Whole blood transcriptional profiles and the pathogenesis of tuberculous meningitis

Mortality and morbidity from tuberculous meningitis (TBM) are common and closely linked to the inflammatory response triggered by Mycobacterium tuberculosis infection, though the mechanisms underlying this association are not well understood. To explore this, we aimed to identify gene modules, hubs, and pathways associated with TBM pathogenesis and mortality, and determine the best predictors of death. Using whole blood RNA sequencing, we analyzed transcriptional profiles from 281 Vietnamese adults with TBM (207 HIV-negative; 74 HIV-positive), 295 with pulmonary TB (PTB), and 30 healthy controls. Through weighted gene co-expression network and pathway analysis, we identified modules, hub genes and pathways linked to TBM severity and mortality, with a consensus analysis identifying consensual patterns between HIV-positive and HIV-negative individuals. Using multivariate elastic-net Cox regression, we selected candidate predictors of TBM mortality, then model prediction performance using logistic regression and internal bootstrap validation to choose best predictors. Increased neutrophil activation and decreased T and B cell activation pathways were linked to TBM mortality. In HIV-positive adults, death was associated with increased angiogenesis, while HIV-negative individuals exhibited heightened TNF signaling and reduced extracellular matrix organization. Four hub genes – MCEMP1, NELL2, ZNF354C, and CD4 – emerged as strong predictors of death from TBM (AUC 0.80 in HIV-negative, 0.86 in HIV-positive). Our findings suggest TBM induces a systemic inflammatory response similar to PTB, but with key gene modules, hubs and pathways strongly associated with death, offering insights for potential therapeutic targets and a novel 4-gene biomarker for predicting TBM outcomes.

Methods
Whole blood samples, collected from participants at enrollment, were stored in PAXgene collection tubes at -800C. RNA samples were isolated using the PAXgene Blood RNA kits (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions, except for an additional washing step before RNA elution. DNA was digested on columns using the RNase-free DNase Set (Qiagen, Valencia, CA, USA). Quality control of the RNA extraction was performed using the Epoch spec for quantity and quality, and Tapestation Eukaryotic RNA Screentape for integrity. RNA-seq was performed by the Ramaciotti Centre for Genomics (Sydney, Australia). One microgram of total RNA was used as input for each sample, using the TruSeq Stranded Total RNA Ribo-zero Globin kit (Illumina). Libraries were generated on the Sciclone G3 NGS (Perkin Elmer, Utah, USA) and the cDNA was amplified using 11 PCR cycles. Libraries were pooled 75 samples per pool and sequenced using NovaSeq 6000 S4 reagents at 2x100bp to generate about 30 million reads per sample. RNA-seq data quality control and pre-processing are described in method section (https://doi.org/10.7554/eLife.92344)