Credit: Philip Patenall/Springer Nature Limited

Over a billion years, prokaryotes have evolved extensive defence mechanisms to avoid falling victim to predatory viruses and other mobile genetic elements (MGEs). More than 50 years ago, mechanisms were uncovered that prevent cellular entry, repress or inactivate functions required for efficient invasion and destroy invading DNA. More recently, it was demonstrated that over half of sequenced bacteria and archaea harbour CRISPR–Cas systems, which confer adaptive and heritable immunity to invasion by MGEs1. CRISPR–Cas action requires that a CRISPR RNA, derived from the genomes of previously invading MGEs, binds a Cas surveillance complex (Cascade) to locate its invading target through DNA–RNA base-pair complementarity, leading to cleavage and destruction of the invading DNA. How does Cascade efficiently distinguish a DNA target from millions of other sequences in the crowded environment of the bacterial cell, when time is of the essence to be protected against invasion by a potentially replicating MGE?

Two complementary papers addressed this search mechanism using single-molecule imaging in live bacteria, quantitative analysis and modelling of different CRISPR–Cas systems2,3. Vink et al.2 tracked single native Cascade complexes in which the Cas8 subunit was fused to a photoactivatable fluorescent protein, to examine search dynamics upon introduction of a target MGE into Escherichia coli cells. The study revealed an exponential relationship between Cascade abundance and CRISPR interference efficiency, with 20 Cascade complexes in a cell providing 50% protection from invasion. Cascade was shown to spend half of its search time rapidly probing DNA in cells (residence time ~30 ms), with short nucleotide sequences known as protospacer adjacent motifs (PAMs) facilitating the target search. The frequency of CRISPR arrays in the host genome and target transcription influenced target search efficiency and consequently, protection from invasion. Jones et al.3 investigated the search process in E. coli of a CRISPR–Cas system naturally found in Streptococcus pyogenes. A target plasmid was introduced into E. coli expressing a catalytically inactive fluorescently labelled Cas9, the single effector protein of this Cascade complex. An advantage of this heterologous system is that it was possible to control accessibility to the target sequence by regulating the binding of a specific transcriptional repressor. Upon release of the repressor, appearance of target-bound Cas9 was observed as distinct foci. Both papers show that the time required for a single Cascade complex to locate one target is >1 hour; therefore, multiple Cascades are required in the search process to ensure efficient protection against invasion. Specificity and speed of search are considered trade-offs, as the need to locally denature the target DNA sequence, to allow Watson–Crick complementarity, slows down the overall search process as compared, for example, with the search by a specific transcriptional activator for its DNA target. Remarkably, the search is optimized by Cascade directly binding PAMs from where adjacent targets can be interrogated. At the same time, the absence of PAMs adjacent to CRISPR in the host genome prevents suicidal host DNA cleavage.

These papers demonstrate the power of single-molecule tracking in live bacteria to probe the mechanisms by which nucleoprotein complexes find their specific targets with high selectivity in an environment with a huge excess of related DNA sequences. The work not only addresses the biological functions of CRISPR–Cas as a bacterial adaptive immunity system, but additionally will inform the development of CRISPR–Cas gene-editing technologies.