Biosensor bacteria for tumor detection

Genetically engineered biosensor bacteria could allow us to detect colon cancer without the need for invasive procedures.

Detecting abnormalities in the gut can be challenging. The gut in an inaccessible and complex environment, which makes detecting changes difficult without invasive techniques.  This holds back research into the gut environment, and also impacts the diagnosis of diseases such as colon cancer. Colon cancer is the third most commonly diagnosed cancers worldwide. As with all cancers, early diagnosis is key to achieve good long-term patient outcomes especially as colon cancer remains the fourth leading cause of cancer-related death. 

Diagnosing colon cancer usually involves a colonoscopy, which is an invasive procedure that is often delayed or avoided by patients. Less invasive procedures do exist, but they are less accurate. One such technique involves testing stool samples for tumour DNA that has been shed into the colon by the tumour and has passed out of the body in stool. Unfortunately, shed tumour DNA is quickly broken down and degraded in the colon environment which reduces the accuracy of this screening technique.

Synthetic biology could provide us with a better solution. “Biosensors” are living genetically engineered bacteria that are capable of detecting a variety of substances, including endogenous (foreign) DNA. Bacteria can be deployed into the gut environment itself, and therefore can come into close proximity with tumours. This close proximity means that any shed tumour DNA has less opportunity to degrade before being detected by the biosensor bacteria. Such a detection technique therefore has the potential to provide greater accuracy than the testing of stool samples and provide valuable diagnostic data without the need for an invasive procedure. 

In a collaboration between the University of California (UCSD) and multiple Australian institutions, researchers developed a bioengineered bacteria that can detect shed tumour DNA while located within the colon. The bacterium chosen for this purpose was Acinetobacter baylyi. A. baylyi is a popular model organism because of its competency; it can take up foreign DNA from its surroundings and then incorporate it into its own chromosomal DNA. While many bacteria have to be subjected to artificial laboratory conditions in order to induce competency, A. baylyi is naturally competent, and actively transports foreign environmental DNA into its cellular environment. 

The researchers genetically modified A. baylyi using CRISPR to express an antibiotic resistance gene to a specific antibiotic only when it took up environmental DNA containing a cancer-associated mutation in a chosen oncogene (a gene that has the potential to cause cancer). If it took up normal DNA without any cancer-associated mutation, the bacteria would not express the resistance gene and so would die on the application of an antibiotic. 

Researchers showed the bacteria was capable of detecting the target cancer-associated sequences both in culture and in mice where tumours with the relevant sequences had been inserted into their digestive tracts. This shows promising clinical applications. In human patients, the detection technique could be altered such that various samples (including stool, blood or urine) could be collected rather than needing to take a sample from the gut environment itself. 

The technique still has limitations. One obvious drawback is detection is only possible with specific known sequences, so the biosensor bacteria will not be able to detect unknown mutations or those that arise will low frequency. Fortunately, colon cancer frequently arises with a small set of common mutations, making it an ideal cancer for such a screening technique.  The technique also relies on the shedding of tumour DNA into the digestive tract, so tumours that are located elsewhere could not be detected. 

The research team also sees more promising applications in the future. Instead of producing an antibiotic resistance gene on the detection of tumour DNA, the biosensor could instead produce a therapeutic agent to kill the tumour cells. This would provide a localised treatment at the site of the tumour, reducing off target effects.  While this technique is only at a preliminary stage, the promising in vivo results demonstrate the suitability of biosensor bacteria for application in humans and the potential to supersede invasive diagnosis techniques. 


4.  - actual study