Biosensors for Colorectal Cancer

The human gut is awash in a sea of microbes that quietly ferment fibers, produce vitamins, and exchange information with the immune system.¹ Now, scientists are tasking bacteria with yet another job as they spelunk their way through the digestive system: cancer detection.

An international team of researchers engineered a bacterial biosensor capable of identifying a cancer-associated DNA mutation, which they published in the journal Science.² The research team included molecular biologists Robert Cooper and Jeff Hasty of the University of California, San Diego and bowel cancer researchers Josephine Wright and Susan Woods at the South Australian Health and Medical Research Institute, and Daniel Worthley at the Colonoscopy Clinic. The study authors hope that this technology will one day aid in the early diagnosis of colorectal cancer, one of the most common causes of cancer-related death globally.

While scientists have previously engineered bacteria to detect inflammation or bleeding in the gut, this is the first bacterial biosensor that detects a specific DNA sequence from host tissues. To accomplish this feat, scientists leveraged Acinetobacter baylyi’s ability to take up extracellular DNA and integrate these sequences into its own genome.

Taking these naturally competent bacteria, detecting DNA changes, and then using that as a biosensor is a really cool advance.
-David Riglar, Imperial College London

“Using living bacteria to sense things in the gut and detect disease is something that I find very exciting,” said David Riglar, a microbiome researcher at Imperial College London who was not involved in the study. “Taking these naturally competent bacteria, detecting DNA changes, and then using that as a biosensor is a really cool advance.”

In this study, researchers wanted to engineer A. baylyi to detect a common colorectal cancer marker: a mutation in codon 12 of the KRAS gene. “At the time, it seemed like a fairly far-fetched idea,” recalled Worthley. By bringing together an interdisciplinary team with expertise in synthetic biology and animal models of colorectal cancer, the researchers achieved this lofty goal.

In their early proof-of-concept experiments, the researchers genetically tinkered with both A. baylyi and the tumor organoids that they wanted the bacteria to detect. They engineered tumor cells with a functional copy of the antibiotic resistance gene, kanR, flanked by KRAS homology arms. The bacteria had matching KRAS homology arms, plus two stop codons that prevented the expression of kanR. When the bacteria gobbled up the donor tumor DNA, the homology arms aligned the DNA sequences and the bacteria integrated the functional kanR into their own genomes, enabling them to grow on antibiotic-laced plates.

To create bacteria that specifically detected mutant KRAS, the researchers harnessed the bacteria’s own CRISPR-Cas machinery, directing these molecular scissors to chop up wild-type, but not mutant KRAS. This would kill any bacteria that acquired the wild-type KRAS.

The researchers then tested these bacteria against colorectal cancer organoids with and without the engineered donor DNA. Only the bacteria cocultured with the engineered tumors acquired antibiotic resistance, showing that the sensor bacteria could discriminate between normal and donor tumors.

See Also "Mutant RAS Proteins Team Up for Oncogenicity"

Next, the researchers tested the biosensors in vivo by administering the bacteria via enema to three groups: mice without tumors, mice with normal colorectal tumors, and mice with the engineered colorectal tumors. Again, only the biosensors administered to the mice with the engineered tumors grew in the presence of the antibiotic, confirming that the bacteria could be used to signal the presence of engineered colorectal cancer in mice.

While these data were promising, human colorectal tumors don’t come engineered with a perfectly placed antibiotic resistance gene for bacteria to acquire. So, the researchers adjusted their strategy to detect natural tumor DNA with the KRAS mutation. This time, they placed a repressor gene inside the KRAS homology arms. This gene prevented the expression of a downstream kanR gene.

When the bacteria swapped their KRAS DNA for the tumor’s KRAS, the repressor was lost, allowing the antibiotic resistance gene to be expressed. As before, wild-type KRAS was targeted for destruction by the CRISPR-Cas system. In vitro, these new biosensors discriminated between mutant and normal KRAS by surviving and becoming antibiotic resistant only in the presence of the cancer-associated mutation. The team named this technique CATCH for cellular assay for targeted, CRISPR-discriminated horizontal gene transfer.

Despite these preliminary successes, Riglar urged caution. “It's important not to run too far ahead in terms of thinking that these systems are ready to go into the clinic,” he said.

“This is absolutely not the endpoint,” Worthley agreed.

The researchers are currently working on strategies to improve the biosensor’s sensitivity to natural tumor DNA in the complex environment of the colon. Due to concerns about administering antibiotic-resistant bacteria to humans, they are also developing other ways for the biosensors to signal the presence of mutant KRAS. To be commercially viable, the biosensor bacteria must be delivered orally, meaning that they will need to survive their journeys through the digestive system and be able to report their findings on the other side.

Since we’ve engineered all the sophistication within the cell, we don’t need such a sophisticated laboratory outside the cell.
-Daniel Worthley, Colonoscopy Clinic

Eventually, however, Worthley hopes that these biosensor bacteria will one day be used as point-of-care diagnostics in remote or low-resource areas such as the Australian outback. “Since we’ve engineered all the sophistication within the cell, we don’t need such a sophisticated laboratory outside the cell,” he said.

Researchers hope for broader applications as well. Instead of turning on an antibiotic resistance gene when they sense tumor DNA, for example, the bacteria could be engineered to turn on production of a genotype-specific small-molecule therapeutic, delivering treatment precisely where it’s needed. Bacteria could be engineered to detect and respond appropriately to various oncogenic mutations, or even difficult-to-treat infections like Clostridium difficile. Worthley sees this potential to marry diagnosis and therapy as the major advantage of these engineered bacteria.

Conflict of interest statement: J.H., D.W., and S.W. have equity in GenCirq Inc., which focuses on cancer therapeutics. D.W., J.H., R.C., S.W., and J.W. have a provisional patent application on this technology.