“Our medical computer might one day be administered as a drug and distributed throughout the body by the blood stream to detect disease markers autonomously and independently in every cell,” said Ehud Shapiro of the Weizmann Institute. “In this way, a single cancer cell could be detected and destroyed before the tumour develops. Even in a late-stage cancer this kind of treatment could reach every secondary growth, however small, and effectively terminate the disease.”

The input, output and software for the computer consist of DNA molecules, while DNA-manipulating enzymes act as the “hardware”. The computer checks the input for the presence of molecular markers: it looks for mutated mRNA sequences, under-expressed mRNA or over-expressed mRNA, all of which can be signs of cancer. The system samples the markers one by one, switching to a “no” state if a marker does not show a disease condition. This becomes the output of the computation, while if all the markers show a disease condition, the output is “yes”. If the output is “yes”, the computer releases a therapeutic short nucleic acid - either DNA or RNA - that can affect the levels of gene expression.

The system can also tailor the amount of treatment released according to the strength of the diagnosis. To achieve this, it uses a number of diagnostic molecules for each computation - some may show a positive output, while others show a negative one. Some of the molecules are designed to release the drug for a positive diagnosis and do nothing for a negative diagnosis, while others do nothing on a positive diagnosis but release a “drug suppressor” molecule for a negative diagnosis. The scientists say this acts as a “check-and-balance” system to prevent mishaps in the treatment.

The team showed that their devices could detect signs of prostate and lung cancer, and release appropriate treatment molecules. The researchers believe that the computer could analyse up to 15 markers - less than is possible with DNA chip techniques but still enough to provide a fairly precise identification. The system also has the potential to release proteins, long DNA molecules or organic molecules as the treatment drug, rather than short nucleic acids.

In this study the scientists tested the computers in a solution containing salts with a pH level for optimum enzyme function. “Unlike an intracellular or even intercellular environment, this solution does not contain the numerous proteins, nucleic acids, lipids and polysaccharides found in living organisms,” said Shapiro. “Some of these compounds may present a significant danger to the components of our computer in its present form or, conversely, our computer could inflict damage on these intracellular components.”

According to Shapiro, that means the prototype will need significant adjustments before it can be tested in an in vivo environment. “It may take decades before a diagnostic molecular computer that is completely safe and can function for long periods without breakdown will be produced and approved for use,” he said.

The scientists reported their work in Nature.