On every visit home, I am greeted and shortly after asked: “how is the cure for cancer coming?” I have yet to find a way to convince my family that I am a microbiologist, I work with bacteria, I don’t try to cure cancer. Well the day has come where microbiologists and virologists have indeed cured cancer. There are numerous examples of viruses that specifically target and kill tumour cells, and several clinical trials to examine this therapy in humans. In fact, my friends in grad school can proudly state they cured cancer with an otherwise, relatively harmless reovirus. One of them went on to start a biotech company using this technology and they are still active (www.oncolyticsbiotech.com).
These cancer killing bacteria and viruses will be the subject of a future post, but here we will focus on the use of bacteria as a cancer detection method. The central observations that motivated this study were that many bacterial species are known to preferentially accumulate and localize in tumours. And not surprisingly, many species of bacteria have been documented to kill diverse types of tumour cells. The general approach to show bacterial growth and localization on tumours is to inject bacteria intravenously into the blood, to facilitate greater access to the organs. This interesting feature of homing in on human tumours is an ability of non-pathogenic bacteria, as well as pathogens. To explain this widespread ability to colonize tumours, it is thought that tumours are sites of suppressed immune surveillance and that are rich in nutrients, both of which support localized bacterial growth.
The authors of this study decided to test the hypothesis that orally administered, harmless probiotic bacteria could pass through the stomach into the normal flow of blood through the hepatic portal vein, to gain access to the liver. The microbiome field has had great success in documenting the abundance and diversity of beneficial bacterial species that consider our body a playground. Not only do they colonize our skin, mouths, respiratory tract, gastrointestinal tracts and the crevices of our toes, but microbiome studies have also reported bacterial DNA signatures in places that were once thought to be sterile, including some organs, the urinary tract, and even the brain. A general conclusion of many microbiome studies is that an altered community of microbes somewhere in the body, is the cause of certain related or unrelated health issues. Although many of these studies identify correlations between healthy and non-healthy microbiomes, and the exact cause of disease is not yet proven, it has definitely opened up the floodgates for the use of probiotics. The well studied probiotic, E coli Nissle 1917 (EcN), was engineered and used as a candidate biosensor for liver tumours, when delivered orally.
A liver cancer model was selected because the liver is the primary site for metastasis of many other tumours. Also, bacteria that are ingested orally, can transit through the stomach or small intestines to access the hepatic portal vein, which supplies the liver with 75% of its blood supply. The resident liver macrophages or Kupffer cells, with help from neutrophils, are the liver’s recycling and filtering system. These resident immune cells clear bacteria present in the blood, and also recycle the iron in red blood cells.
The passage of normal flora microbes from the gut to the liver requires transit across the epithelial barrier of the gastrointestinal tract to the blood, and some researchers have shown that this generally occurs in a disease state. However, in this study using healthy, immunocompetent mice with in tact gut flora, the probiotic E coli Nissle 1917 translocated across the GI tract within 24 hours. EcN is well studied and tolerated as a probiotic.
The authors used a relatively common strategy to engineer EcN to express two different reporter systems: luminescence (luxCDABE) and lacZ. The bacterial-derived lux gene products produce light through enzymatic reactions, which is an ideal reporter system in bacteria. This reporter generates a strong signal in vitro and in vivo during infection or colonization of an animal. LacZ (β-galactosidase) is the work horse of bacterial reporters, and this enzyme cleaves the glycosidic bond between disaccharides into monosaccharides. Numerous substrates have been designed that produce end products upon cleavage that change colour, are fluorescent, luminescent, as well as other output measurements.
True to their hypothesis, oral EcN probiotics indeed targeted the tumours of the liver within 24 hours. The authors employed a unique in vivo imaging application that allowed for the detection of two distinct luminescence signals, one from the firefly luciferase enzyme from the tumour cells, and the other from the localized lux-expressing probiotic strain. The wavelength of light from the firefly luciferase (580-620 nm) varies from that produced by the bacterial luciferase (500-580 nm), allowing for two non-overlapping light signals, very similar to the use of two different fluorophores, such as green fluorescent protein (GFP) and red fluorescent protein (RFP). The distinct light signals were shown to co-localize, thereby confirming tumour-localized bacterial growth. The oral probiotics did not colonize healthy organs (spleen, kidney, brain, heart, lung) and had no ill health effects for up to 12 months. In contrast, the use in intravenous administration of bacteria could lead to colonization of other organs, the recovery of bacteria in the blood, and likely sepsis.
The engineered EcN expressed the lux and lacZ reporters from either a stable integration on the chromosome, or from potentially unstable expression plasmids. They used additional approaches to ensure plasmid stability without any antibiotics to select for plasmid maintenance, including the hok-sok (toxin-antitoxin) plasmid maintenance system and the actin-like protein Alp7 from Bacillus subtilis. This protein forms an actin-like filament that physically pushes plasmids to the poles, thereby ensuring segregation of the plasmid during cell division.
And how did this all come together as a diagnostic test for liver cancer? First, tumours were excised and shown to produce the second reporter lacZ at 5X above the background levels. The next step was to determine if by products of the LacZ enzyme could be detected in the urine, which indicates the presence of tumour-specific bacterial growth in the liver. The mice were administered the LacZ substrate intravenously after the tumour was colonized. Very small volumes of urine (1 μl) could be used to detect lacZ activity, in this case the production of light from cleaving the LuGal substrate. They also showed colour changes in the urine when using an alternative LacZ substrate that visibly changes the urine colour in a positive test.
The report concluded that this probiotic biosensor is sensitive, specific and safe to detect liver tumours in mice. The ultimate test was the detection of probiotic LacZ enzyme activity in the urine. The importance of this technology is that early detection of liver cancer is an unmet and pressing clinical need, and it may improve the current diagnostic methods that rely on imaging techniques. Since other tumours can be exposed to high bacterial concentrations, this approach may also work in detecting colorectal cancers. Although this approach has the advantages of being low cost, nonsurgical and nonradioactive, it will require significant regulatory approval and further testing in humans before approval.
Highlights: Oral administration of probiotic and simple urine test next day to detect liver cancer.
Simple synthetic gene circuits were used but could be improved to increase the signal strength and also the specificity.
Programmable probiotics for detection of cancer in urine. Science Translational Medicine. May 27, 2015, Vol 7, Issue 289. Tal Danino et al.