Request a Quote

Tuberculosis Small RNAs: What Blood Tests Really Score

Tuberculosis small RNAs are the regulatory layer that lets a genetically identical bacterial population behave like a spectrum of drug responses. Bacterial small RNAs tune stress adaptation, dormancy, and drug tolerance without touching the DNA sequence, while host microRNAs, long non-coding RNAs, and circular RNAs tune the macrophage from the other side of the same interface. One review integrating both sides of that cross-kingdom conversation assembled 131 publications over five years. Its sharpest finding is pharmacological: a bacterial small RNA is switched on by rifampicin exposure and then switches on an efflux pump that removes rifampicin from the cell.

Why a regulatory RNA layer matters in a pathogen with no Hfq

Tuberculosis still kills more people than any other single infectious pathogen. In 2024 an estimated 10.7 million people fell ill and roughly 1.23 million died, after three years in which a different respiratory pathogen held that rank. Ten human-adapted lineages circulate, differing in genetic composition, host adaptation, geography, and transmission, and some carry distinct virulence phenotypes.

Protein-coding regulation is only part of that picture. Non-coding RNAs are not translated and act by binding DNA, RNA, or proteins. On the host side, microRNAs of about 18 to 24 nucleotides work post-transcriptionally, long non-coding RNAs above 200 nucleotides work through several mechanisms and can sponge microRNAs, and circular RNAs close into covalently linked loops that resist degradation. On the pathogen side, small RNAs of roughly 50 to 400 nucleotides regulate bacterial genes through target messenger RNAs and RNA-binding proteins.

One detail makes the bacterial side odd. Canonical trans-encoded small RNAs in other bacteria need the chaperone Hfq to locate their targets, and no Hfq or Hfq homolog has been found in this pathogen. The interaction mechanism must therefore be something else, and the review says outright that it is not well delineated. That gap has a practical cost: when the targeting rule is unknown, computational prediction of targets carries more uncertainty than it would in a model organism.

How the evidence was assembled

The scope is deliberate rather than exhaustive. Two databases were searched with a fixed keyword set, work published between 2021 and 2026 was prioritized, and a paper had to address host-derived or mycobacterial non-coding RNAs in tuberculosis with experimental, clinical, transcriptomic, or mechanistic evidence. Duplicates, conference abstracts, editorials, and commentaries were excluded. The search yielded 131 publications.

Class

Typical size

Origin

Main regulatory role

Evidence strength in the review

MicroRNA

18 to 24 nucleotides

Host

Post-transcriptional control of cytokine signaling, macrophage fate, autophagy

Large observational literature, small cohorts

Long non-coding RNA

Above 200 nucleotides

Host

Chromatin and transcript regulation, microRNA sponging

Moderate, frequent in vitro validation

Circular RNA

Closed loop, variable

Host

MicroRNA sponging, transcription and processing control

Smaller literature, high reported discrimination

Bacterial small RNA

50 to 400 nucleotides

Pathogen

Stress adaptation, metabolism, dormancy, drug tolerance

Mechanistic but often prediction-only

The last column is the honest caveat, and the review keeps returning to it. Many proposed small RNA to messenger RNA interactions were predicted with software and never validated experimentally. That does not make them wrong. It does mean the mechanistic layer is thinner than the descriptive layer, and a reader should know which layer a given sentence sits in.

The small RNA that turns rifampicin treatment into an efflux signal

For anyone working with rifampicin, the most consequential finding is a loop rather than a mutation. Rifampicin exposure upregulates a small RNA designated MTS1338, and that upregulation activates the gene for the efflux protein CydC, which lowers intracellular drug concentration. Prediction plus reporter work indicated the small RNA binds the coding region of the efflux transcript and stabilizes it post-transcriptionally.

The functional consequence was measured, not inferred: strains carrying the upregulated small RNA grew in the presence of rifampicin and of a second-line injectable, where control strains did not. The same small RNA is induced by oxidative stress, acidic pH, and nitric oxide, and its induction inside macrophages reflects both transcriptional upregulation through the dormancy and PhoP regulators and slower post-transcriptional degradation.

That is a different resistance logic from the one rifampicin testing usually assumes. A minimum inhibitory concentration on a susceptible strain answers a question about a fixed genome. A regulatory loop answers a question about a population that can raise efflux capacity within hours of first contact with the drug. For a manufacturer or a quality team the implication is narrow but real: a susceptible result on a reference strain does not rule out a regulatory route to lower intracellular exposure in a clinical isolate.

A second small RNA repeats the pattern with a different drug. MTS2823 was reported to modulate isoniazid resistance, and its overexpression was associated with greater intracellular growth in a multidrug-resistant background. Two molecules, two drugs, one mechanism class: small RNAs adjusting the envelope and efflux economy instead of mutating the target.

Adaptation and persistence, measured by what knockout costs

The strongest causal evidence in the review comes from deletion and overexpression rather than correlation. Knocking out MTS1338 and MTS0997 together produced strains that grew more slowly in alveolar macrophage culture medium, the in vitro signal that these RNAs support intracellular survival. In a susceptible mouse strain, the double mutants caused less lung infiltration, fewer neutrophils and macrophages, and less granuloma formation than wild-type infection.

Silencing a second small RNA, Mcr11, in clinical lineage 4 isolates gave the opposite transcriptional signature. Many essential genes fell, spanning lipid synthesis and metabolism, ribosomal proteins, transport systems, respiratory and energy systems, membrane and cell wall components, intermediary metabolism, lipoproteins, and virulence genes. The authors read that as a shift toward slow metabolism or intermediate dormancy under stress, which is precisely the state in which a drug needing active growth struggles to kill.

A third example ties the layer to host immunology. Overexpression of MTS2823 suppressed macrophage secretion of pro-inflammatory cytokines, with recA proposed as the mediator. A pathogen RNA suppressing a host bactericidal program is a different claim from the usual host-microbe narrative, and it would matter for any host-directed adjunct.

The review is careful about the two kinds of evidence. Knockout phenotypes are causal but scarce. Expression correlations are plentiful but weak alone. Treating a predicted interaction as a validated one is the most common overreach in this literature.

Host non-coding RNAs that shape the macrophage response

Host-side regulation is where most of the measured diagnostic performance sits, and it runs through recognizable immune pathways. One microRNA, miR-1236-3p, is elevated in active and latent infection against healthy controls, rises when macrophages are infected, and was shown by reporter assay to target the 3 prime untranslated region of the TLR4 transcript, lowering TLR4 and the downstream MyD88 signal.

Long non-coding RNAs act on macrophage fate rather than on one cytokine. During infection they influence survival, apoptosis, autophagy, pyroptosis, and cytokine output through nuclear and cytoplasmic routes, and some act as competing endogenous RNAs that sponge microRNAs. Circular RNAs are the more stable class because the closed loop resists exonuclease digestion, which is exactly why they were proposed as durable markers.

The stability argument is the most transferable idea on this side. A marker that survives collection and storage is worth more in a blood test than a marker with better biology and worse logistics. Circular RNAs and exosome-packaged bacterial small RNAs both have that property, and both appear in the diagnostic section for the same reason.

How tuberculosis small RNAs perform against other marker classes

The quantitative section is a catalogue of single-cohort area under the curve values. Read together they are more informative than read one at a time, because the range across classes is narrow while cohort size and validation depth are not.

Marker class

Marker

Reported discrimination

Cohort context

Bacterial small RNA

MTS0997 plus MTS1338 plus MTS2823

AUC 0.9104, 92.31 percent specificity, 88.24 percent sensitivity

Blood, patients unable to produce sputum

Bacterial small RNA plus vesicle microRNA

ASdes combined with MTB-miR5

AUC 0.992, 97.62 percent sensitivity, 100 percent specificity

Small extracellular vesicles

Bacterial small RNA plus host microRNA

miR-132 combined with sRNA015

AUC 0.862 for multidrug-resistant disease

Resistant versus healthy comparison

MicroRNA

miR-199b-3p panel

83.33 percent sensitivity, 90 percent specificity, AUC 0.90

Active infection versus healthy

MicroRNA

miR-451a plus miR-1268b

AUC 0.880

Active infection versus healthy

Long non-coding RNA

TCONS_00001838

AUC 0.828, 76.24 percent sensitivity, 72.93 percent specificity

Compared with interferon gamma release assay at 0.860

Circular RNA

circ_0001380

AUC 0.9502, 93.75 percent sensitivity, 87.50 percent specificity

Single cohort

Long non-coding RNA panel

Four transcripts in severe versus mild disease

AUC 0.7820, 0.7288, 0.6647, 0.6615

Severity stratification

Two comparisons inside that table carry more weight than the rest. The long non-coding RNA at 0.828 sits close to an interferon gamma release assay at 0.860 with 91.67 percent sensitivity and 74 percent specificity, and well above a serological antibody test at 0.691 and a skin test at 0.738. That is a fair benchmark because it uses established clinical comparators rather than healthy controls alone.

The bacterial panel at 0.9104 is the more interesting result for a sputum-independent test, and the review is straightforward that it still trails a standard molecular assay on both sensitivity and specificity. A blood marker slightly worse than sputum-based molecular testing but available when sputum is not is a genuine clinical niche, not a replacement.

The counterintuitive finding: the best numbers come from the weakest designs

The obvious reading of that table is that the highest area under the curve marks the best marker. The review's own caveats point the other way.

The highest values come from the smallest and least replicated settings. An area of 0.992 with 100 percent specificity is a strong claim, and such claims usually reflect a small single-site cohort with clean separation between cases and healthy controls. The review itself notes that most findings come from small single-center studies in specific populations, and that variability of signatures plus limited validation across populations are the main barriers to translation. That is a polite way of saying the numbers have not traveled.

The second point is structural. For tuberculosis small RNAs of bacterial origin the specificity problem is smaller in principle, because a pathogen transcript should vanish when the pathogen does, but these markers pay for it in sensitivity. Host markers have the opposite profile: easy to detect, hard to attribute. A host microRNA or long non-coding RNA that reflects immune activation moves in latent infection, in other inflammatory disease, and in treated-but-not-cured states. Half the field is optimizing specificity and the other half sensitivity, and the review presents both without pretending either is settled.

Practical decisions for diagnostics and for susceptibility testing

  • Treat a blood-based non-coding RNA panel as a sputum-independent adjunct, not a replacement, until it has been validated against a clinical comparator in more than one population.
  • Favor markers whose stability survives sample handling, which is why circular RNAs and vesicle-packaged bacterial small RNAs keep reappearing.
  • Report the comparator alongside the area under the curve. A 0.828 against an interferon gamma release assay means something different from 0.95 against healthy controls.
  • When reading rifampicin susceptibility, remember that a regulatory efflux loop can lower intracellular exposure without any resistance mutation, so genotypic and phenotypic results answer different questions.
  • In a development program, measure small RNA induction under drug exposure as a tolerance signal, since the rifampicin-responsive small RNA is induced by the drug it defends against.
  • Treat predicted RNA interactions as hypotheses. The review marks many targets as unconfirmed, and software prediction is not validation.

The fourth point touches release testing most directly. A susceptibility result comes from defined laboratory conditions against a defined inoculum. A strain that raises efflux-regulating small RNA within hours of exposure is invisible to that measurement, and no current pharmacopoeial method is designed to see it. That is a limitation of the assay rather than of the drug, and it belongs in a technical dossier rather than in a footnote.

Limitations

The review covers 2021 to 2026 and excludes earlier mechanistic work except as background, so historical depth is deliberately traded away. It is not systematic and reports no risk-of-bias assessment or pooled estimate. Many interactions rest on computational prediction rather than experiment. The cohorts behind the diagnostic numbers are small, single-center, and rarely replicated across populations differing in ancestry, exposure, or HIV status. Tuberculosis small RNAs of bacterial origin are studied in a handful of laboratory strains and mouse strains. No non-coding RNA marker has been validated prospectively for treatment monitoring, and none is in routine clinical use.

Bottom line

Tuberculosis small RNAs sit where a rifamycin program should care. On the pathogen side, a rifampicin-responsive small RNA upregulates an efflux pump and lowers intracellular drug concentration, while knockouts of two other small RNAs reduce intracellular growth and granuloma formation in vivo. On the host side, microRNA, long non-coding RNA, and circular RNA panels report areas under the curve from 0.66 to 0.99, with the highest values coming from the smallest cohorts. The useful conclusion is not that RNA markers are ready, but that drug tolerance in this pathogen has a regulatory component that conventional susceptibility testing does not measure.

FAQ

What are tuberculosis small RNAs

Tuberculosis small RNAs are short regulatory transcripts of roughly 50 to 400 nucleotides encoded by the pathogen itself, distinct from host microRNAs, long non-coding RNAs, and circular RNAs. They regulate bacterial gene expression by interacting with target messenger RNAs and RNA-binding proteins, and they influence stress adaptation, metabolism, dormancy, persistence, and drug tolerance. Unlike many bacteria, this pathogen has no identified Hfq chaperone, so their targeting mechanism is unresolved.

How does a small RNA contribute to rifampicin resistance

Exposure to rifampicin upregulates a small RNA designated MTS1338, which activates the gene encoding the efflux protein CydC and lowers the intracellular drug concentration. Reporter work indicated the small RNA binds the coding region of the efflux transcript and increases its stability. Strains with the upregulated small RNA grew in the presence of rifampicin where control strains did not. This is regulatory tolerance rather than a target mutation.

Why does the absence of Hfq matter for this field

In many bacteria, trans-encoded small RNAs rely on the Hfq chaperone to find and bind their messenger RNA targets. No Hfq or Hfq homolog has been identified in this pathogen, which means the interaction rule is unknown. When the targeting rule is unknown, computational predictions of which messenger RNA a small RNA regulates carry more uncertainty, and the review marks many such targets as unconfirmed.

What did the MTS1338 and MTS0997 knockouts show

Double knockouts grew more slowly than wild type in alveolar macrophage culture medium, which suggests these RNAs support intracellular survival. In a susceptible mouse strain, infection with the mutants produced less lung tissue infiltration, fewer neutrophils and macrophages, and reduced granuloma formation. Those are causal phenotypes from deletion rather than correlations from expression data, which makes them the strongest evidence in the review.

How do host microRNAs change the macrophage response

Host microRNAs act post-transcriptionally on immune signaling. One example, miR-1236-3p, is elevated in active and latent infection, rises in infected macrophages, and was shown by reporter assay to target the TLR4 transcript, reducing TLR4 and downstream MyD88 signaling. Long non-coding RNAs act more broadly on macrophage survival, apoptosis, autophagy, pyroptosis, and cytokine production, and circular RNAs act partly by sponging microRNAs.

How accurate are bacterial small RNA tests for tuberculosis

A panel of three bacterial small RNAs reported a combined area under the curve of 0.9104 with 92.31 percent specificity and 88.24 percent sensitivity in blood. A combination of a bacterial small RNA with a vesicle microRNA reached 0.992 with 97.62 percent sensitivity and 100 percent specificity. The review notes that these panels still underperform standard molecular testing on both sensitivity and specificity, so the realistic role is a sputum-independent adjunct.

How do RNA markers compare with established tuberculosis tests

One long non-coding RNA reached an area under the curve of 0.828 with 76.24 percent sensitivity and 72.93 percent specificity, against an interferon gamma release assay at 0.860 with 91.67 percent sensitivity and 74 percent specificity. The same comparison put a serological antibody test at 0.691 and a skin test at 0.738. Using an established clinical comparator is more informative than comparing with healthy controls alone.

Why are circular RNAs considered stable markers

Circular RNAs have a covalently closed loop structure with no free ends, so they resist degradation by RNA-degrading factors and are more stable than linear transcripts. That stability is the reason they attract interest as diagnostic markers, because a marker has to survive sample collection, transport, and storage before it can be measured. Bacterial small RNAs packaged into exosomes share that logistical advantage.

Can these markers distinguish latent from active tuberculosis

Only partly. A host microRNA or long non-coding RNA that reflects immune activation changes in both states and can also move with other inflammatory conditions, so specificity against latent infection and against unrelated disease is the weak point. Pathogen-derived small RNAs avoid that problem in principle because they should disappear when the pathogen does, but they lose sensitivity. The review presents both limitations without claiming either is solved.

What evidence is missing before clinical use

Prospective validation in more than one population, replication across ancestry, endemic exposure, and HIV status, standardized sample handling, and comparison against established clinical pathways rather than against healthy controls. Most current findings come from small single-center studies, many interactions are computational predictions rather than validated pairs, and no non-coding RNA marker has been validated for treatment monitoring or is in routine use.

Does this change how rifampicin susceptibility should be read

It adds a caution rather than a replacement. A minimum inhibitory concentration measured on a reference strain answers a question about a fixed genome under defined conditions. A regulatory loop that raises efflux capacity within hours of exposure is not captured by that measurement, and no current pharmacopoeial method is designed to detect it. Genotypic and phenotypic results therefore answer different questions about the same isolate.

Zhengzhou Minzhong Pharmaceutical Co Ltd | Leading Rifamycin API Manufacturer Since 1994

My Instagram