A New Use for an Old Drug: Carmofur Attenuates Lipopolysaccharide (LPS)-Induced Acute Lung Injury via Inhibition of FAAH and NAAA Activities

Acute lung injury (ALI), characterized by a severe inflammatory process, is a complex syndrome that can lead to multisystem organ failure. Fatty acid amide hydrolase (FAAH) and N-acylethanolamine acid amidase (NAAA) are two potential therapeutic targets for inflammation-related diseases. Herein, we identified carmofur, a 5-fluorouracil-releasing drug and clinically used as a chemotherapeutic agent, as a dual FAAH and NAAA inhibitor. In Raw264.7 macrophages, carmofur effectively reduced the mRNA expression of pro-inflammatory factors, including IL-1β, IL-6, iNOS, and TNF-α, and down-regulated signaling proteins of the nuclear transcription factor κB (NF-κB) pathway. Furthermore, carmofur significantly ameliorated the inflammatory responses and promoted resolution of pulmonary injury in lipopolysaccharide (LPS)-induced ALI mice. The pharmacological effects of carmofur were partially blocked by peroxisome proliferator-activated receptor-α (PPARα) antagonist MK886 and cannabinoid receptor 2 (CB2) antagonist SR144528, indicating that carmofur attenuated LPS-induced ALI in a PPARα- and CB2-dependent mechanism. Our study suggested that carmofur might be a novel therapeutic agent for ALI, and drug repurposing may provide us effective therapeutic strategies for ALI.


INTRODUCTION
Acute lung injury (ALI) or its more severe form acute respiratory distress syndrome (ARDS) is a spectrum of lung diseases characterized by a severe inflammatory process, resulting in severe hypoxemia, hypercapnia, diffuse infiltration, and poor pulmonary compliance. ALI may lead to multisystem organ failure and is still a life-threatening disease in critically ill patients (Rubenfeld et al., 2005). Patients diagnosed with ALI are often treated with mechanical ventilation during the course of illness. Although mechanical ventilation is helpful for resolving life-threatening hypoxia and hypercapnia, clinical and experimental studies have shown that mechanical ventilation, if performed incautiously, will further damage the lungs due to overinflation, varotrauma, as well as cyclic closing and reopening of the alveoli, causing ventilator-associated lung injury (Diaz et al., 2010). Moreover, the mechanism of ventilator-associated lung injury may trigger a pulmonary and systemic inflammatory reaction that may further lead to multiple organ dysfunction or multiple system organ failure. Over the past decades, despite considerable progress has been made in understanding the pathogenesis and pathophysiology of ALI, there is still no effective therapeutic strategy for these diseases.
Trauma, pneumonia, acid aspiration, and sepsis resulting from Gram-negative pathogenic bacteria infection are the most common causes of ALI. During acute pulmonary infection, the Toll-like receptor-4 (TLR-4) on epithelial cells and alveolar macrophages detect lipopolysaccharide (LPS), a major component of the outer cell membrane of bacteria, and elicit innate immune responses that eliminate the invading microorganisms (Hussell and Bell, 2014;Kumari et al., 2015). TLR-4 stimulates epithelial and immune cells to produce chemokines and pro-inflammatory cytokines by activation of the nuclear transcription factor κB (NF-κB) pathway, and recruit neutrophils into the lung (Wheeler and Bernard, 1999;Miyake, 2004). Although the engulfed neutrophils kill bacteria during the immune response, it also produces large amounts of free radicals and reactive oxygen species. Uncontrolled inflammation and excessive accumulation of neutrophils can cause alveolar barrier damage and lung dysfunction, leading to respiratory failure and death (Wheeler and Bernard, 1999;Santus et al., 2018). Therefore, regulation of inflammation is critical in protecting ALI, and anti-inflammatory treatment has been proposed as a promising therapeutic strategy for ALI.
Both FAAH and NAAA inhibitors have shown beneficial actions in pulmonary diseases and are potential therapeutic strategies for ALI. FAAH inhibitor URB937 can attenuate onelung ventilation (OLV)-induced lung injury, with an increase in the arterial oxygenation index and decreases in the lung injury score (Yin et al., 2019). NAAA inhibitor ARN726 and F215 also display pronounced therapeutic effects and reduce the tissue damages by suppressing inflammation during the lung injury (Ribeiro et al., 2015;Li et al., 2018). However, it is extremely challenging to push these compounds towards clinical application. One approach to expediting drug development is to discover novel applications for clinically approved drugs, i.e., drug repurposing. With well-known therapeutic window and adverse effects, the old drugs can rapid enter phase II clinical trials. Therefore, we now direct our attention towards clinical drugs and focus on identifying potent anti-inflammatory agents from old drugs.
Carmofur, an antineoplastic drug that has been clinically used for more than 37 years, is also known as an inhibitor of acid ceramidase (AC) (Taguchi, 1980;Kikuchi and Uekama, 1988;Furukawa et al., 2000). AC, a lysosomal enzyme mainly responsible for ceramide degradation, plays an important role in the control of cancer cell proliferation (Realini et al., 2016). We speculated that AC inhibitor carmofur may suppress NAAA activity, since AC exhibits 33-35% amino acid identity with NAAA. Additionally, carmofur contains a urea group, which is commonly seen in many potent FAAH and NAAA inhibitors (Tuo et al., 2017). Carmofur is likely to act as an FAAH and NAAA inhibitor, exhibiting anti-inflammatory activities by regulating the levels of FAEs. As a proof of concept, we examined the capability of carmofur to inhibit FAAH and NAAA, and its therapeutic effects in the LPS-induced lung injury. Our results show that carmofur is an FAAH and NAAA dual inhibitor, and administration of carmofur exhibited beneficial actions in lung injury through a PPARα-and CB2dependent pathway.

Chemicals
All reagents used in the present study were purchased from Sigma (Shanghai, China), seeking the highest grade commercially available unless otherwise indicated. Carmofur, MK886, and SR144528 were purchased from Cayman Chemical (Michigan, USA).

Real-Time Quantitative PCR
Total RNA was extracted from Raw264.7 cells with TRIzol (Invitrogen) and quantified by spectrophotometer (Beckman Coulter, Shanghai, China). cDNA was synthesized from 1 μg of total RNA by using the ReverTra Ace qPCR RT Kit (TOYOBO, Shanghai, China) following the manufacturer's instructions. Real-time quantitative PCR was performed in a 7300 real-time PCR System (Applied Biosystems, CA, USA) using SYBR Premix Ex Taq GC (Takara, Dalian, China). The amplification steps included denaturation at 95°C for 30 s (1 cycle), annealing at 60°C for 60 s (1 cycle), and extension at 72°C for 60 s (26 cycles). RNA levels were normalized using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a reference gene.

Protein Preparation and Enzymatic Assay
HEK293-rNAAA and HEK293-rFAAH cells were incubated with carmofur of a series of concentrations for 30 min. The cells were harvested, washed with PBS, sonicated in 20 mM Tris-HCl (pH 7.5) containing 0.32 M sucrose, and then centrifuged at 800 × g for 15 min at 4°C. The supernatants were collected and subjected to three freeze-thaw cycles at −80°C, and protein concentrations were measured by the BCA protein assay kit. The crude protein containing FAAH or NAAA was used as such without further purification. FAAH activity was measured by incubating protein (30 μg) with AEA (25 mM) in 0.2 ml Tris-HCl buffer (50 mM, pH 8.0, containing 0.05% fatty acid-free BSA) at 37°C for 30 min. NAAA activity was measured by incubating protein (30 μg) with heptadecenoylethanolamide (C17:1 FAE) (25 mM) in 0.2 ml phosphate buffer (50 mM, pH 5.0, containing 0.1% Triton X-100, and 3 mM DTT) at 37°C for 30 min. The reactions were terminated by adding methanol (0.2 ml, containing 1 nmol heptadecanoic acid as internal standards), and the residual hydrolysis products of the substrate were determined by HPLC-MS. NAAA inhibitor (S)-OOPP and FAAH inhibitor URB597 were used as the positive control.

Lipid Extraction and Analysis
Lipids were extracted from Raw264.7 cells and tissue samples using a previously described method . Samples were homogenized in CH 3 OH-H 2 O (1:1) containing C17:1 FAE as internal standards and then extracted with CHCl 3 . The organic phase was dried under N 2 flow, and the residue was purified by solid-phase extraction using CH 3 OH-CHCl 3 (1:9) as eluent. The eluate was dried and reconstituted in methanol (100 μl), and the CH 3 OH solution (10 μl) was used for LC-MS/MS assay. The mobile phase consisted of CH 3 OH and H 2 O (pH 7.4), and the gradient elution was as follows: 85% CH 3 OH for the first 3 min, followed by a linear gradient from 85% to 100% CH 3 OH for 2 min, and then 100% CH 3 OH for 15 min. The elution condition was finally returned to 85% CH 3 OH at a flow rate of 0.7 ml/min. The column temperature was maintained at 40°C. Ion detection was monitored by APCI + -MRM mode. The molecular ions were measured at m/z 348.00/62.00 for AEA, m/z 300.3/62.0 for PEA, and m/z 313.1/62.0 for C17:1 FAE.

Competition Binding Assay
The binding affinity of carmofur to PPARα was detected by PPARα LanthaScreen™ TR-FRET competitive binding assay (Invitrogen) following the kit instruction. Carmofur (20 μM), GW7647 (positive control, 10 μM), or DMSO (negative control) was incubated with a mixture of the GST-tagged PPARα ligandbinding domain (LBD), terbium-labeled anti-GST antibody, and a green fluorescent pan-PPAR ligand (tracer). The mixture was gently shaken for 30 s and then incubated at 25°C for 5 h. The fluorescent emission intensity at 520 nm (λex = 495 nm) was measured.

In Vivo Biological Evaluation of Carmofur
All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health (NIH) and approved by the Animal Care and Use Committees of Xiamen University in China.

LPS-Induced ALI
The LPS-induced ALI model was used to test the antiinflammatory activity of carmofur . Male C57BL/6J mice (20-22 g) were randomly grouped, with eight animals for each group. Mice were anesthetized with chloral hydrate and instilled intratracheally with LPS (5 mg/kg). Carmofur [3 and 10 mg/kg, dissolved in saline with 5% polyethylene glycol 400 (PEG400) and 5% Tween 80] or its vehicle was orally administered twice a day starting from the day of LPS application. Mice were sacrificed 0, 1, and 3 days after LPS instillation.

Analysis of Bronchoalveolar Lavage Fluid (BALF)
On days 0, 1, and 3, mice were euthanized, and lungs were infused five times with 1 ml of PBS (pH 7.5). BALF was extracted, combined, and then centrifuged at 1500 rpm for 10 min at 4°C. The protein content in the supernatant was measured using a bicinchoninic acid (BCA) kit (Pierce). The amount of IL-1β and TNFα was measured in BALF using the murine IL-1β and TNFα ELISA kit according to the manufacturer's instructions. The pellet was resuspended in PBS, and cell differentials were counted by using cytospin and Wright-Giemsa staining.

Evans Blue Dye Extravasation
Evans blue dye extravasation was measured following a previous reported method . Mice were injected intravenously with Evans blue (40 mg/kg) and were euthanized 60 min later. Lungs were perfused to remove intravascular dye, followed by excision and homogenization in PBS. The lung homogenate was diluted with 2 volumes of formamide and incubated at 60°C for 18 h. After centrifugation at 2,000 × g for 30 min, the supernatant was collected and measured at wavelengths of 620 nm and 740 nm using a Tecan reader. The extravasated Evans blue dye concentrations were then calculated using the following formula: A 620 (corrected) = A 620 (actual) − (1.193 × A 740 ) + 0.007.

Myeloperoxidase (MPO) Activity
MPO activity was measured using a previous reported method . Lung samples were homogenized in a potassium phosphate buffer containing 0.5% hexadecyltrimethylammonium bromide, 0.5 mM EDTA, and then subjected to two freeze-thaw cycles at −20°C followed by centrifugation again at 12,000 × g for 30 min at 4°C. Supernatants were collected and added to MPO assay buffer containing potassium phosphate buffer (50 mM, pH 6), H 2 O 2 (w/v, 0.0005%) and o-dianisidine dihydrochloride (0.168 mg/ml). After 25 min of incubation at 25°C, the samples were read at 460 nm for 3 min. The absorbance change during the 30-to 90-s time course was recorded. MPO activity was then calculated using the following formula: MPO activity (U/g) = ΔA 460 × 13.5/lung wet weight (g).

Histology
The whole lungs of mice were harvested and fixed in paraformaldehyde at 4°C, followed by embedding in paraffin. Sections with a thickness of 5 μm were obtained by a microtome. The slices were then stained with hematoxylin and eosin (H&E) after deparaffinized with xylene. Imaging of stained areas was performed with a light microscope (Nikon, Shanghai, China) using ×4 objectives (Dong et al., 2018).

Immunohistochemistry of Lungs Samples
Immunohistochemical analysis was performed on paraffinembedded lung samples. Heat-induced epitope retrieval (Haier, microwave treatment, 600 W) was performed using prewarmed sodium citrate buffer (10 mM, pH 6.0) for 2 × 15 min. In order to block endogenous peroxidase activity, samples were treated with 3% H 2 O 2 /CH 3 OH for 10 min, permeabilized with 0.1% Triton X-100 in PBS for 20 min, and then washed in PBS. Next, slides were blocked with goat serum for 30 min to prevent non-specific binding, followed by incubation overnight at 4°C in the presence of primary antibodies (IL-1β, Abcam, ab200478, dilution 1:100; IL-6, Cell Signaling Technology, 5216, dilution 1:100; TNF-α, Abcam, ab6671, dilution 1:100) (Ji et al., 2018). Sections were then washed with 0.1 M PBS and incubated in the presence of biotinylated goat anti-rabbit IgG secondary antibody (Vector Lab, dilution 1:400) for 1 h at 37°C, followed by incubation with a Vectastain Elite ABC reagent (Vector Lab, dilution 1:200) for 1 h, and then incubated in 2% nickel intensified diaminobenzidine (DAB) for 10 min. Images were taken with an Olympus microscope at 10× magnification. The number of positively stained cells in the whole cartilage area per specimen in five sequential sections was quantified by Image Pro Plus (Media Cybernetics). To confirm the antibody binding specificity for IL-6, IL-1β, and TNF-α, some sections were also incubated with primary or secondary antibody only as controls.

Pharmacokinetics Study
Carmofur were administered orally to C57BL/6J mice with doses of 3 and 10 mg/kg. Three animals per dose and per time point were used. Mice were sacrificed and lung samples were collected at 0,5,15,30,45,60,120,240,480, and 720 min after drug administration. Lung tissues (20-30 mg) were homogenized by ultrasonication in 0.5% HCl solution (1 ml), followed by extraction with chloroform (3 × 3 ml) and then vortex for 1 min. Chloroform layer was separated by centrifugation at 3,000 × g for 10 min and then transferred to a clean V-bottom glass tube and dried under nitrogen (N 2 ) flow. Chloroform (1 ml) was added to resolve dried spots and solid-phase extraction was eluted by methanol/ chloroform (v/v, 1/9). The elution containing carmofur was dried under N 2 and reconstituted in CH 3 CN for HPLC analysis. Briefly, 20 μl of each sample was injected into a reversed-phase column (Agilent, C18 4.6 × 250 mm, 5 μm particle size). Carmofur was eluted by CH 3 CN/H 2 O (50/50, v/v, 0.25% phosphoric acid) with a flow rate of 1.0 ml min −1 at 40°C and determined by UV detection at 260 nm. Carmofur in the lung homogenate with concentrations ranging from 500 to 20,000 ng/ml was prepared for the external standard curve (Kobari et al., 1978).

Data and Statistical Analysis
Randomization was used to assign mice to different experimental groups, to collect and process data. All histology samples were scored blindly and independently by at least two investigators. All statistical analyses were completed using GraphPad Prism version 5.01. Data are shown as the mean ± SEM. Three or more different groups were analyzed by one-way ANOVA with Dunnett's post hoc multiple comparison tests. For all ANOVA, post hoc tests were run only if F achieved P < 0.05 and there was no significant variance inhomogeneity. P < 0.05 was considered statistically significant.

Carmofur is an FAAH and NAAA Dual Inhibitor
As shown in Figures 1A, B, carmofur exhibited potent inhibitory activities towards rat recombinant FAAH (IC 50 = 0.11 ± 0.025 μM) and NAAA (IC 50 = 0.71 ± 0.054 μM) in FAAH or NAAA overexpressed HEK293 cells. However, 5-fluorouracil (5-FU), the active form of carmofur in chemotherapy, showed no inhibitions towards FAAH and NAAA (Figures 1E, F). The hydrolase activities of the carmofur-rFAAH and carmofur-rNAAA interaction complexes were not recovered 6 h after dialysis, suggesting that carmofur may inhibit FAAH and NAAA in an irreversible pattern (Figures 1C, D). To determine whether carmofur can directly regulate the activities of CB2 and PPARα receptors, we examined its ability to bind to CB2 and PPARα receptors by the receptors' competitive binding assay. The results suggested that carmofur did not bind to the ligand-binding domain (LBD) of PPARα and CB2 even at a high dose of 20 μM (Figures 1G, H). Taken together, these results suggested that carmofur was a potent and irreversible FAAH and NAAA dual inhibitor, and it cannot directly active PPARα and CB2.

Effect of Carmofur on LPS-Induced Macrophage Activation
As macrophages take up most of the LPS in the lungs during pulmonary infection (Coers et al., 2017), we next investigated the influence of carmofur on the levels of FAEs in Raw264.7 cells. Carmofur (10 μM) alone significantly suppressed FAAH and NAAA activities and increased the levels of all three FAEs in Raw 264.7 cells (Figures 2A-C, Figure S1). However, in LPS-treated Raw 264.7 cells, carmofur failed to further elevate the AEA levels ( Figure 2C). We then examined whether carmofur modulated the LPS-induced inflammatory responses in Raw264.7 cells. Real-time quantitative PCR and ELISA analysis demonstrated that LPS treatment increased expression of inflammatory cytokines, including IL-1β, IL-6, iNOS, and TNF-α. The increase in cytokine mRNA and protein expression was dose-dependently suppressed by carmofur (1-10 μM) (Figures 2D-G and S2). Moreover, the anti-inflammatory effect of carmofur on Raw264.7 cells was blocked by PPARα antagonist MK886 (10 μM) or CB2 antagonist SR144528 (10 μM), indicating that PPARα and CB2 pathways are involved in the anti-inflammatory action of carmofur in macrophages (Figures 2D-G and S2).
The NF-κB pathway is an important signaling pathway involved in inflammation response in ALI; therefore, we further explored whether carmofur affected the cascade of signaling proteins related to the NF-κB pathway. As shown in Figure 3, LPS stimulation increased the phosphorylation of NF-κB p65 (p-p65) and IκBα (p-IκBα), and enhanced the nuclear translocation of p65 in Raw264.7 cells. Carmofur dose-dependently downregulated the protein expression levels of p-p65 and p-IκBα, and blocked the nuclear translocation of p65 (3 and 10 μM) ( Figures  3A-C). Additionally, the PPARα antagonist MK886 (10 μM) and CB2 antagonist SR144528 (10 μM) both partly abolished the anti-inflammatory effects of carmofur in Raw264.7 cells ( Figures  3A-C). These results suggested that carmofur may alleviate LPSinduced inflammation through the NF-κB signaling pathway, probably mediated by PPARα and CB2 receptors.

Biological Stability of Carmofur in Mice
In previous studies, carmofur showed poor stability in mouse plasma in vitro ; thus, we further examined its stability in the lung. Carmofur were administered orally to C57BL/6J mice, and the drug concentrations in the lung at different time points were determined by HPLC. As shown in Figure 4A, carmofur reached peak levels of 3,460 ng/g tissue (~13 μM) and 10,368 ng/g tissue (~40 μM) in the lung at 30 min after administration at doses of 3 and 10 mg/kg, respectively, and showed a half-life (T 1/2 ) of nearly 45 min in the lung. Interestingly, 8 h after drug administration, considerable amount of carmofur was detected in the lung with concentrations of 190 ng/g tissue (~0.7 μM) and 980 ng/g tissue (~3.8 μM) at doses of 3 and 10 mg/ kg, respectively. It is possible that carmofur bound to proteins in lungs and thus improved the compound biological half-life.
Furthermore, oral administration of carmofur dose-dependently increased PEA, OEA, and AEA levels in lungs (Figures 4B-D). Taken together, these data indicated that carmofur was relatively stable in the lung and was suitable for systemic administration.

Effect of Carmofur on LPS-Induced ALI in Mice
Encouraged by the in vitro anti-inflammatory activity and stability of carmofur in the lung, we then investigated whether carmofur could attenuate LPS-induced ALI. As shown in Figures 4B-D, LPS induced PEA decrease in lungs but had no  Results are expressed as mean ± SEM, n = 3 replicates per group. mRNA levels are presented relative to the GAPDH mRNA expression. **P < 0.01; ***P < 0.001 vs. Sham. # P < 0.05; ## P < 0.01; ### P < 0.001 vs. LPS + vehicle. & P < 0.05; && P < 0.01 vs. LPS + carmofur (10 μM) by one-way ANOVA test. effect on AEA and OEA levels. Carmofur (10 mg/kg, orally) effectively suppressed FAAH and NAAA activities ( Figure S1) and increased the levels of all three FAEs in lungs in ALI mice, which was in agreement with the in vitro results. Accumulation of protein-rich fluid in alveolar spaces is an important symptom of pulmonary injury; thus, BALFs were collected and the protein concentrations were determined. As shown in Figure 5A, 1 day after LPS instillation, significantly elevated alveolar protein and inflammatory cytokine levels in BALF were observed for both vehicle-treated and carmofur-treated ALI mice, indicating the extravascular protein accumulation in the alveolar space. On day 3, the protein and inflammatory cytokine concentrations in BALF from carmofur-treated ALI mice returned to normal levels, while protein and cytokine abundances of BALF from the vehicle-treated group were still high. Immune cell recruitment to the alveolar spaces is another important symptom of lung inflammation. We also observed increased alveolar neutrophil and lymphocytes in BALF from ALI mice on days 1 and 3, while carmofur-treated ALI mice showed fewer immune cells compared to the vehicle-treated group ( Figure 5B). However, there is no significant difference in the number of macrophage between ALI mice and the sham group ( Figure 5B). The capability of carmofur to preserve the integrity of alveolar was also evaluated by Evans blue extravasation ( Figure 5C). The results showed that on day 3, LPS-treated mice showed a more severe leakage of Evans blue in the lung compared with sham mice. Carmofur markedly prevented the increase in Evans blue extravasation. Additionally, the antagonists MK886 and SR144528 significantly inhibited the effects of carmofur on cell recruitment, BALF protein and cytokines, dye extravasation, and MPO activity. Collectively, these data suggested that carmofur reduced the damages to the tissue by suppressing inflammation during the lung injury in LPS-induced ALI mice.
Prolonged neutrophil accumulation is another cardinal feature of ALI, and clearance of neutrophils is associated with inflammation resolution and predicts a good outcome. Therefore, we further examined the neutrophil accumulation by testing the activity of MPO, a marker for neutrophil abundance. The results showed that ALI mice lungs displayed sustained high levels of MPO activity, which was markedly suppressed by carmofur (10 mg/kg, orally) ( Figure 5D). The healing effect of carmofur on ALI mice was further evaluated by histological analysis. Vehicletreated ALI mice showed infiltration of neutrophils and alveolar edema on day 3 after LPS administration (Figures 5E, F). However, ALI mice treated by carmofur (3 and 10 mg/kg, orally) exhibited thinner alveolar walls and less alveolar edema, and inhibited neutrophil accumulations (Figures 5E, F). The inhibition of carmofur on neutrophil accumulations was partially blocked by the PPARα antagonist MK886 (2 mg/kg, i.p.) or CB2 antagonist SR144528 (2 mg/kg, i.p.) (Figures 5E, F). Additionally, to exclude the possibility that the effects of MK886 observed in the present study were due to inhibition of 5-lipoxygenase activating protein (FLAP), we also tested carmofur in PPAR-α knockout mice. As shown in Figure S3A, carmofur had no significant effect on neutrophil accumulation in lungs of PPAR-α knockout mice exposed to intranasal LPS. Moreover, the anti-inflammatory effects of carmofur were also not blocked by FLAP inhibitor Quiflapon sodium ( Figure S3B). These data further confirmed that PPAR-α mediates the effects of carmofur in ALI mice.
Finally, we studied the effect of carmofur on inflammatory cytokines in vivo by immunohistochemistry. LPS-treated ALI mice displayed persistently high levels of IL-1β, IL-6, and TNF-α in lungs, while carmofur but not vehicle dose-dependently reduced these up-regulated cytokines (Figure 6). Antagonist of CB2 or PPARα receptors blocked the anti-inflammatory effects of carmofur ( Figure 6). Additionally, 5-FU (10 mg/kg, i.p.) had no significant effect on neutrophil accumulation and uncontrolled expression of inflammatory cytokines (data not shown). When combined, these data indicated that carmofur could reduce the damages to the tissue by suppressing inflammation during the lung injury in LPS-induced ALI mice.
failed to show therapeutic effects in humans, and a clinical trial of FAAH inhibitor (BIA 10-2474) was stopped because of its toxicity (Huggins et al., 2012; Report by the Temporary Specialist Scientific Committee (TSSC)). Most of the NAAA inhibitors described in literature are still under preclinical study until now (Bottemanne et al., 2018).
One way to expedite drug development is to discover new uses for approved or investigational drugs, i.e., drug repurposing. There are many successful examples for drug repurposing; e.g., gabapentin, originally being used as anti-epileptics, is now commonly used to treat neuropathic pain (Sharlow, 2016). Some clinically approved anti-inflammatory drugs have shown inhibitory activities towards FAAH (e.g., COX-2 inhibitors carprofen and ibuprofen) or NAAA (e.g., IL-1β antagonist diacerein) (Favia et al., 2012;Patel et al., 2013;Petrosino et al., 2015). Therefore, we switch our attention from newly synthesized FAAH and NAAA inhibitors to approved drugs to rapidly develop therapeutic agents for inflammation-related disease. As a proof of concept, carmofur, an antineoplastic drug, was selected, and the capability to attenuate ALI was evaluated in LPS-induced ALI mice. Carmofur is a known inhibitor of AC, an enzyme that exhibits 33-35% amino acid identity with NAAA; thus, we hypothesized that carmofur may also suppress NAAA activity. Our results demonstrated that carmofur is a potent FAAH and NAAA dual inhibitor. Carmofur effectively increased the levels of PEA, OEA, and AEA in macrophages in culture, reduced the mRNA expression of pro-inflammatory cytokines, and downregulated signaling proteins of the NF-κB pathway. Moreover, carmofur reduced the tissue damages in LPS-induced ALI mice via ameliorating the inflammatory responses and suppressing neutrophils infiltration. The pharmacological effects of carmofur were blocked by CB2 antagonist SR144528 and PPARα antagonist MK886, indicating that carmofur attenuates ALI in a CB2-and PPARα-dependent pattern. Representative immunohistochemical analyses of IL-1β, IL-6, and TNF-α in lung sections of sham mice, and ALI mice treated with vehicle (5% PEG/5% Tween-80 in saline, orally), carmofur (3 and 10 mg/kg, orally), SR144528 (2 mg/kg, i.p.), and MK886 (2 mg/kg, i.p.) sacrificed 3 days after LPS instillation. The images represent at least 75% of whole sections. Original magnification × 40; insets, × 100; scale bar, 250 μm. (B) Semiquantitative analysis of the immunohistochemical IL-1β, IL-6, and TNF-α expression. Results are expressed as mean ± SEM (n = 8 mice per group). ***P < 0.001 vs. sham. # P < 0.05; ## P < 0.01; ### P < 0.001 vs. LPS + vehicle. & P < 0.05; && P < 0.01; &&& P < 0.001 vs. LPS + carmofur (10 μM) by one-way ANOVA test. CB2 receptors, highly expressed in the periphery, mainly in immune cells (e.g., macrophages and mast cells), participate in the release of inflammatory cytokines involved in pulmonary injury (Liu et al., 2014). Previous reports showed that carrageenan and LPS-induced inflammation models exhibited CB2-dependent anti-inflammatory effects when FAAH was knocked out or blocked by inhibitors (Lichtman et al., 2004;Naidu et al., 2010;Booker et al., 2012). AEA is an agonist of CB2 receptor (Zygmunt et al., 1999;Di Marzo et al., 2004). Berdyshev et al. have shown that AEA could reduce neutrophil recruitment in the lung, at least in part, by CB2 receptor-mediated pathway in ALI mice (Berdyshev et al., 1998). Therefore, inhibition of the AEA catabolic enzymes FAAH and NAAA by carmofur significantly increased AEA levels in lungs and attenuated ALI. CB2 antagonist SR144528 partially blocked the pharmacological effects of carmofur, suggesting that CB2 may play an important role in the anti-inflammation effects via increasing AEA levels. In addition to increasing systemic AEA, carmofur also elevates levels of OEA and PEA, which are also mainly degraded by FAAH and NAAA. PEA and OEA exert anti-inflammatory via activating the PPARα receptor . In agreement with a previous report that PEA effectively diminished the level of TNFα in BALF (Berdyshev et al., 1998), carmofur significantly reduced the up-regulated TNFα in ALI mice. The PPARα antagonist MK886 partly abolished the therapeutic effects of carmofur in ALI mice, suggesting that PPARα is also involved in the anti-inflammatory action of carmofur.
It is likely that the beneficial action of carmofur in ALI mice is from AC inhibition, since carmofur is a potent AC inhibitor. However, the available literature allows us to exclude this possibility. It has been reported that AC deficiency leads to chronic lung injury and increased immune cell infiltration and inflammatory responses in mice (Yu et al., 2018). Activation of AC, on the contrary, shows anti-inflammatory and anti-apoptotic effects in host cells treated with periodontal bacteria (Azuma et al., 2018). Although carmofur has shown its potential in ALI treatment, toxicity is still the primary concern. The cytotoxic effects of carmofur are mainly attributed to the release of 5-FU, its active drug in chemotherapy, which suppresses DNA synthesis by blocking thymidylate synthase and inhibits AC that hinders cell proliferation Lai et al., 2017). Although the effective therapeutic dose of carmofur for attenuating ALI is 3-10 mg/kg, which is lower than the dose used in classic chemotherapy (2.85-11 mg/kg for human, or 26-100 mg/kg for mice calculated by body surface area) (Taguchi, 1980), the cytotoxic action of carmofur could not be ignored, since the inhibitory potency of carmofur is nearly 10-to 100-fold more potent on AC than on FAAH or NAAA. Repurposing anti-cancer drugs as therapeutic agents for inflammation-related diseases may not be the most ideal approach; therefore, we are still screening other approved drugs to identify potent anti-inflammatory agents. Nevertheless, our study demonstrated that carmofur significantly attenuates ALI in LPS-induced mice, and we are able to identify potent antiinflammatory agents by drug repurposing.

CONCLUSION
In summary, carmofur, an antineoplastic drug, was identified as a potent FAAH and NAAA dual inhibitor, which was further proposed as a therapeutic agent for ALI. In Raw264.7 cells, carmofur effectively increased levels of PEA and OEA, reduced the mRNA expression of pro-inflammatory factors, and down-regulated signaling proteins involved in the NF-κB pathway. More importantly, carmofur significantly attenuated ALI in LPS-induced ALI mice via alleviating inflammation and neutrophil infiltration. The pharmacological effects of carmofur were blocked by PPARα antagonist MK886 and CB2 antagonist SR144528, indicating that carmofur attenuated ALI in a PPARαand CB2-dependent mechanism. Taken together, current work demonstrated that carmofur is a promising therapeutic agent for ALI, and drug repurposing may help us rapidly identify effective therapeutic agents for inflammation-related diseases.

DATA AVAILABILITY
All datasets generated for this study are included in the manuscript and the supplementary files.

ETHICS STATEMENT
All animal experiments were performed in accordance with Guide and Care and Use of Laboratory Animals from National Institutes of Health (NIH) and approved by the Animal Care and Use Committees of Xiamen University in China.

AUTHOR CONTRIBUTIONS
KW, YX, and ZP performed the experiments and collected the data; KW and YL analyzed data; YQ and YL conceived the experiments, designed the experiments, and wrote the manuscript. All authors read and approved the submitted manuscript.

FUNDING
The present study was financially supported by grants from the Natural Science Foundation of China (816001312017 to KW, 81602974 to YL), the Science Technology project of Fujian Province (2017J01360 and 2018D0015 to KW, 2018J05145 to YL), and the Xiamen Science and Technology Program Project (3502Z20189025 to YX, 3502Z20172029 and 2502ZCQ20171000 to KW).