Expression signatures of exosomal long non-coding RNAs in urine serve as novel non-invasive biomarkers for diagnosis and recurrence prediction of bladder cancer

Recently, expression signatures of exosomal long non-coding RNAs (lncRNAs) have been proposed as potential non-invasive biomarkers for cancer detection. In this study, we aimed to develop a urinary exosome (UE)-derived lncRNA panel for diagnosis and recurrence prediction of bladder cancer (BC). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed to screen and evaluate the expressions of eight candidate lncRNAs in a training set (208 urine samples) and a validation set (160 urine samples). A panel consisting of three differently expressed lncRNAs (MALAT1, PCAT-1 and SPRY4-IT1) was established for BC diagnosis in the training set, showing an area under the receiver-operating characteristic (ROC) curve (AUC) of 0.854. Subsequently, the performance of the panel was further verified with an AUC of 0.813 in the validation set, which was significantly higher than that of urine cytology (0.619). In addition, Kaplan-Meier analysis suggested that the up-regulation of PCAT-1 and MALAT1 was associated with poor recurrence-free survival (RFS) of non-muscle-invasive BC (NMIBC) (p < 0.001 and p = 0.002, respectively), and multivariate Cox proportional hazards regression analysis revealed that exosomal PCAT-1 overexpression was an independent prognostic factor for the RFS of NMIBC (p = 0.018). Collectively, our findings indicated that UE-derived lncRNAs possessed considerable clinical value in the diagnosis and prognosis of BC. Electronic supplementary material The online version of this article (10.1186/s12943-018-0893-y) contains supplementary material, which is available to authorized users.

Bladder cancer (BC) is the most common malignancy of the urinary tract worldwide [1]. About 75% of patients are classified as non-muscle-invasive BC (NMIBC), which features a high recurrence rate. Moreover, roughly half of the muscle invasive BC (MIBC) patients will progress to metastasis and die within 3 years [2]. The poor prognosis of BC is partially due to lack of an effective means for early diagnosis. At present, the diagnosis of BC mainly relies on cystoscopy. However, the invasiveness of such procedure limit its use in mass cancer screening. Urine cytology has been proved to be ineffective as a tool to detect low-grade malignancy. Therefore, discovery of effective biomarkers for detection and recurrence prediction of BC can play pivotal roles in improving the prognosis of BC patients.
Exosomes are small vesicles with a diameter of approximately 30-150 nm. They are distributed in nearly all body fluids, including blood and urine. Exosomes can act as vehicles in cell-to-cell communication by transferring oncogenic molecules and play significant roles in tumorigenesis, progression and metastasis [3]. These functional contents are not stochastically packed into exosomes, which may rely on signature motifs, leading to the variation of exosomal contents under different pathological conditions or in different original cell types [4]. Studies have shown that BC cells can secrete exosomes into the urine and lncRNAs have been found to be stably present in exosomes [5]. Therefore, analyzing the expression profiles of urinary exosome (UE)-derived lncRNAs would provide valuable clues for diagnosis of BC. In this study, we systematically analyzed the expression profiles of UE-derived lncRNAs in BC patients and established a three-lncRNA panel for BC detection. Finally, we the explored the prognostic value of the selected exosomal lncRNAs.

Characterization of UEs
Transmission electron microscopy (TEM) showed UEs have a diameter of 60-150 nm with a cup-shaped membrane (Fig. 1a). Western blotting of UEs demonstrated the presence of CD9 and TSG101, which are exosome markers (Fig. 1b). Nanoparticle tracking analysis (NTA) found that particles ranging from 20 nm to 200 nm in diameter accounted for 98.1% (Fig. 1c). The flow cytometry showed that the positive rate of CD63 and CD81 specific antibodies on the exosome surface was 90.9% and 93.6%, respectively (Fig. 1d). Collectively, these data indicated that exosomes existed in urine, which laid a foundation for further study of exosomal biomarkers. The methods are explained in Additional file 1.

Analysis of the stability of identified lncRNAs in UEs
Next, two experiments were performed to verify the stability of UE-derived lncRNAs (MALAT1, PCAT-1 and SPRY4-IT1), considering that this is an essential prerequisite for biomarkers. Firstly, urine samples and exosome isolated nucleic acids were incubated with RNase A for 0, 30, 60 and 90 min. Strikingly, RNase A had no effect  Figure S3 a-c). However, the exosome isolated nucleic acids group was completely degraded by the treatment of RNase A within 30 min (Additional file 6: Figure S3 d-f). Secondly, urine samples were stored at − 80°C for 1, 2 and 3 months. Results indicated that this treatment had no effect on the expression levels of MALAT1, PCAT-1 and SPRY4IT1 in UEs (Additional file 6: Figure S3 g-i). Collectively, our data indicated that the exosomal membrane can protect lncRNAs from being degraded, and their excellent stability makes exosomal lncRNAs ideal biomarkers for tumor diagnosis.

Establishment of the UE-derived lncRNA panel for BC diagnosis
Considering that combinations of tumor markers can improve the diagnostic accuracy, multivariate logistic regression model was performed in the training set to establish the selected exosomal lncRNA panel. The predictive probability of being diagnosed with BC was calculated using the equation as follows: Logit (P) = 0.6577-0.0695 × MALAT1-0.0686 × PCAT-1 -0.0015 × SPRY4-IT1. The AUC of the panel was 0.854 (95% CI = 0.799-0.899, sensitivity = 70.2% and specificity =85.6%, Fig. 2a).

Validation of the diagnostic performance of the lncRNA panel
To further verify the diagnostic performance of the panel, ROC analysis was performed in the validation set. Results showed that the AUC of the panel was 0.813 (95% CI = 0.744-0.870, sensitivity =62.5% and specificity = 85.0%) (Fig. 2b). Currently, urine cytology is widely used in clinical practice, but it has relatively poor sensitivity. Therefore, we compared the diagnostic performance Fig. 2 Evaluation the diagnostic performance of 3-lncRNA panel and urine cytology for BC diagnosis. ROC analysis was used to evaluate the performance of 3-lncRNA panel for the detection of BC in the training set (a) and in the validation set (b); ROC analysis revealed the diagnostic performance of urine cytology for BC diagnosis in the validation set (c) Fig. 3 Recurrence prediction of UE-derived MALAT1 and PCAT-1 expression for NMIBC. Kaplan-Meier curve revealed that overexpression of UEderived MALAT1 (a) and PCAT-1 (b) was relative to a poor recurrence-free survival in NMIBC patients from the validation set between the panel and urine cytology. As expected, the AUC of urine cytology for BC detection was 0.619 (95% CI =0.539-0.694, sensitivity = 25% and specificity = 98.7%) (Fig. 2c), which was significantly lower than that of the panel.
Correlation between the three UE-derived lncRNAs and clinicopathological characteristics Next, we analyzed the correlation between the three UE-derived lncRNAs and clinicopathological characteristics of the BC patients. Results demonstrated overexpression of UE-derived PCAT-1 and SPRY4-IT1 were correlated with advanced TNM stage (all at p < 0.05). However, we did not find any significant association between the three lncRNAs and age, sex, tumor grade or positive lymph node metastasis (all at p > 0.05) (Additional file 7: Table S3).

Correlation between the three UE-derived lncRNAs and recurrence-free survival (RFS)
To explore prognostic value of the three lncRNAs, BC patients were followed-up in the validation set. In the NMIBC group, results showed that patients with up-regulated MALAT1 and PCAT-1 had a significantly lower RFS (p = 0.002 and p < 0.001, respectively, Fig. 3a-b) compared with their corresponding counterparts. Afterwards, univariate Cox proportional hazards regression showed that there was a significant correlation between RFS of NMIBC and PCAT-1 (p = 0.001), MALAT1 (p = 0.005) or tumor stage (p = 0.001). Multivariate analysis revealed that PCAT-1 (p = 0.018) and tumor stage (p = 0.036) were independent prognostic factors for the RFS of NMIBC (Additional file 8: Table S4). However, none of the three lncRNAs were correlated with the recurrence of MIBC patients (all at p > 0.05, Additional file 9: Table S5).
Researchers have revealed UE-derived HOTAIR could serve as a biomarker for BC. However, their study are only conducted in a very small amount of urine samples, which are not verified in a larger population [9]. Moreover, considering that the progression of BC is a complex pathophysiological process, thus no single lncRNA can stand alone as a biomarker, but instead panel consisting of several lncRNAs will be necessary for BC diagnosis.