Prognostic Role of Circulating Exosomal miR-425-3p for the Response of NSCLC to Platinum-Based Chemotherapy

Lung cancer, with non–small cell lung cancer (NSCLC) accounting for approximately 80% of all cases, is one of the leading causes of cancer-related death worldwide (1). Platinum-based schemes are a cornerstone for broad-based treatments of patients with NSCLC, but the clinical and biological heterogeneity of this disease leads to variable responses to such chemotherapy and mixed outcomes (2, 3). Tailoring therapy to the individual patient according to specific predictive factors has the potential to improve outcomes in NSCLC. A number of prognostic biomarkers have been identified, including histologic subtypes and gene expression signatures (4–6). Despite these advances, patients within similar prognostic groups still display heterogeneous outcomes, indicating that current prognostic factors used in NSCLC are suboptimal for predicting fitness to therapy. Therefore, the discovery of novel prognostic biomarkers is one unmet clinical need in NSCLC.

Exosomes are small membrane vesicles (30–100 nm) that contain proteins and nucleic acids, such as miRNAs (7). Exosomes are actively secreted by several cell types, including cancer cells, and can be isolated from peripheral blood. Given the chemical stability and reproducible levels of exosomal miRNAs across many individuals, tumor-derived exosomal miRNAs appear to be attractive biomarker candidates (8–10). They predominantly act as translational repressors by binding to the 3′-UTR regions of their target mRNAs, which are deregulated in most cancer types (11). To date, several miRNAs have been described as differentially expressed and relevant to platinum-based chemotherapy in NSCLC (12–14). However, the clinical significance of circulating exosomal miRNAs in NSCLC has not been examined.

Autophagy is a catabolic process that delivers cellular components such as cytosolic protein aggregates and excess or defective organelles for degradation (15). Activation of autophagy has been suggested to facilitate the survival of tumor cells under different environmental stressors and increase resistance to chemotherapy (16). We and other colleagues have shown that the increase in basal autophagy accompanies the development of cisplatin resistance in NSCLC cells and that inhibiting basal autophagy sensitizes NSCLC cells to cisplatin-induced apoptosis (17–19). In addition, previous studies show that exosomal miRNAs are able to regulate autophagy in recipient cells by targeting autophagy-related genes (20–22). These findings have driven us to investigate the prognostically relevant correlation between basal autophagy contributing to treatment responsiveness and miRNAs derived from circulating exosomes in patients with NSCLC. The resulting miRNAs can be used as potential biomarkers not only for improved predictions of the clinical response to platinum-based chemotherapy, but also for timely adjustment in NSCLC therapeutic regimen.

We obtained 170 serum samples from advanced patients diagnosed at the First Affiliated Hospital of Nanjing Medical University and Affiliated Hospital of Jiangsu University between 2011 and 2016. None of the patients received therapy before the collection of serum samples. In addition, we also collected paraffin sections of lung cancer tissues from a total of 203 NSCLC patients who were diagnosed at the First Affiliated Hospital of Nanjing Medical University between 2008 and 2016. Of these, there were 88 advanced patients in the IIIB or IV TNM stage without surgery, and lung cancer tissues were obtained from aspiration and fiberoptic bronchoscopy biopsies for the first pathologic diagnosis. The remaining 115 patients in the IB, II or IIIA TNM stages received radical surgery, and lung cancer tissues were from surgical resection before any other therapy. The histologic diagnosis of tumors was based on the criteria of the World Health Organization, and the TNM stage was determined according to the 2009 criteria. Patients provided written informed consent in accordance with the Declaration of Helsinki. The study was approved by the ethics committee on Human Research of the First Affiliated Hospital of Nanjing Medical University and Affiliated Hospital of Jiangsu University.

All NSCLC patients were uniformly followed and received standard first-line platinum-based chemotherapy, mainly composed of cisplatin treatment, as the initial or adjuvant treatment after surgery. They were followed up at intervals of 2 months, and the median follow-up period was 36 months (range, 20–98 months). Progression-free survival (PFS) was defined as the interval from the start of treatment to the date of disease progression. Disease-free survival (DFS) was defined as the interval from the date of surgery to the date of recurrence. If recurrence was not diagnosed, patients were censored on the date of death or the last follow-up visit.

On the basis of the PFS, all the advanced patients were divided into two groups, including platinum-resistant NSCLC (PFS ≤ 6 months) and platinum-sensitive NSCLC (PFS > 6 months).

Exosomes in serum samples were isolated and purified using the ExoQuick serum exosome precipitation solution (System Biosciences) according to the manufacturer's instructions. Exosomes from cultured cells were isolated and purified by differential centrifugation as described previously (23). For immunoelectron microscopy analysis, exosomes were prepared by mixing them with an equal volume of 4% paraformaldehyde. Samples were deposited on to Formvar/Carbon Coated electron microscopy grids. Sections were labeled with a mAb against CD63 (Santa Cruz Biotechnology) and then with a secondary gold-labeled rabbit anti-mouse antibody (Santa Cruz Biotechnology). For NanoSight analysis, exosomes were injected into the NanoSight NS300 unit (Malvern Instruments). Capture and analysis settings were manually set according to the manufacturer's protocol. Particle size and concentrations were calculated using Nanoparticle Tracking Analysis (NTA) 2.3 software.

Exosomal RNAs were isolated using a miRNeasy Serum/Plasma Kit (Qiagen) according to the manufacturer's instructions. cDNA was synthesized from 500 ng of total RNA using a miScript II Reverse Transcription Kit (Qiagen). Quantification of miRNAs was performed using a miScript SYBR Green PCR Kit (Qiagen) on the Bio-Rad CFX96 Touch Real-Time PCR Detection System (Bio-Rad). miRNA-specific primers were as followed: hsa-miR-9-5p 5′-ACACTCCAGCTGGGTCTTTGGTTATCTAGCT-3′; hsa-miR-146a-5p 5′-ACACACCAGCTGGGTGAGAACTGAATTCCA-3′; hsa-miR-215-5p 5′-ACATCCAGCTGGGATGACCTATGAATTG-3′; hsa-miR-425-3p 5′-GTACTTCCTGGGATCGGGAATGTCGTGT-3′; hsa-miR-1273h-3p 5′-ACACTGCATATGGGCTGCAGACTCGACCTC-3′;hsa-miR-4755-5p 5′-ACACTCCAGCTCTCTTTCCCTTCAGAGCCT-3′. Because of the lack of a consensus housekeeping miRNA for the qRT-PCR analysis of circulating exosomal miRNAs, we calculated the absolute concentrations of six miRNAs from calibration curves developed with corresponding synthetic miRNA oligonucleotides and normalized the miRNA concentration to the exosomal protein content (24). Total cellular RNA was extracted, and quantitative real-time PCR analyses were conducted as reported previously (25). The mRNA level was normalized to GAPDH. The AKT1 mRNA primers were: forward, 5′-AGCGACGTGGCTATTGTGAAG-3′; and reverse, 5′-GCCATCATTCTTGAGGAGGAAGT-3′.

Circulating exosomes were isolated from two serum pool samples of 10 patients with platinum-resistant NSCLC and 10 patients with platinum-sensitive NSCLC, following the procedure described above. The selected serum samples were randomly selected from the 170 patients with NSCLC. Total RNA was extracted from the purified exosomes, and the amount and quality of the small RNA in the total RNA samples were determined using the Agilent Small RNA Kit with an Agilent 2100 Bioanalyzer. Small RNA libraries were prepared and amplified using the NEBNext Small RNA Library Prep Set (New England BioLabs). Amplified libraries were resolved on a 10% polyacrylamide gel for size selection. The 140- to 160-nucleotide bands corresponding to adapter-ligated constructs derived from the 18- to 30-nucleotide RNA fragments were excised and recovered in a DNA elution buffer. The average size distribution of each library was determined using the Agilent High Sensitivity Chip Kit and quantified on an ABI 7900HT Fast RT-PCR instrument using the KAPA Library Quantification Kit (Kapa Biosystems). Each library was adjusted to a final concentration of 2 nmol/L, pooled, and sequenced using an Illumina HiSeq 4000 sequencer (single-read, 50 cycles) at the Center for Cancer Computational Biology at the Dana-Farber Cancer Institute (Boston, MA). The BCL files were demultiplexed using CASAVA 1.8.2 (Illumina) into FASTQ files. Raw sequencing reads were then analyzed using miRDeep2 to quantify known small RNA species.

The NSCLC cell lines were transfected with either miR-425-3p mimics/inhibitors or ATG5 siRNA, all of which were synthesized by GenePharma Co. Ltd., using Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer's protocol. Briefly, the cells were seeded in 6-well plate at a density of 2 × 10⁵ cells per well in the presence of mimics/inhibitors or ATG5 siRNA (100 pmol/L) and transfection reagent for 24 to 48 hours. The resulting cells were used further for cell viability or Western blot analysis. The sequences of miR-425-3p mimics and inhibitors are 5′-AUCGGGAAUGUCGUGUCCGCCC-3′ and 5′-GGGCGGACACGACAUUCCCGAU-3′, respectively. The sequence of the ATG5 siRNA was 5′-GCAACUCUGGAUGGGAUUGTT-3′. The cells stably expressing the shRNA targeting BECN1 were selected as described previously (19).

The human NSCLC cell lines A549, PC-9, SPCA1, H1299, H1650, H1703, H1975, and cisplatin-resistant A549/DDP were purchased from the Shanghai Institute of Cell Biology (Shanghai, China). They were maintained under standard conditions and media as recommended by the manufacturer. Cell proliferation was determined using the MTT assay in triplicate as described previously (19).

The conventional online programs, including TargetScan (http://www.targetscan.org) and FindTar3 (http://bio.sz.tsinghua.edu.cn), were used to predict the targets of hsa-miR-425-3p. The targets were further analyzed and demonstrated by the following biological experiments.

The pMIR-REPORT miRNA Expression Reporter Vector System (Ambion) consists of a firefly luciferase reporter vector and an associated β-gal reporter control plasmid. The 3′UTR sequences of AKT1 were amplified by PCR from CNE genomic DNA and cloned into the HindIII and SpeI sites of the pMIR-REPORT Luciferase vector. Because there were two predicted target sites for hsa-miR-425-3p in the 3′UTR sequences of AKT1, we examined the four vectors that contained (i) the wild-type 3′UTR; (ii and iii) a mutation of one predicted target site by site-directed mutagenesis (MUT1, MUT2); and (iv) mutations of both predicted target sites (MUT1 + 2). For luciferase reporter assays, the wild-type (WT) or mutated (MUT) luciferase plasmids and miR-425-3p mimics or inhibitor were cotransfected into the cells using Lipofectamine 2000. Transfected cells were lysed 24 hours after transfection, and luciferase activities were assayed following the instructions of the Luciferase Assay System (Promega). The activity of the product of the β-galactosidase gene under the control of a constitutive ACTB promoter was used to normalize the transfection efficiency.

Western blots were performed as described previously(19). The anti-LC3B, anti-BECN1, anti-p-MTOR (S2448), anti-p-AKT (S473), anti-p-AKT1 (S473), anti-p-AKT2 (S474), and anti-AKT1 antibodies were purchased from Cell Signaling Technology. The anti-ACTB, anti-AKT2, anti-AKT3, anti-AKT1/2/3 and anti-TSG101 antibodies were purchased from Santa Cruz Biotechnology. The anti-SQSTM1/p62 antibody was from Abcam. The densitometry of the immunoblots was performed with ImageJ software (NIH, Bethesda, MD).

The IHC analyses were performed using a Real Envision Detection kit from the Gene Tech Company (Shanghai, China). Paraffin-embedded sections of lung cancer tissues were stained with antibodies against AKT1 (Cell Signaling Technology), LC3-I (Novus Biologicals), or ATG5 (R&D Systems) according to the manufacturer's instructions. To quantify protein expression, the integral optical density (IOD) was evaluated by measuring the total area of the tumor section using Image-Pro Plus software (Media Cybernetics). For each slice, 3 photos were taken in random fields, and the average IOD was calculated.

All statistical analyses were performed using the GraphPad Prism5.0 software. The significance of the differences between groups was estimated using the Student t test, χ² test, Fisher exact test, Mann–Whitney U test, or one-way ANOVA with Tukey post hoc analysis where appropriate. The PFS and DFS rates were calculated using the Kaplan–Meier method with the log-rank test applied for comparison. Pearson correlation analysis was performed to investigate the correlation between ATG5 protein expression and PFS. Two-sided P values were calculated, and a probability level of 0.05 was chosen for statistical significance.

Given that circulating exosomal miRNAs can represent specific and stable molecular biomarkers in cancer therapy (8, 26, 27), we compared circulating exosomal miRNAs that could be differentially expressed between patients with NSCLC with low or high responsiveness to platinum-based chemotherapy. The collection of serum samples of patients with advanced NSCLC is described in the Materials and Methods section. On the basis of PFS, they were divided into two groups including platinum-resistant NSCLC (n = 76, PFS ≤ 6 months) and platinum-sensitive NSCLC (n = 94, PFS > 6 months), respectively (Supplementary Table S1; refs. 28, 29). After isolation of circulating exosomes from the serum samples, we confirmed the presence of exosomes by transmission electron microscopy and NanoSight and Western blot analyses. The immunogold labeling for CD63 and the exosome sizes (diameter: approximately 100 nmol/L) were confirmed by electron microscopy and NanoSight analysis (Fig. 1A and B). The expression of exosomal markers CD63 and TSG101 was confirmed by Western blot analysis (Fig. 1C).

To define the content of the exosomes in terms of miRNAs, we performed HiSeq deep sequencing from circulating exosomes of two serum pool samples consisting of 10 patients with platinum-resistant NSCLC and 10 patients with platinum-sensitive NSCLC. Eighty miRNAs with differential expression were detected according to prespecified criteria (i.e., with fold change > 2 or < 0.5, P < 0.01) between the two groups (Fig. 1D; Supplementary Table S2). From the platinum-resistant versus platinum-sensitive comparison, we selected the 6 miRNAs with the largest fold changes, including miR-425-3p, miR-1273h, and miR-4755-5p, which were upregulated, and miR-9-5p, miR-146a-5p, and miR-215-5p, which were downregulated. The expression levels of these 6 miRNAs were further confirmed by qRT-PCR in the 170 circulating exosome samples from patients with advanced NSCLC. The expression levels of miR-425-3p and miR-4755-5p were significantly higher, while that of miR-146a-5p were significantly lower in the circulating exosomes from the patients with platinum-resistant NSCLC versus those from the patients with platinum-sensitive NSCLC (Fig. 1E). To avoid detecting miRNAs that were derived from other nontumor cells, we performed comparisons of NSCLC cell–derived exosomal miRNAs. On the basis of the IC₅₀ values, defined as the concentration of cisplatin that cause a 50% loss of cell viability, a variety of human NSCLC cell lines were divided into cisplatin-resistant cells (IC₅₀ > 3 μg/mL) and cisplatin-sensitive cells (IC₅₀ < 1 μg/mL; Supplementary Fig. S1). When the exosomes from NSCLC cell lines were isolated (Supplementary Fig. S2), we found that miR-425-3p expression levels were dramatically higher in cisplatin-resistant H1299, H1975, and A549/DDP cells than in cisplatin-sensitive SPCA1, PC-9, and A549 cells (Fig. 1F). Moreover, circulating exosomal miR-425-3p expression was upregulated in patients with either platinum-resistant or -sensitive NSCLC compared with expression in the healthy donors (Supplementary Fig. S3). Because our prognostic studies of exosomal miRNAs needed to be restricted to those that were critical for NSCLC tumor–derived miRNAs, we focused the following comprehensive prognostic and experimental analyses only on miR-425-3p. Of note, high miR-425-3p expression in circulating exosomes indicated the disease progression (Fig. 1G, P < 0.0001, log-rank test), suggesting a poor prognosis in the patients with NSCLC.

To confirm the impact of miR-425-3p in circulating exosomes on the NSCLC responsiveness to platinum-based chemotherapy mainly composed of cisplatin, we performed gain- and loss-of-function experiments of miR-425-3p in NSCLC cells. Ectopic overexpression of miR-425-3p remarkably decreased the chemosensitivity of cisplatin-sensitive PC-9 and SPCA1 (Fig. 2A). In contrast, the miR-425-3p inhibitor sensitized cisplatin-resistant H1975 and H1299 cells to cisplatin compared with nontargeting control (NC; Fig. 2B).

To delineate the molecular mechanisms underlying the role of miR-425-3p in conferring chemoresistance to cisplatin in NSCLC cells, we performed a comprehensive bioinformatics analysis of putative miR-425-3p target genes. Among several potential interaction partners, the serine/threonine kinase AKT1, which has been shown to be associated with autophagic regulation and chemoresistance (30–32), was selected. When transiently transfected with miR-425-3p mimics, AKT1 showed a reduction of approximately 60% to 70% of mRNA expression levels in all tested cisplatin-sensitive cell lines (Fig. 3A). In contrast, using a miR-425-3p inhibitor, AKT1 mRNA expression significantly increased by approximately 3- to 5-fold compared with the NC controls (Fig. 3B). Next, we generated the luciferase reporter constructs containing the WT or MUT miR-425-3p binding sequences in the 3′UTR of the AKT1 mRNA. Because there were two predicted target sites, four reporter vectors containing WT, MUT1, MUT2, and MUT1 + 2 were constructed (Fig. 3C). A direct interaction between miR-425-3p and the AKT1 mRNA was confirmed by the observation of reduced luciferase activity after 24 hours when WT AKT1 sequence was cotransfected with the miR-425-3p mimics in the three cell lines (Fig. 3D). Conversely, using the AKT1 3′UTR-mutated construct MUT1 or MUT1 + 2, we observed no changes in luciferase activity. However, the reduction of luciferase activity was also found when using MUT2, suggesting that miR-425-3p could directly bind at the position of 101–124 rather than at the position of 783–802 in the AKT1 3′UTR region. Sequence analysis showed that there are no miR-425-3p binding sequences in the 3′UTR of the AKT2 and AKT3 mRNA around these two target sites (Supplementary Fig. S4). Forced miR-425-3p expression markedly downregulated both the AKT1 and p-AKT1 levels but with little to no effect on the AKT2, p-AKT2, AKT3, or AKT1/2/3 (total AKT) protein levels in cisplatin-sensitive PC-9 and SPCA1 cells (Fig. 3E). Inhibition of miR-425-3p with an inhibitor led to the opposite results (Fig. 3F). In addition, cisplatin-sensitive NSCLC cells, including SPCA1, PC-9, and A549, showed relatively higher levels of p-AKT1 and AKT1 than cisplatin-resistant H1299, H1975, and A549/DDP cells (Fig. 3G), supporting the casual correlation between miR-425-3p and AKT1.

To investigate whether AKT1 is a mediator of miR-425-3p that determines chemotherapy responsiveness, we detected AKT1 expression in the lung cancer tissues from patients with advanced NSCLC by IHC staining. Lower levels of AKT1 expression were observed in platinum-resistant NSCLCs than in platinum-sensitive NSCLCs (Fig. 4A and B). More importantly, low levels of AKT1 expression were associated with a poor PFS in patients with NSCLC receiving platinum-based chemotherapy (Fig. 4C). These results support the hypothesis that AKT1 is a key effector of the miR-425-3p that functions as a prognostic biomarker of responsiveness to platinum-based chemotherapy.

Because inhibition of AKT1 can result in activation of autophagy (33, 34), we evaluated the impact of miR-425-3p expression on autophagic activity in NSCLC cells. Figure 5A shows that in cisplatin-sensitive PC-9, SPCA1, and A549 cells, forced miR-425-3p expression enhanced the LC3B-I to LC3B-II conversion and reduced the expression of SQSTM1/p62 that is selectively incorporated into the autophagosome and degraded upon autophagy induction (35). In addition, the expression of ATG5 was increased, while the phosphorylation of mTOR and AKT was decreased compared with the control cells. In contrast, when using a miR-425-3p inhibitor in cisplatin-resistant H1299, H1975, and A549/DDP cells, the expression of these autophagy-related proteins was the opposite (Fig. 5B). Consistently, the formation of GFP-fused LC3 puncta was increased within PC-9 cells transfected with miR-425-3p mimics, while decreased within H1299 cells transfected with the inhibitor (Supplementary Fig. S5). In addition, treatment with the mTOR inhibitor rapamycin reversed the decrease of the LC3B-I to LC3B-II conversion and the increase of phosphorylated mTOR levels induced by the miR-425-3p inhibitor in the three cisplatin-resistant cells (Fig. 5C), supporting that miR-425-3p functions on the upstream of mTOR.

To investigate whether the upregulation of autophagic activity by miR-425-3p correlates highly with the responsiveness to platinum-based chemotherapy, we collected 88 lung cancer tissue samples from patients with advanced NSCLC in the IIIB or IV TNM stages who received standard first-line platinum-based chemotherapy after diagnosis (Supplementary Table S3). IHC analysis showed that lung cancer tissues from platinum-resistant NSCLCs displayed higher levels of basal autophagy, as determined by the decrease in LC3-I and the increase in ATG5 expression, than tissues from platinum-sensitive NSCLCs (Fig. 6A and B). Further analysis revealed an evident association between LC3-I expression and PFS (Pearson correlation coefficient r = 0.2474; P = 0.0290; Fig. 6C), while ATG5 expression was negatively correlated with PFS (Pearson correlation coefficient r = −0.2310; P = 0.0419). These findings suggested that the levels of basal autophagy were involved in the prognosis of patients with advanced NSCLC following platinum-based chemotherapy. Moreover, we divided the samples into low and high expression groups according to the median value of relevant protein expression. Low LC3-I (P = 0.0045, log-rank test) and high ATG5 expression (P = 0.0372, log-rank test) indicated the disease progression (Fig. 6D). Similar results were obtained in 115 patients with NSCLC in the earlier stages (IB, II, or IIIA TNM) who received standard platinum-based chemotherapy as an adjuvant treatment after radical surgery (Fig. 6E). Low LC3-I (P < 0.0001, log-rank test) and high ATG5 expression (P < 0.0001, log-rank test) was significantly associated with poor DFS. These results suggest that basal autophagy levels in lung cancer tissues might predict the response to platinum-based chemotherapy for patients with either early- or advanced stage NSCLC.

Consistently, we found that cisplatin-resistant NSCLC cells, including A549/DDP, H1975, H1703, and H1299 cells, showed a higher LC3B-I-to-LC3B-II conversion than A549 cells (Supplementary Fig. S6). In contrast, comparable or mildly increased levels of LC3B-II expression were observed in cisplatin-sensitive H1650, PC-9, and SPCA1 cells. When autophagy was suppressed via silencing of an autophagy-related gene, namely ATG5 or BECN1, the resulting cisplatin-resistant H1703 and H1299 cells were more sensitive to cisplatin compared with the controls. When cisplatin-sensitive PC-9 and SPCA1 cells were treated with rapamycin, the increase in the LC3B-I to -II conversion was found, and both cells with the pretreatment were more resistant to cisplatin compared with the controls.

The expression profiles of circulating exosomal miRNAs are of particular interest as novel noninvasive diagnostic and prognostic biomarkers in patients with cancer. However, circulating exosomal miRNAs as prognostic biomarkers of therapeutic response is largely unknown in NSCLC. Here, we analyzed the profile of circulating exosomal miRNAs from patients with platinum-resistant and -sensitive NSCLC within advanced tumors by HiSeq deep sequencing. This approach identified 80 miRNAs with differential expression, including 6 miRNAs with the largest fold change. The further validation of 170 circulating exosome samples by qRT-PCR revealed higher expression of miR-425-3p and miR-4755-5p but lower expression of miR-146a-5p in the platinum-resistant patients relative to the platinum-sensitive NSCLC patients. The other miRNAs of this six miRNA panel had no significant differential changes. miR-146a-5p has been previously reported to be involved in the regulation of resensitization of cisplatin-resistant NSCLC cells (36). However, we failed to observe the impact of miR-146a-5p on the response to cisplatin through overexpression or silencing of miR-146a-5p in NSCLC cells, including PC-9, SPCA1, H1299, and H1975 (Supplementary Fig. S7). On the basis of prior studies in NSCLC cells and our restriction to tumor-derived exosomal miRNAs, we finally confirmed the high expression of miR-425-3p in circulating exosomes as a poor prognostic factor for the responsiveness to platinum-based chemotherapy in NSCLC.

To elucidate possible molecular mechanisms that are regulated by miR-425-3p, we performed a comprehensive bioinformatics analysis to predict putative partners of interest. One of them, AKT1, has been previously shown to be involved as a negative regulator of autophagy (33, 34, 37). And AKT1 plays an important role in chemoresistance of various cancers. (31, 32) A recent study has shown that miR-185 induces autophagy by downregulating AKT1 mRNA expression in hepatocellular carcinoma (37). These data prompted us to focus on AKT1, and we demonstrated that miR-425-3p negatively influences the expression levels of AKT1 and directly interacts with one of the predicted miRNA:mRNA binding sites in the 3′UTR region of the AKT1 mRNA. Modifying the interaction site in the 3′UTR AKT1 abolished the inhibitory effect of miR-425-3p. IHC staining exhibited lower AKT1 expression in lung cancer tissues from platinum-resistant patients with NSCLC and confirmed the linkage between low AKT1 expression and poor PFS. Our study suggests that AKT1 is a mediator of miR-425-3p that determines chemotherapy responsiveness.

In the PI3K–AKT–mTOR pathway, AKT1 activates the mTOR complex, exerting an inhibitory role on autophagy (38). Indeed, miR-425-3p can facilitate autophagic activation by targeting AKT1, consequently leading to the resistance of NSCLC cells to cisplatin treatment. When we investigated the association between basal autophagy in lung cancer tissues and clinical outcome in patients with NSCLC treated with platinum-based chemotherapy, reduced LC3-I, and enhanced ATG5 expression were significantly correlated with low responsiveness and linked with poor PFS in 88 patients with advanced NSCLC. Moreover, a similar relationship was also found in 115 patients with NSCLC in the early stages. High levels of basal autophagy were linked to poor DFS. In agreement, many previous studies, and the results presented in Supplementary Fig. S6, have shown that NSCLC cell lines are characterized by different levels of basal autophagy and that suppression of autophagy in NSCLC cells with active basal autophagy sensitizes them to cisplatin (18, 19, 39). These findings reveal the link between active basal autophagy and the poor response to platinum-based chemotherapy in NSCLC.

Regarding the role of miR-425-3p, a previously published study showed its involvement in predicting the response to sorafenib therapy in hepatocellular carcinoma (40). In addition, a high level of plasma miR-425-3p has been described as a diagnostic marker for early-stage lung adenocarcinoma (41). To our knowledge, this is the first study to show the clinical significance of circulating exosomal miR-425-3p as a valid biomarker for predicting the response to platinum-based chemotherapy in patients with NSCLC. On the basis of the possibility of quantifying exosomal miRNAs in patient blood in real time and its function, such a prognostic biomarker might facilitate treatment decisions and timely adjustment in NSCLC therapeutic regiment.

In summary, our study provides the association between circulating exosomal miR-425-3p and clinical responsiveness in patients with NSCLC treated with platinum-based chemotherapy. This noninvasive prognostic factor might be helpful for the design of personalized therapeutic strategies for patients with NSCLC.

No potential conflicts of interest were disclosed.

Conception and design: D. Yuwen, Q. Xu, Y. Shen

Development of methodology: J. Gao, W. Xue

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D. Yuwen, D. Wang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D. Yuwen, Y. Ma, X. Li, M. Fan

Writing, review, and/or revision of the manuscript: D. Yuwen, Q. Xu, Y. Shen, Y. Shu

Study supervision: Y. Shen, Y. Shu

Y. Shu was supported by a grant from the government of Jiangsu Province (Jiangsu Province Clinical Science and Technology Project, grant no. BL2012008). Q. Xu was supported by National Natural Science Foundation of China (grant no. 81330079). Y. Shen was supported by National Natural Science Foundation of China (grant no. 81573446).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.