CSB iGEM 2006 Poster

Anyone that's familiar with synthetic biology has probably also been bombarded with Keasling's' malaria work and tumor targeting bacteria developed in Voigt's lab.  While his bacteria provided inspiration for this project, they just aren't very smart (at least now), making them unusable.  But what if bacteria were smarter?  What if bacteria could tell the difference between the surface of a cancer cell and the surface of a healthy cell?  What if you could tie population density to cancer cell invasion, but also to effective killing?   And what if you developed a way to get wayward "unsafe" bacteria to commit suicide? 

Targeted Localization

Many modern cancer therapies take advantage of the fact that cancer cells have different proteins or antigens expressed on their surface.  A common approach has been to develop substances that target these markers, namely other proteins called antibodies often linked to a toxin.  These treatments make companies like Genetech big money.  But they often times fail either because delivery is not robust or the accompanying toxin does significant damage in route to target. 

Until very recently our idea of putting antibodies on the surface of bacteria would have been ludicrous.   Conventional antibodies are extremely large and complicated proteins.  Trying to display them on the surface of a cell is tantamount to painting van Gogh's Starry Night while strapped to the hood of a moving car.  However recently the Duke team partnered with Ablynx, a company who produces single domain antibodies called nanobodies, that are much smaller and less complex.   Currently we are attempting to fuse a carcinogenic embryonic antigen nanobody to an auto transporter scaffold protein to allow for display.  At this time identical work is being done at the Center for National Biotechnology in Madrid.

Targeted Localization: The carcinogenic embryonic antigen antibody is expressed on the surface of the bacteria through an auto transporter display system. IgA protease acts as a scaffold onto which nanobodies or scFv antibodies can be attached, allowing for modularity.  The S*Tag protein is used to provide a visual confirmation of surface display.  The PelB leader sequence is necessary to direct the protein to the membrane of the cell and is subsequently cleaved off.

 
Discriminate Killing
If you want to kill something its nice for it to be effective and scalable.  For instance if you want to kill all the mice in your house you'd like your solution to be lethal enough to kill all the mice without eradicating your family members.  Cytosine deaminase (a gene not found in humans) is ideal because it converts a non-toxic molecule to a readily diffusible killer.  The more precursor drug you take the more toxic substance you make.  We are working with a construct from Margaret Black at Washington State University that is 11 times more kinetically favorable towards the substrate molecule.  We are also experimenting with invasion from Voigt's Lab and Ralph Isberg's Lab at Tufts to also allow invasion of tumor cells, since killing from the inside out might be more favorable.

Both genes are linked to quorum sensing to coordinate toxic molecule production and invasion with an expected higher population of "sticky" bacteria surrounding the tumor.  Interestingly it appears that the quorum sensing molecules (AHL) and the toxic small molecules produced by cytosine deaminase both inhibit the same protein (thymidylate synthase) potentially allowing for increased killing efficiency.

Discriminate Killing:Both invasin and cytosine deaminase are put under the control of a receiver device.  Invasin which comes from the pathogenic bacteria Yersinia pseudotuberculosis allows for bacteria cells to invade mammalian cells via b-1 integrin binding. While bacteria capable of invasion   Cytosine deaminase converts the non-toxic small molecule 5-fluorocytosine into the toxic compound 5-fluorouracil allowing control over cancer killing in a dose dependent manner.

Regulated Suicide

While controlling the amount of killer molecules is helpful, uncontrolled bacterial growth can cause septis and be equally hazardous.  We worked with Lingchong You, a pioneer in population control circuits in bacteria, on a circuit that would cause the cell to self destruct when away from the tumor environment (low cell density).

The circuit uses a modified naturally occurring addiction system.   The controlled cell death (Ccd) addiction system plays a crucial role in the stable maintenance of the Escherichia coli F plasmid. It codes for a stable toxin (CcdB) and a less stable antidote (CcdA).   If the plasmid containing the two genes is lost the CcdB out lives CcdA and kills the cell, preventing plasmid loss in the population, confirming an evolutionary advantage.  In our system CcdA is linked to quorum sensing (population dependent) and CcdB is activated externally with IPTG.  At a low population not enough CcdA is produced to save the cell, so it kills itself at low population.  However at a high population the AHL concentration is high enough that sufficient CcdA is produced, allowing for survival.  In vitro we witnessed increased toxicity at higher IPTG/CcdB concentrations and lower amounts of AHL/CcdA.  The antitoxin CcdA rescues the cell when the cells are at a high population in the tumor environment.  Outside the tumor environment cells at low concentrations cells are vulnerable to toxic ccdB.


Regulated Suicide: Lux sender receiver devices are used to link CcdA antitoxin production with quorum sensing.  In sufficiently high concentrations enough CcdA is produced to allow for the survival of the cell.  However at lower concentrations the cell is killed by CcdB.   This allows for bacteria with a high affinity for cancer cells to clump around the tumor and survive while causing bacteria that do not localize to the tumor or leave the immediate tumor environment to commit suicide.

Results


Regulated Suicide:  At a low concentration cells are killed via production of the toxic protein ccdB induced via the addition of IPTG.  However at high cellular density, enough AHL is available to activate production the antitoxin CcdA.  In order to confirm that higher AHL concentration was in fact the reason for increased growth, we added AHL externally.  The high dose of AHL added to a low density starting population resulted in population growth that exceeded that of the originally high density population.  The results confirm that we have an working circuit that diminishes growth capabilities of cells in low density environments.