Non-small cell lung carcinoma cell lines (SK-MES-1, A549, and NCI-H1650), BEAS-2B and THP-1 cells were purchased from ATCC (Manassas, VA, United States) and maintained in DMEM (Thermo Fisher Scientific, Waltham, MA, United States) containing 10% FBS, 1% streptomycin and penicillin (Thermo Fisher Scientific) in an incubator (37°C, 5% CO₂).

IL-4 and IL-13 were obtained from MedChemExpress (MCE, Monmouth Junction, NJ, United States).

The Cancer Genome Atlas Program (TCGA) dataset¹ was used to explore the differentially expressed miRNAs between adjacent normal tissues and NSCLC tissues. The analysis was performed according to the prognosis information of lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) in TCGA. Meanwhile, LUAD and LUSC are the two main subtypes of NSCLC. In addition, TCGA dataset was downloaded and R analysis was performed as previously described (Colaprico et al., 2016).

Non-small cell lung carcinoma cells were transfected with miR-770 agomir or agomir-ctrl (NC; GenePharma, Shanghai, China) by using Lipofectamine 2000 (Thermo Fischer Scientific).

After 48 h of culture, NSCLC cell supernatants were collected by centrifugation (300 × g for 15 min, 2000 × g for 15 min, and 10,000 × g for 30 min). Subsequently, cell supernatants were filtrated and collected to isolate exosomes via ultracentrifugation (120,000 × g for 70 min). The particle sizes of exosomes were investigated by Particle Metrix (PMX), the structure of exosomes was observed by transmission electron microscopy (TEM) and the exosome markers were detected by western blot (Madhyastha et al., 2021; Xing et al., 2021).

Macrophages were treated with PMA (100 ng/ml; Peprotech), and then treated with PBS, A549-Exo-NC, A549-Exo-miR-770 agomir or 20 ng/ml IL4/IL-13 for 24 h. Subsequently, the macrophages surface maker CD206 (BD Biosciences, New Jersey, United States) and CD86 (BD Biosciences) were analyzed by flow cytometry (BD Biosciences).

TRIzol^® reagent (Takara, Tokyo, Japan) Was used to isolate total RNA From cell lines or tissues. the PrimeScript RT reagent kit (ELK bioscience, Wuhan, China) Was used to reverse transcribe total RNA Into cDNA. Subsequently, the SYBR premix Ex Taq II kit (ELK bioscience) was used in RT-qPCR. the protocol Was as follows: 2 min at 94°C, followed by 35 cycles (30 s at 94°C and 45 s at 55°C). the primer sequences Were listed as follows: miR-770 forward, 5′-CTCGCTTCGGCAGCACAT-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′; Arginase-1 forward 5′-AGACCACAGTTTGGCAATTGG-3′ and reverse, 5′-AGGAGAATCCTGGCACATCG-3′; iNOS forward, 5′-GCAG GACTCACAGCCTTTGG-3′ and reverse, 5′-GGCTGGATGT CGGACTTTGT-3′; β-actin forward, 5′-GTCCACCGCAAATG CTTCTA-3′ and reverse, 5′-TGCTGTCACCTTCACCGTTC-3′; MAP3K1 forward 5′-AGCCACGAGTTGTCAAGTCCT-3′ and reverse, 5′-AGCAGGAGGGATTTGCTGAG-3′; And U6 forward, 5′-GTCCACCGCAAATGCTTCTA-3′ and reverse, 5′-TGCTGTCACCTTCACCGTTC-3′. 2^{–Δ Δ t} method Was used for quantification. β-actin or U6 Was used as a loading control.

Non-small cell lung carcinoma cells (5 × 10³ cells/well) were seeded overnight. Then, cells were treated with agomir-ctrl or miR-770 agomir at 37°C for 0, 24, 48 or 72 h. Subsequently, cells were added with CCK-8 reagents (10 μl; Beyotime) at 37°C for 2 h. The absorbance (450 nm) was measured by a microplate reader.

Non-small cell lung carcinoma cells (5 × 10⁴) were centrifuged (956 g, 5 min) and then the residue was resuspended. Subsequently, cells were treated with 5 μl Annexin V-FITC and propidium iodide (PI) in the dark at 4°C for 15 min. Flow cytometer (BD) was used to analyze the cell apoptosis. The data were quantified by FlowJo (BD).

RIPA lysis buffer (Beyotime) was used to isolate total protein from cells. BCA kit (Beyotime) was used to quantify total protein. SDS-PAGE (10%) was used to separate protein (40 μg per lane), and then proteins were transferred onto PVDF membranes (Thermo Fisher Scientific). After blocked with 5% skimmed milk for 1 h, membranes were incubated with primary antibodies overnight as follows: anti-CD63 (Abcam; 1:1,000), anti-TSG101 (Abcam; 1:1,000), anti-CD9 (Abcam; 1:1,000), anti-CD81 (Abcam; 1:1,000), anti-MAP3K1 (Abcam; 1:1,000), anti-p-JNK (Abcam; 1:1,000), anti-JNK (Abcam; 1:1,000), anti-E-cadherin (Abcam; 1:1,000), anti-N-cadherin, anti-vimentin (Abcam; 1:1,000), anti-ERK1/2 (Abcam; 1:1,000), anti-p-ERK1/2 (Abcam; 1:1,000) and anti-β-actin (Abcam; 1:1,000). After that, the membranes were incubated with secondary antibodies (HRP-conjugated, Abcam; 1:5,000) for 1 h. ECL kit (Thermo Fisher Scientific) was used to visualize protein bands. β-actin was used for normalization.

Non-small cell lung carcinoma cells were fixed in methanol for 20 min. Then, cells were incubated with anti-Ki67 (1:1,000; Abcam), anti-CD63 (1:1,000; Abcam) or anti-PKH26 (1:1,000; Abcam). After that, cells were incubated with the secondary antibody (1:5,000; goat anti-rabbit IgG, Abcam). Finally, the result was observed by a microscope (Olympus, Tokyo, Japan).

MAP3K1 containing the putative binding sites of miR-770 was obtained from Genepharma and cloned into the vectors (pGL6; Beyotime) for establishment of reporter vectors MAP3K1 (WT/MT). MAP3K1 (WT/MT) was transfected into NSCLC cells together with vector-control (NC) or miR-770 agomir using Lipofectamine 2000 (Thermo Fisher Scientific). The result was analyzed by the Dual-Glo Luciferase Assay System (Promega).

The upper chamber was pre-treated with 100 μl Matrigel (the migration assay did not include the Matrigel). Subsequently, NSCLC cells (1.0 × 10⁶) were plated into the upper chamber in RPMI1640 medium containing 1% FBS. Meanwhile, the lower chamber was treated with RPMI1640 medium containing 10% FBS. After 24 h of incubation, cells in the lower chamber were stained with 0.1% crystal violet, and then the result was observed and counted by a light microscope.

For the RNA pulldown assay, the Biotin RNA Labeling Mix (Roche, Basel, Switzerland) was used to transcribe and label probe-control or probe-miR-770 in vitro. An RNA structure buffer (Thermo, MA, United States) was used to induce secondary structure formation from the biotin-labeled RNAs. The biotinylated miR-770 and negative control (bio-NC) were generated via GenePharma and coated to streptavidin-conjugated magnetic beads. Cells were lysed and then incubated with the magnetic beads for 6 h. The RNA on the beads was isolated and the enrichment level of MAP3K1 was detected by PCR.

The levels of IL-10 and TGF-β in supernatants of NSCLC cells were assessed by ELISA kit [Multisciences (Lianke) Biotech, Hangzhou, China].

Tissues of mice were fixed overnight, and then cut into thick sections (5 μm). The sections were deparaffinized and rehydrated. For antigen retrieval, sections were heated in a microwave (with sodium citrate buffer). Subsequently, samples were washed with PBS for 5 min. Then, samples were incubated in 3% H₂O₂ for 25 min, washed and incubated in serum for 30 min. After that, the samples were incubated with anti-Ki-67 (Abcam) or anti-CD206 (Abcam) at 4°C, followed with secondary antibody (HRP-labeled; Abcam) for 30 min at 37°C. Finally, diaminobenzidine (DAB) was added and the tissues were observed under a microscope. All the antibodies were obtained from Abcam.

BALB/c nude mice (n = 24; 6–8 weeks old) were obtained from Vital River (Beijing, China). The protocols for animal care and use of laboratory animals were in accordance with ethical committee of Peking University Shenzhen Hospital. A549 cells co-cultured with macrophages, macrophages-exo, macrophages^exo–NC or macrophages^{exo–miR–770 agomir} were subcutaneously transplanted in mice. The tumor volume was investigated once a week as follow: length × width × width. In the end, mice were sacrificed for tumor tissue collection. All in vivo experiments were performed in accordance with National Institutes of Health guide for the care and use of laboratory animals.

Briefly, paraffin sections were washed, permeabilized, and then incubated with 50 μl TUNEL reaction mixtures in a wet box for 60 min at 37°C in the dark. For signal conversion, slides were incubated with 50 μl of peroxidase (POD) for 30 min at 37°C, rinsed with PBS, and then incubated with 50 μl diaminobenzidine (DAB) substrate solution for 10 min at 25°C. Finally, the tissues were observed under an optical microscope.

Data are presented as the mean ± standard deviation. CCK8 assay was performed in quintuplicate. Flow cytometry, western blot, RT-qPCR and ELISA were repeated in triplicate. The other experiments were performed in three times. In addition, Comparisons between two groups were analyzed by the unpaired student’s t-test. All other experiments were repeated three times. One-way analysis of variance and Tukey’s post hoc tests were used for comparisons between ≥ 3 groups. P < 0.05 was considered to indicate a statistically significant difference.

To detect differentially expressed miRNAs in NSCLC, bioinformatics analysis was performed. As illustrated in Figures 1A,B, the differentially expressed miRNAs in NSCLS tissues compared to adjacent normal tissues in TCGA dataset were presented as volcano plots. Overlap among these datasets was indicated by the Venn diagram in Figure 1C. Among these differentially expressed miRNAs in TCGA, 23 were commonly downregulated, whereas 24 were commonly upregulated. Moreover, among the overlap of these differentially expressed miRNAs, miR-770 was closely associated with the tumorigenesis of NSCLC (Zhang et al., 2017). Therefore, miR-770 was selected for further analysis.

In order to investigate the role of miR-770 in NSCLC cells, RT-qPCR was performed. As revealed in Figure 2A, the expression of miR-770 was significantly downregulated in A549, SK-MES-1 and NCI-H1650 cells, compared with that in BEAS-2B cells. Meanwhile, the level of miR-770 was markedly upregulated in A549 and SK-MES-1 cells transfected with miR-770 agomir (Figure 2B). In addition, miR-770 overexpression notably decreased the viability and proliferation of NSCLC cells (Figures 2C,D, Supplementary Figures 1A,B). Moreover, miR-770 agomir significantly induced the apoptosis of NSCLC cells (Figure 2E and Supplementary Figure 1C). Taken together, miR-770 agomir could inhibit the proliferation of NSCLC cells via inducing apoptosis.

In order to detect the effect of miR-770 on NSCLC cell migration and invasion, transwell assays were used. As revealed in Figure 3A, overexpression of miR-770 obviously decreased the migration of NSCLC cells. Consistently, the cell invasion was notably inhibited in the presence of miR-770 agomir (Figure 3B). In summary, miR-770 agomir could inhibit the migration and invasion of NSCLC cells.

It has been reported that tumor-derived exosomes play important roles in tumorigenesis (Rezaei et al., 2021; Xie et al., 2021). Thus, we sought to isolate the exosomes from A549 cells, and then exosomes isolation was detected by TEM. As shown in Figures 4A,B, typical rounded particles ranging from 30 to 150 nm in diameter was observed by TEM, and Nanoparticle Tracking Analysis (NTA) exhibited a similar size distribution of exosomes. Moreover, the expressions of exosomal proteins (CD63, TSG101, CD9 and CD81) were significantly higher in exosomes derived from NSCLC or BEAS-2B cells (A549-exo and BEAS-2-exo), compared with A549 or BEAS-2B cells (Figure 4C). Meanwhile, the expression of miR-770 in A549-exo was notably downregulated, compared with that in BEAS-2-exo (Figure 4D), and exosomes secreted from A549 cells that were transfected with miR-770 agomir (A549/miR-770 agomir-exo) were approximately 100 nm in diameter (Figure 4E). Furthermore, the level of miR-770 was upregulated in A549/miR-770 agomir and A549/miR-770 agomir-exo compared to that in A549 cells transfected with agomir-ctrl (A549/agomir-ctrl) and exosomes derived from A549/agomir-ctrl (A549/agomir-ctrl-exo) (Figure 4F).

To investigate whether exosomes can mediate cell-to-cell crosstalk by transmitting miR-770 between A549 cells and macrophages (PMA-differentiated human THP-1 monocytes), A549 cell derived PKH26-labeled exosomes were co-cultured with macrophages. After 48 h of co-culture, PKH26 lipid dye were observed in macrophages, indicating that A549-exo could be transferred to macrophages (Figure 4G). In addition, macrophages were incubated with CD63-labeled exosomes derived from A549 cells that were transfected with Cy3-labeled miR-770 agomir. Both CD63 lipid dye and Cy3 fluorescence were observed in macrophages (Figure 4H). These results indicated that miR-770 is contained in A549-exo and can be transferred to macrophages. Meanwhile, the level of miR-770 in conditioned medium (CM) of macrophages/miR-770 was limitedly affected by RNase A but obviously inhibited when treated with RNase A plus Triton X-100 (Figure 4I). Besides, A549/miR-770 agomir-exo significantly increased the level of miR-770 in macrophages (Figure 4J). Altogether, miR-770 can be transferred from A549 cells to macrophages cells via exosomes.

In order to detect the role of A549/miR-770 agomir-exo in the polarization of M2 phenotype macrophages, macrophages were treated with 50 μg/ml exosomes derived from A549 cells (IL-4/IL-13 treatment as a positive control), and then flow cytometry was performed. As revealed in Figure 5A, macrophages exhibited a CD86^low/CD206^high phenotype, when cells were co-cultured with A549/agomir-ctrl-exo; however, when macrophages were incubated with A549/miR-770 agomir-exo, cells exhibited a CD86^high/CD206^low phenotype (Figure 5A). In addition, macrophages treated with A549/agomir-ctrl-exo expressed notably more M2 representative cytokines Arginase-1, IL-10 and TGF-β, but there were no significant changes of M1 representative cytokine iNOS. In contrast, macrophages incubated with A549/miR-770 agomir-exo showed markedly reduced level of Arginase-1, IL-10 and TGF-β compared to those incubated with A549/agomir-ctrl-exo (Figures 5B,C). These results suggested that exosomes with upregulated miR-770 could inhibit the polarization of M2 macrophages.

In order to investigate whether exosomal miR-770 could inhibit the migration and invasion of NSCLC cells by inhibiting M2 macrophages polarization, transwell assay was performed. As indicated in Figure 6A, A549 cells after co-culturing with M2 macrophages induced by A549/agomir-NC-exo exhibited increased cell migration and invasion abilities; however, this phenomenon was notably reversed by exosomes containing miR-770 agomir. Meanwhile, exosomes with upregulated miR-770 significantly reversed the effect of macrophages incubated with A549/agomir-NC-exo on vimentin, N-cadherin and E-cadherin levels (Figure 6B). Taken together, exosomal miR-770 could inhibit the migration, invasion and EMT process of NSCLC cells via suppressing the polarization of M2 macrophages.

To find the target mRNA of miR-770, targetscan (version 7.2²) was used. As indicated in Figure 7A, MAP3K1 might be the potential target of miR-770, and the relative luciferase activity of reporter vectors containing WT-MAP3K1 was significantly inhibited by miR-770 agomir (Figure 7B). However, miR-770 agomir had very limited effect on luciferase activity of reporter vectors containing MT-MAP3K1 (Figure 7B). Meanwhile, the level of MAP3K1 in macrophages was significantly downregulated by miR-770 overexpression (Figure 7C), and the enrichment of MAP3K1 was significantly upregulated by miR-770-probe, compared with probe-control (Figure 7D). In addition, the levels of MAP3K1, p-JNK, and p-ERK1/2 were significantly increased in macrophages incubated with A549/agomir-ctrl-exo; however, this phenomenon was reversed when macrophages were incubated with A549/miR-770 agomir-exo (Figure 7E). Furthermore, macrophages co-treated with A549/miR-770 agomir-exo and MAP3K1-OE notably increased the viability, migration and invasion of A549 cells, compared to macrophages treated with A549/miR-770 agomir-exo (Figures 7F,G). Taken together, exosomal miR-770 could inhibit the polarization of M2 macrophages via inhibition of MAPK signaling.

To investigate the function of exosomal miR-770 in NSCLC in vivo, xenograft mice model was established. As revealed in Figures 8A–C, A549-exo-treated macrophages significantly increased the tumor volume and tumor weight of mice, while macrophages treated with A549/miR-770 agomir-exo significantly reversed these phenomena. Meanwhile, the positive rate of Ki67 and CD206 staining in tumor tissues was notably downregulated in A549 + M^{Exo–miR–770 agomir} group, compared with the A549 + M^Exo group (Figure 8D). To sum up, exosomal miR-770 could inhibit the tumor growth of NSCLC in vivo via suppressing the M2 polarization of macrophages.

In order to further investigate the effect of exosomal miR-770 on tumor growth of NSCLC, TUNEL staining was performed. The data showed that the positive rate of TUNEL staining was significantly increased in macrophages treated with A549/miR-770 agomir-exo, compared with the A549 + M^Exo group (Figure 9A). Consistently, the level of miR-770 was notably upregulated in macrophages treated with A549/miR-770 agomir-exo, compared with the A549 + M^Exo group (Figure 9B). Meanwhile, A549-exo-treated macrophages significantly increased the level of MAP3K1 in tumor tissues of mice, while this phenomenon was significantly reversed by macrophages treated with A549/miR-770 agomir-exo (Figure 9B). In summary, Exosomal miR-770 significantly induced the apoptosis in tumor tissues of mice.