DNA repeats turn a brief cellular signal into a record that can be read later

Hundreds of DNA repeats made it possible to reconstruct a past cellular signal from a single cell

On August 7, bioRxiv published a preprint describing a system that records cellular signals in natural DNA repeats. In cultured human cells, the system distinguished four predefined signal levels from the record in a single cell. In mouse tissues, it accumulated records of gene activity over several weeks.

A brief cellular signal is difficult to capture. Researchers usually track changes by collecting samples at different timepoints, but the cells in each sample cannot then be measured again. A molecular recorder works differently. The researcher first links a selected signal to a DNA editor. When the signal occurs, the editor makes small changes in the DNA, which can later be read by sequencing.

In 2016, Reza Kalhor and colleagues developed a changing DNA barcode that recorded cell history at a single genomic locus. A guide RNA, which provides a short instruction to the Cas9 protein, directed Cas9 to the locus encoding that same instruction. A single locus can hold little distinguishable information, so earlier systems had to reconstruct signal history from many cells.

The authors of the new study use natural genomic repeats. One guide RNA directs the editor to hundreds of similar sites. A single pair of primers, which are short DNA fragments that initiate copying, then allows all these sites to be read by sequencing. In the human cell prototype, the researchers read 304 sites, of which 234 could be distinguished by sequence.

Each copy responds to the same signal at a different rate. Fast sites distinguish brief exposures but reach saturation sooner, while slow sites preserve differences during prolonged exposure. The combined pattern of changes therefore contains information about both signal strength and duration. When the authors exposed cells to four levels of a control signal, this pattern correctly identified the previous level in 640 of 747 individual cells, or 85.7% of cases. By comparison, sites present in one or two copies produced correct answers in about one third of cases.

For the mouse experiment, the researchers selected a repeat with 188 copies. They engineered activation of Fos and Npas4, genes that respond rapidly to cellular excitation, to initiate expression of the recorder guide RNA. Over three and ten weeks, the recorder accumulated distinct patterns in the liver, cortex, hippocampus, thalamus, and cerebellum. In a seizure model, the Fos recorder registered more editing in the cortex over four weeks. Analysis of individual sites detected differences in the cortex and cerebellum.