Tumor-derived exosomal miR-155-5p and miR-658 in serum as diagnostic biomarkers for early-stage non-small cell lung cancer

Exosomal microRNAs (ExmiRNAs) provided a non-invasive and ideal method for cancer diagnosis. However, few studies identied the role of specic serum ExmiRNAs proles in early non-small cell lung cancer (NSCLC) diagnosis, especially for 0 and I stage. Herein, the present study was designed to validate the novel serum ExmiRNAs as diagnostic biomarkers for early-stage NSCLC. Serum exosomes were collected from the healthy donors and NSCLC patients by ultracentrifugation, and characterized with qNano, TEM, and western immunoblotting. Exosomal RNAs were subjected to miRNA array for evaluating the expression levels of miRNAs. Partly differently expressed serum exosomal miRNAs were veried by large-scale samples from 312 healthy donors and 318 NSCLC patients.

NSCLC patients present with advanced stage at diagnosis, and have missed the best opportunity of accepting operation. If NSCLC patients can be found and diagnosed in early stage, there will be more treatment options to ensure better overall survival. Liquid biopsy is a currently promising technology for early diagnosis of NSCLC based on blood contents, such as circulating tumor cells [5], circulating tumor DNA [6], and exosomes [7].
Exosomes are known as small membrane vesicles, with a diameter of 30-100 nm, which can be released by most types of cells, including cancer cells [8,9], and contain heterogeneous genetic materials, such as proteins, lipids, DNAs, non-coding RNAs (long non-coding RNAs, microRNAs), and other small molecules.
Increasing evidences revealed that exosomes, as the intercellular communication media, can be delivered to target cells [10,11]. Moreover, after being absorbed, exosomes can signi cantly change the biological behaviors of recipient cells with altering tumor microenvironment and facilitating epithelial-tomesenchymal transition (EMT) progress, further leading to tumor occurrence, invasion and metastasis.
MiRNAs are short, single-stranded and relatively conservative RNA molecules with about 22 nucleotides in length [15,16], representing a class of small nonprotein coding RNAs. MiRNAs, parceled in exosomes, are considered to be stabilized from RNase and highly enriched contrasted to the plasms and serum [17,18]. Given that miRNAs are related to the origination and progression of different diseases such as cancers [10], and are dysregulated in several categories of cancer patients. So exosomal miRNAs can be used as molecular biomarkers for cancer diagnosis [19,20]. For instance, plasma exosomal miR-141-3p and miR-375 were signi cantly over-expressed in rectal cancer patients with synchronous liver metastasis, and can be employed as diagnostic markers of rectal cancer aggressiveness and systemic dissemination [21]. In addition, the expressions of serum exosomal miR-99b-5p and miR-150-5p were obviously decreased in colorectal cancer patients, and these two serum exosomal miRNAs could act as critical biomarkers for diagnosing colorectal cancer patients [22].
In our study, we evaluated differential serum exosomal miRNAs pro le by miRNA array between NSCLC patients and healthy donors. Consequently, serum exosomal miR-155-5p and miR-658 were selected, and their diagnostic potential and relationship with clinical characteristics for early-stage NSCLC were analyzed. Thus, these serum exosomal miRNAs could be served as novel diagnostic biomarkers for NSCLC, especially for early-stage NSCLC.

Materials And Methods
Patients and samples A total of 318 NSCLC patients were recruited between March 2019 to July 2019 from Shandong Cancer Hospital, A liated to Shandong First Medical University (Jinan, China). None of NSCLC patients have received treatment, including surgery, radiotherapy and chemotherapy. Clinical and pathological characteristics of patients, including age, gender, smoking, TNM stage, (according to the eighth edition American Joint Committee on Cancer) were obtained from the patients' records. In addition, a total of 312 healthy donors, who did not have any other tumor disease, were acquired from the same hospital.
Exosomes from 5 healthy donors and 10 NSCLC patients were subjected to miRNA array, and then 318 patients and 312 controls were subjected to veri cation. The peripheral blood from patients and healthy donors were collected and then RT-PCRcentrifuged at 1, 000× g for 10 min at 4°C. The serums were then transferred to fresh tubes and stored at -80°C for exosome isolation.

Isolation of Exosomes
Serum exosomes were isolated using ultracentrifugation according to the procedures as described previously [20,23]. Brie y, the serum was thawed on ice and centrifuged at 10, 000×g for 30 min at 4°C to remove the cellular debris. Then 1 mL of supernatant was ultracentrifugated (Beckman Coulter, Brea, CA, USA) at 100, 000× g for 2 h at 4°C to pelletize exosomes. Exosome pellets were characterized with qNano, TEM, and western immunoblotting, followed by miRNA sequencing and RNAs extraction. Transmission Electron Microscopy (TEM) assay 15 µL of exosome samples were added on formvar-coated copper grid. After 1 min, the remaining liquid was removed by lter paper and the samples were placed in 15µL of 2% uranyl acetate staining solution for 1 min at room temperature. Subsequently, excess liquid was wiped off, and the copper grid was baked under the lamp for 10 min, followed by observing the exosomes by FEI Tecnai T20 transmission electron microscope (FEI Company, USA).

Western immunoblotting
Exosomes or cellular proteins were separated using 10% SDS-PAGE and transferred to PVDF membranes (Millipore, Billerica, MA, USA). After blocked with 5% evaporated skimmed milk in Tris-buffered saline containing 0.1% Tween 20 (TBS-T) for 1 h, the membranes were probed with rabbit primary antibodies against CD63, TSG101, and GM130 overnight at 4 o C. Subsequently, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. The immunoreactive bands were visualized by ECL blotting detection reagents (Bio-Rad, USA).

MiRNA pro ling and Data Analysis
Quali ed RNA from exosomes was used to label miRNA by the miRCURY™ Hy3™/Hy5™ Power labeling kit (Exiqon, Vedbaek, Denmark) according to the manufacturer's instruction. The Hy3™-labeled samples were hybridized on to the miRCURYTM LNA Array (v. 19.0) (Exiqon) according to the instruction manual.
Afterwards, the wash buffer kit (Exiqon) was used to wash the slides for several times, followed by scanning with the Axon GenePix 4000B microarray scanner (Axon Instruments, Foster City, CA). Images were uploaded into GenePix Pro 6.0 software (Axon) for extracting data and aligning the grid. MiRNAs, with intensities >=30, were selected to calculate the normalization factor. After normalizing the expressed data, miRNAs, those were signi cant differentially expressed by fold change >=2.0, were ltered. MiRNA expression levels between the samples were distinguished by hierarchical clustering.

RNA isolation and Real-Time PCR (RT-PCR)
All processes were carried out in an RNase-free area. Total RNAs were isolated from serum exosomes using TRIzol reagent (Thermo Fisher Scienti c, Carlsbad, CA, USA) according to the manufacturer's instructions. Then, the Mix-X miRNA First Strand Synthesis Kit (TaKaRa Bio, Nojihigashi, Kusatsu, Japan) was used to reverse-transcribe RNA into complementary DNA (cDNA) according to the manufacturer's protocol. RT-PCR was performed using TB-Green Premix Ex Taq II Reagent (TaKaRa Bio) according to the manufacturer's instructions on a Roche LightCycler480 System (Roche Diagnostics, Basel, Switzerland) [22,24]. The relative expression levels of miRNAs were calculated using the following equation: ΔCT = CtmiRNA-CtU6 as described previously [25]. Each sample was analyzed in duplicate.
Statistical Analysis SPSS 22.0 (IBM, Ehningen, Germany) and GraphPad Prism 6.0 (San Diego, CA, USA) were used for statistical analysis. The different expressions of miRNAs among groups were determined using the Mann-Whitney unpaired test or paired t test. The association between miRNAs and the clinical characteristics was evaluated by χ2 test or paired-samples t test. Receiver operating characteristic (ROC) and area under the curve (AUC) ware applied to assess the diagnostic power of the candidate predictors. p < 0.05 was considered to be statistically signi cant.

Results
Identi cation of isolated serum exosomes Serum exosomes, isolated from NSCLC patients and healthy donors by ultracentrifugation, were characterized by qNano, TEM and immunoblotting. As shown in Figure 1A, qNano analysis identi ed that the size of isolated serum exosomes distributed from approximately 50 nm to 150 nm in diameter. This was in agreement with the results of TEM, that isolated serum exosomes had a spherical morphology with a diameter of 50-150 nm as indicated in Figure 1B. In addition, western immunoblotting con rmed the two well-known protein markers, including CD63 and TSG101 [19,22], which were enriched in exosomes but absent in cell extracts ( Figure 1C). Besides, GM130 as a negative control was only detected in the cell extracts but not in exosomes [20]. These data demonstrated that the vesicles isolated from serum by ultracentrifugation were truly exosomes.
Exosomal miRNA Pro le Serum exosomes from 5 healthy donors and 10 NSCLC patients were subjected to RNA isolation and miRNA sequencing, the results of raw exosomal miRNA expression pro le showed that a total of 1,937 differential has-miRNAs were screened, and 22 miRNAs (10 upregulated and 12 downregulated, Tables 1) were selected based on >2.0-fold difference between the healthy donors and NSCLC patients ( Figure 2).

Correlations between serum exosomal miRNAs and clinical pathology characteristics
The relationship between serum exosomal miRNAs and the clinical pathology characteristics of NSCLC patients were conducted. As shown in Table 2, the results demonstrated that both serum exosomal miR-155-5p and miR-658 were not correlated with gender, age, smoking, drinking and histological type. Interestingly, serum exosomal miR-155-5p can segregate NSCLC patients with metastasis from patients without metastasis, and expression levels of serum exosomal miR-155-5p were signi cantly decreased in patients with lymph node metastasis and distant metastasis (p = 0.0018 and p = 0.0077, respectively) ( Figure 5A and 5B).

Discussion
Of all new cancer diagnoses, lung cancer accounts for about 14%, and the mortality of lung cancer occupies approximately 25% of all cancer deaths [26]. In addition, survival rates of NSCLC patients decrease dramatically with the advanced progress of diagnosis. On the contrary, assuming that diagnosis for NSCLC is at an early stage, nearly 90% of mortality can be prevented [4]. Thus, it is important and urgent to establish sensitive, speci c and non-invasive biomarkers for the differential diagnosis of NSCLC patients, especially for early-stage patients. Therefore, the present study was designed to con rm the vital role of serum exosomal miRNAs in diagnosis of NSCLC.
Many reports con rmed that miRNAs have better stabilization in vesicles such as exosomes, and highly enriched contrasted to the plasms and serum [17,18]. A study conducted by Tanaka et al. demonstrated that miR-21, stemmed from exosomes, was higher-expressed than that in serum [17]. Additionally, Takuma et al. revealed that the levels of miR-191, miR-21 and miR-451a, derived from exosomes, were signi cantly increased in pancreatic cancer and intraductal papillary mucinous neoplasm (IPMN) patients, while the expressions of these three miRNAs had no obvious changes among the healthy control, pancreatic cancer and IPMN groups [27]. All of these ndings indicated that exosomal miRNAs, as diagnostic markers, were more suitable.
MiR-155-5p has been reported for its multiple functional roles in in ammation, immunity and especially in human malignancies [28][29][30]. It was found that miR-155-5p contributed to down-regulate the expression of ADME gene, drug metabolizing enzymes and transporters under the condition of liver in ammation, which was conducive to better understand and predict drug metabolism of in ammatory diseases [28]. Moreover, another study suggested that miR-155-5p were signi cantly increased in early stages of chronic lymphocytic leukemia (CLL), and was a modest biomarker for predicting the risk of CLL [29]. Besides, miR-155-5p also could function as a tumor suppressor factor by targeting IGF2 in Wilms tumors (WT), and up-regulation of miR-155-5p could suppress cell proliferation, migration and invasion and induce cell apoptosis [30]. Additionally, miR-155-5p was over-expressed in HPV E6/E7 mRNA positive tissues, and could increase the risk of cervical cancer. Thus, miR-155 could serve as a complement approach for diagnosing and predicting the progression of cervical cancer [31]. Moreover, exosomemediated miR-155-5p could distinguish breast cancer patients with these patients suffering an increased risk of developing breast cancer brain metastasis, and could also enhance the chemoresistance in breast cancer cells, providing potential novel targets for breast cancer therapy [32]. Taken together, these previous studies revealed the oncogenic features of miR-155-5p, which was bene cial to support our ndings. In our study, we found that serum exosomal miR-155-5p was signi cantly lower-expressed, and could distinguish NSCLC patients, especially 0 and I stage NSCLC patients from healthy controls, with the AUC value of 0.716. Thus, serum exosomal miR-155-5p could be a potential diagnostic biomarker for early-stage NSCLC patients. Interestingly, serum exosomal miR-155-5p was also a promising non-invasive diagnostic marker for distinguishing metastatic NSCLC patients from non-metastatic NSCLC patients.
At present, there are few studies on the miR-658. In previous studies, miR-658 was down-regulated in gastric cancer with no metastasis compared with gastric cancer with distant metastasis. Further, elevated miR-658 facilitated the migration and invasion of gastric cells, and was closely associated with of gastric cancer with no metastasis by activating PAX3 [33]. Moreover, another study revealed that exosomal miR-658 derived from lung cancer cells with ge tinib-resistant could convert sensitive cells into a resistant phenotype, providing a novel mechanism and target for therapy in cancer patients suffering with resistance against ge tinib [34]. In sum, these previous ndings testi ed that miR-658 played an essential role in human cancers and sustained our present research. Our study demonstrated that serum exosomal miR-658 were signi cantly down-regulated, and could also distinguish NSCLC patients, especially 0 and I stage NSCLC patients from healthy controls, with the AUC value of 0.735.
Serum biomarkers for diagnosis of NSCLC patients have been widely studied and applied in clinical, such as CEA, CA125, NSE, CYFRA21-1, and CA153 [19,20]. However, there is no any effective biomarker for diagnosis of early-stage NSCLC patients. CEA and CYFRA21-1, as meaningful prognostic markers and predictors for therapeutic e cacy or metastasis and recurrence, were selected in the present study. Our data identi ed that CEA and CYFRA21-1 could be combined with serum exosomal miRNAs for diagnosis of early-stage NSCLC patients, and serum exosomal miR-155-5p and miR-658 combined with CEA or CYFRA21-1 could improve the diagnostic power for early-stage NSCLC patients with AUC values of 0.899 and 0.867, respectively.
To the best of our knowledge, this study was the rst time to demonstrate the novel serum exosomal miR-155-5p and miR-658 as biomarkers for diagnosing NSCLC patients, and to identify serum exosomal miR-155-5p a potential non-invasive diagnostic marker for distinguishing metastasis from non-metastatic NSCLC. Our study enrolled enough specimens, which was considerable convincing. However, the present study had several limitations. On the one hand, due to lack of long-term clinical follow-up data, we could not evaluate the prognostic power of serum exosomal miR-155-5p and miR-658. On the other hand, this study focused on healthy donors and NSCLC patients, and did not include cases with begin lung diseases, such as pneumonia, pulmonary brosis, and so on. Thus, a further analysis should be conducted to demonstrate that these biomarkers can be employed to distinguish NSCLC patients from begin lung disease patients and healthy individuals.

Conclusion
In conclusion, down-regulation of serum exosomal miR-155-5p and miR-658 was critical for diagnosing NSCLC patients, especially with early-stage patients (0 and I stage). Combination of serum exosomal miR-155-5p and miR-658 could improve the diagnostic power. Besides, serum exosomal miR-155-5p was a potential non-invasive diagnostic marker for distinguishing metastatisis from non-metastatic NSCLC.

Availability of data and materials
The data that support the fndings of this study are available from the corresponding author upon reasonable request.

Ethics approval and consent to participate
This study was approved by the Medical Ethics Committee of the Shandong Cancer Hospital, A liated to Shandong First Medical University (Jinan, China), and written informed consent was obtained from each patient.

Consent for publication
This manuscript has been seen and approved by all authors for publication.    Exosomal miRNA Pro le of NSCLC patients. Exosomal miRNA pro le of the NSCLC patients. A heat map was generated after supervised hierarchical cluster analysis. The differential miRNA expression is shown in red (upregulation) vs. green (downregulation).