitive antagonist of the AA-derived LTB₄ and inhibits the activation of its receptor, LTB₄R1, limiting leukocyte recruitment and activation (71). The increases in pro-resolving mediators observed herein underscore the body’s efforts to counteract pro-inflammatory pathways triggered by endotoxemia, especially as the immune response progresses over time.

The correlations observed between respiratory and heart rates with specific OXL may suggest that these potent lipid mediators actively influence key symptoms associated with endotoxemia. The strong positive correlation between respiratory rate RvD5_n-3,DPA—an OXL known for its anti-inflammatory, pro-resolving actions—implies that elevated respiratory rates may be driven by upregulation of resolving pathways in response to inflammatory stimuli. RvD5_n-3,DPA has been shown to enhance macrophage-mediated pathogen clearance, which may reflect an adaptive attempt to restore homeostasis in the microenvironments of various organs and limit tissue damage driven by endotoxemia (72, 73). Likewise, the positive association with the DHA-derived RvD- and PD- precursor, 17-HDoHE, further supports OXL roles in respiratory modulation, as prior studies in mice suggest these compounds reduce pulmonary inflammation and promote pathogen clearance by neutrophils and macrophages (74).

Conversely, the negative relationship between respiratory rate and pro-inflammatory 5-HETE suggests that this molecule may intensify inflammatory processes that challenge respiratory function. Past research demonstrates that 5-HETE stimulates constriction of the vasculature and accumulation of fluid in the lungs, which may lead to respiratory distress (75). Moreover, this molecule is known to slow contraction of smooth muscle in pulmonary bronchioles, which may limit gas exchange and blood oxygenation during endotoxemia (76).

Heart rate demonstrated moderate positive correlations with OXL such as TXB₂, 18-carboxy-dinor-LTB₄, and 5,6-DiHETE, suggesting that these pro-inflammatory mediators may drive tachycardia and vascular responses commonly seen in endotoxemia. TXB₂, a stable metabolite of thromboxane A2, induces vasoconstriction and platelet aggregation, potentially increasing heart rate and cardiac workload as part of an acute vascular response to inflammation (77). Additionally, LTB₄ and its derivatives, including 18-carboxy-dinor-LTB₄, have been linked to heightened leukocyte activation and endothelial interactions that could intensify systemic inflammation, subsequently increasing heart rate to meet metabolic demands (78). Taken together, these findings indicate that specific OXL may actively drive respiratory and cardiovascular adaptations during endotoxemia, aligning with their known roles in modulating inflammation and vascular responses.

Our study is not without limitations. While we quantified temporal changes in OXL profiles and their magnitudes in plasma, their tissues of origin remain unknown. Additionally, the transcriptional, translational, and activation statuses of oxidizing enzymes which drive OXL production were not directly quantified in this study. Moreover, the decision to analyze samples at +2H and +12H may have limited our ability to detect changes in plasma OXL with shorter half-lives or delayed response times. While previous evidence by our group demonstrates that 3 h of LPS exposure alters adipocytes’ synthesis of lipid-based mediators and transcription of key oxidizing enzymes (e.g., COX), this model cannot replicate chronic inflammation in vivo (33). In addition, previous research indicates that mitochondrial degradation of OXL is upregulated during LPS-induced inflammation, framing the need for additional in vivo studies to fully elucidate the dynamic balance between synthesis and degradation of these lipid mediators (79). Although this study has a relatively small sample size, we observed consistent patterns across individuals, and the use of robust statistical methods supports the reliability of our findings. Additional studies with larger cohorts are needed to elucidate mechanistic details about specific OXL pathways. Importantly, as used in the present study, single-infusion models may not capture the full effects of repeat or chronic exposure to LPS on the OXL profile, however, the initial evidence provided herein supports the implementation of a multiple LPS infusion model in cows during negative energy balance for use in future studies.

Our findings demonstrate that LPS infusions induce distinct clinical, physiological, and biochemical responses in dairy cows, recapitulating hallmark inflammatory symptoms of endotoxemia, thereby providing a robust model for studying systemic inflammation. Our results show that dynamic alterations occur in response to endotoxemia, with time-dependent shifts in key FA-derived mediators that suggest they play distinct roles in the initiation, progression, and resolution of inflammation. The observed differential modulation of COX-, LOX-, and CYP-derived metabolites from AA, EPA, DHA, and LA underscores the complex, time-dependent nature of OXL pathways in both promoting and resolving inflammation. While pro-inflammatory AA-derived metabolites persisted throughout our study’s timeframe, the spikes observed in DHA- and EPA-derived pro-resolving mediators suggest an incomplete and, perhaps, impaired resolution of inflammation. Additionally, our results underscore the potential role of AT in OXL production during endotoxemia, with adipocytes serving as both a source and modulator of inflammatory mediators. The distinct responses observed between in vivo plasma and in vitro adipocyte models point to a potential protective mechanism within AT that may buffer local inflammation while supporting systemic responses necessary for pathogen clearance. Altogether, these insights into the OXL response to LPS provide valuable directions for future therapeutic strategies aimed at targeting lipid mediator pathways to mitigate systemic inflammation and improve outcomes in human and veterinary patients challenged with endotoxemia.