Technical Comments

Response to Comment on “RNA-guided DNA insertion with CRISPR-associated transposases”

See allHide authors and affiliations

Science  05 Jun 2020:
Vol. 368, Issue 6495, eabb2920
DOI: 10.1126/science.abb2920


Rice et al. suggest that the CRISPR-associated transposase ShCAST system could lead to additional insertion products beyond simple integration of the donor. We clarify the outcomes of ShCAST-mediated insertions in Escherichia coli, which consist of both simple insertions and integration of the donor plasmid. This latter outcome can be avoided by use of a 5′ nicked DNA donor.

Rice et al. (1) raise an important point about the mechanism of integration of the type V CRISPR-associated transposase (CAST) systems we recently characterized (2). Although CAST loci belong to the Tn7-like class of transposons, additional analysis indicates that type V CAST loci are most closely related to transposons of the Tn5053 family (3, 4), a distinct family of transposons that has not yet been mechanistically characterized. In particular, the Tn5053 family transposons lack TnsA, the enzyme responsible for the 5′ donor cleavage in the Tn7 transposon (5, 6), and result in a co-integrate product that is resolved by the site-specific recombinase TniR (4). Thus, in CAST systems, the 5′ donor ends might not be cleaved, resulting in a co-integrate product containing duplicated cargo DNA and the donor backbone. Alternatively, a CAST protein might either cleave the 5′ donor end or help resolve the co-integrate to yield a simple insertion.

To investigate the exact structure of the insertion product, we performed nanopore sequencing of 15 Scytonema hofmanni CAST (ShCAST)–mediated genome insertions in E. coli and found nine simple insertions and nine co-integrates across several target sites (Fig. 1A). Similarly, a genetic assay using a plasmid target revealed about 20% co-integrate insertions (Fig. 1B). A tentative model is that the initial insertion product is a co-integrate that can be resolved by the cellular DNA recombination and repair systems. However, we cannot rule out that CAST components contribute to 5′ donor cleavage or co-integrate resolution in E. coli or the native host. Notably, all instances of CAST insertion in cyanobacterial genomes identified to date are simple insertions. We will continue to investigate these as a part of our ongoing studies.

Fig. 1 Characterization of CAST insertion products.

(A) Schematic of genome targeting experiment and summary of nanopore sequencing results. (B) Genetic assay for plasmid targeting. pInserts were retransformed and selected on CmR+ and CmR+KanR+ plates to determine the fraction of co-integrate insertions. The total insertion frequency was determined by droplet digital PCR and used to calculate the co-integrate rate. (C) In vitro reactions with purified CAST proteins using plasmid donor or PCR-amplified linear donor.

Use of linear or 5′ nicked DNA donors prevents co-integrate formation (Fig. 1C), thereby providing an approach for applying CAST for homologous recombination–independent genome engineering. Continued investigation of the mechanism of CAST is expected to yield a deeper understanding of the biology of this remarkable DNA integration system and propel its development as a molecular technology.


Stay Connected to Science

Navigate This Article