Nature, 2017, Vol 547, p345-349

By George Church out of the Department of Genetics and the Wyss Institute for Biologically Inspired Engineering at Harvard University

This project was funded primary by multiple NIH grants

DNA is an excellent medium for archiving data. Recent effortshave illustrated the potential for information storage in DNAusing synthesized oligonucleotides assembled in vitro1–6. Arelatively unexplored avenue of information storage in DNA isthe ability to write information into the genome of a living cellby the addition of nucleotides over time. Using the Cas1–Cas2integrase, the CRISPR–Cas microbial immune system storesthe nucleotide content of invading viruses to confer adaptiveimmunity7. When harnessed, this system has the potential towrite arbitrary information into the genome8. Here we use theCRISPR–Cas system to encode the pixel values of black and whiteimages and a short movie into the genomes of a population ofliving bacteria. In doing so, we push the technical limits of thisinformation storage system and optimize strategies to minimizethose limitations. We also uncover underlying principles of theCRISPR–Cas adaptation system, including sequence determinantsof spacer acquisition that are relevant for understanding boththe basic biology of bacterial adaptation and its technologicalapplications. This work demonstrates that this system can captureand stably store practical amounts of real data within the genomesof populations of living cells.

My takeaways:

1. While the application they demonstrated in the main manuscript is cool, the real value of this paper is in the details of how they were able to achieve data reading and writing using DNA. They conducted many more control and mechanistic experiments than the actual reading and writing of data experiments. They identified a viable pathway to using bacteria and DNA to store information. This will have many obvious applications in data storage and protection. 
2. While the technology is exciting, it is still very much in its infancy. I don’t expect to see this technology hit the markets for at least a decade.