Walt et al from Tufts University describe a method of sending timed and on-demand released messages using printed arrays of bacteria. Or as they call it: Steganography by Printed Arrays of Microbes (SPAM). The researchers published their results in Proceedings of the National Academy of Sciences. The system they describe uses genetically engineered strains of Escherichia coli with added fluorescent proteins. These living organisms can carry a message and release the information when conditions selected prior match the environment. They use 7 strains containing fluorescent proteins of different wavelengths. This provides a septenary numeral system that can be translated into text.
The sender encodes a message using a septenary code. Bacteria are printed on an agar plate and copied on a nitrocellulose membrane. The membrane is sealed and sent by post to the receiver. As long as the proteins are not expressed, the message will not be visible. The recipient stamps the membrane on agar which enables the bacteria to grow and form colonies again. When the appropriate inducer is added, the fluorescent proteins will be expressed and the colonies can be read by measuring the fluorescent reflection.
Several methods are described to secure the secret message. Bacteria tend to lose their extra genetic information (fluorescence) without selective pressure. The fluorescent proteins are added to the genome of E. Coli together with an antibiotic resistance gene. This creates a selective pressure and only bacteria that contain the resistance gene, and thus the fluorescent protein, will survive when grown on a plate with antibiotics. With no appropriate antibiotics the bacteria will lose their fluorescent properties and the message will be lost. Also the inducer, which in the experiments was IPTG, should be known by the receiver to express the fluorescent proteins. Without the proper inducers the proteins will not be expressed by E. coli as they are not essential for their growth.
Every code can be cracked, and some options to reveal the code include sequencing the genomes or running a molecular test like PCR looking for known fluorescent protein genes. The Tufts team is not the first hiding secret messages in bacteria. Craig Venter, known for sequencing the first whole human genome, once embedded the names of three authors of their publication in a synthetic genome of Mycoplasma genitalium as a watermark.
Abstract in PNAS: InfoBiology by printed arrays of microorganism colonies for timed and on-demand release of messages
(hat tip: Wired)
Image: Walt et al./PNAS