Published online 26 September 2011 | Nature | doi:10.1038/news.2011.557

News

Bacteria encode secret messages

Undercover E. coli act as updated invisible ink.

E. ColliThe name's Coli. Escherichia coli.photolibrary.com

For millennia, people have written secret messages in invisible ink, which could only be read under certain lights or after developing with certain chemicals. Now, scientists have come up with a way of encoding messages in the colours of glowing bacteria.

The technique, dubbed steganography by printed arrays of microbes (SPAM), creates messages that can be sent through the post, unlocked with antibiotics and deciphered using simple equipment. It is described in the Proceedings of the National Academy of Sciences1.

For years, scientists have been able to encode messages in biological molecules such as DNA or proteins. Biologist Craig Venter wrote his name, along with several quotations, into the DNA of the partially synthetic bacteria that he unveiled last year2.

Manuel Palacios, a chemist at Tufts University in Medford Massachusetts, and his colleagues took a simpler approach by encrypting messages using seven strains of Escherichia coli bacteria. Each one was engineered to produce a different fluorescent protein, which glows in a different colour under the right light.

"We really wanted to use easily observable traits," says Palacios. Techniques such as Venter's require sophisticated equipment to sequence the DNA and unlock the message. "In our case, light-emitting diodes and an iPhone would do," notes Palacios.

Colonies of bacteria are grown in rows of paired spots, every combination of two colours corresponding to a different letter, digit or symbol. For example, two yellow spots signify a 't', whereas an orange and a green spot denote a 'd'. Once grown, the pattern of colonies is imprinted onto a nitrocellulose sheet, which is posted in an envelope. The recipient can use the sheet to regrow the bacteria in the same pattern and decipher the message.

By choosing the right E. coli strains, people could send messages that appear after specific periods of time, or slowly degenerate like the self-destructing memos from the television programme and film series Mission: Impossible.

Palacios has also developed ways to turn antibiotics into keys that unlock the hidden messages, by linking the genes for the fluorescent proteins to the ability to resist specific antibiotics. To prove the principle, he created a SPAM sheet that, when exposed to the antibiotic ampicillin, read "this is a bioencoded message from the walt lab @ tufts university 2011". If he used the drug kanamycin instead, the bacteria glowed in different colours, translating to "you have used the wrong cipher and the message is gibberish".

Coded colonies

The project was funded by the US Defense Advanced Research Projects Agency (DARPA), which challenged labs to find ways of transmitting coded information using chemical signals rather than electrical ones. But some cryptography experts think that the bacteria would never make it out of Q's gadget lab.

ADVERTISEMENT

"It's a clever concept, but as a method of hiding information it's not especially practical," says Meredith L. Patterson, a security researcher at network-security firm Red Lambda, based in Orlando, Florida. There are so few antibiotics that a code-breaker could easily apply all of them to a SPAM sheet to uncover the message. "It's trivial to brute-force the solution," adds Patterson.

A clever cryptographer could get around this problem by encoding different but plausible messages with each antibiotic. "If kanamycin yields 'We will storm the beaches at Pas de Calais' and ampicillin yields 'We will storm the beaches at Normandy', Hitler still has to decide where he's going to send his troops," says Patterson.

Coded messages may appeal to DARPA, but in practice, says Palacios, "that's the last application this would be used for". He is more interested in the potential for watermarking genetically modified organisms with 'biological barcodes', to trace their provenance and prevent counterfeits.

The team is now trying to encode messages in more robust microbes such as yeast or spore-forming bacteria; and more complex organisms, such as plants, by exploiting variation in the shapes of leaves or the patterns of roots. "The more traits you have, the more information you can embed at the time," says Palacios. 

Commenting is now closed.