Non-Coding RNAs Shape Tuberculosis Infection Across Host and Pathogen Boundaries

Tuberculosis remains one of the world’s most formidable infectious threats, caused by members of the Mycobacterium tuberculosis complex and responsible for an estimated 10.7 million new cases and 1.23 million deaths globally in 2024. Following a temporary displacement by COVID-19 across three prior years, the pathogen has reclaimed its status as the leading cause of mortality from a single infectious agent worldwide. A comprehensive scientific review published in Frontiers in Cellular and Infection Microbiology by researchers from the University of KwaZulu-Natal, Stellenbosch University, and the University of Western Cape synthesizes evidence from 131 primary experimental, transcriptomic, and clinical investigations published between 2021 and 2026. The authors establish that non-coding ribonucleic acids function as decisive post-transcriptional regulators across both sides of the infection interface. Rather than serving as passive transcriptional byproducts, eukaryotic host transcripts and mycobacterial regulatory molecules engage in an intricate cross-kingdom molecular contest. These untranslated regulators directly determine whether host macrophages successfully restrict intracellular bacilli through autophagy and regulated cell death, or whether the pathogen adapts to aggressive host defenses, enters protective metabolic dormancy, and acquires resistance to frontline antibiotics.

The review delineates four prominent structural classes of non-coding RNAs that execute distinct functional operations during tuberculosis pathogenesis. Host cells mobilize small, evolutionarily conserved microRNAs measuring approximately 18 to 24 nucleotides in length, which mature through canonical and non-canonical biogenesis pathways to post-transcriptionally silence target messenger RNAs. Alongside them, long non-coding RNAs exceed 200 nucleotides and mature through processes resembling messenger RNA biogenesis, utilizing diverse nuclear and cytoplasmic mechanisms to bind chromatin, proteins, and competing RNAs. Circular RNAs represent an intrinsically durable class formed by back-splicing or exon skipping of pre-messenger RNAs, whose covalently closed-loop structure lacks free 5-prime caps and 3-prime polyadenylated tails, conferring resistance to conventional RNA-degrading exonucleases. On the microbial side, Mycobacterium tuberculosis produces specialized small regulatory RNAs that fall into two structural categories: cis-encoded antisense transcripts spanning 75 to 10,000 nucleotides that share absolute complementarity with their overlapping genomic templates to guide ribonuclease-III degradation, and trans-encoded sRNAs spanning 30 to 300 nucleotides that exhibit imperfect complementarity across multiple genetic targets. Because the bacterial genome lacks the canonical Hfq RNA chaperone found in other bacterial species, mycobacterial sRNAs rely on alternative, unresolved cellular mechanisms to execute stress adaptation and rewire metabolic pathways.

Host microRNAs exert rigorous control over inflammatory signaling cascades during mycobacterial invasion, frequently serving as molecular levers that intracellular bacilli manipulate to escape destruction. Profiling of peripheral blood from patients with active and latent tuberculosis demonstrates marked upregulation of miR-1236-3p, a phenotype reproduced during in vitro infection of human THP-1 macrophages. Bioinformatics analysis using TargetScan and validation via dual-luciferase reporter assays confirmed that miR-1236-3p directly binds the 3-prime untranslated region of the toll-like receptor 4 gene. By dampening toll-like receptor 4 expression, the microRNA downregulates myeloid differentiation primary response 88 and tumor necrosis factor receptor-associated factor 6, inhibiting nuclear factor kappa B phosphorylation. This repression curtails macrophage production of interleukin-6, interleukin-1-beta, tumor necrosis factor-alpha, reactive oxygen species, and nitric oxide, disabling host bactericidal activity. Conversely, infection-induced downregulation of miR-26a releases the histone deacetylase sirtuin 6, which transcriptionally represses the glycolytic genes glut1, ldha, and aldolase a, concurrently reducing succinate dehydrogenase-dependent interleukin-1-beta release through hypoxia-inducible factor-1-alpha and nuclear factor kappa B pathways. Other regulatory axes, such as miR-18b suppression of hypoxia-inducible factor-1-alpha and miR-502-3p targeting of Rho-associated coiled-coil-forming protein kinase 1, further demonstrate how microRNA modulation alters macrophage antimicrobial competence.

Long non-coding RNAs function extensively as competing endogenous RNAs that intercept and sponge microRNAs, thereby preserving downstream messenger RNA translation and governing innate defense networks. Experimental models evaluating negative-pressure therapeutic interventions revealed that mechanical pressure downregulates the long non-coding transcript X-inactive specific transcript in infected macrophages. Similarly, the long non-coding RNA differentiation antagonizing non-protein coding RNA acts as a master regulator of antimycobacterial autophagy. By sponging miR-1301-3p and miR-5194, this transcript prevents the post-transcriptional repression of core autophagy-related genes including ATG4D and ATG5, alongside STAT3, STAT5B, and RHEB. This sponge interaction accelerates both autophagosome formation and autolysosome fusion, successfully restricting intracellular mycobacterial replication. In parallel, small nucleolar RNA host gene 16, which is notably depleted in active tuberculosis patients but preserved during latent infection, promotes autophagic clearance by sequestering miR-302a-3p and miR-15a-5p, stimulating vitamin D receptor-dependent production of antimicrobial cathelicidin and beta-defensin 2 while priming lymphocytes to secrete interferon-gamma.

Circular RNAs provide an additional layer of post-transcriptional regulation by sequestering microRNAs within macrophages, dictating the balance between pathogen clearance and intracellular survival. Infection of THP-1 macrophages with the virulent clinical strain 1458 induces robust expression of circ_0004771, which directly sequesters miR-3921 as validated by quantitative real-time PCR, dual-luciferase reporter assays, and fluorescence in situ hybridization. De-repression of the downstream target, triggering receptor expressed on myeloid cells-1, enhances p65 phosphorylation, stimulates interleukin-1-beta secretion, and curtails intracellular bacterial viability. Similarly, circ-ZNF277 sponges miR-378d to upregulate the Ras-related protein Rab10, driving the release of pro-inflammatory cytokines that facilitate pathogen elimination. In contrast, mycobacterial infection exploits other circular transcripts to promote persistence: circRNA_SLC8A1 binds miR-20b-5p to elevate SQSTM1 and p62 expression, activating nuclear factor kappa B signaling in a manner that inhibits host-protective apoptosis. In the autophagic arena, circTRAPPC6B expression correlates positively with ATG16L1 mRNA levels by sequestering miR-874-3p, whereas pathogen-induced circTRAPPC6B downregulation represents an immune evasion tactic that blunts autophagosome maturation. Additionally, monocyte-derived hsa_circ_0045474 exhibits profound downregulation in tuberculosis patients, coordinating macrophage autophagy via the miR-582-5p and Tankyrase-2 regulatory axis.

The orchestration of programmed host cell death represents another pivotal battleground where non-coding RNAs determine the clinical course of infection. While classical apoptosis restricts mycobacterial dissemination by packaging bacilli within intact apoptotic bodies for efficient antigen presentation, necrotic rupture releases viable bacteria and promotes extensive tissue damage. Natural killer cell-derived exosomes deliver miR-1249-3p into infected macrophages, stimulating intrinsic and extrinsic apoptotic cascades by elevating Caspase-3, Caspase-8, Caspase-9, poly-ADP-ribose polymerase, bax, bid, and caspase-1 while downregulating bcl-2 and its direct target skor1. In stark contrast, mycobacteria employ the surface lipoprotein PstS1 to provoke macrophage exosomal delivery of miR-122, which suppresses Beclin 1 and LC3-II via the PI3K/AKT/mTOR axis, simultaneously inhibiting autophagy and provoking destructive tissue inflammation. Bacterial virulence factors also engage long non-coding RNAs; the secretory protein ESAT-6 drives upregulation of the transcript NEAT1, which sponges miR-125-5p to boost tumor necrosis factor-alpha-mediated cell death. When necrosis occurs, the cytoplasmic long non-coding RNA NR_003508 sequesters miR-346-3p to upregulate receptor-interacting serine/threonine-protein kinase 1, assembling the RIPK1/RIPK3/MLKL necrosome, triggering mitochondrial reactive oxygen species production, inducing calcium ion influx, and sustaining necrotic tissue destruction.

Recent investigations highlighted in the review link host non-coding RNAs to newly defined, non-apoptotic regulated cell death modalities, specifically iron-dependent ferroptosis and copper-dependent cuproptosis. Pathogen survival in human macrophages is profoundly shaped by metabolic perturbations, with tuberculosis lesions exhibiting hallmark signatures of ferroptosis including glutathione peroxidase 4 repression, iron overload, and toxic lipid peroxidation. A newly identified transcript, lncRNA-cytoplasm-regulating ferroptosis and tuberculosis survival, is significantly upregulated during infection. Mechanistic assays confirmed that this long non-coding RNA sponges miR-515-5p and miR-519e-5p, relieving repression on spermidine/spermine N1-acetyltransferase 1. Elevated enzyme activity subsequently induces arachidonic acid 15-lipoxygenase expression, driving lipid peroxidation and ferroptotic cell death through iron metabolic pathways independent of the canonical cystine/glutathione/GPX4 axis, thereby establishing a permissive environment for mycobacterial persistence. Simultaneously, investigations into copper toxicity reveal that nuclear factor erythroid 2-related factor 2 coordinates the expression of cuproptosis genes including Ferredoxin 1, dihydrolipoyl transacetylase, SLC31A1, and HSP70 in infected THP-1 cells. Among seven candidate long non-coding transcripts targeting Ferredoxin 1, lncRNA-Gm5532 was shown through bioinformatics modeling to sponge miR-7232-5p, promoting copper accumulation, dihydrolipoyl transacetylase lipid acylation, mitochondrial membrane collapse, and cuproptotic macrophage lysis.

Mycobacterium tuberculosis deploys its own battery of small non-coding RNAs to navigate host-imposed nutrient depletion, oxidative bursts, and antimicrobial chemotherapy. Under nutrient starvation and chronic macrophage infection, the bacterial sRNA F6 exhibits sharp upregulation, repressing the HrcA-regulated chaperonins groES and groEL2 to slow metabolic activity and enforce a state of protective dormancy. The stress-responsive sRNA MTS1338, also known as DrrS, is transcriptionally governed by the dual virulence regulators DosR and PhoP, which bind its upstream core promoter while post-transcriptional stability is maintained through the depletion of the ribonuclease YbeY. Upregulated MTS1338 enhances bacterial catalase activity to detoxify host reactive oxygen species, stabilizes the virulence factor whiB6 mRNA against ribonuclease degradation, interacts with Rv3679 to regulate glycerol metabolism, and confers pathogenic phenotypes when heterologously expressed in Mycobacterium smegmatis. Similarly, sRNA MTS2823 binds the coding region of recA in multidrug-resistant clinical lineage 2 strains, suppressing macrophage nitric oxide and cytokine release while consolidating resistance to isoniazid.

Circulating host and bacterial non-coding RNAs have emerged as viable candidates for non-invasive diagnostic, prognostic, and treatment-monitoring biomarkers in liquid biopsies. Plasma extracellular vesicle profiles combining miR-451a, miR-1908-5p, and miR-1268b demonstrated diagnostic accuracy with an area under the curve of 0.970 against healthy controls, while pairing miR-451a with miR-1268b discriminated asymptomatic tuberculosis with an area under the curve of 0.880. For drug-resistant tuberculosis, combining host miR-132 with mycobacterial sRNA015 elevated diagnostic accuracy to an area under the curve of 0.862. Among circular RNAs, serum circ_0001380 achieved an area under the curve of 0.9502 with 93.75 percent sensitivity and 87.50 percent specificity, while the long non-coding RNA TCONS_00001838 reached an area under the curve of 0.828, closely approaching the 0.860 benchmark of commercial interferon-gamma release assays. Furthermore, exosomal bacterial sRNAs detected in human peripheral blood—specifically MTS0997, MTS1338, and MTS2823—yielded a combined area under the curve of 0.9104, while combined ASdes and MTB-miR5 attained an area under the curve of 0.992, successfully identifying infected individuals who tested negative by conventional sputum smear microscopy and culture.

Despite these promising molecular and diagnostic insights, the authors emphasize that substantial translational hurdles must be overcome before non-coding RNA technologies can enter clinical practice. An overwhelming majority of published mechanistic studies rely strictly on avirulent H37Ra or virulent H37Rv laboratory control strains, which fail to reflect the extensive genetic diversity, distinct lineage methylomes, and differing immunogenic profiles characteristic of the ten globally circulating Mycobacterium tuberculosis lineages. Furthermore, an acute methodological imbalance persists across the literature: while microRNA pathways benefit from extensive experimental validation, numerous circular RNA networks and bacterial sRNA-mRNA interactions remain grounded solely in computational predictions from tools such as TargetRNA2, IntaRNA, CopraRNA, and TargetScan without confirmation by Northern blotting or Rapid Amplification of cDNA Ends PCR. In addition, existing biomarker cohorts are predominantly small, single-center studies drawn from restricted geographic regions, notably China, South Korea, and Indonesia, creating publication bias and leaving biomarker performance largely untested across high-burden settings where endemic HIV co-infection, diabetes, and severe malnutrition alter immune baselines. Finally, the relative scarcity of longitudinal studies investigating latent infection leaves critical gaps in understanding how non-coding RNA switches govern progression to fulminant disease, underscoring the urgent necessity for prospective multicenter clinical trials.

Subject of Research: Cross-kingdom regulatory roles of host and mycobacterial non-coding RNAs in tuberculosis pathogenesis and clinical biomarker development

Article Title: Mycobacterial and host non-coding RNAs in tuberculosis: immunoregulatroy functions, adaptation strategies, and biomarker potential

Article References: Nyide, A. N. G., de Oliveira, T., & Mvubu, N. E. (2026). Mycobacterial and host non-coding RNAs in tuberculosis: immunoregulatroy functions, adaptation strategies, and biomarker potential. Frontiers in Cellular and Infection Microbiology, 16, Article 1937599. https://doi.org/10.3389/fcimb.2026.1937599

Image Credits: AI Generated

DOI: 10.3389/fcimb.2026.1937599

Keywords: tuberculosis, non-coding RNA, microRNA, long non-coding RNA, circular RNA, small RNA, Mycobacterium tuberculosis, biomarkers, Mycobacterial, host, non-coding, RNAs