The blood samples from 29 healthy children and 30 HAdV patients used in the present study were obtained from Guangzhou Women and Children's Medical Center (Guangzhou, China) between January 2015 and January 2019. The ages of all patients and healthy volunteers, including 33 males and 26 females, ranged from 1 to 13 years. HAdV pneumonia was diagnosed using the following criteria: i) Lower respiratory and/or systemic symptoms of infection; ii) computed tomography (CT) scan or lung infiltration on chest radiography; and iii) positive results for HAdV immunoglobulin M (IgM) antibody in serum and/or HAdV DNA by PCR in throat swabs and/or BAL fluid. We also excluded the samples from mixed infection patients who were infected with HAdV together with other microorganisms. Serum samples were prepared by centrifugation at 1,000 × g for 10 min. The supernatants were collected and stored in aliquots at −80°C. Study approval was granted by the Ethics Committee at Guangzhou Women and Children's Medical Center (approval no. 2014121815), and all the parents or legal guardians of the patients signed written informed consent forms and agreed to its content.

Serum from all participants was collected for exosome isolation using the ExoQuick Exosome Precipitation kit (System Biosciences, LLC, Palo Alto, CA, USA) according to the manufacturer's protocol (16).

Exosome suspensions in PBS (100 µl) were added to a copper mesh placed on a clean wax plate. After 4 min, the copper mesh was removed and placed in 2% phosphotungstic acid for 5 min, while the mesh was dried on filter paper. The morphology of the exosomes was examined using a JEM-1230 transmission electron microscope (JEOL Ltd., Tokyo, Japan).

The exosome pellet was dissolved in radioimmunoprecipitation assay lysis buffer (Cell Signaling Technology, Inc., Danvers, MA, USA) for 30 min at 4°C, and the protein concentration was determined using a Bradford protein assay kit (Bio-Rad Laboratories, Inc., Hercules, CA, USA) (17,18). The proteins (20 µg/lane) were separated via 15% SDS-PAGE and then transferred to polyvinylidene difluoride membranes. Membranes were blocked in 5% non-fat dried milk for 1 h, followed by incubation for 1 h with anti-CD9 (1:1,000; cat. no. 13174, Cell Signaling Technology, Inc.) and anti-heat shock protein 90α (HSP90α; 1:1,000; cat. no. 8165, Cell Signaling Technology, Inc.) primary antibodies, and subsequent incubation for 1 h with the secondary antibodies (horseradish peroxidase-conjugated anti-rabbit immunoglobulin G; 1:1,000; cat. no. 7074, Cell Signaling Technology, Inc.). The bands were visualized using the SuperSignal chemiluminescence system (Thermo Fisher Scientific, Inc., Waltham, MA, USA). All steps were performed at room temperature.

RNA was extracted from the exosome pellets using TRIzol^® (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol (17).

Small RNA libraries were generated using NEBNext^® Multiplex Small RNA Library Prep Set for Illumina^® (New England BioLabs, Inc., Ipswich, MA, USA); all subsequent steps were conducted according to the manufacturer's protocol. Briefly, the multiplex 3′ small RNA (SR) adapters, which the reverse transcription (RT) primers were hybridized to, were ligated to the miRNA. The RT step was performed upon ligation of the multiplex 5′ SR adapter, and 15 cycles of PCR were performed to enrich those DNA fragments that had adapter molecules on both ends. PCR was performed as follows: 30 sec at 94°C, then 15 sec at 94°C, 30 sec at 62°C, and 15 sec at 70°C for 15 cycles. The library fragments were size-selected by 6% PAGE extraction. The purified DNAs of ~140 nt constituted the miRNA library. The libraries were quantified with using a Qubit 3.0 fluorometer (Invitrogen; Thermo Fisher Scientific, Inc.) and validated with an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA, USA), prior to being sequenced on an Illumina HiSeq 2500 Sequencing System (Illumina, Inc., San Diego, CA, USA) for 50 cycles.

The adapter sequences were removed from all reads. Reads with Phred quality scores were <10 were truncated at their first nucleotides. Reads <17 nt were discarded. The remaining reads were mapped to a human miRNA reference sequence in the miRBase database (http://www.mirbase.org/) using the FANSe2 algorithm with parameters-L60-E2-U1-S10 (FANSe v2.0; http://bioinformatics.jnu.edu.cn/software/fanse2/). miRNA expression was quantified as reads per million. The differentially expressed miRNAs were detected using the edgeR package version 3.8 (R, http://bioconductor.org/packages/release/bioc/html/edgeR.html). miRNAs with >2 or <0.5-fold change, and P<0.01 (no correction for Type I error was applied) were considered as differentially expressed miRNAs (19,20).

RNA was extracted from exosome pellets as aforementioned and the isolated RNAs were reverse transcribed using a SuperScript™ First-Strand Synthesis System for RT-PCR kit (cat. no. 11904018, Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. qPCR was performed using the Power SYBR Green PCR Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc.) (17). Each reaction was performed in a 20-µl volume system including 5 µl cDNA, 10 µl Power SYBR Green PCR Master Mix, 0.5 µl each primer and 4.0 µl double distilled water. PCR was performed as follows: 5 min at 94°C, then 20 sec at 94°C and 30 sec at 60°C for 39 cycles. The relative expression of miRNAs normalized to other miRNAs was calculated using the 2^−∆∆Cq method (15,21). All RT-qPCR primers used in the study are listed in Tables I–III.

Data was analyzed using SPSS software version 18.0 (SPSS, Inc., Chicago, IL, USA). Paired t-tests and two-way analyses of variance were performed to analyze the relative expression of exosomal miRNAs. P<0.01 was considered to indicate a statistically significant difference.

All patients enrolled in the present study were diagnosed with pneumonia based on clinical symptoms and lobar infiltrates on chest radiography. The main clinical symptoms were fever and cough, while two patients exhibited wheezing and tachypnea. All patients were HAdV-positive, with results from the HAdV IgM antibody test using serum and/or HAdV DNA PCR analysis using throat swabs and/or BAL fluid. The clinical characteristics of the cases are summarized in Table IV. Abnormal chest radiographs were also observed in all cases. Chest scans revealed diffuse infiltration of the lungs. Representative high-resolution CT scans of the chest of healthy children and patients are shown in Fig. 1.

Exosomes were isolated from the serum of four healthy children and five adenovirus-infected patients. The patients were included in the study according to the criteria in Table IV. The isolated exosomes were characterized by TEM, and appeared as spherical vesicles of 30–100 nm in diameter in all samples (Fig. 2A), which is consistent with previously reported characteristics of exosomes (20,22). In addition, exosome protein markers, namely CD9 and HSP90α (22,23) were also detected in our exosome samples (Fig. 2B). CD9 and HSP90α were highly expressed in the isolated exosomes compared with their levels in serum. These results suggest that exosomes were successfully isolated in the present study.

To compare the different miRNA signatures of patients with pneumonia caused by adenovirus infection with those of healthy volunteers, a total of 300 ng RNA isolated from 400 µl serum of each of the samples was used for miRNA-seq to compare the differentially expressed miRNAs in HAdV-infected patients. The expression of the miRNAs analyzed by miRNA-seq was compared between samples from four healthy controls and samples from five patients with HAdV infection (Fig. 3A). The majority of miRNAs exhibited significantly altered expression in HAdV-infected patients, providing a clear distinction of HAdV pneumonia. Among the miRNAs evaluated, 18 were significantly upregulated while 10 were downregulated in patients compared with the normal controls (Fig. 3B). These results indicated that these miRNAs may serve as biomarkers for HAdV-infected patients.

The present study further analyzed the miRNAs with significantly different expression in HAdV-infected patients compared with healthy children. In total, the top nine differentially expressed miRNAs in order to analyze their association with pneumonia caused by HAdV infection. The expression of these nine miRNAs was evaluated in samples from five healthy children and five HAdV-infected patients. Importantly, an internal reference is essential in RT-qPCR experiments. However, in the case of miRNAs isolated from exosomes, an internal reference is not available. To solve this issue, pairwise analysis of each miRNA was used. It was observed that miR-103b-5p/miR-98-5p and miR-450a-5p/miR-103a-3p exhibited opposite expression trends. miR-103b-5p and miR-450a-5p were upregulated in HAdV samples while miR-98-5p and miR-103a-3p were downregulated in HAdV samples, indicating that this ratio may serve as a biomarker for HAdV pneumonia (Fig. 4). Notably, miR-103b-5p is a serum biomarker for colorectal adenocarcinoma (24). Thus, the present results suggested that the selected miRNAs may reflect HAdV infection.

To verify the diagnostic capability of serum miRNAs as potential biomarkers in the present study, blood samples from 20 healthy children and 20 HAdV-infected patients were collected. The RT-qPCR results of the two pairs of miRNAs identified above were analyzed for all the samples in the validation cohort (Fig. 5). Relative Cq values for the miRNAs in the two pairs were subtracted from one another [Cq(miR-450a-5p)-Cq(miR-103a-3p) and Cq(miR-103b-5p)-Cq(miR-98-5p)]; subtracted values were used instead of ratios to simplify clinical analysis. The two pairs of miRNAs analyzed between healthy controls and HAdV patients were clearly separated as two clusters, suggesting that the miRNAs selected in the present study could be considered as good diagnostic biomarkers for HAdV pneumonia, at least in the cohort experiment. Importantly, neither pair of miRNAs could independently distinguish HAdV-infected patients from healthy children, highlighting the requirement for combining the two miRNA pairs identified in the present study.