These new tests, conducted in a single tube within minutes, could enable at-home testing for various diseases. By incorporating CRISPR technology, the test achieves high reliability by distinguishing between false and true positives.
Scientists at the University of Florida have harnessed the power of artificial intelligence (AI) to develop a simplified test capable of detecting both
Led by Piyush Jain, a professor of chemical engineering at UF,
To achieve this, Jain's team revitalized a one-pot reaction system, with the entire testing process takes place in a single small test tube. Leveraging a technology called reverse transcription-loop-mediated isothermal amplification (RT-LAMP), these tests amplify specific sections of a virus's genetic material and generate a visible signal upon detection. Interpreting the test results can be as straightforward as observing a color change or utilizing a small device designed to detect alterations in the test tube.
Although the FDA has authorized some at-home one-pot tests for COVID-19, they exhibit a relatively high rate of false positives, diminishing their reliability. Jain's team incorporated
By utilizing CRISPR's ability to target specific genetic sequences, the test distinguishes between false and true positives. A positive result is only indicated when the genetic sequence unique to the targeted virus, such as hepatitis C, is present.
CRISPR-based diagnostic platforms have gained prominence during the COVID-19 pandemic, offering a faster and potentially more efficient alternative to the time-consuming quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR) method.
Recent advancements in CRISPR-based tests have led to the development of one-pot detection assays, combining pre-amplification steps like RT-LAMP with CRISPR-Cas reactions. However, one challenge has been the lack of Cas enzyme thermostability at the optimal temperature range for RT-LAMP.
Jain and his team engineered thermostable variants of the BrCas12b enzyme, expanding the range of detection and enabling more versatile primer design for improved diagnostic assays.
To achieve enzymatic activity at high temperatures and cover the optimal range of RT-LAMP, de novo structural analyses were employed to engineer a more thermostable BrCas12b variant. Predicted structures of the wild-type BrCas12b were generated using AlphaFold and SWISS-MODEL, which provided insights into potential mutations that could enhance thermostability. Stability prediction tools, such as HotSpotWizard and DeepDDG, aided in identifying mutations that improve stability, catalytic activity, and specificity.
However, a challenge arose due to the differing temperature requirements of RT-LAMP and CRISPR technologies. While RT-LAMP necessitates 150 degrees Fahrenheit, CRISPR functions optimally at 100 degrees. Overcoming this hurdle, Jain's team is working to develop a CRISPR system capable of withstanding higher temperatures.
Through the analysis of a heat-tolerant CRISPR enzyme discovered in a heat-loving bacterial species, the researchers employed AI tools to identify modifications that would enable the enzyme's functionality at 150 degrees. Several changes suggested by AI algorithms were tested in the lab, ultimately resulting in 4 modifications that allowed the enzyme to operate at the desired temperature.
Termed SPLENDID (Single Pot, Long Enzyme, No Discrimination), the simplified test was validated using clinical samples from patients with hepatitis C or COVID-19. The test exhibited 97% accuracy for SARS-CoV-2 and 95% accuracy for the most prevalent strain of hepatitis C globally.
Although its performance against less common hepatitis C variants was suboptimal, Jain believes that simple adjustments to the test will swiftly improve its accuracy. The assay exhibits high sensitivity and specificity, providing results in as little as 20 minutes.
The engineered thermostable BrCas12b variants pave the way for more versatile and efficient CRISPR-based diagnostic assays. By expanding the temperature range of detection, these advancements contribute to improved primer design and enhance the sensitivity and specificity of nucleic acid detection for infectious diseases.
This groundbreaking development not only brings us closer to reliable at-home testing for multiple diseases but also demonstrates the tremendous potential of AI in accelerating the refinement of diagnostic methods.
