Male C57BL/6 mice (8–10 weeks) were applied in experiments, which was approved by the Committee of Animal Care and Use of Southeast University (Protocol number 20180106007). Mice (n = 5–7 per group) were administered with lipopolysaccharide (5 μg/g; MCE, USA) via intratracheal injection to create an ARDS model. Agomir, agomir NC (dose: 5 nmol/mouse), antagomir, or antagomir NC of MiR-199a-3p (dose: 50 nmol/mouse) was administered by caudal vein injection and followed by the establishment of the ARDS model 24 h after this administration. Normal saline (NS) was applied as the negative control and the vehicle for LPS administration. After 24 hours, bronchoalveolar lavage fluid (BALF) was acquired. Lung tissue from the mice was collected for HE staining, and the levels of pro-inflammatory cytokines of lung tissue were measured with an enzyme-linked immunosorbent assay (ELISA) kit (Ela Science, People’s Republic of China), which included TNF-α, IL-12, IL-10, and IL-6. We also conducted in vitro experiments to mimic an in vivo environment to preclude the effect of endogenous RNase/DNase on agomir and antagomir in this study ( Figure S1 ).

RAW264.7 cells were acquired from the Cells Resource Center for Biological Science (Shanghai, China). RAW264.7 cells were maintained in a DMEM medium (Wisent Biotechnology, China); 10% fetal bovine serum and 1% penicillin-streptomycin were applied, and we cultured the cells at 37°C in an incubator with 5% CO₂.To assess the influence of lipopolysaccharide-stimulated on activating RAW264.7 cells, two groups of cells were treated with phosphate-balanced solution or LPS (MCE, USA) at different times. To ensure PAK4 inhibition by pretreatment, we add PF3758309 (20 nM; Sigma) into the medium 24 h before the administration of LPS. Mouse AMs were isolated using previously described protocol (Soni et al., 2016).

To assess the cytotoxic influence of lipopolysaccharide on macrophages, we conducted the CCK-8 assay and trypan blue exclusion assay. The cells were seeded into 96-well plates (2000 cells/100 µl), incubated at 37°C overnight, and then treated with 0, 10, 100, or 1000 ng/ml LPS for 24 h. Then, 10 µl CCK8 (Beyotime, China) was added to each well, and the cells were incubated for 4 h. Afterwards, the absorbance of the cells at 450 nm was measured using a microplate reader. We also conducted the trypan blue exclusion assay to evaluate the cytotoxic effect of LPS. After administration with LPS of different concentrations for 24 h, the viability of RAW364.7 cells was evaluated by counting the number of viable cells.

The negative controls (NC), mimics, or inhibitors of MiR-199a-3p (50 nM; Genepharma, Shanghai) were transfected into the RAW264.7 cells based on the manufacturer’s procedure. RAW264.7 cells were seeded into a 6-cm dish at a density of 1 × 10⁵/ml. The transfection was performed by adding PoweverFectTM transfection reagents (SignaGen, USA) with MiR-199a-3p mimic, MiR-199a-3p inhibitor, or NC. The optimal concentration of the transfection reagent was determined by detecting the intensity of green fluorescence generated in the transfected cells. After the transfection, the cells should be washed by phosphate-balanced solution for three times. After 48 h, the medium ought to be replaced, and the efficacy of transfection was detected by RT-qPCR. On the other hand, RAW264.7 cells were also transfected by the RNAi of Rab8a (RiboBio, Guangzhu), pursuant to the manufacturer’s procedure. The efficiency was measured by Western blotting analysis. Moreover, after transfection, we conducted the CCK8 assay to assess the cytotoxic influence of siRNA, mimics, inhibitors, and negative controls ( Figure S2 ).

We collected secretory autophagosomes in accordance with the methods published previously (Xu et al., 2022). In order to remove debris, the cell culture supernatants or bronchoalveolar lavage fluid samples were acquired and centrifuged at 400×g for 5 min at first; then, we collected the supernatants for centrifugation of 12,000×g for 30 min. After that, the supernatants were removed, and then we washed the SAP-containing pellets with phosphate-balanced solution for three times. Subsequently, the pellets were purified with LC3b antibody attained magnetic beads (Miltenyi Biotec; Cell Signaling Technology). Finally, we resuspended the acquired pellets in phosphate-balanced solution and stored them in a −80°C refrigerator for further experiments.

We tested concentrations and sizes of secretory autophagosomes by NTA with ZetaView PMX (Particle Metrix, Germany). First, we diluted the samples with 1× phosphate-balanced solution. Then, NTA measurements were conducted at 11 different positions; 110 nm polystyrene particles were applied for calibration of the ZetaView system. The temperature should be maintained from 23°C to 30°C.

Firstly, we deposited the sample on the formvar-carbon grid and hatched it at room temperature for 5 min. Next, 1% uranyl acetate was added for 5 s. Then, we put three drops of water on the parafilm to wash the grid for three times. When the grid had dried, we applied the transmission electron microscope to observe the sample at 80 kV.

We applied the RIPA lysis buffer to obtain proteins from cell lysates and SAPs. The concentrations of protein were determined by the bicinchoninic acid (BCA) protein assay (Beyotime, China). Protein was separated by 10%–15% SDS-PAGE gel for electrophoresis and transferred to PVDF membranes (Bio-Rad, American). Then, we blocked the membranes with 5% bovine serum albumin (BSA) for 1 h at room temperature and hatched with primary antibodies overnight at 4°C, which included antibodies against F4/80, LC3B, β-actin, Rab8a (1:1000; #6975, #43566,#4970 and #30325, CST), and PAK4 (1:1000, pS474, Absin). Subsequently, the membranes were then hatched with the horseradish peroxidase-conjugated secondary antibody (1:5000; Beyotime, China) for 1 h at 4°C and then the protein bands were visualized with the electrochemiluminescence detection kits (Beyotime, China).

The levels of pro-inflammatory cytokines were quantified by ELISA kit pursuant to the manufacturer’s protocols (Ela Science, China), which included IL-6, IL-12, TNF-α, IL-10, and IL-1β. We applied the microplate reader to acquire the OD450 values of each sample.

We used TRIzol^® (Takara, Japan) to acquire total and quantified RNA with a NanoDrop ND-2100 spectrophotometer (Thermo Scientific). Then, the RNA was reversely transcribed with the Transcriptor First Strand cDNA kit (Ribobio), and U6 was applied as the internal control. Subsequently, we performed the qPCR with specific primers (Ribobio), and the primer sequences included: MiR-199a-3p, 5′-CAGTGTTCAGACTACCTGTTC-3′; U6: 5′-GTGCTCGCTTCGGCAGCACATAT-3′. Afterwards, we conducted qPCR amplification with SYBR Green PCR Master Mix (Ribobio) on a LightCycler 96 system (Roche, Switzerland).

HEK293 cells were used to conduct the luciferase reporter assay. The cells were cotransfected with the reporter plasmids and MiR-199a-3p mimics or MiR-NC plasmids with either the wild-type or mutant PAK4 3′-UTR (Promega, USA) using the Lipofectamine 2000 reagent (Invitrogen, USA). Then, after 24 h, luciferase activity was measured with a luciferase analysis reagent kit, pursuant to the manufacturer’s instructions (Promega).

All experiments in this study were repeated for at least three times. We applied GraphPad Prism 8.0 (GraphPad Software, USA) to conduct data analysis. Results were signified as the mean ± standard deviation (SD) and compared by using Student’s t-test or one-way analysis of variance (ANOVA) with a cutoff of p < 0.05. Kaplan-Meier curves were used to demonstrate survival rates, while survival curves were compared using the log-rank test.

First, we detected the cell viability after LPS administration in order to identify the cytotoxic influence of lipopolysaccharide on RAW264.7. We found that the cell viability of RAW264.7 cells decreased significantly with 100 ng/ml LPS ( Figures 1A , S3 ). Then, we added 100 ng/ml LPS to the medium at different times to evaluate the influence of LPS on activating RAW264.7 cells. The results illustrated that lipopolysaccharide treatment for 24 h could increase the levels of the pro-inflammatory cytokines (IL-1β and TNF-α) of the cell supernatants remarkably ( Figures 1B, C ). Therefore, we considered 100 ng/ml LPS induction for 24 h as an adequate condition to provoke inflammation in RAW264.7 cells. Additionally, we noticed that the expression of MiR-199a-3p in both AMs and RAW264.7 cells increased significantly after LPS administration ( Figures 1D, E ). Furthermore, we quantified MiR-199a-3p expression in the BALF of mice with LPS-induced ARDS. Our results pointed out that, compared with control group, the MiR-199a-3p expression level in the BALF significantly increased ( Figure 1F ). Therefore, MiR-199a-3p expression increased after LPS administration, both in vitro and in vivo.

To explore the impact of MiR-199a-3p in mice with ARDS, we injected the MiR-199a-3p agomir or antagomir into mice via the caudal vein before the establishment of the ARDS model by intratracheal LPS administration. The MiR-199a-3p expression in the BALF of mice with agomir increased significantly ( Figure 2A ). Compared to mice of control group, over-expression of MiR-199a-3p aggravated the pathological injury of lung tissues ( Figures 2B, C ). The results of ELISA demonstrated that proinflammatory cytokine (TNF-α and IL-6) levels increased after agomir administration, while the levels of anti-inflammatory cytokines (IL-10 and IL-12) decreased significantly ( Figure 2D ). In addition, compared with the NC groups, the mortality of mice injected with agomir increased remarkably ( Figure 2E ). Thus, MiR-199a-3p over-expression probably aggravates LPS-stimulated ARDS in mice.

To investigate the mechanism of lung injury exacerbation with MiR-199a-3p overexpression, we isolated vesicles from the BALF of mice and verified the double-membrane structure with TEM ( Figures 3A , S4 ). We also conducted NTA to assess the size of vesicles from BALF, which was consistent with that of autophagosomes ( Figure S5 ). The marker of autophagosome LC3 and the marker of macrophages F4/80 was detected by Western blotting ( Figure 3B ), and this indicated that the vesicles isolated from BALF could be considered as SAPs, which were at least partially released from macrophages. As the proinflammatory cytokine IL-1β was the vital molecule inside SAPs that caused lung injury in mice (Soni et al., 2016), we applied the ELISA kit to test the level of IL-1β in secretory autophagosomes from BALF after breaking the SAP membrane with Triton X-100. These results indicated much more IL-1β in secretory autophagosomes from the BALF of LPS-induced ARDS mice than those in the control group ( Figure 3C ). Following the injection of agomir, agomir-NC, antagomir, and antagomir-NC of MiR-199a-3p, the results of NTA revealed that the quantity of particles increased significantly in bronchoalveolar lavage fluid obtained from mice in ARDS with MiR-199a-3p overexpression against that in the NC group ( Figure 3D ). The results of Western blotting illustrated that LC3 was much more sufficient in the BALF from mice with MiR-199a-3p overexpression ( Figure 3E ). Furthermore, IL-1β levels in BALF noticeably increased with MiR-199a-3p over-expression, especially for SAPs in BALF ( Figure 3F ). Therefore, MiR-199a-3p over-expression markedly promotes SAP release in mice with ARDS. Therefore, MiR-199a-3p over-expression markedly promotes SAP release in mice with ARDS.

To explore the influence of MiR-199a-3p on the SAP release from LPS-provoked RAW264.7 cells, cells were transfected with the MiR-199a-3p mimic or inhibitor before LPS administration, and the efficiency was determined by fluorescence microscopy and qRT-PCR ( Figures 4A, B ). The results of NTA demonstrated that the quantity of vesicles increased prominently in the supernatant of RAW264.7 cells with MiR-199a-3p overexpression compared with that in the NC group ( Figure 4C ). Western blotting analysis revealed that the level of LC3 was much more sufficient in the RAW264.7 cell supernatant with MiR-199a-3p overexpression ( Figure 4D ). Additionally, IL-1β levels in SAPs markedly increased with MiR-199a-3p mimic transfection ( Figure 4E ). Therefore, SAP secretion from macrophages could be induced by MiR-199a-3p over-expression.

To ascertain the relevant signaling pathways involved in MiR-199a-3p-regulated SAP secretion, Rab8a pathways were investigated because of the vital role in SAP secretion, as determined in our previous study (Soni et al., 2016). The expression levels of Rab8a increased after the transfection of the MiR-199a-3p mimic in RAW264.7 cells ( Figure 5A ). In addition, Rab8a in RAW264.7 cells transfected with the MiR-199a-3p mimic or inhibitor was knocked down by siRNA, and the efficiency was validated by Western blot analysis ( Figure 5B ). Both NTA and Western blot analysis demonstrated that Rab8a silencing significantly inhibited the secretion of SAPs ( Figures 5C, D ). Notably, Rab8a silencing reversed the promoted influence of the MiR-199a-3p mimics on SAP secretion in LPS-stimulated macrophages. Therefore, these results imply that MiR-199a-3p mimics stimulate LPS-induced SAP secretion via the activation of Rab8a in vitro.

Rab8a is a small guanosine triphosphatase (GTPase) that belongs to the Ras superfamily, which could be regulated by PAK4. We used TargetScan to recognize latent target genes of MiR-199a-3p. According to assumed target sequences at positions 136–155 of the PAK4 3′-UTR ( Figure 6A ), PAK4 was considered as a latent target of MiR-199a-3p. Western blot analysis showed that MiR-199a-3p mimics suppressed the expression of PAK4 in RAW264.7 cells ( Figure 6B ). Next, a luciferase reporter analysis showed that MiR-199a-3p repressed the luciferase activity of cells transfected with wild-type PAK4 3′-UTRs, but did not change the luciferase activity of cells transfected with mutated PAK4 3′-UTRs ( Figure 6C ). These results confirm that PAK4 is a direct target for MiR-199a-3p. To clarify the involvement of PAK4 in SAP secretion, we inhibited PAK4 expression by using the PAK4 inhibitor PF-3758309 in RAW264.7 cells transfected with MiR-199a-3p mimic or inhibitor. The results demonstrated that PAK4 inhibition with PF-3758309 in RAW264.7 cells transfected with the MiR-199a-3p mimic or inhibitor significantly promoted SAP secretion. Additionally, MiR-199a-3p mimics, combined with PAK4 inhibition, synergistically enhanced SAP secretion, compared with other groups ( Figures 6D, E ). Furthermore, MiR-199a-3p inhibitors markedly suppressed Rab8a expression in LPS-induced macrophages, yet failed to suppress Rab8a expression after PAK4 inhibition ( Figure 6F ). These findings indicate that MiR-199a-3p over-expression activates Rab8a by directly decreasing PAK4 expression.