Lpar3 ^-/- mice were gifts from Professor Jerold Chun (41), Wide-type (WT) littermates were served as controls. All the breeding colony was bred and housed in a specific pathogen-free barrier facility maintained by the State Key Laboratory of Cardiovascular Disease. The genotypes were verified by PCR using primer sets in Supplementary Table 1 . All animal experiments were approved by the Laboratory Animal Management and Use Committee of Fuwai Hospital, Chinese Academy of Medical Sciences (animal application approval number FW-2019-0009). The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996) and the ‘Regulation to the Care and Use of Experimental Animals’ of the Beijing Council on Animal Care (1996).

For LPS-induced sepsis model, 8-week-old male Lpar3 ^-/- and WT littermate mice with an average weight of 20–25 g was injected intraperitoneally with LPS (7 mg/kg, Escherichia coli 055: B5, Sigma-Aldrich, L2880) or an equal volume of phosphate-buffered saline (PBS) as a control. Survival rate was recorded every 4 hours for a 64 hours period after LPS injection.

For Cecal ligation puncture (CLP) sepsis model, 8-week-old male Lpar3 ^-/- and WT littermate with an average body weight of 20-25 g were prepared and removed hair by using hair removal cream one day before the experiment. The mice were anesthetized with isoflurane and operated under isoflurane maintenance. By avoiding blood vessels, the abdominal skin and muscles of the mice were carefully incised (approximately 1 cm), the mouse cecum was slowly clamped out by reaching into the abdominal cavity with sterilized forceps, 50% of the cecum was ligated with 6-0 silk, and a 21G needle was used to puncture near the ligature (two holes). The ligated appendix was carefully returned to the abdominal cavity and the abdominal muscles and skin were sutured separately. After awakening, the mice were placed back in their cages and kept in a 12-hour fast. Sham-operated mice underwent the same procedure without ligation and puncture of the exposed cecum. Survival rate was recorded every 6 hours for a 96 hours period after CLP.

For DNase I-treated mouse model, DNaseI (Roche, 11284932001) was dissolved into a 20 U/g solution with PBS, and injected through the tail vein at 10 μL/g at 1-hour post-LPS-injection.

For (2S)-OMPT treated mouse model, 8-week-old male WT mice were injected intraperitoneally with LPS (15 mg/kg, Escherichia coli 055: B5, Sigma-Aldrich L2880), At 1-hour after LPS injection or CLP, (2S)-OMPT (10 mg/kg, Sigma-Aldrich, 857235P.) was dissolved in 3% (w/v) BSA/PBS, and injected through the tail vein. Survival rate was recorded every 4 hours for a 64 hours period after LPS injection or CLP.

Murray score was used to evaluate lung injury (42). The scale includes 4 items, each item scores 4 points, a score of 0 signified no injury, a score of 1 signified injury in 25% of the lung, a score of 2 signified injury in 50% of the lung, a score of 3 signified injury in 75% of the lung, and a score of 4 signified injury throughout the lung. The higher the score, the more serious the lung injury. The individuals who graded the severity of lung injury were blinded to the genotype.

MPO-DNA complexes were identified using a capture ELISA. As the capturing antibody, 5 μg/mL anti-MPO polyclonal antibody (Merck Millipore; 07-496-I) was coated to 96-well microtiter plates overnight at 4°C. After blocking in incubation buffer, 40 μL of mouse plasma or culture supernatants was added per well in combination with the peroxidase-labelled anti-DNA monoclonal antibody (component NO.2 of cell death detection ELISA kit; Roche, CatNo:11774425001) following the manufacturer’s instructions. After 2 hours of incubation at room temperature (RT) on a shaking device, the samples were washed. 100 μL ATBS (one piece of component NO.7 added to 5 ml component NO.6) was added per well and incubation at 37°C in the dark for 25 minutes, 100 μL ATBS stop buffer (component NO.8) was added per well to terminate the reaction. The absorbance at 405 nm wavelength was measured by the automatic enzyme mark analyzer (Infinite_M2000, TECAN, Switzerland).

Concentration of ds-DNA was measured using the Quant-iT™ PicoGreen^® dsDNA Reagent and Kits (Invitrogen, P7589) according to the manufacturer’s instructions.

LPA levels in mouse plasma were quantified by an enzyme-linked immunosorbent assay (ELISA) kit (USCN Life Science, CEK623Ge), according to the manufacturer’s instructions.

IL-1β, IL-6, IL-8 detection kits customized from Bio-Rad (Bio-Rad, 17001201) were used to determine the cytokine levels in mouse plasma by Luminex technology according to the manufacturer’s instructions.

Lungs from control and LPS-challenged mice were embedded in paraffin, sectioned (5 μm), and mounted on glass slides. After dewaxing, antigen retrieval was performed using citrate buffer and then permeabilized with 0.1% Triton X-100 for 10 minutes. After being blocked by hydrogen peroxide, the slides were immunostained with Anti-Fibrinogen antibody (1:100, Abcam, ab34269), overnight at 4°C and rewarmed at 37°C for 30 minutes. Subsequently, sections were incubated with ready-to-use undiluted secondary antibodies conjugated with HRP for 30 minutes at 37°C. Subsequently, Diaminobenzidine (DAB) staining was used for 5 minutes at room temperature and the nuclei were stained with hematoxylin for 2.5 minutes at room temperature. The stained sections were observed under a light microscope. Images were analyzed using the Image-J software.

The slicing preparation process is described above. After dewaxing, antigen retrieval was performed using citrate buffer and permeabilized with 0.1% Triton X-100 for 10 min, then specimens were blocked with goat serum. The sections were incubated with primary antibodies, specifically anti-citrullinated-histone H3 (1:100; Abcam; ab5103) and Anti-Myeloperoxidase antibody (1:200; Abcam; ab208670), followed by detection with the samples were incubated with Alexa Fluor-594-coupled (1:500, Invitrogen; A-11012) and/or Alexa Fluor-488-coupled (1:500, Invitrogen; A-11034) secondary antibodies overnight at 4°C. Coverslips were mounted with a VectaShield medium containing DAPI to detect DNA. The sections were imaged using a Zeiss inverted fluorescence microscope (AXI0; Zeiss) that was equipped with a Zensoftware or using a laser-scanning confocal microscope (SP8; Leica) with a 20×water immersion objective. Images were analyzed using the Image-Pro Plus 6.0 software (Media Cybernetics, Inc. Rockville, MD, USA).

To determine microcirculatory perfusion during abdomen, mice were removed hair by using hair removal cream one day before the experiment. Microcirculatory perfusion was monitored when the mice were anesthetized and in a state of uniform respiration, abdomen was scanned using laser Doppler flowmetry (Perimed, Sweden) for 5 min.

The prothrombin time (PT) and fibrinogen (FIB) was measured by using SF-400 semi-automatic coagulometer (Beijing Success Technology Inc) according to the manufacturer’s instructions.

Neutrophils were isolated using Neutrophil isolation kit (Solarbio, P9201) according to the manufacturer’s instructions. The purity of the neutrophil preparations (consistently > 95%) was routinely verified with Giemsa staining, and cell viability (>97%) was verified by trypan blue exclusion assay. Monocytes were isolated using a mouse monocyte isolation kit (STEMCELL Technologies, 19861) according to the manufacturer’s instructions.

Isolated neutrophils were determined and counted by hemocytometer and diluted to 10^5/L, SYTOX Green Nucleic Acid Stain (Thermo Fisher Scientific, S7020) was added to isolated neutrophils with final concentration of 5 μM. Take a 96-well plate (black walls and white bottoms) and spread into premixed neutrophils. added 50 μL plasma to the corresponding wells and incubated at 37°C for 4 hours. Measured the fluorescence of each well at 520 nm by a microplate reader and imaged using a Zeiss inverted fluorescence microscope (AXI0; Zeiss) that was equipped with a Zensoftware.

Neutrophil pretreatment procedure is described above. Take a 96-well plate (black walls and white bottoms) and spread into premixed neutrophils. Added 50 μL isolated monocytes premixed with LPS (final concentration 0.5 mg/mL) to the corresponding wells and incubated at 37°C for 8 hours. Measured the fluorescence of each well at 520nm by a microplate reader and imaged using a Zeiss inverted fluorescence microscope (AXI0; Zeiss) that was equipped with a Zensoftware.

The total RNA was extracted from monocytes isolated from septic model by using TRIzol^® RNA isolation kit (Invitrogen; 15596026). The quality of isolated RNA was checked on an agarose gel and quantified using the ultraviolet (UV) spectrophotometry. cDNA was synthesized from 2 μg of total RNA using PrimeScript™ RT Master Mix (Takara, RR036A). Real-time PCR was performed with SYBR Green detection. An ABI Prism 7300 sequence detection system (Applied Biosystems) was used for the PCR cycling reaction, real-time data collection, and analysis. GAPDH was selected as the reference gene. The relative transcript levels were quantified by the 2^-ΔΔCT method. The qRT-PCR primers are listed in Supplementary Table 1 .

Isolated Monocytes were lysed with RIPA lysis buffer (Beyotime Biotechnology, P0013B) containing protease inhibitors (Roche, 04693132001) and phosphatase inhibitors (Roche, 04906845001), and lysed on ice for 30 minutes. The lysate was clarified by centrifugation at 15000 g at 4°C for 15 minutes. Protein concentration was measured by the bicinchoninic acid assay (Beyotime Biotechnology, P0009). Separate equal amounts of protein from all samples by SDS-PAGE and transfer to PVDF membrane (Merck Millipore, ISEQ00010). The membranes were blocked with 5% (w/v) non-fat milk in PBS containing 0.1%(v/v) Tween-20 for 1 hour and incubated at 4°C overnight with primary antibodies, CD14 (1:1000; Abcam; ab221678), MyD88 (1:800; Abcam; ab219413), p65-NFκB (1:1000; Abcam; ab32536), p65-NFκB phospho S536 (1:1000; Abcam; ab76302), GAPDH (1:5000; Sigma-aldrich; G9545). Immunoreactive bands were detected by horseradish peroxidase–labeled secondary antibodies using SuperSignal™ West Pico Plus (ThermoFisher Scientific, 34577). Prestained protein ladders were used to estimate the molecular weights (ThermoFisher Scientific, 26616). Merge image and images of each protein hybridization membrane is shown in Supplementary Figure 3 . Protein band intensity was measured using Image J software.

Statistical analysis was performed using GraphPad Prism 7 Software. To compare two groups, independent sample T-test was used for western blot and immunohistochemistry analysis of human tissue samples. To compare more than two groups, a one-way ANOVA test with Tukey’s multiple comparisons post-hoc test was used. Data are expressed as means ± SEM of triplicate samples drawn from a minimum of three independent experiments, and a value of P<0.05 was considered statistically significant.

To explore the potential role of LPA₃ in sepsis, Lipopolysaccharide (LPS) or an equal volume of phosphate-buffered saline (PBS) as a control was injected intraperitoneally into wide-type (WT) and Lpar3 ^-/- mice. The survival rate was monitored every 4 hours for 64 hours. When mice were administered 7 mg/kg LPS, all Lpar3 ^-/- mice with sepsis died within 32 hours, with a survival rate of 0%. WT mice with sepsis continued to survive until the end of the test period (64 hours) with a 100% survival rate ( Figure 1A ). To further confirm this result, another sepsis model, cecal ligation and puncture (CLP) model were also made and survival rate was recorded every 6 hours for a 96 hours period. CLP mice also showed differences in survival rates, with 50% in the WT mice and only 10% in the Lpar3 ^-/- mice to the end of the test period (96 hours) ( Figure 1B ). This result suggests that LPA₃ deficiency significantly reduces the survival rate of mice with sepsis and that LPA₃ plays an important role in the pathogenesis of sepsis, especially endotoxin-induced sepsis.

To further investigate organ damage in mice with sepsis, organs were collected 12 hours after LPS injection. Hematoxylin and eosin (H&E) staining showed significant hyperplasia of the splenic red marrow in Lpar3 ^-/- mice with sepsis ( Supplementary Figure 1 ), suggesting that LPA₃ deficiency caused a more intense inflammatory response in sepsis. In addition to the spleen, H&E staining revealed increased levels of lung injury in Lpar3 ^-/- mice with sepsis ( Figures 1C, D ), and microthrombi were present in the small vessels of the lungs of Lpar3 ^-/- mice with sepsis ( Figure 1E ). To further verify this assumption, we measured fibrin/fibrinogen content in the lungs. Immunohistochemical staining revealed that LPA₃ deficiency resulted in a significant increase in fibrin/fibrinogen content in the lungs of mice with sepsis ( Figures 1E, F ). Given this information, we hypothesized that LPA₃ deficiency may cause DIC in mice with sepsis. During sepsis, the ongoing activation of coagulation pathways can overwhelm the anticoagulant systems and result in consumptive coagulopathy or DIC. Blood coagulation measurements demonstrated that, compared with that in WT mice with sepsis, Lpar3 ^-/- mice with sepsis had significantly longer plasma prothrombin time (PT) ( Figure 2A ), less fibrinogen levels (FIB) ( Figure 2B ), and more D-dimer content ( Figure 2C ). This data seemed to confirm that the mice had experienced a severe coagulation event, resulting in excessive thrombin consumption. In addition, we observed that Lpar3 ^-/- mice with sepsis had almost no epidermal blood flow ( Supplementary Figures 2A, B ), and microcirculatory perfusion was seriously blocked. These results suggest that LPA₃ plays an important role in the maintenance of normal coagulation during sepsis.

NETs play an important role against infection in septic patients and animal models. However, there is growing evidence that excessive NETs formation during sepsis contributes to organ injury, intravascular thrombosis, and DIC (29, 32–35). Recently, we found that LPA regulates thrombotic stability in vitro by inducing the release of NETs from neutrophils (39). Therefore, we hypothesized that LPA receptors might be involved in the regulation of NETs release. To confirm that LPA₃ affects the production of NETs during sepsis, we isolated plasma from WT and Lpar3 ^-/- mice with sepsis, co-cultured with isolated WT neutrophils separately. The results showed that the plasma of Lpar3 ^-/- mice with sepsis had an enhanced ability to induce NETs compared to that of the WT controls ( Figures 3A, B ). Ds-DNA and MPO-DNA complexes are major indicators of NETs in plasma, detection of dsDNA and MPO-DNA levels in the plasma revealed its increased levels in Lpar3 ^-/- mice when compared to that in the WT controls after LPS injection ( Figures 3C, D ). We also performed immunofluorescence assays for NETs markers in lung tissue and found that the expression of CitH₃ and MPO were much higher in Lpar3 ^-/- mice than in WT controls after LPS injection ( Supplementary Figures 2C–F ). These results imply LPA₃ is associated with the release of NETs during sepsis.

One of the major components of NETs are extracellular DNA networks, making DNaseI an effective NETs scavenger. To explore the role of NETs in the development of sepsis in Lpar3 ^-/- mice, DNaseI (or PBS alone as a control) was injected into LPS-induced septic mice through the tail vein. DNaseI-treated Lpar3 ^-/- mice with sepsis exhibited significantly higher survival rates than the PBS-treated Lpar3 ^-/- mice with sepsis ( Figure 4A ). Similarly, the score of lung injury was significantly reduced compared to that in the PBS-injected Lpar3 ^-/- mice ( Figures 4B, C ). Moreover, there was no difference between Lpar3 ^-/-and WT mice with sepsis in lung injury, PT and FIB after DNaseI treatment ( Figures 4C–E ). Together, these results showed that excessive NETs generation in Lpar3 ^-/- sepsis mice may be the main cause of their aggravated sepsis injury.

The plasma of Lpar3 ^-/- mice with sepsis is more capable of inducing NETs, suggesting the presence of more inducers. Recently, we identified LPA as an inducer of NETs (39). Therefore, we considered whether LPA₃ deficiency would cause changes in the content of LPA and thus induce the production of excess NETs. We measured plasma LPA levels by enzyme-linked immunoassay (ELISA) and found that plasma LPA levels were significantly increased in septic mice, but LPA₃ deficiency did not significantly affect plasma LPA levels in mice with sepsis ( Figure 5A ). As previously mentioned, LPA₃ deficiency caused an increased inflammatory response in sepsis ( Supplementary Figure 1 ); so, we examined the expression of IL-1β, IL-6 and IL-8, which are the major pro-inflammatory cytokines reported to be capable of directly inducing NETs in mouse plasma. The results showed that the plasma levels of IL-6 and IL-8 were higher in Lpar3 ^-/- sepsis mice than in WT mice ( Figures 5B, C ), while IL-1β had no difference ( Figure 5D ). Excess IL-6 and IL-8 could stimulate neutrophils and increase NETs levels. Therefore, these results suggest plasma IL-6 and IL-8, but not plasma LPA and IL-1β, are responsible for the excessive release of NETs in Lpar3 ^-/- mice with sepsis.

Monocytes are one of the main sources of IL-6 and IL-8. Thus, we speculate LPA₃ from monocytes plays a role in the development of sepsis. We co-cultured WT mouse neutrophils with monocytes from Lpar3 ^-/- or WT mice under stimulation with LPS or PBS control. SYTOX-Green staining showed that neutrophils co-cultured with Lpar3 ^-/- monocytes were more likely to release NETs after LPS treatment ( Figures 6A, B ), and also had higher levels of the MPO-DNA in the cell culture supernatant ( Figure 6C ). These results indicated that LPA₃ in monocytes resists the release of NETs during sepsis.

We investigated the effect of LPA₃ on expression of IL-6 and IL-8 in monocytes. We found that LPS treatment increased the RNA expression levels of IL-6 and IL-8 in monocytes isolated from septic model, whereas LPA₃ deficiency further increased the expression levels of IL-6 and IL-8 in monocytes ( Figures 6D, E ). This result indicates that monocyte LPA₃ plays an important role in the elevation of plasma IL-6 and IL-8 levels in LPS-induced Lpar3 ^-/- mice with sepsis.

In monocytes, LPA was found to enhance the ability of LPS to induce CD14 expression (43). For this reason, we investigated whether LPA₃ deficiency affected the level of CD14 on the surface of monocytes in mice with sepsis. qRT-PCR and western blotting revealed that CD14 expression was significantly increased in monocytes from Lpar3 ^-/- mice with sepsis ( Figures 6F–H ). As an LPS receptor, increased expression of CD14 leads to further amplification of LPS signaling.

CD14 binds to Toll-like receptor 4 (TLR4) on the monocyte cytosolic membrane and further activates downstream signaling pathways, such as NFκB and MAPK through the MyD88-dependent pathway or TRIF-dependent pathway, exerting a wide range of biological effects. To elucidate the changes in the molecular mechanisms of monocytes in Lpar3^-/- mice with sepsis, we examined the related signaling pathway proteins. The results revealed that monocyte MyD88 protein and p-P65 NFκB protein expression were significantly increased in Lpar3 ^-/- mice with sepsis ( Figures 6I, J ). These results indicate LPA₃ deficiency may induce excess NETs through the MyD88-NFκB signaling pathway.

The absence of LPA₃ induced excessive NETs and caused increased organ damage and more thrombosis. Therefore, we hypothesized that (2S)-OMPT, an LPA₃ selective agonist, could have therapeutic implications for sepsis and sepsis-related thrombosis. We constructed LPS-induced and CLP-induced sepsis model in WT mice and injected (2S)-OMPT into septic mice via tail vein 1 hour after LPS injection or CLP to investigate its therapeutic effect on sepsis. Specifically, in the LPS-induced model, to reflect the difference in mortality in WT mice, we increased the LPS dose to 15 mg/kg. (2S)-OMPT injection significantly improved the survival rate in both LPS-induced and CLP-induced sepsis ( Figures 7A, B ) and decreased lung injury in LPS-induced septic mice compared with that in the controls ( Figures 7C, D ). Furthermore, (2S)-OMPT injection shortened the PT, increased the FIB content ( Figures 7E, F ), and reduced the degree of microcirculatory prefusion embolism ( Supplementary Figure 4 ) in mice with sepsis. This significantly reduced the risk of abnormal coagulation and sepsis-related thrombosis. As previously mentioned, aggravated septic injury is related to LPA₃ due to excessive NETs induced by increased levels of IL-6 and IL-8. (2S)-OMPT injection decreased the levels of plasma IL-6 and IL-8 ( Figures 7G, H ) and reduced the formation of NETs ( Figures 7I, J ). These results demonstrate (2S)-OMPT has therapeutic implications for sepsis-related tissue damage and thrombosis through inhibition of IL-6 and IL-8 induced NETs formation, possibly contributing to the future clinical treatment of sepsis.