Want to know what a cell gets up to? How it responds to extracellular signals and/or a transient drug exposure? I think we all do! Looking at which and by how much different genes are transcribed in a cell is a good way to answer these questions, but so far, current approaches, namely RNA-sequencing, only provides us with a snapshot of the RNA present at one moment in time. To fully appreciate complex cellular behaviours, a history of the genes transcribed over time can provide a more in-depth understanding to answer these questions. Well, whilst the technology is still not quite there yet, Schmidt et. al 1 have made great progress in establishing a new method in bacteria that exploits CRISPR-technology to store RNA that is transcribed by the cell. Their findings, published recently in Nature, combine this CRISPR-mediated acquisition of RNA with deep sequencing to create Record-Seq, a tool that may have great future potential to keep track of transcriptional history from a population of cells.
A need for Record-seq
Genes transcribed by cells can already be detected by extracting RNA from cells and sequencing it (RNA-seq). But whilst this approach can give accurate readings on the level and identity of genes transcribed, it only provides a snapshot of what the cell was up to. Cells are extremely complex, transcribing gene X highly one moment, and gene Y the next. Record-seq may provide a way to detect these temporal changes in gene expression which could help identify key genes involved in orchestrating cellular responses. But how?
Exploiting the CRISPR- acquisition system
Whilst CRISPR is often associated with gene-editing, it is important to remember that CRISPR is a bacterial adaptive immune system, mediated by Cas proteins. The ‘immune system’ works on the bacteria storing DNA fragments in a CRISPR array that can be transcribed to recognise foreign viruses, enabling the viral DNA to be targeted and destroyed. I touched upon this mechanism in my first ever blog post (2), but will briefly recap.
There are three key steps;
- Integration of (foreign) DNA fragments into a CRISPR array (ACQUISITION). Spacers are added in a temporal order; most recent nearest the leader sequence (Figure 1). This stage is mediated by the proteins Cas1 and Cas2.
- Transcription of the spacers with the CRISPR repeats. In complex with Cas proteins, the spacer sequence is used to recognise matching DNA fragments (BIOGENESIS)
- Recognised DNA is destroyed (INTERFERENCE). This is the stage exploited in genome editing.
(For more information, check out this video which explains very well ‘What is CRISPR’ – https://www.youtube.com/watch?v=MnYppmstxIs )
The key feature exploited by Record-seq is the fact that new spacers are acquired in temporal order – new spacers are added proximal to the leader sequence (Figure 1). This means that older spacers can be distinguished from more recently integrated spacers. Cas1-Cas2-mediated spacer acquisition has been exploited before, as seen in Shipman’s groups work (3), where they electroporated DNA fragments into a population of bacteria over time. The fragments were stored in the CRISPR array across the bacterial population. The array could then be sequenced, retrieving the fragment sequence information. In one case, Shipman’s group had encoded 5 frames of the classic movie by Muybridge of the galloping horse into a series of DNA fragments. This information was successfully retrieved after using the CRISPR acquisition system.
The key steps then are; ENCODE, STORE, RETRIEVE
However, these methods depended on electroporating external DNA sequences one by one into a population of bacteria. To record a cells transcriptional history, a mechanism to encode RNA present inside a cell is required. Schmidt’s team overcame this by exploiting a naturally occurring RT-Cas1 protein*, which in combination with Cas2, can convert RNA to DNA before integrating the DNA into the CRISPR array as a new spacer. Newly acquired spacers can then be determined by sequencing. The name of this technique is Record-seq.
*RT = reverse transcriptase, and is the component that converts single stranded RNA to double stranded DNA.
More RNA, more spacers
To get integrated into the CRISPR array, the RNA must be recognised, converted to DNA and inserted into DNA. It makes sense therefore that RNA that is more abundant is more likely to get integrated. Nonetheless, the team proved this to be the case by increasing the transcriptional output of certain genes through an inducible system – after Record-seq was performed, they saw a corresponding increase of those genes in spacer acquisition.
One of the aims of Record-seq was to provide a technique that could rival RNA-seq by providing a read out for changes in gene expression over time. By exposing cells transiently to paraquat and comparing results to unexposed and continuously exposed cells to paraquat, the team showed that Record-seq, but not RNA-seq, was able to detect an influence on transcription when there was only a transient exposure to paraquat.
Record-seq can store a cells history, but does it have a future?
This theory behind this technique is just one of the amazing applications of CRISPR systems. However, a major limitation with Record-seq is the low efficiency of spacer acquisition. Only a minor fraction of the bacterial population acquired more than one spacer. Compared to the thousands of different RNAs that could be present, this results in a huge lack of data of the cells transcriptome. This problem would only be extrapolated further if the technique could be used in mammalian cells which have many more genes. Moreover, it would limit the temporal resolution of examining fluctuations in gene expression. Nonetheless, this is a great step forward and with more improvement, Record-seq could become a general technique to understand transcriptional histories.
- Schmidt, F., Cherepkova, M. Y. & Platt, R. J. Transcriptional recording by CRISPR spacer acquisition from RNA. Nature (2018). doi:10.1038/s41586-018-0569-1
- Shipman, S. CRISPR-Cas encoding of a digital movie into the genomes of a population of living bacteria Nature 547, 345-349 (2017)