Asian Scientist Magazine (Aug. 14, 2023) — For researchers working on long-lasting data storage, DNA is an attractive medium. It carries genetic information with high fidelity across generations and through multiple cell divisions in multicellular organisms. It is so sturdy that DNA fragments have been found in fossils as old as over two million years.

Hard drives save data in a binary code, with information coded in a string of 1s and 0s. DNA, on the other hand, has a four-letter nucleotide code—ATCG. Storing any information in DNA requires translation of the binary codes into a string of nucleotides followed by synthesis of the DNA sequence. This way, scientists have stored a movie, a music album, and even all of Wikipedia in DNA.

Encoding data directly into DNA eliminate the need for DNA synthesis but such efforts have been ineffective in encoding 2D information like images. In a recent study published in Nature Communications, researchers from the National University of Singapore have been able to do just that: Captured images directly into the DNA of living bacteria using light.

Explaining their choice of method, Cheng Kai Lim, a graduate researcher and the first author of the paper said that “light is a cheap source of information, it’s easily programmable, and you don’t need any fancy equipment.”

The team’s living camera, named BacCam, makes use of two routinely used techniques: optogenetics and DNA barcoding. Optogenetics is leveraging light-sensitive molecules to manipulate cells; DNA barcodes are short specific sequences of DNA that act as tags.

The researchers used 96-well microtiter plates having 12 rows and 6 columns of tiny wells. A monochrome image of resolution 12 X 8 pixels was projected onto the plate, each pixel corresponding to a well. For “light” pixels, a light-activated optogenetic system makes a cut at a particular location in the bacterial DNA in those wells. Additionally, samples in each well were tagged with a different barcode.

Well-specific barcodes, or well-codes as the team called them, ensured that samples from each well could be pooled together for storage. DNA sequences contained information about which pixel they corresponded to and what the state of the pixel was (light or dark). Then the pooled DNA was sequenced to retrieve the image.

Demonstrating the durability of this method, the researchers were able to retrieve images after a week in different harsh conditions. These included freezing cells at -20 °C, keeping them in an oven at 60 °C, and drying them into a powder.

Next, the researchers looked at the possibility of storing multiple images together, with one microwell plate for each image. They pooled together samples from the five plates, after tagging all samples in each plate with an image-specific barcode. In this case, each DNA sequence contained information about i) which image it corresponded to, ii) the location of the pixel in the image, and iii) its light or dark state.

Sequencing the pooled DNA reconstructed all five images with over 90% of the pixels in each image being read correctly. The image retrieval was highly accurate even when the samples were diluted to a factor of 100.

In future work, the research team plans to capture colored pictures. Colors are generally coded in RGB (red, green and blue) values, with different values of the three primary colors. Putting together optogenetic circuits that react to the red, green, and blue lights, could allow them to capture the full spectrum of light, said Lim. Practically, this would require highly sensitive optogenetic circuits that are able to encode levels, and not just the presence, of light.

The next frontier for DNA data storage could be capturing video directly into DNA. “People have developed ways to encode information in a temporal manner. We are exploring how those methods can be incorporated into our system,” Lim added.

—

Source: National University of Singapore; Images: Yipei Lieu/ Asian Scientist Magazine