One Vial, Two Targets: Portable CRISPR-Based Tuberculosis Diagnosis with 100% Sensitivity

A paper about a portable tuberculosis test that can be done directly from sputum, without complex equipment, was published in Science Advances. Isothermal DNA amplification and CRISPR reading are combined in a single test tube; two conservative M. tuberculosis insertions (IS6110 and IS1081) are tested at once, plus a human gene as an internal sample control. In a small “blind” test on clinical samples, the test gave 100% sensitivity (6/6) and 100% specificity (7/7) compared to culture; the detection limit is ~69–81 CFU/ml in simulated sputum. The result can be seen on a paper test strip, and the reagents are lyophilized (storage without a “cold chain”).

Background

According to WHO, TB will kill about 1.25 million people in 2023; TB has once again become the leading infectious cause of death, with an estimated 10.8 million cases. This makes accessible and rapid diagnostics critical, especially in resource-limited settings.

• Current tests are a compromise between accuracy, speed, and price. The “gold standard” culture is very accurate but slow; microscopy is fast but insensitive; PCR platforms like Xpert MTB/RIF Ultra are significantly more sensitive (LOD ≈ 15.6 CFU/ml), but require expensive equipment and cartridges, which limits coverage.
• Why isothermal and CRISPR. Isothermal RPA amplification works at 37–42 °C and does not require thermal cyclers — convenient “in the field”. CRISPR reading (Cas12/13) adds species specificity and allows visual lateral flow (“strip”) without complex optics. In total, this is the path to portable and cheap PVR tests.
• Why two MBT targets at once (IS6110 + IS1081). IS6110 is a multicopy insert of the M. tuberculosis complex, but some lines have few copies, and tests for IS6110 alone may “miss.” Adding a second IS1081 insert reduces the risk of false-negative results.
• Why internal human control. Breath samples can contain inhibitors and "bad" samples. Endogenous human control (e.g. RNase P/genomic DNA) confirms that the material is adequate and the reaction is not suppressed - otherwise the result cannot be considered negative.
• The one-pot format is important in itself. It reduces steps, reduces the risk of contamination, and facilitates work outside the lab. Such approaches for TB have already shown viability; the new work extends the idea to a dual-target format with internal controls, plus demonstrates lyophilization of reagents and a strip reader.
• What is the value of the current article? The authors demonstrated detection directly from sputum after the most simplified sample preparation, the detection limit at ~70–80 CFU/ml in simulated sputum, and 100% sensitivity/specificity on a small blind set of clinical samples — a good “tech demo” for further multicenter validations.

How does this work

• The scientists combined RPA (rapid isothermal amplification of genetic material at 37 °C) with the "cutting" enzymes Cas13a/Cas12a. Having selected the guide RNAs, they configured the system to target two MBT targets simultaneously, and in parallel to human DNA (checking that there is material in the sample at all and the reaction has not "stalled").
• All the chemistry goes in one test tube; after incubation, the result is read either with a fluorimeter or on a lateral flow strip - essentially like an express test.
• Sputum processing has been simplified to heating and a brief centrifuge—without nucleic acid extractors. For the most limited conditions, the authors even discuss manual alternatives to the centrifuge.

What the tests showed

• Detection limit: 69.0 CFU/ml (strain H37Rv) and 80.5 CFU/ml (BCG) in "spike" sputum. No cross-reactivity with other bacteria/fungi was detected.
• Clinical setting (blinded samples): on 13 samples from real practice - 100% sensitivity (6/6) and 100% specificity (7/7) relative to seeding. For comparison, on the same kit GeneXpert Ultra showed 100%/86%, respectively.
• Technical nuances: of the two reading options, Cas13a worked better (for the "one-pot" format, it is more sensitive than Cas12a). Plus, parallel testing of two Mtb targets reduces the risk of false results.

Why is this necessary?

Today’s tests are a compromise between accuracy, speed, and availability: culture is very accurate but slow; swabs are fast but inaccurate; PCR systems like GeneXpert are accurate and fast but require expensive instruments and cartridges. The new CRISPR approach aims to close the gap: field diagnostics that operate at 37°C, with paper strip readings, and the potential to store reagents without refrigeration.

Limitations and what's next

This is an early demonstration on a small clinical set - large, multi-center tests are needed (including HIV coinfection, in children, with paucibacillary forms and different MBT lines). The authors separately note that in the "field" version itself, the tabletop centrifuge will have to be replaced with a fully manual solution. But the architecture itself - "two targets + internal control" in one test tube - has already proven its operability and provides a clear path to refinement for mass use.

Source: Alexandra G. Bell et al. A streamlined CRISPR-based test for tuberculosis detection directly from sputum, Science Advances, online August 6, 2025 (Vol. 11, Issue 32). DOI: 10.1126/sciadv.adx206