Through assembly of the target binding module, anti-tumor drug synthesis module and regulatory module, we would introduce our genetically engineered Bacillus subtilis, B. hercules, as an anti-colon-tumor agent to provide direct tumoricidal effect during cancer therapy. It would be applied as an oral medicine which should retain viability in digestive tract and execute anti-tumor activity when and only when it binds to colon cancer cell.

Driven by pVeg constitutive promoter, RPMrel, the colon-tumor specific peptide should be displayed on the cell wall of B. hercules under the facilitation of LytC cell wall displaying system before oral ingestion. Once ingested, B. Hercules would be expected to reach colon after 8 hours without retaining in or colonizing the gastrointestinal tract. However, upon reaching the colon, it should hold up around colon tumor cells and colonize around them, and should await signals to produce anti-tumor molecule, BMP-2, which would be released to the local environment.
The expression of BMP-2 would be controlled by xylose inducible promoter. Since normally no xylose is present in colon, xylose would be taken orally or injected from anus. At that moment, B. Hercules should has successfully localized around colon tumor, and the arriving xylose should induce the production of BMP-2. The locally concentrated BMP-2 is expected to suppress colon tumor grow and trigger the apoptosis of the tumor. The BMP-2 level in colon should be low, and should have little adverse effect on normal colon tissues.
The concentration of xylose might overshoot, which might result in excessive production of BMP-2. If that happened, the excess xylose would induce high level of expression of the toxin YdcE. Once the amount of YdcE breaches the protection threshold from antitoxin YdcD, uninhibited YdcE will lead to the destruction of B. hercules. This would hopefully avoid the overdosing effect of xylose and the overproduction of BMP-2, and their subsequent adverse effect.