Technical Comments

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

See allHide authors and affiliations

Science  05 Jun 2020:
Vol. 368, Issue 6495, eabb2022
DOI: 10.1126/science.abb2022


Strecker et al. (Research Articles, 5 July 2019, p. 48) described a system for exploiting a Tn7-type transposon-encoded CRISPR-Cas system to make RNA-guided, programmable insertions. Although this system has great promise, we note that the well-established biochemistry of Tn7 suggests that the particular system used may insert not only the transposon but also the entire donor plasmid.

The Scytonema hofmanni CRISPR-associated transposase (ShCAST) system described by Strecker et al. (1) is based on a Tn7-like transposon that carries its own nuclease-defective CRISPR-Cas system, which it uses for RNA-guided target choice (2). The canonical Escherichia coli transposon Tn7 uses two enzymes to create double-strand breaks at the donor-transposon junctions: TnsB to cleave one strand (and subsequently attach it to target DNA) and TnsA to cleave the other (second) DNA strand, leading to excision of the transposon from a donor plasmid followed by its insertion into a target plasmid. When TnsA is mutated, however, the products are not these simple insertions of the transposon, but rather co-integrates in which the entire donor plasmid, flanked by copies of the transposon, is inserted into the target (3). Similar co-integrate products are also created by systems such as Mu transposase, which is a TnsB homolog that lacks a TnsA partner (4).

The ShCAST Tn7-like transposon used by Strecker et al. is naturally lacking in TnsA. Although some DDE transposases harbor insertion domains that trigger other mechanisms for cleavage of the second strand, ShCAST TnsB does not. Homologies to the other systems noted above therefore suggest that their products may not be solely simple insertions as cartooned, but instead may include fusions of the donor plasmid and target. Unfortunately, no experiments were reported that would distinguish between the two possibilities, such as measuring the size of purified “pInsert” product plasmids containing the transposon or testing for the presence of donor DNA in the final products. Although the ShCAST integration system is suggested to bypass the need for homologous recombination–mediated DNA repair normally found in some CRISPR/Cas applications, processing of co-integrate products to remove the donor DNA would likely still require these recombination functions unless another nuclease processed the second strand at the donor site.


Stay Connected to Science

Navigate This Article