Urinary exosomal microRNA profiling in intermediate-risk prostate cancer

MicroRNAs (miRNAs) of urine exosomes have emerged as biomarkers for urological cancers, owing to their high stability. MiRNAs have been linked to factors associated with aggressive prostate cancer such as biochemical recurrence (BCR) and metastasis. In this study, we aimed to identify urinary exosomal miRNAs as prognostic markers associated with BCR in intermediate-risk prostate cancer. We profiled the expression levels of miRNAs via next generation sequencing in urinary exosomes from 21 non-BCR patients and 6 BCR patients of intermediate-risk prostate cancer. A total of 21 urinary exosomal miRNAs were found to be differentially expressed (> twofold) in BCR patients compared to non-BCR patients. For external validation, we validated these results using quantitative reverse transcription PCR in an independent cohort of 28 non-BCR patients and 26 BCR patients. A validation analysis revealed that three miRNAs (miR-26a-5p, miR-532-5p, and miR-99b-3p) were upregulated in exosomes from BCR patients. The univariate and multivariate Cox regression analyses showed that miR-532-5p was an important predictive factor for BCR of intermediate-risk prostate cancer. In conclusion, miR-532-5p in urine exosomes might be a potential biomarker for predicting BCR, which is a poor prognosis in patients with intermediate-risk prostate cancer. Further research is needed on the biological functions and mechanisms of this miRNA.


Scientific Reports
| (2021) 11:7355 | https://doi.org/10.1038/s41598-021-86785-z www.nature.com/scientificreports/ stably detected, because they are highly protected from degradation by RNase 14 . MiR-21 which is one of the representative oncogenic miRNAs was upregulated in urinary exosomes of prostate cancer patients compared to healthy controls, and it was suggested that urinary exosomal miR-21 may be a diagnostic biomarker for prostate cancer 15,16 . In another study with regard to prostate cancer, Huang et al. identified that plasma exosomal miR-1290 and miR-375 could be prognostic biomarkers for castration-resistant prostate cancer 17 .
We have recently reported that it is possible to successfully distinguish prostate cancer from benign prostatic hyperplasia (BPH) with high specificity and selectivity using the exosomes isolated from urine with the aqueous two-phase system (ATPS) method. The ATPS is an exosome isolation method that is simple and has a higher recovery efficiency than the other exosome isolation technique, ultracentrifugation, because it can separate exosomes with ~ 100% efficiency within ~ 30 min 18 . In this study, we performed miRNA profiling using urinary exosomes of prostate cancer patients with intermediate risk who underwent RP and investigated its potential as a predictive biomarker for biochemical recurrence (BCR).

Results
Clinical characteristics. Study subjects were selected as a discovery cohort of 27 patients and a validation cohort of 54 patients in clinically intermediate-risk prostate cancer who underwent RP. Their baseline demographics and tumor characteristics of non-BCR and BCR are described in Table 1. As shown in Table 1, all patients showed a preoperative PSA less than 20 ng/ml, a clinical Gleason score of 7, and a clinical stage of T2. There was no statistically significant difference between the non-BCR and BCR groups with respect to age, preoperative PSA, Gleason score, pathologic stage, margin status, or prostatic intraepithelial neoplasia (PIN) (p > 0.05).

Next generation sequencing (NGS) of urinary exosomes from prostate cancer patients.
To compare the miRNA profiles of non-BCR and BCR patients, we carried out small RNA sequencing using urinary exosomes from 21 patients with non-BCR and 6 patients with BCR. Hierarchical clustering analysis showed that the samples did not cluster into specific groups (data not shown). However, a total of 21 urinary exosomal miRNAs were significantly upregulated in BCR patients compared to non-BCR ones (p-value < 0.05, foldchange > 2). Among the microRNAs with significantly increased differential expression levels, we selected 10 miRNAs as candidates due to their read counts ( Table 2, Discovery cohort).
Validation of the RNA-sequencing data by RT-qPCR analysis. The expression levels of the 10 candidate miRNAs were analyzed in the smaller test cohort of 10 patients with non-BCR and 9 patients with BCR using RT-qPCR. The RT-qPCR results showed that miR-26a-5p, miR-532-5p, miR-99b-3p, and miR-363-3p were expressed at levels more than 1.5-fold higher in BCR patients ( Table 2, Test cohort). To clarify this, the expression levels of four miRNAs in 28 patients with non-BCR and 26 patients with BCR, including previous the smaller test cohort patients, were further analyzed. Based on validation results, three miRNAs (miR-26a-5p, miR-532-5p, and miR-99b-3p) were significantly upregulated in urinary exosomes from BCR patients ( Fig. 1; Table 2, Validation cohort).
To assess the clinical significance of miRNA expression, a BCR-free survival assay was performed on the validation cohort of 54 patients. Kaplan-Meier analysis showed that patients with high three miRNA (miR-26a-5p, miR-532-5p, and miR-99b-3p) expression in their urinary exosomes had significantly shorter BCR-free survival after RP than patients with low miRNA expression did (Fig. 3).

Discussion
Intermediate-risk prostate cancer remains a difficult disease for absolute prognostication, as it consists of a heterogeneous population of patients with variable prognoses. The biochemical and clinical recurrence rates of these patients after RP, RT, or brachytherapy vary from 2 to 70%, and so there has been increasing effort to develop prognostic biomarkers to reflect the actual risk of recurrence more accurately. In fact, various studies have been conducted to identify biomarkers that are able to predict BCR after RP precisely at the DNA, RNA, and protein levels 19,20 . Wei et al. showed that miR-1 in prostate cancer tissue, a novel miRNA-based BCR biomarker, is a potential predictive factor 21 . In addition, Zhao et al. suggested 5 miRNAs (miR-30c-5p, miR-31-5p, miR-141-3p, miR-148a-3p, and miR-221-3p) as a tissue-based multiple miRNA panel for prognostic biomarkers 22 . The BCR biomarkers reported to date are mainly drawn from studies using tissues obtained after RP, but it is difficult to use easily as a testing tool because there are several disadvantages to obtaining the tissue. In contrast, urine is a good tool for biomarker testing because it can be easily obtained in a non-invasive manner.
Our study aimed to investigate urinary exosomal miRNA as a potential predictive biomarker for BCR after RP in intermediate-risk prostate cancer patients. We found 21 upregulated miRNAs in the urinary exosomes of BCR patients compared to the urinary exosomes of non-BCR patients. Next, we validated 3 upregulated miRNAs in independent cohorts using RT-qPCR, and finally miR-532-5p, a single miRNA, was identified as a marker for predicting BCR after RP. To the best of our knowledge, this is the first study to identify miRNA as a predictive biomarker for recurrence using NGS in the urinary exosomes of patients at intermediate risk. Bryant et al. reported that miR-375 and miR-141 are increased in both serum exosomes and microvesicles of recurrent prostate cancer patients compared with non-recurrent prostate cancer patients following RP 23 . Fredsøe et al. found urine-based miRNA biomarkers that could accurately predict the likelihood of recurrence after RP. This three-miRNA model (miR-125b-5p*let-7a-5p/miR-151-5p) significantly predicted time to BCR independently of routine clinicopathological variables such as PSA, clinical T stage, and Gleason score 24 . Although this was also a study based on urinary exosomes, our study has some differences. The exosome was effectively isolated using the ATPS method, providing high recovery efficiency compared to the conventional method, and we successfully performed miRNA profiling in the urinary exosome using NGS. Moreover, we identified a single miRNA that could discriminate between non-BCR and BCR patients.  www.nature.com/scientificreports/ MiR-532-5p has been reported to be involved in tumorigenesis in various cancers, and interestingly, it has both oncogenic and tumor suppressor roles, depending on the type of cancer. For instance, miR-532-5p was upregulated in breast cancer tissues compared with normal tissues, and the downregulation of miR-532-5p decreased the cell proliferation and migration of breast cancer cells 25 . For colorectal cancer (CRC), miR-532-5p was increased in CRC tissues, and furthermore, miR-532-5p overexpression promoted CRC cell growth 26 . The results of these studies indicated that miR-532-5p might be an oncogenic miRNA. Meanwhile, miR-532-5p was downregulated in renal cell carcinoma (RCC) tissue and cell lines. Moreover, miR-532-5p not only inhibited RCC cell proliferation in vitro, but it also suppressed tumorigenicity in vivo 27 . In addition, the expression of miR-532-5p was decreased in bladder cancer, and loss of miR-532-5p promoted the proliferation and invasion of bladder cancer cells 28 . These findings suggested that miR-532-5p functions as a tumor suppressor. However, the role of miR-532-5p in prostate cancer remains unknown. Therefore, further study is required to clarify the roles of miR-532-5p in prostate cancer.
Our study has some limitations. First, this study had a relatively small sample size, with samples collected from a single center, and it lacks external validation. Second, we have proposed that miR-532-5p is a predictor of prostate cancer recurrence in urinary exosomes, but the mechanism associated with this miRNA is still unclear in prostate cancer. Therefore, further study is needed on the roles and related mechanisms of miR-532-5p.

Materials and methods
Patients. The study protocol was approved and carried out in accordance with the approved guidelines by the Institutional Review Board at the Catholic University of Korea, Seoul St. Mary's Hospital (IRB approval No. KC14SISI0213). Urine samples of clinically localized prostate cancer adult's patients who underwent RP were obtained from the Korea Prostate Bank (Seoul, Republic of Korea) with informed consent. This included 49 patients with non-BCR and 32 patients with BCR. BCR was defined as an increase in PSA level to ≥ 0.2 ng/mL after RP. This study was approved by the Institutional Review Board of the Catholic University of Korea, College of Medicine. The characteristics of the recurrent and non-recurrent patients are summarized in Table 1.
Urine was catheterized during RP, and was centrifuged at 2500 rpm for 20 min at 4 °C. The supernatant was transferred to new tubes and stored at − 80 °C.
Urinary exosome isolation. Urinary exosomes were isolated using ATPS (Exo2D, EsosomePlus, Seoul, Republic of Korea) according to the manufacturer's instructions. Briefly, samples were thawed at 4 °C and then were centrifuged again at 2500 rpm for 5 min to remove any residual cell debris. Next, 10 mL of the supernatant were transferred to 15-mL tubes for exosome preparation. Then, 2 mL of Exo2D reagent B were added to the clarified urine and the mixture was incubated at 4 °C for 30 min. The precipitated exosomes were recovered by centrifugation at 3000×g for 30 min at 4 °C. Exosome pellets were resuspended with 200 μL of phosphatebuffered saline.
Exosomal RNA extraction. Exosomal total RNA was extracted from the urinary exosomes using a miRNeasy Serum/Plasma Kit (Qiagen, Hilden, Germany) 29 . The urinary exosomes were disrupted and homogenized in 1 mL of QIAzol lysis reagent, and the rest of the procedure was performed according to the manufacturer's protocol. The extracted RNA was eluted with 14 µL of RNase-free water. The quantity and purity of the RNA was then measured using a NanoDrop 2000 spectrophotometer system (Thermo Fisher Scientific, Waltham, MA, USA). The RNA integrity was assessed by an Agilent 2100 bioanalyzer using the RNA 6000 Pico Chip (Agilent Technologies, Amstelveen, The Netherlands).
Library preparation and sequencing. For control and test RNAs, the construction of a library was performed using a NEBNext Multiplex Small RNA Library Prep kit (New England BioLabs, Inc., Ipswich, MA, USA) according to the manufacturer's instructions 30 . Briefly, for library construction, total RNA from each sample was used at 1 µg to ligate the adaptors and then cDNA was synthesized using reverse-transcriptase with adaptor-specific primers. PCR was performed for library amplification, and library clean-up was carried out using a QIAquick PCR Purification Kit (Qiagen) and AMPure XP beads (Beckman Coulter, Inc., Brea, CA, USA). The yield and size distribution of the small RNA libraries were assessed by the Agilent 2100 Bioanalyzer instrument for the High-Sensitivity DNA Assay (Agilent Technologies). High-throughput sequences were produced by a NextSeq 500 system as way of single-end 75 sequencing (Illumina, San Diego, CA, USA). Data analysis. Sequence reads were mapped by the Bowtie 2 software tool in order to obtain BAM files (alignment files). A mature miRNA sequence was used as a reference for mapping. Read counts mapped on the mature miRNA sequence were extracted from the alignment file using bedtools (v2.25.0) and Bioconductor 31 , which uses the R (version 3.2.2) statistical programming language (https:// www.r-proje ct. org/). Read counts were used to determine the expression levels of the miRNAs. The quantile normalization method was used for comparisons between samples.