Genome-Wide miRNA Analysis Identifies Potential Biomarkers in Distinguishing Tuberculous and Viral Meningitis.

Liping Pan,Zeyu Ding,Jinli Zhang,Qi Sun,Weibi Chen,Zongde Zhang,Xuehua Liu,Mailing Huang,Aiying Xing,Boping Du,Jing Li,Rongrong Wei,Hongyan Jia,Fei Liu,Yajie Wang

doi:10.3389/fcimb.2019.00323

Abstract

Tuberculous meningitis (TBM) is the most common and severe form of central nervous system tuberculosis. Due to the non-specific clinical presentation and lack of efficient diagnosis methods, it is difficult to discriminate TBM from other frequent types of meningitis, especially viral meningitis (VM). In order to identify the potential biomarkers for discriminating TBM and VM and to reveal the different pathophysiological processes between TBM and VM, a genome-wide miRNA screening of PBMCs from TBM, VM, and healthy controls (HCs) using microarray assay was performed (12 samples). Twenty-eight differentially expressed miRNAs were identified between TBM and VM, and 11 differentially expressed miRNAs were identified between TBM and HCs. The 6 overlapping miRNAs detected in both TBM vs. VM and TBM vs. HCs were verified by qPCR analysis and showed a 100% consistent expression patterns with that in microarray test. Statistically significant differences of 4 miRNAs (miR-126-3p, miR-130a-3p, miR-151a-3p, and miR-199a-5p) were further confirmed in TBM compared with VM and HCs in independent PBMCs sample set (n = 96, P < 0.01). Three of which were also showed significantly different between TBM and VM in CSF samples (n = 70, P < 0.05). The receiver operating characteristic curve (ROC) analysis showed that the area under the ROC curve (AUC) of these 4 miRNAs in PBMCs were more than 0.70 in discriminating TBM from VM. Combination of these 4 miRNAs could achieve better discriminative capacity [AUC = 0.893 (0.788–0.957)], with a sensitivity of 90.6% (75.0–98.0%), and a specificity of 86.7% (69.3–96.2%). Additional validation was performed to evaluate the diagnostic panel in another independent sample set (n = 49), which yielded a sensitivity of 81.8% (9/11), and specificity of 90.0% (9/10) in distinguishing TBM and VM, and a sensitivity of 81.8% (9/11), and a specificity of 84.6% (11/13) in discriminating TBM from other non-TBM patients. This study uncovered the miRNA profiles of TBM and VM patients, which can facilitate better understanding of the pathogenesis involved in these two diseases and identified 4 novel miRNAs in distinguishing TBM and VM.

Highlights

Tuberculous meningitis (TBM) is one of the serious chronic infections of central nervous system (CNS) caused by Mycobacterium tuberculosis (M.TB)
The TBM, viral meningitis (VM), healthy controls (HCs) and other non-TBM were defined as previous studies (Li et al, 2017; Sun et al, 2018): (1) TBM: acid-fast bacilli were detected in the cerebrospinal fluid (CSF), CSF culture was positive, or the CSF XpertMTB/RIF test was positive for M.TB. (2) VM: a viral etiology was confirmed, or the clinical outcome was favorable with supportive and/or antiviral therapy, and bacterial, fungal and other noninfectious causes of meningitis were ruled out (Hristea et al, 2012; Li et al, 2017)
Another 32 TBM, 30 VM, and 34 HCs were assigned to the first validation set

Summary

Introduction

Tuberculous meningitis (TBM) is one of the serious chronic infections of central nervous system (CNS) caused by Mycobacterium tuberculosis (M.TB). About more than half of TBM patients die despite anti-tuberculosis chemotherapy in developing countries (Bidstrup et al, 2002). TBM was definitely diagnosed by Ziehl-Neelsen (ZN) staining of cerebrospinal fluid (CSF) smears and CSF M.TB culture. The smear has very low sensitivity (10–20%), whilst M.TB culture lacks sensitivity and requires appropriately 6–8 weeks obtaining the result (Thwaites et al, 2002; Pai et al, 2003). WHO has recommended XpertMTB/RIF test prior to microscopy and culture as the preferred diagnostic test for suspected TBM patients, the use of this test was limited due to the high cost and low positive predictive value in areas with low prevalence of rifampicin resistance (Torok, 2015). Identification of novel biomarkers is urgently needed for TBM diagnosis

Methods

Results

Conclusion