Exploring Mycobacterium tuberculosis rRNA Transcriptomic Signatures as Response to Anti-TB Treatment in Whole Blood RNA-seq Data

Current clinical tools for monitoring tuberculosis (TB) treatment response rely mostly on sputum culture and chest X-ray, which are unreliable in patients with low bacterial loads and lack the desirable promptness. There is a need for biomarkers able to provide earlier and more accurate insights into pathogen viability and response to therapy. We analyzed publicly available whole-blood RNA-seq data from 79 TB patients sampled at diagnosis, and weeks 1, 4, and 24 of standard anti-TB treatment. After aligning human reads and filtering, remaining reads were mapped to the Mycobacterium tuberculosis H37Rv genome to quantify rRNA subunit transcripts (16S, 23S, 5S, ITS1, ITS2). Microbiome profiles were assessed using Kraken2/Bracken, with alpha/beta diversity analyses and differential abundance (ANCOM-BC2). 16S and 23S rRNA transcripts were consistently detected across all treatment times, with 23S reads dominating in diagnosis and early stages of treatment shifting toward a significant predominance of 16S reads at week 24. 5S and ITS1 were inconsistently detected, whereas ITS2 was undetectable. Alpha diversity (Shannon index) increased during treatment (significant at weeks 1 and 4), while beta diversity showed significant shifts over time despite no significant differences in total M. tuberculosis abundance. Our findings suggest that it may be feasible to detect M. tuberculosis rRNA signatures in blood RNA-seq and suggest dynamic transcriptomic changes during treatment. The 16S/23S ratio and minor rRNA units may serve as complementary biomarkers for treatment monitoring and other transcriptomic-based biomarkers. Future work should validate these findings in larger cohorts using optimized RNA-seq protocols focusing on the pathogen.